STP 442a-1978

MANUAL ON WATER
(Fourth Edition)
Sponsored by
Committee D-19
on Water
AMERICAN SOCIETY FOR
TESTING AND MATERIALS
ASTM SPECIAL TECHNICAL PUBLICATION 442A

C. E. Hamilton
Dow Chemical Company
editor
List price $28.50

04-442010-16
AMERICAN SOCIETY FOR TESTING AND MATERIALS

1916 Race Street, Philadelphia, Pa. 19103
This page intentionally left blank
Editoral Staff
Jane B. Wheeler, Managing Editor
Helen M. Hoersch, Associate Editor
Ellen J. McGlinchey, Senior Assistant Editor
SheHa G. Pulver, Assistant Editor
Copyright © BY AMERICAN SOCIETY FOR TESTING AND MATERIALS 1978
Library of Congress Catalog Card Number: 78-51579
NOTE
The Society is not responsible, as a body,
for the statements and opinions
advanced in this publication.
Printed in Tallahassee. Fla.

July 1978
Foreword
ASTM Committee D-19 on Water had long believed that something more
than standard methods of testing are needed by those who look to the Society
for aid and technical guidance. It was felt that general discussions of the
nature and uses of water should illuminate these prescribed procedures and
specifications. Hence, the ASTM Manual on Water.
Preparation of the first edition of the manual was begun in 1946. The first
manual task group consisted of R. C. Adams, chairman and editor, A. A.
Berk, T. H. Daugherty, O. M. Elliot, J. A. Holmes,' C. E. Imhoff,1 A. K.
Light, F. U. Neat,1 F. R. Owens, and A. H. Reynolds.1 Each member drafted
an assigned portion of the manual with the assistance from his associates and
other members of Committee D-19. These drafts were modified, critiqued,
correlated, and assembled by the entire task group so that the resulting
manual was a group product rather than a compilation of individual
authorships. Many persons not in the task group also contributed to the first
edition. Particular acknowledgment should be made for the individual
contributions of Max Hecht,1 L. K. Herndon, E. O. Patridge,1 and C. K.
Rice.
The second edition task group consisted of R. D. Hoak, 1 chairman and
editor, F. N. Alquist,1 R. A. Baker, G. D. Beal, F. E. Clarke, E. F. Davidson,
B. W. Dickerson, D. K. French,1 A. S. Goldin, J. E. Kinney, W. W. Leathen,
S. K. Love, W. A. Moore, F. R. Owens, J. K. Rice, C. C. Ruchhoft, 1 R. K.
Scott,1 J. M. Seamon, J. B. Smith, F. N. Speller,1 and R. F. Weston.
Substantial contributions were also made by C. K. Rice, D-19 Standards
Advisor.
The third edition task group consisted of J. F. Wilkes, chairman and
editor, R. A. Baker, F. E. Clarke, E. F. Davidson, R. D. Hoak,1 C. E.
Imhoff, 1 R. S. Ingols, J. K. Jacobsen, Pratt Johnson, Bernd Kahn, F. N.
Kemmer, R. A. Larrick, S. K. Love, J. A. Martucci, W. L. Neiland, E. P.
Partridge,1 J. H. Phillips, E. A. Savelli, J. S. Speer, and S. L. Williams.
Again, many of those named had worked on the previous edition. Prepara-
tion of the manual was a collaborative effort, involving the entire group plus
many other D-19 members not assigned to the group. The major decision to
remove the standard methods of analysis of water from the manual was
made. These methods and specifications are published annually as the
Annual Book of ASTM Standards, Part 31. Each year all newly accepted
methods and revised methods are incorporated into the reissued book.
'Deceased.
Contents
Introduction 1
Chapter 1—Water Sources and Supply 3
Chapter 2—Uses of Industrial Water 18
Chapter 3—Production and Preservation of Ultrapure Water 28
Chapter 4—Effects of Composition on Industrial Use 38
Chapter 5—Treatment of Process Water and Waste Water 46
Chapter 6—Technology of Industrial Water Reuse 84
Chapter 7—Self-Purification and Other Natural Quality Recovery
Mechanisms 97
Chapter 8—Thermal Loading of Water Supplies 107
Chapter 9—Sampling and Flow Measurement of Water 118
Chapter 10—Water Quality Monitoring 145
Chapter 11— Analysis of Water and Waste Water 177
Chapter 12—Sensory Examination of Water 254
Chapter 13—Sampling and Identification of Deposits in Steam and
Water Systems 276
Chapter 14—Chemical Analysis of Deposits 307
Chapter 15—Sediment in Streams and Other Water Bodies 319
Chapter 16—Radioactive Nuclides in Water 327
Chapter 17—Nuclear Water Technology 352
Chapter 18—Practices for Measurement of Radioactivity 387
Appendix—Reference Tables and Curves 443
Index 459
STP442A-EB/Jul. 1978
Introduction
This manual is intended as brief reference source of information on water.

It will not replace an adequate library on the subject, but it does provide
basic information for routine use and cites references to the technical
literature, thus serving as a point of departure for more specific and detailed
studies.
For executives and administrators, it offers information on the influence
of water on industries in which it is used as a raw material or in conjunction
with manufacturing operations. The influence of water on various industries
ranges from that in the paper industry, where large volumes of water of good
quality are used, to industries where only incidental quantities and qualities
of water are needed. The manual should serve as a guide to the nature of
water planning required at the executive and investment levels of manage-
ment. At no level, however, can the quantity and quality of available water
be disregarded as it relates to the requirements of each process or manu-
facturing operation for today and for the foreseeable future.
Operations personnel will find the manual to be a guide to the significance
of the treatment being applied. Combined with the general discussion of the
problems arising from water use are the specific details of control procedures
and instructions for such critical operations as sampling water under the
various conditions and in the several forms in which it is used. The manual
may be useful as a text for plant operator training and for the indoctrination
of technologists from other fields. Chemists and other technologists who
have special knowledge of water can use the manual as a reference for
specific information not published elsewhere.
The manual may be useful as a text or reference resource in technical
schools, universities, and continuing education short courses or seminars.
While this was not the prime purpose for its preparation, the material
presented should be suitable for classroom use. Despite the extensive use
that can be made of the manual, it should not be expected to replace
competent and well-trained technologists. It will present general information
to some and detailed information to others, but the design and efficient
application of the treatments and techniques discussed require experience
that no books can supply.
This edition of the manual represents a substantial expansion of the
1
Copyright© 1978 by ASTM International www.astm.org
2 MANUAL ON WATER
previous editions. It has been developed to satisfy the growing need for
dependable information about the characteristics of water and the problems
its use and reuse entail. New chapters have been added to highlight problem
areas and to suggest new approaches for their identification and solution.
The five new chapters are: Production and Preservation of Ultrapure
Water, Chapter 3, which discusses this problem area and the need for the
definition of solutions thereof; Chapter 15, Sediments in Streams and Other
Water Bodies, describes the state of the art in this newly developing field;
Nuclear Water Technology, Chapter 17, pulls together information and
experience in this relatively new water field; and Chapter 18 deals with
practices for the measurement of radioactivity.
Water Quality Monitoring, Chapter 10, while not all new material, has
been expanded extensively with the addition of sections on surveillance and
sample conditioning. The other thirteen chapters have been updated and
revised to a lesser extent over the previous edition. The Appendix has also
been revised but still contains much useful information for the water analyst
or engineer.
A new task group is now being organized to consider further new material,
revisions, and updating of the current edition. If you have comments or
suggestions of material which you think would be helpful to you or to other
users of the manual, please contact the ASTM D-19 Committee on Water.
C. E. Hamilton
Dow Chemical Company,
Midland, Mich. 48640;
editor.
Chapter 1—Water Sources and Supply
Water is one of the most important engineering materials used by

industry. Without an adequate supply of water of acceptable quality,
virtually no industry can exist. In a very real sense, man's industrial progress
can be traced by charting his utilization of the water resources of the earth.
Our future industrial progress will depend in large measure on our rational
use of water.
A reliable water supply is a basic consideration in the selection of a site for
any industrial plant. This chapter will discuss the sources and distribution of
water available for industrial use in the United States.
For the purpose of discussion, the total water supply may be subdivided
into precipitation, runoff, groundwater, and saline water.
Precipitation—The Basic Source

Precipitation on conterminous United States averages about 76 cm/year
(30 in./year). This totals about 1.66 X 10'° m 3 /day [4400 U. S. billion
gal/day (Bgd)]. Of the 76 cm (30 in.), 55 cm (21'/2 in.) is evaporated (in-
cluding evapotranspiration). The remaining 21 cm (Sl/2 in.) is our man-
ageable water supply, about 4.7 X 109 m3/day (1250 Bgd). Of this 21 cm
(S1A in.) we used in 1960 a little less than 5 cm (2 in.), of which about 1.25 cm
(l/2 in.) eventually evaporated. The remainder (3.75 cm) (V/i in.) joined the
unused flow to make a total of 20 cm (8 in.) that flowed into the oceans,
about 4.4 X 109 m3/day (1165 Bgd).
The distribution of rainfall is very uneven. In the humid East it ranges
from about 76 to 152 cm (30 to 60 in./year) and in the semiarid West
from about 25 to 50 cm (10 to 20 in.). Extremes range from 380 cm
(150 in.) or more in the Olympic Peninsula in Washington to less than 2.7 cm
(5 in.) in Death Valley, California. If we include the State of Hawaii we
find the maximum reported rainfall is over 1143 cm/year (450 in./year)
on the island of Kauai.
Runoff—Source of Surface Supplies

As would be expected from the wide variability in precipitation, the
4.4 X 109 m3/day (1165 Bgd) runoff also is distributed unevenly. For
3
4 MANUAL ON WATER
example, in the North Atlantic Slope area, runoff averages about 5.1 x 102
m 3 /day (136 Bgd); in the Mississippi River basin the average is about
1.5 X 109 m 3 /day (400 Bgd); in the Colorado River basin it is about 5.7 X
107 m3day (15 Bgd); and in the Columbia River basin, about 8.4 X 10*
m 3 /day (223 Bgd).
In 1960 the national average runoff was equivalent to about 25 000
litres per capita per day (led) (6500 gal per capita per day) (gcd). Based on
population growth projections, this figure would be about 18 000 led (4800
gcd) in 1980 and 13 000 led (3500 gcd) in the year 2000. If the total runoff
to the oceans could be apportioned equally to each person, there would be
enough water from this source on a national basis for the foreseeable future.
But based on actual current and anticipated future distribution of popula-
tion and available water resources, the per capita figure may range from as
little as 11 000 litres/day (3000 gal/day) (gpd) to as much as 110 000
litres/day (30 000 gpd).
Groundwater—Invisible Source
In the discussion on runoff, the figures given for runoff to the oceans
include groundwater. The water falling as precipitation to the ground
follows three courses. Some runs off directly to streams and rivers, some
infiltrates to the ground water reservoirs, and the rest evaporates or
transpires through vegetation.
The water in the ground may be thought of as a temporary natural
reservoir. It is invisible and consequently to many people is shrouded in
mystery. Nearly all groundwater is in constant though slow motion toward
streams or other surface water bodies. In fact, in dry periods streams
maintain their flow because of groundwater discharge. Thus the water
discharging through streams to the oceans includes most of the runoff from
the groundwater reservoir. The remainder discharges by underground
routes.
Groundwater reservoirs store far more water than is or conceivably could
be stored in surface reservoirs. The available and accessible supply of ground-
water in the United States is estimated to be in excess of 2.2 X 1014 m3
(59 000 000 billion gallons) [7].' However, when large quantities are being
sought it usually is more difficult and expensive to locate, to study and
evaluate, to develop, and to manage groundwater than surface water.
Furthermore, the management of groundwater is complicated by defi-
ciencies and incorrect concepts of groundwater hydrology and by much of
existing water law. Development of groundwater as a phase of coordinated
multiple-purpose water management can be achieved on a large scale only if
adequate hydrologic information is available [2].
'The italic numbers in. brackets refer to the list of references appended to this chapter.
CHAPTER 1 ON WATER SOURCES AND SUPPLY 5
Saline Water—Prospective Future Source

Natural saline water (usually seawater or mixtures of' seawater and
freshwater) is used by industry only where adequate supplies of freshwater
are unavailable. Nearly all industrial use of saline water is for cooling at or
near the seacoast. The quantity of seawater is unlimited. Converted saline
water may become an important new source of freshwater when and where
the cost of conversion including conveyance to the point of use becomes
competitive with sources of freshwater. Desalination of brackish water in
arid regions also is practiced.
Beginning in 1952 the Federal Government began a program of research
in converting saline water into freshwater. This effort has resulted in the
construction of five demonstration plants. The location, process of con-
version, and capacity of these plants are given in Table 1.
TABLE 1 Saline water demonstration plants.

Plant Capacity,
Location Source of Water Conversion Process .gpd'
Freeport, Tex. Gulf of Mexico multiple-effect distilla- 1 000 000
tion, falling film
San Diego, Calif." Pacific Ocean multistage distillation. 1 000 000
Hash
Webster. S. Dak. wells electrodialysis (mem- 250 000
brane process)
Roswell, N. Mex. wells distillation, vapor 1 000 000
compression
Wrightsvilie Beach. N.C. Atlantic Ocean free/ ing 250 000
"Relocated. 1964. at U. S. Naval Base. Guanta*namo Bay. Cuba.
''Metric conversion factor: 1 gpd = 0.003784 m3/day.
The cost of converting saline water has been reduced substantially during
the past 20 years. Cost at the conversion plant ranges from about $0.60 to
$1.50 per thousand gallons depending on the process used, the salinity of the
raw water, the capacity of the plant, and other factors. Distribution cost
would have to be added.
In areas where adequate quantities of freshwater are not available, certain
kinds of industries and domestic uses probably can justify the cost of
converted saline water. Reuse of fresh or converted water will keep cost of
industrial water to a minimum.
Water Demands
Demands for water for all purposes are constantly increasing. Estimates of
present and future demands will suggest how the available water resources
should be utilized.
6 MANUAL ON WATER
Water Withdrawals
Many estimates have been made about the quantities of water that are
withdrawn from our surface and underground sources for useful purposes.
Among those frequently quoted are estimates compiled by the U. S. Senate
Select Committee. The total water withdrawn in 1954 for all uses, excluding
hydropower, was about 3.4 X 108 m3 (300 Bgd), which was equivalent to
about 6800 led (1800 gcd). Of this amount 58 percent was used for irrigation,
36 percent for industry, and 6 percent for public supplies, rural and other
purposes. A projection by the Select Committee [3] to the year 2000
anticipates total water use of about 3.1 X 109 m3 (900 Bgd), of which over 70
percent will be used by industry. However, only about one fifth of the water
withdrawn would be consumed and most of this by agriculture.
In another estimate [4} the total water withdrawn in 1960 was 1.0 X109 m3
(270 Bgd), of which 51 percent was for industry, 40 percent for irrigation, 6
percent for public supplies, and 1 percent for rural use.
From 90 to 95 percent of industrial water is used for cooling. After cooling
is accomplished, the water is returned to the water system for reuse. The heat
removed in cooling processes is transferred from the industrial equipment to
the water as it reenters the hydrologic system. The higher temperature tends
to upset the delicate balance established by living plants and animals
indigenous to the receiving streams. The heated water also reduces oxygen
solubility and increases the biochemical oxygen demand, thus producing an
oxygen sag in the streams. In extreme cases, other pronounced and complex
quality changes will occur as a result of the temperature change. As larger
quantities of water are employed for cooling, some of the other uses of water
will be affected to some degree.
Consumptive Use
Consumptive use of water has been defined as ".. . the quantity of water
discharged to the atmosphere (evaporated) or incorporated in the products
of the process in connection with vegetative growth, food processing, or
incidental to an industrial process" [5].
Industry consumes only about 2 percent of the water it uses. This is
divided principally between water evaporated and water incorporated in
products of manufacturing. The heaviest consumption is in connection with
irrigation, where 60 percent or more of the water withdrawn is lost to the
water system through evaporation and transpiration. About 17 percent of
water used for public supplies is consumed.
Storage
Water is stored in reservoirs to control floods, develop power, provide
water supplies for municipalities, industries, and irrigation, dilute wastes,
stimulate recreation, and for other beneficial uses. In 1954 [6] a total of
3.4 X 10" m^ (278 000 000 acre-ft) (91 000 billion U. S. gal) was stored in
conterminous United States having usable capacities of 6.2 X.106 m3 (5000
acre-ft) (1.6 billion U. S. gal) or more.
The distribution of these storage capacities by states is given in Table 2.
TABLE 2— Summary of reservoirs showing usable storage.

Usable Storage
U. S.
Billion
0
State Acre-Feet Gallons'
Alabama 2 522 900 822
Arizona 33 246 800 10 834
Arkansas 9 561 800 3 116
California 17 700 900 5 768
Colorado 3 826 700 1 247
Connecticut 413 500 135

Delaware 6 800 2.2
Florida 2 752 300 897
Georgia 4 792 300 1 562
Idaho 9 530 100 3 105
Illinois 203 700 66.4

Indiana 236 000 76.9
Iowa 458 000 149
Kansas 4 008 700 1 306
Kentucky 8 451 000 2 754
Louisiana 1 307 600 426

Maine 5 180 500 1 688
Maryland 472 600 154
Massachusetts 1 791 200 584
Michigan 521 900 170
Minnesota 3 822 500 I 246

Mississippi 3 825 900 1 247
Missouri 5 744 500 1 872
Montana 23 888 100 7 784
Nebraska 3 842 400 1 252
Nevada 673 300 219

New Hampshire I 069 900 348
New Jersey 216 100 70.4
New Mexico 3 530 700 1 150
New York 4 383 600 1 428
North Carolina 3 112 600 1 014

North Dakota 18 786 100 6 121
Ohio 3 089 200 1 007
Oklahoma 10 248 400 3 339
Oregon 3 747 000 I 221
Pennsylvania 1 739 500 567

Rhode Island 117 700 38.4
(Continued)
8 MANUAL ON WATER
TABLE 2—Continued.
Usable Storage
U. S.
Billion
State Acre-Feet " Gallons b
South Carolina 3 759 800 1 225

South Dakota 22 476 600 7 324
Tennessee 10 053 700 3 276
Texas 18 852 900 6 143

Utah 2 053 200 669
Vermont 351 700 115
Virginia 2 440 800 795
Washington 11 364 600 3 703
West Virginia 932 800 304

Wisconsin 2 156 500 703
Wyoming 4 855 800 1 582
Total 278 121 200 90 626

!
"Metric conversion factor: I acre-foot = 1233.49 m .
^•Metric conversion factor: 1 gal (I. 1 . S.) - 0.003785412 nv.
Reusable Supplies
Most of the water withdrawn is not consumed but is returned to a
downstream or underground source. Thus the total supply is depleted by
only a portion of the entire amount withdrawn. For this reason the water
returned augments the original supply and is available for reuse. Were it not
for reuse, some areas would have run out of water. It must be kept in mind,
however, that each use of water (except for generation of hydropower) will
degrade its quality to some extent. Impoundment of water for hydropower
or other uses can cause significant changes in water quality.
Quality Considerations in Industrial Supplies

Water is relatively pure when formed in clouds, but it absorbs gases from
the air, mainly oxygen, carbon dioxide, S02 and NOx, during its fall as rain or
snow. Upon reaching the earth, water dissolves materials with which it
comes in contact. The amount of dissolution depends upon the solubility of
the soil or rocks in the area and the length of time that the water and the
earth-environment are in contact. Water also becomes polluted with in-
dustrial, municipal, and agricultural wastes.
General Quality Factors

Both the total concentration of dissolved solids and the concentration of
the more important constituents determine the suitability of water for
specific industrial applications. Hardness also is an important factor.

Surface waters in general are softer and less concentrated than ground-
waters. However, hardness and dissolved solids in surface waters vary
considerably over the period of a year and in many streams from day to day.
Surface waters are frequently turbid, some excessively so, and with few
exceptions must be filtered before use for public supplies or by industry.
In a given stream, hardness and concentration of dissolved solids generally
vary inversely with the volume of water discharged. This is because the
proportion of groundwater, which usually is more concentrated than surface
water, is much smaller during high water than during low water periods.
During periods of low flow, groundwater discharge to a stream channel may
account for the total water in a stream.
The average concentration of dissolved solids in major rivers of the nation
ranges from about 60 to 700 mg/litre (ppm). The lower concentrations are
found in streams in the Atlantic and eastern Gulf Coasts, in the Pacific
Northwest, and in the Great Lakes. Higher concentrations are found in the
western Gulf, Midcontinent, and Southwestern streams.
Groundwaters are characterized by essentially constant composition and
temperature and ordinarily can be used without filtration. They normally are
harder and contain more dissolved minerals than surface waters and thus are
less desirable for certain industrial purposes.
The varying quality characteristics of surface and groundwaters are
illustrated by the analyses given in Table 3. These analyses represent the raw
water used for public supplies of eight of the larger cities in the United
States. Analyses 1 through 4 are for surface waters and 5 through 8 are for
groundwaters. A more extensive tabulation is given in the Appendix.
Hardness
Because of the importance of the hardness characteristic of industrial
waters, this topic will be discussed in greater detail. For many industries, the
hardness of the intake water is the principal basis for classifying water.
Hardness is defined in accordance with the ASTM Definitions of Terms
Relating to Industrial Water and Industrial Waste Water (D 1129) [7], as a
characteristic of water generally accepted as representing the total con-
centration of calcium and magnesium. Although other constituents, such as
iron, manganese, and other polyvalent cations cause hardness, they usually
are present in such small concentrations that they do not materially affect
the hardness of industrial water. Only calcium and magnesium are included
in the term hardness as used in this discussion.
Hardness may be divided into two types, carbonate and noncarbonate.
Carbonate hardness includes that portion of the hardness which is equiva-
lent to the carbonate and bicarbonate present in the water. Any hardness in
excess of this amount is called noncarbonate. Carbonate hardness generally
is not objectionable in industrial use of water. Where it is excessive, however,
it can be reduced or removed rather simply and inexpensively. Noncar-
TABLE 3—Analyses of typical surface and groundwaters in the United States. [From U. S.
Geological Survey Water Supply Paper 1812 (milligrams per litre).]
Analysis number? 1 2 3 4 5 6 7 8
Date of collection 8/22/61 9/1/61 10/19/61 1961 8/9/61 9/15/61 8/23/617/25/61
Silica (SiCh) 5.9 2.1 5.5 8.7 25 8.3 9.1 21
Iron (Fe) 0.00 0.17 0.40 0.03 0.72 0.10 0.01
Manganese (Mn) 0.00 0.14 0.00 0.01 0.06 0.00
Calcium (Ca) 8.5 28 37 84 65 12 57 26
Magnesium (Mg) 2.6 7.0 8.9 28 23 6.1 32 6.2
Sodium (Na) 3.6 4.1 17 92 14 7.5 3.4 86
Potassium (K) 1.5 0.9 5.1 4.0 1.8 0.7 1.7 1.9
Bicarbonate (HCOa) 25 92 128 140 179 78 332 254

Carbonate (COj) 0 0 0 1 0 0 0 0
Sulfate (SO4> 9.0 18 48 285 112 3.8 12 11
Chloride (Cl) 5.5 8.0 10 83 18 3.0 3.5 40
Fluoride (F) 0.1 0.0 0.4 0.4 0.8 0.4 O.I 0.7
Nitrate (NOs) 4.7 0.5 3.2 1.4 0.2 1.2 0.2 0.0
Dissolved solids 59 129 222 657 410 87 284 318

Hardness as CaCO.i 32 99 129 323 256 55 274 90
Noncarbonate hardness as CaCOj 11 24 24 206 110 0 2 0
Specific conductance (micromhos at
25°C) 91 213 324 1040 535 137 498 533
PH 6.4 7.6 7.7 8.4 7.9 6.8 7.6 7.2
Color 5 3 5 5 5 1 0
"Analyses numbers are identified as follows:
1. Baltimore, Md., North Branch Patapsco River (raw).
2. Detroit, Mich., Detroit River (raw).
3. St. Louis, Mo., Mississippi River (raw).
4. Los Angeles, Calif., Colorado River (raw).
5. Jacksonville, Fla., Composite of several wells (raw),
6. Memphis, Tenn., Allen Well Field. Composite of several wells 122 to 183 m (400 to 600 ft) deep (raw).
7. Rockford, 111., Well No. 15, 413 m (1355 ft) deep (raw).
8. Houston, Tex., Southwest Well Field. Composite of 12 wells 149 to 610 m (490 to 2000 ft) deep (raw).
bonate hardness, on the other hand, especially where sulfate is the pre-
dominant anion, may result in formation of hard, adherent scale in boilers
and other heat-exchange equipment. This type of hardness is more difficult
and more costly to remove.
Any descriptive or numerical classification of hardness of water is
arbitrary. A water that is termed hard in some areas may be considered soft
in other areas. The U. S. Geological Survey uses the following classification:
Range of Hardness
mg/ litre (ppm) Descriptive Classification
I to 60 soft
61 to 120 moderately hard
121 to 180 hard
Above 180 very hard
Figure 1 shows the hardness in the foregoing ranges for the finished public
supplies of the 100 largest cities in the United States in 1962. The numbers in
the map identify the cities listed in Table 4. The tables also gives, in figures,
the approximate average hardness in these cities.
Other Quality Parameters

Other quality parameters also have a bearing on the suitability of water
for particular uses.
Silica in water ranges in concentrations from less than 1 to over 75
FIG. 1—Hardness of treated public water supplies of the 100 largest cities in the United
States, 1962 (average weighted by population served).
1-2 MANUAL ON WATER
TABLE 4—Hardness of finished water supplies of 100 largest cities in the United States, 1962.
[From\J. S. Geological Survey Water-Supply Paper 1812 (hardness in milligrams
per litre of calcium carbonate).]
Number on map Average
State (Fig. 1) City Hardness
Alabama 1 Birmingham 88
2 Mobile 35
3 Montgomery 42
Arizona 4 Phoenix 250

5 Tucson 185
California 6 Fresno 82
7 Long Beach 110
8 Los Angeles 110
9 Oakland 58
10 Sacramento 65
11 San Diego 235
12 San Francisco 35
13 San Jose 205
Colorado 14 Denver 78
Connecticut 15 Bridgeport 31
16 Hartford 28
17 New Haven 42
^District of Columbia 18 Washington 115
Florida 19 Jacksonville 280

20 Miami 61
21 St. Petersburg 92
22 Tampa 115
Georgia 23 Atlanta 23
24 Savannah 105
Hawaii 25 Honolulu 60
Illinois 26 Chicago 140

27 Rockford 280
Indiana 28 Evansville 155

29 Fort Wayne 85
30 Gary 140
31 Indianapolis 285
32 South Bend 350
Iowa 33 Des Moines 95
Kansas 34 Kansas City 250

35 Topeka 105
36 Wichita 105
Kentucky 37 Louisville 110
(Continued)
TABLE 4—Continued.
Number on map Average
State (Fig. 1) City Hardness
Louisiana 38 Baton Rouge 5
39 New Orleans 91
40 Shreveport 60
Maryland 41 Baltimore 55
Massachusetts 42 Boston 13
43 Springfield 12
44 Worcester 15
Michigan 45 Detroit 100

46 Flint 85
47 Grand Rapids 140
Minnesota 48 "Minneapolis 65
49 St. Paul 90
Mississippi 50 Jackson 50
Missouri 51 Kansas City 85

52 St. Louis 105
Nebraska 53 Lincoln 180

54 Omaha 150
New Jersey 55 Jersey City 40

56 Newark 35
57 Paterson 55
New Mexico 58 Albuquerque 99
New York 59 Albany 52

60 Buffalo 135
61 New York City 36
62 Rochester 86
63 Syracuse 110
64 Yonkers 78
North Carolina 65 Charlotte 24

66 Greensboro 47
Ohio 67 Akron 130

68 Cincinnati 160
69 Cleveland 125
70 Columbus 100
71 Dayton 105
72 Toledo 70
73 Youngstown 85
Oklahoma 74 Oklahoma City 130

75 Tulsa 93
Oregon 76 Portland 7
(Continued)
14 MANUAL ON WATER
TABLE I—Continued.
Number on Average
State map (Fig. 1) City Hardness
Pennsylvania 77 Erie 120
78 Philadelphia 105
79 Pittsburgh 110
Rhode Island 80 Providence 28
Tennessee 81 Chattanooga 87
82 Memphis 40
83 Nashville 92
Texas 84 Amarillo 215
85 Austin 88
86 Corpus Christi 135
87 Dallas 90
88 El Paso 97
89 Fort Worth 145
90 Houston 60
91 Lubbock 310
92 San Antonio 230
Utah 93 Salt Lake City 180
Virginia 94 Norfolk 77
95 Richmond 60
Washington 96 Seattle 20
97 Spokane 150
98 Tacoma 15
Wisconsin 99 Madison 290
100 Milwaukee 130
mg/litre and averages from 5 to 15 mg/litre. Under certain conditions, silica

forms hard scale in boilers and on turbine blades.
Iron, manganese, and aluminum in surface waters usually range from 0.01
to 0.5 mg/litre. Organic matter can form complexes with these metals, thus
permitting higher concentrations to occur. These metals frequently are
present in groundwater in substantially higher concentrations than in surface
water. Values from one to several milligrams per litre (parts per million) in
groundwater are not uncommon. Iron and manganese are especially ob-
jectionable in many industrial processes primarily because they impart
undesirable colors and stains to manufactured products.
Other minor elements occur in small concentrations commonly in the
micrograms per litre (parts per billion) (parts per 1000 million) range.
Sodium and potassium are not important industrially in low concentra-
tions. The sodium content of water varies over a wide range from below 5
mg/litre in freshwater to many thousands of milligrams per litre (parts per
million) in brines. Seawater contains about 10 700 mg/litre of sodium.
Potassium is ordinarily present in freshwater in low concentrations, seldom

in excess of 10 mg/litre.
Alkalinity in water supplies usually represents the bicarbonate ion,
HCO3-, although it also includes carbonate,CO3—, and in water receiving
certain alkali wastes, hydroxide, OH-. Alkalinity customarily is expressed
as calcium carbonate, CaCOj. Practically all natural waters have alkalinity
unless neutralized by acid from industrial effluents or from mine drainage.
The alkalinity is related to the buffer capacity of the water, that is, the
capacity of the water to receive acid without substantially lowering the pH.
It also is troublesome in certain industrial processes.
Sulfate concentrations range widely in natural waters. Sulfate may be
derived from leaching of gypsum, anhydrite, and shale deposits, or from
oxidation of sulfides. Waters in arid or semiarid regions, such as South-
western United States, may contain relatively large quantities of sulfate.
Sulfate, when combined with calcium, forms troublesome scale in industrial
heat exchangers and is difficult to remove.
Chloride is found in all natural waters, and occurs over a wide range of
concentrations. Streams along the Atlantic Coast, eastern Gulf Coast, and in
the Pacific Northwest normally contain less than 25 mg/litre of chloride,
except in tidal reaches or where polluted. Streams in arid or semiarid regions
may contain several hundred milligrams per litre of chloride. Seawater
contains about 19 000 mg/litre of chloride, and other brines may contain
much higher concentrations.
Chloride in concentrations less than 50 mg/litre has no detrimental effects
on most industrial uses, and water supplies with higher concentrations are
used rather widely. High-chloride waters promote corrosion of metal
surfaces.
Dissolved gases, including carbon dioxide (CCh), oxygen (02), hydrogen
sulfide (H2S), or ammonia (NFb), frequently are important, depending on
the final use of the water. While some of these gases may be combined as
ions, they are usually considered as dissolved gases because they may be
separated as gases by heat or agitation, and they leave no residue on
evaporation. Gases must be removed from water for certain industrial uses,
especially where steam is generated in boilers.
Carbon dioxide is a respiration product of aquatic plants and animals and
is one of the by-products of aerobic and anaerobic decomposition of organic
matter. Acids from natural sources or pollution liberate carbon dioxide from
bicarbonate. Most of the carbon dioxide in water is derived from these
sources. Air contains only about 0.04 percent carbon dioxide, and the
concentration in water in equilibrium with air will approach 0.7 mg/litre.
Streams normally contain low concentrations of carbon dioxide, but some
groundwaters contain 20 mg/litre or more in their natural environment.
Dissolved oxygen ranges between about 3 and 12 mg/litre in surface
waters and there is little or none in most groundwaters. Polluted surface
waters may contain less than 3 mg/litre, even zero in extreme situations.
16 MANUAL ON WATER
Hydrogen sulfide may range from 0.0 to several mg/litre in groundwaters

but usually does not exceed 1 or 2 mg/litre. It is seldom found in surface
waters unless they are polluted. As little as 0.1 mg/litre in water causes
perceptible odor.
Ammonia ranges between 0.0 and about 4 mg/litre and occurs naturally in
both surface and groundwaters. Ammonia also results from the presence of
sewage or industrial wastes.
Hydrogen ion concentration, or pH (which is a measure of the hydrogen
ion activity), is related to the acidity or alkalinity of a water. The pH of most
surface waters ranges between 6 and 8. Waters containing free mineral acids
from mine drainage or other industrial wastes may have pH values below 4,
depending on the buffering capacity of the receiving water. Some ground-
waters have pH values above 8 and some below 5, but most are in the 6 to 8
range. The pH values of waters have a very definite bearing on the utility of
supplies for many industrial purposes.
Turbidity is a characteristic of most surface waters. It is caused by
particulate matter in suspension and results from erosion of land surfaces.
Turbidity may also result from inadequately stabilized or filtered treated
water. Groundwaters usually are not turbid. However, some groundwaters
that are chemically unstable may develop turbidity on exposure to the
atmosphere.
Color usually results from the presence of organic matter; it may come
from decayed vegetation or from municipal or industrial wastes.
Microorganisms are present in nearly all surface waters, and they are
commonly a problem. They may be classified as bacteria, fungi, and algae.
They are part of natural life processes of the hydrologic, geologic, and
ecologic environments. Microorganisms also are influenced by man-made
activities and may increase or decrease in population in response to waste
materials discharged into the streams.
Other properties of importance to industry are taste, corrosiveness, and
radioactivity. All of these quality parameters and their effects on industrial
processes will be discussed at greater length in succeeding chapters.
Summary
The country is not likely to run out of water in the foreseeable future. The
total water supply is essentially the same now as in the past and will be in the
future. The distribution of water, however is highly variable as to time and
place. In addition, the needs for water on a countrywide basis cannot be
correlated with its availability.
Industrial water requirements must be carefully assessed in terms of the
quantity and quality of water that can be obtained at the plant site. If the
available supply is not of suitable quality, the cost of conditioning must be
weighed against the costs of making available other water sources or other
plant locations.
Reuse of water is being practiced widely in industry. The practice will

undoubtedly increase as industry uses an increasingly large share of the
available supply.
Converted brackish and saline water is an additional source of water.
Although the cost of desalted water is greater than that of available
freshwater, increasing demands in water-short areas may justify its use.
Industry is expected to make maximum efficient use of the available
sources of water supply and to return the used water to the hydrologic
system with as little impairment in quality as possible.
References
[/] Nace, R. L., "Water Management. Agriculture and Ground Water Supplies," II. S.
Geological Survey Circular 415, 1960.
[2\ McGuinness, C. L., "The Role of Ground Water in the National Situation," U. S. Geolog-
ical Survey Water-Supply Paper 1800, 1963.
[3] Select Committee on National Resources, U. S. Senate, "Water Supply and Demand,"
Committee Print No. 32, 1960.
[4] MacKichan, K. A. and Kammerer, J. C., "Estimated Use of Water in the United States,
1960," U. S. Geological Survey Circular 456, 1961.
[5] Task Group Report, American Water Works Association, "Water Conservation in
Industry," Journal of the American Water Works Association, JAWWA, Vol. 45, 1953,
p. 1249.
[6\ Thomas, N. O. and Harbeck, G. E., "Reservoirs in the United States," U. S. Geological
Survey Water-Supply Paper 1360-A, 1955.
[7] Annual Book of ASTM Standards. Part 31.
Chapter 2—Uses of Industrial Water
Since earliest times, men have settled along the shores of rivers and lakes to
avail themselves of freshwater for their homes and industries. The American
colonists built their mills along streams which could furnish power and
supply the quality of water necessary for the manufacture of such goods as
textiles and paper.
In modern times, public awareness of the importance of water is mounting
as headlines warn of a growing problem of water quality and availability.
The expanding industry must carefully examine the limitations which water
may impose on it: How much raw water will be available ten years hence?
Can a portion of the plant's usage be reduced by recycle? What quality water
might be required by new uses? How can this quality be achieved economi-
cally while impurity levels in the raw water supply continue to rise? What
degree of waste treatment is required to preserve the stream water quality?
To help bring these problems into perspective, this chapter on the uses of
industrial water outlines general principles and practices rather than specific
details of water usage by individual industries.
There are few, if any, significant uses of water which are peculiar to a
single industry. The major uses—steam generation, cooling, and pro-
cessing—are common to all industries. Of the various categories of process
uses of water, such as conveying and washing, many appear in a variety of
plants unrelated in any other way.
These general principles and practices are presented to orient the engineer
toward good water management in his own industry. The examples illus-
trating water use in several specific plants were selected for their universal
appeal. In them, engineers in all branches of industry will find practices
related to their own operations.
Some Thermodynamic Properties of Water

The polar nature of the water molecule makes water a unique material for
the transfer of heat from one system to another. The heat capacity of water is
so great that when 0.45 kg (1 Ib) of water absorbs 1055J (1 Btu) of heat from
a system which it is employed to cool, its temperature rises only 1 deg F. By
comparison, if oil or alcohol were used for the same cooling service, the
18
CHAPTER 2 ON USES OF INDUSTRIAL WATER 19
temperature rise would be twice as great or twice the volume would be

needed to keep within the same temperature limits. If air were used, almost
3000 volumes of air would be needed to replace a volume of water at the
same temperature rise.
Because water molecules tend to "adhere" to one another by the process
known as "hydrogen bonding," considerable energy is required to convert
water to steam. But once in the form of steam, water vapor can convey large
quantities of heat from one system, such as a steam boiler, to another, such
as a steam-heated cooking vessel. When the steam gives up its heat to the
material in the vessel, each pound liberates about 1023 kj (970 Btu).
Since this quantity of heat, the heat of vaporization, is associated with a
change in state from liquid or vapor, the evaporation of a stream of water
sprayed into the air cools the water-jet as it gives up energy to the water
vapor escaping into the air. Thus, evaporation can be used either to supply
heat, as steam is produced, or to cool water for extraction of heat. A third
property of water related to heat transfer is its heat of fusion, about 337
kJ/kg (144 Btu/lb). This is considerably greater than the heat released by
solidification of other common liquids, another consequence of the unique
structure of the water molecule.
Basic Water Uses

Of the many services which water can perform for our society, those which
may be defined basic to industry are heat transfer, power generation, and
process uses.
Heat Transfer
Industry capitalizes on the large heat capacities of water for economical
heating or cooling of process units.
The most common method of supplying heat to various areas of an
industrial complex is by steam generation. The steam is produced from fuel
at a central generating station and distributed at relatively high pressure and
velocity throughout the plant. Temperature at each unit is easily controlled
by regulating flow or pressure.
In addition to the steam generated by direct firing of a boiler with coal, oil,
or gas, additional steam may be produced by firing waste products or by
recovering heat from a high-temperature process stream. Some typical
process heat boilers (sometimes called "waste-heat boilers") are described in
Table 1.
Although steam historically has been the major conveyance of industrial
heat because of high heat capacity and heat-transfer coefficients, hot water is
replacing it in some of the newer plants. This is due to the lower installation
20 MANUAL ON WATER
TABLE 1 — Some industrial process heat hollers.

Type of Plant Description of Heat Source
Cement lime hot gases leaving kiln
Sulfuric acid hot SO2 from sulfur burners
Ammonia synthesis gas after exothermic reaction
Sugar burning bagasse (pressed cane stalks)
Wood pulp burning bark; burning black liquor
Steel burning blast furnace gas; hot gases leaving
open hearths or basic oxygen furnace
Oil refining burning CO from cat-cracker; hot
catalvst from cat-cracker
cost of modern circulating hot-water systems. However, many factors are

involved in the choice between these two systems, a major one being the
possibility of on-site generation of power from steam which may then be
used a second time at lower pressure for heating.
Circulating water has been used for cooling process equipment for many
years. Early industry, with a plentiful supply of water, simply pumped cool
water from a nearby stream or lake through the plant and discharged the
heated water in any convenient manner. Even today there are areas where
this practice is still the most practical means of moving heat. This is
particularly suited to coastal locations, with plants designed to withstand the
corrosive attack of seawater.
The volume of water which must be withdrawn from a freshwater source
for cooling can be tremendously reduced by adding an evaporative cooling
tower or pond to the system. For example, if a plant withdraws and uses 1.26
m3/s (20 000 gal/min) of freshwater, heating this water 6 deg C in cooling its
process equipment on a once-through system, a reduction in withdrawal to
about 0.015 m 3 /5 (250 gal/min) freshwater may be practical by recirculation
of 1.26 m 3 /s (20 000 gal/min) over a cooling tower. This is due to the high
heat of vaporization of water, which cools the recirculated water about 6 deg
C for each 1 percent evaporated.
The temperature to which the recirculated water can be reduced by this
process is limited by the wet-bulb temperature of the air used to induce
evaporation. This temperature of recirculated water will usually be higher
than that of the fresh makeup source. If further temperature reduction is
needed, individual refrigeration may be added to those units in the plant
needing the additional cooling.
In the chemical industry, hot hydrocarbon gas streams are often passed
through a spray of water which serves the combined purpose of quenching
the high-temperature vapor, condensing heavy fractions, and washing
unwanted material from the stream.
A method of cooling that is not widespread but sometimes practical is the
introduction to ice into a system. This method is used to quench the heat of
certain organic chemical reactions.
Power Generation
By far the greatest share of our total power generation comes from
steam plants. The public utilities today produce over 9 x 1012 kJ/day
(2.5 billion kWh/day) of electricity by this route. Most of these plants
recover the steam with condensers, so that the water makeup is usually
extremely low—less than 1 percent. However, some of these utility plants
have been located adjacent to industrial plants to whom they supply both
electricity and steam. This steam is often consumed or contaminated, so that
such plants may require as much as 50 percent makeup. A typical arrange-
ment of this kind involves a utility station adjoining a refinery. The refinery
supplies the utility water from its private source and fuel, and the utility
supplies its neighbor electricity and low-pressure steam.
The industrial plant generatinfflteteam for power differs from the utility
station in the relatively greater extent of water loss, with higher makeup, in
the variation and degree of condensate contamination, and in the tempera-
ture of the returned condensate.
A major reason for the relatively larger makeup rate in industrial plants is
the unfavorable cost of installing piping to return condensate from a large,
complex plant. A typical case is the petroleum refinery where some of the
steam is sent to the outlying tank farm to heat heavy, viscous oils and keep
them pumpable. Not only is it costly to run douple piping to return
condensate, but, in this case, recovery is unattractive because of the
likelihood of oil contamination. It is certain that the future will bring greater
pressure to reclaim a large share of industrial condensate which until now
has appeared not worth recovering.
The relatively high temperature of industrial process condensate is due to
the steam being condensed at high temperatures. In the utility station, steam
is expanded through a turbine to subatmospheric pressure and finally
condensed in a condenser by cooling water. The condensate returned to the
system is only a few degrees warmer than the discharged cooling water. In
the industrial plant, the condenser is the process equipment, and it usually
operates at pressures above atmospheric. Therefore, the industrial plant
typically returns condensate at 65 to 121° C (150 to 250° F) as compared to 38
to 49° C (100 to 120°F) utility condensate.
The industrial steam plant may use steam to drive machinery, to generate
electricity, or both. The paper mill is a good example of this. A typical mill
will produce steam at 6205 kPa (900 psi). This may be used through a
number of turbines to drive pumps, generators, compressors, paper ma-
chines, and auxiliary equipment. Some of the steam will exhaust to an 862-
kPa (125-psi) steam line, used to heat the large drier rolls over which the
paper passes. Other drier rolls may operate at 413 kPa (60 psi). This steam
will also go to cooking the wood chips in a digester to make pulp. Some of
the turbines in the power plant may expand the high-pressure steam all the
way down to vacuum, as a utility operation, to obtain maximum thermal
efficiency. A typical paper mill system is shown in Figure 1.
22 MANUAL ON WATER
FIG. 1—Paper mill steam system.
In some geographical areas, the electrical energy cost may be so low that
the industrial plant cannot afford to generate electricity even though the
steam is available. Such is the case, for example, with a paper mill in the area
served by Tennessee Valley Authority. All turbines discharge steam at 413
kPa (60 psi) or above, none operate with condensers, and only mechanical
power is produced from the steam.
Even the tax structure plays an important part in the decision as to the
best use to make of steam in an industrial plant. Corporate tax rates tend to
favor higher operating costs as against extra capital costs which could lower
operating costs. As a result, industry tends to depreciate capital equipment
at a more rapid rate than do the utilities. Because of this, industry has not
been encouraged to install the higher-pressure (over 8274 kPa) (1200 psi)
steam generating plants which utilities are finding so economical to operate.
/The generation of mechanical power from water flowing through hy-
draulic turbines recalls the early days of the industrial revolution and the use
of the water wheel to mill grains. A number of industrial plants today
continue to use natural hydraulic power for mechanical jobs or for
generation of electricity. In some parts of Canada, the excess of hydro-
electric power and its low cost is such that at times of the year it is actually
cheaper to generate steam in electric boilers than in fuel-fired boilers.
In addition to making use of natural hydraulic power, industry also uses
artificially generated hydraulic power for special applications. The most

common of these involves high-pressure water-jets used to focus a high
concentration of kinetic energy on a small area. Steel mills blast mill scale
from steel sheet in this manner. Lumber mills strip the bark from gigantic
redwood logs rotating on a lathe with water jetting from a nozzle at 17 237
kPa (2500 psi). The mining industry has employed jets for many years to
fragment ore lodes and separate the desirable minerals from the surrounding
burden of soil.
Role of Water in Industrial Processes

In process applications, water has valuable duties to perform in conveying
other materials, in various washing procedures, as a raw material itself, and
in numerous other uses which may be peculiar to only one industry or even
to a single plant.
Conveying—Just as our watercourses convey a natural burden of sus-

pended matter, water streams constrained by piping or flumes in a plant can
be put to work transporting materials from one area to another. The pulp and
paper industry is one of the largest users of water for this purpose.
An integrated mill of 900 t/day (1000 tons/day) capacity producing a
variety of grades of paperboard, for example, may withdraw 7.57 to 11.36 X
104 m 3 /day [20 to 30 million gal/day (Mgd)] of water, of which 80 to 90
percent will be required for process uses. After the pulp is produced, it may
be screened to remove oversize material and debris at a "consistence" of 6 to
8 percent, or about 92 to 94 percent water. It will then be dewatered to
reduce volume for storage at about 12 percent consistency. It is then diluted
again to about 8 percent for pumping to the paper mill area. After refining
and further processing, the pulp may finally be diluted to a consistency of
perhaps 0.5 to 1.5 percent (98.5 to 99.5 percent water) as it is fed onto the
Fourdrinier wire.
Water is then removed on passage over the paper machine by drainage,
pressing, and heating, each succeeding step requiring inore elaborate
machinery and costing considerably more per pound of water removed. The
finished sheet may contain only 2 to 3 percent water. Water removed at the
machine by-€rainage is recovered and returned for another cycle of dilution
and conveying. A typical flow diagram is shown in Fig. 2.
A number of mining processes use water to dissolve or melt underground
mineral deposits and convey the material to the surface for further process-
ing. Trona is brought up from depths of thousands of feet below the surface
in Wyoming by pumping water down one pipe and concentrated sodium
bicarbonate liquor up another. Salt deposits are also mined in this fashion.
The Frasch process for mining sulfur is somewhat more complex in that
the water must be heated to 177°C (350° F) to melt the surfur below ground.
24 MANUAL ON WATER
FIG. 2—Flow diagram for paper mill processes.
Air must then be used to help lift the water-sulfur mixture to the surface,
since the sulfur is insoluble in water.
The mounting cost of oil exploration and oil well development has led to
secondary and tertiary recovery of remnant oil from older fields by water
and steam flooding. Specially treated water, brine, or water-steam mixture
is pumped underground to displace oil from zones where it had escaped
capture by primary methods. Water must be treated for this purpose by
rather sophisticated means to avoid plugging the underground formation
with debris.
Other examples of water's job as a conveyor can be found in such diverse
industries as food, where water carries vegetables in canning plants and beets
in the beet sugar mills, and coal, where pulverized coal is transported long
distances by pipeline as a water slurry.
To some extent, almost every industry uses water as a conveyor and
diluent for wastes, just as man does in his own society through sewerage
systems.
Even in the vapor phase, water is an effective conveyor. In steam
distillation processes, the bubbling of steam in a vessel containing an organic
liquid causes a mixture of water vapor and hydrocarbon to be discharged in
a molecular ratio equal to the ratio of the individual vapor pressures of the
two materials. Steam is also used for the evacuation of other gases from an
enclosure. An example is the use of steam jet eductors on turbine condensers
to continually remove noncondensible gases from the condenser shell to
prevent gas-binding the heat-transfer surfaces.
Water for Washing—Water is convenient and economical to use for
general cleanup of industrial equipment. Aside from the aesthetic view,
washup of equipment is important from such practical considerations as
safety (personal injury may result from slipping on debris accumulated on
the operating floor), quality (trash may find its way into the products), and
production (equipment may be damaged by dirt). The nuclear industry
probably has the most elaborate procedures of any industry for washup.
Equipment exposed to accumulations of radioactive particles is throughly
cleansed with detergents and rinsed copiously with water.
Washing of production materials takes a variety of forms, each suited to a

particular type of operation. These may be generally categorized as washing
by dilution, displacement, or extraction.
The electroplating industry provides a unique example of washing by
dilution, since the items being plated pass through many chemical solution
tanks, and carryover from one to the next must be avoided by intermediate
rinsing. Figure 3 shows a typical plant of this type. Modern plating plants
are equipped with carefully designed countercurrent rinse tanks to provide a
thorough washing with a minimum loss of water or production of waste
water. The washing effect is accomplished in the rinse tank by dilution of the
surface film of chemical from the previous reaction tank. Diffusion, of
course, plays a part, too, in an effective rinsing operation.
FIG. 3—Electroplating plant rinse system.
The microelectronic industry requires water of extremely high quality.

The washing of the miniature parts must be done with water that is ion free
and contains no suspended matter because the residue from a drop of
evaporated water on a miniature part would make it defective. The
production of this high-quality water requires the use of extensive filtration
and demineralization techniques.
The paper industry provides an example of washing by displacement in
the removal of residual cooking liquor from pulp on brown stock washers.
These units are essentially vacuum filters, with the filter drum rotation partly
submerged in a vat of digested pulp at about 15 percent consistency. The
pulp forms a mat on the drum as the liquor flows through the screen
covering the drum to a collection pipe inside. When the revolution of the
drum has carried the wet mat out of the vat, upward, and toward the
26 MANUAL ON WATER
opposite side, flat sprays of water wash the pulp and displace the remaining
liquor through the mat to the collection pipe for recovery and reuse.
A similar displacement process is used to remove and recover caustic soda
from mercerized cotton fabric. The recovered caustic is then freed of
cellulosic material by washing it in a dialyser. In this apparatus, caustic on
one side of a membrane is extracted by water on the other side, leaving the
large carbohydrate molecules behind.
Washing by extraction is practiced in many industries. It is used in most
domestic oil refineries to remove salt from crude oil so that the salt will not
foul distillation columns. Water is pumped into the crude oil, amounting to
about 4 percent of the oil flow, and thoroughly mixed for maximum salt
extraction. Emulsion breaking chemicals may then be added, just before the
oil passes into a vessel having electrodes to break the water out of the oil and
cause it to separate by gravity.
Refined sugar is also washed by water extraction during centrifuging. The
wash water removes the film of molasses from the crystals of sugar so that
the final product will be white and free-flowing.
The widespread growth of air conditioning has led to thfc use of water for
washing debris from air. Often the temperature of the wash water will be
closely controlled to provide the correct humidity in the cleaned air. This has
become increasingly important in the aerospace industry, where entire
assembly rooms must be completely dust-free to avoid any dirt accumulation
on machined surfaces or control devices.
Hydraulic classification may be considered as a special form of water-
washing. Water acts as a medium for separation of a desired material from
its contaminants, as in the separation of slate from coal and in froth flotation
in the mining industry.
Water as a Raw Material—Many a schoolboy has been startled by the
demonstration in his chemistry class of the sputtering reaction of sodium
metal with water, evolving caustic soda and hydrogen. Mercury cell caustic
soda is manufactured in much the same way, the sodium in the amalgam
reacting thus:
This is one of many industrial chemical reactions in which water participates

as a raw material.
In some processes, the reactions take place in the vapor phase. This is
illustrated by the first reaction in the chain of events leading to the
manufacture of ammonia, in which steam is reacted with methane in natural
gas to produce the needed hydrogen:
In this reaction, oxygen is split from the water molecule. Coal also reacts
with steam in this way in gas producers to manufacture illuminating gas.
CHAPTER 2 ON USES OF INDUSTRIAL WATER. 27
Water of high quality becomes the bulk of the finished product in the
beverage industry. Each beverage has its own quality standards, ranging
from an exceptionally clear water for carbonated beverages (even traces of
suspended matter greatly reduce the degree of carbonation) to a fully
demineralized water for deproofing of alcoholic beverages.
Water is similarly an important raw material for the drug industry. It
becomes a part of many liquid formulas, including parenteral solutions.
Water may be considered as a convenient vehicle for many chemicals.
Many acids are aqueous solutions of acidic gases in water, such as hydrogen
chloride (HC1), sulfur trioxide (S(b), and phosphoric anpydride (P^s).
Household ammonia is a solution of ammonia gas in water. Caustic soda,
which was described earlier as manufactured by reacting sodium with water,
is kept in an excess of water and handled entirely as a solution by a number
of plants. This saves the cost of evaporation necessary to produce solid
sodium hydroxide.
Miscellaneous Uses—Water is such an inexpensive and convenient mate-

rial that industry has put it to work in countless odd jobs. It is widely used as
a barrier material, such as sealing liquid on the glands of centrifugal pumps
and on gas holders.
Used in large hydraulic clutches, it can be injected and withdrawn from
the system to start and stop a driven load or to adjust its speed. The heat
developed in this service must be dissipated to avoid deposit formations, and
the water must be treated for both scale and corrosion control.
Water is a neutron moderator, making it a valuable shield in the nuclear
industry for protection of personnel from radiation. Radioactive elements
are cut or fabricated in a pool of water by special machines, observed by an
underwater television camera which transmits the picture to the operator.
The operator manipulates the controls at the operating floor, shielded by
perhaps 6 to 9 m (20 to 30 ft) of water from the radioactive object on the
floor of the pool. The water must be of extremely high clarity.
Summary
Water is essential to most industries. Its value is earning increased
recognition. Each plant is becoming cognizant of the need to be alert to its
use of water and to the quality required of each use. A water balance sheet
should be prepared and kept up to date by every water-using industry both
for its own protection and as a part of its obligation as a good neighbor.
The aspects of water sources, quality standards, use, and reuse are
explored in other chapters of this manual. Of one thing we may be certain:
There will be no end to discovering new jobs for water to perform for
industry, nor will its importance in existing jobs diminish.
Chapter 3—Production and

Preservation of Ultrapure Water
The word water probably has as many meanings as there are people
discussing the term. This chapter attempts to define and limit a particular
type of water called "ultrapure" water so that the reader will be in a better
position to draw up his own definition within the scope of his own use of the
material.
History and Background

The preparation of pure water probably ranks among the oldest of all
known chemical processes, having been noted by Aristotle (384-322 B.C.),
who made mention of the fact that pure water was produced by the
evaporation of seawater. Thus, distillation as a method of producing pure
water appears to have been used by the early experimental scientists. Later
on, Pliny the Elder (A.D. 23-79) described methods of designing a still which
included the use of wool in the upper condensing section to collect various
oils as well as water droplets. The Alexandrians refined this system further
by adding a head to the still, enabling them to prepare larger quantities of
distilled liquids, including water and turpentine.
The Arabians made further improvements on this apparatus by cooling
the condenser section or head with water, which led to the production of
essential oils by the distillation of plants as well as alcohol from the
distillation of wine. Distilled water was also produced by this system in
practical quantities for more widespread use. The development of this still
design also made possible the study by the alchemist of hydrochloric, nitric,
and sulfuric acids in the relatively pure state. Evidently, the production of
distilled water was limited to relatively small systems until the 17th century,
when the production of drinking water from seawater became somewhat of a
challenge. In 1683, Fitzgerald patented a process for "the sweetening of sea
water." In 1739, Hales wrote a history of the early attempts to produce
quantities of pure water from seawater in his book Philosophical Experi-
ments.
In the 19th century, the production of pure water by various distillation
systems became quite widespread with patents having been given to Chaplin
28
CHAPTER 3 ON PRODUCTION AND PRESERVATION 29
of Glasgow, Scotland; Rocher of Nantes, France; and Galle and Mazeline of

Havre, France. These patents were aimed mainly at solving the drinking
water problem for the different navies.
In the last half of the 19th century, distillation began to become more
refined with the introduction of the Liebig condenser and the reflux
condenser design invented by Kolbe and Frankland. Vacuum distillation
was developed at about this time by two individuals independent of each
other, Dittmar and Anschutz. Shortly thereafter, in 1855,-Wurtz developed a
special fractional distillation condenser, with refinements on such con-
densers following in 1871 by Linnemann and in 1874 Le Belle andHenninger.
Although much of this development work was related to the distillation of
liquids other than water, the references consistently seem to refer to the fact
that the chemists were attempting to find methods of making water of higher
and higher purity.
In 1878, the first company known to have a business of producing water
stills among its products was founded by a William Barnstead in Boston,
Massachusetts. Shortly thereafter, in 1893, Kohlrausch and Heydweiller
conducted the classic experiment to determine the resistance of absolute
water. They redistilled water 28 times from quartz stills and found that the
resistance was 26.8 megohms at 18°C.
At the turn of the century, the interest in producing pure water increased
not only because of the advances in chemistry, but also because of the
emerging scientific fields of biology and medicine. In 1905, Delavierre was
awarded a patent on a process for the purification of water by subjecting the
water to electrolysis. In 1906, Frank Candy of Westminister, England,
received a patent on the manufacture of special water filters by heating
basalt scum and basaltic lava with a carbonaceous matter in a reduced
atmosphere. This development was the basis of the beginning of a company
which has become very large and is still in existence. That same year a man
named James H. Bridge of Philadelphia was awarded a British patent on the
sterilization and purification of water by the use of ozone. In 1907, another
British patent was issued to an American company, the Sanitary Waterstill
Company of Brooklyn, New York, for what appears to have been a still with
a combination condenser and special storage container. A year later the
Schmeisser compound still was developed in Germany. This was essentially a
multiple-effect concentric-walled still designed to produce water with a
reduced energy input.
With the exception of the development by Candy noted in the foregoing,
the purification techniques to this point were limited almost entirely to
various methods of distillation. In 1909, however, a Doctor Gans, working
in the laboratory for soil culture at the Royal Prussian Geological Institute
in Berlin, succeeded in producing artificial zeolites by melting silicates of
alumina with sodium carbonate and quartz. These synthetic ion exchange
materials were given the name "permutit." The development of synthetic ion
exchange materials has continued from that time on, with notable progress
30 MANUAL ON WATER
having been made by Adams and Holmes in England in 1936 with the
development of the synthetic gel-type anion resins. During the decade
starting in 1940, patents were issued or applied for, or both, on most of the
basic cation and anion exchange materials in use today. Similarly, appli-
cations were made on the four-bed process and the mixed-bed process of
utilizing these various materials, which formerly had been employed only in
single- or two-bed configurations.
In addition to distillation and ion exchange, the historical methods of
producing pure water include the preparation of small quantities by burning
pure oxygen and hydrogen and the use of freeze crystalization techniques.
Both of these techniques will be covered briefly later under methods of
preparation; however, neither was employed to produce any significant
quantities of ultrapure water. Over the past 20 years, numerous other
improvements have been made in the design of stills, ion exchange systems,
various membrane processes, and in the chemical structure of ion exchange
materials. However, with a few exceptions, references to these improvements
will be omitted. The reader interested in pursuing such information is
referred to Properties of Ordinary Water Substance by Ernest Dorsey,
which is not only a very valuable source of reference information, but is the
most complete compilation available of information on water properties.
Also, a review by Greppin on the preparation and preservation of pure water
produced by distillation and ion exchange contains over 100 references.
Uses of Ultrapure Water

Perhaps a short discussion of some of the uses of ultrapure water might
help to give a proper perspective to the reader interested in an understanding
of the term. The original use for distilled water was the requirement for
improving drinking water from brackish and biologically contaminated raw-
water supply. This was, of course, a medical use in the final analysis since it
became obvious that sickness resulted from drinking certain naturally
occurring waters. This medical use continued to expand as the alchemist
developed remedies for the sick, compounding many of their elixirs with
distilled water. As the art of surgery developed, doctors recognized that
incisions or wounds which were washed with potable water oftentimes
became badly infected, generally with unpleasant end results. It is not
surprising that the use of boiled water for sterilization prior its use on
patients soon became recognized as standard medical practice. In this
connection, ultraviolet irradiation of water was discovered and practiced at
about the turn of the century.
With the advent of ether as a general anaesthetic during the 1840's, more
complicated general surgery became commonplace, which led to the dis-
covery of a phenomenon which to this day remains a mystery and continues
to plague hospital patients. It was noted that certain patients who had
undergone surgery which included the irrigation by sterile water of an open
cavity in the body suddenly developed strange symptoms of shock, including

a relatively rapid increase in their body temperature. In many instances, the
temperature would rise to a lethal point for reasons not at all clear to the
doctors in attendance. Eventually, it was discovered that the water which
was introduced into the bloodstream of the body not only had to be sterile,
but also had to be distilled in such a way that the fever-producing substances
were removed. Because the temperature-producing symptom was the most
obvious result of the injection of these impurities, they were termed
"pyrogens" and the reaction on the body became known as "pyrogen shock."
It is said that two ounces of concentrated pyrogenic material properly
diluted for injection would cause pyrogen shock and perhaps death to the
entire population of the United States.
Since pyrogens are the by-product of bacterial metabolism, they are
almost always found in the presence of biological impurities, but because
their detection is more qualitative than quantitative, the mere presence of
bacteria does not necessarily guarantee the presence of detectable pyrogens.
Similarly, because pyrogens are not destroyed by conventional sterilization
techniques, the fact that water has been distilled and sterilized does not
insure freedom from pyrogens. This difficulty has led to the classic hospital
procedure whereby only freshly distilled water from a special still designed to
remove pyrogens may be used for intravenous injections. This observation
was made by Burlage and McAllister in 1936 and has been the procedure
followed by the hospital and pharmaceutical industry in the United States
for the past quarter century. The attitude in Europe has been slightly
different, as indicated by Bauer, who pointed out in 1957 that the use of
potable water that had been boiled and^ filtered had been permitted for the
preparation of injectables up to that time. His investigation indicated that
only distilled water should be used for that purpose, which is the practice
followed by most hospitals and pharmaceutical companies in the world
today.
With the improvement in storage techniques involving ultraviolet irra-
diation and special air filters, pyrogen-free water may be stored for extended
periods of time, making the sizing of distillation apparatus dependent upon
daily and weekly production rather than hourly production. Other methods
of preparing pyrogen-free water have also been investigated, including a
method involving filtered demineralized water which has been electrolyzed
using a direct current of 0.1 mA per square millimeter at 7500 V. Such a
process evidently oxidizes the pyrogens. Another method developed by
Smith involves caustic digestion of demineralized water followed by neutral-
ization with acid.
The widespread use of pyrogen-free water in the preparation of solutions
for intravenous injection has been one of the bases of the present remarkable
improvement in medical progress which has been noted over the past 20
years. Intravenous feeding not only prevents dehydration of the patient, but
also gives the physician a method of maintaining the nutrient intake which
32 MANUAL ON WATER
prevents much of the weakening of the patient that formerly was so

debilitating following surgery. The use of the many chemotherapy methods
of medication which utilize massive intravenous injections is also becoming
commonplace. All of the water used for these procedures must be pyrogen-
free by law and must have been produced by distillation.
Another allied use of pure water is in the preparation of media which are
employed in research involving the growth of live tissue outside a living host.
This entire science is referred to as tissue culture and was the basis of the
development of the polio vaccine, the measles vaccine, and is presently one
of the most important processes used in the study of cancer. The definition
of water purity for this application cannot be stated since the impurities that
prevent the development of cells under certain conditions are not even
known, nor are the quantities measurable. For example, it has been
demonstrated that water which has been processed through ion exchange
resins prior to distillation may not be suitable for successful tissue culture
experiments while raw water passed through the same still works quite well.
Certain heavy metals, such as copper and silver, are extremely harmful while
others are an absolute necessity. The tissue culture medium itself contains
many of the elements normally found in the body tissue, such as potassium
and sodium. Since silica and boron are also evidently relatively harmless,
glass stills are often used as redistillers by technicians working in this field. It
should be noted that impurities harmful to the growth of tissue in some cases
are not ionized to the extent of the normal ionization of the water molecule
and, therefore, cannot be detected by conductivity techniques.
Since the previous uses do not involve really large volumes of ultrapure
water, one might think that the uses for such water are quite limited. On the
contrary, one of the largest users of pure water is continually upgrading its
purity requirements while at the same time increasing its use of the material.
The power industry was probably the original volume user of ultrapure
water and continues to maintain its reputation in this regard. Both the fossil
and nuclear plants are making demands on the water purification industry
for a continuous source of ultrapure water having total solids not including
organic matter and volatile matter in the range of one to two parts per
billion. This concept of purity was originally requested by the atomic energy
industry but is now in demand by the industry in general, particularly where
boilers are operating in the critical range. The seriousness of this purity
problem is probably best exemplified by a 650-M W power plant of modern
design which operates at a boiler feedwater rate of 292 500 kg (650 000 Ib)
per minute. On this basis, two parts per billion of impurity in the feedwater
will amount to 3.78 t (4.2 tons) of scale or residue in the system per year.
As the various disciplines of science become more aware of the materials
or the changes in the substances they are working with, the contaminants in
the materials under study become more important. These disciplines include
chemistry, biology, botany, bacteriology, nuclear science, physics, bio-
chemistry, and electronics. The research conducted in this last general area
has been heavily dependent upon ultrapure water for its success since
essentially all semiconductor and microelectronic development and experi-
mentation has involved the use of ultrapure water as a cleaning agent. It is
conceded that molecular quantities of contaminants can ruin the yield of
production runs of electronic components and make the development of
others completely impossible. It has been postulated that one part of copper
in 10 million billion (1015) parts of water can be the difference between
success and failure. The reason for this critical water requirement is based on
its use as a solvent or final rinse in almost every chemical processing step
used in the production of basic electronic components. The industry has
developed to the point where complete electronic packages are much smaller
than the head of a pin. The circuits within these packages are in the
dimension range of one micrometre, which means that any particulate
matter much larger than 0.2 yum can cause failure of a device.
Prior to World War II, pure water was used in the rinsing of components
of electron tubes. However, it is well known that these tubes failed after a
relatively short life of perhaps two or three years. Following World War II,
the importance of contamination control in the production of electron tubes
became obvious and relatively large water purification plants were installed
to improve the reliability of tube production. This was particularly notice-
able in the television tube industry since each tube produced required
approximately 190 1 (50 gal) of ultrapure water. During this period, it also
became evident that contamination-free electron tubes could be produced
which would have an estimated life of perhaps 20 years. This type of life
expectancy was essential for all components being used in such systems as
the repeater stations installed in the transatlantic telephone cables. It is
important to recognize that, without a source of ultrapure water as a final
rinse for the components of these devices, their production would have been
impossible. As the industry became aware of the importance of con-
tamination-free water in the production of electronic tubes, the semi-
conductor industry was developed in the late 1950's, which added to the
requirements for much larger volumes of water of even higher quality. The
state of the art is fast approaching the point where absolute water, free of
molecular quantities of all types of impurities, will become desirable if not
essential to the production of electronic components.
Description of Types of Pure Water

Industry, the various areas of science, and medical and pharmaceutical
investigators have been defining pure water since the writing of definitions
and standards began. As noted previously, the definition of pure water in
most cases is a very subjective one since the investigator using the water is
often particularly interested in the absence of only certain classifications of
impurities or perhaps individual substances rather than in the general
absence of all matter. Traces of sodium chloride in water being prepared for
34 MANUAL ON WATER
intravenous injection would not be of serious consequence whereas traces of

biologicals and pyrogens would be of the utmost importance. Similarly,
traces of organic matter would be of minimal concern to the chemist running
inorganic analyses. Rather than attempt to define ultrapure water in terms of
resistance or parts per trillion of impurity, it is more logical to refer the
reader to the Annual Book of ASTM Standards, Part 31, Method D-l 193,
Specification for Reagent Water, and to the U. S. Pharmacopia, which list
several different grades of waters stipulated for certain uses. Briefly, these
various standards are as follows:
Type 1 grade of reagent water shall be prepared by the distillation of
feedwater having a maximum conductivity of 20 pmho per centimetre at
25° C (77° F) followed by polishing with a mixed bed of ion exchange
materials and a 0.2 /urn membrane filter.
Type II grade of reagent water shall be prepared by distillation using a still
designed to produce a distillate having a conductivity of less than 1.0 /xmho
per centimetre at 25° C (77° F). This may be accomplished by double
distillation or the use of a still incorporating special baffling and degasing
features.
Type HI grade of reagent water shall be prepared by distillation, ion
exchange, or reverse osmosis followed by polishing with a 0.45 -/xm
membrane filter. Such water shall have a conductivity of less than 1.0 //mho
per centimetre at 25° C (77° F).
Type IV reagent grade water may be prepared by distillation, ion
exchange, reverse osmosis, or electrodialysis and shall have a conductivity of
less than 5.0 nmho per centimetre at 25° C (77° F).
The U. S. Pharmacopia definitions are as follows:
Purified water is obtained either by distillation or by ion exchange
treatment. It cannot contain more than 10 ppm total solids and must meet
the United States Public Health Service regulations for potable water with
respect to bacteriological purity.
Water for injection is water purified by distillation which meets the test
requirements for Purified Water plus the Pyrogen Test.
Sterile water for injection is water for injection that is sterilized and
suitably packaged.
Future definitions of ultrapure water will undoubtedly depend upon the
development of test procedures as well as various combinations of systems of
preparation. Since the contamination of water by molecular quantities of
impurity is becoming of significant interest to certain users, it is likely that
ultrapure water of molecular purity may be defined based upon tests which
have been promulgated by the user himself: These special test procedures
may be based upon the use of the water in carrying out a standard procedure
with known standard methods and materials in such a way that the water
itself will be tested. For example, the purity of water might be defined as that
water which when used in the preparation of tissue culture media would
support the growth of monkey brain tissue for a period of 60 days or longer.
Water passing this specification might then be used to grow another type of
tissue of a sensitive nature. The rinsing of standard semiconductor devices
could also be used to detect molecular contaminants and could be the basis
of a definition of molecular-free water.
Since molecular quantities of impurities may be measured by various
theoretical techniques, it is quite possible that the actual number of
molecules in the water may be used to define the grade of water. Another
measure of the purity or type of water may depend upon the polymetric
structure of the water molecule, which may vary depending upon the
preparation techniques and the previous history of the water. The various
isotypes of hydrogen and oxygen combined to form water molecules can
also be of significance and may be another basis for a series of definitions.
Classification of Impurities
Any discussion of pure water or ultrapure water must include some
parameter on the type and limit of impurity; otherwise, the entire discussion,
whether on the method of purification, the method of handling, or the
method of testing for impurities, will be meaningless. Ultrapure water to
many, especially to those involved in central power station operations and
the pharmaceutical and electronics industries, has been pretty well defined
by the industry. This chapter may, therefore, seem somewhat theoretical and
impractical; on the other hand, it is possible that individuals within the
industries may wish to revalue their definition based upon some of the
changes in the methods of preparation and preservation which are outlined.
Since any investigator considering the use of ultrapure water usually has an
idea about the types of impurities which cause difficulty to his experiment or
process, he should govern his method of preparation and preservation in
such a way as to prevent contamination from the undesirable constituents.
The preparation and preservation based on the elimination of contaminants
might be conducted considering those contaminants which can generally be
classified as liquids, solids, or gases.
In some instances the classification may be unclear. In general, however,
the physical state of the impurity can be postulated. A further breakdown of
the contaminants may be made by classifying the materials âs organics,
which generally are all compounds containing carbon in any form, or as
inorganics, which are essentially all substances which do not contain carbon
in their molecular structure.
Generally speaking, liquid impurities are the most difficult of all to
remove if one is attempting to produce ultrapure water, particularly if the
impurities fall in the classification of miscible organic substances. Examples
of this impurity might include alcohol, liquid detergents, and organic acids
such as acetic or citric acid. Many of these substances that ionize may be
removed by ion exchange, as would be the case with acetic acid, and might
also be removed quite effectively by adsorption. Large-molecular-weight
36 MANUAL ON WATER
impurities including many of the naturally occurring organic liquid con-

taminants present in part-per-million quantities in natural waters can be
removed quite effectively by reverse osmosis and ultrafiltration techniques.
Fractional distillation may be considered in combination with adsorption
for the removal of the liquids such as alcohol; however, if part-per-million
quantities are involved, such systems for removal may be of questionable
value. Under these conditions, destructive digestion using solutions of
potassium permanganate and potassium hydroxide may be far more effec-
tive.
Fortunately, there are very few miscible inorganic impurities. However,
hydrogen peroxide and bromine, in small quantities, are examples. Both of
these materials should be removed by destruction of the molecule as such in
the case of the hydrogen peroxide or by combination of the molecule with
other substances in the case of bromine. Chemical decomposition or
combination should always be considered as a final step in any removal
technique to be considered.
The immiscible organic liquids include all of the naturally occurring oils
which may be derived from contact of a person's skin with the water to be
purified or from petroleum products which may contaminate naturally
occurring waters in a myriad of ways. Sometimes these oils may be separated
by filtration employing special adsorbents which attract hydrophobic im-
purities or by membrane filtration varying from the reverse-osmosis pore
size materials to the membrane filters having pore sizes in the range of one
micrometre. Distillation may also effectively remove these compounds,
particularly if an oxidizing digestion step precedes the distillation process.
Inorganic liquids as such are not common and generally fall into the
classification of low melting metals such as mercury and the synthetic
silicones, which are similar in many respects to organic oils except that silica
is used as the building block instead of carbon. The metallic impurities, of
course, can be removed quite easily by physical separation of the elements
followed by a combination of filtration, ion exchange, and distillation. If the
elements concerned contaminated the water with ionized impurities, ion
exchange would be quite adequate. However, generally such elements would
probably impart colloidal oxide impurities which would dictate the desir-
ability of distillation as a preferred method of removal. The oil-like silicone
impurities should be removed by adsorption, distillation, digestion, and
membrane filtration, as was indicated in the case of the organic oils.
Molecular solid organic impurities form one of the more troublesome
classifications since they cannot readily be removed by conventional ion
exchange and filtration processes. Examples of such impurities include
sugars, impurities from plastics, residues from plastic containers, pyrogens,
and viruses depending upon the size of the viruses involved. Molecular
organic substances which may be present in small quantities in the range of
one or two parts per million may be removed by digestion with strong
oxidizing agents followed by distillation, as was the foregoing case with the
organic matter. The development of reverse-osmosis membranes has made
the use of this system practical.
Chapter 4—Effects of Composition

on Industrial Use
Water is essential to practically all industrial operations, but it rarely

should be used just as it is drawn from stream, lake, or well. The impurities
introduced into water by either natural or man-made pollution can seriously
affect both equipment and product if appropriate precautions are not taken.
Natural Contamination
Absolutely pure water is a laboratory curiosity. Even in the absence of
municipal sewage and industrial wastes, raw water in nature always carries
various foreign substances. The most common impurities in water are
dissolved gases, such as carbon dioxide and oxygen, and soluble mineral
matter, including such metal ions as calcium, magnesium, iron, and sodium
in chemical balance with such anions as sulfate, bicarbonate, carbonate,
hydroxyl, chloride, and others. These substances are dissolved as water flows
over or percolates through the ground. In addition to dissolved substances,
many surface waters carry solids ranging from sand, coarse enough to settle
rapidly, to colloidally dispersed particles indefinitely stable.
Depending upon its source, raw water also carries variable amounts of
disolved or dispersed organic material. Obvious examples are drainage from
peat bogs and cypress swamps, runoff from farms, and decomposition
products from aquatic plants and animals killed by changes in weather or
environment. Most organic matter persists over relatively short distances in
streams because of the purifying capacity of stream organisms that utilize
organic matter as food. Some organic substances, for example, humates and
tannates, are highly resistant to biochemical destruction, and color from
them may persist almost indefinitely.
The natural accumulation of impurities in water is a result of the chemico-
physical constitution of our world, and very little can be done to prevent it.
On the other hand, pollution from the discharge of municipal and industrial
wastes can be controlled.
Industrial and Municipal Pollution

Pollution of streams has been going on ever since the appearance of settled
communities. People found it expedient to dispose of unwanted materials in
38
CHAPTER 4 ON EFFECTS OF COMPOSITION ON INDUSTRIAL USE 39
natural watercourses because they were quickly carried away. Stream

pollution did not become a problem in the United States until the Industrial
Revolution and the introduction of water carriage of sewage around 1885
began to exert their effects. Pollution steadily increased as the manu-
facturing industry expanded and communities grew larger.
Pollution by municipal sewage and industrial wastes has come to be
considered a national problem. Solution of the "national" problem requires,
however, the working out of an untidy mess of local problems for each
drainage basin.
While control of pollution is important to each individual as he drinks,
washes, swims, fishes, boats, or merely gazes upon water, in this chapter it is
pertinent to consider only the adverse effects which impurities in water may
exert upon its use by industry.
Types of Pollution
A watercourse can be polluted either directly or indirectly, and the causes
may be physical, chemical, biological, thermal, and radiological. Although
these species of pollution can be identified, it would be an unusual case that
could be classified wholly under one particular form. The only exception to
this might be instances of pollution by radioactive substances.
Physical pollution can be caused by sludge deposits, floating debris, scum
and foam, turbidity, color, taste, odor, and temperature, to name some of
the more common kinds.
Chemical pollution can take many forms. Increases in hardness, salinity,
and acidity or basicity are commonly encountered. Discharge of toxic metal
ions, such as chromium, copper, lead, mercury, silver, and zinc, is not
uncommon, although their value may often warrant recovery in waste
treatment plants. Iron, manganese, and aluminum are present in most mine
waters and various industrial wastes. Cyanides, flourides, and odorous
organic compounds may be present in waste water. In fact, every chemical
manufactured or used in industry may be present to some extent in streams
receiving industrial waste water.
Biological pollution can be important in water for industrial operations,
as well as in water for drinking. Slime organisms, for example, reduce heat
transfer when they develop on the tubes of condensers and coolers, or impair
the quality of paper.
Heat is regarded as a significant pollutant. Temperature increases in some
streams have seriously impaired their value for cooling water, have increased
their capacity for assimilating wastes beyond the reaeration capacity of the
stream, and have destroyed aquatic life.
The advent and rapid expansion of uses for radioactive elements have
generated an awareness of the possibility of pollution of water by radio-
activity: The great potential danger of this kind of pollution was promptly
recognized and the problem of waste disposal was studied in advance. As a
result, no cases of serious pollution have occurred.
40 MANUAL ON WATER
Effect of Pollution on Industrial Operations

A specific pollutant may be generally undesirable in many industrial
operations. For example, silt resulting from erosion must usually be
removed from water to be employed in process operations, in the cooling of
all but direct once-through equipment, and in the generation of steam. Some
other pollutant may be considered inoffensive except in some unique
situations. Thus, little attention is normally paid to chloride ion up to
concentrations of several hundred milligrams per litre. An extreme effort has
been made, however, to exclude it from the water in nuclear reactors because
of its effect in accelerating stress-corrosion cracking of stainless steel.
An exhaustive catalog of each specialized industrial system in whicli each
possible pollutant found in water has produced an undesirable effect is
beyond the scope of this chapter. For a broad look, it is possible, however, to
consider separately the effect of pollutants on corrosion, on the generation
of steam, on the cooling of industrial processes, on the performance of water
as a constituent in a process for reacting, washing, separating, or trans-
porting materials, and as an ingredient of a product.
Corrosion
Corrosion is probably the most ubiquitous and expensive problem in
water collection and distribution systems, water treatment equipment, and
water-using equipment. Its effects range from such spectacular hazards as
boiler explosions to the downgrading of product quality as a result of the
presence of corrosion products in process water.
The total cost of corrosion in the United States has been estimated at $50
billion annually. The ruptured boiler must be replaced, as must the leaking
water pipe, but there are many other direct or indirect costs resulting from
corrosion. Among these are the increased power required for pumping, the
reduced product yields resulting from inefficient cooling, and the increased
costs from the use of equipment with heavier walls and expensive alloys to
compensate for corrosion.
Although the fundmental corrosion reactions are simple, the form and
rate of corrosion are influenced by subtle and complex interactions of
design, construction, and operation that make corrosion prevention and
control a far from simple task. Corrosion is generally understood as the
deterioration of a metal in its environment. Although nonmetallic materials
such as plastics and wood also deteriorate in their environment, this
deterioration is generally not considered as corrosion. More specifically,
corrosion is the solution of some portion of a metal in its environment as the
result of a chemical, electrochemical, or physical process.
The following tend to intensify corrosion at the metal-water interface,
either by increasing the rate or by increasing the localization of attack.
Since corrosion is essentially a chemical reaction, the rate increases as the
temperature increases. Therefore, corrosion rates are likely to be higher on

heat-transfer surfaces than on non-heat-transfer surfaces for the same water
composition. The corrosion rate for most metals is greater under acid
conditions. Amphoteric metals such as aluminum and zinc will be attacked
in either acid or alkaline environments.
The presence of a depolarizing agent such as oxygen in the water will tend
to accelerate corrosion. Although it is generally true that the higher the
oxygen concentration the more rapid the corrosion will be, a point can be
reached at which the oxide film is maintained and the corrosion reaction is
retarded.
Extremes of velocity, either high or low, affect the corrosion rate.
Erosion may be caused by excessively high water velocity even when there is
no suspended matter present. Softer metals, such as copper and many of its
alloys, are particularly susceptible to erosion even at relatively low velocities.
Low velocity also encourages localized corrosion and ineffective distribution
of corrosion inhibitors. Conditions that cause one portion of the surface to
differ in properties from another will tend to localize corrosion. Deposits
may lead to differential oxygen attack of the metal surface as localized
corrosion. Corrosion products and a wide variety of biological organisms
including bacterial slimes, algae, and mollusks are frequent deposits.
Construction, such as crevices, often create conditions almost identical to
those deposits. The coupling of two different metals tends to accelerate the
electrochemical corrosion of the more anodic metal. Conductivity is the
water property which has the greatest influence upon the rate of electro-
chemical corrosion. The higher the conductivity, the greater the corrosion
current that can flow.
Generally speaking, under any given set of conditions, the rate of
electrochemical corrosion will increase as the difference between the coupled
metals on the electromotive series or the galvanic series increases. The area
relationship of the anode and cathode has a direct effect on the corrosion-
rate. Failure will usually come much more rapidly in a system which has a
small anode and large cathode. For example, the seawater exposure of a
joint prepared by riveting two sheets of copper with steel rivets will lead to
rapid failure, whereas two sheets of steel joined with copper rivets will last
much longer. Mechanical stresses on a section of the metal will tend to
promote localized corrosion.
Most of the economically feasible methods for combating corrosion do
not prevent it but, rather, reduce its rate. Generally, the following control
measures should be considered: selection of corrosion-resistant materials
that provide protective coatings; appropriate mechanical and process design;
and modification of water composition to permit use of a less resistant
metal.
The selection of corrosion-resistant materials of construction is the most
effective control measure, but, unfortunately, it is usually the most ex-
pensive. Even when the general water composition is known, minor details
42 MANUAL ON WATER
of construction or water constituents or conditions can cause failures of the

materials that otherwise would appear to be satisfactory.
Steam Generation
More than a few micrograms per litre of any substance is regarded as an
undesirable pollutant in water to be fed to a boiler producing steam at a
pressure in the range above 13 790 kPa (2000 psi). Even for the pressures of
4137 to 10 342 kPa (600 to 1500 psi) in many contemporary power plants,
dissolved salts as well as suspended solids are often limited to the milligrams
per litre range in the makeup. At lower pressures, operation with appreciable
dissolved solids is accepted practice. General values for water quality
guidelines for boiler feedwater and boiler blowdown quality as developed by
the ASME Research Committee on Water in Thermal Power Systems are
listed in Table 1.
Cooling of Industrial Processes

Almost any quality of water, including seawater, has been used for
industrial cooling in once-through systems, where heat removed from
process equipment increases the temperature of the water on its way to some
other use or to waste. In many such systems, it is not even considered worth
the cost to clarify the large flow of cooling water to remove suspended solids.
When silt accumulates on heat-transfer surfaces sufficiently to impair their
performance, it is removed intermittently by either mechanical or chemical
means. In many areas, reduction of temperature in water discharged from
once-through cooling systems now is required before returning effluent to
the original source.
A more or less universal nuisance in once-through cooling systems is the
accumulation of slime organisms, unless they are kept under control by
periodic addition of chlorine to the water.
TABLE 1 — Boiler-water concentration limits.

Boiler Pressure Total Solids, Alkalinity Suspended
at Outlet, psig" mg. litre (Total), mg litre Solids, mg litre Silica, mg litre
0 to 300 3500 700 350 125
301 to 450 3000 600 250 90
451 to 600 2500 500 150 50
601 to 750 2000 400 100 35

751 to 900 1500 300 60 20
901 to 1000 1250 250 40 8
1001 to 1500 1000 200 20 2.5

1501 to 2000 750 150 10 1.0
to 2001 500 100 5 0.5
"Metric conversion: 1 psig - 6.895 kPa.
Inorganic coatings, frequently of a slimy character, also result from the

deposition of hydrous oxides of iron or manganese, or both, from the
cooling water. Harder deposits of calcium carbonate scale are produced only
if the once-through cooling water is practically saturated as it reaches the
heat-transfer surface.
Where heat is removed from the cooling water by evaporation, as in a
spray pond or a cooling tower, solids originally dissolved in the initial supply
are concentrated in the water recirculated to the cooling operation. As a
result, the conditions necessary for deposition of a scale of calcium
carbonate are more frequently attained.
Since a cooling tower inevitably functions as a giant air washer, a
recirculating cooling system is always being seeded with airborne bacteria
and other minute living organisms. It also accumulates dust, becomes
saturated with oxygen, and collects such gases as carbon dioxide, sulfur
dioxide and trioxide, hydrogen chloride, hydrogen sulfide, or ammonia
released by the burning of fuel or by chemical processing. Fouling of heat-
transfer surfaces, corrosion of metal lines and equipment, and damage to the
surfaces of cooling towers by chemical attack, bacterial attack, or both, are
commonly controlled by a variety of water-conditioning techniques.
Processing Materials
Some industrial processing operations can be conducted with almost any
quality water. An example is the washing of fines from bituminous coal. In
other processes, only demineralized water of the highest quality is employed.
The manufacture of microelectronic components is an example often cited,
but the production of electrolytic tin plate free of pinholes requires water of
similar high purity. Between these extremes lie many special situations.
There is no standard approach to providing water for industrial operations.
Consider a few of the many criteria:
Color due to the natural decomposition of vegetation, to the oxidation of
dissolved iron or manganese, or to various constituents of municipal or
industrial wastes must be minimized in water used for processing natural and
synthetic fibers, washing and dyeing the fabrics woven from them, and
manufacturing clear plastics or quality paper of high brightness.
Suspended solids in a range of particle size from those which can be
readily removed by sedimentation to those so small that they remain
dispersed for long periods of time are an almost universal pollutant of river
water. In relation to the trouble they cause in most process operations, the
cost of their removal is sufficiently low as to justify clarification, either as an
initial step or as an additional benefit of softening by precipitation.
Dissolved salts in water that might be considered of the highest quality for
drinking must nevertheless be considered as pollutants if they cause oper-
ating difficulties, impair product quality, or impose costs for treatment to
remove them.
44 MANUAL ON WATER
Water for rinsing steel surfaces to be electroplated or painted should not

leave hygroscopic salt deposits on the dried metal because subsequent
blistering of the finish in humid atmospheres may otherwise occur.
Iron and manganese cause a general staining of all water fixtures, dulling
of shades in dyeing of textiles, and introduce undesirable color in washing of
clear plastics, white paper, and textile fibers and the products woven from
them.
Calcium and magnesium in water used for washing textiles tend to cause
uneven dyeing. Water low in hardness is required for some electroplating
processes, washing and bleaching of paper, pulp, leather tanning, and
laundering.
Copper and mercury at levels so low as to be barely detectible cause
cracking of alumimum alloys.
Incorporation in Product
Taste and odor are vital considerations in water used for brewing beer or
bottling nonalcoholic carbonated beverages, for food processing, and phar-
maceutical industries.
Color is a critical characteristic in water for club soda.
Dissolved salts should be low in water used for making artificial ice to
minimize opaque accumulations of solids in the last water frozen.
Calcium and magnesium in water used for cooking peas and other
vegetables for canning toughen them undesirably.
Keeping Water Reusable

From the very beginning of the Industrial Revolution, industry has had to
accept costs due to natural impurities in water. The early history of the steam
boiler contains references to the "fur" deposited from hard water, which
must be removed frequently to avoid blistering and failure of the boiler shell
directly above the furnace. Deposits of relatively insoluble compounds of
calcium and magnesium—the "hardness" in the water—as an insulating scale
on heat-transfer surfaces in contemporary condensers and coolers as well as
in boilers are prevented only by controlled pretreatment and chemical
conditioning of the water.
Not only dissolved substances present in natural waters, but also silt from
erosion of the land, as well as slime produced by bacteria present in water,
foul industrial equipment, interfere with heat transfer, or impair the quality
of products.
Even the oxygen dissolved in water from the air, vitally necessary to the
normal balance of life in a stream or lake, accelerates failure of pipelines and
equipment by localized corrosion.
The fact that a water supply either in the natural state or after treatment in
a municipal water plant meets the standards set for drinking water does not
mean that this water is suitable for industrial use. Further treatment usually
is needed to minimize its total cost.
While natural impurities can always cause some kind of trouble wherever
water is used by industry, man has compounded his problems exponentially
in two ways. While multiplying his numbers, he has increased tremendously
his control of power and his capabilities for manufacturing products to meet
many new wants in addition to old needs. Throughout the industrial
expansion of the past two centuries, he has conveniently flushed his sanitary
and his industrial wastes into the nearest stream or lake.
From the strictly economic viewpoint for our society as a whole, the
optimum control of this pollution would be that which leads to the lowest
total of all the costs to all the groups using and reusing the water through its
complete path back to the reservoir of the oceans. But water also has
aesthetic and recreational values which are leading us as a society to pay
more than the minimum number of dollars which would be needed for the
optimum level of control on a strictly economic basis.
Keeping water reusable provides a promising alternative to transportation
of water from a diminishing number of sources at increasing distances from
the point of use, and to the recovery of usable water from the. oceans.
However, the cost of controlling entry of the wastes of society into our
presently available water will not be small.
Chapter 5—Treatment of Process

Water and Waste Water
The purification of water for industrial use can be very complex or rela-
tively simple, depending upon the properties of the raw water and the degree
of purity required. There are many methods and combinations in use, but all
are covered in three basic processes: physical, chemical, and physicochemical
treatment.
A fourth basic process, biological treatment, is often employed where
waste water must be purified before it may be discharged. This process takes
advantage of the ability of living microorganisms to induce a variety of
chemical and physicochemical reactions. Treatment of waste water is usually
a much more complicated operation than the production of process water of
appropriate quality.
This chapter discusses these processes as they are used in the treatment of
industrial water and waste water and outlines some of the technical problems
and basic considerations involved.
Physical Treatment
Physical treatment pertains to those processes in which there is separation
of impurities from water without chemical change. The common methods
are sedimentation, straining and filtration, multiliquid phase separation,
degasification, dilution, removal of entrainment, distillation and stripping,
underground discharge, and ocean discharge.
Sedimentation
Sedimentation is the process by which the force of gravity acts on particles
heavier than water and causes them to move downward and settle out.
Surface waters contain varying amounts of suspended material, and this
process is used to clarify raw water either by simple sedimentation or with
the addition of chemical coagulants. The tanks in which this action takes
place are called sedimentation basins.
Particle size, weight, shape, frictional resistance, and viscosity play an
important part in the design of such tanks. Theoretical calculations usually
are based on modifications of Stoke's law and the assumption that the
CHAPTER 5 ON TREATMENT OF PROCESS WATER AND WASTE WATER 47
particles are spheres. Many factors must be considered in design, and all of
these must be modified as judgment and experience dictate. The theory of
design will not be given here; it will suffice to state that effective clarification
can be accomplished in tanks that are properly designed and operated.
Settling basins may be constructed of earth, wood, concrete, or steel. They
may be rectangular or circular. The usual period of retention ranges from 4
to 12 h. Where the water contains large amounts of settleable solids, the
basins often are provided with mechanical scrapers which move the settled
sludge to a sump from which it is forced out by the hydraulic head of water
in the basin. In other instances, manual sludge discharge valves are provided
to remove some of the settled material; it is necessary to remove such units
from service at regular intervals for thorough cleaning.
Inlet and outlet distribution of the fluid must provide for uniform flow in
the basin if efficient operation is to be attained. In rectangular tanks, the
influent is distributed by a trough equipped with adjustable sluice gates or
plug valves, and the effluent is usually discharged over a weir or series of
weirs. Circular units are provided with a center distributor and an overflow
launder at the periphery of the tank.
Short-circuiting often plagues the operation of rectangular basins,, thereby
decreasing the theoretical retention time. This can be overcome by im-
proving the inlet distribution or by effective use of muffle boards at the inlet
and sometimes just ahead of the effluent weir.
Depending upon the character of the silt and suspended matter, plain
sedimentation may provide as much as 70 percent removal; where coagu-
lants are employed, removals may be as high as 95 percent.
Many industrial waste waters contain organic and inorganic settleable
solids that must be removed before final discharge. These solids respond to
the process just described, and settling basins of similar design are used for
the purpose. Retention periods are usually much shorter than for raw
waters, rarely exceeding 2 to 3 h. Long retention periods are detrimental
where the solids are all organic, because bacteria} action may take place.
This may result in anaerobic conditions and the generation of odors. The
settled sludge will have a tendency to float if considerable gas is evolved, and
this will tend to nullify the operation.
Air flotation is very effective for removing flocculent solids that settle very
slowly or have a tendency to float. The air is dispersed in the liquid under
pressure and the pressure released just prior to discharge into a receiving
basin. The finely divided air floats the jsolids to the surface, where they are
skimmed off. Effective separation is provided at minimum cost by this
method.
Straining and Filtration

Strainers and filters may be used where it is necessary to remove floating
or suspended solids from water, either as a further step after sedimentation
or where space precludes installation of settling basins.
48 MANUAL ON WATER
Several types of strainers are in commercial use. Cloth, metal gauze,

rotary drum, and rotary disk are the usual kinds. Cloth and metal gauze
filters are used where the suspended materials are fairly fine and in low
concentration. Cheesecloth or similar material is rolled around a perforated
mandrel and installed in a cast iron housing. The water passes through the
filter into the drum and the finely suspended material is retained on the
cloth. When the pressure drop reaches a predetermined value, the unit is
taken out of service and the cloth washed or replaced. These units will do a
good job of straining but are bulky and have a fairly high loss of head.
Metal gauze is becoming popular as a strainer. Here the size of the mesh
can be rigidly controlled down to several micrometres and selective straining
can be performed.
Rotary drum and disk strainers are quite widely used for coarse separa-
tions. The drum type comprises a rotating drum covered with a screen of the
proper mesh size for the purpose and installed inside a housing. The drum is
driven at slow speed by an /electric motor. Valving is provided to backflush
automatically a portion of the screen in its rotation and thereby keep it
clean. These units are built to handle up to several thousand litres per minute
at a low pressure drop. This unit may be modified by replacing the screen
with perforated porcelain disks.
A drum strainer is used where floating solids must be removed from waste
water. This is a drum frame, covered with screen and rotated in a concrete
channel. The water flows through the screen and out of the center of the
drum. Provision is made for backwashing the screen as it rotates. This unit is
for gravity flow only.
The disk unit is a perforated disk set at an angle in a channel where the
water must flow through the disk. The solids are scraped or brushed off the
portion of the rotating disk that is above the water surface. This is a gravity-
flow unit.
Filtration of water is carried out in either pressure or gravity units
containing graded sand, calcite, magnetite, anthracite, charcoal, or finely
divided, relatively insoluble materials such as diatomaceous earth. Silica
sand is the most common filtering medium for cold, neutral waters because
of low solubility, low cost, and the relatively high specific gravity of sand
particles. Pressure filters are usually designed for a downflow rate of 81 to
122 litres/min/M2 (2 to 3 gal/min)/ft 2 of surface area, and a reverse-flow
backwash rate up to 68 litres/min (18 gal/min). The filter medium usually is
specified on the basis of effective, size and uniformity coefficient. The
effective size of filter medium is usually specified to be not more than 1.65
and is always greater than 1.0.
Crushed anthracite coal, because of its relatively low solubility, is usually
specified as the filter medium where hot, strongly acid, or strongly alkaline
waters must be filtered. The anthracite filter medium usually is backwashed
at lower velocities than sand of the same size. Anthracite filter medium
weighs only about 80 kg/m 3 (50 lb/ft3), approximately one half the bulk
density of sand.
Diatomaceous earth is often used as a filter medium. In one type, the

diatomaceous earth is supported by a porous stone or wire mesh element and
removed by backwashing. In other types, the filter material is deposited on a
rotating filter drum in a fairly thick layer. A doctor blade continuously
scrapes off a thin film and thus exposes a fresh surface. Precoated filters
have the advantage of high-quality effluent, low weight, high capacity for
their size, and sometimes lower first cost. They have the disadvantage of high
pressure drop, high operating maintenance and repair costs, and rapid
depreciation and obsolescence. Poorly designed units are often unde-
pendable in operation even in the care of a competent and reliable operator.
A diatomite filter must have two to four times the filter area of a sand filter
of equivalent capacity. Even with the larger area, the pumping power needed
may be two to four times higher than for the equivalent sand filter. There are
numerous types of precoat filters which employ a variety of precoat
materials.
It should be understood that filtration will not remove true color from
water. This can be eliminated only by use of a suitable coagulant dispersed in
the water prior to filtration.
Multiliquid Phase Separation

One or more nonmiscible liquids can often be physically separated from
water as a result of differences in specific gravity. Gravity separations of this
kind are performed on a small scale in laboratory separatory funnels and on
a commercial scale in specially designed equipment [/].' The oil and water
separator for oil refinery waste water is an example of one large-scale
application of this process [2].
Frequently there is not much difference in the specific gravities of the
liquids, and one of them may be dispersed in a finely divided, globular state.
Thermal convection currents under such conditions may be strong enough to
prevent efficient operation of the equipment. It is therefore always desirable
to minimize thermal currents by careful design of gravity separator systems.
An equalizing basin ahead of the treatment unit will minimize temperature
and other fluctuations that may disturb proper operation. Agglomeration of
the dispersed phase by chemical or physical means will greatly increase the
efficiency of the separator.
Centrifugal equipment is used for multiliquid phase separations where the
value of the separated product justifies the relatively high capital cost of the
equipment. Use of a centrifugal separator to remove traces of suspended
water and solids from lubricating oil is an economical application of this
process. The differential separating force between the two or more liquids
can be greatly increased with such equipment. The strength and cost of the
materials from which centrifugal separators are constructed place a limit on
their size and capacity.
Air flotation is widely used in this type of separation. The oily water is put
'Italic numbers in brackets refer to the list of references appended to this chapter.
50 MANUAL ON WATER
under a pressure of 207 to 345 kPa (30 to 50 psi) and air is diffused into the
liquid. The pressure is released just prior to discharge into a separating tank
where the fine air bubbles rise to the surface carrying the particles of oil with
them. A rotating skimmer sweeps the floated oil to a collecting trough from
which it flows into a receiving tank. These systems can easily produce 90
percent separation with very short retention periods, sometimes of no more
than 30-min duration [3].
Desgasification
Stripping operations, such as physical deaeration and aeration, are
processes for removing undesirable dissolved gases from water. One or more
dissolved gases, such as oxygen [4], carbon dioxide [5], ammonia, or
hydrogen sulfide, are removed by exposing large surfaces of the liquid to a
vapor phase deficient in the gases to be eliminated.
Degasifying equipment can be classified as ebullition, tray or drip, and
spray units. In the ebullition type, a suitable gas is bubbled through the
liquid for a period sufficient to change the vapor pressure equilibrium and
enlarge the liquid surface. This principle was widely used in the past for
hydrogen sulfide and carbon dioxide removal; it is now employed mainly to
separate carbon dioxide and oxygen in deaerating feedwater heaters, usually
operating under a few pounds pressure.
Tray or drip gas-stripping equipment exposes large surface areas to the
desired vapor pressure conditions by slowly flowing the water in thin layers
over plates, channels, slats, or packing and allowing it to fall or trickle in
droplets or thin sheetsy^Deaerating feedwater heaters operate on this
principle to remove dissolved oxygen and carbon dioxide from boiler
feedwater. The heaters are usually operated under a few pounds pressure and
at temperatures above 100°C (212°F). Degasifiers used to remove carbon
dioxide where alkalinity is controlled by acid addition, or by blending
hydrogen zeolite effluent with sodium effluent or after cation exchange,
operate on this principle. Closed wooden tanks containing many stacks of
trays allow the water to drip down while air blown up through the tank
escapes at the top.
Spray and spray-ebullition gas strippers utilizing atomizing nozzles,
ebullition, and entrainment to divide the water into fine droplets for
exposure to the desired vapor pressure conditions are finding wide use, not
only in deaerating feedwater heaters, but also with vacuum deaeration of
carbon dioxide and oxygen from cold water and ammonia from hot water.
Vaccum deaeration is carried on in one or two stages, usually in vertical steel
tanks where the vacuum is maintained by steam-jet ejectors.
With the exception of vacuum deaeration, some means must be employed
to maintain the partial pressure of the undesirable gases well below the point
at which stripping will cease. In a gas-stripping basin or tower, the air that is
pumped or blown through the equipment carries the undesirable gas away to
the atmosphere. Steam with a low partial pressure of free oxygen is the
stripping agent in deaerating feedwater heaters. A vent condenser is usually
included to permit the incoming feedwater to absorb most of the heat in the
stripping steam.
Dilution
High salt concentrations in waste water from the oil and alkali industries
present a difficult disposal problem. Similar problems occur in other
chemical plants where sodium sulfate and organic solids are present in high
concentrations in effluents. Regulatory agencies have set limiting concen-
trations in the receiving streams that restrict the amount of dissolved matter
and suspended matter that may be discharged.
Where inorganic salts such as sodium chloride, calcium chloride, and
sodium sulfate must be discharged, there are two methods in use. One is by
dilution with uncontaminated water, and the other is by controlled dis-
charge. In those plants where the waste stream is small and there is available
considerable uncontaminated waste water (for example, from cooling
operations), the two are blended to conform with the concentration allow-
able in the plant discharge. Where the amount of clean water is insufficient
for diluting all the concentrated waste, any surplus of the latter is held for
controlled discharge.
Controlled discharge takes advantage of the increased stream flows during
winter months or rainy seasons to dispose of concentrated wastes that would
otherwise overload the receiving water. The wastes are stored in ponds and
released during periods of heavy runoff at a rate that will not exceed the
allowable concentration in the river. This operation may require large areas
of land for storage lagoons, but it is often the most economical method of
handling the problem.
When there is sufficient flow in the receiving stream to assimilate the
biochemical oxygen demand (BOD) load from organic matter, dilution can
provide satisfactory stream assimilation and eliminate areas of heavy oxygen
depletion. This method has certain limitations but it has been used success-
fully in some areas.
The dissolved salts and sludge in evaporators, boilers, and cooling towers
are concentrated in the circulating water and these concentrations must be
controlled by blowdown. Blowdown is thus a dilution process whereby
concentrated solutions and suspensions are replaced by more dilute water.
Intermittent boiler blowdown usually discharges directly to an atmospheric
drain tank, with the result that all of the heat of the liquid and of the flashed
steam are lost. Continuous blowdown systems make it possible to recover
this heat. Heat-recovery systems are available for both high- and low-
temperature operation. Heat-recovery equipment for high-temperature
blowdown may consist of a flash tank to salvage flash steam or of a surface
heat exchanger for transferring the blowdown heat to the feedwater just
52 MANUAL ON WATER
before it enters the boiler. Heat-recovery flash tanks or surface exchangers

for low-temperature blowdown usually transfer the blowdown heat to the
cold feedwater in or ahead of the deaerating feedwater heater. Heat-recovery
systems for high- and low- level blowdown can be used in series where the
overall heat balance of the plant makes such equipment economical.
Figure 1 shows the percentage of the boiler water that is flashed to steam
as it is throttled from various pressures to atmospheric pressure. It also
shows the total heat of the boiler water at various pressures.
Blowdown from cooling towers is often called purge water. Windage or
other water losses are sufficient in some cases to limit the concentration of
salts of the cooling tower water. Where additional dilution is required, a
continuous stream of water is run to the waste water system.
FIG. 1—Flashed steam recoverable from continuous blowdown systems.

Removal of Entrainment
One of the fundamental problems in boiler design and operation is the
mechanical separation of solutions and solids from the steam. All boilers
must include entrainment separators. The simplest method is to provide
sufficient disengaging space above the normal waterline in the steam drum.
It is economical for industrial steam generators to use dry pipes, baffles, or
special types of separating devices to reduce the size of the steam drum. Most
modern steam drums are furnished with carefully designed and proved
internal equipment. Steam separator equipment is mechanical in nature and
operates by gravity, centrifugal force, and surface tension or by a com-
bination of these forces to remove liquid and solid contamination. Pre-
liminary washing of the. steam by feedwater or steam condensate may be
done in steam separating equipment. Devices for overcoming entrainment of
liquids in gas will not appreciably separate a mixture of gases, such as steam
and carbon dioxide, or steam and hydrocarbon vapors.
Entrainment separation efficiency in a steam boiler is controlled not only
by the mechanical equipment, the steam rate, and the water level in the
boiler drum, but also by the surface tension of the dissolved solids and the
liquids in the boiler water. There is usually a limit for dissolved solids above
which, at normal steam rates, priming and carryover will occur. No general
standard for allowable dissolved solids can cover all types of steam
generators. While operators at one plant will insist that concentrations must
be kept below 1000 mg/litre dissolved solids, operators at another plant with
identical equipment, but with different water conditions, may be able to
operate satisfactorily with 5000 mg/litre.
The same general principles apply to reduction of solids carryover in the
NOTE—For boiler pressures between 689 and 5516 kPa (100 and 800 psig), use "flash" pres-
sure curves slanting from lower left-hand to upper right-hand corner and the bottom axis. For
boiler pressures above 5516 kPa (800 psig), use "flash" pressure curves slanting from lower
right-hand to upper left-hand corner and top axis.
This chart is used to calculate the percentage of boiler water discharged by a continuous
blowdown system that can be flashed into steam at a reduced pressure and recoverable as low
pressure steam for heating or process.
EXAMPLE—A boiler operates at a pressure of 3103 kPa (450 psig). Continuous blowdown
amounts to 4540 kg/h (10 000 Ib/h). What percentage of blowdown water can be recovered as
flashed steam at 5 psig pressure?
SOLUTION—Locate 3103 kPa (450 psi) on left-hand axis. Follow horizontally toward the
right to the intersection with 34 kPa (5 psi) "flash" curve (point A). Drop vertically downward
to the bottom axis and read 25.5 percent. Twenty-five and a half percent of 4540 kg/h (10 000
Ib/h) blowdown = 1158 kg/h (2550 Ib/h) of flash steam at 34 kPa (5 psig) pressure.
These curves have been prepared from the formula:
Flashed steam, percent
where:
HB- heat of liquid at boiler pressure in Btu/lb,
HF= heat of liquid at flash pressure in Btu/lb, and
VF- latent heat of vaporization at flash pressure in Btu/lb.
54 MANUAL ON WATER
vapor from evaporators in process operations. This is especially important

where barometric condensers are used to provide vacuum, because the tail
water is usually discharged directly to the sewer, and minimum solids are
desirable to keep product losses and the BOD of the waste water to a
minimum.
Distillation
Distillation is the time-honored method for preparing high-quality pure
water. Both dissolved and suspended solids can be removed almost com-
pletely by this purely physical process of evaporation and condensation. The
condensate from a water distillation unit with good steam separator
equipment, and at normal evaporating rates, should contain no more than a
few milligrams per litre of dissolved solids and may contain only a fraction of 1
mg/litre. The suspended and dissolved solids of the feedwater remain in the
evaporator salines and are removed by blowdown and descaling operations.
The high quality of its product gave the distillation process an initial
advantage over other processes for the preparation of boiler feedwater.
Distillation transfers a large part of the boiler water treatment problem from
the plant boilers to low-pressure evaporators where it can be more success-
fully handled. Single-effect and multiple-effect evaporators have found wide
use in plants where large quantities of low-level heat or exhaust steam
otherwise would be wasted. Low-level, or so-called waste heat, is not
worthless; its cost is too often underestimated in the overall plant heat
balance and in the original plant design.
Compression distillation equipment may be economical where low-level
heat is not available in sufficient quantity or is being put to more valuable
use. These units, employing the heat-pump principle of salvaging the latent
heat of evaporation, have the heat economy of 15 or more effects of
multiple-effect evaporation. Compression stills may be electrically driven or
may be powered by other prime movers. Diesel-driven compression dis-
tillation units can produce from 67 to 112 kg (150 to 250 Ib) of distilled water
per pound of fuel oil.
In industries where only a few milligrams per litre of dissolved solids are
allowable in the process water, either distillation or one of the deionizing
processes (discussed later in the "Chemical Treatment" section) can be used.
The choice will depend largely upon the concentration of dissolved solids in
the raw water.
Distillation equipment should be so designed that the scale and sludge
problem can be handled with minimum labor. The problem is similar to that
of a boiler except that the temperature usually is lower. Heat-transfer rates
at lower temperatures are more vulnerable to the insulating effects of scale
and sludge deposits.
Extremely high-quality distilled water can be obtained by multiple or
fractional distillation. In the latter process, part of the product is condensed
and refluxed to scrub entrained salts and solids from the steam being
produced in the vapor generator. It is doubtful whether this quality will

surpass that from a well-operated mixed-bed deionizer.
Underground Discharge
Underground discharge is a method for disposal of liquid wastes into
permeable strata in the earth. It is a satisfactory method only in areas where
other methods are extremely expensive or unavailable. The permeable strata
must be so located that the liquid waste pumped into them will not
contaminate strata used for water supply. These permeable layers are very
deep, about 1200 to 1500 m (4000 to 5000 ft) being usual. Furthermore,
injection wells are very expensive.
The waste water must be sterile and free from suspended solids. Extensive
pretreatment is essential to remove silt, suspended solids, oil, and emulsions.
Chlorination is generally provided to make sure that it is free from micro-
organisms. Additional pressure above that developed by the depth to the
aquifer is necessary to force the water into the strata; this may be as high as
14 000 to 20 000 kPa (about 2000 to 3000 psi).
This method has been used in only a few locations because of the extensive
study required and the high cost of the installation. It does have its
applications, however, and should not be overlooked.
Ocean Discharge
Ocean discharge of waste water is practiced by plants in coastal locations.
Actually, it is a dilution method. Consideration must be given to the path of
ocean currents in the area of the outfall to ensure that the waste water will
not be carried to the shore and contaminate bathing beaches. The major cost
for this method is construction of an outfall to such distance that safe
disposal will result.
Desalination
It has been estimated by many experts that by 1980 we will require nearly
2/2 X 109 m 3 /day (600 U. S. billion gallons of water per day), which is double
/ur present demand. Recognizing the need, and the benefits to be gained
from saline water conversion, Congress enacted the Saline Water Act in
1952, and extended it by amendment in 1955 to provide for a program of
research and development having as its primary objective the development
of economical methods of producing freshwater from sea and brackish
waters. The Department of the Interior, through the Office of Saline Water,
sponsors scientific research and development through Federally financed
contracts and grants, by research in Federal laboratories, and by stimulating
and coordinating private activities in this field.
In 1958, a new law authorized construction and operation of not less than
56 MANUAL ON WATER
five demonstration plants to test some of the more advanced conversion

processes. The methods selected were:
A. Multiple-effect long-tube vertical (LTV) distillation.
B. Multistage flash distillation.
C. Forced circulation vapor compression distillation.
D. Electrodialysis.
£". Freezing.
These demonstration plants have been built and operated. Complete data
including costs on these plants have been published elsewhere. A brief
description of each will be given here.
Method A—The seawater plant at Freeport, Tex., has a capacity of 3785
m 3 /day (1 Mgd) and uses multiple-effect LTV evaporation.
Like many distillation processes, this plant is based on a system of
multiple effects to achieve the degree of heat recovery necessary for a
favorable economy ratio. The plant regularly operates at a ratio of 10+1 or 4.5
kg (10 Ib) of product water to 0.45 kg (1 Ib) of steam. The basic unit in the
multiple-effect process is the LTV evaporator, which operates with a
downward flow of seawater as a film inside the vertical tubes. Steam heats
the outside of the tubes, and the vapor given off by the first effect becomes
the steam source for the second effect. Brine effluent concentration is three-
to four-fold. Product water regularly runs less than 50 mg/litre total
dissolved solids.
Method B—The seawater plant at San Diego, Calif.,2 has a capacity of
3785 m 3 /day (1 Mgd) and uses multistage flash evaporation. The process
involves flashing vapor from the brine at successively lower temperatures
and correspondingly higher vacuums and condensing the vapor on tubes in
the upper portion through which the incoming seawater is being carried.
Method C—The brackish water (about one-half the salinity of seawater)
plant at Roswell, N. M., has a capacity of 3785 m 3 /day (1 Mgd) and uses
forced-circulation vapor compression'. Start-up heat is added by means of a
boiler. During operation, however, additional heat for the evaporator is
added by a large compressor that compresses the vapor, adds heat, and
vaporizes more brine. Inlet pressure to the compressor is essentially atmo-
spheric at 103 kPa and 100°C (15 psia and 213°F); outlet pressure is 186 kPa
at 111°C (27 psia at 232°F).
Method D—The brackish water plant at Webster, S. D. (total dissolved
solids, approximately 2000 mg/litre), has a capacity of 946 m 3 /day (250 000
gpd) and uses the electrodialysis process. This process involves the transport
of salt ions through permeable membranes driven by an electric current. The
positive ions move toward the negative poles, the negative ions toward the
positive poles, and the membranes are permeable to either positive or
negative ions. Water spaces between the membranes in the stack are thus
alternately enriched and depleted of salt ions.
Method E—The seawater plant at Wrightsville Beach, N. C., has a
Relocated, 1964, at U. S. Naval Base. Guantdnamo Bay. Cuba.
capacity of 757 m 3 /day (200 000 gpd) and employs the controlled crystal-
lization freezing process. This process uses direct-contact heat exchange
between a secondary refrigerant and seawater, and centrifugal separation of
ice and brine.
Solar distillation of seawater as a source of freshwater is an old process
that includes all those methods in which the energy of the sun is collected
and used immediately, or after a period of time, in evaporation of freshwater
from seawater. While energy from the sun is abundant, it is intermittent and
of low intensity. In general, solar distillation units consist of sloping glass or
plastic plates covering a long, narrow, black-surfaced basin. The transparent
sloping surfaces transmit solar energy which is absorbed in the basin. The
seawater in the basin, warmed to a temperature as high as 71°C (160°F),
vaporizes into the air space, condenses on the cooler transparent cover
surfaces, and runs down the sloping cover to troughs and storage.
In addition to the processes described in the foregoing, salt removal by
reverse osmosis is being evaluated in several demonstration plants, and this
process shows considerable merit for brackish water.
Chemical Treatment
Chemical treatment pertains to those processes in which separation of the
impurities from water involves chemical alteration of the contaminating
material. It may include precipitation, ion exchange, sequestering reactions,
gas removal, oxidation and reduction reactions, chemical control of bio-
logical growths, and sterilization.
Precipitation
When certain soluble salts are added to an aqueous solution, some of the
free ions may react to form comparatively insoluble compounds. The
precipitate can be separated by filtration or by decanting the clear liquid
after settling. Precipitation occurs in accordance with definite laws govern-
ing the combining weights of the reactants and their solubility products [6\.
Many of the methods described in the following depend upon precipitation
of impurities. ,
Softening—The earliest commercial chemical precipitation process was
the addition of hydrated lime (Ca(OH)2) to water for removal of bicarbonate
hardness. Lime converts bicarbonate hardness to relatively insoluble calcium
carbonate:
The process can be carried out in either batch or continuous equipment. In

either type, close control of the lime dosage is necessary to approach the
theoretical minimum hardness. Chemical cost is usually law if the lime
particles are sufficiently small and if good agitation is provided. Fine
58 MANUAL ON WATER
particles result from proper slaking of quicklime or through dispersion of

hydrated lime [7].
Removal of carbonate hardness by heating the water to drive off the
dissolved and half-bound carbon dioxide accomplishes the same result as
adding lime:
Water containing noncarbonate hardness is only partially softened by

adding lime alone. Noncarbonate hardness can be removed by adding soda
ash (sodium carbonate):
Soda ash usually is combined with lime to remove both kinds of hardness in
one step. The two reactions can be carried out simultaneously, hot or cold, in
batch or continuous equipment. A coagulant generally is used to improve
the separation of the solid phase.
"Magnesium can be precipitated in a lime-soda softener by increasing the
hydroxide alkalinity of the effluent to the point at which magnesium
hydroxide is relatively insoluble. This can be accomplished with hydrated
lime or caustic soda:
If magnesium is precipitated by lime, there is no appreciable reduction in

hardness unless additional soda is used to precipitate the calcium that
replaces the magnesium. It is economical in some locations to use caustic
soda alone instead of lime and soda for precipitating magnesium. This
eliminated lime-slurry feeding problems and particle size difficulties.
Theoretically, magnesium can be precipitated only as described in the
foregoing. At several installations, however, where high-calcium hydrated
lime alone was being used for calcium bicarbonate precipitation, it has been
found that about 60 percent of the magnesium was being precipitated, even
though only a few milligrams per litre of hydroxide alkalinity were present.
This small amount was insufficient to precipitate the magnesium as the hy-
drate, but certain unusual equilibrium conditions produced the unexpected
result. Total reduction of hardness was within normal limits in these instances,
but the amount of calcium precipitated was less than normal. The operations
were carried out in the cold in sludge-blanket units.
Most continuous-process installations for lime-soda or caustic soda
softening utilize accumulated sludge in suspension to promote growth of
larger precipitation particles. This arrangement carries the chemical reaction
and absorption processes to practical completion in a short time. The unit
for performing this operation is known as a sludge-blanket precipitator. In

older plants, settled sludge was mixed with the incoming water to produce
the same effect. However, efficiencies were not as high as with the new units,
and the older method has been more or less abandoned.
Temperature plays a very important part in chemical reaction. Where low
hardness is desired and hot water can be used, as in boiler feed, temperatures
from 104 to 146°C (220 to 295° F) are employed. A hardness of about 70
mg/litre is the best that can be produced at temperatures of 10 to 27°C (50»to
80° F), but this can be reduced to about-25 to 30 mg/litre at 104°C (220° F)
and to'10 mg/litre at 146°C (295°F).
Zero hardness is desired for boiler feed makeup, and it is evident that the
lime or lime-soda method will not provide this. Phosphate softening
following hot lime-soda treatment is often used as a second softening stage;
it is used for single-stage softening when the initial hardness is 30 mg/litre or
less [8]. The theory involved is the precipitation of the calcium ion as the
triphosphate, an insoluble material:
Any of the phosphates may be used; they are added in aqueous solution.
Where the calcium hardness is mainly bicarbonate, phosphoric acid provides
alkalinity control as well as lower total solids in the effluent. The reaction is
carried out at a pH of 10 or higher with maintenance of a few mg/litre of
hydroxide residual. The resultant hardness in the treated water is about 1
mg/litre as CaCOx Caustic soda is used for precipitation of magnesium ion
when phosphates are used to precipitate calcium.
In the past, especially in European practice, barium carbonate was used
instead of lime for softening. This method has not found favor in America,
where it is used in only a few specific instances.
Internal Boiler-Water Softening—Usually it is unsatisfactory to depend
upon external softening alone for boiler makeup water, especially for high-
pressure boilers. Internal boiler water treatment may be required to convert
scale-forming salts to soft sludges to be removed by boiler blowdown.
Chemical compounds whose solubilities increase with a rise in temperature
can be made to form nonadherent sludges [9]; those whose solubilities
decrease with increase in temperature often form hard, adherent scale.
Internal boiler-water treatment in actual practice is far more complex than
these general principles indicate. This results from the wide variety of
compounds and mixtures of compounds that can be formed in the boiler,
and from the effects of organic matter and rates of evaporation on crystal
structure [10-13].
It is not desirable to precipitate large quantities of minerals in a boiler, and
internal boiler-water treatment preferably should be applied as a supplement
to external chemical or physical treatment. Internal treatment alone may be
satisfactory for low-pressure boilers, for which only moderate amounts of
reasonably soft feedwater are required.
60 MANUAL ON WATER
Bicarbonates and carbonates introduced with the feedwater or with

internal treatment decompose in the boiler to produce carbon dioxide in the
steam and to form hydroxides in the concentrated boiler water:
The hydroxides produced in this manner can precipitate the magnesium in

the boiler water as magnesium hydroxide:
During the decomposition of the bicarbonates and carbonates, some of the

calcium of the boiler water is precipitated as calcium carbonate:
The rate of decomposition of the bicarbonates and carbonates may be so

rapid at boiler pressures above (34.5 kPa) (5 psig) that the calcium hardness
is entirely precipitated and the hydroxide concentration becomes excessive.
One method of solving this difficulty is by using phosphates [/¥]. Another
method of solving this difficulty is by using chelating agents such as the
sodium salts of ethylenediaminetetraacetic acid (EDTA) and nitrilotriacetic
acid (NTA). These agents form highly soluble complexes with metallic ions,
thereby preventing scale formation on heating surfaces.
Iron, Manganese, and Chromium Removal—Iron and manganese can be
removed by precipitation, using lime or caustic soda to adjust the pH to the
proper range. Groundwaters often contain iron and manganese that must be
removed. These waters usually have very little color or turbidity except that
due to those minerals. The iron is generally bound with carbon dioxide as
the bicarbonate, and aeration is a necessary pretreatment to release the
carbon dioxide and precipitate some of the iron. The addition of alkali
converts the ferric iron to a hydroxide which settles out. For iron alone, a
pH in the range of 8.5 is sufficient, but when manganese is present a pH
above 9 is required. A coagulant is generally required to aid subsidence of
the precipitate. Chlorine is often used to accelerate the rate of oxidation.
Retention basins are necessary for separation of the precipitate. These may
be followed by sand filters to provide a satisfactory final effluent.
Waste waters from plating operations may contain chromium, usually in
the hexavalent form. Chromium in this form is toxic to aquatic and human
life; it must be removed prior to final discharge.
The chromium is first converted to the trivalent form by lowering the pH
to 2 with acid and adding a reducing agent such as ferrous sulfate, sodium
bisulfite, or sulfur dioxide. The chemical reaction is as follows:
The trivalent chromium is then precipitated as the hydroxide with lime

and pH adjustment to some value between 7.8 and 8.5. Residual chromium
can be reduced to very low limits by this method.
Emulsion Breaking or Demulsification

Organic oil emulsions in waste waters present a disposal problem because
the emulsion will contain considerable organic matter with a high BOD, and
breaking the emulsion will release free oil which may not be discharged to
surface streams. In many cases these emulsions can be broken by acidifica-
tion below pH 4, but the resultant acid water is extremely corrosive and must
be handled in acid-resisting structures. pH correction is required after
sedimentation and before discharge to the receiving stream.
Demulsification with calcium chloride and adjustment to pH 9 with soda
ash is a method that has gained acceptance. The calcium chloride dosage is
high, about 2000 mg/litre, but exceptional breaking and precipitation of the
emulsion have been obtained with a resultant high-percentage recovery of
oil. The chemistry of this procedure is not fully understood, but full-scale
installations have proved the feasibility of the process.
Ion Exchange
It may be necessary to remove the cations or anions, or both, from water
to provide a satisfactory supply for process and boiler feed uses. A preceding
section described removal of cations by precipitation; this process had no
effect on the anions present, and other means are required to remove them.
Certain insoluble substances possess a capacity for exchanging ions bound
in their molecular structure with other ions in water. The exchanged ions are
released by regeneration of the exchange resin. Negatively or positively
charged ions may be exchanged, depending upon the nature of the resin. A
wide variety of solid materials possess this reversible property. A charac-
teristic difference between ion exchange and precipitation processes is that
the former produces only solutions for waste disposal while the latter yields
both liquid and solids. This important difference may be the deciding factor
in a choice of process. Disposal of sludge from a precipitation process may
make it uneconomical as compared with ion exchange. On the other hand,
waste water from an ion exchange process may present a disposal problem.
Natural and synthetic substances (zeolites) have been used for many years
to remove calcium and magnesium ions from water. These ions are then
replaced by sodium ions when the zeolite is regenerated:
Softening:
62 MANUAL ON WATER
Regeneration:
The process will operate equally well if the calcium and magnesium ions
are replaced with potassium ions or with any other positive ions that can
exist in a relatively concentrated neutral solution.
In the early years of the industry, zeolites were developed from natural
greensand. Suitable treatment opened up the pores in the grains and
hardened the surface. Exchange capacity, which was dependent upon pore
area, was relatively low, about 6.4 kg/m\ Capacities of 0.292 kg were
developed by either acid or heat treatment. The natural zeolites were not
resistant to high- or low-pH water, and pretreatment of the raw water was
often necessary because they were easily fouled by iron or organic matter.
Synthetic gel zeolites were later produced and had higher capacities, but
their use was rather restricted and they have gradually been abandoned in
favor of the newer synthetic resins.
With advances in the art of ion exchange, carbonaceous and synthetic
resin zeolites were developed that would withstand both acid and alkaline
solutions and provide both cation and anion removal. Carbonaceous zeolites
are limited to about 18.3 kg/m 3 while the resins have a maximum capacity of
about 1.9 kg on the cation cycle.
In hydrogen cycle operation, the material exchanges hydrogen ions for
calcium, magnesium, sodium, and other positive ions; it is regenerated with
sulfuric or hydrochloric acid:
The effluent from hydrogen ion exchangers contains free mineral acidity,
and it may be desirable to operate such installations in parallel with sodium
cycle exchangers or to mix the effluent with untreated water to neutralize the
acidity. The combination of sodium and hydrogen cycle ion exchange makes
it possible to remove the hardness and reduce alkalinity at the same time.
The bicarbonate hardness is converted into dissolved carbon dioxide, which
can be removed by degasification.
Where the anions must be removed from the water, the process usually is
operated in two stages. The water first flows through a hydrogen-regen-
erated cation unit, and the acid effluent is then passed through an acid-
absorber anion exchanger.
In this two-step process, the acid effluent is usually degassed to remove
free carbon dioxide before passing through the acid-absorbent exchanger to
complete the deionization. The degassing step can be eliminated if a strongly
basic, acid-absorbent exchange medium is used. If the hydrogen exchange of
the first step is complete, the product from the final unit will be neutral and
low in ionized solids. If the effluent from the hydrogen cycle contains
positive ions, these ions with their negative components will pass through the
acid-absorbent units and appear in the final effluent. It is therefore essential

to have good operation of the hydrogen cycle in this process. Depending
upon the solids in water, the two-step units will produce an effluent with a
conductivity of about 10 to 50 ^mho. Two or more stages, or a mixed-bed
unit, will be necessary if higher purity is required.
Weakly ionized materials in water passed unaffected through the older
type of ion exchangers. Thus, silica was not removed by the original two-step
hydrogen exchanger and weakly basic acid absorber. If the silica is converted
into the strongly ionized fluosilicic acid by addition of fluorine compounds
at some point in the process, the fluosilicic acid can be removed by the weakly
basic acid absorber. The development of strongly basic acid absorbers has
solved economically the problem of ion exchange removal of silica without
the use of fluorine compounds.
In general, silica removal has been carried on along with anion removal
using highly basic anion resins. Where softening and silica removal are
required but not complete demineralization, the use of strongly basic anion
exchangers following normal sodium-regenerated cation exchangers has
served the purpose economically. The capacity of anion resins for silica
removal has been found to be a function of the caustic soda dosage used for
regeneration. Silica reductions to 1 mg/litre or less from 35 mg/litre have
been obtained by this method [75].
Initial cost, process, and control problems of two-step deionization are
now being reduced by the use of mixed-bed deionizers. The hydrogen
exchange and acid-absorbing materials are mixed intimately in a single bed
in these units. A mixed-bed unit performs like multiple stages of the two-step
process and can deliver water with extremely low ion content. The break-
though of ions at the end of.the operating cycle can be made extremely sharp
with mixed-bed units. During backwashing, the two media separate by
differential density, with the hydrogen exchanger below the acid absorber.
The acid absorber can be regenerated with caustic soda and rinsed through
the spent hydrogen exchange material. The hydrogen exchanger is then
regenerated by introducing acid through a distributor located just below the
anion absorber medium. Another method of regeneration is to introduce
simultaneously the caustic soda at the top of the anion resin and acid at the
bottom of the cation resin bed, removing the spent caustic and acid together
through a common collector located at the interface of the beds.
Ion exchange must complete with distillation for the production of high-
purity industrial water. The choice between the two processes depends
largely upon the total concentration of ionized and nonionized substances in
the water supply, and whether sterile water is required from the process.
Other forms of mixed-bed or multiple-bed ion exchangers are of com-
merial importance. One involves simultaneous sodium and chloride ex-
change, in which both kinds of ions in the regenerating solution serve a
purpose. The sodium ions regenerate the cation exchange material, libera-
ting calcium and magnesium ions, while the chloride ions regenerate the
anion exchange material, liberating bicarbonate ions. These units will
64 MANUAL ON WATER
replace hardness with sodium ions and bicarbonates with chloride ions. The
two ion exchange materials need not be completely mixed; the unit can be
operated with the material in the layers that develop during backwashing.
This method is feasible, but it is not widely used. Two stages are employed
more often, the first to remove calcium and magnesium, utilizing sodium
chloride for regeneration. In the second stage, the anion material absorbs the
bicarbonate ion and replaces it with chloride ion. Here a mixture of sodium
chloride and caustic soda is used for regeneration. Although the salt
performs the regeneration, the caustic has been found to lengthen the life of
the resin, which results in higher capacity [7(5].
Ion Exchange in Waste Treatment—Ion exchange is being used to recover
chromium from waste waters. The chromic acid waste is first passed through
an anion exchanger to remove the chromium in accordance with the
equation:
The actual removal is believed to occur in two or more successive

reactions:
The anion material is regenerated with caustic soda, which releases the
chromium as sodium dichromate:
The products of regeneration are then passed through a cation exchanger

to convert the sodium dichromate back to chromic acid:
This treatment recovers the chromic acid and concentrates it for reuse.
The chromic acid content of plating waste waters varies from 100 to 500
mg/litre CrOs. The recovered chromic acid has a concentration of about
30 000 mg/litre or 3 percent CrOa, which is sufficiently high for reuse. The
waste water has been rid of dilute chromic acid and it can be reused or
discharged to a stream.
Ion exchange has also been used for recovery of certain ionized organic
compounds. These include nicotine from water used to wash the gases from
cigarette drying, and tartaric acid from winery wastes. Nonionized fruit
sugars have been recovered from cannery wastes and citrus peels, and pectins
from grapefruit wastes, by ion exchange. These are special applications but
show the wide range of applicability of this process to the treatment of waste
waters for product recovery.
Ion Exclusion—A technique, known as ion exclusion, has been evolved
which has many possibilities in waste-water treatment and certain process
operations. It allows separation of ionic and nonionic materials, such as
sodium chloride from ethyl alcohol, sodium chloride from formaldehyde,
and sodium sulfate from glycerin. Thus, separations that were never
economically possible are now quite feasible.
The method provides for alternate feeding of the aqueous solution of ionic
and nonionic materials and water into an exchange column. One volume of
feed solution is usually followed by one or more volumes of water. An
aqueous solution of ionic material comes out of the column first, followed by
an aqueous solution of the nonionic material. The separation of the ionic
material from the nonionic material may be quite sharp.
Separations of this kind are possible because of the peculiar behavior of
an ion-exchange bead in an aqueous solution of a mixture of an electrolyte
and a non-electrolyte. In such solutions, the electrolyte tends to be more
concentrated around the bead, while the nonelectrolyte_tends to have the
same concentration within and outside the bead. This difference in con-
centration between the two solutes is the basis for the process. The ionic
material appears to be rapidly flushed through the void spaces between the
individual ion beads, while the nonionic material is retained in the column
for a longer period since it diffuses through the resin beads as well as around
them.
Resins used in ion exclusion do not need to be regenerated because they
never become exhausted; the only operating cost is for water and pumping
[17].
Ion Retardation—Another ion-exchange process for separating water-
soluble materials is ion retardation. This has resulted from development of a
resin that has both anionic and cationic exchange sites located in the same
bead. The resin will therefore adsorb both anions and cations from the feed
solution. The resin sites are so closely associated, however, that they
partially neutralize the electrical charges in each other; the adsorbed ions are
thus weakly held and may be displaced from the resin by water alone.
Ion retardation is a column operation in which alternate portions of feed
and rinse water are passed through the column, as in ion exclusion. It
permits separation of two electrolytes, such as ferrous sulfate from zinc
sulfate, sodium hydroxide from sodium chloride, and ammonium chloride
from zinc chloride. It will also separate electrolytes from nonelectrolytes,
and salts from very large nonelectrolyte molecules.
Like ion exclusion, this technique holds promise of adding another tool to
the fractionation of process streams and treatment of waste waters.
Sequestering
The foregoing subsections have been devoted to water-treating processes
66 MANUAL ON WATER
in which chemical constituents are removed as precipitates or by ion

exchange. A third chemical method is that of sequestering the impurity into
soluble complexes. In the sequestering process, certain normally positive
ions, such as calcium and magnesium, are so strongly held in a complex
negative ion that the equilibrium concentration of the free metal ion is
extremely low. The concentration of free metal ions can be depressed by this
process to the point where the insoluble soaps will not form. Water so
treated may have a soap hardness of zero, and it can even redissolve
precipitated calcium soap.
Nearly 100 years were required between Thomas Graham's first studies of
sodium metaphosphate and the general recognition that a soluble negative
complex ion of calcium does exist. The discovery [18, 19] of the soluble
phosphate-calcium complex was made with sodium hexametaphosphate.
Many of the poly-phosphates display this property in various degrees. Other
chemicals have been found in recent years which are able to sequester
hardness in true soluble complexes. Polysilicates are available that form
soluble complexes with iron [20].
Relatively large concentrations of treating chemicals usually are required
to sequester effectively even moderate concentrations of undesirable con-
stituents. Thus the sequestering process finds its most economic application
as a supplementary or final treatment after the bulk of the undesirable
constituents have been removed by a process having a lower unit cost.
Oxidation
Surface- and groundwaters may contain odor- and taste-producing sub-
stances that make them unfit for drinking and for some process uses. These
substances are usually organic, and often they can be removed by oxidation.
Spray ponds provide one method for accomplishing this. Water is
dispersed in the air as fine droplets by spraying it over a pond through
nozzles. The intimate contact between water and air provides the desired
oxidation. Usually any dissolved gases are released at the same time. A
forced-draft degassing tower may be used in place of a spray pond; it should
provide about the same efficiency.
Oxidation with chlorine, chlorine dioxide, or ozone can be used to convert
objectionable organic matter into innocuous compounds. A short contact
period, usually not over 10 min, is required for complete oxidation. Certain
waste waters may contain taste-producing materials, such as phenol, in small
volume but high concentration. Biological treatment under these conditions
would entail a very large capital expenditure, but the operating cost would
be low. The capital investment can be reduced materially by using chlorine
or chlorine dioxide, because the operation would be the equivalent of a
waste treatment unit. The waste-water flow would be the water supply to the
chlorinator and it would be discharged as a treated effluent. The operating
cost for such an arrangement might be rather high, but the low capital
investment would make it attractive in many instances. The effluent will

contain excess chlorine, and some retention is necessary before release to a
stream.
Normal biological methods are usually prohibitive where waste waters
contain more than about 3 percent of soluble organic matter, because of the
large dilution necessary for effective purification. A method known as wet
combustion has been developed which provides efficient air oxidation
through operation at elevated pressure and temperature [27]. The waste
water is heated to not less than 350° C and the pressure raised to 13 800 kPa
(2000 psi). Air is blown through the liquid under these conditions, and
effective oxidation of the organic material takes place, with purification
efficiency well above 90 percent. Power requirements for this process are
very high, but with effective power and heat recovery, the economics of the
operation are satisfactory.
A catalytic oxidation method has been developed for organic wastes plus
water that are produced in the vapor phase and that can be prevented from
condensing prior to treatment [22]. The waste feed is raised to 300° C and,
together with air in the ratio of 300 to 700 percent of the requirement for
oxidation, also heated to 300° C, is fed into a converter containing copper or
manganese chromite catalyst maintained at about 600° C. The reaction is
exothermic, and means must be provided to remove enough heat to hold the
catalyst at 600° C. Conversion efficiencies of 90 percent and above can be
achieved by this method, with organic concentrations as high as 8 percent;
the method is uneconomical at concentrations below 1 percent.
Iron in water above 0.3 mg/litre and manganese over 0.2 mg/litre are
generally objectionable. Both can be removed from most waters by lime-
soda or caustic soda precipitation process. They also can be removed by
oxidation, precipitation, and filtration because of the comparative in-
solubility of hydrated ferric and manganic oxides. Aerated water is passed
through contact filters in which the filter medium acts as a catalyst for the
oxidation reaction. Manganese is more dependent on catalytic contact than
is iron in this process. Silica sand is often a suitable contact medium for iron
removal, but complete manganese removal usually requires pyrolusite
(manganic oxide) or special manganese zeolite. These manganese zeolite
units are not cation exchangers but are contact oxidizers that are re-
generated by treating zeolite with manganous sulfate and potassium per-
manganate. The higher oxides of manganese are deposited on the zeolite
granules during regeneration. Contact with this material oxidizes ferrous
and manganous ions to insoluble hydrated oxides, which are then removed
by filtration. Such units are employed most economically as a final
treatment to remove the last trace of iron and manganese from effluents that
already have been treated by a less expensive, but less efficient, process.
The sodium-cycle ion-exchange process will remove ferrous and man-
ganous salts simultaneously with hardness if these metals are present as
bivalent ions.
68 MANUAL ON WATER
Neutralization
Neutralization of acidity or basicity may be necessary to provide satis-
factory process water or for treatment of waste water. Many groundwaters
contain such high concentrations of carbon dioxide that the water is quite
acidic. Aeration will remove most of the carbon dioxide, but not all of it, and
the pH may still be so low that the water is corrosive for general purposes.
Accordingly, neutralization is provided to raise the pH to the desired value.
Calcite filters, which consist of pressure vessels filled with pulverized
marble, can be used if the quantity of water is small. The carbonic acid is
neutralized by the calcium carbonate and the water is stabilized at the same
time. Stabilization means that the alkalinity-pH ratio is automatically
adjusted to provide a water neither corrosive nor scale-forming.
Alkali feeds are used where large volumes of water are needed. The
solution feeder comprises equipment for continuously feeding a water
solution of caustic soda, soda ash, or lime slurry. Provision is made for
controlling the rate of chemical addition to maintain the desired pH value.
The control may be manual or automatic, through a pH recorder-controller.
Neutralization is not instantaneous, and a contact period must be provided,
usually not more than 5 min. The alkali solution may be fed by gravity or by
pumping into the stream to be treated. Liquid storage tanks are usually
designed for 12- to 24-h operation between refills.
Chemicals may also be fed dry. Only soda ash or lime may be used in dry
feeders because caustic soda is too hygroscopic to permit uniform feeding.
The dry-feed machines operate either on the volumetric principle, in which a
measured amount is displaced each stroke, or gravimetric, whereby the flow
of dry material is weighed continuously. The dry chemical discharges into a
mixing box where it forms a solution or a suspension. It is then allowed to
flow by gravity, or is pumped, to the point of application. These machines
can be made to operate in proportion to flow rate or to pH response.
Hoppers over the machines provide chemical storage of 8 to 24 h.
Where lime or lime-and-soda methods of softening are employed, the
effluent will contain caustic alkalinity. This excess alkalinity generally must
be neutralized. This is done by adding acid, usually sulfuric or carbonic. The
amount of acid fed is determined by the resultant pH value desired. The
usual control is by a pH recorder-controller.
A 10 percent acid solution is normally used to treat relatively small
volumes of water. Concentrated acid is fed if the volume is large. A metering
pump is preferred for acid feed in large-volume operations. The pump can be
controlled for ratio-flow control, pH control, or a combination of both.
Acid-resistant storage tanks, piping, and pumps must be used.
Both alkaline and acidic waters are encountered in waste treatment. Free
mineral acids and free alkalies must be neutralized before discharging such
waters. This means neutralization to the methyl orange or the pheno-
Iphthalein end point, respectively. Rarely is it necessary to go above pH 4.6
or below pH 8.3. The neutralized waste water will not reduce the total
alkalinity in the receiving stream. It is very difficult to maintain pH values

above 5 since slight increments in alkali feed cause wide swings in pH as
neutrality is approached.
Where acid and alkali waste waters are discharged in the same plant,
blending these may be all that is necessary for satisfactory neutralization.
Any deficiency of one waste can be made up by a separate chemical feed, and
the final blended effluent can be discharged at the desired pH. The blending
should be carried out in an acidproof tank providing sufficient time for full
reaction. Neutralization studies in the laboratory will establish the period of
retention required.
Consideration must be given to the type as well as to the concentration of
acids in waste waters that must be neutralized. Such acids as hydrochloric or
nitric can be neutralized with lime without trouble since the salts produced
are soluble in high concentrations. With sulfuric acid, there is a very definite
limit if high-calcium lime is to be used; the acid concentration must be held
below 0.3 percent if calcium sulfate deposits are to be avoided. Concentra-
tions up to 1 percent can be handled successfully if dolomitic lime is used. If
other acids are present, the resultant depression of calcium sulfate solubility
must be taken into account. An example of this is a blend of neutralized
nitric and sulfuric acids in which the calcium sulfate solubility was reduced
to 1100 mg/litre from about 2000 mg/litre.
Acids can be neutralized by upward flow through granular limestone
beds [23]. The flow rate is usually 3340 to 5350 litres/min/m2 (25 to 40
gal/min/ft 2 ) of bed, and beds 0.9 to 1.5 m (3 to 5 ft) in depth are placed in
acid-resistant tanks. These beds are quite effective because of the turbulence
created as the water is forced to flow upward. Sulfation must be avoided
where sulfuric acid is neutralized, or the limestone particles *will cement
together.
Gravity flow is necessary for handling large water volumes if high
pumping costs are to be avoided. Usually, both rate of flow and acid
concentration will vary widely. The multistage neutralizer [24] can be used to
deal with this difficulty. The neutralizer consists of several chambers in
series, up to six in number, each provided with a mechanical mixer. The
waste water flows from the top of one chamber to the bottom of the next
one, thereby providing ample reaction time without short-circuiting. Lime
slurry is fed at alternate chambers; the feed rate is governed by pH recorder-
controllers. The first chambers of the multineutralizer function as equalizing
units and the alternate ones as reaction vessels. Before designing such a
system, detailed reaction-rate studies must be made to establish total
retention time, reaction time, and rate dosage.
Silo storage of burned lime is usually provided together with gravimetric
feeders and slakers and effective slurry storage. The slurry, usually 10
percent, is circulated past the feed valves and back to the slurry storage by
centrifugal pumps. Provision must be made for holding constant back
pressure on the feed valves. Excellent effluent control can be achieved with
70 MANUAL ON WATER
an installation of this kind for treating waste flows as high as 37 850 to

56 775 m 3 /day (10 to 15 Mgd) with lime requirements of 108 t (120 tons) per
day.
Degassing
Dissolved Carbon Dioxide—Dissolved carbon dioxide can be chemically
removed by adding lime, soda ash, or caustic soda to form soluble
bicarbonates or relatively insoluble carbonates. This method requires careful
chemical control. Approximately the same result can be obtained with less
control and supervision by passing the water through a bed of material with
which the dissolved carbon dioxide will react chemically to form a neutral or
alkaline product. The most common material for this process is crystalline or
granular calcium carbonate, such as crushed calcite or limestone. The
calcium carbonate slowly dissolves by reaction with the carbon dioxide to
form additional bicarbonate hardness in the water:
Such a unit also acts as a filter, and periodic backwashing may be required.
The initial investment is the principal cost of the process; chemical replace-
ment and operating costs are small. The additional hardness added to the
water by the process may be objectionable [25].
Chemical Removal of Dissolved Oxygen—To supplement physical dea-
eration, dissolved oxygen also can be removed from water by chemical
means. Sodium sulfite is used extensively as a final treatment to consume the
small trace of dissolved oxygen which is not removed by deaerating
feedwater heaters. Sulfite reacts rapidly with dissolved oxygen at boiler
water temperature to form sulfate:
It is easy to determine sulfite in boiler water, and the concentration and

consumption of sulfite can be used as a convenient chemical integrator for
checking the performance of the physical deaerating equipment. Sulfite at
boiler water temperature may reduce certain chemicals, however, such as
iron and copper salts, and a material balance may be slightly inaccurate.
Ferrous iron salts and hydrazine have been used as oxygen scavengers.
Colloidal iron, tannin, and other chemicals are also used for their oxygen-
consuming properties.
Contact Reduction of Dissolved Oxygen—Steel wool contact filters have
been used in the past for dissolved oxygen removal through oxidation of the
metal, the products of corrosion largely remaining in the filter. Steel wool
filters in some waters liberate ferrous salts, which consume additional
dissolved oxygen beyond the filter. The slow rate of oxygen consumption by
steel wool contact filters has limited their commercial use. Contact reduction
of dissolved oxygen can be brought about by a reversible oxidation

reduction process using an amine exchange resin. Inasmuch as this material
can be regenerated by chemicals, it is similar to ion exchange.
Dechlorination and Removal of Gases by Chlorine—Sulfite, sulfur di-
oxide, and thiosulfate or bisulfite can combine with dissolved chlorine in a
dechlorination process. The reactions may be represented by the equations:
Chemical methods usually are employed only for removing low chlorine
residuals. Chlorine gas, on the other hand, may be used to convert ammonia
or ammonium ions to chloramine, which can be removed from solution by
aeration.
Sterilization and Disinfection

In general, all waters, whether surface or ground, contain bacteria, the
type and number depending upon existing conditions. Groundwaters taken
from sand strata usually are free of pathogens but may contain a fairly high
count of soil organisms not harmful to man. Water taken from rock wells
may or may not contain pathogenic organisms.
Surface streams are prone to bacterial contamination from both soil and
man, the amount of the latter increasing greatly with sewage discharges.
Other types of bacteria can be present in large numbers where industrial
pollution is present.
Soil bacteria must be more or less completely removed from water for
municipal supply, and pathogenic organisms must be absent. In industrial
waters, small numbers of organisms can usually be tolerated, depending
upon process use. It is thus apparent that these organisms must be
eliminated if satisfactory potable water supplies are to be provided.
Where small volumes require sterilization, as in the laboratory, heating to
boiling temperature is sufficient. When continuous treatment is necessary, as
in a water purification plant, bactericidal agents are employed. The most
common of these 4s chlorine, either as a gas or as hypochlorite.
Feeding gaseous chlorine requires specially constructed equipment and,
while the gas may be diffused into the water to be treated, the general
practice is to feed an aqueous solution of it. This results in full utilization of
the gas and better control. Gas chlorination is generally employed when 6.75
kg (15 Ib) or more is required per day, because there is no saving in the cost
of the feeding equipment below this quantity.
Bleaching powder, high-test calcium hypochlorite powder, or sodium
hypochlorite solution are used where only small amounts of chlorine are
needed. These solutions are fed by gravity or pumped into the system by a
72 MANUAL ON WATER
diaphragm pump. The available chlorine in these compounds is much more

expensive than gaseous chlorine, but the low equipment cost and simplicity
of operation more than offset the difference.
The amount required to provide a residual for sterilization is in general
dependent upon the amount of organic matter in the water, because this
must be oxidized before there can be a residual. Most groundwater is low in
organic matter, and the effluent from a filter plant is usually very low. Feed
of 0.5 to 1 mg/litre under such conditions will provide a residual of 0.2
mg/litre as combined chlorine. Feed rates of 10 to 40 mg/litre may be
required if the organic matter is high.
Chloramines are used where there is a large distribution system or a
reservoir open to sunlight, because chlorine alone may be dissipated too
quickly to ensure sterile water. Chloramines are formed by adding ammonia
in a 1+4 ratio of ammonia to chlorine. The Chloramines are not as active as
free chlorine, and a minimum contact time of 15 min is required. Chlora-
mines are very persistent, however, and they provide more satisfactory
sterilization in long distribution systems.
It has been found that a combined-chlorine residual will not always
destroy all of the microorganisms in the system and that tastes and odors
sometimes occur in sequence. This problem is combated by maintaining a
free-chlorine residual. This is done by feeding sufficient chlorine to oxidize
all the organic matter, thereby providing a residual of free chlorine. This
practice is now followed at most waterworks.
Chlorine dioxide is a product that is finding favor in eliminating tastes and
odors, for it is also an excellent sterilizing agent. This is produced by reacting
sodium chlorite solution with a chlorine solution at pH 3.0. The resultant
solution contains both chlorine and chlorine dioxide. Its use has been
successful in eliminating tastes and od9rs in many cases when other methods
have failed, and the residual appears to persist well.
Accessory equipment is required for making a solution of the chlorite and
for feeding it. Furthermore, a reaction chamber is necessary for mixing the
two solutions. The chlorine dioxide and chlorine combination is more
expensive in terms of actual available chlorine, but with a difficult taste and
odor problem it provides the only way to eliminate it.
Sterilization of water is also effected with ozone generated by passing an
electrical discharge through dry air. It must be generated at the site in
expensive equipment with a high electrical demand. The gas and air mixture
is bubbled through the water as in a normal aeration system. The cost of
installation and operation is considerably higher than for chlorinators,
except in large waterworks. In addition, measurement of residual ozone is
nearly impossible; as a result, it has been used in very few locations.
Bromine possesses the same sterilizing properties as chlorine but it is a
more difficult material to feed and control. So far, it has been used
experimentally in some swimming pool installations.
Ultraviolet light can kill bacteria in clear water. The light must be
generated at the point of use, and the equipment is complex. A few
experimental units have been installed in the past, but the method has not
been generally accepted.
Silver iodide has excellent disinfection properties when fed as a soluble
salt. Its high cost precludes its use except during war emergencies. It was
made in tablet form during World War II and was extensively used by the
armed forces.
Control of Biological Growths
Various biological growths in water may require chemical treatment,

either to inhibit their growth, to kill them by sterilization, or to stimulate
them with hydroponics. The objectionable organic growths in industrial
water systems may include pathogenic and nonpathogenic bacteria, algae,
slimes, iron and manganese bacteria, sulfur bacteria, and a host of other
microscopic and macroscopic organisms. Some organic growths are danger-
ous to health, others cause bad tastes and odors, and still others cause
clogging of equipment and piping.
Chlorination processes are effective for control of most biological growths
in water [26]. Chromates and salts of copper, zinc, and silver also are used
for this purpose. Organic compounds such as phenol, chlorinated phenol
derivatives, naphthenates, quaternaries, and many others under a wide
variety of trade-names are available for biological growth inhibition. Copper
sulfate, with or without aliphatics such as citrates to keep the copper in
dilute neutral solution, is used for the economical control of algae in large
bodies of water such as reservoirs and lakes. Treatment with considerably
less than 1 mg/litre of copper sulfate is usually satisfactory, depending upon
the type of algae [27] and the care that must be taken not to kill fish. The
amount of the treatment need only be sufficient for a limited top water layer
and not for all of the water in the reservoir. For large quantities of water
used only once, the economic choice for biological growth control is usually
limited to copper sulfate, chlorination, and ozonation.
In smaller bodies where the water is reused, as in circulating cooling water
basins, treatment with other chemicals may be economical. Intermittent
chlorination may be economically satisfactory for control of biological
fouling in cooling water systems. Residual free chlorine in the order of 0.5 to
2 mg/litre is satisfactory in most cases for control of slimes and the spat or
larval stage of larger organisms. Intermittent treatment with bromine also is
used to a limited extent, especially for the control of such macroscopic
organisms as mussels, which are resistant to chlorine.
Incineration
Incineration is not usually considered as a waste water disposal method,
but there are certain cases where it is applicable. When the volume of waste
water is under 40 litres/min (10 gal/min) and the soluble organic concentra-
74 MANUAL ON WATER
tion is above 4 percent, the waste can be sprayed into the boiler furnace for
evaporation of water and combustion of organics. The boiler furnace must
be large enough to handle the vapor load. Some additional fuel will be
required, and the fuel-feed and burners must be of sufficient size to handle
this increase. This "method can generally be used only with oil-fired boilers. If
the waste water contains toxic materials, a modified unit provided with its
own burner and fuel supply can be used effectively for destroying the
compounds.
A great deal of study is necessary, followed by pilot operation, before
incineration can be used effectively, but the technique does provide a
satisfactory answer for certain waste water problems.
Physicochemical Processes
Several important water treatment processes depend upon combined
chemical and physical action. These are coagulation, adsorption and
absorption, crystal growth deactivators, additives for changing surface
tension, and corrosion inhibition.
Coagulation
Most surface waters contain suspended matter and color. Much of the
suspended matter consists of particles too small to settle rapidly, even in
quiet water. Color is caused by absorption of light by suspended and
dissolved organic and mineral substances. A chemical coagulant, such as
alum, ferrous or ferric sulfate, or sodium aluminate, will cause the finely
divided and colloidally suspended materials to be gathered together into
larger particles, called floe, which will settle rapidly. A flocculertt precipitate
which has enormous surface area per unit of volume is produced by
coagulants. This precipitate removes suspended and colloidal matter from
the water by electro-physical attraction, adsorption, absorption, and phys-
ical entanglement. For example, aluminum sulfate reacts with hydroxyl ions
in water to form complex aluminum hydroxides that are relatively insoluble
in the pH range from 6 to 8. The particles of precipitated aluminum
hydroxide, with the help of suitable mild agitation, agglomerate into visible
clumps which settle by gravity. The simultaneous formation, agglomeration,
adsorption, and inclusion of the suspended matter and color into the floe is
the physicochemical process of coagulation.
Each coagulant has a pH range in which it is most effective. The addition
of the coagulant usually changes the pH of the solution at the time the floe is
formed. Acidic coagulants such as aluminum and ferrous and ferric sulfates
precipitate as basic compounds and reduce the pH of the water. When the
natural alkalinity of the water is not sufficient to maintain the minimum pH
requirement after coagulation, it is necessary to add lime or soda. The
optimum range for coagulation with aluminum sulfate is between pH 5.5
and 7.0. Above 7.0, there is a tendency for some of the alum to remain
unprecipitated, and a small amount of residual aluminum will remain in the
final filtered water. When the pH of the raw water gets above 7.2, the
amount of coagulant necessary to drop the pH into the optimum band may
be too great economically, especially if the coagulation pH is 6.3 to 6.4 for
the water. Accordingly, sulfuric acid is added ahead of or with the aluminum
sulfate to reduce the pH to the proper range for good coagulation. Sulfuric
acid will lower the pH much faster and at less expense than the coagulant.
Ferric sulfate is an effective coagulant over the wide pH range of 5.5 to 10.
It has an acidic reaction and lowers the alkalinity of the water. Ferrous
sulfate, while effective over a wide range, is generally used between pH 8.5 to
10. In addition to requiring alkalinity for coagulation, there must be an
oxidizing agent to convert the ferrous iron to the ferric state. Usually there is
enough dissolved oxygen in the water to complete the oxidation; if
insufficient, some form of aerating device is employed to increase the oxygen
content. Chlorine also is frequently added for this purpose.
Many factors affect the amount of coagulant needed for optimum results.
Among these are type and amount of suspended matter, water temperature,
rate of mixing, and retention time. Coagulant dosage may vary from 5
mg/litre for a good water very low in turbidity and color to 100 mg/litre or
more for a highly turbid one. Presettling is usually provided when turbidities
continually run over 400 to 500 mg/litre, to reduce the coagulant require-
ment.
Continuous and proportionate feeding of the coagulating chemical is
necessary for effective coagulation. In addition, control of the pH in the
optimum range is essential. The coagulant can be fed dry or in solution,
depending upon the amount to be fed and specific conditions. Usually
quantities up to 45 kg (100 Ib) per day are fed as a solution. Above this, dry-
feed machines are employed with a dissolving tank so that a dilute solution is
fed; with proper physical arrangement the dry material can be fed directly
into the water.
The coagulant should first be thoroughly dispersed in the stream to be
treated. This is accomplished in a high-speed mixing chamber having a
retention of from 0.5 to 1.0 min. After the dispersion, the finely formed floes
require agglomeration by very gentle agitation so they can develop into large
masses. Slowly revolving paddles in a chamber providing from 40 to 60 min
retention will bring the floe to maximum development and produce one that
will settle rapidly and provide a satisfactory supernatant for filtration.
Adsorption and Absorption

The processes of adsorption and absorption of organic and colloidal
materials into flocculent precipitates with a resultant increase in sedi-
mentation rate are used to remove color, taste, and odors from water. High
silica concentrations also can be economically decreased by ferric and
76 MANUAL ON WATER
magnesium floes. These processes are more efficient in terms of chemical

consumption when the flocculent precipitate develops in the water than
when a preformed precipitate is added.
Adsorption on ferric hydroxide can reduce the silica content to 2 or 3
mg/litre. The initial reduction occurs rapidly, but suitable mixing and
retention time are required to attain minimum concentrations. The process
operates best in the cold. Magnesium hydroxide also will remove silica and
often is used in the lime-soda precipitation process for silica adsorption and
precipitation. Magnesium can be added to the water as magnesium car-
bonate, magnesium bicarbonate, or magnesium oxide. Magnesium hydrox-
ide requires a longer contact time than ferric hydroxide for maximum silica
reduction. It differs from the ferric hydroxide process in that warming the
water improves silica reduction. The pH for magnesium hydroxide forma-
tion must be above 10.2, which is the usual condition with lime-soda
softening. On the other hand, the ferric hydroxide process is usually carried
out separately because the pH for its most economical silica reduction is
about 9.0.
Taste, odor, and some dissolved gases often can be removed from water by
adsorption on activated carbon. This material can be added as a powder to
the filter influent and removed from the filter bed during backwashing.
Activated carbon filters of granular or powdered activated carbon are
available for filtered water service. Activated carbon may be partially
regenerated by steam or other gases. Activated carbon filters are also used
for dechlorinating water.
Crystal Growth Deactivators

The surface deactivation properties 6*f numerous chemicals are employed
in a variety of treatment processes for retarding or inhibiting crystal growth.
The polyphosphates have this property in various degrees. Tannin, lignin,
and many other organic materials have surface deactivation properties.
Crystal surfaces are believed to be covered with molecular or submicroscopic
films of the surface deactivator, thereby reducing, altering, or inhibiting
further crystal growth.
Threshold treatment is an example in which calcium carbonate deposits
are inhibited, although concentrations above the normal solubility of the
scale-forming material are present. Calcium phosphate precipitation in
feedwater equipment can be inhibited in the same manner. This technique is
also used to retard precipitation of ferric iron in water containing dissolved
oxygen; thus red-water troubles frequently can be reduced by crystal
deactivation.
Surface deactivation is used as an aid in keeping trace elements in solution
for hydroponics, in limiting crystal size, and for many other special uses. An
ever-increasing number of surface active chemicals, under a wide variety of
trade names, are available to the water conditioning engineer.
Additives for Surface Tension Changes

Numerous chemicals have the property of significantly changing the
natural surface tension of the water. Chemicals which markedly decrease
surface tension are called wetting agents, and those which materially increase
surface tension are often spoken of as foam suppressors. Wetting agents in a
wide variety of commercial forms are available for many industrial uses,
such as ore-separation processes, laundry operation, fire fighting, acid
cleaning, and a host of others. Numerous chemicals that increase the surface
tension of water also are commercially available for reduction of foaming
and priming difficulties in steam generators.
Corrosion Inhibition
Water conditioning for corrosion control involves both physical and
chemical processes operating together. Protective films that inhibit metal
deterioration are physical barriers. The formation and maintenance of such
films may be the result of chemical reactions. These physicochemical
corrosion inhibition processes are briefly described in the following three
subsections.
Cathodic Protection—The electrolytic theory of iron corrosion, as postu-
lated by Whitney [28] and demonstrated by Walker and associates [29], is
essential for understanding the process of cathodic protection [30]. Ac-
cording to this theory, metals in contact with water tend to go into solution,
forming positively charged ions. Positive ions can go into solution in an
electrically neutral system only if an equivalent number of positive ions of
some other element are plated out. In the case of pure water in contact with a
metal, such as iron, hydrogen ions are plated out at the interface. The
reaction may be written:
Metal Ions Ion Atoms
The liberated hydrogen forms a thin, insulating film on the metal which
prevents further corrosion. Thereafter the corrosion reaction proceeds at a
rate that is a function of the rate of removal of the layer of atomic hydrogen.
The observed effects of dissolved oxygen, temperature, pH, salt concentra-
tion, and turbulence on this protective hydrogen film on iron and steel have
been found to agree with this theory [31]. Any method for maintaining the
protective hydrogen film will reduce corrosion.
Galvanizing steel provides a form of cathodic protection. Here the
protective hydrogen film is maintained on the exposed steel cathode at the
expense of the zinc anode, which slowly goes into solution. Magnesium and
zinc rods or plates suspended in water and electrically connected to the steel
tank are further examples of cathodic protection. In other forms of this
78 MANUAL ON WATER
process, direct current is supplied from an outside source to maintain the

protective hydrogen film on the steel cathode, and a corrosion-resistant
anode is used. A sacrificial anode, such as an abandoned pipeline, also can
be used. Various types of commercial equipment are available to provide
cathodic protection. Cathodic protection often is used to inhibit corrosion in
cold-water tanks and pipelines. By a variation of this process, hulls of steel
ships can be protected from electrolytic corrosion caused by copper-alloy
propellers.
Physicochemical Corrosion Barriers—Any type of continuous film that is
chemically inactive in water and is tenaciously attached to the metal will
inhibit aqueous corrosion. Painting and molten-metal plating are examples
of physicochemical corrosion barriers. A corrosion-inhibiting film also can
be developed from the metal itself by chemical action. Metal oxide finishes
and chromate, silicate, phosphate, and tannate films are examples of
chemically produced corrosion barriers. The art of chemically producing
and maintaining uniform corrosion-resistant films on the various engineer-
ing materials is a complex field.
A well-maintained steel boiler might be defined as a clean, thin, con-
tinuous film of water-insoluble material, reinforced by steel, and used to
generate vapors. The development and maintenance of this thin continuous
film, usually consisting of metal reaction products, depend on many factors.
Raising the pH of the water in contact with steel to between 10.5 and 11.7 to
develop a more uniform protective film is common practice for steam
boilers. Other conditions such as the removal of dissolved oxygen and
carbon dioxide by physical or chemical deaeration will improve the uni-
formity and tenacity of the protective film. Volatile nitrogen compounds
(neutralizing amines) are used to neutralize carbonic acid [52, 33]; filming
amines are used to provide a monomolecular protective film [34]. Venting of
such gases as carbon dioxide [35] from steam condensate systems results in
an increased pH and improves the protective film. In open cooling-water
systems, chromates [36] and silicates [37] are widely used for the develop-
ment of thin, uniform protective films on steel. Organic materials such as
tannin and lignin compounds are also used for this purpose.
The polyphosphates [38] under controlled conditions can form protective
films on steel and cast iron. Polyphosphates, organics, and chromates-are
often used together for cooling-tower corrosion inhibition. Where serious
deposits of calcium carbonate scale will not develop, it is sometimes
desirable to increase the pH in cold, hard-water lines [39] by a controlled
lime treatment to reduce corrosion. All the aforementioned processes are
used for the purpose of developing and maintaining a thin, uniform,
tenacious film to keep corrosion of pipe and equipment at minimum levels.
Inhibition of Biological Corrosion—The chemistry of biological corrosion
is still in its infancy. It is believed that dilute sulfuric and nitric acids can be
produced on the surface of metals by biological reactions in polluted waters.
Elimination of the acid-producing organisms should prevent biochemical
corrosion of ferrous and copper-alloy tubing. These organisms can be

controlled by the same general processes used in physical and chemical
regulation of biological growths.
Liquid-Phase Separation
With the steadily increasing demand for water, considerable thought has
been given to the purification of seawater, and it is here that liquid-phase
separation is being widely investigated.
The development of ion-selective membranes has led to their use in this
work. Membrane electrolysis employs a stack of alternating cationic and
ionic selective membranes which provides a series of parallel cells through
which the raw water flows. An electric potential across the stack causes the
cations and anions to move in opposite directions and to pass through the
appropriate selective membranes. As a result, ions tend to leave alternate
cells and to concentrate in the cells between them. Thus the stack con-
tinuously discharges two streams, one relatively demineralized and the other
proportionately concentrated in the abstracted ions. The process shows
promise as an economical method for desalting brackish water.
Osmoionic demineralization is another method being studied for purifying
salt water with ion-selective membranes. No electricity is required, because
the method uses osmotic pressure to move ions through the membrane. The
procedure is effective for desalting brines with chloride concentrations in the
range of 3000 mg/ litre.
Dialysis and electrodialysis have been used in the treatment of certain
waste waters. These methods separate ionizable salts from colloidal sus-
pensions. The ionized salt passes through a membrane and leaves the
colloidal material behind. Depending upon the material to be separated and
the conditions at hand, electricity may be used as an additional driving force.
These methods are being used for alkali and acid separation and recovery
from process liquors.
Biological Treatment
Interest is awakening in the time-honored process of percolation of raw
wastes through a bed of graded media.
Sand percolation is one of the oldest recognized methods of water
purification. The process had the ability to remove color, turbidity, and
bacteria by means of biological activity within a gelatinous growth on the
media grains. The biologically active material provided the major puri-
fication; the graded media, or sand, served as a support for the bio-mass
(known as schmutzdecke). Good purification resulted at application rates of
2300 to 2800 litres of filter per square metre per day (2.5 to 3.0 million
gallons per acre of filter per day). Intermittent dosing or trickle-spray fluid
application of the raw waste may be used to increase the throughput of
80 MANUAL ON WATER
percolation units, but higher rates tend to favor more advanced processes.
Influent pretreatment may also provide minor improvement in application
rates.
Biological purification is widely used for treatment of waste water
containing dissolved organic matter. Bacteria break down the complex
compounds into simple, stable ones; the usual end products are carbon
dioxide, water, nitrate, and sulfate. This change is brought about through
the metabolism and cell synthesis of the microorganisms present. Usually
these processes are carried out in the presence of excess dissolved oxygen,
and the operation is then known as aerobic decomposition. Another group
of microorganisms can thrive in an environment devoid of dissolved oxygen.
Under these conditions the process is known as anaerobic decomposition.
Although most biological treatment today is aerobic, there are some wastes
which respond better to anaerobic decomposition.
There are two broad types of bio-oxidation equipment, trickling filters
and activated sludge. The trickling filter is a bed of graded rock, usually 0.9
to 1.8 m (3 to 6 ft) deep, but sometimes as deep as 3 m (10 ft). The bed
diameter may vary from 4.5 to 61.0 m (15 to 200 ft). The rock is supported by
an underdrain system for collection of the liquid flowing over it. A rotary
distributor on the top provides for uniform spreading of the waste water on
the bed. The underdrain system has air ducts which allow air to circulate
through the bed by natural draft; this provides the oxygen necessary for
good biological action. The stone provides a large surface for growth of the
organisms that accomplish aerobic decomposition. The units are classed
high or low rate depending upon the rate of operation. Up to 4700
litres/m 2 /day (5 million gallons per acre per day) (Mgad) is considered low
rate, while 9400 to 28 200 Htres/m2/day (10 to 30 Mgad) is high rate.
Biochemical oxygen demand is the design criterion. The volume of the filter
is based upon decomposition of 0.34 to 1.72 kg/m 3 (1 to 5 Ib of BOD/yd3) of
stone per day. The diameter is fixed by the rate of application and volume of
waste water to be handled per day. Actual design should be developed only
after pilot plant studies of the waste water in question. Efficiencies of single
filter units range from 35 to 85 percent removal of BOD. Where higher
efficiencies must be provided, a second unit is installed to operate in series
with the first to provide two-stage treatment. Under such conditions, 80 to
95 percent purification can usually be obtained.
In the activated sludge process the waste water to be treated is held for a
period of time in a tank in which air is blown through the liquid. This
develops a lush bacterial suspension which brings about aerobic decom-
position. The retention time will usually range between 4 and 24 h. BOD in
kilograms per day is the design criterion for these units and the loading will
vary between 0.24 to 2.4 kg/m3 (15 and 150 Ib/1000 ft3) of aeration tank
volume. The air required is supplied by rotary compressors or centrifugal
blowers. The difuser system may be perforated pipe, nozzles, carborundum
plates, or tubes. Turbomixers of various design characteristics are also used.
Purification efficiencies will range from 75 to 95 percent depending upon the

waste water and the concentration of bacteria in the aeration tank water.
Higher efficiencies can generally be obtained with higher bacterial con-
centrations, but this must be balanced against an increased air requirement.
The concentration of bacterial slime (reported as suspended solids) will
range from 2000 to 5000 mg/litre.
Biological purification produces sludge, whether by trickling filters or by
activated sludge. This must be removed from the effluent before final
disposal. Settling basins providing from 1 '/$ to 5 h retention permit the solids
to settle. Scrapers move the settled material to a central point for removal.
The sludge produced by trickling filters is removed completely from the
system. In the activated sludge process, a portion is returned and mixed with
the raw waste water entering the aeration tank. The excess over that
necessary to maintain a satisfactory concentration of suspended solids in the
aeration tank is removed from the system.
The wasted sludge must be disposed of in some manner. This can be a
problem because of its fairly large volume, inability to dry readily on a sand
bed, and evolution of offensive odors. Bacterial decomposition of the wet
sludge under anaerobic conditions will yield a sludge that dries well and has
little odor; this process is widely used for sludge treatment. Closed concrete
tanks providing from 25 to 60 days retention and heated to about 32° C
(90° F) hold the raw sludge and allow bacterial decomposition to occur. The
volatile solids are broken down from about 75 to 80 percent to about 50
percent and, in so doing, carbon dioxide, methane, and nitrogen are given off
as gases. If a sufficient volume of gas is evolved, it can be burned for heating
the tanks and also used for operating gas engines in the plant.
There have been many adaptations of the activated sludge process to serve
particular purposes. Among the best known of these are the Guggenheim,
Kraus, contact areation, dispersed-growth aeration, and biosorption meth-
ods. These have all been described in various texts and they will not be
discussed here.
Where the effluent from the treatment plant must discharge into a very
small stream, the amount of organic material remaining in the treated water
may be too large for inoffensive assimilation; in such cases, additional
purification becomes necessary. Here use is made of the oldest type of
biological unit, the slow sand filter. This is simply a sand bed which is dosed
intermittently with the effluent from the treatment plant. Slime growths
develop on the sand and additional purification and straining take place.
The effluent flowing from the underdrains is clear and has a very low BOD.
The rate of flow through these filters is extremely low and extensive land
areas are required, but they can do a most effective job.
Modifications of the slow sand filter are the percolation bed and spray
irrigation. These use large areas of land where the soil is sandy and little or
no clay is present. Percolation beds make use of the sandy soil above the
ground water table and require 1.5 to 2.5 m (5 to 8 ft) of depth for good
82 MANUAL ON WATER
operation. Low dikes are built and means provided for dosing the areas once
a day Dosing is normally equivalent to flooding to a depth of about 5 cm (2
in).
The waste is usually diluted with water to provide a BOD of 300 to 500
mg/litre and applied to give a BOD loading of 0.11 to 0.22 kg/m 2 /day (100
to 200 Ib per acre per day). Under such conditions, purification efficiencies
of 90 to 98 percent have been obtained. In the sand, biological growths
develop which utilize the soluble organic material in the waste water as food
and thereby provide purification. Air is drawn into the soil during the dosing
period; this provides the oxygen needed by the organisms.
Spray irrigation utilizes the same general principle. Rotating sprays or
fixed sprays spread the diluted waste over a large area, which can be crop
land or wooded areas. The spray period usually lasts from 8 to 12 h per 24 h.
Sandy soil is a requisite, and purification is effected by biological growths in
the soil. The sprayed waste water should be low in suspended solids to avoid
plugging of the soil. While this method has been developed primarily for
disposal of cannery and milk wastes, it is applicable to almost any easily
oxidized, nontoxic waste.
A modification of the activated sludge process is the oxidation pond. In
this adaptation, the waste water is held in a shallow pond for a period of
days. Aeration from the atmosphere takes place and bacterial growths
develop. These ponds can remove 95 percent of the BOD and kill 99 percent
of the coliform bacteria; thus purification is generally sufficient to allow
discharge of the effluent into streams. These stabilization ponds have been
quite widely used in the paper industry and they are gaining acceptance in
the textile and chemical industries. They are used as a low-cost method for
municipal sewage treatment.
Summary
An attempt has been made to enumerate the principal water treatment
processes serving industry at the present time. New processes will un-
doubtedly be developed in the future. The discussion has been limited to the
main technical and economic features. Literature references have been given
for readers desiring greater detail.
Reference
[7] Ingersol, C, Petroleum Refiner, PEREA, June 1951, p. 1064.
[2] "Waste Water Containing Oil," Manual on Disposal of Refinery Wastes, American
Petroleum Institute., Section I, 4th ed., 1949.
[3] Beebe, A. H., Sewage and Industrial Wastes, SIWAA, Vol. 25, 1953, p. 1314.
\4\ McDermet, J. R., Transactions, American Society of Mechanical Engineers, TASMA,
Vol. 42, 1920, p. 267.
[5] Amorosi, A. M. and McDermet, J. R., Proceedings, American Society for Testing and
Materials, ASTEA, Vol. 39, 1939, p. 1204.
[6] McKinney, D. S., Proceedings, American Society for Testing and Materials, ASTEA, Vol.
39, 1939, p. 1191.
[7] Hoak, R. D., Water & Sewage Works, WSWOA, Vol. 100, 1953, p. 468.
[8] Feller, E. W., Power, POWEA, Vol. 91, Dec. 1947, p. 71.
[9] Hall, R. E. et al, Industrial and Engineering Chemistry, IECHA, Vol. 17, 1925, p. 283;
Ibid, Vol.21, 1929.jp.JJ26; Techinical Bulletin, Carnegie Institute, Vol. 24, 1927.
[JO] Barkley, J. F., "Questions and Answers on Boiler Feedwater Conditioning," Bureau of
Mines, U. S. Government Printing Office, Washington, D. C, 1943.
[11} "Central Boiler Plants," Technical Manual TM5-650, Chapter 4, War Department, U. S.
Government Printing Office, Washington, D. C., 1947.
[12] "Boiler Feedwater and Feedwater Apparatus," Bureau of Ships Manual, BSJA, Chapter
56, Navy Department, U. S. Government Printing Office, 1948, p. 56.
[13] McKinney, D. S. in Chemistry of Engineering Materials, McGraw-Hill, New York, 1942,
pp. 1-32.
[14] Hall, R. E. and Jackson, H. A., U. S. Patent No. 1,903,041,1933.
[15] Applebaum, S. B. and Dickerson, B. W., "Silica Removal by Salt Splitting Without
Demineralization," 16th Annual Meeting, American Power Conference, 24-26 March
1954.
[16] Applebaum, S. B., Êxperiences with Chloride Anion Exchangers for Reducing Alka-
linity Without Acid," Proceedings, 13th Annual Water Conference, Engineers' Society of
Western Pennsylvania, 20 Oct. 1952.
[17] Wheaton, R. M. and Bauman, W. C., Industrial and Engineering Chemistry, IECHA,
Vol. 45, 1953, p. 228.
[18] Hall, R. E., U. S. Patent No. 1,956,515, 1934, reissue, 19,719 1935; U. S. Patent No.
2,087,089, 1937.
[19] Industrial and Engineering Chemistry, IECHA, Vol. 29, 1937, p. 584; Vol. 31, 1939, p. 51,
Vol. 32, 1940, p. 1572.
[20] Henry, C. R., Journal, American Water Works Association, JAWWA, Vol. 42, Sept.
1950.
[21] Zimmerman, F. J., Chemical Engineering, CHEEA, 25 Aug. 1958, p. 117.
[22] Green, R. V. and Moses, D. V., Sewage and Industrial Wastes, SIWAA, 1952, p. 288.
[23] Gehm, H. W., Sewage Works Journal, SWJUA, Vol. 16, 1944, p. 104.
[24] Dickerson, B. W. and Brooks, R. M., Industrial and Engineering Chemistry, IECHA,
Vol. 42, 1950, p. 599.
[25] Ryan, W. J., Water Treatment and Purification, McGraw-Hill, New York, 1937, p. 221.
[26] Dobson, J. G., Transactions, American Society Mechanical Engineers, TASMA, April
1946.
[27] Hale, F. E., "The Use of Copper Sulfate in Control of Microscopic Organisms," Phelps
Dodge Refining Corporation, 1939.
[28] Whitney, W. R., Journal, American Chemical Society, JACSA, Vol. 25, 1903, p. 395.
[29] Walker, W. H., Cederholm, A. M., and Bent, L. N., Journal, American Chemical
Society, JACSA, Vol. 29, 1907, p. 1251.
[30] Pearson, J. M., "Fundamentals of Cathodic Protection," Corrosion Handbook, Wiley,
New York, 1948, p. 923.
[31] Speller, F. N., Corrosion, Causes and Prevention, 3rd ed., McGraw-Hill, New York,
1951.
[32] Speller, F. N., Proceedings, PDHAA, National District Heating Association, Vol. 24.
1933, p. 203.
[33] Berk, A. A., U. S. Bureau of Mines, Report of Investigation No. 3754, 1944.
[34] Kahler, H. L. and Brown, J. K., Proceedings, 10th Annual Water Conference, Engineers'
Society of Western Pennsylvania, 1949, pp. 115^126.
[35] Guernsey, E. W., Transactions, American Society of Heating and Ventilating Engineers,
TAHVA, Vol. 51, 1945.
[36] Darrin, Marc, Industrial and Engineering Chemistry, IECHA, Vol. 37, 1945, p. 741.
[37] Lehrman, L. and Shuldener, H. L., Journal, American Water Works Assocation,
JAWWA, Vol. 43, March, 1951.
[38] Hatch, G. B. and Rice, O., Industrial and Engineering Chemistry, IECHA, Vol. 37, 1945.
[39] Powell, S. T., Industrial and Engineering Chemistry, IECHA, Vol. 26, 1934.
Chapter 6—Technology of Industrial

Water Reuse
The growing water shortage in many parts of the world has made water
reuse a popular topic of concern. However, reuse actually is an ancient
practice. Since there is a fixed amount of water on the earth, it is clear from
the law of conservation of matter that the same water must be used over and
over again to fulfill the many needs. The perpetual cycling of water from
atmosphere to earth to ocean and back again to the atmosphere of course is
not uniform throughout the world. In some places, water evaporated from
the oceans is returned rather quickly by mountain streams with little or no
use. In other cases, precipitation may percolate into groundwater systems,
from which much of it may not emerge again for thousands of years. But
sooner or later the natural laws which control the hydrologic cycle guarantee
that all water will return to a point of use. Reuse is the inevitable result of
this process.
The current emphasis on reuse of water results from the fact that water
users and water uses are increasing throughout the world. Industrialization
is responsible for the greatest increase in water requirements. Since the
amount of water is fixed, it is obvious that the available supply must be
handled more often between the time it precipitates from the atmosphere
and the time it evaporates from the ocean. Each successive user along its
path automatically becomes a reuser, and the more numerous the reusers,
the more important becomes the subject of technology of water reuse.
Significance of Terms "Use" and "Reuse"
Considerable hysteria has been created by writers who imply that all water
drawn from a resource for a given purpose is lost to the needs of other
prospective water users. A thoughtful person will recognize immediately that
the terms "water withdrawal," "water use," and "water consumption" are not
synonymous. An industry may draw large quantities of water from a surface
stream or a well, use it only once in a closed cooling circuit, and pass it on to
the next user downstream. Another industry may draw very little water from
the source, but use it over and over again with relatively little actual water
84
CHAPTER 6 ON TECHNOLOGY OF INDUSTRIAL WATER REUSE 85
loss. Occasionally an industry will consume virtually all of the water it

withdraws. The carbon black industry is an example, because the process of
making carbon black results in evaporation of most of the water back to the
atmosphere. Despite this variation, the average water consumption (loss to
the atmosphere, etc.) for industries as a whole, including steam-electric
power plants, is only about two percent of the water withdrawn from the
original source. In other words, 98 percent of all industrial water is simply
passed through the plant and becomes reuse water for the next operation
downstream.
The requirements of individual industries vary markedly, as one might
expect. Major users, such as the chemical and metal industries, withdraw 25
to 30 times as much as minor industries, like the leather products industry.
Within a given industry, one factory located in a particular part of the
country may use 10 to 25 times as much water per unit of product as similar
plants located elsewhere. New factories are likely to be more efficient water
users than old ones, because of increasing competition for available supplies.
Large installations tend to use somewhat less water per product unit than
small operations, and factories in arid regions are far more water conscious
than similar ones in humid areas. This pattern is evidence that industries are
reasonably flexible in tailoring their water requirements to fit the local
situations. It is safe to say that almost every individual factory can reduce its
primary water needs by improving its water-use philosophies, introducing
conservation practices, and increasing reuse techniques. Where cooling is the
principal use of water, and this generally is the case, substitution of a recycle
system for a once-through system can reduce intake requirements by as
much as 90 percent. This, and other reuse practices, could easily counteract
the twofold increase in industrial water intake predicted for the year 1980.
Today, industry as a whole uses five times more water than it draws from
the original water source. In other words, it practices 500 percent reuse.
Tripling the amount of reuse would make it possible to double our industrial
production without increasing water intake. Of course, the quantity of water
lost through evaporation and wastes would need to remain the same. In view
of this possibility, it is not surprising that the number of reuse systems is
increasing. A 1959 Census of Manufacturers, published early in 1961 and
summarized in the Journal of Industrial and Engineering Chemistry later that
same year [7],1 showed that overall water reuse in industries other than
steam-electric generation increased by about 45 percent in the five-year
period between 1954 and 1959. An increase in reuse of 25 percent was
observed among industries which had not reported reuse during earlier
surveys. While this census was restricted to industries withdrawing 75 000 m3
(20 million gallons) or more of water per year, it included the major users, who
account for 90 percent or more of the total quantity of water withdrawn
(excluding steam-electric generation). This remarkable increase in efficiency
'The italic numbers in brackets refer to the list of references appended to this chapter.
86 MANUAL ON WATER
of water use resulted from the variety of return flow and on-site reuse
systems discussed in the following text.
Reuse of Returned Flow

About 98 percent of the 5.3 X 109 m 3 /day (140 billion gallons) of water
withdrawn daily by U. S. industry is returned to the sources from which it
came. Roughly 80 percent of this is used as cooling water and thus is
contaminated only with heat and concentrated dissolved solids by evapora-
tion. The remaining 20 percent carries a great variety of contaminants in an
equally great variety of concentrations. The nature of the industrial complex
and the receiving stream play a major role in determining the significance of
the quality of this return. If the receiving stream is sluggish or if the ratio of
withdrawal to streamflow is great, the chemical, physical, and biological
processes which normally restore a water to acceptable quality may be
relatively inefficient, and downstream users may receive water of relatively
poorer quality. As an extreme example, the quantity of water withdrawn
from the Verdegris River in Kansas during the low-flow period is approxi-
mately 17 times the natural flow.
Regardless of the volume of flow in the source stream, concentration of
industries along its banks can create quality problems by steadily increasing
the ratio of wastes to available dilution water and reducing the distance
and recovery time between users. A survey in 1959 showed that there were
254 industrial installations of 10 different types on the Delaware River and
estuary. Such a complex yields a return flow with sundry contaminants in
relatively low concentrations which must be removed from large volumes of
water before the supply will meet certain critical process needs of down-
stream users. The expected increase in industrial activity in the next several
decades will markedly increase the volume and variety of return-flow
contaminants. This could be a serious problem to the return-flow reusers
unless waste treatments keep pace with the increasing waste discharges.
Organizations such as the Ohio River Valley Water Sanitation Commission,
and numerous industries, already have made considerable progress toward
evaluating and eliminating discharge quality problems. Standardization
work by ASTM Committee D-19 on Water and similar groups is providing
the tools for improved evaluation.
One should not assume that it always will be possible or even desirable to
eliminate significant treated waste water discharge from all surface streams.
Groundwater
Short-term reuse systems are not nearly as applicable for groundwater
sources as for surface-water sources. However, underground water may be the
principal, it not the only, supply in certain areas with limited rainfall and few
surface streams. This is true in parts of the southwestern United States. In
such areas, pumping groundwater at rates exceeding recharge of the
groundwater reservoir (aquifer) results in declining water tables and land
subsidence. In coastal areas, declining water tables can result in serious

saltwater intrusion. These circumstances have increased interest in ground-
water recharge operations. Returning treated waste water to groundwater
aquifers also is attractive because the infiltration process further restores
water quality. In 1955, artificial groundwater recharge operations with
treated waste water came to about 190 000 m 3 /day (50 million gal/day).
Long Island, New York, has a requirement for returning all cooling waters
to the groundwater aquifers to maintain water level and prevent seawater
encroachment.
Groundwater recharge can be accomplished either through direct injection
wells, or by spreading water in ditches, ridge-and-furrow systems, or
recharge basins. Recharging through injection wells is three to seven times as
expensive as the surface operations, and involves the risk that water may
clog the aquifer or contaminate the supply because of unfavorable chemical
or geochemical reactions. Bypassing the corrective influence of the soil
mantle also is a deficiency of the well recharge method. In its present stage of
development, well injection is best suited to cooling water return where heat
is the only significant pollutant involved. Even here, differences in pH, gas
content, or oxidation-reduction potential might create significant problems,
particularly if unstable constituents, such as ferrous ion, are present in the
groundwater supply.
Returning water to the ground by spraying, ponding, and spreading is
considerably better insurance that reusable groundwater will result. Of
course these methods are applicable only to areas with favorable percolation
properties. In Wisconsin, ridge-and-furrow irrigation with cheese processing
wastes and spray disposal of paper mill wastes are carried on extensively
with little trouble from groundwater pollution. Spray disposal of food
processing wastes also is practiced extensively in New Jersey. Israel's
national water system provides for recharge with both flash flood water and
waste waters. A major portion of the country's total supply (both surface
water and groundwater) is combined in a flexible integrated system which
provides for maximum efficiency in use and reuse.
The years to come will see more efforts directed toward returning water to
groundwater reservoirs for the multiple purposes of reducing evaporation
loss, restoring quality, and preventing seawater encroachment and land
subsidence. It is important to recognize that this type of reuse is hindered by
the rather serious limitations of flow through porous media as compared
with the free flow of open channels. However, water technology must make
maximum practicable use of groundwater rechargd.
Cooperative Reuse Systems (Tandem Systems)

Sewage
A cooperative or tandem reuse system is a special case of return flow in
which one user passes waste water from his operations directly to another
88 MANUAL ON WATER
user instead of returning it to a general source of supply. The most common

waste water used in such an operation is domestic sewage effluent. This is
because sewage plant effluent often is readily available and nearly always is
of relatively qood quality. A classic example of a tandem reuse system
involving reuse of municipal sewage water is Bethlehem Steel Company's
reuse of Baltimore city sewage effluent. Bethlehem originally used well water
for critical cooling operations, which required quality better than that of
Baltimore harbor water. Eventually, increasing cooling demands and decline
in water table in the aquifer beneath Sparrows Point, Md., necessitated
search for another supply. Thorough study showed reuse of city sewage
effluent to be the most practicable solution to the problem. This system of
tandem reuse was started in 1942, and has increased from 9.46 to 47.3 X 104
m3/day (25 million to 125 million gal/day), in 1960.
Tandem use of sewage effluent has the advantage of reducing demand on
the primary supply and thereby extending its life. Supplying Bethlehem Steel
from Baltimore city's resources would scarcely be possible. Another ex-
cellent example of the same situation is Texaco's Amarillo refinery, where an
agreement was made with the city to operate the refinery on sewage effluent
rather than to compete with the city for the limited groundwater resource.
The cost is lower than for city water in this case.
Numerous industries use sewage effluent from their own housing develop-
ments, particularly where source water is in short supply. For example, the
Santa Rita Copper Company, in New Mexico uses sewage waste in
precipitation processes and then reuses it in leaching operations. In-
formation on numerous other installations has been published [2].
Use of Industrial Waste

Tandem reuse of industrial waste water also is practicable to some extent
and probably will be practiced increasingly because of the heavy concentra-
tion of industry in certain areas. Where limited water resources pose serious
restrictions on the amount of water which can be withdrawn, this plan often
is preferable to a plan involving competition among the industries for
primary withdrawal. Where industries are located close together and the
receiving stream is highly contaminated, water withdrawn, treated, and used
by an upstream plant sometimes is of better quality than water drawn
directly from the stream. If this is true, it is to the advantage of the
downstream user to draw the waste directly from the first user than to pump
from the original source. In the 1959 Census, the chemical industry reused
4.1 X 108 m3 (109 billion gallons) in this manner. English coal mines practice
tandem reuse when water-rich and water-poor mines are located close
together.
The quality requirements of the second user in a tandem reuse system
determines the treatment that must be applied to the waste-water effluent of
the first user. Before being used as Bethlehem Steel Company's cooling
water, Baltimore's sewage effluent must be subjected to supplementary

settling and chlorination, in addition to the normal combined trickling filter-
activated-sludge treatment.
On-Site Reuse Systems

The figures cited earlier for withdrawal and reuse show that there is
considerable in-plant recycle of water after it is taken from the water source
and before it is discharged as waste. This is accomplished with "on-site reuse
systems," which may be divided into several principal types—simple recycle,
multiple recycle, and cascade—but combinations of two or more generally
are used.
Simple Recycle System

The condensate return system of a steam electric generation plant is the most
common and familiar type of a simple on-site recycle system. Because of the
care which must be taken in treating boiler feedwater, economic considera-
tions require that the water be used to the maximum practicable extent once
it is in the boiler system. This is accomplished by returning the steam
condensate to the boiler feedline by way of the condenser, and by way of any
waste heat exchangers which may be included in the system to improve
economy. The returning condensate is partially degassed in the condenser
and frequently is further degassed in a pressure-closed deaerator to minimize
dissolved oxygen. It also may be subjected to other treatments, such as ion
exchange or special filtration processes, to ensure maximum purity before it
is returned to the boiler feedwater line. Were it not for inevitable losses
through steam leaks and the need to waste a small part of the boiler water to
carry away corrosion products and reaction products of internal boiler-
water treatments, this recycle would satisfy the entire water requirement.
Since these loses cannot be avoided, the returning condensate is fortified
with a small amount of treated makeup water from the main water source.
Condensate recycle in a well-kept boiler plant constitutes about the most
perfect reuse system available.
Multiple Recycle System

In a multiple recycle system, parallel circuits carry various grades of
process and cooling water of qualities suitable for the processes involved.
This plan has the advantage that the several waste streams are kept separate
for simplified, individual treatment and product reclamation. It also allows
for individual control of temperature, pressure, and flow rate. While the
circuits of a multiple cycle may operate with individual cleanup systems (ion
exchange, etc.), it is more common to maintain quality in each circuit by
continuously blowing a part of the circulating water to waste or to a less
90 MANUAL ON WATER
critical use in the same industrial plant. Where the waste water is passed to
another process or cooling circuit, the system becomes a compound
multicycle-cascade type. The Dow Chemical Company's Midland, Mich.,
plant has an elaborate multiple recycle system which handles high-pressure
service and cooling water, domestic water, high-grade cooling and process
water, and demineralized boiler feedwater, in five separate supplies in-
volving many miles of distribution lines. This complex system uses eight
times as much water as it draws from the original water source. Eventually
all of the circuits are integrated to form a semicascade system and the final
waste water is mixed and treated in a central location. In such complex
operations, it is common to reclaim chemicals as by-products from the
various circuits as well as from the final cascade. It is also common to mix
the discharges to gain advantages from reactions of components. Thus an
alkaline waste might be neutralized by mixing it with an acid waste. Final
waste that cannot be reclaimed economically is discharged into a biological
waste treatment plant or into disposal wells.
Cascade System
A cascade system differs from a multiple recycle system in that a single
effluent stream flows from one process to another, increasing its waste
content as it goes until it finally is no longer useful for any part of the
operation. The water withdrawn from the source of supply first flows to the
equipment or process which requires the coldest or cleanest water, and from
here it passes to successive operations where progressively higher tempera-
tures and poorer water qualities can be tolerated. Thus, the waste of one
operation becomes the supply of the next. Either single passes or recycle
circuits may form the various steps of the cascade, but considerable recycle is
characteristic of most cascade systems. Almost every industry employs
cascade to some extent. A cascade system does not reduce water loss or
decrease the waste which eventually must be treated or discarded. On the
other hand, it generally reduces the number of water-cooling and waste-
treatment facilities required, and, of course, increases the percentage of
reuse. Its principal shortcoming is the need for designing heat exchangers
and processing equipment to tolerate the relatively poor quality of water in
the later stages. It also is more difficult to reclaim specific by-products from
the integrated waste stream that results from a cascade system. The Kaiser
Steel plant at Fontana, Calif., is the classic example of this type of reuse
system. Here a mixed plant supply of well water and local company water
is used for stepwise cooling of motors and heating furnaces; mill cooling and
scale flushing; and cooling open hearths, blast furnaces, and coke plants.
After this, it is divided into parallel streams for further cascading. These
streams handle less critical operations like gas washing, precipitative flush-
ing, and blast furnace cooling. The streams finally are mingled and used for
quenching slag. The diagram of the circuit presented in Fig. 1 shows that
some water is diverted for use in the tin mill, and that some is reclaimed from
NUMBERS INDICATE COOLING

MAIN RESERVOIR TIRE PROTECTION
DOMESTIC
WATER FROM WELLS POWER PLANT
RESERVOIR
AND LOCAL COMPANY AND MISC.
INDUSTRIAL
RESERVOIR
ZA MOTOR DOMESTIC
GALVANIZING ROOM AND HEATING USES
FURNACES
SEWAGE EFFLUENT
ZB MILL COOLING 10 COOLING

ACID WASTE SYSTEM
AND SCALE FLUSH
1 OPEN HEARTH
BLAST FURNACE
AND COKE PLANT
8 BLAST FURNACE 12 BLAST FURNACE TIN MILL

COOLING COOLING PROCESS
5 GAS WASH PPT.

9 GAS WASH 11 SLAG
FLUSH AND
AND PPT. FLUSH QUENCHING
BLAST FURNACES
SKULL SLAG QUENCH

SLAG PITS
CRACKER SPRAYS
PIT RUNOFF
Kaiser's Fontana, Calif., system of multi-recycling cascade is so effective that the plant with-
draws only 5320 litres (1400 gal) of water per ton of steel produced, whereas as much as 247 000
litres (65 000 gal) are required in some plants. It can be said that the water is wrung dry in this
3000 percent reuse system.
FIG. I—Example of water reuse system.
the slag pit. Much of the final waste water is lost by evaporation from the
slag pit.
While a cascade system does not reduce the amount of water lost by
evaporation and may actually increase it to some extent, it does markedly
92 MANUAL ON WATER
increase the ratio of water used to the water withdrawn. In the case of
Kaiser's plant, each gallon of water is used 30 to 40 times, so that the plant
withdraws only 5320 litres (1400 gal) of water per ton of steel produced,
whereas as much as 247 000 litres (65 000 gal) of water per ton are withdrawn
in some plants. Such a cascade plant is more costly to build, but it has the
advantage of ability to operate in an arid zone where it might be impossible
to operate otherwise.
Examples of cascade reuse can be found in almost every industry. In
Holland power plant blowdown (boiler waste) is used for coal washing.
Stack gas wash water and mine waste are used for the same purpose. Any
number of stepwise combinations can be used to advantage where shortages
provide the incentive.
Process Integration Systems

In some cascade and recycle systems, production of particularly trouble-
some waste necessitates high dilution of final concentrated effluents or
extremely costly waste disposal systems. The black liquors produced by
sulfite and neutral sulfite semichemical (NSSC) paper mills are examples.
Until recently there was no profitable process for reclaiming raw materials
from such liquors, and the relatively high percentage of organic matter
contained in them caused rapid depletion of oxygen and consequent
pollution in receiving streams. Use of the waste in road-binding operations
and spray disposal on wastelands were the most practicable solutions to the
black-liquor problem. The latter occasionally caused serious iron contam-
ination of groundwater systems due to complexing actions of organic
components.
Occassionally, troublesome wastes can be handled by ingenious integra-
tions of industrial processes. Sonoco Products Co. of Hartsville, S. C, has
perfected such a system for handling black-liquors. Formic and acetic acid
are first reclaimed from the waste water, which is then combined with kraft
mill waste to generate raw materials useful in the kraft process. It is possible
to reclaim 98 percent of the NSSC black-liquor water in evaporators. A flow
sketch of this integration system was shown in Industrial and Engineering
Chemistry [3]. Numerous other possibilities for improving reuse through
process integration can be and have been exploited in controlling industrial
waste problems.
Quality Problems in Reuse
Heat Dissipation
The extent to which water can be used is determined in large part by the
effectiveness of quality control treatments in the reuse system. Since
approximately 80 percent of all industrial water is used for cooling,
elimination of heat contamination is a major consideration. In recycle

systems, cooling is accomplished with spray ponds, cooling towers, evapo-
rative condensers, and air-cooled heat exchangers. All but the last depend on
the cooling effect of evaporation. Approximately 1 percent of the circulating
water is evaporated for every 6 deg C temperature reduction, and the
loss is the same regardless of the type of system in which the evaporation is
taking place, provided the relatively small windage losses are disregarded.
Even when there is a considerable wind-loss factor, cooling tower water
consumption generally is no more than twice that of a once-through system,
in which the water flows directly to the source stream. The effectiveness of an
evaporative cooling system, like cooling of heated water in a river, is
determined by the wet-bulb temperature of the environment, since this is the
absolute lower limit to which the water can be cooled by evaporation. The
actual terminal temperature may range from a degree or two below
atmospheric temperature, at high humidity, to -1°C (30° F) or more below
atmospheric temperature when the air is very dry. Thus, evaporative coolers
are most effective in arid regions.
Forced-draft cooling towers with bottom fans and countercurrent air flow
are gaining favor over induced-draft (top-fan) and natural-draft types. All
types of cooling towers are more efficient than spray ponds for industrial
cooling applications because of the better contact between water and
circulating air.
Evaporative condensers combine cooling-tower and process-cooler func-
tions in a single operation. In a typical installation, cooling water is sprayed
on condenser tubes of a refrigerating system to condense refrigerant and
release heat to a circulating air stream, thus dissipating heat. The plan is best
suited to unit process cooling operations. A single installation in General
Electric's Electronics Park Site has 68 evaporative condensers.
Air-cooled heat exchangers waste no water by evaporation, but they can
cool only to within a few degrees of atmospheric temperature, and thus are
limited to relatively high-temperature applications. Compounding systems
to cool as far as possible with air and then to continue temperature reduction
in a cooling tower or evaporative system of another type often is an eco-
nomical way of handling cooling loads.
Organic Deposits
Open cooling systems are likely to develop algae and bacterial slime
growths. In the United States, organic growths generally are controlled with
chlorine or with nonoxidizing bactericides and algicides, whereas the British
chimney-like cooling towers depend on light-tightness for elimination of
some of the nuisance organisms. Antirot and fungal control treatments are
available for protection of construction materials and fill in wooden towers,
but slow deterioration is inevitable.
94 MANUAL ON WATER
Corrosion and Mineral Deposits

Corrosion also is a pertinent consideration in the metallic components of
recycle systems. This is handled by use of resistant components and by
inhibition with soluble oils, phosphate or chromate compounds, silicates,
zinc compounds, and dianodic mixtures as required. Organic and mineral
acids frequently are fed to maintain proper pH, and minimize deposit
formation. It is also feasible to control pH by diverting a part of the cooling
water through an acid-type cation-exchange resin instead of by direct feed of
acid. The so-called "starvation process" employed in Great Britain is an
interesting form of this treatment. Here a carboxylic acid type of exchange
resin is used which reacts only with carbonate alkalinity, thus making it
impossible to overacidify the circuit or to generate free mineral acid. The
process requires no more acid for regenerating exchangers than would be
required to treat the circuit directly.
Troublesome scale and sludge deposits will accumulate in any recycle
system if evaporation yields concentrations of water components above
saturation values, or if circulation rates are inadequate to keep undissolved
solids in suspension. Deposit formation generally is avoided by draining a
part of the circulating water to waste to maintain acceptable concentration
ranges and by replacing this waste with fresh makeup water. The percentage
of waste water can be calculated from a knowledge of the composition of
feed water and behavior of the wastes involved. Information for calculating
tolerable calcium sulfate contents is given in Industrial and Engineering
Chemistry \4\.
Safe disposal of concentrated blowdown wastes is a pertinent con-
sideration in every reuse system. A variety of organic and some inorganic
wastes can be separated by standard processes like coagulation, sedi-
mentation, demulsification, skimming, flue gas stripping, filtration, centrif-
ugation, and biological and chemical oxidation. Certain soluble inorganics
require more costly treatments, like evaporation, electrodialysis, and ion
exchange.
Profitable Aspects of Water Reuse

Wherever possible, an attempt should be made to recover by-products
from waste waters of reuse systems in order to defray a part of the cost of
operation. The extent of by-product recovery is limited only by the resource-
fulness of the technical staff. Paper mills reclaim large amounts of usable
fiber from wash waters; refineries salvage oils from their integrated
wastes; mines recover much coal from washers; platers obtain copper from
chemical baths and spent ion exchangers; steel mills dredge oxide from scale
pits in sizable quantities; and brewers recover from process waters vitamins
and antibiotics that otherwise would go to waste.
In some instances, by-product reclamation processes also generate high-
quality water for reuse. For example in the plating industry, ion exchange
resins simultaneously remove metal ions continuously from circulating wash

waters and generate freshwater for continued washing. The trapped metals
are valuable for recharging the plating bath, as mentioned in the foregoing.
The benefits derived from by-product recovery can be very sizable.
Whether or not such reclamation is economically practicable depends on the
size of the operation.
Reuse systems also can save money in initial water costs. Price advantage
in reuse will increase as increased industry results in greater contamination
of primary water supplies and primary treatment costs increase accordingly.
Depletion of local water supplies and need for long-distance transport also
will generate a price advantage for reuse systems.
One should not overlook the intangible profits of reuse, including
reduction in chemical, thermal, and biological pollution; flexibility for plant
expansion without providing larger water mains and sewers; relative sim-
plicity of treating and otherwise handling concentrated waste waters which
result from recycle; and maximum freedom from water pollution by up-
stream users.
Planning a Reuse System

The time will come when every existing industry must think seriously
about improving its water use facilities and this inevitably will mean
increased reuse. By the same token, new industries should give serious
consideration to the role of reuse in their proposed operations. In either case,
the following elements will be pertinent to developing a sound plan:
(a) Quantitative and qualitative water requirements of all process opera-
tions.
(b) Relative cost of alternative water supplies.
(c) Water use efficiencies of various plans and counterplans.
(d) Adaptability of various reuse systems to the operations involved.
(e) Overall economic advantage of alternative systems.
(/) Severity of ultimate waste disposal problems.
In any evaluation of this type, it is important to remember that water use
is a very dynamic operation because of our rapidly changing industrial
economy. Thus, flexibility should be an integral part of plant design and
continuous appraisal a recognized aspect of water use.
Future Outlook on Industrial Reuse of Water

Anyone with industrial experience will be quick to recognize that reuse of
water can involve some very troublesome features. Pages already have been
written on the effects of small quantities of detergents in return-flow
operations, and these troubles have been generated by minute quantities of
the contaminant in question. Certain components of process waters could
easily generate corrosion problems that require much more careful attention
96 MANUAL ON WATER
to water-handling equipment than now is required. Experience shows that

initial investment almost always is higher in a plant designed primarily for
high-percentage reuse.
One should also be careful not to overlook the very real problem of
disposing of concentrated wastes which all recycle systems must generate
sooner or later. While these wastes occur in relatively small volumes, their
disposal to streams in periodic slugs can produce serious local effects, which
are often more serious than continuous discharge of dilute waste from a
once-through system. Much of the success of reuse depends on discovering
ingenious methods of disposing of these wastes without their being returned
to the water courses. Processes like the slag quenching employed by Kaiser,
or deep-well injection employed by Dow Chemical Co., are steps in this
direction.
These disadvantages will not discourage water reuse. Increasing demands
for industrial water and expanding legislation for pollution control will
cause a steady increase in the cost of industrial water supplies and effluent
treatment. As the cost of the primary supply approaches the cost of recycle
and cleanup, reuse systems will become increasingly attractive and more
prevalent. To keep pace with this trend, more attention must be given to
methods for sampling and testing waters and waste waters. The work of
ASTM Committee D-19 is aimed in this direction.
References
[7] Clarke, F.. E., Industrial and Engineering Chemistry, IECHA, Vol. 54, Feb. 1962, pp.
18-25.
[2] Council, C. H., "Municipal Wastewater for Industrial Supply," presented at the Meeting
of the Process Industries Division of the American Society of Mechanical Engineers,
Houston, Tex., Oct. 4-6, 1961.
[3] Weaver, D. G., Industrial and Engineering Chemistry, IECHA, Vol. 53, Oct. 1961, pp.
773-778.
[4] Denman, W. L., Industrial and Engineering Chemistry, IECHA, Vol. 53, Oct. 1961, pp.
817-822.
Chapter 7—Self-Purification and

Other Natural-Quality
Recovery Mechanisms
Water is the most plentiful liquid on earth. It is a requirement for all life,
whether it is used in metabolic processes, as a solvent of minerals, or in the
disposal of waste. Indeed, it is difficult to name a natural phenomenon in
which water, in one form or another, has not had an essential role.
Nature has provided for the conservation of this abundant and valuable
liquid by endowing it with a considerable capacity to rid itself of foreign
substances. This process, known as self-purification, is an inherent system
for water. Man's only problem with the system is that the water does not
always clean itself quickly and completely enough for the next intended reuse.
Almost all of the processes used by man to clean the water are accelerated
natural-quality recovery mechanisms. The water may be moving in streams,or
be relatively static as in lakes or the oceans, but the natural processes of
purification are proceeding toward completion. Advantage is taken
of this intrinsic ability of water to purify itself by using it as the final
polishing treatment for all municipal and industrial waste discharges.
Streams contaminated by natural, human, or industrial wastes are
rendered stable by self-purification. Eventually, this water becomes innoc-
uous and is again reused by man or nature. This purification is accomplished
by combinations of the physical, chemical, and biological forces present in
the water environment. The extent to which each force operates is dependent
upon the contaminating substance and the particular water environment in
which it is present. Physical forces separate suspended solids; chemical forces
cause reactions which neutralize the nocuous wastes; and biological forces
stabilize the water by using the contaminating substances as nutrient sources
and by enchancing both physical and chemical phenomena. Generally, the
biological forces are the most active and the most important in the self-
purification of a stream.
The Dynamic Equilibrium

Water, as vapor, is present in the atmosphere in amounts that we express
as relative humidity. Condensation of this vapor causes the formation of
97
96 MANUAL ON WATER
droplets of water which are pure until they fall through a contaminating
atmosphere as rain, hail, or snow. The purity of the water is further affected
by contact with the soil or other surface.
The atmosphere usually contains a number of contaminants, especially in
areas of heavy air pollution. Such pollutants may be in the gaseous, liquid,
or solid state. Some solids, such as dusts, are laden with bacteria, yeasts, and
molds. In addition, the air contains the spores of microorganisms which
exist in the atmosphere without benefit of being attached to solid particles,
or entrapped in aerosoled liquids. Some bacteria and viruses, also occur in
the atmosphere as free-floating agents. Thus the atmosphere possesses a
distinct and variable microbiological flora.
The microflora of the atmosphere provide one of the sources of living
matter constituting a normal stream biota. When the water of the atmos-
phere falls to earth, the microorganisms and other matter present in the air
are washed out and settle upon the ground. Many of these microorganisms
become entrapped in the soil, while others are carried directly into streams
with the rainfall runoff. In addition, the water picks up microorganisms that
are indigenous to the soil, which nearly everywhere is teeming with these
microscopic life forms. In this manner, water accumulates a wide variety of
microorganisms by the time it flows as a natural stream. The higher aquatic
biota vary widely depending not only on the microflora but also upon the
geochemistry, temperature, and the land character of the stream banks
through which the stream flows. Thus the runoff from a mountain is totally
different chemically, physically, and biologically to the runoff from a
swamp.
After water has formed one of the surface streams of the earth, it
establishes very characteristic biological flora, chemical composition, and
physical properties. The stream continually comes to equilibrium with its
environment, or changes due to the area through which it runs. The stream
biota reflects changes in the qualities of water received by the stream. For
example, if such waters contain domestic wastes (those derived from human
populations), the stream biota will respond with increases in the population
of organisms which are able to use these substances as food until either the
food is consumed or some other growth requirement becomes limiting. As
though to rid their environment of the unnatural or unusual added material,
the complete biota of the stream changes. This ability of the stream to
change its biota, its chemistry, and its physical properties in order to restore
its original state of purity may be termed "the dynamic equilibrium." The
forces of self-purification—physical, chemical, and especially biological—are
constantly interreacting to improve the quality of the water.
The Natural Balance

Nature provides each stream a balanced biota, whether a clean mountain
stream with little contamination, or a heavily laden swamp stream, or a
CHAPTER 7 ON SELF-PURIFICATION AND OTHER RECOVERY MECHANISMS 99
heavily polluted stream. Each of these streams is characterized by a well-

balanced flora and fauna, both in the water and on the bed of the stream. This
condition may well be compared to those conditions existing in a balanced
aquarium. This natural order is not always in concert with man's desires for
further use of this water resource, but the balance is maintained.
The many forms of aquatic life from the smallest bacterium to the largest
fishes, are distributed in the stream in accordance with the environment. The
environment influences both the number and the species present in any
portion of the stream. By creating such a balance, nature provides for
enough of each species of aquatic life so that the water quality is improved.
Each type of plant and animal life is dependent upon the existence of other
living organisms within the same realm. For example, the presence of
organic or other oxidizable matter is required for bacterial growth. Protozoa
depend upon such bacteria, as well a algae, for continuation of life. The
rotifera and crustaceae use all these microscopic forms as a source of food,
while, in turn, the large aquatic plants and animals require certain forms of
the smaller members of the stream biota for their existence. Thus nature
provides for an interdependence and balance of life forms in a stream
environment.
Pollution
The definition of water pollution generally accepted is: Any impairment of
water quality by man that is prejudicial or injurious to its suitability for
other defined uses. Note that this definition does not require that something
be added to the water, but considers as pollution any degradation in quality
as a result of man's use of the water. Thus, a river downstream from an
impoundment may be degraded if the stored water discharged has lost its
dissolved oxygen or if its temperature is too low for summer recreational
uses such as swimming, even though its temperature possibly has improved
for a trout habitat. Pollution through increased turbidity may be caused by
withdrawal of sand and gravel from the river bottom, with the waste wash
water and silt returned to the stream.
Although the effects of temperature increase or decrease are presented in
detail in Chapter 8, it needs to be pointed out that not all temperature
changes of normal ambient river water are "thermal pollution," since the
change may not be degrading the quality for other defined uses. When a
water discharge temperature is considerably below that which is normal in a
given stretch of river, it has been observed that temperature recovery to
normal values is much slower than the concurrent rise in dissolved oxygen to
saturation in the same stretch of river. Under conditions where the dissolved
oxygen content recovers in 8 km (5 miles), the temperature of very cold
water discharged from an impoundment may require as much as 64 km (40
miles) for recovery.
Degradation in water quality may occur below impoundment discharges
100 MANUAL ON WATER
at certain times of the year, for other reasons. Water in the lower layer
(hypolimnion) of thermally stratified impoundments may become anaerobic
for several months of the year. Discharged water entering the streams below
the impoundment may be low in oxygen during warm summer months,
when dissolved oxygen content may be most critical for a waste discharge.
Because of anaerobic conditions in the hypolimnion, the stored water may
contain hydrogen sulfide, iron, and manganese. Hydrogen sulfide lost to the
atmosphere below the dam may be a nuisance there. Iron generally oxidizes
rapidly and is precipitated on underwater surfaces; these surfaces become
visible during low flow periods. Manganese creates more difficult problems
of removal or treatment for users downstream. Cold water discharge from
the hypolimnion may increase the efficiency of power plant or industrial
condensers and may promote trout fishing, but the low temperatures will
prevent swimming, and also change the balance of numbers and species of
aquatic organisms.
When mineral acids or alkalies are added to a stream, either intentionally
or accidentally, many natural factors may be involved in recovery mech-
anisms. Observations of a stream reduced to pH 2.5 by release of a mixture
of hydrofluoric and fluosilicic acids indicated surprisingly rapid recovery to
a normal pH. There was complete absence of obvious toxicity in the river a
few miles downstream from the point of hydrofluoric acid addition,
apparently because of reaction with silicates in river bed minerals. The
reaction of hydrofluoric acid with silica apparently produced silicon tetra-
fluoride, which was lost to the air. This self-purification process was
chemical, rather than biological, but development of a normal biological
pattern quickly followed disappearance of the acid, there being ample soil
organisms available for reinnoculation of the river area, once excess
hydrogen ion toxicity had been removed.
A balance of biological forms is normal for all streams. Although various
species of plant and animal life may predominate in each stream, no two
streams are identical. Clean, pure streams, however, have certain char-
acteristics in common. One property of such a stream is its ability to dispose
of a reasonable amount of pollution. This may be accomplished without an
appreciable disturbance of the normal biota. If a stream receives an
unusually large amount of waste, the biota will change in an attempt to
stabilize such pollutants. However, if the quantity of such foreign materials
exceeds the capacity of the stream biota to render them innocuous, the
higher aquatic life may be completely destroyed for a stretch of the river. As
the unusual food concentration decreases, a new and different group of
organisms develops as a response to the new levels of oxygen and food.
Recovery of Polluted Streams

Forces of Self-Purification
Fortunately, the forces of self-purification usually prevent such complete
destruction of one of mankind's greatest assets. These forces, closely related
CHAPTER 7ON SELF-PURIFICATION ANDOTHERRECOVERYMECHANISMS 101
and mutually dependent, are physical, chemical, and biological in nature.

The physical forces include the action of gravity, light, aeration, dilution,
and turnover. By gravity, the heavier suspended impurities are removed by
sedimentation of individual particles. Finely divided suspended solids and
colloidal matter settle out of the water as aggregates or coagulated masses.
Photosynthesis causes the removal of carbon dioxide and increases the
concentration of dissolved oxygen. Color is removed from water by the
bleaching action of sunlight and by absorption on colloids. Turbulence in a
stream, as by flowing over a rocky botton, causes aeration which adds
oxygen to the water by adsorption from the atmosphere. Carbon dioxide
and gases of decomposition escape to the atmosphere by the same phe-
nomenon. Influents consisting of water from tributaries, domestic sewage,
and industrial wastes are diluted by the receiving stream. Such dilution
assists in the rapid recovery of a stream. The turnover of pools in a stream,
similar to that which occurs in lakes, is another physical factor which aids
stream recovery.
The chemical forces of stream self-purification include oxidation, reduc-
tion, neutralization, and coagulation. Oxidations render suspended and
dissolved organic matter harmless, the products generally being mineral
matter and gases. Incompletely oxidized mineral matter may be oxidized to
completion, or hydrolyzed and deposited, by precipitation, on the bed of a
stream. Likewise, chemical reductions may result in the liquefaction or
gasification of organic matter in the process of anaerobic decomposition.
Mineral matter may be put into solution by such reductions. Ground or
stream water containing dissolved carbon dioxide, in contact with carbonate
rocks, dissolves calcium and magnesium carbonates as their bicarbonates.
This contributes alkalinity to the water, serving either as a pH buffer or as an
actual neutralizing agent for organic or mineral acids. A stream is usually
dependent on entering water of recent subsurface origin for maintenance of
its alkaline reserve. These neutralizations tend to maintain optimum con-
ditions for the biochemical reactions of stream purification. Coagulation
causes the precipitation and deposition of dissolved and colloidal substances
on the stream bed. The removal of such materials permits other forces of
stream purification to become more active.
The biological forces are not only the most important but are the most
active forces; they are closely associated with the food requirements of the
microbiota present.
Bacteria known as heterotrophs can cause the degradation of practically
any organic substance with sufficient time. Newly synthesized economic
poisons, however, constitute especially refractory substances. Instead of
degrading within a reasonable time frame, they tend to accumulate within
the protoplasm of the biota. Many of the economic poisons, including
endrin, chlordanes, and DDT, are very sparingly soluble in water. Instead of
bacterial degradation, the molecules are taken into the fauna and migrate to
the fatty tissues. The molecules in the fatty tissue are relatively innoccuous
until the organism experiences stress or starvation, or both, when death may
102 MANUAL ON WATER
occur. It is an interesting philosophical question whether the reduction in the

concentration of the insecticide in the flowing water by biological uptake
constitutes stream purification where the molecules may still endanger the
biota or one who uses them for food, or both.
Less toxic, newly synthesized organic molecules will persist for days rather
than for hours with normal food substances. Very few studies have followed
the fate of such compounds as trichlorophenol or dichlorotyrosine.
Many groups of the heterotrophs have rather specific nutritional require-
ments so that the identity of the bacteria which respond depends on the
presence of specific molecules of normal food. Thus, the genus Klebsiella is
found in streams below industries processing cellulose, because the members
of the genus can decompose partially degraded cellulose while very few other
genera of bacteria can do it. The ability of different groups to respond to a
wide variety of foods greatly assists in the purification process. Many groups
of such bacteria have relatively specific nutritional requirements, and their
demands for sources of food material greatly assist in the purification of a
stream. Fats, carbohydrates, and proteins are all converted by aerobic or
anaerobic processes to products of a simpler chemical nature. Often, several
groups of bacteria cooperate to degrade such compounds in a series of
interdependent steps.
Other bacteria, known as autotrophs, live in inorganic environments and
complete their life cycle without the benefit of preformed organic matter.
The sole sources of carbon for such microorganisms are carbon dioxide,
carbon monoxide, or carbon suboxide. Minerals essential for the continua-
tion of life are obtained from inorganic salts. The autotrophs are divided
into groups that have specific nutritional requirements. Some live by
oxidizing elemental sulfur, sulfides, thiosulfate, or tetrathionate. Others
reduce sulfates to sulfide or to elemental sulfur. Nitrifying bacteria convert
ammonia to nitrate. Iron organisms oxidize ferrous iron to the ferric state.
Other groups of autotrophs are capable of oxidizing phenols and cyanides.
In fact, as our knowledge of autotrophism grows, we realize that there are
few, if any, unstable inorganic compounds that cannot be converted to stable
products by such microorganisms. As in other reactions caused by bacteria,
this conversion often proceeds through a series of individual reactions
resulting in a specific product which, in turn, is utilized by another group of
microorganisms. An example of this is the oxidation of ammonia to nitrate,
where Nitrosomonas oxidizes ammonia to nitrite, and Nitrobacter oxidizes
the nitrite to nitrate. The autotrophic bacteria thus can convert many
unstable compounds to stable material which can be utilized by higher forms
of life.
The action of bacteria-destroying agents, known as bacteriophages, has a
very important role in ridding a stream of pathogenic microorganisms. Such
bacteriolytic viruses are present in sewage and in the intestinal tract of man
and animals. They specifically destroy, by lysis, certain pathogenic bacteria,
especially the etiologic agents of the typhoids and dysenteries. Also, the pus-
forming bacteria, staphylococci and streptococci, are rapidly destroyed
under favorable conditions. Present knowledge does not indicate that

bacteriophages have any effect upon the microorganisms normally occurring
in streams.
The algae utilize carbon dioxide, an end product of bacterial decom-
position, and liberate oxygen. This aids in the reaeration of streams and
induces rapid stream recovery. In addition, certain inorganic and nitro-
genous substances, also occurring as the result of bacterial activity, are
converted to cellular material. The algae then serve as a source of food for
many of the higher forms of aquatic life.
The protozoa live on organic and nitrogenous matter. Many species prey
upon bacteria, while other species find that the algae provide all of the
nutrients necessary for their life process. Some protozoa contain chloro-
phyll, and are similar to the algae.
The rotifera, the next higher form of aquatic life, prey upon the protozoa
and consume algae. Many are saprophytic organisms and feed upon
decaying organic material. This group serves as a source of food for other
members of the stream biota.
The crustaceae, consisting of aquatic arthropods, feed upon many of the
lower forms of life, especially on the algae and smaller aquatic plants. Such
animals serve to maintain a balance between the lower and higher forms in
the stream biota.
Rooted forms of aquatic plants utilize the organic and mineral matter in
bottom deposits. Also, such forms assist in the aeration of the stream by
utilizing carbon dioxide and liberating oxygen. Aquatic plants create
excellent habitats for fish and other forms of stream life.
Large aquatic animals burrow in mud deposits in search of food. The
bottom sediments are disturbed and thus made more satisfactory for growth
of smaller organisms. One aquatic animal will survive by devouring another.
By such carnivorousness, the animal life of a stream is kept in balance.
Living things in streams are dependent upon many variables. If some
extraneous factor causes an unusual increase in the population of a
particular group of the flora or fauna, another portion of the biota will rise
to diminish this abnormality. The ability of the forces of self-purification to
return a stream to its normal condition, or to adjust itself to a new condition,
represents a real dynamic equilibrium.
Zones of Pollution and Self-Purification

A whole series of events occur when pollutants, such as domestic and
industrial wastes, are discharged into a stream. These changes take place
along the course of the stream, usually in zones that are not sharply divided
but that gradually blend into one another. Such areas shift with meteoro-
logical and hydrographic conditions. Despite this variability, classical zones
of differentiation have been established that are of interest in the natural
purification of streams.
The first area after the entrance of a pollutant is the zone of degradation.
104 MANUAL ON WATER
If the pollution is a domestic or an organic industrial waste, the water

becomes dark and turbid. Sunlight is shut out. Coagulation may occur and
suspended matter settles to the bottom. A deposit of sludge is formed and
decomposition starts. The bacterial flora increases greatly and reduces the
supply of dissolved oxygen. There is an increase in carbon dioxide. Fish may
be very active, especially if the waste will serve as food. If the pollutant is an
inorganic industrial waste, its toxicity might inhibit most of the aquatic life.
When this occurs, dilution must take place before any stream can begin to
recover. The water flora typical of this zone appears as decomposition
progresses. Specific groups of bacteria will predominate, depending upon
the type of waste. Algae may develop; if so, they will consume carbon
dioxide, and help to replenish the supply of oxygen. The presence or absence
of odors will depend upon the efficiency of reaeration. The growth of green
plants is inhibited by the color and turbidity of the water, which reduce
photosynthetic activity. The number of microorganisms, other than bacteria,
decreases because of unfavorable conditions.
Eventually the zone of degradation gives way to a zone of active
decomposition. Here, the dissolved oxygen is nearly depleted and aerobic
microorganisms gradually disappear. An anaerobic condition may result,
and putrefaction will occur, with accompanying foul odors. Bacteria will
continue to flourish, anaerobic microorganisms taking up the work of the
aerobes as the supply of dissolved oxygen is entirely used up. But as soon as
the stream becomes partially reaerated, the aerobes will again predominate.
The growth of algae will rid the stream of excess carbon dioxide and help to
reestablish a favorable oxygen balance. Large aquatic plants and animals
will be present only in areas of favorable oxygen supply.
Following this zone is the area of recovery and clean water. The water
gradually becomes clear, and the bottom deposits do not evolve offensive
odors. The dissolved oxygen increases probably to saturation. There is
enough carbon dioxide to supply a normal aquatic flora. Bacteria decrease
in number as the predatory microorganisms develop. Various protozoa,
rotifera, and crustaceae are present. Algae become established, and the
larger aquatic plants reappear. The more tolerant fish are present, until the
zone of clean water is reached. The clean water is characterized by the
normal flora and fauna previously discussed.
In natural streams, there is a considerable overlapping of these zones.
Some zones are completely eliminated. At times, certain zones may cover
extensive portions of a stream. The whole cycle of self-purification is
adaptable to innumerable conditions. The process may be retarded, inter-
rupted, or accelerated by the entrance of new waste materials or tributary
waters. Each stream operates under conditions peculiar to it alone.
Effect of Pollution on Dynamic Equilibrium

The dynamic equilibrium, or the capacity of a stream to recover after
receiving pollutants, is affected by domestic and industrial wastes. Streams
receiving domestic wastes such as sewage undergo the classical cycle of zonal
self-purification. The same is true, in most instances, for organic industrial
wastes. A notable exception is that some wastes are toxic and inhibit, or
prevent, the operation of stream purification processes. The same is true
of many inorganic wastes.
Acid mine drainage is a pollutant that is indigenous to bituminous coal
regions. It is characterized by being highly acidic and containing large
concentrations of ferrous iron. It is formed by the oxidation of sulfuritic
materials (represented by the formula FeS2) of both operating and aban-
doned bituminous coal mines. A chemosynthetic autotrophic bacterium
(Ferrobacillus ferrooxidans) enhances the reaction several fold where the pH
of the water is in the range from 2.8 to 3.5. Dissolved oxygen is utilized by
these ferruginous effluents through the oxidation of ferrous iron to the ferric
state. Owing to the presence of Ferrobacillus ferrooxidans, which increases
the oxidation rate, the area in which dissolved oxygen is being consumed is
kept to the shortest possible stream distance. After the deposition of iron, as
basic ferric sulfates and hydrates, the stream makes a rapid recovery and
supports most of the normal forms of aquatic life.
Importance of a Well-Adjusted Stream

From the public health aspect alone, it is essential to have well-adjusted
streams. Such a stream, owing to its dynamic capacity to respond to the
entrance of foreign materials, will quickly rid itself of pollutants. It will
remain a clean, healthy stream, and will be a distinct asset to the community.
But the capacity of the stream to respond to pollutants can never be
exceeded without paying the penalty of retarding the self-purification
process.
A normal stream, through its inherent capacity for purification, will be
free of disease-producing bacteria, and will not be toxic for any normal
aquatic flora or fauna. In addition, the recreational benefits derived from
healthy streams will be well worth the care and effort required for their
maintenance.
Water Quality Criteria

It is difficult to establish exact criteria for the measurement of water
quality. There is a wide variation in the mineral content of natural waters.
Water from an unpolluted well in the Corn Belt, pumped into an Ap-
palachian stream, would constitute a contaminant in the minds of most
persons familiar only with the Appalachian groundwater. Water entirely
suitable for human consumption may require treatment for use as boiler
feedwater. Essential industries may produce critical wastes. The use of non-
man-contaminated natural water may need to be hedged with about as many
restrictions as those applying to sanitary or industrial waste.
Biological reactions provide one of the most important means for stream
106 MANUAL ON WATER
purification. A water treatment plant or a waste disposal plant, sewage or

industrial, requires machinery and control devices, trained operators, and
waste land for disposal of residues. A stream, given half a chance, performs
the function of these man-built plants with no equipment or supervision
whatever. All that is required is some intelligence in controlling the
concentration and rate of addition of domestic or industrial effluents to a
stream.
Chapter 8—Thermal Loading of

Water Supplies
The discharge of heated water to streams, lakes, esturies, and groundwater

can cause various effects. Some of these are detrimental while others are
beneficial for subsequent uses of the receiving water. The heated discharge
has been recognized as a problem in water quality and it is attracting
increasing attention. Although ill effects are relatively minor at present, they
may eventually become serious where thermal loading is tending to increase.
The term "thermal pollution" has been used to identify the detrimental
effects of heated discharges. This is not a particularly good term because
heat itself is not a pollutant. All matter contains heat, except when it has a
temperature of absolute zero. The heat content of a substance is directly
proportional to its mass, whereas its temperature is a measure of the
intensity of the kinetic energy of its molecules. Thus it is the temperature of
water that is of primary concern, rather than its heat content.
There are two related reasons for the interest in thermal loading. The first
is the fact that a change in temperature will modify the environment of
aquatic flora and fauna, resulting in a change in the species that can live and
propagate in a given body of water. This is caused largely by an increase in
the rate of biological reactions that accompanies a rise in temperature. The
second is a fear that the tendency for thermal loading to increase will, in
time, exceed the capacity of surface water to dissipate the added heat and
thus create permanent temperature elevations over considerable areas.
There is a threefold difficulty in dealing rationally with thermal loading.
In the first place, it is often assumed that any artificial temperature elevation
is necessarily detrimental to a body of water. It will be shown later this is not
always the case. Secondly, many people feel quite strongly that an artificial
change in the aquatic environment is wrong if it injures any living thing. This
is a primitive idealism that fails to comprehend the need for compromise
between the desirable and the essential. Yet it contributes to minunder-
standings about practical control measures. Finally, the scantiness of reliable
data on the biological effects of thermal loading greatly hinders objective
approaches to the problem.
107
108 MANUAL ON WATER
Sources of Thermal Loads

Heat is added to surface- and groundwaters in many ways. Some of these
are natural, some artificial. Heat is also dissipated by several mechanisms; if
this were not so, the temperature of water would continue to rise.
Stream Sources
The physiographic features of surface streams vary greatly across the
country. Width and depth, channel configuration and gradient, size of
drainage basin and climate—all influence stream temperatures in some
degree, both singly and in combination.
Excluding heat from hot springs for the moment, all natural heat in
streams comes directly or indirectly from the radiant energy of the sun. The
effect of solar radiation can be illustrated by the fact that the summer
temperature of large rivers in western Pennsylvania rises 1° to 2°C from
morning to afternoon, depending upon the amount of cloud cover. The sun
also contributes heat indirectly by heating watersheds, thereby warming the
rain that falls on them. The addition of heat from this source is probably a
relatively minor factor because the specific heat of earth and rock is only
about a fifth of that of water. Further, the cooling effect of a higher rate of
evaporation would tend to increase the rate of heat dissipation. This is a
complex mechanism that is difficult to generalize because it depends upon
the humidity gradient above the surface.
Water flowing off the land can raise stream temperatures in some areas.
This is especially true for the irrigatedjands of the Far West. Water in
irrigation ditches is heated by the sun. A variable proportion of this water
runs off into natural streams. The effect is likely to be small where the water
drains to large rivers.
The electromagnetic spectrum of the sun covers an enormous range (0.001
to 3 X 1014 A), but practically all radiation of wavelength shorter than the
near ultraviolet is absorbed by the outer atmosphere of the earth, while
radiation beyond the near infrared is absorbed by dust and water vapor at
lower levels. The radiant energy that reaches the earth's surface is partly
reflected and partly absorbed. That which is absorbed is reemitted as long-
wavelength radiation. This dissipation of energy tends to lower the tempera-
ture of surface water and it accounts, in part, for the diurnal change in water
temperature.
Evaporative cooling is another important factor in the dissipation of heat.
The rate of evaporation of water is proportional to the air-water temperature
difference and the unsaturation of the surrounding air. When the layer of air
adjacent to a water surface is saturated with water vapor (at its dew point),
the net evaporation is zero because the number of water molecules leaving
the water surface is exactly balanced by the number condensing into it. Wind
has an important effect on evaporative cooling because it sweeps away the
CHAPTER 8 ON THERMAL LOADING OF WATER SUPPLIES 109
saturated air and replaces it with air containing less moisture. The latent heat
of vaporization of water is 580 kg-cal at 30° C (1045 Btu at 86° F). Thus
evaporation of a cubic metre (35.3 ft3) of water would absorb 580 000 kg-cal
(2.3 X 106 Btu). This would have the theoretical effect of lowering the
temperature of 580 m3 (2.05 X 104 ft3) of water 1°C (1.8°F).
Water is also cooled by loss of sensible heat through direct transfer to cool
air in contact with it. Aside from back-radiation of energy, about 20 percent
of the temperature lowering results from loss of sensible heat and the
remainder from latent heat of vaporization.
Climate has a distinct effect on surface water temperature. Stream
temperatures in summer range from about 18°C in the Northeast to 13^C on
the Northwest Coast and parts of the Great Basin to 32° C in the Southwest
and Southeast. These temperatures reflect the influence of air tempera-
ture as it varies across the country.
Groundwater Sources
It has been estimated that some 60 percent of the mean flow of streams is
contributed by groundwater discharged from adjoining aquifer systems.
Below the depth where the temperature of aquifers is unaffected by
insolation and stream infiltration, the temperature of groundwater is
practically constant in any given locality. Thus groundwater discharge tends
to ameliorate stream temperatures because it is warmer in winter and cooler
in summer. Shallow groundwater temperatures normally vary inversely with
latitude, ranging from 3°C in northern Minnesota to 22° C in the Southwest
and 25° C in southern Florida.
Below the depth at which the temperature of the earth's crust is affected by
insolation, the primary control on temperature in a particular region is the
relatively constant flow of heat from the interior. Other factors include the
thermal conductivity of rocks and their contained fluids, volcanic or
magmatic activity, and tectonic movements. The temperature of ground-
water at a few tens of feet is about the same as the mean annual air
temperature. Several authors report an average value of the geothermal
gradient, usually measured to some depth below the zone of active ground-
water circulation, of 1°C for each 30 m of depth.
Water from thermal springs can raise the temperature of streams in some
areas. There are many such springs in the United States, most of which are
west of the 105th meridian. The western springs arise from precipitation that
has percolated into the earth and come in contact with hot geological
formations. Those in the Appalachian range spring from very deep forma-
tions where the water has been warmed by heat conducted from the interior
of the earth. The U. S. Geological Survey has listed 1059 springs with normal
temperature at least 6°C above the mean annual air temperature. A few
springs have temperatures of 100°C or more. Less than a third of the thermal
springs have flows of 0.38 m3/min (100 gal/min) or more. Thus this source
110 MANUAL ON WATER
of thermal loading is of minor significance. Warm Spring in Montana is the

largest one on record; it discharges 303 m3/min,(80 000 gpm) at 20° C. The
aggregate flow of all thermal springs is probably less than 1900 m 3 /min
(500 000 gal/min).
Lakes and Reservoirs Sources

The temperature of natural and artificial impoundments is governed by
physical factors similar to those that influence the temperature of streams.
These bodies of water are usually quite deep as compared with streams. They
therefore resemble deep, quiet rivers in certain respects. Their surface is
warmed by solar radiation and cooled by evaporation and convective heat
transfer. Under summer conditions in temperate climates, unless rather
shallow, lakes differ from streams in that they consist of two dissimilar
layers. A surface layer of variable depth lies upon a relatively stagnant layer.
The surface layer is affected by insolation and the several cooling mecha-
nisms, whereas the bottom layer is insulated from these effects and maintains
a uniform low temperature. Some deep lakes, however, are nearly isothermal
because of wind-induced mixing. Where cool water enters a lake, it tends,
because of its higher density, to underflow the surface layer. Thus tempera-
ture variations are sometimes complex and difficult to predict.
Reservoirs can affect the temperature of streams in a manner that is not
always recognized. When a reservoir is drawn down to augment low stream
flows or to prepare for floor storage, the water is often discharged from the
stagnant bottom layer. This cold water can have a deleterious effect on the
stream biota for two reasons. In the first place, a sharp decrease in
temperature can -harm aquatic life just as an increase can. It is the sudden
change in temperature that causes the ill effect. Secondly, stagnant reservoir
water is often completely devoid of dissolved oxygen. Such water can
destroy aerobic biota, notably the fishes, which are killed by asphyxiation.
Industrial Sources
Most water withdrawn for industrial operations is used for cooling. This
water is discharged substantially unchanged except for an increase in
temperature. The effects of these discharges are highly variable. They depend
upon the temperature difference between the discharge and the receiving
water, the size and configuration of the receiving stream, the rate of heat
dissipation, the quality of the receiving water, and the downstream uses that
could be affected. There are also certain economic considerations that are
mentioned in a later section.
The generation of hydroelectric power is a special industrial use of water.
This operation merely converts the potential energy of water into electric
power, but it can adversely affect conditions downstream. In the first place,
the operation of hydro plants is highly cyclical in many areas because the
output of electricity is varied to meet changes in the demand for power. As a

consequence of the variation of flow through the turbines, downstream
water levels rise and fall markedly. This can cause many undesirable effects.
Secondly, water flowing through the turbines is drawn from a low level in
the impoundment to take advantage of the greatest possible hydrostatic
head. Stagnant reservoir water may contain little or no dissolved oxygen and
is generally quite cold. It can cause the effects described in the preceding
section. For example, a hydro dam on the Chattahoochee River has lowered
the summer temperature of the downstream stretch to 10°C from
30° C. The bass that formerly propagated in the river have disappeared
because they cannot spawn in such cold water. Recently, the river has been
successfully stocked with trout. On the other hand, the water is too cold for
swimming. This is an example of negative thermal loading.
The electric power industry has been mentioned specifically only because
it has been subjected to some of the sharpest criticism. Other industries with
heavy cooling requirements contribute to the thermal loading of surface
water. Their problems are similar to those of the power industry and in some
cases are every bit as acute.
Municipal Sources
Domestic sewage, with or without treatment, is usually discharged to
surface water at the most convenient location. This waste flow normally has
a temperature well above that of the receiving water. Although it contributes
measurably to the thermal load, it has consequences greater than the
physical effect of raising stream temperatures. This method of disposing of
sewage relies upon natural phenomena to render the material innocuous.
The principal natural process involves utilization of the dissolved oxygen in
surface water to oxidize the organic substances in sewage to relatively inert
compounds. The rate of aerobic oxidation varies directly with temperature.
Thus, as the temperature of the receiving water rises, the demand upon its
limited dissolved oxygen resources increases. The increase may be great
enough to use up all the dissolved oxygen. Anaerobic conditions then
develop, and decomposition of the organic matter generates foul gases. At
the same time, all aquatic life that depends upon aerobic conditions is
destroyed. This represents a condition where a rise in temperature can have a
profound effect on water quality. The situation is of course not confined to
municipal sewage; the same result can occur where a heated industrial waste
containing organic matter is discharged.
Many industries and some municipalities store partly treated waste water
in lagoons. Some lagoons are equipped with devices that provide for
discharge of the stored water in an appropriate proportion to the flow of the
receiving water. In other cases, the lagoons are drawn down only during
periods of local flood flows. These lagoons are generally shallow basins and
their summer temperature will usually be higher than that of surface streams.
112 MANUAL ON WATER
Data are not available on this source of thermal loading but the potential
effects should not be ignored.
Thermal Patterns
The rate of natural distribution and dissipation of heat is a major
consideration in deciding upon the size of a thermal load that can safely be
added to surface water.
Field tests have indicated that the die-away of temperature in streams is
roughly exponential; that is, as distance downstream from a source increases,
the stream temperature approaches the natural temperature asymptotically.
Thermal loads are dissipated in several ways, including back-radiation,
evaporative, and conductive cooling.
Heated water discharged below the surface of a stream usually mixes
rapidly with the body of water in a vertical direction. If the discharge is at the
water surface, a substantial temperature gradient may persist for a con-
siderable distance downstream, especially where the stream flow is tranquil.
Actual die-away patterns can be plotted by making a series of temperature
traverses at regular intervals along the course of a stream. In predicting
thermal patterns, it is necessary to measure solar radiation, wind velocity,
and stream flow, as well as the decline in temperature from the source. Quite
off thermal patterns are found in canalized rivers receiving a number of
heated discharges. In a survey of the Monongahela River in Pennsylvania, the
decrease in temperature was measured in terms of stream distance at several
locations at which thermal loading had raised the stream temperature to
32° C or higher. The rate of decrease ranged from 0.5° to 3°C in several 300-
m reaches. The rate of fall varied more with location than with time of year.
Systems for temperature predictions on complex thermally loaded streams
are not yet fully developed, but progress is being made.
The winter temperature pattern is isothermal in typical temperate-zone
lakes or reservoirs under open-water conditions. There is complete vertical
circulation in consequence of the uniform density of the water at about 4°C.
In spring, solar heating warms the water surface, decreasing its density
sufficiently to prevent mixing of the warm, upper water with the colder,
denser water below. As the surface becomes progressively warmer, the
density gradient steepens and the depth to which wind can mix the upper
waters is diminished. The result is a separation of the lake into two isolated
masses consisting of a circulating surface volume (epilimnion) floating above
a relatively stagnant volume below (hypolimnion). Separating the two is the
thermocline, a stratum of water of variable vertical thickness, in which the
temperature decreases rapidly with increasing depth. This stratification
usually persists until autumn, when cooling of the surface increases its
density and leads to breakdown of thermal stability. When this occurs, the
lake "overturns" under the influence of wind, and circulation of all the water
results.
Wind will keep the water in circulation unless a persistent ice cover forms.
In that event, an inverse type of thermal stratification will result until it is
destroyed by warming and wind action. If winter stagnation occurs, it will be
followed by a spring overturn and a second period of circulation.
The most significant changes, in groundwater temperatures are caused by
induced infiltration of surface water or the injection of warm water.
Infiltration from shallow wells in several Central and Eastern locations has
caused an annual variation in groundwater temperature of 8° to 18°C. In
some coastal areas, it is necessary to inject water from air conditioners to
restrain the intrusion of saltwater into aquifers. An example is Long Island,
N. Y., where air^conditioner water is injected at temperatures 2° to 11°C
higher than the water pumped from aquifers.
The rate of thermal dispersion at any point in an aquifer is mainly a
function of the transmissibility of the aquifer and its extent. The thermal
conductivity and specific heat of the aquifer matrix are relatively unimpor-
tant. Heat in water discharged to an aquifer through a well, or recharged
from a surface water source, is dispersed both vertically and horizontally in
the same manner as a dissolved solid; its rate of vertical dispersion is
enhanced by convective movement due to the density contrast between the
formation water and the recharge water. The pattern of thermal dispersion is
also a function of the temperature of the recharge water, because the
viscosity of water varies about 2 percent/deg C in the range 5° to 38° C. This
causes a significant increase in the rate of groundwater flow in the thermally
affected area. The thermal loading of groundwaters results in extremely
complex hydraulic problems because both density and viscosity have
transient values throughout the affected aquifer.
Effects of Temperature on Beneficial Uses

Surface waters provide a habitat for a great variety of flora and fauna. But
the number of species that will live and propagate in a particular body of
water is dependent upon the conditions of the aquatic environment. Water
temperature is possibly the single most important condition of the environ-
ment because of its effect on the concentration of dissolved oxygen, the
biochemical reaction rate, pH, toxicity, and the physical activity of aquatic
animals.
Much of the concern over thermal loading of streams arises from its
potential effect on fish. Millions of people indulge in the pleasure of sport
fishing. It is a relaxing activity that helps to relieve the strains of modern
living. One can easily sympathize with the asperity of dedicated fishermen
when their favorite streams are threatened with a discharge of heated water.
Most species of fish can adapt themselves to a fairly wide range of water
temperature, provided that temperature changes occur slowly enough. Sharp
changes are often destructive, whether the temperature rises or falls rapidly.
Both cold-water and warm-water species are found in the same latitude
114 MANUAL ON WATER
because of their adaptability to various temperature levels. There is an upper

temperature level for each of the two types, however, beyond which they
cannot survive. Brook trout, for example, live in a temperature range from
about 4° to 24° C, whereas large-mouth bass will withstand the same low
temperature but will not be harmed by an upper limit of about 35° C. Both
species will survive temperatures higher than these upper limits for a
relatively short time. The eggs of brook trout will not hatch at temperatures
above 15°C, whereas bass will not spawn in cold water. It should be noted
that fish acclimated to a temperature higher than the normal level for the
species will be killed when the temperature falls to a value that is somewhat
higher than the low-temperature survival limit.
There is one aspect of thermal loading upon which fishermen and aquatic
biologists do not agree. In the South, as well as in the North, condenser
water from power plants greatly improves fishing during the colder months
of the year. Fish tend to congregate in warm water during cold weather and
thus provide ideal conditions for wintertime fishing. Some aquatic biologists
deplore this as an unnatural condition. The increased fish population in a
limited volume of water is thought to deplete the food supply, thereby
making it easier to catch fish. Large catches under such abnormal conditions
would tend to reduce the size of the resident crop. As a practical matter,
however, fish are plentiful where this situation has persisted for many years.
One benefit is the fact that warm water serves as a refuge for species
susceptible to winter killing, and some species spawn earlier in warm water
than in areas beyond the heated zone.
The effect of thermal loading on dissolved oxygen concentrations is a
subject upon which there is considerable misunderstanding. Oxygen is only
slightly soluble in water. Equilibrium concentrations, which range from 14.6
mg/litre at 0°C to 7.1 mg/litre at 35° C, vary inversely with temperature.
An adequate supply of dissolved oxygen is essential for a healthy stream
because desirable aquatic biota live aerobically. Organic matter is the food
supply of bacteria and their utilization of it is an oxidative process which
consumes dissolved oxygen. Bacteria multiply rapidly in the presence of an
ample food supply. There organisms are the natural food of protozoans
which, in turn, are devoured by more advanced creatures. This is the food
chain that provides nourishment for the fishes. All the elements in this chain
consume oxygen. The dissolved oxygen in streams is furnished primarily by
two mechanisms: reaeration and photosynthesis. In healthy streams, the
consumption of oxygen by the food chain is balanced by continual
replenishment from these sources.
Aquatic biologists have agreed that, for some species of fish, the dissolved
oxygen concentration should not fall below 5 mg/litre. In view of the inverse
relationship between temperature and oxygen solubility, they have con-
cluded that any increase in water temperature automatically results in a
corresponding decrease in dissolved oxygen. This implies the assumption of
equilibrium conditions. But static equilibria rarely occur in streams, and

there is evidence for a substantial lag in the adjustment of dissolved oxygen
concentrations to changes in temperature.
Many 24- and 48-h surveys have been made on large rivers and small
streams in western Pennsylvania at all seasons of the year. These bodies of
water receive a considerable amount of organic matter from municipal and
industrial sources, but dissolved oxygen concentrations have ranged upward
from 6.9 mg/litre. Also, oxygen resources were at least as high in summer as
they were in winter. Bacterial metabolism and the respiration of aquatic
plants and animals are continuous, whereas photosynthesis ceases at night.
This has led to the conclusion that dissolved oxygen must necessarily fall to
a low level at night in streams carrying a substantial load of organic
pollution. Nevertheless, these surveys showed that the diurnal variation in
dissolved oxygen in this area is relatively slight. However, streams with high
populations of algae and other plant life will cause significant day-to-night
variations in dissolved oxygen. For example, a stream with an average
dissolved oxygen concentration of 7.0 mg/ litre varied from a daytime
photosynthesis peak of 10.4 mg/litre to a minimum nighttime sag point of
4.7 mg/litre dissolved oxygen concentration. This is not to imply that
organic loads never cause a severe drop in dissolved oxygen. For example,
below the outfall of a large sewage treatment plant, dissolved oxygen fell to
1.2 mg/litre from 7.1 over a distance of about a half mile downstream. These
findings apply only to the area surveyed and are not to be regarded as
typical. They do suggest that speculation is a poor substitute for analysis in
the field.
Much of the difficulty in evaluating the effect of temperature on fish in a
natural environment derives from the limited amount of reliable published
data. Temperature effects are a major interest of aquatic biologists, and a
great deal of data are available on lethal temperature maxima. These values
have come, for the most part, from carefully controlled laboratory studies
which may have no real relationship to any natural habitat. In defense of the
aquatic biologist, however, it is necessary to recognize the extreme difficulty
of getting good data on the effects of thermal loading on a dynamic complex
of living things in a variable environment. This is a field that is now
beginning to receive much needed research attention.
Recent years have witnessed a steadily growing popularity of aquatic
sports. This can be attributed to the instant mobility of most of the populace,
the wherewithal to purchase or rent the needed equipment, and the means to
visit areas that were once financially out of reach. These recreational
activities take many forms, from sailing to water skiing, water polo,
canoeing, or swimming. Devotees of aquatic recreation object to an increase
in temperature because swimming in warm water is not very exhilarating.
Effluents from municipal sewage treatment plants are generally dis-
charged into nearby bodies of water. This disposal method is quite ac-
116 MANUAL ON WATER
ceptable where the degree of treatment conforms with the volume of diluting
water and the forces of natural purification. Aerobic biological activity is the
most important of the purification forces because it converts complex
organic matter into innocuous substances. Within certain limits, the rate of
biological activity roughly doubles with each 10° C rise in temperature. This
places a corresponding burden on the dissolved-oxygen resources of the
water. A rise in temperature is advantageous where the water contains
enough dissolved oxygen to accommodate the increased demand, because
purification will occur more rapidly. But noxious conditions will develop
wherever the demand for dissolved oxygen is greater than the supply. A rise
in temperature, however, will generally stimulate the proliferation of
photosynthetic organisms, especially the phytoplankton, which will help to
make up the oxygen deficit during daylight and cause an added deficit
during nighttime due to respiration. Warm water can therefore be beneficial
in some cases, but detrimental in others.
The ability to draw a glass of cool water from a faucet would often be
desirable, but the almost universal use of refrigeration makes this less
important than it once was. On the other hand, the warmer the supply, the
lower the cost of heating water for the laundry and other purposes. The
lower the temperature of the water entering a purification plant, the more
difficult the treatment becomes. For each 5 deg C rise in raw-water
temperature, the cost of treatment drops 30 to 50c per million gallons
processed. But warm water might promote the growth of algae that can
cause tastes and odors. Thus a rise in the temperature of municipal water
supplies has both good and bad features.
Industry is an extensive user of water for cooling purposes. This suggests
that a source of cool water would confer a benefit so valuable that it would
be a major factor in plant location. But such items as raw materials, markets,
transportation, supply of workmen, and community services are usually so
much more important to a successful enterprise that the temperature of the
local water supply becomes a minor factor. Industry resorts to well water or
refrigeration where it needs water cooler than the surface supply.
The steam-electric power industry might constitute an objection to this
generalization about the limited importance of a cool-water supply. There is
an inverse relationship between the temperature of cooling water and the
efficiency of power generation; this is a matter of real concern in some areas.
This is necessary to protect subsequent water uses, and many kinds of
specific regulations have been promulgated to ensure effective downstream
protection.
It is easy to accept the abstract principle of reasonable use of streams for
waste disposal because it represents economic logic. But problems arise in
applying the principle in specific situations because of the difficulty of
reaching agreement on the reasonableness of a particular use. This is a
consequence of limited quantitative information on the actual effects of
waste waters, especially those at elevated temperatures.
Summary
This chapter has described thermal loading, and the principal mechanisms
by which the added heat is dissipated. The disparate effects that can result
from discharges of heated water have been discussed in some detail. In
particular, considerable attention has been given to the effect on fish because
this commonly provokes the greatest public distress.
Due to Federal regulations, industries have or are facing the technical and
economic problems associated with elimination of heated discharges.
Chapter 9—Sampling and Flow

Measurement of Water
A satisfactory sample of water or waste water must meet two basic

requirements: It must accurately represent the large or small mass sampled
and it must be of adequate size for subsequent laboratory examination.
Since industrial water and waste water must be sampled under a great
variety of conditions, there is no single method that can be universally
applied. The method, location, and time of sampling must be coordinated so
the results obtained will serve the purpose for which the sample is intended.
Waters or waste waters are not fixed in their compositions but show
appreciable changes depending upon their sources, seasonal precipitation,
temperatures, wastes, and processes. This chapter discusses variations in
composition of industrial water and waste water related to environment and
the mechanics of the sampling operation. These should serve as guides in
determining the frequency of sampling, the appropriate points, and suitable
sampling apparatus. '
In the study of water as an engineering material, and in the handling,
treatment, and disposal of industrial waste water, flow measurement is an
essential function demanding applications of many types of flow-measuring
devices and equipment to suit specific conditions. This chapter includes a
brief summary and description of the classes and categories of flow-
measuring devices and equipment to suit specific conditions, the important
points to consider in their selection, and the fundamentals of flow measure-
ment.
Manual Sampling
Types of liquid samples are broadly classified into two major groups, one
called instantaneous, spot, snap, or grab samples; and the other integrated,
continuous, or composite samples. A grab sample is a collected single
portion of water or waste water which serves as a spot check and is useful for
investigating abnormal characteristics. A series of grab samples is necessary
for tracing constituent variations. The volume of individual samples will
depend on the analysis requirements. If samples are to be composited, the
volume collected must be more than needed for the proportional part of the
118
CHAPTER 9 ON SAMPLING AND FLOW MEASUREMENT OF WATER 119
composite. To prevent serious errors in suspended solids concentration in

composite samples, the original grab samples must be kept well mixed while
transferring from one container to another.
Composite samples correlated with flow show the average conditions. The
results are particularly useful in computing quantities of waste material
discharged over a specified period of time. If the rate of flow is constant, the
composite is made up of a number of uniform portions collected at frequent
regular intervals. When the rate of flow varies, a weighted composite is
collected. For this case, the volume of each portion is proportional to the
rate of discharge at the time it is collected. Since it is generally desirable to
minimize the amount of analytical work, samples are composited over 4-, 8-,
or 24-h periods, depending upon conditions.
Manual sampling provides basic information useful in preparing speci-
fications for mechanical devices designed for automatic and continuous
sampling of industrial water and waste water.
Continuous Sampling
If it is necessary for sampling to extend over a considerable time, or when
a continuous record of conditions at a given sampling point is required, it
will be most practical to install automatic continuous sampling equipment.
Correctly designed and installed, a continuous sampling device will provide
more frequent samples than is practical by the manual method, will eliminate
errors due to the human element, reduce costs for supervision, and eliminate
the tedious routine work required by manual sampling.
Since the scope of this chapter does not permit complete description of the
many automatic devices suitable for continuous sampling of industrial water
and waste water, the reader is referred to other sources for additional
detailed information [7,2].'
Sampling at Atmospheric Pressure

In general, industrial water has its origin in rain and melted snow reaching
streams by direct runoff over the surface of the ground and by underground
seepage. It contains very little dissolved material before starting to flow on
the surface or through the earth. Where the soil and surface rocks are
practically insoluble, the surface and seepage waters dissolve very little
mineral matter. Mountain streams usually show little variation in com-
position for miles. The sampling of such water, therefore, is relatively simple
except during flood stages when there are rapid changes in turbidity. This is
important in the operation of a filtration plant. Normally, a sample taken at
almost any location will meet the two basic requirements of satisfactory
sampling. The data in Fig. 1 [3] are an excellent example of this.
120 MANUAL ON WATER
FIG. 1—Composition of Davidson River water.
Usually, water entering rivers as a result of seepage contains larger

quantities of minerals in solution than water flowing over the surface of the
ground, since the contact time is much longer. Hence, during periods of
drought, the concentration of dissolved material in a river is high because
most of the water flows from aquifers. On the contrary, during flood stage,
the dissolved mineral content of a river water is low because it is receiving
the bulk of water from surface runoff. The reverse is true with respect to
suspended solids. The nature of the soil and rocks over which a river flows is
an important factor in determining the kind and quantity of dissolved
minerals which the stream carries.
The problem of sampling becomes quite involved for rivers flowing
through highly industrialized or mining areas. Here there may be little
mixing, even over a considerable distance. This absence of mixing is obvious
at the junction of a muddy stream and a clear one, or of two streams of
different color. An extreme example of the variation in composition from
bank to bank is the Susquehanna River in Pennsylvania. The eastern branch
drains the northern anthracite coal fields and is usually highly acid. The
western branch drains agricultural and forested lands and is usually alkaline.
More than 75 miles below their confluence, these two streams still are not
thoroughly mixed. The eastern branch maintains Us identity as an acid water
along the left bank; the western branch retains its alkaline characteristics
along the right bank. Figure 2 presents typical results of cross-channel
sampling of the Susquehanna River about 60 miles below the confluence [4\.
Complete study of water composition in a river basin would require a
large number of sampling stations and samples taken at frequent intervals.
FIG. 2—Cross-channel sampling of Susquehanna River below Sunbury, Pa.
The cooperative studies by the Pennsylvania Department of Commerce and

the U. S. Geological Survey [5] comprise such an extensive sampling
program.
The sampling effort for one plant at a river location would not be so
extensive. Usually, a single station and daily sampling are adequate.
Occasionally, more frequent sampling is required, for example, during flood
stage when the turbidity of river water changes rapidly. Figure 3 shows the
course of such a change during a period of heavy rainfall.
The characteristics of water in a stream flowing through semiarid country
usually change more rapidly than in more humid sections. Here, the surface
rocks and soil are fairly soluble, and the runoff of each storm quickly
dissolves appreciable amounts of minerals. Streams in such areas should be
sampled at frequent intervals during times of changing stream flow.
For industries located on streams near tidewater, the water taken from the
FIG. 3—Turbidity of Miami River at Cleves, Ohio, 4 to 7 Jan. 1949.

122 MANUAL ON WATER
stream must be checked frequently at various depths to ensure the desired

quality for the particular industrial use. Plants located far enough upstream
to be free of salt during flood stage may encounter seawater intrusion during
low-water periods. In the lower part of some estuaries, the saltwater will
remain at or near the bottom, while the freshwater flows near the top. In
such locations, an elaborate schedule of frequent sampling from many
stations at various depths is required to make certain that seawater will not
intrude unexpectedly.
Large lakes usually remain constant in composition over long periods of
time. The Great Lakes, for example, need be sampled only three or four
times a year to define their characteristics for most purposes. An exception
to this would be for an industrial operation located near the discharge of a
river or for some type of pollution.
Representative sampling of most small lakes will depend on the stream
that flows into the lake basin. Some small lakes may be fed by more than one
stream. The characteristics of the water in the streams therefore will govern
the sampling schedule for the lake.
The composition of the water in a deep lake may vary considerably with
depth, as shown in Fig. 4. The most favorable quality for a given industrial
use should be determined by sampling at different levels [6\. Seasonal
changes may make several sampling stations necessary. The concentration of
dissolved oxygen and carbon dioxide varies considerably with both depth
and the season [7]. Generally, the dissolved oxygen decreases as both depth
and temperature increase.
Another factor that must be considered in the sampling of deep lakes is
FIG. 4—Variation in concentration with depth, Lake Mead, Arizona and Nevada (1 ft -
3.048 m).
that they turn over during the fall. As the surface water cools and becomes
more dense, the lower layers, being warmer and less dense, rise and replace
the former top layers. Usually this turnover is rapid, and the resulting
changes in composition of the water may be considerable.
Shallow lakes do not vary much in composition from top to bottom
because mixing by the winds is generally good. The problem of sampling is
therefore decided by the number and composition of streams feeding such
lakes.
Reservoirs can usually be considered as lakes, but the phenomenon of a
density current frequently is characteristic of deep reservoirs and occas-
ionally of shallow basins. A density current is a stream of water that
maintains its identity, with respect to both chemical composition and
suspended solids, for long distances beneath the surface of the reservoir. It is
caused by difference in density of the flowing water, either as a result of a
difference in temperature or of the material in solution or in suspension or
both. When encountered, such a density current requires a careful sampling
schedule.
Sampling of Water Under Pressure
The composition of most well waters varies only slightly from year to year,
and a single sampling station is sufficient. Usually two or three samples per
year provide satisfactory coverage. There are exceptions, such as deep wells
that have been drilled through strata containing water high in mineral
content into a deeper stratum that contains water of better quality. Leaking
casings or poor construction can lead to contamination of the good water. A
shallow well receiving water by percolation from a nearby stream may
fluctuate in composition. Such conditions necessitate more frequent samp-
ling.
The groundwater in some areas varies in composition with the season.
Such cyclic changes are usually related to solution channels in limestone,
gypsum, or other soluble rock, and to variations in rainfall. Rainwater
quickly reaches the water table and dilutes the higher concentrations of
dissolved minerals. The concentration of the groundwater is thus lower
during the rainy season than during the dry season. Figure 5 shows the
composition of such a well-water supply.
A new well, or one that has been dormant for an extended period, should
be pumped sufficiently before sampling, to guarantee that the delivered
water is representative of normal conditions. This may require several days
or weeks.
For sampling water from a pipe or conduit, it is desirable to choose a
single sampling station. For example, in a cooling system with or without
cooling towers, a single sampling station is desirable for chemical-treatment
control. Satisfactory selection of the station may require a fairly extensive
124 MANUAL ON WATER
FIG. 5— Variation in the composition of well water near Matanzas, Cuba.
initial investigation. The most desirable location is chosen by checking

samples from a number of outlets.
A factor often neglected in locating a sampling station in pipelines is the
possibility of skin samples. An outlet that is flush with the inside wall
frequently will not withdraw a sample that is representative of the cross-
sectional flow. Differences become more pronounced with increasing dia-
meter of the pipe. This difficulty may be overcome by having several
sampling outlets extending different distances into the pipe, using a per-
forated sampling nozzle such as the one shown in Fig. 6. The inlets of such
nozzles must face upstream. The total cross-sectional area of the holes
should be such that the flow of water in the sampling line is proportional to
the flow in the pipe or conduit.
Frequency of sampling should depend upon the magnitude of variations
in composition of the water; that is, the schedule of samples must afford
satisfactory coverage of water changes. A sampling station should not be
located in any section of a pipe where the water is stagnant, such as a dead
FIG. 6—Sampling nozzle for pipe or conduit.

end. A sample from such a location cannot be satisfactory. Physical

conditions often limit the installation of sampling stations, but the best
possible ones should be chosen.
Petcocks and drain valves provide convenient sampling outlets. They are
used extensively and may provide satisfactory samples from small lines or
dead-end samples. It is advisable to test this type of sample outlet for
representativeness, instead of accepting it merely for convenience.
Industrial water in a pressure system may, in many instances, be satis-
factorily sampled using weir boxes, sumps, drains, and overflows as
sampling stations. Such sampling devices should not be used if unstable
constituents are to be determined.
Sampling for Control of Water-Treatment Processes

The sampling of water to control the operation of a treatment process is of
major importance to all types of water-treatment plants. The nature of the
raw water and the quality of the treated effluent govern the number of
samples taken and the type of chemical and physical tests required. If the
composition of the raw water is subject to wide and frequent changes in
quality, the frequency of sampling and tests must be increased accordingly.
Thus changes in chemical dosages can be performed before the quality of
treated effluent is affected substantially. Similarly, sampling and testing of
the treated water must be frequent enough to provide data for maintenance
of water quality and to indicate and deficiency in treatment due to change in
chemical dosage or variation in quality of influent.
Following samples of water permit the use of instruments for continuously
recording certain characteristics, such as pH, conductivity, temperature,
turbidity, color, chlorine residual, oxygen, hydrogen, and silica content.
Instruments may be justified by the saving in costs over manpower formerly
needed for time-consuming laboratory testing, and their feasibility should be
examined.
All types of water treatment plants generally require sampling points on
the raw water, treated water, and other miscellaneous points if the operation
is to be efficient. High-flow clarification and softening plants with sludge
contact or sludge blanket treatment units generally require periodic manual
sampling to measure the sludge concentration and sludge level within the
unit. The frequency of sampling will depend upon the variability in
concentration of the constituent used as the basis for control. Samples
drawn at 8-h intervals are usually satisfactory for a water supply of fairly
constant composition. On the contrary, it may be necessary to sample at less
than 1-h intervals with a supply whose composition is changing rapidly or
with a process that is especially sensitive. Figure 7 illustrates the two types of
water.
Ion-exchange units, including zeolite softeners operating on the sodium
and hydrogen cycles, and demineralizing units of multiple-bed and mixed-
126 MANUAL ON WATER
FIG. 7— Waters of constant and varying hardness.
bed types, may require several sampling points for satisfactory control and
for termination of the operating run before treated water quality has
deteriorated significantly. Since this type of treatment plant lends itself to
full or semiautomatic operation, flowing samples are generally employed
together with monitoring instruments for continuously recording pH, differ-
ential conductivity, and hardness. Complete reliability in silica monitoring
instruments must be attained in terminating the service run of demineralizers
as the exhaustion point of the ion-exchange material is approached. A
practical silica recorder accurate in the range of 0.02 to 0.05 mg/litre and
manual sampling are important to the successful operation of this specific
treatment method. Manual samples taken toward the end of an operating
run and again following regeneration and return to service are usually
necessary, depending upon the type of installation and the degree of
instrumentation. The time for removing these two samples must be estab-
lished by sampling over a short-time interval or on a volume-of-water basis,
near the end of two or three demineralizing cycles and two or three rinsing
cycles. The composition of some water treated by ion exchange changes
rapidly toward the end of the service cycle and may require a sampling
interval as short as 1 min.
Sampling at Elevated Temperatures

The requirements for sampling just discussed may also apply to hot-water
service, with the addition that the sample must be cooled to approximately
room temperature. Failure to cool will result in flashing and an unsatis-

factory sample. Figure 8 shows a suitable type of cooling coil. Another
common type is a pair of concentric tubes coiled to any convenient helical
diameter. The cooling water flows through the outside tube and the hot-
water sample flows in the opposite direction through the inner tube.
Sampling of water from boilers, feedwater heaters, or evaporators also
must be accomplished under pressure using a cooling coil. Manufacturers
frequently equip pressure vessels with satisfactory sampling outlets. The
continuous blowdown line of a modern boiler is often used. In such cases, to
avoid flashing, it is advisable to locate the sampling outlet between the boiler
and the flow-control valve of the continuous blowdown line. It is sometimes
necessary to check various possible sampling stations on some boilers to
locate the most desirable one for the purpose.
A satisfactory sample of steam, particularly from boilers operating at high
pressure and temperature, is extremely important. The ASTM Methods for
Sampling Steam (D 1066-75) [#], gives complete details of the apparatus,
including several types of sampling nozzles, piping, fittings, sample con-
denser, and cooling coils, necessary for getting a steam sample.
Because of the ease with which the approximate purity of steam con-
FIG. 8— Water-cooling coil (1ft - 3.048 m; 1 in. = 25.4 mm).

128 MANUAL ON WATER
densate can be determined by conductivity measurement, this method is

preferable to evaporation of large volumes of sample and is used almost
exclusively for continuously recording sample purity. lonizable gases such as
carbon dioxide and ammonia dissolved in the condensed steam sample
contribute to the conductivity measured, and corrections for their pressure
are necessary. Chemical analyses may be made to find their concentration in
the sample. Automatic steam degassers are available which will practically
remove these dissolved gases.
The schematic diagram of Fig. 9 shows the principal parts of one type of
apparatus used in steam purity analysis in which inaccuracies in conductivity
resulting from ionizable gases and volatile amines are eliminated.
Condensate analyzers employing ion-exchange degassing increase the
sensitivity of detection of inorganic contaminants and eliminate interference
by amines but not by carbon dioxide. Carbon dioxide, however, is usually at
very low concentration in the modern power plant.
A steam sample is condensed at the atmospheric boiling point in a
condensing chamber where practically all the free carbon dioxide and a
small percentage of the ammonia are removed by venting. The effluent from
this chamber passes through a conductivity cell and a small hydrogen
exchange unit. Hydroxide and carbonate alkalinity, volatile amines, and
ammonia are removed in this step. Cations of mineral salts are replaced by
hydrogen ions. Passage through a reboil chamber eliminates bound carbon
dioxide and provides a constant-temperature final sample for a second
conductivity cell. Since the conductivity of the corresponding acids is several
times greater than the original mineral salts, the sensitivity of conductivity
measurement for mineral carry-over in steam is increased.
For maximum accuracy of steam purity determination, the sodium
content of a condensed sample of saturated or superheated steam is
measured by flame photometric techniques. It is possible to measure the
sodium content of condensed steam with a precision of 0.4 /ig/ litre where
suitable apparatus and proper technique are used. The ASTM Tests for
Sodium and Potassium in Water and Water-Formed Deposits by Flame
Photometry (D 1428) [#], describe the method and apparatus. The sodium
electrode method of determining steam and condensate purity is another
method coming into increasing use or supplemental use in the modern power
plant [the ASTM Methods for Continuous Determination of Sodium in
Water by Ion Selective Electrode (D 2791)] [8].
Sampling at Subatmospheric Pressure

It is often necessary to sample water at reduced pressure, as in vacuum-
return systems or condenser hot wells. Here, a different physical arrange-
ment is required for sampling, and extreme precautions must be taken
against leakage. Minor outleakage from a pressure system will not neces-
sarily impair the sample, but inleakage to a sample under subatmospheric
FIG. 9—Simplified schematic steam analyzer.

130 MANUAL ON WATER
pressure will contaminate it. Leakage from pressure systems is visible, but
inleakage to a subatmospheric sample is not.
Deep Sampling
To facilitate collection of samples in deep bodies of water, two types of
samplers, depth integrating and point-type samplers, are in general use.
Depth integrating samplers consist only of a weighted metal frame or
other mechanical device for holding or clamping a glass bottle securely while
it is lowered at a uniform rate throughout the vertical profile of the body of
water.
For the collection of water at a specific depth, the Foerst-modified
Kemerrer-type water sampler is frequently used. The sampler is lowered to
the desired depth in the open position; then a messenger weight which trips
the closing mechanism is run down the suspension line.
The messenger weight must meet the triggering device on top of the
sampler squarely. Hence, a set of messenger weights drilled to accommodate
suspension lines of different diameters permits the use of the sampler with a
wide variety of ropes and cables.
When samples are required for the determination of dissolved gas content
and constituents susceptible to aeration, special equipment and technique
are required. The Foerst sampler can be utilized for this purpose by
transferring the sample by means of a tube connected to the outlet valve
projecting into the bottom of the sample bottle. Several displacements of
water in the bottle should be made to assure that a representative sample is
obtained. ASTM Method D 3370, Sampling Water [#], describes sampling
methods and apparatus.
Sampling of Radioactive Water

In sampling water for the determination of its radioactivity, whether for
gross or for radionuclide assay, the normal principles of sampling apply.
However, because of the amount of radioactive substances in the water and
the potential hazards related to working with water containing radioactive
nuclides, special handling of samples may be required. The ASTM Practices
for Sampling Water, (D 3370) give complete details which cover handling
precautions, apparatus, frequency and duration of sampling, and prepara-
tion of sample containers. It is necessary to make certain that the sample
drawn is actually representative. This may require a sample composited
proportionally to flow rate or, where adequate mixing is not assured, a
number of samples from different points and various depths in river, lake,
well, reservoirs, processing tanks, or reactor cooling water containers.
When sampling for radioactivity determinations, precautions should be
taken to reduce the magnitude of activity adsorbed on the container walls.
When the nuclide amount is in the order of 10~12 g or less, normal adsorption
CHAPTER 9 ON SAMPLING AND FLOW MEASUREMENT OR Wftffft t«r
losses on container walls may amount to an important percentage of thfe

activity. Plastic materials or wax-coated containers give less difficulty in this
respect than do glass or metals. Adsorption of radioactive materials on
container walls often can be minimized by addition of carrier materials of
chelating agents or suitable acidification. The addition of liquids such as
acids, however, may change the distribution of radioactivity between the
dissolved and suspended constituents.
In sampling streams or other natural waters, adsorptive losses are
relatively unimportant since in such waters the easily adsorbable materials
are generally carried on the surfaces of whatever suspended solids are
present. In this case, treatment to prevent adsorption may result in undesired
transfer of radionuclides from the suspended to the dissolved phase.
Sampling of Waste Water

Many of the problems in sampling waste water are similar to those
discussed in this chapter for sampling water. In particular, information
covering sampling in rivers and streams (flowing waters) entering larger
bodies of water, in deep and shallow lakes (standing waters), and in tidal
basins is applicable. These discussions should be reviewed as prerequisite to
the added precautions outlined in this section. Objectives may differ
somewhat in the case of waste water, because it is necessary not only to
obtain representative samples, but to study the waste load, methods of
treatment, and the possibility of recovering materials at a profit. Rigid
specifications—National Pollution Discharge Elimination System permits—
for the quality of waterborne wastes have been formulated by state and
Federal agencies to control pollution throughout the drainage area receiving
an industrial waste discharge, making a carefully supervised sampling
program of plant waste waters mandatory.
Objectives
Common objectives in a sampling program should serve a number of
purposes, including the determination of quantities of industrial wastes to
be discharged, location of major sources within a plant, character of wastes,
establishment of a basis for waste treatment, potential recovery of valuable
material, and the effect of discharged wastes on the receiving stream.
Sampling Procedures
It would be rare if two waste problems were alike in all details, and the
approach to waste sampling must deviate somewhat from the procedures for
water; each must be handled as an individual case. Sampling must be
tailored to fit the operation of each manufacturing plant and the character-
istics of the waste produced. Special procedures are necessary to handle
132 MANUAL ON WATER
problems in sampling wastes that vary considerably in composition, de-

pending upon the type of plant. Thus, suitable sampling locations must be
decided upon as well as the frequency and type of sample to be collected.
Examination of drawings showing underground sewers and manholes in
manufacturing areas will help to locate appropriate points for sampling.
The degree of variation in rate of flow will determine the time interval for
sampling. This interval must be short enough to provide a true representa-
tion of the flow. Although the rate of flow may vary only slightly, the
concentration of waste products may range widely. Frequent sampling, that
is, uniform 10- or 15-min intervals, permits estimation of the average
concentration during the sampling period. This is necessary if a representa-
tive sample is to be obtained.
Sampling Points
Sampling points for waste waters are located preferably where flow
conditions encourage a homogeneous mixture, such as the downstream side
of a hydraulic jump, Parshall flume, or weir.
In sewers and deep narrow channels, samples should be removed from a
point one third the water depth from the bottom. In wide channels, the point
of collection should be rotated across the channel. In any event, the velocity
of flow at the sample point should be sufficient to prevent deposition of
solids. Creation of turbulence during collection of samples may liberate
dissolved bases and volatiles and make the sample unrepresentative. A
representative sample of a heterogeneous mixture of waste water, parti-
cularly one containing oil, is difficult to collect, but collection of a composite
and analysis of an entire individual sample will give a fairly accurate estimate
of the characteristics of the waste.
Recognized testing and precautionary measures [9] should be followed in
taking samples from deep manholes to guard against accumulations of toxic
and explosive gases.
Sampling Equipment
Figure 10 shows a diagrammatic arrangement of a continuous sampler
suitable for uniform flows of homogeneous waste water.
Sample Preservation
Considerable research on the problem of preventing chemical, physical,
and biological changes in waste water samples which might affect the
biochemical oxygen demand (BOD), suspended solids, dissolved gases,
cyanide, and phenol values has failed to perfect a universal treatment or
method, or to formulate a set of fixed rules applicable to samples of all types.
Prompt analysis is undoubtedly the most positive assurance against error
from sample deterioration.
This simple jar-and-tube setup samples waste

effectively when flow is nearly constant. As water
drains from the upper carboy, the vacuum created
siphons waste into the lower one. The rate of flow
is regulated by the pinch clamp to fill the lower
carboy during the sampling period.
FIG. 10—Continuous sampler.
Where it is imperative that stabilization and preservation be attempted,

for example, when a 24-h composite is collected, it is recommended that the
sample be refrigerated at 4°C.
Other methods applicable to many types of waste waters include chemical
treatment, acidification, and alkalinization. Probable errors due to deterio-
ration of the sample should be designated in reporting the analytical data.
Sampling Precautions and Equipment

The volume of water required for a sample depends on the number of
determinations to be made. This is covered in detail in ASTM Method D
3370 and ASTM Method for Reporting Results of Analysis of Water (D
596) [8]. Generally, a minimum of 2 to 3 litres should be collected. A volume
of 4 litres is desirable for the determination of all mineral constituents. In
any event, the quantity submitted to the laboratory should include a surplus
for check analysis. If only a few sample determinations are required, the
sample volume may be reduced correspondingly. When a sample has
settleable solids, the container should be large enough to allow some air
space to facilitate mixing in the laboratory.
There are certain minimum requirements for sampling equipment which
should be observed. For example, the sample container must not change the
sample composition. It must be clean and, if glass, should be of chemically
resistant type. Wide-mouth bottles are preferred for waste waters.
134 MANUAL ON WATER
Rubber stoppers may be attacked by some solvents in wastes and thereby

contaminate the sample. Plastic bottles with screw tops of the same material
may be necessary. These are preferred for samples of condensed steam and
other water in which the silica concentration is important.
Equipment for sampling purposes, cooling coils, degassers, steam sample
nozzles, piping, and tubing intended for collecting acid waste water must be
constructed of material resistant to the water being sampled. Such equip-
ment must also withstand the temperature and pressure involved. Sampling
equipment and piping for water or steam not in continuous use must be
flushed sufficiently to remove all stagnant water.
A new system must be flushed sufficiently to remove foreign material,
such as rust and oil, before placing it in service. The time required for
cleansing depends on the piping and sampling equipment required. It is
possible to be guided by the physical appearance of the effluent in most cases
or by individual checking of samples. The sample outlet should discharge
from the bottom of the sample container; this decreases the chance for
atmospheric contamination. Overflowing the sample receiver also provides
additional desirable flushing of the sample container. A volume of water
equal to at least four to ten times the volume of the sample container should
be overflowed to waste before taking the sample.
The rate of sample flow is highly important. If the rate of flow is too rapid,
gas bubbles may be entrained; if too slow, the sample may become
contaminated by diffusion. When samples are taken for determination of
unstable constituents, unrepresentative or contaminated samples may pro-
vide highly misleading information. Continuous samplers for waste waters
and steam require frequent inspection and cleaning to be reliable.
The foregoing discussion on sampling industrial water and waste water
does not cover the multiplicity of situations that may be encountered. It
offers partial direction for obtaining samples which accurately represent the
mass to be sampled, and which meet laboratory needs for sample size. In all
sampling procedures, the sample collector must exercise good judgment in
deciding whether the sample is representative and if laboratory examination
will provide pertinent and usable data.
Flow Measurement
Flow measurement is an essential function in the study of water as an
engineering material. Since it involves principles of hydraulics and fluid
mechanics that are applicable to all problems of flow measurement, a
knowledge of the theory, methods, and types of equipment for measuring
flow is essential. A simple measurement of the liquid depth in a standard
container per unit of time serves conveniently for obtaining information on
the amount of precipitation on the earth as rainfall. More complicated
measurements involving rather intricate formulas and flow-measuring de-
vices are necessary for determining the volume of surface runoff and
underground seepage collected in flowing streams, flows in distribution

mains and piping, and flows in sewers and open channels used for the
disposal of water as industrial waste.
Row measurement is of inestimable value in the conservation of one of
man's most valuable material assets, water. Economic considerations pro-
vide an outstanding reason for flow measurement of water and waste water.
Heavy costs are often involved to cover the material and labor for water-
gathering systems and storage facilities, which include wells, dams, im-
pounding reservoirs, tanks, pumps, and piping for transporting water.
Installation and operating costs for treatment facilities are important factors
in providing municipal and industrial water supplies. Flow measurement is
an essential item in determining the cost of water delivered to consumers.
Flow measurement is important in the handling, treatment, and disposal
of industrial waste water and domestic sewage, for control of discharge and
making sewage service charges, and to provide records for pollution control
agencies authorized to enforce laws governing disposal of waste. Flow
measurement is a necessary accessory to sampling techniques; without
accurate measurement, very little progress can be made with treatment
methods or to establish the volume of waste that may be discharged without
damage to the receiving waters.
Basic Requirements
The network of piping, underground sewers, and open channels used to
convey water and waste water in modern plants presents many problems in
flow measurement. There is no universal procedure, because flows must be
measured under a variety of conditions. The measurement methods must be
applicable to specific conditions, The type of flow device, its location, cost of
installation, quality of flow data, and suitability for the service must be
considered. This discussion covers briefly the kinds of flow-measuring
devices, fundamentals of flow measurement, description and design of flow
devices, and important points to consider in their selection. These should
serve as a guide in choosing the equipment and method best suited to the
purpose and in establishing a suitable location for obtaining data. The
subject of hydraulics cannot be covered completely in this brief discussion,
and it is suggested that reference books [10] be reviewed to obtain basic
fundamental concepts.
Points of Flow Measurement

The points of flow measurement for water may include the sources of
supply, treatment processes, and distribution systems. Collection points at
sewer inlets, sewer outfalls, treatment plants, open channels, and discharges
to receiving waters are recommended for waste waters.
136 MANUAL ON WATER
Devices and Methods for Flow Measurement [\ \]

Devices used for flow measurement may fall into one of several classes: (1)
head-area meters, (2) functional meters, (3) pumps, (4) displacement meters,
(5) constant differential meters, and (6) inertial flowmeters. Other methods
used frequently for flow measurement include surface floats, salt concentra-
tion, thermal techniques, and radioactive tracers.
Head-Area Meters—Head-area meters are used only in open channel flow
or partly filled pipes; they cannot be used on lines under pressure, but are
useful to measure flow from reservoirs. These meters operate on the
principle that a constriction or controlled barrier in the flow channel will
back up the liquid and create a higher level (head) than the level below the
barrier; that head will be a function of the velocity of flow and therefore of
the flow rate. Head-area meters include weirs, measuring flumes, and open-
end nozzles. Several types of weirs and the basic elements of a Parshall flume
are shown in Figs. 11 and 12, respectively. A typical installation of head-area
meter and V-notch weir is shown in Fig. 13.
Functional Meters—Functional meters measure some function of fluid in
movement and, with supplementary mechanical devices, convert that meas-
urement into rate of flow or total flow. This principle is illustrated
diagrammatically in Fig. 14. The components of a simple mechanical meter
designed to function by differential pressure developed across on orifice
installed in a pipe is shown in Fig. 15.
There are three types of functional meters:
(a) Differential-head devices produce a difference in head between two
points of the device, and this differential is a function of velocity. These
devices include the venturi tube, Dall flow tube, insert nozzle, Gentile flow
tube, and orifice. Typical examples of each are shown in Figs. 16 through 20.
FIG. 11—Types of weirs.

FIG. 12—Basic elements of Parshall flume.
FIG. 13— Typical installation of head area meter for V-notch weir.
138 MANUAL ON WATER
FIG. 14—Pressure characteristics at a concentric orifice in a pipeline (1 in. - 25.4 mm).
These devices are used in main-line metering and in metering flows at any
point where water moves under pressure in a confined pipe.
(b) Velocity meters consist of a device by which a vane or propeller turns
in direct ratio to the rate of flow past the propeller. These devices, similar to
the one shown in Fig. 21, are used in main-line metering, pramp discharges,
etc.
(c) Pitot tubes consist of a vertical tube with a right-angled tip inserted in a
FIG. 15—Simple differential-pressure flowmeter.

FIG. 16—Schematic diagram of Venturi tube.
flow, with the tip facing upstream. The height to which the fluid rises in the
tube is a function of velocity and therefore of rate of flow.
Pumps—Pumps can be used as measuring devices only if the number of
repeat cycles can be determined. Controlled-volume reciprocating pumps
can be used easily for this purpose, but centrifugal pump discharges must be
calibrated for each speed of operation, suction lift, or discharge head. Pumps
are not generally used or recommended for flow measurement and should be
considered only when other means prove to be impractical.
Displacement Meters—Displacement meters operate on the fill-and-draw
principle. Fill-and-draw measurements of reservoirs and tanks can be made
if flows into and out of the tank do not occur simultaneously. When the fill-
and-draw principle is applied to a closed device, a compartment is alter-
nately filled and emptied.
Constant-Differential Meters—Constant-differential meters, commonly
called rotameters, consist of a tapered tube, smaller at the bottom than at the
top, in which is confined a movable element termed a "float." The float rises
or falls in the tube in proportion to changes in the rate of flow of the fluid.
This device, which is illustrated diagrammatically in Fig. 22, is not used for
measuring large flows; it has its greatest usefulness in measuring gas flows.
FIG. n—Dallflow tube.

140 MANUAL ON WATER
FIG. 18—Schematic diagram of flow nozzle.
Inertial Flowmeters—Inertial flowmeters measure mass rather than

volume. They operate by indicting the inertial force of a moving mass of
fluid.
Selection of Primary Units

The device that actually measures the flow of a fluid is called a primary
unit (weir, Parshall flume, open nozzle, venturi tube, Dall flow tube, orfice,
etc.). The primary unit is only one of two parts of a metering device. Before
discussing secondary units, it is desirable to comment on factors in the
selection of primary units. Weirs are infrequently used in permanent
installations for measuring industrial wastes. The following three sections
are devoted to the other principal primary units.
Factors that have a bearing on the selection of the primary unit include
type of service, engineering considerations, installation conditions, acces-
sories, comparative characteristics, and operator's convenience. Not all of
the factors and items mentioned hereafter apply to all types of primary units;
where they do apply, however, full consideration must be given to the
evaluation of each factor.
Engineering considerations include such factors as allowable loss of head;
FIG. 19— Gentile flow tube.

FIG. 20—Orifice plates: concentric, eccentric, and segmental.
throat size of the constriction; acceptable differential pressure; upstream and

downstream flow conditions; type of fluid to be measured; laying length of
the unit (that is, length of pipe it will replace in the line); types of ends
(flange, bell and spigot, concrete coupling); distance to secondary unit; type
of unit to transmit the measured data to the secondary unit; corrosion
resistance required; cleanouts, inspection holes, and manholes; construction
costs as affected by pits for float tubes; and forms for flumes.
Installation factors cover the position of the device (vertical or hori-
zontal), piping arrangements, vaults, and possible bypass installations.
Accessories required might consist of straightening vanes, flushing devices
(intermittent or continuous), vents, vent cleaners, air chambers, floats, float
tubes, sediment traps, seals, etc.
Comparative characteristics of flow-measuring devices comprise accuracy,
low cost, high capacity, range, permanent loss of head, life expectancy,
suitability for solids-bearing liquids, freedom from upstream disturbances,
ability to handle high pressures, and ease of installation.
Secondary Instruments
A secondary instrument or receiver is the device used to convert meas-
urements sensed by, and transmitted from, the primary measuring unit into
FIG. 21—Velocity meter (Builders' Propeloflo).

142 MANUAL ON WATER
FIG. 22—Typical rotameter.
observable information on flow. Liquid level and pressure also can be shown
on secondary instruments.
Secondary instruments are classified in two ways: (a) according to the type
of information desired, and (b) according to the method of transmission
from primary to secondary unit.
Classes by Type of Information Desired

Indicator—The indicator shows the rate of flow at the moment of
observation, in cubic feet per second, gallons per minute, gallons per hour,
millions of gallons per day, etc.
Totalizer—The totalizer shows the running total to time of observation, in
cubic feet, gallons, millions of gallons, etc.
Recorder—The recorder draws a record on a chart, showing the rate of
flow as a continuous record for each instant during a particular period
(hour, day, week).
Combinations are available as indicator-recorder, totalizer-indicator,
totalizer-recorder, and totalizer-indicator-recorder. Indicators have a single
arm and dial. Indicators-recorders may have the indicator arm separate from
the chart, or attached to the pen arm.
Classes By Type of Transmission

Mechanical Transmission—Mechanical transmission exists when the re-
ceiver is operated by cables directly connected to float wells, which are in
turn connected to the pressure takeoffs (taps) of the primary device.

Mechanical transmission also may be obtained by connecting the primary
device pressure taps directly to mercury wells in the receiver unit.
Electrical Transmission—Electrical transmission is used where the re-
ceiver cannot be located near the primary unit or where there is insufficient
pressure for direct connection to the secondary unit. Electrical transmission
is of the time-impulse type (Chronoflo Telemeter) whereby the length of time
that the signal is transmitted is proportional to the magnitude of the function
being measured (rate of flow, pressure, level, etc.) This signal can actuate a
receiving instrument, or control pumps, chemical feeders or valves, or both.
Electrical transmission is possible over any distance from feet to miles, over
telephone circuits, private a-c or d-c lines, or multiservice wires. A se-
quencing unit (sequence unit transmitter) will transmit signals from several
transmitters consecutively over the same pair of wires. Sequencing is useful
in the operation and control of well-field pumps and booster pumping
stations.
Pneumatic Transmission—Pneumatic transmission is useful up to 457 m
(1500 ft); it generally makes use of a varying controlled air pressure ranging
from 21 to 105 kPa (3 to 15 psi). Its principal features are an absence of
wiring, a continuous output signal (pressure), extreme sensitivity to change
in the variable being measured, and immediate response to change in the
variable.
The three types of transmission are illustrated in Fig. 23.
FIG. 23—Types of transmission: (left to right) mechanical, electrical, and pneumatic.
References
[7] Environmental Protection Agency, Training Manual: Industrial Liquid Waste Surveys,
Cincinnati, Division of Manpower and Training, 1969.
[2] New York State Department of Health, Industrial Wastewater Discharges, Albany, Health
Education Service, 1969.
[3] Private communication from Cyrus William Rice and Company, Inc., Pittsburgh, Pa.
144 MANUAL ON WATER
[4] White, W. F., "Chemical Quality of Industrial Surface Waters of Pennsylvania," Paper
presented at the Tenth Annual Water Conference, Engineering Society of Western
Pennsylvania, Oct. 1949.
[5] "Industrial Utility of Water in Pennsylvania," Publication 17, Pennsylvania Department of
Commerce, State Planning Board, Aug. 1947.
[0] "Quality of Surface Waters in the United States," Paper 1022, U. S. Geological Survey,
1944.
[7] Dewey, E. S., Jr., "Life in the Depths of a Pond," Scientific American, SCAMA, Vol. 185,
Oct. 1951.
[g] Annual Book of ASTM Standards, Part 31.
[9] "How Sewer Manholes Are Checked for Toxic and Explosive Gases," Public Works, Vol.
85, May 1954, p. 149; "Gaseous Fuels," American Gas Association Book, 1954, p. 334.
[10] King, H. W. and Brater, E. F., Handbook of Hydraulics, 5th ed., McGraw-Hill, New
York, 1963; Russell, G. E., Textbook on Hydraulics; King, H. W., Wisler, C. O., and
Woodburn, J. G., Hydraulics, 5th ed., Wiley, New York; and Gibson, A. H., Hydraulics
and Its Applications.
[11] Symons, G. E., Water and Sewage Works, WSWOA, Vol. 101, Sept. 1954, p. 395.
Chapter 10—Water-Quality
Monitoring
The broad objectives of water monitoring include the conservation of

water quality and quantity to assure an adequate supply of water suitable in
quality for both public and industrial uses and for the maintenance of fish
and wildlife. Water monitoring is the continuous sampling, measurement,
and analysis of the quantity and quality of various liquid streams. These
streams may include waste-water streams or plant effluents; watercourses
such as rivers, lakes, and estuaries; groundwater; recirculated streams such
as cooling water; power plant streams such as boiler feedwater or con-
densate; or process effluents. Monitoring of public waters and waste waters
protects the integrity of a sewer system, may prevent the upset of municipal
waste treatment processes, especially biological processes, and preserves a
safe public drinking supply and a healthful recreational environment. In
addition to providing many of these same public safeguards, the monitoring
of industrial processes provides for better products at reduced cost, im-
proved process efficiency, and a lesser quantity of waste water that has to be
treated or reclaimed.
Monitoring is no longer a solely voluntary procedure. In the United States
especially, monitoring is a legal requirement of a regulatory agency carried
out to determine compliance with the Federal Water Pollution Control Act
Amendments of 1972, Public Law 92-500. Passage of the 1972 amendments
authorized the establishment, under the Environmental Protection Agency,
(EPA), of national programs for the prevention, reduction, and elimination
of pollution—these to include establishing, equipping, and maintaining a
water quality monitoring system for all waters. Owners and operators of any
point sources are now required to establish and maintain records, to make
reports, and to install, use, and maintain monitoring equipment and
methods in such a manner as the EPA should provide. Under Public Law 92-
500, point-source discharge into waters is prohibited unless the discharge is
authorized by permit. A substantial monitoring program may be necessary
to provide the information required by the permit. Such a monitoring system
would also be able to provide data to answer inaccurate accusations of
harmful or illegal waste-water discharges. Adequate monitoring records can
145
146 MANUAL ON WATER
document that a facility was operating in conformance with permit re-

quirements during any particular period of time.
Earlier methods of monitoring, which largely involved wet chemical
procedures, have, in the past few years, been replaced by more reliable
instrumental techniques which produce results rapidly and accurately. Since
no operator is required, results are more objective, and accuracy can be
checked easily by proper calibration with known standards. Not only are
samples analyzed shortly after collection, but the interval between sampling
and results is minimized. Samples no longer need to be transported to a
laboratory for analysis, eliminating several sources of error. Since the use of
preservatives is unnecessary, these variations are avoided. In general, on-line
instrumental monitoring obviates the inconvenience, the inaccuracy, and the
delay of laboratory procedures.
Wherever possible, and particularly where continuous results are nec-
essary, monitoring should be automated. Besides providing a greater degree
of process control, data may be displayed immediately and still be re-
trievable for historical record. In addition to the advantages just mentioned,
an automatic monitoring system may be less costly than analysis of repeated
grab samples. Installation of a highly sensitive, multiparameter system when
only a quick positive or negative readout is required will incur needless
expense. Good reliable systems require a minimum of attention and
therefore can also be operated and maintained at a reasonable cost. In
general, careful selection of rugged, dependable monitoring systems will
keep costs to a minimum. In an economical, automated system, the level of
sensitivity of the tests should match the sensitivity of the required results.
Considerations when Employing Monitoring Instrumentation

Automatic monitoring of water characteristics provides information often
not available from other techniques. One basic advantage is the promptness
with which the sample is analyzed after it is collected. Other advantages are
the short or zero time interval between samples, and the rapidity with which
the data become available for use in evaluation or control. Some or all of
these factors often contribute to the development of better methods of
controlling water quality, because a complete set of the variables of interest
is provided.
Economics
The economics of automatic monitoring can be attractive. Whereas a
fairly large capital expenditure must be made for the monitoring equipment,
the operating expenses, including maintenance, are much less than the costs
of analyzing repeated grab samples in a control laboratory. Because they are
completely objective, the results are more reliable, and their accuracy can be
determined easily with the proper calibration and checks against standards.
CHAPTER 10 ON WATER-QUALITY MONITORING 147
The ease of checking known standards with automatic instrumentation

permits more frequent restandardizations, thus increasing the reliability of
analytical results.
Process Control
The monitoring of industrial process streams not only offers the direct
advantages of better products at reduced costs, but also the indirect
advantage of improved process efficiency, thus producing less waste material
that must be treated or reclaimed. An example of such process control may
be found in the steel industry, where sheet steel is pickled in acid baths. In
one typical case, it was found that continuous monitoring and control of free
acid concentration and iron content of the pickling baths led to an improved
product, with reduced acid consumption. The decrease in acid usage meant
that less excess acid had to be neutralized when the spent pickling baths were
treated to precipitate ferric hydroxide, and clarified, before the effluent
could be discharged to a natural body of water.
Control of Waste Treatment Systems

The foregoing example points up another important aspect of monitoring,
which now is becoming widespread. This is the continuous monitoring of
industrial waste treatment discharges, as well as the quality of receiving
streams or other natural bodies of water, to indicate performance of
treatment systems, to control waste discharges, and to prevent water
pollution. Programs for monitoring various properties of the water in major
surface supplies such as the Ohio and Delaware Rivers have been in
operation for a number of years and have produced much valuable data.
Monitoring River Quality

Monitoring programs for quality determination on rivers and other
surface supplies have several major objectives. Among these are the collec-
tion of data that can point up changes in stream quality, that indicate the
approximate area of causes of change, and that develop data that provide a
background of water quality for present or possible future uses. An
automated monitoring program of proper design will provide as much data
as desired, or, conversely, the minimum amount of data consistent with the
control desired—another economic advantage. Such programs provide the
mechanics of reporting data so as to permit data utilization either on a
continuing basis or at times when specific causes for change can be
determined. Moreover, data can be evaluated as to significance for the area
or river in question, rather than be limited to comparison of one river with
another. One of the major advantages is the fact that data reported represent
actual conditions rather than averages.
148 MANUAL ON WATER
When properly designed and operated, automated stream monitoring

programs provide the necessary flexibility to permit change in analyses, or in
frequency of analyses, when suggested by evaluation of data. Through
automated monitoring, it is possible to maintain agreement between various
agencies collecting data, by developing uniform mechanics for analytical
techniques and reporting of results. Review of data provided through
monitoring systems will pinpoint the factors that cause changes in observed
analytical results and will suggest determinations that reveal more clearly
either the reasons for or the significance of changes. At present, many
agencies operate control points or sampling stations from which samples
collected are analyzed regularly or intermittently. Some of these analyses are
comprehensive; others cover only a few parameters. All laboratories do not
use identical procedures, or even modifications of the same procedures.
Means of reporting results differ widely, as do sampling procedures. Many
samples are composites collected over periods of hours, days, or weeks.
Neither the period of compositing nor the basis are uniform at all locations
and agencies. Composites may be based on conductivity measurement
agreement, equal volumes of samples collected, or on the ratio of volume to
flow at the time of sampling.
And if the preceding variables are not enough to cause problems, consider
other factors contributing to lack of agreement in results between various
laboratories and agencies. Among these may be cited the varying intervals
between sampling and analysis, the use of preservatives or variation in
means of preserving samples, errors in analysis, or in calculation of results,
plus a lack of appreciation of limits of accuracy and precision of analytical
procedures. Failure to question the reasonableness of results at the time of
analysis, or to warn the analyst of conditions that can influence analytical
techniques, leads to trouble. The collection of samples at different flow
conditions or the reporting of average or medium concentrations from a
number of analyses will provide data of only limited value, as will
determination of average load based on average flow for a time period and
an average of analytical results.
The preceding subsections indicate the numerous aspects of the operation
that could be stabilized and reduced to uniformity by automated stream qual-
ity monitoring programs. The monitoring units could be supplemented by
alarm systems to be actuated by changes outside the normal range, and by
simple checkpoint installations upstream that permit relating changes at
major control points to sources causing the changes. Recorded data would
permit development of inventory maps showing runoff sources that can
influence water characteristics, that indicate the effect of impounded water
quality and the manner of release (low-temperature reservoirs, mine water
storage, oxygen depletion by excessive organic or metallic impurity loadings,
etc.), and that permit ready evaluation of past water-quality records. Water-
quality data could be directed toward appraisal of water suitability for
current and future uses. Load data clearly defining contaminants carried by
CHAPTER 10 ON WATER-QUALITY MONITORING 1^
rivers and tributaries could be used for material balances and could permit
evaluation of the effect of a stream on another river or body of water!
Types of Measurements
The types of measurements used in evaluating water characteristics fall
into two major classifications, physical and chemical.
Physical measurements can be considered broadly (with some ambiguity
and overlap) to encompass properties of temperature, density, conductance,
surface tension, refractive index, radioactivity, pH, redox potential, and
some others employing chromatography, colorimetry, turbidimetry-nephel-
ometry, and specific ion electrodes. Some of these determinations are made
simply by immersing a sensor in the water, while others require a portion of
the water to be taken as sample and diverted through the sensor portion of
the instrument.
The second major category of measurements is those involving chemical
analysis procedures. Here, chemical analysis is defined as the addition of
reagents, causing chemical reactions to occur, before a measurement can be
obtained. Examples of chemical analysis measurements are all those that
add reagents and employ various sensors which respond to pH, redox
potential, color, turbidity, or conductance changes to read out the result. A
brief description of these methods is presented later in this chapter.
Sample portions are collected and measured at programmed intervals
when a continual analyzer is used. The time between continual analyzer
results can vary from seconds to hours.
It must be understood, however, that the type of sampling system required
is dictated to a great extent by the type of instrument to be used and by
stability of the component to be measured. Furthermore, the instrument
output or readout can be no more accurate or representative than is the
sample supplied. If the instrument does not receive a representative sample,
it cannot produce representative data, regardless of how well it is designed or
operates.
When an instrument is to be used for direct control of water or process
streams, as well as for monitoring, the delay interval between the time of
sample collection and detection of the signal must be accounted for and
offset by varying the modes of a process controller.
When an instrument is to be used for process control, the delay interval
between the sampling time and the detection of the analytical result must not
be significantly greater than the time required for the sampled flow to reach
the control point.
Environmental Considerations
Misapplications, such as installing a monitoring instrument in an environ-
ment where it cannot function properly, will lead to unsatisfactory operation,
150 MANUAL ON WATER
high maintenance costs, and production of erroneous data. Monitoring in-

strument manufacturers generally specify optimum conditions for the location
and operation of their equipment. If the application cannot meet these con-
ditions, it may be necessary to use an alternative instrument, or to choose
an alternative measurement for which an instrument that can operate under
the prevailing conditions is available. For example, a hardness monitoring
instrument located in an area that is too hot might cause degradation or
decomposition of a colorimetric hardness reagent. The best solution to such
environmental problems is to locate the instruments in a building or cubicle
that can be air conditioned, or where atmospheric conditions otherwise can be
controlled. The added cost of such housing should be insignificant when the
value of the data and reduced costs of maintenance are considered.
Sampling Considerations
Preparation for the installation of an instrument consists of much more
than (1) speculating on what analysis will suffice, (2) obtaining bids from
several vendors and accepting the lowest one, and (3) setting the instrument
on a concrete pad and flipping on the switch. It is alarming to learn that this
is typical in far too many applications. This type of philosophy is normally
associated with ill-defined measurement objectives and may be the result of
pressure from a regulatory agency to provide immediate data on a particular
pollutant.
Beginning with the premise that the correct analyzer has been chosen for a
given application, the following aspects of sampling require in-depth
consideration if a functional support package is to be assembled.
1. Measurement objective
2. Sample point
3. Sample probe
4. Pumping system
5. Transfer lines
6. Sample conditioning
7. Analyzer interface (stream selection)
Measurement Objective
The definition of measurement objectives may appear to be elementary;
however, in numerous instances, measurement systems are incorrectly
applied. An example of a misapplied measurement system is the use of a
highly sensitive analyzer when something less sophisticated with a high-level
alarm device is all that is required. The majority of environmental analyzers
are used for either record-keeping or alarming, while only a few are used for
control purposes. Illustrations of control instruments would be automatic
titrators, biological activity monitors, and smoke monitors. The analyzer
used for record-keeping should be sensitive and extremely reliable if the
recorded data are necessary for demonstrating compliance.
Sample Point
The location of the sampling point is a direct outcome of the decision on
measurement objectives. Sampling for the purpose of determining the total
carbon discharge into a river is at a completely different location than that
intended to detect a leak at a tank farm. Questions of major importance in
locating a sample point are:
(a) Can a representation sample be obtained at this location? The more
representative sample will be obtained immediately downstream from a
turbulent area. When several streams flow together, causing minimal
agitation, it is best to induce turbulence if a representative sample is to be
obtained. In Fig. 1, three sewers are depicted flowing into a common sewer,
one of which is partially submerged. Note the location of the sample point in
relation to the area of turbulence.
(b) Can a measurement at this point quickly isolate the problem? Do not
expect an analyzer located downstream from the junction of three sewers to
be capable of identifying, during a period of high discharge, which sewer is
responsible.
(c) Is the measurement to be used for determining the magnitude of
discharge, to alarm in case of abnormal discharge, or to control pollution
abatement equipment?
(d) Is the sample point only one of several which will be monitored by a
single, centrally located analyzer?
(e) Will the sample point location allow sufficient lead time for ap-
propriate action?
FIG. 1—Multiple sewer junction.

152 MANUAL ON WATER
Sample Probe
The type and positioning of the sample probe are to some degree dictated
by the measurement objective. If the measurement is not concerned with
solids, particularly heavy solids, then the probe should not be located too
near the bottom of a sewer. The reverse applies if you do not desire to
encounter oils, greases, etc., which normally are surface materials.
Probe design should not be taken lightly. A well-designed probe not only
can minimize the maintenance cost, but also reduce the frustation of trying
to operate an analyzer without an adequate sample.
Stream contents should dictate how the probe is built! If the sewer has a
low level of solids and is pressurized, it is possible to use an enlarged version
of a gas sample thief. When a high concentration of randomly sized solids is
present in an unpressurized sewer, it is best to mount vertically a large-
diameter probe. The suction velocity should be low enough to permit the
heavier objects to fall back down through the probe.
The use of relatively fine screens at the probe inlet is not recommended.
This practice usually results in a high incidence of plugging.
Vertical water probes used with low-pressure safety switch protection on
the pump are capable of self-cleaning. When the probe inlet becomes
plugged, the pump automatically shuts off, permitting the column of water
to backflush the probe. This procedure is not always adequate; if plugging
becomes frequent, it may require the use of pressurized backflushing.
Pumping System
A cursory market survey on pumps will be sufficient to confuse the most
inquisitive individual. Pumps vary in operating principle, dimensions,
specifications, and price.
Experience has demonstrated that price should be the last factor con-
sidered when attempting to establish if a pump can perform the desired task.
The following items require consideration when searching for that "right"
pump for the job.
(a) Does the application demand explosion-proof equipment?
(b) Must the pump be submersible?
(c) Must the pump be self-priming, or will it have a flooded inlet? What is
the suction lift height?
(d) What are the discharge pressure at the desired elevation and the
distance to where the sample is to be delivered?
(e) Is the pump required to handle solids, strings, rags, wood, stones,
sand, etc.?
(/) Are the materials of construction suitable for the waste material the
pump will be handling?
Although open impeller pumps lack efficiency as compared to a positive-
displacement pump, they are far less apt to foul or clog. Also, some sewers
are so difficult to pump that it becomes necessary to use prehomogenization

or delumping hardware prior to pumping.
Transfer Lines
Selection of the proper transfer line requires more thought than merely
using some surplus tubing that happens to be available. The diameter of the
line should be considered in relationship to the length in order to minimize
the pressure drop across the line. The line should have adequate size to
provide sufficient sample to the analyzer without an excessive amount of
discharge pressure. The stream velocity should be high enough to maintain
solids suspension, help reduce wall buildup, and minimize transfer time.
Specific corrective action might be required to reduce or periodically
eliminate biological growth lining the pipe. Backflushing or the use of a
bactericide might be necessary to keep the line open. Again, the materials of
construction should be properly selected.
Sample Conditioning [7]1

Up to this point, the discussion has been concerned with the extraction of
a sample from the sewer and its transfer to a location adjacent to the
analyzer. If extensive sample conditioning is required, it should be ac-
complished at this point. The sample conditioner will very possibly take on
the appearances of an instrument in itself. The components used in the
conditioner will be dictated by the measurement objectives, and the con-
ditioner's ability to perform satisfactorily will play a major part in obtaining
the desired measurement.
Sample conditioners can be considered standard only to a point, after
which they must be customized for a particular application. For example,
measurements requiring only the soluble content can be achieved by phase
separation, skimming, and solids reparation. Obtaining a measurement of
the total amount of a material, both soluble and insoluble, is not an easy
task. It requires effective homogenization and blending in order to get the
solids reduced to a size compatible with the sample ports in the analyzer.
Mechanical reduction of solids to a size which approaches a true solution 1.0
nm mean diameter is not considered feasible.
Functions that should be considered when designing a sample conditioner
are as follows.
Stream Splitting and Solids-Liquid Separation—The technique of stream
splitting can be extremely useful in sample conditioning. It is an ideal
method of extracting a very small sample from a larger volume and making a
pressure reduction all in one step, without the use of valves or pipe reducers.
Experience has taught that the use of valves, pipe reducers, 90-deg elbows,
154 MANUAL ON WATER
and rotometers should be avoided if at all possible in streams containing any

appreciable amount of solids. These components are usually the cause of
plugged sample lines.
Properly designed, a stream splitter can also perform a solids-liquid
separation, thereby minimizing filter loading. Illustrated in Fig. 2 are various
types of stream splitters which have been used with satisfactory results. The
high-pressure splitter and the multiple-stream selector are ideal for streams
with low solids content. Each uses a metal finger inserted into a high-
pressure stream to deflect a small sample into a secondary sample line. The
V-type splitter is bent to achieve the desired volume of sample off one side
FIG. 2—Sample stream splitters and selector.

while the remaining sample is deflected to a return sewer line. The solids-
liquid separator uses the principle of surface tension. The water adheres to
the curved plate while the larger solids continue their free fall through the
pipe.
Use of the simple and inexpensive dropout pot can greatly improve a
sample conditioning system. This device is not as maintenance-free as the
others, but it can greatly reduce the loading on a filter system if operated
correctly.
Dilution—Sample dilution can serve a threefold purpose. It can reduce
the solids concentration in a sample, relieving the load on the filter; it can
reduce the incidence of valve plugging; and it can provide analyzer range-
ability on high-concentration streams.
This approach has been used for the past several years in total carbon
measurements in an effort to overcome the nonlinear response of infrared
analyzers. If the carbon content in the sample can be held below 200
mg/litre, the response of the analyzer is reasonably linear. The calibration
curves in Fig. 3 show the degree of nonlinearity that is encountered when the
carbon concentration increases from 100 to 1000 mg/litre.
FIG. 3—Calibration curves (infrared analyzer).

156 MANUAL ON WATER
An example of how a single-stream system, based on volumetric dilution,

works is shown in Fig. 4.
Homogenization and Blending—Homogenizers are used primarily to
macerate solids and put them into a stable suspension. In order to minimize
equipment size and energy consumption, it is preferable to have a batch
operation with a small sample. This homogenization procedure should be
accomplished just prior to sample analysis, and it is considered a polishing
procedure.
Blending is used principally to maintain sample homogeneity. It should be
utilized at any stage in sample transfer and preparation where stratification
or settling might occur — the most important location being just prior to
sample extraction. Blending can be accomplished by mechanical stirring or
induced turbulence. In the dilution system just described, efficient blending
is obtained by recirculating the diluted sample with a centrifugal pump. The
operation is shown in Fig. 5 with a top view of the dilution chamber. Notice
the swirling action that is induced as the sample stream is reentering the
blending chamber.
Filtration—Although the use of filters is widely accepted in pollution
technology, considerable caution should be exercised when choosing a filter.
Incorrectly applied filters can do more harm than good. They are com-
parable to sample probes in being responsible for insufficient sample
delivery. In addition to plugging, they can cause serious analyzer errors due
to tailing and sample contamination. Filters which absorb materials rather
FIG. 4—Dilution system.

FIG. 5—Sample blender.
than separate them should not be used in analytical methods unless the filter
is guaranteed selective.
Five different filtering assemblies are shown in Fig. 6. The Cuno Auto-
Clean, which utilizes a stack of metal disks, has been reasonably successful in
past applications. The second type of Cuno is a cartridge filter, and it offers
no mechanism for self-cleaning other than replacing the cartridge. The sand
filter has been and is still being used in many locations. If a periodic
backflush is employed, it can do a fair job of solids separation, particularly
on silty solids. The Swirlklean offers an entirely different approach to
filtration from the other ones shown. The sample is injected tangentially into
the chamber between the two side membranes. By differential pressure, a
certain volume of sample is forced through the filtering membrane. This
system can be backflushed by reversing the filtered sample flow through the
membrane, permitting the swirling action of the incoming sample to remove
the solids from the membrane surface.
The single-pass system is an approach similar to the sketch. By sequencing
a filter tape, a small volume of sample, for example, 100 to 500 ml, can be
filtered periodically without consuming a large amount of tape. In this
system, each sample would be filtered through a clean section of tape. This
technique appears frequently in air pollution instrumentation.
No filter can perform indefinitely without some form of maintenance. It
must be either cleaned periodically or changed periodically. Dependence on
human performance for maintaining a clean system never seems to work
satisfactorily. If the cleanup can be automated, by all means do it.
158 MANUAL ON WATER
FIG. 6—Filters.
Analyzer Interface
Interfacing single or multiple streams into a single analyzer is no problem
if two important rules are followed.
(a) Never deadhead a stream; keep it flowing all the time.
(b) If a manifold is used to gather multiple streams, keep it close to the
analyzer. This will reduce the amount of sample required to flush out the
previous sample and thereby minimize the chance of cross-contamination.
Both of these points are illustrated in Fig. 7 with a simple three-stream-
manifold arrangement. The manifold is short and is mounted vertically to
ensure drainage. The valves are always three-way, providing a constant flow
all the time.
FIG. 7—Sample manifold.
Instrument Selection and Installation

It is essential to choose the proper type of monitoring instrument. It must
provide the desired measurement with the necessary specificity, sensitivity,
accuracy, and precision, plus whatever other specific requirements are
essential for a given application. Instrument designs and specifications
should be studied to determine the amount of maintenance required for
satisfactory operation. The cost and availability of spare parts, and the time
required to secure emergency service by factory-trained instrument repair-
men, when needed, must be taken into account as essential items when
potential maintenance costs are being assessed.
Stream and System Dynamics

The dynamics of the stream to be measured also must be considered. A
water supply or process stream that can change characteristics widely within
a few minutes cannot be satisfactorily monitored with an instrument that
reports results 15 min after a sample is taken. The reverse is also true, since it
is not necessary to run analyses on samples of a stable water supply or
process every few seconds when the system monitored will not change
characteristics over a period of several hours. Under these conditions, a
continuously operating monitoring unit may furnish repetitive data of
limited value, and, depending upon the instrument type, waste substantial
amounts of costly reagents.
Readout Systems
Instrument readout systems are many and varied in concept. They range
from simple "Go-No go" lights or the calibrated scale of a thermometer to
the typewritten results of a digital computer, as well as to signal outputs and
160 MANUAL ON WATER
impulses which directly change the parameters of the system being mon-
itored. The basic types of outputs used are analog, digital, alarm, and
process control signals.
The analog readout systems involve the indication or recording of a
varying signal from the monitoring instruments. A wide variety of indicating
meters for monitoring instruments is available. In many cases, indicating
meters are built into the instrument, and the buyer has no choice of
manufacturer or type. This type of indicating readout can be used only
where an operator is available to observe the results.
Analog recording devices are available as an option for most instruments
producing some kind of electrical output. Recording furnishes a permanent
record of the system being monitored and allows evaluation of the results at
any time, without introducing subjective human errors. There is an option
and selection of recorder style to be determined by the user, depending on
how he plans to evaluate and store the data.
Digital readout systems differ from analog systems in that the readout
data are displayed as numbers. Digital indicating instruments also may be
supplied with some type of printer or recorder. Digital records vary from the
electrically actuated mechanical counter, with embossed digits for stamping
numbers on a paper tape, to high-speed magnetic tape systems which store
the information in code, to be read out or retrieved by a computer whenever
desired. Digital recorders not only have the advantage of numerical display,
but also offer the further potential of facilitating data use, in conjunction
with a digital computer, for process control or for storage and future
evaluation of the data by a computer.
Alarm systems can be built in, as they are in the less-expensive "go-No go"
type of instruments. Also, they can be attached to any other kind of readout
device to give an audible or visible signal at the instrument or at a remote
location when a preset maximum or minimum value is reached.
Surveillance Systems
Surveillance systems in the broadest terms are used for the collection of
water-quality data. Automatic monitoring instruments can be used effec-
tively to gather continuous or continual data on water systems under
observation, determining short-term discrete changes that are not obtainable
by composite sampling and laboratory analysis techniques.
Surveillance systems for data collection on surface waters such as rivers
determine changes in stream quality; the nature, location, and degree of
change; and pinpoint within monitoring system limits the cause of radical or
undesirable changes. Within industrial plants, surveillance systems have
been used to detect spills or leaks that would result in losses of expensive
product or raw materials. In other applications, monitoring instruments are
used to protect costly equipment (such as high-pressure boilers) or processes
(waste treatment plants) from harmful waterborne constituents.
Automated monitoring stations have been installed by Government

agencies and quasi-government agencies to monitor river quality. The Ohio
River Sanitation Commission (ORSANCO) and the U. S. Geological
Survey (USGS) have successfully operated systems for many years. The data
collected by these and other surveillance systems are available for qualified
use from the EPA "STORET" group. The historical data accumulated can
present a scorecard on the effectiveness of pollution control efforts.
The most recent thrust of surveillance systems has been to monitor
effluent quality and to regulate the amount of waterborne dumping per-
mitted under threat of defined economic levies or fines against the offender.
It is here that the automatic monitoring systems will have an important
application. When properly designed and operated, these systems provide
the flexibility in test parameters and frequency required to detect spills etc.
(Fig. 8). More importantly, both the regulating agency and the effluent
contributor, be they industrial or municipal groups, can use the same or
equivalent testing devices that are readily standardized and will eliminate
from test performance the hazards of different sampling techniques, analy-
tical methods, and the human element. In case of litigation, the printed
record of the unbiased automatic monitor can be used to provide uniformly
acceptable data.
A total network of surface water monitoring systems has been proposed
and partially completed. Recorded data from such an automated system
would permit the development of inventory maps showing the influence of
runoff sources, impoundment, mine water storage, and oxygen depletion,
and permit logical evaluation of past water-quality records.
Effluent Systems
Effluent monitoring systems have resulted from the regulations and limits
established by the pollution control laws noted early in this chapter. The
effluents considered here are the ultimate discharges from industrial or
municipal facilities to public watercourses.
The purpose of effluent monitoring is to control contaminant discharges
to public watercourses. The monitoring system may be installed for de-
fensive or punitive purposes.
Parameters measured vary with each specific effluent. For some param-
eters such as total nitrogen or chlorinated hydrocarbons, specific process
monitors are not available. In these cases, composite or grab samples must
be taken and analyzed in the laboratory with the associated hazards of
sampling accepted.
In the case of many parameters, instruments are available for continuous
or continual monitoring. These include mercury, chromate, temperature,
dissolved oxygen, conductivity, pH, turbidity, total organic carbon, and
total oxygen demand.
The monitors are capable of a good degree of reliability with low
FIG. 8— Use of TOD-GC system.
maintenance. Systems employing these measurements include sample pre-

paration and data handling functions. The data signal can be used to activate
alarms and control equipment.
The effectiveness of a good effluent monitoring system is shown in Fig. 9.
The Dow Chemical Company in Midland, Mich., made a positive com-
mitment to minimize or eliminate plant organic spills by installing a
monitoring and control system on waste streams. They installed at total-
oxygen-demand (TOD) monitoring network that ultimately operated stream
diversion sytems, transferring flow to holding ponds when inevitable spills
occurred. The results achieved are dramatic and make the case for continual
monitoring. Many of the hazardous spills detected by the system were of too
short a duration to be found by grab sampling techniques, and 24-h
composite samples would have masked the severity of the spills.
Process Systems
Physical and chemical parameters that are important to the operation and
control of industrial processes or to the protection of the process equipment
are automatically monitored by many industries. Typical physical param-
eters include temperature, pressure, flow rate, and liquid level. Chemical
parameters are pH, specific anions and cations, along with general and
specific organic compounds. Techniques employed in measuring the re-
quired parameters include, among others, automatic titrators, selective ion
electrodes, chromatography, spectrophotometry, radiography.
FIG. 9—TOD monitoring results for clean-water sewers.

164 MANUAL ON WATER
Monitoring systems for power plant water quality was one of the pioneer
applications for automated analytical equipment. Several different systems
specifically designed for such service have been introduced and are giving
excellent service, with minimum maintenance expense and substantial
savings in technical manpower costs. Being automated and internally
rechecked against standards, and automatically recalibrated at frequent
intervals, they produce data free from human bias and subjective errors.
Since results are recorded automatically, they are continuously available for
inspection to determine causes of observed variations, and are essentially
tamperproof. Samples can be analyzed as often as needed, at minimum
expense. Moreover, control systems for chemical feed, blowdown, and other
adjustments can be tied into the analytical system and used to activate these
subsystems whenever necessary, or to adjust feed and blowdown rates in
accordance with steam demand and water quality.
Some of the parameters now being monitored by automated systems
include silica (in boiler water, steam, and makeup), Fig. 10; hardness, pH,
conductivity and silica in demineralizer effluents or evaporator output;
phosphate and alkalinity in boiler water; pH, chromate, residual chlorine,
phosphate, and zinc in recirculating cooling water; dissolved solids or
conductivity of boiler water and cooling water; trace metals in boiler
makeup; dissolved oxygen in deaerator effluent and feedwater; hydrazine
residual in boiler water; ammonia in condensate; and corrosion rates.
Nuclear power plants place even greater emphasis on monitoring, and must
add radioactivity detection and measurement to the parameters listed earlier.
Obviously, it would be impractical or too costly to set up such an
extensive analytical control program based on manual grab sampling and
analysis. Only the availability of rugged, continuous monitoring systems at
reasonable cost has permitted the degree of sophistication and control of
water analysis currently being employed in modern power plants. The
monitoring equipment must perform all the necessary functions, including
continuous metering of sample and reagents, mixing, purification (by
dialysis, filtration, etc.), heating, digestion, or reaction, measuring, and
continuous reporting or recording of results, plus adjustment or control of
processes, where indicated. It is a tribute to the skill of those who design and
develop monitoring systems that they are able to incorporate the necessary
sensitivity, accuracy, and reproducibility in units that are sufficiently rugged
and maintenance-free to permit continuous service in severe industrial
applications where they are operated by relatively unskilled personnel or
technicians with minimum background experience.
Specific process applications in other industries include on-line titration of
acidity and iron content in pickling operations to maximize the efficiency of
the operation while minimizing acid dumping and associated neutralization
costs.
Continuous control of rinse tanks with backflow cycling in multirinse
stages reduces water usage and minimizes metal loss and pollution in the
electroplating industry.
FIG. 10— Typical flow diagram for automated silica determination on demineralized water supply.
166 MANUAL ON WATER
Leak detection sensors coupled with automatic dump valves in case of

leaks allows recirculation of condensate and cooling waters previously used
on a once-through basis in many industries
Process losses have been reported where the automatic monitoring system
paid for itself in a single day by early detection of the leak of an expensive
product.
The industry that is now actively pursuing automatic process control is
one of the oldest. Waste treatment plant operation has been brought out of
the dark ages by the recent push for pollution control. Efforts to balance
plant loading rates, nutrient additions, coagulants, chlorine, dissolved
oxygen, and effluent quality, along with early warning of toxic influents,
have resulted in many studies and proposals for automatic control systems.
Figure 11 shows a good composite of the current measurements available.
This plant would feature feedforward or feedback control (Fig. 12) of
chemical additions and have holding pond capacity to divert and treat toxic
influents.
Instruments Dependent On Physical Measurements

A brief description of some principles of direct-measuring instruments
may be of value as a guide in the selection and use of these instruments. It is
not within the scope of this chapter to give all details of the principles
involved, nor to cover all types of available monitoring instruments.
Therefore, the following discussion will survey only the more common
instrument types. Also, the several properties will be considered only where
pertinent to instruments used to monitor water characteristics.
Temperature
Temperature-measuring instruments probably have been available longer
than any other type. The several types of thermometers are well known; they
are manufactured in many rugged versions designed for industrial applica-
tions. Recording thermometers generally are actuated by expansion or
contraction of an enclosed gas or liquid. In deforming its container, the
enclosed fluid provides energy to operate a mechanical linkage connected to
a pen writing on moving paper driven by a clock motor. The pen also can be
actuated by the bending of a bimetallic strip.
Temperature can be measured electrically, or determined by measuring
the voltage generated at the junction of two wires of different metals, as in
the common thermocouple. Outputs of thermocouples are measured either
on millivolt meters or on potentiometric recorders equipped with the proper
type of reference junctions. More recent devices in the electrical tempera-
ture-measuring field are resistors made of a ceramic material on which
electrical resistance varies with temperature. These are called thermistors
and, in general, have a much higher sensitivity than other temperature
FIG. 11 — Waste treatment plant.
168 MANUAL ON WATER
FIG. 12—Automatic control loop for a typical phosphate removal system in a sewage
treatment plant.
measuring devices. However, they require accurate calibration, a stable

electrical power supply, and are limited to temperatures below 300° C.
Conductivity
The conductance of water is another very common direct measurement. It
is defined as the reciprocal of the resistance in ohms, measured between
opposite faces of a centimetre cube of an aqueous solution at a specified
temperature. It is normally expressed in micromhos per centimetre. Pure
water has a theoretical minimum conductance of approximately 0.05 yum /cm
at 25° C. As ionizable substances are added, the conductance will increase,

thus providing- a measure of the total ion concentration of a water sample.
Since the property of conductance varies with temperature, industrial
monitoring instruments should include temperature compensation provi-
sions. The necessary potential is supplied to the conductivity cell by a stable
a-c power supply, usually operating at 60 or 1000 Hz, and various methods
are used to measure the current flow through the cell.
Recently, electrode-less conductivity instruments have been developed.
Here, the solution to be monitored is passed through an electrically insulated
tube with inductance coils or capacitor plates around each end, and
conductance is measured by inducing a current in the solution. This type of
unit is especially useful for monitoring corrosive liquids or those with
suspended solids, and which would either attack or deposit an immersed
electrode surface. Conductance instruments are very useful in monitoring
the total ion content of water, but cannot be used for a specific ion analysis
except under special conditions. Useful data on conductivity measurement
may be found in the ASTM Tests for Electrical Conductivity of Water
(D 1125) [2].
Density
The density of water is the basis of direct methods for monitoring the
amount of dissolved or suspended solids in water samples. Techniques
employed include the use of hydrometer floats in overflow chambers, chain
balance float units, liquid purge differential-pressure units, or air bubbler
systems. With these basic approaches, electrical position detectors or
differential-pressure sensors are used to initiate and transmit signals to
indicators or recorders. Since density varies with temperature, industrial
monitoring instruments for water density also must have temperature
compensation. Careful study of the specifications of available equipments
will be necessary to decide upon the type of instrument best suited to a
specific application.
Refractive Index
The extent to which fluids of different optical densities will bend or refract
light (their index of refraction) is the basis of another monitoring system.
Refractive index measurements can be used to monitor water characteristics
and to give direct indications of dissolved liquids and solids in water
samples. This method applies mainly to binary mixtures of liquids, but also
can be used to determine the total impurities in water. The measurement of
refractive index is temperature-sensitive, so industrial monitoring devices
using this principle must have proper temperature compensation. Various
photometric means are used to detect the extent to which the light beam is
refracted and to send the appropriate signal to an indicator or recorder.
170 MANUAL ON WATER
There are specific applications in which this type of instrument can be very
accurate and sensitive.
Surface Tension
Certain organic impurities such as modern detergents have a measurable
effect on the surface tension of water, and this effect can be used as the basis
for monitoring contamination. Surface tension effects are measured by
determining the force necessary to lift a special platinum-iridium ring off a
liquid surface. This type of tensiometer device is very sensitive to impurities
collecting on the equipment, so the metal ring and vessels must be kept
scrupulously clean, and turbid samples must be carefully centrifuged.
Information on surface tension measurement may be found in the ASTM
Test for Surface Tension of Water and Waste Water (D 1590) [2].
Suspended Solids
Suspended solids can be monitored by turbidimetric or nephelometric
analyzers, depending upon the sensitivity required. A turbidimeter measures
the amount of light passing directly through a sample, while nephelometry is
based on light-scattering properties (Tyndall effect) and measures the
amount of light scattered by suspended solids at approximately 90 deg from
a light beam passing through the sample. Nephelometry is the more accurate
procedure, especially in low turbidity ranges. These instruments are affected
by variation in particle color and size, but can give accurate results for
specific suspended materials. Information on turbidity measurement may be
found in the ASTM Tests for Turbidity of Water (D 1889) [2].
Color
Flow-type colorimeters can be used to monitor variations in colored
components and impurities in water. These instruments are available for
measuring the absorption of light rays, from ultraviolet through visible and
into the infrared wavelengths. Industrial monitoring colorimetry instru-
ments usually are nondispersive. That is, they do not use a grating or prism
to reflect light into its spectrum. Instead they use interference or colored
glass filters which allow only the wavelength band of that filter to pass
through the sample. Nondispersive infrared instruments also can have filters
of standard liquids or gases to isolate wave bands and to compare samples
with a reference standard. The wave bands to be passed by given instruments
are selected at a wavelength where the colored components of interest absorb
the maximum light. The amount of light absorbed at the selected wavelength
then will be a function of the concentration of the component of interest in
the sample. Photoelectric detectors of various designs are used to sense the
amount of light transmitted. A number of different electrical circuits may be
used to measure the output of the photodetector and convert it to a useful

signal for an indicator or recorder (Fig. 13).
Electrochemical Monitoring Methods

The next three types of monitors to be discussed are different from those
listed earlier in that they are based on electrochemical determinations rather
than on straightforward physical measurements. The first type of monitoring
unit is a pH measuring instrument. pH is a measure of the concentration of
hydrogen ions in water, and commonly is defined as the negative logarithm
of the hydrogen ion concentration. A number of electrode systems have been
developed to measure this property of solutions, but the one that has proven
to be the most practical is the glass electrode. It is an electrochemical half cell
based on the potential generated across a glass membrane with a fixed
hydrogen ion (H+) concentration on one side and the unknown (H+) ion
concentration on the other. When used in conjunction with a standard half-
cell reference electrode, the glass electrode passes a very-high-impedance
potential to a high-impedance amplifier from which the output indicates or
records the pH value.
For an industrial pH monitoring instrument, a temperature-sensitive
resistance element is immersed in the sample, and connected in the amplifier
FIG. 13—Schematic diagram showing components of one indicating-alarm monitoring system

for hardness control.
172 MANUAL ON WATER
circuit to compensate for temperature variations. These instruments, when

properly installed with suitable flow cells, and continuously maintained, can
monitor accurately and reliably the hydrogen ion concentrations of water
supplies and other samples. Antimony electrodes have been used in place of
standard glass electrodes. While they have a lower impedance, they are
attacked by a number of materials found in water; as a result, antimony
electrodes do not have widespread application. The ASTM Test for pH of
Water and Waste Water (D 1293) [2] covers pH measurement by means of
the glass electrode.
Oxidation-Reduction Potential
Another electrochemical measurement uses a platinum electrode in con-
junction with a standard reference electrode to measure the oxidation-
reduction potential of industrial water supplies and aqueous process streams.
The readout system is similar to that used for a pH meter except that the scale
reading is in millivolts (+ or -). The sign of the potential is important in this
case, as it indicates whether the solution has an oxidizing or reducing
potential. These monitoring units are classified as direct-measuring
instruments and, in general, are simple and maintenance-free. At the same
time, they usually cannot provide the specificity necessary to monitor
individual components. In other applications, oxidation-reduction systems
may comprise the detector section of more complex chemical analysis
instruments. The ASTM Test for Oxidation-Reduction Potential of Water
(D 1498) [2] describes the apparatus and procedures for this measurement.
Specific Ion Electrodes

The third type of electrochemical monitoring unit is the specific ion
electrode. Electrodes are available to measure the activity of Na+, K+, Ag+,
Cu++, Ca++, Mg++, other divalent ions, F-, S=,C1~, CN~, and NOr ions,
among others. These electrodes measure ion activity, not concentration, so
that appropriate corrections must be made (especially at high activity) to get
a reading of ion concentration. Various designs are employed in these
electrodes. Some are stable and maintenance free, while others are not
suitable for industrial monitoring. The specific application must be studied
carefully and the manufacturer's suggestions followed.
Chemical Analysis Instruments

Chemical analysis instruments have developed out of a need for specific or
sensitive analysis methods for monitoring and controlling water character-
istics. Under proper conditions, many components of interest can be made to
undergo chemical reactions that produce definite reaction products. Meas-
urement of the amount of reaction products, or the amount of reagent
necessary to complete the reaction, is the basis of chemical analysis

instruments. In some cases, the function of analytical instrumentation is
automation of a laboratory or field test procedure. In other applications,
they are the miniaturization of a process reactor. Such analytical instru-
mentation often is relatively simple compared with other modern types of
process control systems. With proper service and maintenance schedules,
automatic continuous analytical instruments can provide valuable informa-
tion that is essential for water monitoring and quality control.
Colorimetric Analyzers
The first of the basic instruments to be described is the colorimetric
analyzer. Such units include systems for adding color-forming reagents to
the sample. For maximum accuracy, the colorimetric analyzer should
incorporate the additional features of a comparison cell, a system for
reacting and measuring comparison reagents (or blanks), and an automatic
zero or calibration reset. Provisions for necessary pretreatment such as
heating, extraction, filtration, digestion, or dialysis can be provided. In
many applications, the colorimetric chemical analyzer has the potential of
being the most specific and sensitive water-monitoring instrument available
(see Fig. 13).
Galvanic and Electrochemical Reactions

Galvanic and electrochemical reactions provide the basis for a number of
chemical analysis instruments such as those used to measure hydrazine, or
dissolved oxygen and other water-soluble gases which have an effect on
electrode reactions when they are added to the electrolyte of a galvanic cell.
The physical variations of these instruments are too broad to permit a
generalized description of them. Temperature compensation is required for
such monitoring units. In dissolved gas analysis, the gas of interest usually is
diffused through a plastic membrane to allow long, unattended service. The
electrolyte, and in some cases the electrodes, eventually become spent and
these must be replaced. Depending upon the physical design, these electro-
chemical monitoring units often require the least maintenance of any type of
chemical analysis instrument.
Coulometric Analyzers
Coulometric instruments are based in principle on Faraday's law, which
states that a definite quantity of direct current will cause a valence change in
a specific number of ions in a solution. If the amount of current passed is
known, then the number of ions changed also is known. Instruments based
on this principle have an absolute calibration rather than relative calibration
methods based on standards, typical of most other instruments. The
174 MANUAL ON WATER
generation of the new valence state can be accomplished either directly in the
sample solution, or in a separate vessel, in a pure chemical reagent precursor
which is then reacted with the sample. The latter allows some freedom from
interferences by other ions in the sample which also may undergo a change
of valence state when they are near the generating electrodes.
The coulometric instruments include a liquid section containing gen-
erating and detecting electrodes which indicate to the circuit that sufficient
current is passing to react with the ions in the solution. The readout section
includes indicating or recording instruments which measure current sent to
the electrodes. The instrument can use either a flow cell or a fixed-volume
measuring cell that performs batch-type analysis. The batch type more often
are laboratory rather than field monitoring instruments.
Coulometric instruments require periodic cleaning, depending upon the
sample stream. In some types, where generating electrodes are the material
for the formation of the reacting valence state, the electrodes must be
replaced at regular intervals. The electrical generating and measuring circuits
require the normal maintenance for circuits of these types, and maintenance
problems may vary according to the electrical components used.
Corrosion Measurement
The corrosivity of water is measured by the chemical action of the water
on a standard sample of a given material sensitive to the corrosive substances
of interest. The amount of chemical reaction occurring in a period of time
may be estimated by several different methods, with various instruments. In
one case, the current generated by solution of the metal is measured; in
another, the change in the resistance of a metal element due to reduction in
cross section of the metal probe is measured. The primary maintenance
problems with this type of instrument are requirements for periodic re-
placement of the standard sample units or probes, and leakage, which
permits internal resistance changes or short-circuiting of probes.
Titrators
One type of chemical analysis instrument having a broad field of
application is the titrator. These units are based on conventional volumetric
analysis methods. Such methods are adequately covered in ASTM standards
and in many other texts on analytical chemistry, so the details of analytical
procedures will not be discussed. The titrator-type instruments include
devices for measuring the volume of a standard solution or titrant added to a
measured volume of sample, and systems to determine when the added
volume of standard solution has provided enough reagent to be equivalent to
the component it is reacting within the sample.
The titrator instrument most often is an intermittent or batch type;
however, continuous units are available. A continuous titrator instrument
usually operates by pumping the sample through the system at a constant

rate, and varies the rate of titrant addition so that it will be stoichio-
metrically equivalent to the component of interest in the sample. Several
different approaches are used for measuring and indicating the volume of
titrant used. The detecting system to a large extent is dictated by the method
of chemical analysis necessary. Detector systems which may be used can be
based on pH, redox potential, colorimetric, turbidometric-nephelometric,
and conductometric measurements. The maintenance requirements and
adaptability of these titrator instruments vary as widely as do their design.
The specifications of all such instruments available and economically
acceptable should be studied thoroughly to determine which type will
perform most effectively in a desired application.
Operation and Maintenance of Monitoring Instrumentation

Because of the great variation in types of monitoring instruments, no
attempt will be made here to provide specific operating and maintenance
instructions. In most cases, the instrument manufacturers supply specific
operating and maintenance information for their equipment, this informa-
tion being derived both from experience during development and that gained
during field operation. Failure to follow the manufacturer's environmental
control specifications and operating instructions will lead to unsatisfactory
results and equipment malfunction more quickly with monitoring instru-
ments than with any other type of process control equipment.
While many types of monitoring instrumentation are relatively complex,
and sometimes challenging to operate and maintain, the data provided are
very valuable in defining water characteristics, and justify extra maintenance
effort.
The operator's attitude and approach to the operation of monitoring
instrumentation, and particularly to its preventive maintenance, must be
different than it is to maintenance of control valves, pumps, or other units of
process hardware. The monitoring instrument is designed to furnish in-
formation about the characteristics of water in the operator's system which
can be obtained in no other way. Such instrumentation can be considered as
an extension of the human senses and memory which allows the user to
extend his natural abilities and technical skills. Through automated in-
strumentation, more efficient use of manpower can be made, together with
accumulation of increased knowledge, leading to more efficient processes
and better control of aqueous systems. Monitoring equipment then should
receive all the care and attention to operating and maintenance details which
it deserves.
For maintaining water-monitoring instruments, a general recommenda-
tion is to set up a regular preventive-maintenance program at the time the
equipment is installed, and follow up to make sure it is carried out. Any
system allowing instruments to operate without scheduled maintenance until
176 MANUAL ON WATER
a component fails is unsound, and will prove both time-consuming and

expensive. Preventive-maintenance programs will vary widely according to
the type of instruments used and the nature of their environment. In-
formation needed to establish maintenance programs can be obtained from
manufacturer's operating and maintenance manuals, or directly from the
nearest manufacturer's service representative.
Obviously, it is not possible in this chapter to explain the detailed
functions of every type of water monitoring instrument. The two main
reasons are space limitations, and the unusually rapid developments in the
monitoring field which soon would make detailed descriptions obsolete.
Therefore, for background data, a list of suggested reference texts is
appended. In addition, much useful information can be obtained by a survey
of the manufacturers' current literature on monitoring systems.
References and Bibliography
References
[7] Sugar, J. W. and Brubaker, J. H., "Development of Sample Conditioning Systems for
Automatic Environmental Instrumentation," presented at the 19th Annual Instrument
Society of America, Analysis Instrumentation Symposium, St. Louis, Mo., April 1973.
[2] Annual Book of ASTM Standards, Part 31.
Bibliography
Considine, D. M., Ed.-in-Chief, Process Instrument and Controls Handbook, McGraw-Hill,
New York, 1957.
Siggia, Sidney, Continuous Analysis of Chemical Process Systems, Wiley, New York, 1959.
Minnar, E. J., Ed., ISA Transducer Compendium, Instrument Society of America, IFI/
Plenum Data Corp., New York, 1963.
Holzbock, W. G., Instruments for Measurement and Control, 2nd ed., Reinhold, New York,
1962.
Liptak, B. G., Ed., Instrument Engineers' Handbook, Chilton, Philadelphia, Pa., 1969.
Chapter 11—Analysis of Water

and Waste Water
Water quality is a function of both the source of the water and its
intended use. The quality of a given stream can be defined only in terms of
the requirements of the system in which it is being received or from which the
water is discharged. The waste from one use may well be the water supply for
another use. Before one can control water quality for a use, a great deal must
be known about both the chemical and physical characterizations of the
possible water sources, and the exact requirements for the planned use.
Knowing these data, one can set out to develop the proper treatment for the
selected water supply. Once the treatment program is started, analytical data
are needed to determine its effectiveness. Overtreatment causes unnecessary
expense, and undertreatment also will cost money because of equipment
inefficiency or failure, plus resulting downtime, maintenance expense, and,
potentially, reduced product quality. One may conclude that analytical data
are needed for effective planning of the system and for monitoring per-
formance of the system subsequently. The Annual Book of ASTM Stand-
ards, Part 31 [7],1 presents a wide variety of analytical methods to determine
the various constituents and characteristics of significance to water quality
evaluation.
This chapter provides the analyst with criteria necessary for the selection
of the best method to check for each constituent, and guidance to properly
interpret results of the analyses, in relation to the water system under
consideration. To accomplish this goal, the chapter discusses (1) the value of
a well-defined, systematic program of analysis; (2) sampling methods for
various needs; (3) the basis for selection of the test methods to achieve the
desired results; (4) the cross-checking of analytical data; and (5) the
interpretation of results.
Although the determination of water quality is done to select the proper
treatment program, the description of specific treatment methods is beyond
the scope of this chapter. (Treatment methods are discussed in Chapter 5.) In
selecting sites for new plant construction, where water is required for plant
operations and for use in manufacturing products, the quality and quantity
177
178 MANUAL ON WATER
of water available, and its analysis, play important parts in the economics of
plant operation.
In order to design an effective and comprehensive treatment program, it is
necessary to determine all the variables which may be important in the use of
the finished water. At first, a complete analysis would be wise, including
consideration of seasonal changes; thereafter, perhaps, only selected tests
would be necessary on a routine basis to monitor the more important
variables and maintain the water treatment program. Variations in certain
characteristics often can be predicted on the basis of the source of water. For
example, deep well water would not be expected to vary appreciably in
organic matter or hardness, while river water may vary rapidly and widely in
organic matter, ammonia, and other constituents, depending the nature of
water use upstream. Lake water will vary in organic loading depending on
depth and size of the lake, changes in temperature, microorganism growth,
and pollution factors.
Rivers represent the most commonly used water sources. The impurities
encountered can come from diverse sources, including agricultural runoff,
municipal wastes, mine drainage, industrial waste, and natural forest runoff.
A survey upstream will show several of the specific impurities and often the
contributing sources which affect the water quality at the plant site.
Treatment required for a river water, depending on the proposed use, may be
extensive or minimal as indicated by interpretation of the analytical data.
For example, a river water for cooling tower purposes may require as little
treatment as a chromate corrosion inhibitor and only minimal control
analyses. The same water source used for boiler feedwater may need
coagulation, sedimentation, filtration, de-ionizing and de-oxygenation to be
sure that corrosion and deposit formation are minimal. Frequent sampling
and extensive analysis will be needed for proper operation. A municipal
water supply form this same source may require softening filtration,
activated carbon treatment, and chlorination to meet the drinking water
standards and other end-use criteria.
In many cases, comparisons are made between two or more potential
supply sources to determine which source offers the best possibility of
efficient and economical operation. A comprehensive analysis of the water
from each source is the key to the selection. The economic evaluation must
consider all treatment costs, as well as the costs of obtaining and ensuring
continued availability of an adequate volume for the needs today and for
some time in the future.
Nature and Scope of an Analytical Program

The requirements of the system with respect to the quality of the supply
water determines the extent of analysis. The number of tests and accuracy
required to maintain control are not the same for all systems. For example,
adequate control of water in a low-pressure boiler may require analysis of
CHAPTER 11 ON ANALYSIS OF WATER AND WASTE WATER 179
only four or five items on samples taken once per day or once per shift from
a single sampling point, with sensitivity requirements on each analysis falling
in the milligram-per-litre range. On the other hand, control of boiler water in
a high-pressure steam generator may require eight or ten determinations of
items measured in micrograms per litre, on samples taken at various spots
throughout the system, every hour or two. The difference in the cost of these
two control programs would be substantial. Time and money can be saved
by designing the analytical program to collect only essential data. However,
the program must be complete enough to ensure protection of expensive
equipment from corrosion or deposit formation.
There are several factors to consider when selecting an analytical method,
including the sensitivity required, the interferences present, and the varia-
tions in precision and accuracy that can be tolerated. Even the background
of the analyst to perform the test must be considered, since many procedures
are satisfactory only in the hands of a skilled operator. The discussion of
water analysis methods in this chapter primarily is designed to assist in the
proper selection of individual procedures for specific applications. The test
methods described in the latter part of the chapter are grouped according to
the constituents for which tests are to be made. For example, all the methods
for dissolved oxygen in water are discussed together; so also are methods for
iron, and methods for silica. This is done to emphasize the importance of
recognizing differences between the various methods available for deter-
mining individual constituents, and to illustrate the importance of selecting
the proper procedure for a given application. The discussion of methods also
includes a brief review of the relative importance of each element in different
water treatment and water use situations, to direct the analyst to the
selection of the proper method. For example, a series of samples to be
analyzed (for hardness, silica, alkalinity, iron, and oil) from a once-through
cooling system would be handled differently than similar samples from
boiler feedwater. The hardness that can be tolerated in cooling water
ordinarily will be high compared with the allowable hardness in boiler
feedwater. Thus, one must select a very sensitive hardness test for boiler
feedwater, whereas a less sensitive but accurate procedure is adequate for
most cooling systems. The amount of silica in a feedwater for a high-pressure
boiler ordinarily is kept below 1 mg/litre, and the test method must be very
accurate in this low range. Silica in cooling water is not as critical with
respect to sensitivity requirements, but the high degree of accuracy still is
required for optimum operation. While the test for alkalinity is equally
important in both systems, the method used for boiler feedwater may not be
applicable to treatment programs used in cooling systems. Cooling water
from open recirculating systems often is maintained with pH below 7, and
corresponding alkalinity levels, whereas boiler feedwater normally is alka-
line.
The concentration range and possible interferences such as highly colored
materials present in the sample because of treatment will have to be
180 MANUAL ON WATE R
considered in selecting a method. When iron is found in boiler feed water, it

may mean either that corrosion is taking place, or that suspended iron from
reclaimed condensate or the raw source water is not being completely
removed. In either case, the iron entering the boiler can contribute to
deposits which will cause heat losses and eventual failure of boiler tubes.
Feed water with 10 /ug/ litre dissolved or suspended iron entering a boiler
producing a million pounds of steam per hour can deposit 4.5 kg (10 Ib) or
more of iron oxides every 1000 h of running time. Thus a few micrograms
per litre of iron would indicate potential trouble, and a sensitive test is
required. Iron in cooling water is not quite as critical, but it must be
controlled in the low milligrams-per-litre range, for it reacts adversely with
the treatment chemicals, promotes growth of iron bacteria, and accelerates
deposition of fouling layers.
The methods selected for analysis of boiler condensate water will differ
from those applicable for untreated water. One would expect boiler con-
densate to contain extremely small quantities of dissolved or suspended
constituents such as calcium, iron, magnesium, and sodium. Thus, test
methods for these constituents must be very sensitive. On the other hand,
untreated water will contain these same constituents in large quantities,
requiring methods with much higher concentration range. The same raw
water may contain certain significant constituents such as fluoride or
cyanide in very low concentrations, but which may be important in the
treatment program. In that case, the raw water will require highly sensitive
analytical methods for these constituents, and less sensitive methods for the
more standard mineral determinations. Brine or brackish water supplies will
require a different analytical approach than freshwater. Brines or brackish
waters contain very high concentrations of dissolved salts. A method with a
high concentration range and freedom from interferences must be selected.
Freshwater samples usually will be lower in dissolved solids; thus a more
sensitive test with lower concentration range may be utilized.
Process water supplies will dictate different methods of analysis than will
industrial waste-water samples. Process water may contain added water
treatment chemicals to control certain elements. The analysis must check on
the effectiveness of added treatment chemicals. Certain constituents, when
present, are objectionable and must be checked. The presence of objec-
tionable substances may require chemical treatment to be altered, and
additional checks made.
Industrial waste control introduces another type of analytical program,
since waste solutions are very different from the water received for use. The
U. S. Environmental Protection Agency (EPA) and other regulatory bodies
fiffve set limits (National Pollutant Discharge Elimination System) (NPDES)
for materials which may be discharged into a receiving water. For a given
source of waste water, the potential extent of contamination and accept-
ability of the waste for discharge will depend on the concentration of the
particular constituents in the waste water. The appropriate analytical
methods are dictated in the regulation, with the option of proving another
method equivalent or better. Methods appropriate for the receiving water
may not be useful for waste waters, because of concentration differences,
interferences, and other factors.
The foregoing discussions show that one must determine not only the
specific constituents tests, but also the type of water system being con-
sidered, and the required tolerances of the method with respect to precision,
accuracy, sensitivity, and interferences.
Planning the Program of Analysis

The planning of a sound program of analysis is the first necessary step if
appropriate information is to be realized from the data gathered. This
planning is done with the objective of assuring that standards of water
quality and treatment established for a particular system are met and
continuously maintained. Some care is required in planning. A program that
is too extensive, providing unneeded data, causes unnecessary expenditure
of time and money. On the other hand, an inadequate program of sampling
and analyses can lead to incorrect interpretation of data and improper
treatment, with resultant damage to costly equipment or products. The test
performed should give the greatest economy of time and motion while still
providing the necessary information.
There are four primary considerations in planning a program of analysis:
(1) determining what stages of the system need control and, therefore,
analysis; (2) determining which constituents at each stage need control and,
therefore, how thorough an analysis is required; (3) determining what
method will be used in measuring a particular constituent; and (4) deter-
mining how frequently the constituent must be measured.
In general, an adequate plan will include provisions for regular analyses of
the supply water and any reused water (such as condensates and recircula-
ting cooling waters); analysis of treated waters; analysis of waste-water
discharges for compliance with regulatory agency standards; and periodic
performance checking of the entire system. Included in the planning must be
the sampling procedures and location. Reasonable forethought will result in
effective representative sampling.
Selection of the constituents to be measured, and therefore the extent of
analysis on a particular sample, will be dictated by the requirements of the
system. A complete analysis is required where full knowledge of the various
constituents is needed. For example, supply waters initially would require a
complete analysis in order to determine the suitability of the supply for a
process use. Provisions for determination of substances that will interfere
with or damage the process, along with those substances which may interfere
with the treatment, are important considerations in a supply water. How-
ever, an abbreviated analysis program can be used where only a few
constituents are critical. In a treated cooling water, for example, measure-
182 MANUAL ON WATER
ment of two or three constituents may suffice for maintaining proper

treatment level and water conditions. A combination of an initial complete
analysis followed by subsequent abbreviated analyses is common. Variations
in a few key constituents may be used to signal the need for further system
checking.
In supply waters, an abbreviated analysis may be made once the total
constituents are identified. In this case, the constituents measured must be
carefully chosen, as surface supplies are subject to seasonal changes and
periodic contamination; the constituents measured must reflect these changes.
Of considerable aid in determining constituents which must be considered, or
are present in troublesome amounts, is the use of qualitative or semi-
quantitative analytical methods. For example, optical emission techniques
can provide rapid analysis of a large number of constituents present in water.
In this manner, water supplies can be checked conveniently at regular
intervals, to pick up variations which may arise. If emission equipment is not
available, it is sometimes 'profitable to send samples to a consultant
laboratory, to supplement other information available for the analytical
program.
After the constituents to be measured are determined, the appropriate
analytical method must be chosen. The ruling considerations here are the
sensitivity, accuracy, and precision desired, the interferences present, and the
speed and the cost of the analysis. High-precision methods are used where
accuracy and precision are important considerations, and where many
interferences must be considered. However, they usually are more com-
plicated and time-consuming than alternative methods, and the accuracy
obtained may not justify the time Consumed. In many instances, the
highprecision methods may not provide the sensitivity needed. Less de-
manding, simplified methods are appropriate for use where applicable, and
can save appreciable time and effort when certain interferences are absent.
Also, some analytical methods provide alternative procedures that meet
sensitivity requirements. In many instances, accuracy is not the most
important consideration.
In the discussion of the various procedures, later in this chapter, other
tests mentioned (not included in the Book ofASTM Standards, Part 31 [7]
may be applicable to special cases. In choosing the analytical method, the
choice made must be an appropriate balance of the foregoing considerations
to meet the needs of the system and the process most effectively, and with
maximum economy. An important first step in every analysis is the
recording of the sample source and physical appearance. The sample should
be labeled appropriately; the analyst must know the type of water,
treatment added, if any, and the sampling conditions. The analyst's own
description is also important, such as the solution color, turbidity, the
presence and color of any solids, and whether magnetic or nonmagnetic,
whether the container appears oily or greasy, and the presence of unusual or
unexpected odors. All these factors can be used as a check on interpretation
of final results, or as a guide to potential analytical difficulties or inter-

ferences. In planning analytical methods to use, brief qualitative or semi-
quantitative tests again can be very helpful. They may be used to indicate the
concentration ranges of the constituent sought, or of interferences present,
thus guiding the choice of methods.
The fourth consideration in a program of analysis is the frequency of
testing, that is, how often a particular analysis is required. The demands for
control of a particular system again will determine the choice. Considera-
tions of the type of water needed, the nature of the supply, the methods of
treatment, the response time of the control system, and the overall cost of the
program are significant to an effective operation. Costs of full protection of
equipment and processes must be weighed against the hazard of reduced
testing. After a water system and treatment program are fully stabilized,
further reduction of the frequency of analysis can be considered. Simplified
procedures also may be adopted with periodic checks to verify the accuracy
with standardized tests.
Table 1 and associated Tables 2 and 3 suggest an order of application of
methods which permits the determination of existing properties and con-
stituents in a logical sequence. The flow diagrams separate the samples
affected by air contact from those not affected by air contact. Further
divisions are indicated for determinations to be made on flowing samples,
separate samples, and special samples. With these considerations in mind, an
analytical program can be planned which will conserve time and sample.
Often, a simple test for a given constituent or property can be used to detect
its presence to determine whether it should be analyzed by the standard
method. Such preliminary tests often will save considerable time. One or
more constituents can be determined on a single sample by a scheme of
successive analyses of nitrates or aliquots, while other constituents require a
separate sample arid special handling. Many of the tests can be made while
others are in progress. With some practice, the analyst will acquire a sense of
timing which will permit him to carry out the necessary operations with the
greatest possible speed consistent with appropiate accuracy and precision.
Furthermore, he will learn to connect the presence of one constituent with
the necessary absence of another. For example, ferrous iron would not be
present in water that contains residual chlorine.
Special Apparatus for the Laboratory Methods Presented

Advances in analytical instrumentation have made available to the analyst
a wide variety of sophisticated apparatus ranging from highly versatile to
highly specialized application. The choice of instruments should be governed
by the nature and frequency of the analytical information that is desired.
Accuracy, sensitivity, and available funds must be considered. Highly
precise, sophisticated instruments are expensive; many require special
laboratory conditions such as controlled humidity and temperature, and
184 MANUAL ON WATER
TABLE 1—Outline of principal sample groups.

Sampling procedures shall be in accordance with the applicable ASTM methods.
SAMPLES FOR PROPERTIES AND CON-
SAMPLES FOR PROPERTIES AND CON- STITUENTS UNAFFECTED BY AIR
ST1TUENTS AFFECTED BY AIR CONTACT CONTACT
In order to obtain the true concentration of proper- Samples for which these properties and constitu-
ties and constituents affected by air contact during ents are to be determined may or may not be filtered
sampling, or during the interval between sampling in the laboratory prior to analysis, depending upon
and analysis, it is necessary to employ special meth- the nature and the amount of undissolved material,
ods or equipment, or both, for sampling, and, pre- the method of analysis used, and the specific in-
ferably, to carry out the determinations immedi- formation required. If the analysis is to be
ately. If determinations are not made immediately representative of the sample as collected at the
in the field, it must be realized that laboratory resource, the amount and composition of the un-
sults reported for these constituents are based on dissolved material present should be determined.
the sample in an as-received condition, and are not Spectrographic and X-ray diffraction examination
necessarily representative of the water sampled, of the undissolved material is of value in con-
since the laboratory usually has no control over, or nection with these procedures. (Note 1.)
knowledge of, the sampling methods used.
(Note 1.)
SEPARATE POR-
SEPARATE TIONS OF A SEPARATE
FLOW SAMPLES SAMPLES SINGLE SAMPLE SAMPLES
(See Table 2) (See Table 3) Separate portions of Samples for the de-
Flow samples shall be Samples for which a single sample may termination of extract-
used in determining these properties and be used for determin- able matter and floating
these properties and constituents are to be ing these properties liquids must be collected
constituents at the determined require and constituents in- separately in glass-
sample source. This that the sample con- dividually or, in some stoppered flasks of
information may be tainers be sealed cases, in a sequence chemically resistant
continuously and au- against air contact of analyses. The cus- glass. Samples may also
t o m a t i c a l l y deter- during the interval tomary precautions be collected in cork-
mined, indicated, and between sampling and against contamina- stoppered flasks, pro-
recorded. analysis; or the sam- tion by a i r b o r n e vided the stoppers are
ples must be chem- solids, or too long wrapped in metal foil.
ically fixed immediate- storage in unsuitable
ly after sampling. containers (Note 1),
must be observed in
the case of samples
to be analyzed for
these constituents.
NOTE 1—If the property or constituent to be determined may be affected by reaction of the
sample with the sample container, a separate sample should be collected in a special container
that will not itself contaminate the sample.
specially trained personnel. It is necessary to consider whether their use will

justify initial and continuing costs.
An example will help illustrate some of the factors influencing the choice
of instruments. Suppose a particular water is to be analyzed routinely for
iron content. A decision is made as to what concentration of iron is
significant. ASTM Tests for Iron in Water and Waste Water (D 1068) [7]
present a colorimetric method applicable to several concentration ranges. If
the concentration of interest is greater than 0.1 mg/litre, Nessler tubes can
TABLE 2—Properties and constituents affected by air contact.

Flow samples shall be used in determining these properties and constituents at the sample source.
This information may be continuously and automatically determined, indicated, and recorded.
Electrical Conductivity: D 1125
Hydrogen: D 1588
Oxygen: D 1589, D 888
pH (Hydrogen Ion): D 1293
be used and the color comparison made visually. If sensitivities down to 0.02
mg/litre are needed, a filter photometer is required. Below 0.02 mg/litre, a
spectrophotometer will be required. It will be noted that the greater the
sensitivity need, the more costly the instrumentation becomes. However, the
instrument also becomes more versatile. The spectrophotometer in this case
can be applied to the upper ranges as well, which may be a consideration in
other procedures that must be run. Convenience also may be a considera-
tion. In the example just given, when the iron concentration of interest is
greater than 0.1 mg/litre it still may be profitable to use photometer instead
of visual comparison methods. The frequency of tests required may make it
a considerable waste of time and effort to prepare the series of standards
needed each time for visual comparison. A photometer, with a standard
curve, then can be a timesaving convenience.
The choice between manual and automatic or semiautomatic instrumenta-
tion will be governed primarily by the frequency and speed of analysis
required. The needs of the analytical program will determine the choices of
TABLE 3—Properties and constituents affected by air contact.

Samples for which these properties and constituents are to be determined require that the sample
be sealed against air contact during the interval between sampling and analysis; or the samples
must be chemically fixed immediately after sampling.
Acidity and Alkalinity: D 1067
Ammonia and Ammonium Ion: D 1426
Calcium and Magnesium Hardness: D 511; D 1126"
Carbon Dioxide, Carbonate, Bicarbonate; D 513
Chlorine Residual: D 1253, D 1427
Chlorine Requirement: D 1291
Hydrazine: D 1385
Hydroxide Ion: D 514
Iron: D 1068
Nitrite Ion: D 1254
Dissolved Oxygen: D 888, D 1589
pH:D1293*
Phenolic-Type Compounds: D 1783, D 2580
Sulfides and Hydrogen Sulfide: D 1255 *
Sulfur Dioxide, Sulfite, and Bisulfite: D 1339 b
"These properties and constituents are usually determined on separate portions of a single
sample that has not been sealed against air contact, because, in most instances, it is not justified
to obtain separate, sealed, or chemically fixed samples for these determinations. When these
determinations are requested on samples, a note should be made qualifying the analytical results
in this regard.
*In order to calculate these constituents correctly, the pH value of the sample must be noted
simultaneously with the sampling. The truest pH value would be in the flowing source.
186 MANUAL ON WATER
instrumentation. A time-and-cost evaluation may become necessary when

considering instrumentation selection. Of course, any fairly extensive
analytical program will require the usual laboratory facilities. The quantity
and types of analytical balances, fume hoods, steam tables, ovens, laboratory
bench space, sample storage, and a host of ancillary equipment must fit the
analytical program. Certain analyses will require special sampling apparatus;
this too must be considered.
Not to be overlooked is safety of the laboratory or testing area. Control
measures for toxic gas, fire, and explosion hazards must be planned.
Flammable compounds must be safely stored. Personnel safety involves
mandatory use of protective clothing, safety shields and glasses, a knowledge
of the chemicals and reactions involved, and an aggressive laboratory safety
education program.
Laboratory Practices
To help ensure the success of an analytical program, certain practices

should be observed in the laboratory and become part of the accepted
routine. Analytical methods may represent sound scientific principles and
years of development, but if they are misused or abused, the effort is wasted.
Similarly, the most modern facilities and equipment must be properly used
in order to yield meaningful information. A review of some of these factors
should be helpful.
Sample and reagent containers must be selected with care. The choice of
clear or colored storage containers, soft or borosilicate glassware, plastic
containers, and the type of stoppers or caps used can influence accuracy.
Generally, a method or sampling procedure will indicate the type of
container required; these requirements should be adhered to rigidly. Clean-
liness of glassware and containers is of the utmost importance in water
analysis. Since concentrations of some constituents may be sought in the
micrograms-per-litre range, even slight traces of extraneous material can
cause appreciable error. Glassware may be segregated for specific uses, but it
is better to treat the entire lot for the severest case.
To eliminate or recognize errors traceable to reagent variations, operator
habits, and similar factors, blank determinations should be run. Spot checks
on the reagent water [see ASTM Specifications for Reagent Water (D 1193)]
[/] should be made, and the there should also be occasional checks on
known samples to establish relative accuracy among analysts in the labora-
tory and between laboratories.
To eliminate other potential sources of error, periodic checks on cali-
bration or standardization of analytical equipment or solutions must be
made. With equipment such as spectrophotometers, standard curves should
be checked frequently. New dye lots should always be checked because batch-
to-batch variations may shift calibration curves. pH meters must be
calibrated frequently. Even thermometers should not be assumed to be
correct, but should be compared with one known to be accurate.
Another important aspect of laboratory practice is the quest for new and
improved methods. The analyst should not cease to search once he has found
a satisfactory one. Methods offering greater specificity and reliability are
continually being developed. He should be aware of these, evaluate them,
and introduce them, if they are improvements over existing procedures.
Since the analyst is an important part of any method, safety must be
considered in laboratory practices. Knowledge of potentially dangerous
reactions and conditions is essential, and due precautions should be part of
the laboratory routine.
Analytical Sampling
The best analytical methods are of no value if poor sampling techniques
are employed. The sample must be representative of the water to be
examined, and the concentration of the constituent of interest must remain
the same until the analytical tests are made. The technique of sampling
varies, depending on the type of water and its use.
After the program of analysis is planned, consideration must be given to
obtaining representative samples. In this respect, the proper location of
sampling points and auxiliary equipment such as cooling coils or degassers is
important. Several methods in the Annual Book of ASTM Standards, Part
31, cover the equipment and general techniques of sampling:
D 1192, Specifications for Equipment for Sampling Water and Steam.
D 1066, Sampling Steam.
D 887, Field Sampling of Water-Formed Deposits.
D 3370, Practices for Sampling Water.
These range from the selection of valves, fittings, and equipment to the
methods of sampling of specific waters, such as steam and boiler waters.
Method D 3370 discusses the sampling of water and includes a list of sample
volumes required for various determinations. All of these methods should be
consulted when setting up the sampling procedures and locations, and the
appropriate method consulted when sampling. A complete discussion of
sampling principles is presented in Chapter 9.
Many samples require special techniques and containers. For example,
trace iron or copper analysis requires specially cleaned containers and
acidification; samples for silica analysis should be taken in plastic bottles;
each sample for dissolved oxygen must be flushed through the bottle with at
least ten times its volume of sample and then carefully stoppered to prevent
air entrapment. Each ASTM method will specify the technique of sampling
required and should be followed accurately to ensure meaningful analytical
results.
Discussion of Methods
A variety of methods is generally available for each determination. In
addition to the use to be made of the results, the choice of procedures is
188 MANUAL ON WATER
governed by the probable range of concentrations, the accuracy required, the

presence of interfering substances, the skill of the analyst, and the apparatus
available.
This chapter is intended as an aid in making the best compromise between
speed on the one hand and completeness and accuracy on the other.
Therefore, the discussion of each constituent indicates, in most cases, several
available methods for determining that constituent and provides some basis
for making choices in specific cases.
The discussion of methods in this chapter is presented in alphabetical
order by constituent or property in each of the following categories, of which
Table 4a-e gives a complete listing:
Organic constituents (Table 4a)
General properties (Table 4b)
Radioactivity (Table 4c)
Bacteriological (Table 4d)
Inorganic constituents (Table 4e)
TABLE 4—Categorical listings of ASTAf methods discussed in this chapter.

(a) ORGANIC CONSTITUENTS
Concentration Methods
D 2910-75 ~ Organic Matter, Removal of, by Activated Carbon Adsorption
D 2778-70 (1974) Organic Matter, Solvent Extraction of
General Methods
D 2579-74 Carbon, Total and Organic, by Combustion Infrared
D 2908-74 Organic Matter, Volatile, by Gas Chromatography
D 1252-67(1974) Oxygen Demand, Chemical (Dichromate)
D 3250-77 Oxygen Demand, Total, by Combustion
D 3560-77 Oxygen Demand, by Combustion-Infrared Analysis
Specific Constituents
Adenosine Triphosphate (ATP) Content of Microorganisms, proposed
D 2327-68 (1974) Amines, Primary and Secondary; Octadecyl- and Dioctadecylamine
D 2330-68 (1974) Alkyl Benzene Sulfonate
A. Direct Methylene Blue Colorimetric
B. Interference Limited Methylene Blue Colorimetric
D 2036-75 Cyanides
A. Total, After Distillation
B. Amenable to Chlorination
C. Amenable to Chlorination Without Distillation
D. Cyanogen Chloride
E. Spot Test for Sample Screening
D 2036-75 Cyanogen Chloride
D 2909-74 Cyclohexylamine, Colorimetric with Diazotized p-Nitroaniline
D3113-75 EDTA, Tetrasodium Salt of,
A. Total (Chelated and Unchelated)
B. Unchelated
D 3478-75 T Herbicides, Chlorinated Phenoxy Acid, by Gas Chromatography
TABLE 4—Continued,
(a) ORGANIC CONSTITUENTS (continued)
Specific Constituents Continued
D 1942-74 Morpholine, Colorimetric by Thiocarbamate

D 3371 - 74 T Nitrites, by Gas-Liquid Chromatography
D 3086-72 T Pesticides, Organochlorine
D 1783-70(1974) Phenolic Compounds
A. Chloroform Extraction
B. Direct Colorimetric
D 2580-68 (1974) Phenols, by Gas-Liquid Chromatography
D 3534-76 T Test for Polychlorinated Biphenyls (PCB's) in Water
Waterborne Oils
D 3415-75 T Identification
D 3325-74 T Preservation of Samples
D 3326-74 T Preparation of Sample for Identification
D 3327-74 T Analysis for Selected Elements
A. Nitrogen
B. Sulfur
C. Nickel and Vanadium
D 3328 - 74a T Petroleum Oils, Comparison of, by Gas Chromatography
D 3414-75 T Infrared Analysis
(b) GENERAL PROPERTIES

D 1067 - 70 Acidity or Alkalinity
A. Electrometric Titration
B. Electrometric or Color-Change Titration
C. Color-Comparison Titration
D. Color-Change Titration after Boiling
E. Color-Change Titration after H2O2 and Boiling
D 1291-75 Chlorine Requirement
D 2035 - 74 Coagulation-Flocculation Jar Test
D 1125-77 Conductivity, Electrical
A. Laboratory
B. Field
C. Saline Waters by Electrodeless Method
D 2776 - 72 Corrosivity, in Absence of Heat Transfer (Electrical Methods)
A. Electrical Resistance
B. Linear Polarization
D 807-52(1976) Corrosivity, Embrittlement Detector Method
D 2688 - 70 Corrosivity, in Absence of Heat Transfer (Weight Loss Methods)
A. Steam Condensate (NDHA)
B. Cooling Water (Coupon)
C. Cooling and Distribution Water (ISWS)
D. Steam Condensate (USBM)
D 1126-67(1974) Hardness
A. Gravimetric
B. Titrimetric
D 1888-67(1974) Matter, Paniculate and Dissolved
A. More than 25 ppm
B. 25 ppm or Less (Automatic Evaporation)
D 1292-65(1975) Odor
Continued.
190 MANUAL ON WATER
(b) GENERAL PROPERTIES (continued)
D 1498-76 Oxidation-Reduction Potential
D 1252-67 (1974) Oxygen Demand, Chemical (Dichromate)
D 3250 - 77 Oxygen Demand, Total, by Combustion
D 1293-65(1970) pH
Solids (see Matter, Paniculate and Dissolved)
D 1429-76 Specific Gravity
A. Pycnometer Method
B. Erlenmeyer Flask Method
D 1590 - 60 (1972) Surface Tension
D 1345 - 59 (1977), Toxicity, Freshwater Fishes
D 2037-68 (1973) Toxicity, Inhibitory, to Diatoms
D 1889-71 (1976) Turbidity
A. Jackson Candle
B. Nephelometric
C. Absolute
(c) RADIOACTIVITY
Measurement of Radioactivity
D 3085 - 75 Activity, Low-Level
*D 1943-66(1971) Alpha Particle Radioactivity
D 3084 - 75 Alpha Spectrometry
*D 1890 - 66 (1971) Beta Particle Radioactivity
*D 1690-73 Gamma Radioactivity
*D 2459 - 72 Gamma Spectrometry
*D 2470-70 (1975) Neutron-Emitting Fission Products
Specific Radkmuclides
*D 2038 - 74 Barium
*D 2577-72 Cesium
*D 2334-73 Iodine
A. Heterogeneous Exchange
B. Distillation
C. Extraction
*D 2461 - 69 (1975) Iron-59
*D 2039 - 74 Manganese
D 3357-74 T Nickel
D 2460 - 70 (1975) Radium, Radionuclides of
D 3454 - 75 T Radium-226
*D 2476-75 Tritium
E 318 - 69 (1975) Uranium, Colorimetric
D 2907 - 75 Uranium
A. Direct Fluorometric
B. Extraction
D 3315-73 T Zirconium
Heavy Water
*D 2032 - 68 (1975) Deuterium Oxide

*D 2033 - 73 Deuterium Oxide, Permanganate Consumption by Impurities in
*D 2184-72 Deuterium Oxide, Testing
(d) BACTERIOLOGICAL
Adenosine Triphosphate (ATP) Content of Microorganisms, proposed
D 3508 - 76 T Fecal Coliform Recovery, Evaluation of Membrane Filters
D 1128-60 (1974) Microorganisms, Microscopic Matter, Identification of
F 60-68 (1973) Microbiological Contaminants, Detection and Enumeration
D 932 - 72 Iron Bacteria
D 993 - 58 (1976 Sulfate-Reducing Bacteria
A. lodimetric
B. Lead Acetate Paper
(e) INORGANIC CONSTITUENTS

D 857-69(1976) Aluminum
A. Fluorometric
B. Spectrophotometric with 8-Quinolinol
C. Spectrophotometric with Ferron
D 1426-74 Ammonia Nitrogen
A. Distillation \
B. Direct Nesslerization
Arsenic by Differential Pulse Polarography, proposed
D 2972 - 74 Arsenic, Spectrophotometric with Silver Diethyldithiocarbamate
D 513-71 Bicarbonate and Carbonate, Carbon Dioxide
A. Precise CO: Evolution
B. Abridged CO2 Evolution
C. Bicarbonate Titration
D. Differential Titration
E. Direct Titration of Free CO2
D 3082-74 Boron
A. Carminic Acid Colorimetric
B. Mannitol Potentiometric
D 1246-77 Bromide and Iodide
A. Iodide, Colorimetric-Catalytic
B. Bromide, Colorimetric-Catalytic
C. Iodide and Bromide, Titrimetric
D 2576-70 (1976) Cadmium by Atomic Absorption Spectrophotometry
Cadmium and Lead by Differential Pulse Anodic Stripping
Voltammetry, proposed j
D 2576-70 (1976) Calcium by Atomic Absorption Spectrophotometry
D 3557 - 77 Cadmium by Atomic Absorption Spectrophotometry
A. Atomic Absorption, Direct
B. Atomic Absorption, Chelation Extraction
D 511-76 Calcium and Magnesium
A. Gravimetric
B. Complexometric Titration
C. Atomic Absorption Spectrophotometric
D 1126- 67 (1974) Calcium and Magnesium Hardness
A. Gravimetric
B. Titrimetric
D 513-71 Carbon Dioxide, Bicarbonate, Carbonate
A. Precise CO2 Evolution
B. Abridged CO2 Evolution
C. Bicarbonate Titration
D. Differential Titration
E. Direct Titration of Free COa
Continued.
192 MANUAL ON WATER
(e) INORGANIC CONSTITUENTS (continued)
D 512^67(1974) Chloride
A. Mercurimetric Titration
B. Silver Nitrate Titration
C. Colorimetric with Ferric Thiocyanate
D 1253-76 Chlorine, Residual, in Water
A. Direct Amperometric Titration
B. Starch-Iodide Titration (Direct and Back Titration)
C. Amperometric Back Titration
D 1427 - 68 (1974) Chlorine, Residual, in Waste Water
A. Starch-Iodide Titration
B. Amperometric Titration
D 1687 - 77 Chromium, Total
A. Permanganate Oxidation
B. Hypobromite Oxidation
C. Alkaline-Persulfate Oxidation
D. Atomic Absorption
D 2576-70 (1976} Chromium by Atomic Absorption Spectrophptometry
D 2576-70 (1976> Cobalt by Atomic Absorption Spectrophotometry
D 3558 - 77 Cobalt by Atomic Absorption Spectrophotometry
D 1688-77 Copper
A. Colorimetric with Neocuproine (High Range)
B. Colorimetric with Neocuproine (Low Range)
C. Colorimetric with Cuprethol
D. Atomic Absorption, Direct
E. Atomic Absorption, Chelation Extraction
D 2576-70 (1976) Copper by Atomic Absorption Spectrophotometry
F 57-68(1973) Copper in High-Purity Water
D 1179-72 Fluoride
A. Photometric SPADNS
B. Ion Selective Probe
D 1385 - 67 (1972) Hydrazine, Colorimetric with p-Dimethylaminobenzaldehyde
D 1588-60 (1974) Hydrogen, Dissolved and Gaseous
A. Thermal Conductivity
B. Volumetric
D 514 - 67 (1974) Hydroxide, Titrimetric
D 1246-77 Iodide and Bromide
A. Iodide, Chlorimetric-Catalytic
B. Bromide, Colorimetric-Catalytic
C. Iodide and Bromide, Volumetric
D 1068-77 Iron
A. Photometric, Orthophenanthroline
B. Photometric, Bathophenanthroline
C. Atomic Absorption, Direct
D. Atomic Absorption, Chelation Extraction
D 2576-70 (1976) Iron by Atomic Absorption Spectrophotometry
D 2576-70(1976) Lead by Atomic Absorption Spectrophotometry
D 3559 - 77 Lead by Atomic Absorption Spectrophotometry
Lead and Cadmium by Atomic Spectrophotometry, proposed
D 3561 - 77 Lithium, Potassium, and Sodium in Brackish Water, Seawater, and
Brines by Atomic Absorption Spectrophotometry
D 2576-70 (1976) Magnesium by Atomic Absorption Spectrophotometry
D 511-76 Magnesium and Calcium
A. Gravimetric
B. Complexometric Titration
C. Atomic Absorption Spectrophotometric
D 858 - 77 Manganese, Colorimetric as Permanganate
A. Colorimetric
B. Atomic Absorption, Direct
C. Atomic Absorption, Chelation Extraction
D 2576-70(1976) Manganese by Atomic Absorption Spectrophotometry
D 3223 - 73 Mercury by Cold Vapor Atomic Absorption Spectrophotometry
D 2576-70 (1976) Metals by Atomic Absorption Spectrophotometry
D 2576-70 (1976) Metals by Atomic Absorption Spectrophotometry
A. Cadmium, 0.05 to 5 mg/ litre
B. Calcium, 0.3 to 15 mg/litre
C. Chromium, 0.2 to 20 mg/litre
D. Cobalt, 0.25 to 50 mg/ litre
E. Copper, 0.1 to 20 mg/litre
F. Iron, 0.1 to 20 mg/litre
G. Lead, 0.34 to 40 mg/litre
H. Magnesium, 0.05 to 3.5 ing/lit re
I. Manganese, 0.04 to 10 mg/litre
J. Nickel, 0.08 to 20 mg/ litre
K. Zinc, 0.02 to 3 mg/litre
D 3372 - 75 Molybdenum by Atomic Absorption Spectrophotometry
D 1886-77 Nickel
A. Photometric, Carbamate
B. Photometric, Glyoxime
D 2576-70(1976) Nickel by Atomic Absorption Spectrophotometry
D 992-71 Nitrate, Colorimetric with Brucine-Sulfanilic Acid
D 1254-67(1974) Nitrate
A. Colorimetric by Diazotization
B. Volumetric with Permanganate
D 1589-60(1974) Oxygen, Dissolved in Waste Water
A. Alsterberg (Azide)
B. Rideal-Stewart
C. Pomeroy-Kirshman-Alsterberg
D. Polarographic
D 888-66(1977) Oxygen, Dissolved in Water
A. Colorimetric-Indigo Carmine
B. Dead-Stop End Point
C. Potentiometric End Point
D. Thiosulfate Titration
E. Thiosulfate Titration-Starch Indicator
D 515-72 Phosphorus
A. Colorimetric, Ascorbic Acid Reduction
B. Colorimetric, Amino Reduction
C. Colorimetric, Molybdovanadophosphate
D 3561 - 77 Potassium, Sodium, and Lithium in Brackish Water, Seawater, and
D 1428-64(1971) Potassium and Sodium by Flame Photometry
A. Analysis of Water
B. Trace Amounts of Sodium in Low-Solids Water
C. Analysis of Water-Formed Deposits
D 3081-72 T Selenium, Colorimetric with Diaminobenzidine
Continued.
194 MANUAL ON WATER
D 859-68 (1974) Silica
A. Gravimetric
B. Colorimetric as Molybdenum Blue
C. Colorimetric as Molybdenum Yellow or Blue
D. Colorimetric, Sulfite Reduction
Spectrochemical Analysis by Rotating Desk Technique Using An
Optical Emission Spectrometer, proposed
D 3561 - 77 Sodium, Potassium, and Lithium in Brackish Water, Seawater, and
D 1428-64(1971) Sodium and Potassium by Flame Photometry
A. Analysis of Water
B. Trace Amounts of Sodium in Low-Solids Water
C. Analysis of Water-Formed Deposits
D 2791 - 77 Sodium, Continuous Determination by Ion Selective Electrode
A. Sodium Ion-Electrode
B. Flame Photometry
D 3352 - 74 Strontium, in Brackish Water, Seawater, and Brines
D 516-68(1974) Sulfate
A. Gravimetric
B. Turbidimetric
C. Volumetric
D 1339-72 Sulfite, Volumetric
A. 0.1 to 6 mg/litre NaaSOs
B. 6 mg/litre Na2SO3 or higher
C. 3 mg/litre Na2SO3 or higher
D 2333-68 (1974) Thorium, Colorimetric with l-(o-Arsonophenylazo)-2-Naphthol-3,6-
Disulfonic acid
D 2907-75 Uranium
A. Direct Fluorometric
B. Extraction
D 3373 - 75 Vanadium, Colorimetric-Catalytic
D 1691-77 Zinc
A. Colorimetric with Zincon, High Range
B. Colorimetric with Zincon, Low Range
D 2576-70 (1976) Zinc by Atomic Absorption Spectrophotometry
Analysis of Organic Constituents in Water

The ever-increasing number of organic constituents which have been
detected in water supplies has become a significant concern and problem in
determining water quality. The usual methods of analysis and control are
not sensitive enough to monitor the concentrations of these organic con-
taminants. Such substances as insecticides, herbicides, and other organics in
the micrograms-per-litre concentration range may impair water quality for
several important intended uses. Drinking water use may be impaired by
deteriorated taste and odor of the water. Toxicity to aquatic organisms,
including food-fish, may destroy or seriously limit this natural-resource use.
However, the long-term effects of such organic contaminants are largely
unknown.
New techniques which are required for analysis are being developed and
standardized rapidly. Chromatography methods are being used to separate
the constituents for identification by infrared, ultraviolet, and mass spectro-
scopy. Solvent extraction and activated carbon adsorption have been used to
separate and concentrate the organic substances from water. Thin-layer and
paper chromatography are helpful in separating the organics that cannot be
separated by gas-liquid chromatography. Identification of specific organic
compounds is limited to the compounds for which there are pure standards
to compare elution times in gas and liquid chromatography. Even these
identifications must be considered only tentative until confirmed by at least
two other columns or, in liquid chromatography, two different developing
conditions. It is advisable to further confirm the organic structure by
spectroscopy. Fluorescent spectroscopy has been used as a detector with
paper and thin-layer chromatography. The chromatographed spots are
caused to fluoresce by ultraviolet light excitation, and the spectrum of the
fluorescent emission is used to identify the compound. The intensity of the
fluorescence can be related to the concentration of the compound present in
the spot.
Analysis of organic contaminants in water in nanogram- and picogram-per-
litre concentrations is receiving much attention, and new techniques are
being developed. Table 4a shows the methods published in the Annual Book
of ASTM Standards, Part 31, for organic constituents of water and waste
water.
Chemical Oxygen Demand

Chemical oxygen demand (COD) is defined as the amount of oxygen,
expressed in milligrams per litre, consumed under specified conditions in the
oxidation of organic and oxidizable inorganic matter in waste water,
corrected for the influence of chlorides. In itself, the chemical oxygen
demand is not a reliable index as to the quality or quantity of pollution, yet
its determination can give worthwhile information when compared with
previous results for the same water, or when considered in the light of other
knowledge about the constituents of the water. Not all organic substances
are oxidized completely or at the same rate; therefore, use of precisely
reproducible techniques is important when comparisons are needed.
ASTM Method D 1252 uses a standard potassium dichromate solution in
50 volume percent sulfuric acid to oxidize organic and oxidizable inorganic
material. The excess dichromate is then titrated with a standard ferrous
ammonium sulfate solution, using o-phenanthroline ferrous complex as an
internal indicator. Silver sulfate can be added as a catalyst to facilitate the
oxidation of straight chain organic acids and alcohols.
Other oxidizing agents such as potassium permanganate and potassium
iodate have been used to determine COD, but the dichromate method is
more reproducible and generally superior. On relatively pure waters,
196 MANUAL ON WATER
however, the permanganate COD may be preferable because much more

dilute solutions can be used.
General Properties of Water

Properties of water not usually measured in the laboratory are not
included in Table 4b. Of these, temperature is probably the most important.
Acidity and Basicity (Alkalinity) [2]2

Acidity and basicity (alkalinity) are defined as the quantitative capacity of
aqueous media to react with hydroxyl or hydrogen ions, respectively, as
distanct from pH, which is an intensity factor. Acidity or basicity in water is
the result of dissociation or hydrolysis of its solutes. The strength of the
acidity or basicity gives a good indication of the potential corrosiveness of
the water and also furnishes a further guide in choosing appropriate
treatment of either raw water or plant effluents.
Alkalinity in water may be reported in terms of equivalent titratable
amounts of bicarbonate, carbonate, or hydroxide ions, or in terms of
equivalent amounts of calcium carbonate. The term alkalinity asxused in
water analyses is not in full accord with generally accepted chemical
terminology. In the discussion of pH, 7.0 is considered to be the neutral
point. In water analysis about 4.5 is the end point for titration of alkalinity.
Thus, a water with a pH of 6 could be called acid on the basis of the pH, but
might still have titratable alkalinity. Some water chemists prefer to avoid the
general term alkalinity, and speak of carbonate, bicarbonate, and hydroxide
instead. The determination of these ions, however, is an acid-base titration
and does not differentiate the ions that enter the reaction.
Sources of Alkalinity
The anions that may contribute to alkalinity are those which form acids
that are only weakly dissociated in solution, and which thus enter into
hydrolysis reactions. Chloride, sulfate, and nitrate ions do not affect the
alkalinity. Cations that form weakly dissociated bases take part in hydrolysis
reactions which tend to produce acidity rather than alkalinity. The presence
of certain organic materials may have a considerable effect upon the
alkalinity determination.
Because of the relative abundance of carbonate minerals and because
carbon dioxide, which enters into equilibria with them in water solution, is
readily available, bicarbonate and carbonate are to be expected in most
waters. The presence of hydroxide ions in natural water in amounts
sufficient to affect the alkalinity determination directly is very rare, unless
artificial contamination has occurred.
2
Much of the following discussion is taken from Ref 2.
Hydroxide may occur in water that has been softened by the lime or lime-
soda process. Water in contact with fresh concrete may have an excessively
high pH and apparent hydroxide alkalinity, due in part to silicates. Concrete
may be a source of hydroxide alkalinity in samples from wells that are under
construction.
Under any system of reporting titrated alkalinity now in use, all the effects
of the anions entering into hydrolysis reactions are lumped together and
reported as an equivalent amount of a single substance, or as two more or
less postulated ions. Thus, although alkalinity data may suggest the presence
of definite amounts of carbonate, bicarbonate, or hydroxide, the ions have
not been directly determined as such and the results may include the
equivalent of all or part of such other anions as may tend to hydrolize. These
ions include silicate, phosphate, borate, and possibly fluoride. Some other
ions not ordinarily found in natural water, such as arsenate, aluminate, and
certain organic anions which may occur in colored waters, also could
increase the alkalinity.
Chemistry of Alkalinity Determination

The alkalinity of water is determined in the laboratory by titration with
standard acid to a definite pH. In the titration, conversion of dissociated
carbonate ions (CO3~) to undissociated carbonic acid takes place in two
steps. In the first step, the carbonate ions in solution each take up one
hydrogen ion and become bicarbonate ions (HCOa"). The hydrogen ions
needed are supplied by the acid added in the titration. As the titration
proceeds, the amount of unchanged carbonate decreases until the carbonate
ions are nearly changed over. At this point in the titration, a small addition
of acid will produce a comparatively large reduction in pH, indicating the
essential completion of the first step. This end point usually is considered to
be at a pH of 8.2 and can be observed either with a pH measuring device or
an indicator that changes color in this range, such as phenolphthalein.
Theoretically, at this pH level, the ratio of HCOs" ions to COs" ions is more
than 100 to 1, as computed from dissociation constants for water and
carbonic acid.
In the second step, each of the bicarbonate ions takes up one hydrogen ion
to form molecules of undissociated carbonic acid. This requires additional
quantities of acid. When nearly all the HCO3~ ions have been taken care of,
the addition of a small amount of acid again causes a large decrease in pH.
At a pH of 4.5, the ratio of H2CO3 molecules to HCOj" ions is theoretically
more than 100 to 1, and this is chosen as the end point. It can be detected by
a pH measuring device or by a suitable indicator such as methyl orange or
methyl red.
In the interpretation of the titration, it is assumed that all alkalinity
present in the water was in the forms of HCOs", CDs', and OH~. If no OH"
is present, the acid required to complete Step 1 is a measure of the COs"
198 MANUAL ON WATER
content. The amount required for Step 2, when corrected for any HCOy
produced in the carrying out of Step 1, is a measure of the original HCOjT
content. Waters containing OH~ require more acid for Step 1 than for Step 2
and the magnitude of this difference is a measure of the OH" concentration.
It is assumed throughout that OH~ and HCO.r cannot exist in the same
solution.
. . No matter what ions may be involved, if they are titrated stoichio-
flietrically by lowering the pH to 4.5, alkalinity values in equivalents per
million (epm) which give a cation-anion balance are obtained. Silicate in the
form SiO>~ should not be present in waters in measurable amounts unless
the pH is at least 11. The HSiO.r form, however, might represent about half
the silica present at a pH of 9 and about a tenth of it at pH 8.
Dissociated silica can contribute to titrated alkalinity between pH values
of 8 and 9, but the magnitude of the effect in this range is uncertain. Silicate
ions in either of the foregoing forms would be fully titrated in the regular
alkalinity determinations. They would then be reported in terms of equiva-
lent quantities of bicarbonate or carbonate. Other types of silicate ions also
might affect the alkalinity.
Orthophosphate may occur in water in three forms, PO4=, HPO4=, and
H2PO4~. The trivalent form is present in appreciable amounts only above a
pH of 10. The HPO4= ion predominates from a pH of 7 to a pH of 2. The
phosphate ions are not fully converted to orthophosphoric acid even at a pH
of 1, but practically all the HPO4= ions would be converted to H2PO4" ions
in an alkalinity titration stopped at a pH of 4.5. Waters that contain enough
phosphate to affect the alkalinity appreciably are rare in nature. If such a
water were titrated for alkalinity in the usual way, a part of the phosphate
(about a third) would appear as an equivalent amount of bicarbonate or
carbonate.
Borate ions also hydrolyze in water to affect the pH. Theoretically these
ions should all be converted to boric acid in the regular alkalinity titration.
The form of such ions in solution in natural waters is not certain, however,
and the effect of berates on alkalinity titration is difficult to evaluate fully.
Waters are seldom found in which measurable alkalinity is contributed by
borate ions, because boron concentrations in water are normally very low.
Borate in the dihydrogen form (H2BO3~) is theoretically possible in the ratio
of 1 part to about 17 parts of the undissociated acid at a pH of 8. At a pH of
7 and lower, dissociated ions are essentially absent.
Fluoride ions hydrolyze, but weakly, and probably they are rarely present
in large enough amounts to affect the alkalinity.
The values given for dissociation constants in the literature hold for dilute
solutions of single salts. Natural waters are not so simple, and exact
conformance with theoretical behavior, therefore, is not to be expected. The
computations, however, do give useful indications of the behavior of various
ions that can take part in hydrolysis and contribute to alkalinity.
Relation to pH
Many users of chemical analyses are confused by the distinction between
pH value and titratable alkalinity (because the titration for alkalinity is
carried to a pH of 4.5, whereas at the same time a pH of 7 is stated to be the
neutral point).
If all the hydrolysis in a water is assume^ to be due to the system CO3 =
-2LLHCO, JtLHzCOj; the pH provides a means of deciding how far to the left
the reaction is displaced. A pH below 8.2 indicates the absence of the normal
carbonate phase and a pH below 4.5 indicates the absence of both the
normal carbonate and bicarbonate phases, but otherwise the pH is only
qualitative. By titration of the system with acid, a quantitative determination
of the amount of each phase can be made, if the original assumption as to the
ions present is correct. To complete this determination the equilibrium must
be shifted all the way to the right and the pH lowered to 4.5. In pure water at
neutrality the pH is 7, but this represents an absence of ions that would take
part in hydrolysis and is an entirely different condition.
Range of Concentration
Alkalinity reported as hydroxide is ordinarily absent from uncontam-
inated natural water. Alkalinities reported as carbonate are common,
particularly in groundwaters, or concentrated surface waters that have a
large amount of sodium in proportion to calcium and magnesium. The usual
stream water has no apparent carbonate alkalinity and should have a pH
below 8.2 if it contains an appreciable amount of calcium. Some surface
waters have a pH greater than 8.2, either from CO?= or as a result of silicates
or other anions in solution.
Alkalinity reported as carbonate usually is present in small amounts,
generally less than 10 mg/litre. In waters high in sodium, higher values are
sometimes encountered, but concentrations over 50 mg/ litre are extremely
unusual.
In acid waters, there is no bicarbonate alkalinity. In alkaline waters (all
those whose pH is 4.5 or over), the range in alkalinity reported as
bicarbonate is from zero up to more than 1000 mg/ litre, although concentra-
tions much higher than 500 mg/litre are unusual. Waters highly charged
with carbon dioxide may contain large amounts of alkalinity as bicarbonate.
The reduction of sulfate to sulfide with the assistance of certain types of
bacteria produces carbon dioxide as a by-product, and may cause high
bicarbonate concentrations in waters associated with petroleum.
Acidity
As a corollary to the property of alkalinity, there is the opposite property
of acidity which is found in some natural waters. If the usual alkalinity end
200 MANUAL ON WATER
point is used, a water would not be reported as having acidity unless its pH is
below 4.5.
Sources of Natural Water Acidity

Free acids such as HC1 or gases like SC>2 combine with water to form acid.
Certain hot springs yield strongly acid water that has dissolved these
substances. Free acids may also be added to streams or groundwaters
through waste disposal, or through oxidation of sulfides. The latter source is
an important one in mining districts. Drainage from mines in areas where
pyrite or other metallic sulfides occur in the ore, or in associated strata, may
be strongly acid, containing not only free sulfuric acid but also iron and
aluminum. This effect is especially noticeable in coal mining areas.
The oxidation of pyrite in air, which may be aided by some types of
.bacteria, produces ferrous sulfate and sulfuric acid. The ferrous sulfate may
be further oxidized to ferric sulfate, and eventually precipitates as ferric
hydroxide. The hydroxide ions thus are removed from solution and the
water becomes more acid through the process of hydrolysis. Aluminum
cations in water also hydrolyze to give an acid reaction. A considerable
oxygen supply is required for oxidation of pyrite, and ordinarily it would not
take place very far below the water table, or in other situations where oxygen
is deficient. Ferrous and ferric salts are added to surface streams in some
types of industrial wastes.
Ammonium ions occasionally occur in natural waters. These ions are
capable of hydrolysis, contributing to the acidity of samples.
Chemistry of Acidity Determination

In those waters which have a pH below 4.5, the presence of acidity is
reported. The acidity may be the result of actual free acids such as sulfuric or
hydrochloric in solution. It can also be the result of the hydrolysis of certain
cations in solution. The free-mineral-acid determination usually is ac-
complished by titration with standard alkali to a pH of about 4.5 in the cold
solution. An approximation of acidity can be obtained also by means of a
calculation based on the pH. These two procedures give meaningful results
only in the absence or near absence of iron and any other cations that might
cause hydrolysis.
Total acidity generally has been determined by titration to the phenol-
phthalein end point (pH 8.2) at the boiling temperature. This supposedly
gives the net effect of hydrolysis as well as the free-acid content. Actually the
end point is somewhat arbitrary and may not be stoichiometric for all forms
of iron or other cations involved in hydrolysis. No method for accurate
determination of free acid in the presence of salts which hydrolyze to lower
pH has yet been described, and lack of cation-anion balance is to be
expected in waters where these conditions exist.
Range of Concentration
Acidity can range from zero at the pH level that is taken as neutrality to
several hundred milligrams per litre for pH levels around 2.5. Springs and
waters strongly affected by mine drainage or industrial wastes may at times
have lower pH values. The total acidity has some significance in terms of
treatment needed, but is of less value in the representation of actual
conditions in the water than reporting of alkalinity as carbonate and
bicarbonate is for alkaline water.
Methods
In ASTM Methods D 1067, the high-precision procedure involves elec-
trometric titration with 0.02 N hydrochloric acid or sodium hydroxide either
to specific pH end points, or by developing a titration curve for the sample
and noting the pH values at the inflection points of the curve. The latter
procedure is used for evaluation of the buffering capacity of water if the
buffering salts are other than carbonates and bicarbonates.
Three routine procedures were included in Methods D 1067. Method A
involves titration to a predesignated pH, the end point being determined
electrometrically or by color change of an internal indicator. In Method B,
the end point is determined by comparison of the color developed by an
added indicator with the color of a standard buffer solution containing the
same added indicator. In routine Method C, particularly applicable to acid
waste waters, the sample is boiled to attain equilibrium, then titrated to a
designated pH, the end point being determined by color change of an
internal indicator. The latter procedure is applicable for waters highly
polluted with acid industrial wastes, containing slowly hydrolyzable or
buffering materials, or materials that will interfere by reason of color,
precipitation, and other factors.
Alkalinity or acidity in the intermediate range between pH 3.5 and pH 9 is
arbitrarily regarded as resulting from the presence of weak bases or weak
acids. Alkalinity above this range, or acidity below it, is considered to be due
to the presence of relatively strong bases or strong acids, respectively.
Special procedures said to be sensitive and accurate to 0.05 mg/litre have
been developed for measuring extremely low alkalinities [3].
Accuracy and Reproducibility of Results

It is evident from the foregoing that for some waters, at least, alkalinity
data are largely empirical. Relative amounts of the different forms of
alkalinity reported by analyses should not always be considered significant,
as changes can occur from the time the sample is collected until the bottle is
opened and the alkalinity titration is made. The total alkalinity of a sample
should be a reproducible figure, provided that similar conditions of sample
treatment are followed. Agreement much closer than 2 to 5 percent,
202 MANUAL ON WATER
however, cannot be expected on duplicate samples, and differences this size

should not be considered significant in making interpretations.
Because of the uncertainty regarding chemistry of acid waters, and their
tendency to change in storage, the data on acidity also should be interpreted
cautiously.
Chlorine Requirement
Often it is desirable to establish the quantity of chlorine required to
achieve a specific objective in the treatment of a water by chlorination. This
objective might be elimination of tastes and odors; reduction of biochemical
oxygen demand by destroying or modifying decomposable organic sub-
stances; separation of grease in waste water; or destruction or modification
of oxidizable constituents in waste water.
ASTM Method D 1291 provides a procedure for finding the chlorine
requirement of process water and waste water. It involves adding a chlorine
solution of known strength, in increasing increments of chlorine concen-
trations, to different portions of the sample water. The contact time and pH
are maintained as specified for the particular object of chlorination. The
chlorine requirement of the water is found by interpolation of a plotted
curve of chlorine dosages against the results of chlorine residual tests or
other tests which determine whether the specific object of the chlorination
has been reached in the different sample portions. The chlorine requirement
is the amount of chlorine, expressed in parts per million, required to achieve
the objectives of chlorination.
Color
Color is a common constituent of many natural waters. It generally results
from natural metallic ions, and from lignins, humic acids, and other
vegetation products thought to be in colloidal suspension. In drinking water,
color is objectionable for aesthetic reasons, but in several manufacturing
processes, virtually colorless waters are required. Occasionally, pollution by
industrial wastes will impart unusual or unnatural colors to water supplies.
Artificial color scales, using standards prepared from potassium chloro-
platinate and cobaltous chloride solutions, are often used to measure natural
color in water [4\. Suspended matter must be removed by centrifuging before
the comparisons are made. Addition of calcium chloride [5] to assist in
centrifuging is desirable with some types of turbidity. Spectrophotometric
and photometric methods have been developed [6], some especially for
industrial waste water \4}.
Continuous measurement of color has been practiced to a very limited
extent. Methods for measuring color, with particular emphasis on photo-
metric methods which offer greatest promise for continuous analysis, are
discussed by Staats [7].
Electrical Conductivity
Electrical conductivity has special significance because it is a quick and
convenient method for measuring electrolyte concentration. The specific
conductance of a dilute solution of an electrolyte is almost directly propor-
tional to the ionic concentration of the electrolyte, and the total conductivity
of a given water is equal to the sum of the several conductivities resulting
from the electrolytes present. Because of the unusually high conductance per
weight unit of the hydrogen or hydroxide ion, samples frequently are
neutralized before the electrical conductivity is measured. Conductivity
measurements can therefore be related to dissolved solids concentration.
Also, they sometimes are required for special investigations of corrosion in
water.
ASTM Method D 1125 describes the apparatus and procedures for
determining electrical conductivity of industrial waters. The high-precision
procedure included requires careful control of temperature, all measure-
ments being made at 25° C. The procedure designed for more routine use
incorporates a temperature correction chart.
Conductivity is a useful guide to the purity of distilled water and is much
used for evaluation of steam purity in the power industry. It is important to
note, however, that dissolved gases (particularly carbon dioxide and am-
monia) have an important effect on conductivity. When determining steam
purity by this method, it is generally necessary to reduce the concentration of
these gases and to correct for the residuals present [8,9]. Special apparatus is
available for removing gases from the water before the conductivity
determination is made [10,11]. A good approximation of the dissolved solids
in distilled water and condensate can be made by boiling a sample down to
less than half the original volume to drive off the dissolved gases, deter-
mining the conductivity of the cooled remainder, and by correcting this
value for the reduction in volume [72,73].
Electrical conductivity cannot be precisely related to dissolved solids
because some dissolved substances such as silica and certain types of organic
matter contribute little or nothing to the conductivity.
Continuous measurement of conductivity of flowing samples is often used
for monitoring process waters [14], and is essential in monitoring quality of
high-purity supplies, such as demineralized water.
Hardness
Hardness is now generally accepted as representing the total concentration
of calcium and magnesium ions. Originally hardness was understood to be
the capacity of a water for precipitating soap. It was measured by the
amount of a standard soap solution required to produce a stable lather.
Therefore, other polyvalent ions (in addition to calcium and magnesium)
capable of precipitating the soap were included as part of the hardness.
204 MANUAL ON WATER
Hardness has traditionally been classified as two types: temporary hard-

ness and permanent hardness. That portion of hardness which disappears
upon boiling is referred to as temporary hardness, and is essentially the
bicarbonates of calcium and magnesium which are precipitated as car-
bonates by the loss of carbon dioxide in heating. The hardness remaining
after boiling is permanent hardness, including sulfates, chlorides, and
nitrates of calcium and magnesium. Most often the total hardness concen-
tration is the primary interest. Most analytical methods do not differentiate
between the foregoing hardness classifications, but determine only total
hardness, with provisions for determining the individual calcium and
magnesium concentrations. The various properties and effects of hardness in
industrial water are discussed under Calcium and Magnesium in this
chapter.
ASTM Methods D 1126 provide two methods for determining total
hardness. The more precise procedure is a gravimetric method that is
cumbersome but dependable with waters of unknown composition. Calcium
is determined by precipitation as the oxalate, which is then ignited to
calcium oxide. Magnesium is determined by the precipitation of magnesium
ammonium orthophosphate, which is then ignited to magnesium pyrophos-
phate. Total hardness is calculated from the sum of calcium and magnesium
in the ignited residues.
The more rapid method is a volumetric titration in which a buffered water
sample is titrated with an organic sequestering agent, sodium ethylenedia-
mine tetraacetate, in the presence of an indicator dye. By titrating a second
aliquot of the sample in the presence of a different indicator and buffer, a
separate determination is made for calcium, thus differentiating between the
calcium and magnesium. If interferences to not exceed the specified limits,
the volumetric method is as accurate as the gravimetric method.
Odor
Odor is a characteristic of many natural waters and can be the result of
many contributing factors. Among these are pollution by sewage and trade
wastes, the presence of living matter (bacteria, fungi, algae, and animal life),
the decomposition of various forms of living matter, and the presence of
colloidal vegetable or earthy matter.
Odor tests can be useful in many respects. For example, they may be used
in checking the quality of raw and treated waters, in determining the
effectiveness of deodorizing procedures, and as a method of tracking sources
of contamination or leaks in industrial processes. Odor is difficult to
measure quantitatively.
ASTM Method D 1292 is a method for measuring odor in waters for
comparative or control purposes. The threshold odor number of water is
determined by using the sense of smell and a series of dilutions of the sample
with odor-free water. The odor threshold of a sample represents the number
of volumes of dilution water required to yield the minimum definitely

perceptible odor. Psychological influences are eliminated as much as
possible by having one person dilute the sample and label the flasks in code,
and other persons conduct the odor tests, examining dilutions in random
order. The average results of two or more odor testers is suggested. ASTM
Method D 1292 also includes a table for classifying odors by chemical types.
Chapter 12 of this manual presents a comprehensive discussion on odor
testing and other sensory examination methods.
Oxidation-Reduction Potential
The oxidation-reduction potential of a solution is defined as the elec-
tromotive force (emf) developed by a platinum electrode immersed in the
water, referred to the standard hydrogen electrode. Every oxidation-
reduction reaction involves an exchange of electrons; the emf necessary to
oppose this flow of electrons gives a measure of the oxidation-reduction
potential. ASTM Method D 1498 describes the apparatus and procedure for
measuring the oxidation-reduction potential of all types of industrial waters.
It does not deal with the interpretation or application of the results.
Empirical testing is necessary to adapt this method of measurement to
specific ions in a particular application.
Some of the applications of oxidation reduction potential measurements
in water have been summarized by Rosenthal. In one stage of the seawater
bromine process, the concentration of free bromine in acidified seawater
treated with chlorine is obtained by measuring the oxidation-reduction
potential. A similar measurement on a plating plant waste is used as a gage
of dichromate ion concentration for controlling the addition of a reducing
agent. Continuous measurement of low chlorine concentrations is possible
by this method.
Specific Gravity
Specific gravity is the ratio of weight of a given volume of sample to the
weight of an equal volume of water, under a fixed set of conditions. It is
rarely an important property of industrial supply or process waters, but may
be significant in waste analysis. It can be used to estimate the concentration
of certain solutions used in industrial processes such as caustic or brines, and
the concentration of dissolved solids in waters by evaporation.
ASTM Method D 1429 covers an accurate gravimetric determination
using a specific-gravity bottle (pycnometer) in which a known volume of
water is weighed at a temperature of 15.5°C (60° F). This method has an
accuracy of ±0.0005. A less-precise method is included for determining the
specific gravity of sludges and muds. This is done by weighing an Erlenmeyer
flask filled first with water and then with the sample, and comparing the two
net weights.
206 MANUAL ON WATER
Specific gravity also has been determined to the fourth decimal place with
a balance of the Westphal type which measures buoyancy of the water on a
totally immersed plummet. Specific gravity of concentrated brines and
caustic solutions can be measured with appropriate hydrometers with
sufficient accuracy for control purposes. A quick approximation of the
specific gravity of any water solution can be made by weighing a pipetted
sample and comparing it with the weight of the same volume of distilled
water at the working temperature.
Surface Tension
Surface tension is a little-used property in the investigation and control
analyses of industrial waters. Occasionally, it is used to indicate contamina-
tion by some water-soluble organic solvent or as an indication of the
presence of surfactants. This can be helpful, for example, in locating a
possible cause of boiler water foaming.
ASTM Method D 1590 uses a ring tensiometer to measure the surface
tension of a liquid. It measures the amount of force required to lift a
standard platinum-indium wire ring from the surface of the sample. The
method does not disclose any direct relationship between the surface tension
and the concentration of surfactant solutions.
A rough indication of surface tension can be obtained with a slender glass-
stoppered glass tube which is filled to a standard mark with the sample. The
tube is stoppered and placed in a horizontal position, and the length of the
air bubble is compared with the length of the air bubble when distilled water
is used in place of the sample. With samples of lowered surface tension, the
bubble is longer than that obtained with distilled water. The tube is
calibrated by filling to the standard mark with distilled water, stoppering,
placing a horizontal position, and marking the tube at the two ends of the air
bubble.
Suspended and Dissolved Matter

Suspended solids in supply and process waters are an important con-
sideration in water treatment. They add to deposition and fouling problems,
and in high-velocity water contribute to the erosion of metallic surfaces. The
determination of suspended solids often is used as a guide to the satisfactory
operation of plant processes and to the final disposition of industrial wastes.
The amount of dissolved solids present in an industrial water is one of its
more important properties, since it is one of the factors that determine use
and reuse of the water. In boiler wasters, there is a dissolved solids limit
above which carry-over and steam contamination sharply increase. The test
is useful in complete analysis because the result compared with the sum of
individual constituents can be used as a check on the analysis.
Methods for determining suspended and dissolved matter in water are
presented in ASTM Method D 1888. Method B of D 1888 is intended for

waters containing less than 25 mg/ litre solids. In Method B, more than 4
litres of sample are required and the evaporation is done slowly, with special
precautions to exclude contaminants. This method requires special equip-
ment, including an automatic evaporation feature which may not be
available in most laboratories. As an alternative procedure, it is possible to
obtain an approximate measure of dissolved solids in high-purity waters by
measuring the electrical conductivity (see discussion under Electrical Con-
ductivity).
Method A of D 1888 is applicable for waters with a dissolved solids
content of 25 mg/ litre or more. It involves evaporating sufficient sample to
yield at least 25 mg of residue, which is dried and weighed. Provisions also
are made for determining volatile solids and fixed solids.
Turbidity
Turbidity in waters is due to the presence of suspended solids dispersed
throughout the water, resulting in a reduction of its transparency. Measure-
ment of turbidity is therefore a measure of opacity or light-scattering by
suspended solids, and is reported in arbitrary units. ASTM Method D 1889
covers the determination of the turbidity in industrial water and industrial
waste water. Three procedures are presented, the Jackson candle, nephelo-
metric, and absolute turbidity methods.
A Jackson candle turbidimeter consists of a graduated glass tube, a
standard candle, and a support for the candle and tube. The water being
tested is poured into the tube until the image of the candle flame just
disappears. The depth of water at this point is related to'turbidity values by
standardization of the apparatus; usually it is converted directly into
turbidity values by reference to tables supplied with the instrument. Waters
of high turbidity, more than Jackson turbidity number 1000, are diluted
before the test is made.
When turbidity is lower than Jackson turbidity number 25, it is desirable
to use a relative turbidity method such as nephelometry. This is based on the
light-scattering properties (Tyndall effect) of the suspended solids. Nephelo-
meters are available commercially, and most of these instruments measure
light scattered at approximately 90 deg, or at 90 and 270 deg. Turbidity
standards have been prepared by thoroughly mixing suitable amounts of
finely divided titanium dioxide into partially polymerized polystyrene.
Calibrated commercial standards in sealed tubes can be purchased. Results
are reported in terms of turbidity units.
Neither of the preceding methods measures absolute turbidity. Instru-
ments are commercially available for measuring absolute turbidity and are
based on an integrating sphere which sums up most of the source light that
has been scattered by the sample. This method is not subject to particle size
limitations. Absolute turbidity is defined as the fractional decrease of
208 MANUAL ON WATER
incident monochromatic light intensity, due to scattering, in traversing 1 cm

of fluid.
In some industrial research and control applications, it is desirable to
know the dissymmetry of the light-scattering at various angles in order to
determine the average linear dimensions of the suspended particles. A
goniophotometer can be used to obtain the geometric light-scattering
pattern for this purpose.
Radioactivity (Table 4c)

Radioactivity measurements are becoming increasingly important as the
"Nuclear Age** progresses. Instrumental methods are used which may
indicate only the intensity of radiation for safety purposes, or which may
provide information on the nature of the radiation, the intensity, and the
half-life to permit quantitative determination of individual isotropic species.
These measurements may be important to industrial water analysis from
two standpoints: (1) analysis for nonradioactive constituents by radio-
activation methods; and (2) analysis to detect radioactive contamination of
supply, process, or waste waters, or the concentration of radioactive isotopes
that have been added as tracers. There are three methods for measuring the
separate types of radioactive emission: ASTM Methods D 1943 , D 1890,
and D 1690. ASTM Methods D 2459, D 2470, D 3084, and D 3085 are
additional radioactivity measurements. Specific radionuclides are deter-
mined by ASTM methods as shown in Table 4c. Trace quantities of some
nonradioactive elements can be determined by the neutron activation
method. The sample is exposed to neutron radiation in a nuclear reactor or
an accelerator to produce a radioactive isotope of the element to be
determined. Since this radioisotope has its own characteristic radiations and
mode of decay, its concentration can be determined. Some laboratories
equipped with the extensive and expensive equipment required will perform
this type of analysis on a fee basis. This analysis can determine only the
elemental concentrations and not the chemical form of the element. Chap-
ters 16, 17, and 18 should be consulted for more information on the subject.
Biological or Other Tests (Table 4d and e)
Iron Bacteria
Iron bacteria is a general classification for microorganisms that utilize
ferrous iron as a source of energy and are characterized by the deposition of
ferric hydroxide in their mucilaginous sheaths. The process is continuous
with these growths, and over a period of time large accumulations of slimy
brown deposits can occur. Iron bacteria may clog water lines, reduce heat
transfer, and cause staining; objectionable odors may arise following death
of the bacteria. The organic matter in the water is consequently increased,
and this in turn favors the multiplication of other bacteria.
ASTM Method D 932 describes the procedure for the detection and
identification of iron bacteria by examination under a microscope. Several
genera are characterized, and photographs are provided to aid in their
identification.
Microorganisms
An examination of a water supply for microorganisms is done for one or
more of the following reasons: to detect contamination; to identify and
estimate the amount of algae, bacteria, mold, protozoa, and other plankton;
to elucidate the results of chemical examination and microbicidal treatment;
and to ascertain the causes of undesired odor, color, or turbidity. Chapter 5,
Treatment of Process Water and Waste Water, mentions the effects of, and
treatment for, various microorganisms occurring in water.
ASTM Method D 1128 covers the identification, by microscopic examina-
tion, of bacterial cells, mold filaments, algae, protozoa, and other small
animals, and such particles of inert material as may be found in process
water and waste water.
Sulfate-Reducing Bacteria
Sulfate-reducing bacteria are microorganisms capable of reducing sulfate,
or other fonris of oxidized sulfur, to hydrogen sulfide. This reaction takes
place only under anaerobic conditions; thus, the bacteria may be found in
quantity in anaerobic deep wells, lake or marsh bottoms, reservoir bottoms,
in the soil, and beneath deposits or other bacterial growth in aerated
systems. Their proliferation and the amount of hydrogen sulfide produced
depend upon the sulfate concentration. Because of the production of
hydrogen sulfide, these bacteria cause corrosion of metals and deposition
problems.
ASTM Methods D 993 include a procedure for identification and
estimation of sulfate-reducing bacteria, based on iodometric determination
of the hydrogen sulfide produced by the bacteria in a suitable medium under
anaerobic conditions. A simplified procedure provides for qualitative deter-
mination based on the use of lead acetate paper to detect the hydrogen
sulfide produced.
Toxicity to Freshwater Fishes

With an ever-increasing variety of manufacturing processes using chem-
icals and discharging waste products to rivers and streams, the protection of
fish life from deleterious pollution is an important consideration. ASTM
Method D 1345 is a bioassay procedure applicable for use as a guide in
evaluating acute toxicity of wastes to freshwater fishes, and in estimating
safe concentrations of such industrial waste discharges. It is a batch method
which evaluates the acute toxicity of industrial wastes and other water
210 MANUAL ON WATER
pollutants, and can be applied to the determination of the toxicity of pure

chemical compounds in water solution as well. Acute toxicity is defined as
any direct lethal action of pollution to freshwater fishes that is demonstrable
within 96 h or less. Test fish are exposed to dilutions of the sample being
evaluated for specified periods, and median tolerance limits are calculated.
This test does not evaluate long-range toxic effects such as substances that
may interfere seriously with growth and reproduction.
Aluminum
The control of aluminum in industrial waters is desirable because of its
potential scale- and sludge-forming characteristics. Its presence may be due
to suspended clays which are present in the raw water or from coagulants
used in the pretreatment of raw waters. Corrosion and erosion of aluminum
tubing in condensers and feedwater heaters also can be a source of
aluminum. While the corrosion-erosion rate of these tubes is low, the large
surface area involved eventually can lead to sizeable concentrations in
boilers.
Aluminum in the form of alumina or the aluminum ion can be trouble-
some in boilers. If hardness treatment of the feedwater is not properly
controlled and maintained, scale deposits of calcium and magnesium
aluminate may form. Also, in carbonate- or phosphate-treated feedwaters,
the presence of aluminum in excess may cause hard adherent aluminum-
silica complex deposits in the boiler. This sodium aluminum silicate scale
called analcite sharply reduces the heat-transfer ability of boiler tubes. Thin
films of analcite deposits can lead to several overheating and metal failure.
Alumina can be troublesome in cooling waters, because it precipitates at
pH levels usually maintained. Alumina may be present because of carry-over
from coagulation and clarification procedures. Dissolved aluminum may
also react with cooling water treatment chemicals, such as phosphates, to
form precipitates.
When analyzing raw waters for aluminum, attention should be given to
the presence of insoluble clays and other aluminum-containing suspended
matter. In boilers, these clays will become soluble forms of alumina at
normal operating levels of alkalinity and temperature, and so will become
available to contribute to deposition.
Fluorometric and spectrophotometric techniques both offer extreme
sensitivity for determining trace amounts of aluminum. One rapid method
[75] based on the fluorescence of aluminum 8-hydroxyquinolate in chloro-
form solution is capable of detecting 5/ug of aluminum per litre of water
solution. Another micrograms-per-litre range test [75] for aluminum in high-
purity waters uses morin reagent. In this test for aluminum concentrations of
0.25 to 2.0 /xg/litre, the standard deviation ranges from 0.09 to 0.06. These
two methods now are included in ASTM Methods D 857, Methods B and A,
respectively, Spectrophotometric Method C using Perron (8-hydroxy-7-
iodo-5-quinoline sulfonic acid) determines only soluble aluminum in the

concentration range of 0 to 2.0 mg/litre.
Several other reagents exist for the colorimetric determination of alumi-
num. Among these, aluminon (aurin tricarboxylic acid) probably is the most
widely used. Improvements in the aluminon test in recent years have
increased its reliability and sensitivity [77], and some authors believe it to be
superior to other colorimetric tests, including hematoxylin [18,19]. Methods
using aluminon reagents [77] have a limit of detection of 0.02 mg/litre.
Another test [20] for aluminum using Solochrome Cyanine R is pur-
portedly capable of detecting 0.004 mg/litre aluminum. It is relatively simple
and reliable, but fluoride and large amounts of orthophosphate, pyrophos-
phate, and hexametaphosphate interfere. Ascorbic acid is used to eliminate
ferric iron interference.
Ammonia and Ammonium

Ammonia is a gas which is highly soluble in water and is found in many
surface and groundwaters. In mineral waters, it is mostly the result of
percolation through strata which contain ammoniacal salts; however, am-
monia can be the result of reduction of nitrate in waters by prolonged
contact with metallic piping and tanks, or percolation through some
reducing media such as ferruginous sand. Ammonia in water is sometimes
derived from decomposition of sewage, and from agricultural runoff. This
ammonia is readily oxidized to nitrate by certain organisms. The presence of
ammonia is often regarded as indicative of fresh pollution by nitrogenous
wastes.
Concentrations of ammonium ion, as distinct from total ammonia
content, can be calculated readily from the latter value if sufficient in-
formation is available. The total ammonium content is distributed among
ammonium ion, free ammonia (ammonium hydroxide), and other un-
ionized ammonium compounds. When the last can be neglected, the
concentration of ammonium ion at any given temperature depends on the
pH of the water.
Generally speaking, the value is useful only for estimating the effect of
small concentrations of ammonia on the conductivity of steam condensate.
For this purpose, distribution may be obtained from a table in which
percentages of ionized and un-ionized ammonia are given for the pH range
from 1 to 13. The difference between total ammonia and ammonium ion also
has some significance in the prediction of losses of ammonia at vents.
Ammonia is often used in water treatment to control or adjust pH. For
example, in condensate systems, ammonia can be fed to control corrosion by
carbonic acid. In this respect, ammonia may be objectionable if the
condensate is to be used as boiler feedwater. Ammonia, especially in the
presence of oxygen, can be corrosive to the copper and copper alloy
equipment in steam condensing and preboiler systems. In cooling waters,
212 MANUAL ON WATER
ammonia can be a nutrient for microorganisms, giving rise to objectionable

amounts of microbiological growth. Also, the conversion of ammonia to
nitrates by certain microorganisms can cause a reduction in the pH of the
water.
The two procedures of ASTM Method D 1426 cover the determination of
ammonia and ammonium ions. The high-precision procedure is applicable
to all types of industrial waters. It involves buffering the sample with a
mixture of phosphates to pH 7.4 to inhibit hydrolysis of organic nitrogen
compounds, and distilling it into a boric acid solution. Ammonia nitrogen is
measured in the distillate by Nesslerization or by titrating with standard
sulfuric acid. The routine procedure uses direct Nesslerization (without
distillation) for rapid, routine determination of ammonia nitrogen. Inter-
ferences in both procedures are discussed in the method cited. Determina-
tions should be carried out at room temperature (25 to 30° C).
Nessler reagent is a strongly alkaline solution containing a complex iodide
of mercury which reacts with ammonia to form a brownish complex amine.
The reaction is so sensitive that less than 0.1 mg/ litre of ammonia can be
measured easily. Color comparisons with standards can be made roughly in
bottles, but more exactly in Nessler tubes. Color comparators or colori-
meters are also used. Standards can be made from known concentrations of
ammonia, or permanent standards may be prepared in Nessler tubes from
solutions of potassium chloroplatinate and cobaltous chloride. Samples
containing colored or turbidity-forming impurities frequently can be diluted
with ammonia-free water until they no longer interfere with the determina-
tion. Barium, calcium, magnesium, and other metals that produce turbidity
can first be precipitated in the water sample and the supernatant used for
ammonia analysis. Cloudiness due to the formation of magnesium hy-
droxide and calcium carbonate can be prevented by using Rochelle salts or
glassy polyphosphates to sequester the alkaline-earth ions.
The sodium phenolate method [21] is reportedly free from the annoyance
caused by turbidity when determining low concentrations of ammonia
colorimetrically. This procedure is based on photometric measurement of
the intensity of the blue color formed when sodium hypochlorite is added to
a sample containing a low concentration of ammonia plus sodium phenolate
solution. It has been reported that the presence of amines does not interfere
with this method.
The total nitrogen content of a water includes nitrate, nitrite, organic, and
ammonia nitrogen. The latter two comprise what is regarded as the total
Kjeldahl nitrogen. A Kjeldahl nitrogen involves a digestion step in which the
organic nitrogen present is converted to ammonium bisulfate by the action
of sulfuric acid and a catalyst. Nitrate and nitrite are not included by this
procedure, but can be determined as ammonia if a reduction method is used
[22].
Organic nitrogen compounds are produced by biological processes; hence,
an increase in the organic nitrogen content of a water may indicate increased
pollution. The organic nitrogen can be determined directly with the Kjeldahl
method if any ammonia present is first removed by distillation from a
buffered sample.
In all of the foregoing procedures, after digestion, the resulting ammonia
is distilled from an alkaline medium and can be measured as given in ASTM
Test of Ammonia on Industrial Water and Industrial Waste Water (D 1426),
or titrated if the amount is large.
Barium and Strontium

Barium and strontium are usually negligible and, therefore, unimportant
in most industrial waters. However, water supplies in occasional geographic
locations may contain significant quantities. Barium and strontium may be
introduced into the water system as minor impurities in chemicals used in the
softening and pretreatment of water supplies. The reactions of barium and
strontium in boiler waters and cooling waters are similar to those of calcium
and magnesium and, therefore, seldom cause problems in industrial systems.
Both strontium [23] and barium [24] have been determined by precipita-
tion with chromate. The precipitate is separated, redissolved, and the
chromate determined colorimetrically. The sample should contain at least 10
mg of the constituent sought, and suitable precautions must be taken to
effect its quantitative precipitation as the chromate.
Barium of the order of 50 mg/ litre can be determined quickly by titration
with sodium sulfate [25] in the presence of a suitable indicator. However,
when special accuracy is required, the gravimetric method, in which barium
sulfate is precipitated and weighed, is recommended. A more rapid method
is the estimation of barium from turbidity formed upon the addition of a
small quantity of sulfuric acid to the water, in comparison with similarly
prepared standards. Small quantities of barium in natural waters con-
taminated by industrial wastes were determined by Dzhaparidze [26\ in this
manner. One milligram per litre of barium could be detected.
For an accurate determination of strontium, it is essential that it be
separated from calcium. When strontium is present, the standard gravi-
metric oxalate method determines calcium and strontium oxides together.
The oxides are dissolved in nitric acid and the solution is evaporated to
dryness. The residue is then heated to 160°C for 2 h, cooled, and a mixture
of equal parts of absolute alcohol and ether is added to dissolve calcium
nitrate. Evaporation, heating, and extraction are repeated on the strontium
nitrate residue until extraction of calcium nitrate is substantially complete
(there is little or no residue when the extract is evaporated). Strontium then
is converted to the sulfate, which is ignited and weighed. Quicker but
considerably less reliable results can be obtained by using the control
methods given for barium.
Several workers [27-29] have developed flame photometric methods for
strontium. Two milligrams per litre strontium is the lower limit on most of
214 MANUAL ON WATER
these methods; however, Rains et al [30], using a high-resolution flame

spectrophotometer, determined strontium in concentration of 0.1 to 1.0
mg/ litre in the presence of several thousand-fold excess of calcium, without
prior separation. Strontium and barium have been determined by direct
flame photometry by Nakajima [37] and co-workers, in amounts as little as
10 ,ug.
Both strontium and barium have been determined spectrographically [32],
The range for quantitative determination is from 0.001 to 0.05 percent of the
residue remaining after evaporation of the sample. Tungsten was used as the
internal standard. Strontium in concentrations of 0.005 to 1.0 mg/litre was
determined by Skougstad [33] using a copper-spark spectrographic pro-
cedure and an internal standard of lanthanum.
Bicarbonate, Carbonate,
Carbon Dioxide
Carbon dioxide dissolves in water in varying amounts depending upon
temperature, pressure, and mineral content of the supply. In water, free
carbon dioxide increases corrosivity, causing acid attack upon metals with
which it comes in contact. In naturally occurring waters, dissolved carbon
dioxide is responsible for the dissolution of many minerals picked up by the
water in its percolation through the ground. Limestone, for example, is
dissolved in this manner, resulting in soluble calcium bicarbonate. Free
carbon dioxide present in process waters causes corrosion to the system. In
steam condensate, carbon dioxide forms carbonic acid; it is very corrosive
and appropriate steps must be taken for its elimination or neutralization.
The total available carbon dioxide content of water is considered to be the
sum of that combined as carbonate and bicarbonate, and free carbon
dioxide present. These ions exist in equilibrium with each other, their
relative concentrations depending upon the pH. Procedures for carbon
dioxide determination are given in ASTM Methods D 513. In the most
precise method, all the carbon dioxide, whether present originally in
combined form as carbonate and bicarbonate, or as free CCh, is separated
from the sample and collected in standard barium hydroxide; the excess of
the base is found by titration and calculated to carbon dioxide. In a more
routine method, the sample is treated to convert all carbon dioxide to
strontium carbonate; the carbonate is then dissolved in hydrochloric acid
and the excess is titrated and related to the original carbon dioxide content.
Curves are included to enable estimation of the percentage composition of
the sample in terms of carbonate, bicarbonate, and free carbon dioxide at all
pH levels.
Free carbon dioxide can be estimated very quickly by titrating a sample
with standard sodium carbonate or sodium hydroxide until only bicarbonate
is present in the solution (phenolphtalein end point); this procedure is much
used for carbon dioxide in well waters. Standard sodium hydroxide solution
must be used for free CO: determinatons on acid waters such as de-
mineralizer cation exchanger effluents; sodium carbonate is not an accept-
able titrant in this case. Alkalinity titrations and pH determinations along
with the use of nomographs can also be used to determine the concentrations
of the various forms of carbon dioxide.
Carbonate and bicarbonate ions are the chief sources of alkalinity in
naturally occurring waters, and are important when considering such waters
for plant use. Major quantities of soluble calcium and magnesium hardness
will be present as the bicarbonates. However, when such supplies are heated
sufficiently to drive off part of the combined COz, forming carbonates,
precipitation of calcium carbonate may occur. The scale-forming character-
istics of calcium carbonate may be useful at times. A thin layer laid down
inside metallic piping can provide a protective coating, preventing corrosion.
In such cases, the control of alkalinity (carbonate, bicarbonate, and carbon
dioxide content) is very important.
In boiler water applications, knowledge of the carbonate and bicarbonate
content of the water supply is essential for proper treatment. The concen-
trations of these ions must be accurately controlled and adjusted to obtain
optimum removal of hardness in lime or lime-soda ash pretreatment
systems. In low-pressure boilers, maintenance of a properly controlled
carbonate reserve in boiler water is desirable to prevent deposition of
calcium sulfate, which forms a very hard scale. Since calcium carbonate is
much less soluble than calcium sulfate at boiler water temperatures, the
presence of the carbonate ion ensures that residual calcium will be precipi-
tated as the carbonate, rather than sulfate.
In feedwater of higher-pressure boilers, control of the bicarbonate content
is essential to minimize breakdown and release of carbon dioxide, which will
go off with the steam. Carbon dioxide will dissolve in condensate to form
carbonic acid, which is corrosive to steam-condensing equipment and to the
condensate return system.
The concentrations of carbonate and bicarbonate alkalinity in water
supplies can be calculated, if the total carbon dioxide and pH of the supplies
are known. ASTM Methods D 513 include procedures for the determination
of total carbon dioxide, and provide graphs and tabular data for estimation
of carbonic acid, carbonate, and bicarbonate ions at all pH levels from 2.0 to
13.0.
For control purposes, approximations of reasonable accuracy can be
made from alkalinity and pH determinations using data compiled in the
form of nomographs [34]. These nomograph methods are not applicable
when the water supply contains appreciable concentrations of phosphates,
silicates, sulfites, and certain kinds of organic matter.
There are additional methods of somewhat restricted accuracy. The
titration of samples successively to the phenolphthalein and methyl orange
(or other suitable indicator) end points [4] can be used to indicate carbonate
and bicarbonate concentrations. The interfering ions mentioned in the
216 MANUAL ON WATER
foregoing also affect this titration method. On boiler waters, color from
tannins or lignin derivatives will interfere with detection of the titration
end points. A variation of this approach [35] in which the sample is titrated
between definite pH limits, before and after the carbon dioxide has been
boiled off, avoids principal titration errors and eliminates corrections by
coupling them with the blank determination.
Another quick method for approximating carbonate in boiler waters
depends on the insolubility of barium carbonate and barium phosphate in
alkaline solution. An aliquot of the sample is titrated to the phenolphthalein
end point, and the titration is then repeated on a second aliquot, to which the
difference between the two titrations represents carbonate and phosphate
alkalinity and may be calculated to carbonate if the phosphate concentration
is known.
Bromide and Iodide

Bromide and iodide seldom are significant in industrial water supplies.
Bromides are found primarily in seawater and may also be present in well
waters in coastal areas suggesting the intrusion of seawater. Iodide ions are
found in almost all waters, but primarily in infinitesimal traces. Aside from
considerations of physiological importance, iodide is insignificant in most
waters.
Bromides that may be present in freshwater supplies probably are
contributed by certain industrial discharges. Bromine also is used as a
germicide in the treatment of some metropolitan drinking supplies and in
many swimming pools. This is also a source of bromides although the
practice is not widespread. Certain connate waters may contain relatively
high concentrations of bromide and iodide ions and can be a source of
contamination to groundwater supplies.
ASTM Methods D 1246 determine iodide and bromide separately, as well
as combined iodide and bromide. In Method A, the catalytic effect of iodide
in the reduction of eerie ions with arsenious acid is used to estimate small
quantities of iodide. The reaction rate is directly proportional to the iodide
concentration, and the eerie ion not reduced is measured indirectly. In
another procedure based on reduction of eerie ions [4], ferrous ammonium
sulfate is added to stop the reaction, the eerie ion remaining oxidizing the
ferrous iron to an equivalent amount of ferric iron. The reaction of the ferric
ions with thiocyanate ions that are added is the basis of the colorimetric
measurement. Time and temperature control are very important.
Method B for bromide, based on the catalytic effect of bromide upon the
oxidation of iodide to iodate, is capable of determining 0.01 to 0.1 mg/litre
bromide [36]. The amount of iodide oxidized in a given period of time is
inversely proportional to the amount of bromide present. Optimum con-
ditions for the oxidation are established and limits on potential interferences
given. Up to 10 mg/litre iodide causes no significant error.
Bromide in the presence of iodide can be determined colorimetrically

using phenol red in the range from 0.1 to 1.0 mg/ litre. The bromide is
oxidized to bromine, then the bromine reacts with the phenol red to give a
blue color. Timing and reagent concentrations are very important; following
the recommendations of Goldman and Byles [37], results are reported to be
good.
In Method C, iodide is determined by oxidizing the iodide to iodate with
bromine in a buffered solution. The excess bromine is subsequently reduced
by adding sodium formate. Iodine equivalent to the iodate then is liberated
from added potassium iodide and titrated. Combined iodide and bromide
are measured by oxidizing them to iodate and bromate with hypochlorite,
the excess hypochlorite subsequently being reduced by adding sodium
formate. Iodine equivalent to the combined iodate and bromate is then
liberated from added potassium iodide and titrated. Bromide is calculated
from the difference in the two titrations.
There are several variations of the method for determining iodide which
involve extraction into various organic solvents and use of alternate
oxidation techniques. One method [38] capable of detecting 0.1 to 1.0
/*g/ litre of iodide uses preliminary procedures very similar to those of
Method C of D 1246. At the last step, however, phosphoric acid instead of
sulfuric acie is used to acidify the sample. Then potassium iodide and starch
are added. The iodine-starch complex is purported to be stable in the
phosphoric acid medium and the absorbance of the complex is measured
spectrophotometrically.
Calcium and Magnesium

Calcium and magnesium are the two major scale-forming contaminants in
most raw-water supplies. Practically all boiler water treatment methods have
as their major objective the prevention or reduction of calcium and
magnesium deposits in the boiler. These two ions from heavy voluminous
deposits ranging from hard scales to soft precipitates on heat-transfer
surfaces and other internals of the boiler system. Should untreated feedwater
be used in a boiler, heat-transfer surfaces soon would become scaled with
calcium sulfate (which produces a hard crystalline deposit) and calcium
carbonate, which can occur either as a soft deposit or as a hard scale.
Calcium carbonate is a major constituent of deposits in preboiler systems.
Magnesium salts, such as magnesium silicate and magnesium hydroxide,
generally form soft deposits and sludges. These hardness deposits, among
others, seriously interfere with heat transfer and reduce efficiency. If such
deposits are not prevented or removed at appropriate intervals, tube failures
frequently occur.
Since calcium and magnesium react differently to the chemical conditions
present in boilers, both should be determined in feedwater analysis, instead
of being lumped together as a total hardness determination. Concentrations
218 MANUAL ON WATER
of calcium and magnesium encountered in industrial waters will vary from

several hundred milligrams per litre in raw-water supplies to 1 /ug/ litre or
less in the feedwater for high-pressure boilers.
Incrustation and scale deposits in open cooling systems can form because
of the concentration of dissolved solids during evaporative cooling. Calcium
carbonate precipitates in the form of scale or sludge when solubility limits
are exceeded. In some open recirculating systems, calcium sulfate also may
be deposited unless bleedoff is properly adjusted.
Calcium and magnesium ions are important considerations in the dyeing
of textiles. They may precipitate certain dyestuffs, producing streaky or
spotty dyed goods, or they may act as mordants, causing greater dyestuff
absorption and deeper shades.
The standard method for calcium and magnesium determination, ASTM
Method D 511, employs conventional gravimetric procedures in which
calcium is precipitated as the oxalate and magnesium as the phosphate.
Weights of residues from ignition of the precipitates are converted to
constituent concentrations.
Time-saving modifications have been used widely for routine testing. For
example, the reprecipitation step in each gravimetric procedure may be
omitted at a considerable sacrifice in accuracy, perhaps as much as 10
percent of the value found. The precipitated calcium oxalate may be
dissolved in sulfuric acid and the resulting oxalic acid titrated with potas-
sium permanganate; the titer is directly related to the calcium content.
Finally, titration methods for hardness can be used to determine calcium and
magnesium separately in most industrial waters.
Several rapid colorimetric tests exist for the determination of calcium and
magnesium. Titan yellow has been used as a reagent for the determination of
magnesium, but this colorimetric procedure has been found unsatisfactory
by many investigators. A variant called thiazole yellow is reported [39] to be
a bftter reagent; it provides a reproducible color that is not affected
materially by iron, aluminum, manganese, phosphorus, or calcium. The
optimum range is 1 to 6 nig/litre of magnesium, with an accuracy within 5
percent. Magnesium in amounts up to 180 mg/litre has been determined [40]
spectrophotometrically using p-nitrophenyllazo-1-naphthol; this reagent
forms a color complex with magnesium having absorption bands at 405 and
580 nm. Other alkaline earths do not interfere, but iron interferes. Xylidyl
Blue II has also been investigated [41] and proposed as a useful colorimetric
reagent for magnesium.
Small amounts of calcium have been determined using glyoxal bis (2-
hydroxyanil), which forms a complex with calcium in alkaline solution and
is stable in an aqueous alcoholic solution. It has a maximum absorption at
520 nm. Kerr [42] investigated optimum conditions, including organic
solvents and the effect of foreign ions, and established a method suitable for
determining 0 to 4 mg/litre of calcium. Magnesium up to 50 mg/litre does
not interfere. If hexanol is used, a quantitative chloroform extraction of the
complex, and re-extraction of the calcium back into an acidified aqueous
phase, also has been proposed as a preliminary separation [43] prior to

developing the color. Small amounts of calcium also can be determined [44]
gravimetrically by precipitating and weighing calcium potassium nickel
hexanitrite.
Chloride
Chloride is present (generally as sodium chloride) in practically all waters.
Occasionally chloride will be present in combination with magnesium or
calcium, although higher concentrations of these salts suggest contamination
by seawater, brines, or industrial wastes. The concentration of chloride in
natural waters varies from a few milligrams per litre, or less, to several
thousand milligrams per litre in some groundwaters. Reasonable amounts of
chlorides can be tolerated in many industrial waters and in potable supplies.
Although chloride usually is not the direct cause of corrosion, the chloride
ion is small, highly mobile, and will accelerate corrosion reactions in
proportion to its concentration, because of its ability to penetrate and
increase solution rate of iron, and its contribution to electrolyte content. For
this reason, brines can corrode a system at a rapid rate. Chlorides can
accelerate stress corrosion of stainless steel alloys, so chloride concentration
must be carefully controlled in systems where these metals are used. In low-
pressure boilers and cooling waters, chlorides present few problems.
Chloride remains unchanged in process water, and this property makes
chloride analysis very useful. Chloride concentrates in boiler waters and
evaporative cooling waters in an exact ratio to the amount of evaporation
that takes place. Thus, a chloride analysis of makeup water compared with
that in the boiler water or cooling water gives a simple procedure for
estimating the amount of concentration that is occurring. Comparative
chloride analyses in conjunction with nitrate analyses also can be used to
indicate contamination of groundwater by sewage wastes.
ASTM Methods D 512 include three high-precision procedures. The first
is a mercurometric titration applicable to high-purity industrial water and
water relatively free of heavy metal ions. The second is a silver nitrate
titration for use with low-purity industrial supplies, process, and waste
waters; and the third, a mercuric thiocyanate colorimetric method. In
Method A, dilute mercuric nitrate solution is added to an acidified sample in
the presence of mixed diphenylcarbazone-bromphenol blue indicator. The
end point of the titration is the formation of the blue-violet mercury
diphenylcarbazone complex. In Method B, the sample is adjusted to pH 8.3
and titrated with silver nitrate in the presence of potassium chromate
indicator. The end point is indicated by persistence of the brick-red silver
chromate color. In Method C, mercuric thiocyanate reacts with the chloride
present, releasing an amount of thiocyanate ion that is equivalent to the
chloride. Added ferric iron forms a red color with the thiocyanate ion, which
is measured photometrically or by visual comparison.
Method A without sample concentration is for chloride concentrations
220 MANUAL ON WATER
down to 0.1 mg/litre. Method B is for chloride concentrations above 5

mg/litre in waters where interferences from color or high concentrations of
heavy metals are present. Method C can detect 0.02 mg/litre chloride, and
primarily is for relatively pure waters where low concentrations of chloride
must be accurately known.
Titration methods using a small aliquot are normally employed when the
chloride concentration is high. However, concentrations above 1000 mg/-
litre probably can be determined somewhat more accurately by a more
cumbersome gravimetric procedure in which a slight excess of silver nitrate
is used to precipitate silver chloride. The precipitate is coagulated by
digestion (but not boiling), filtered through a prepared porous crucible,
washed free from silver nitrate (hot water), dried at 150°C, and weighed.
Strong light will decompose the silver chloride precipitate, causing serious
errors.
Chlorine
The determination of chlorine is important because an increasing pro-
portion of industrial water is chlorinated. In addition to use for sterilization
of potable water, chlorine treatment is used for control of algae and
prevention of bacterial slime growths throughout cooling systems. In cooling
towers, continuous contact with free chlorine can cause some deterioration
of lumber; in large concentrations, chlorine can accelerate corrosion of
metals, especially in the presence of oxygen.
Chlorine in solution exists as free available chlorine such as hypochlorite
ions (OC1) or hypochlorous acid (HOC1), and as combined available
chloride, consisting of chlorine combined with ammonia nitrogen or nitro-
genous compounds. The sterilizing action of combined chlorine is not as
potent as that of free chlorine, but it possesses a longer residual time in
water; that is, it is more stable. In a given contact period, about 25 times as
much combined chlorine as free chlorine is required to yield the same
sterilization effect. Thus, the ammonia content of water is an important
consideration in chlorination treatment.
Chlorine residuals may be present in some surface waters where chlori-
nated industrial and sewage wastes are discharged.
Two methods for chlorine determination are presented in the Annual
Book of ASTM Standards, Part 31 [7]: Methods D 1253; and D 1427.
Method D 1253 includes a high-precision and two more routine methods for
determining total available chlorine residual. Two of these methods are
suitabi' ~ differentiating between free and combined residuals. The high-
precision method is an amperometric titration with phenylarseneoxide.
Routine Method A (which also differentiates between free and combined
residuals) is used for rapid control determinations of chlorine residuals up
to 10 mg/litre, but is most sensitive below 1 mg/ litre. Routine Method B
is applicable for chlorine residuals above 1.0 mg/ litre, and is particularly
useful for measuring high residuals (10 to 300 mg/litre). Both routine
methods are colorimetric procedures using orthotolidine to form a yellow
color.
The two methods of D 1427 permit determination of chlorine residuals in
the presence of most interfering constituents found in industrial waste water.
Method A is a starch-iodide titration, and Method B is an amperometric
titration. Both cover total chlorine measurement, but only the latter is
applicable to determination of free available chlorine.
A rapid method for determining free chlorine residuals above 1.0 mg/litre
in cooling waters involves titration of an acidified sample with a standard
methyl orange solution. Free chlorine decolorizes the methyl orange until
the chlorine residual is exhausted; then the color of the methyl orange
persists, indicating the end point. Chromate in low concentrations does not
interfere, but at higher levels masks the end point. Xylene cyanole has been
used instead of methyl orange, giving a readily discernible green end point in
high-chromate waters.
Samples to be analyzed for chlorine should be kept away from direct
sunlight prior to analysis. The analysis should be made as soon as possible
(not more then 5 min after sampling), because a chlorine residual may
decrease with time as a result of any chlorine demand of the sample.
Chromium
Chromium in water is of interest principally in the hexavalent form, in
connection with its use as a corrosion inhibitor. It also is a frequent
component of industrial wastes, in which it may be present in either the
trivalent or hexavalent form, or both. It is of minor importance, however, in
water supply for power plants. If introduced as a contaminant into boilers,
chromate will be reduced to trivalent chromium and precipitated as chromic
hydroxide.
Chromate and dichromate are not found naturally in water, but may be
present because of industrial contamination. The presence of chromium
compounds in surface water supplies usually indicates discharge of metal
plating solutions and similar industrial wastes, or the use of chromate and
dichromate-based corrosion inhibitors in cooling water. Obviously, the
manner of disposal of blowdown water from cooling towers using chromate
inhibitors presents a problem. Discharge of chromates into surface waters
should be avoided to prevent potential contamination of irrigation and
drinking water supplies, damage to aquatic life and other adverse effects.
The U. S. Public Health Service (USPHS) limit for hexavalent chromium in
drinking water is 0.05 mg/litre maximum.
Alkaline chromates and dichromates have long been used as corrosion
inhibitors in cooling systems. In ferrous metal systems, the chromate or
dichromate ion reacts with ferrous ions released at anodic areas, hexavalent
chromium being reduced to trivalent chromium and precipitated to form a
222 MANUAL ON WATER
blocking mixture (thought to be a mixed ferricchromic hydroxide) which

seals pores in the iron oxide film and prevents further corrosion. The
resulting protective film is tough, adherent, and self-healing, when adequate
chromate concentrations are maintained in solution. The maintenance of an
adequate chromate concentration to provide substance for the self-healing is
essential. Chromate and dichromates are particularly effective as corrosion
inhibitors in refrigerator brine systems.
A method for amounts of chromium in excess of 100 mg (in the absence of
organic matter) involves oxidation of chromium to the hexavalent form with
permanganate or silver nitrate or persulfate, destruction of excess oxidant,
and titration with standard ferrous sulfate solution. Excess ferrous sulfate
may be used and back-titrated with dichromate, using o-phenanthroline, or
with permanganate.
Graham proposed a method which depends on the reaction of hexavalent
chromium with diphenylcarbazide in acid solution to produce a reddish-
violet color. The optimum range is up to 5 mg/litre and the accuracy is about
4 percent. Phosphoric acid has been shown to mask iron, the most important
interfering element \46\. A similar method has been proposed for the
determination of chromium in samples of polluted water, industrial wastes,
or sewage [4]. Organic material is decomposed by fuming with strong acids,
and other interfering metals are extracted. Chromium is oxidized to the
hexavalent state with permanganate, the excess permanganate destroyed
with sodium azide, and the color developed with diphenyl-carbazide. See
ASTM Methods D 1687 for similar procedures using either permanganate
or hypobromite for oxidation. Goodman [47] simplified the diphenyl-
carbazide color development by preparing the reagents as a single mixture of
dry powders rather than separate solutions.
Carmoazine reagent [(2-sulfo-l-napthylazo)-l-naphthal-4-sulfonic acid]
can be used to determine chromium [48] both visually and photometrically
in the microgram-per-litre range. According to the author, no other ions
interfere. Polyanskii [49] used 1-naphthylamine to determine chromium.
The sensitivity of this method was increased by extracting the colored
complex with isoamyl alcohol. The sensitivity was 0.1 mg/litre.
Copper
Copper is rarely found in natural waters and, when it exists, is usually
derived from industrial wastes. Copper sulfate and stabilized copper com-
pounds now are being used extensively in the treatment of reservoir water to
control the growth of algae and diatoms, providing another source of copper
in various waters.
Knowing the copper content of industrial waters is important mainly
because of corrosion and deposition processes in which it is a factor. Copper
can be deposited in metallic form in pipes and boiler tubes, presumably with
the dissolution of an equivalent amount of iron. So far, the role of copper in
boiler corrosion has not been satisfactorily explained; however, copper

deposits in boilers as a spongy mass or is laid down in sheets. Corrosion in
the preboiler system usually is the source of such copper. Some authors
believe copper is not the primary cause of corrosion in boilers, but
accelerates corrosion once it has started. Very small amounts of copper will
also attack aluminum, particularly in hard water.
Analysis for copper content of water also can be used to monitor
corrosion of copper, brass, and other copper alloy fittings and tubes.
Concentrations of ammonia above 10 mg/litre in water along with dissolved
oxygen can result in fairly significant copper dissolution. Carbon dioxide
and sodium bicarbonate also can cause corrosion of copper and copper
alloys.
Several colorimetric agents for use in determining copper are in existence.
Sodium diethyldithiocarbamate, which produces a yellow color with copper,
lacks sensitivity and selectivity. Extraction of the colored product with a
suitable organic solvent increases the sensitivity and simplifies the removal
of interferences. Ammonium thioglycolate added during extraction will help
interferences from iron, nickel, manganese, and chromium. Ethylenedia-
minetetraacetic acid (EDTA) can be used to eliminate interference from zinc
and lead. The method can be used in the range from 0.1 to 1 mg/litre. If
interfering substances are not present in troublesome concentrations, deter-
mination can be made directly on the sample.
The procedure using (2-hydroxyethyl) dithiocarbamate (cuprethol) is
superior to that utilizing sodium diethyldithiocarbamate; routine Method C
of ASTM Methods D 1688 employs this reagent for the determination of
copper in relatively unpolluted waters. No extraction step is involved.
Precision Methods A and B of D 1688 use neocuproine to develop a color
with copper, as a basis for its determination. The copper is reduced to
cuprous, and the color complex then extracted with chloroform. Table 5
compares the three methods presented and indicates their applicability.
A procedure using bathocuproine offers promise of being an improved
method for copper [4], and is recommended for unknown and polluted
waters because of its relative freedom from interferences.
A very selective and sensitive reagent, nickel diethyl dithiophosphate, has
been used [50] for the determination of copper. To determine copper in
water, 1 to 2 litres of sample is acidified and then extracted three times with
small portions of 0.005 M nickel diethyl dithiophosphate and carbon
tetrachloride. Copper in the amount of 0.01 mg was determined at a dilution
of 1 to 109. This procedure for trace amounts of copper was modified by
Nemodruk and Stasynchenko [57] to be faster, simpler, and more accurate.
Oxalyldihydrazide has been used [52] to determine from 0.02 to 1 mg of
copper per millilitre of sea water without using an extraction. To 10 ml of
sample was added 2 ml of saturated oxalyldihydrazide, 10 ml of ammoniacal
5 percent ammonium chloride, and 10 ml of 40 percent aqueous acetalde-
hyde. A 1-h waiting period was required for complete color development. At
TABLE 5—Copper analysis methods.
Method D 1688 Precision Method A (Neocuproine) Precision Method B (Neocuproine) Routine Method C (Cuprethol)
Basis of measurement colorimetric colorimetric colorimetric
Range 0.05 mg/ litre Cu or over 0.002 to 0.100 mg/ litre Cu 0.05 mg/ litre Cu or over
Application all types of industrial water and in- industrial waters such as steam con- industrial waters relatively unpol-
dustrial waste waters containing 0.05 densates and deionized water luted. Does not involve organic
mg/ litre Cu or more extraction; therefore more rapid
than Method A
Summary of method the copper is reduced with hydroxyla- same as Method A except choice is cupric ions form a yellow-colored
mine hydrochloride and the pH ad- given between chloroform and iso- chelate with cuprethol [bis-(2-hy-
justed to 4.0 to 6.0. The cuprous amyl alcohol as the organic solvent droxyethyl) dithiocarbamate]. The
reacted with neocuproine and the used for extraction colored compound formed at a pH
yellow color extracted with chloro- between 5 and 6 is read or com-
form is read or compared pared
Interferences none from the ions commonly found in none of the ions commonly found in excessive concentration of interfer-
industrial water or industrial waste .low-solids industrial water ing ions will rarely be encountered
water. Interfering ions are removed in unpolluted water. If Bi, Co,
Hg +2 > Ni, CN, and sulfidel ions
present, use Method A
Instrumentation or special Nessler tubes, filter photometer, or Nessler tubes, filter photometer, or Nessler tube, filter photometer, or
equipment spectrophotometer (capable of spectrophotometer (capable of spectrophotometer (capable of
reading at 457 nm) reading 454 to 457 nm) reading at 435 nm)
low concentrations of foreign metals, only chromium interfered. Zinc

dibenzyldithiocarbamate [53] has been used for the determination of copper
in potable water. A single extraction of an acidified sample is made with a
chloroform solution of zinc dibenzyldithiocarbamate and can be applied to
colored or turbid waters. Other materials likely to be present do not
interfere, and the method is suitable for routine use. Copper in the amount
of 0.02 mg/litre can be detected.
Low concentrations of copper are determined also by the potassium
ferrocyanide method [54]. Copper in a sample from which lead has been
removed as lead sulfate is first concentrated by precipitation as copper
sulfide. Any iron present is removed after the precipitate has been converted
to nitrate, and potassium ferrocyanide is added to develop the reddish color,
which is then compared with standards.
The dithizone method proposed for copper (also used for zinc and other
metals) depends upon formation of stable, colored metal dithizonates from
extraction of aqueous solutions with dithizone in carbon tetrachloride or
chloroform. By adjustment of pH, use of complexing agents, and other
techniques, separation of metal dithizonates is made. This method is time-
consuming and it is less accurate than the cuprethol or cuproine methods.
A direct Zincon method, described later under Zinc, also has been adapted
for determination of copper.
An electrolytic method for copper also has been used. It involves
deposition of copper metal on a platinum cathode and subsequent weighing
of the electrode. This method is an excellent one for waters of high copper
content, such as plating wastes. ASTM Method E 53 [55] can be used for the
electrolytic determination after first concentrating the sample.
Cyanide
Cyanide does not occur naturally in water, but is the result of contamina-
tion. Waste waters from certain chemical processes and metal plating
industries are the chief source of the contaminant in surface waters. Residues
for insecticides and rodenticides can also be a source.
The chief reason for concern with cyanide in water is its highly toxic
character. In low concentrations it can kill marine life and useful bacteria,
thus upsetting the natural purification abilities of streams and promoting
buildup of other organic pollutants. Cyanide also may be detrimental to
sewage treatment processes.
The major problem in analyzing for cyanide is the removal of interfering
substances. ASTM Methods D 2036 include procedures for removing
interferences and making the determination. Methods for removal of
sulfides, fatty acids, and oxidizing agents are outlined, followed by distilla-
tion procedures for removing most other interferences. After interferences
are removed, cyanide is determined either by a modified Liebig titration
method using p-dimethylaminobenzalrhodanine internal indicator, or by a
colorimetric method using pyridine-pyrazolone if the concentration is below
226 MANUAL ON WATER
1 mg/litre. Provisions also are made for determining the concentration of

cyanide that is amenable to chlorination treatment.
Fluoride
Fluoride is present in almost all natural water supplies. The fluoride
content varies considerably with the source, and with the extent of con-
tamination by waste discharges. Flourine has little significance in industrial
waters, but in drinking water in amounts of 1 to 1.5 mg/litre it is an effective
preventive of dental caries, if ingested during the period of permanent tooth
development. Above this amount, fluorine may cause mottling of teeth. For
potable water supplies, fluoride in excess of 1.0 to 1.5 mg/litre should be
removed by appropriate treatment processes.
Fluoride occasionally is used in cooling water treatment to prevent
precipitation of aluminum ions. In this application it functions by forming a
soluble complex, and can be considered a deposit inhibitor or antifoulant.
Fluoride also will precipitate in the presence of higher concentrations of
calcium and magnesium. In waters containing fair amounts of the fluoride
ion, the fluoride content will be significantly reduced during lime-soda
softening.
ASTM Methods D 1179 provide two procedures for the determination of
fluoride. Precision Method A [56] based upon the bleaching effect of the
fluoride ion on zirconium-organic complexones uses SPADNS reagent
(sodium 2-(p-sulfophenylazo)-l ,8-dihydroxy-3,6-naphthalene disulfonate).
This method tolerates many interferences and determines fluoride in the
range from 0.00 to 1.40 mg/litre. No waiting period is involved for color
fading; the reaction is immediate, color is stable and follows Beer's law in the
concentration range mentioned. The high-precision method provides for the
elimination of interferences by incorporating a distillation procedure, and
should be the method of choice on unknown waters.
If interferences are known to be absent, or present in sufficiently low
concentrations, the distillation may be omitted, and routine Method B used.
Method B depends upon the bleaching effect of the fluoride ion on
zirconium alizarin color.
A direct photometric method [57] using the wine-red cerium (Hl)-alizarin
complexone to produce a blue color with fluoride also can tolerate relatively
high amounts of interferences.
Hydrogen
Hydrogen occurs in the atmosphere in very minute quantities but is found
free in nature in significant amounts in the gases of volcanoes and fumeroles.
It has also been found occluded in rocks such as granite and has been
reported to be sometimes present in natural gas. Hydrogen dissolved in
natural waters is practically nonexistent and in the past was not regarded as
an important constituent of industrial water. It has now been recognized that

the rate of corrosion reactions in expensive boiler systems can be followed by
comparing hydrogen concentrations in the water and vapor in various parts
of the system. The hydrogen produced is a by-product of the dissolution of
iron in water (corrosion reactions). The methods used for measurements of
hydrogen depend on elaborate instrumentation [58,59], since the concentra-
tion sought is of the order of micrograms per litre.
ASTM Methods D 1588 present two methods. The method of highest
precision is based on the separation of hydrogen and measurement of its
thermal conductivity compared with the thermal conductivity of air under
the same conditions. Zero to 40 /ug/litre can be determined with an accuracy
of ±5 percent. The alternative method is a volumetric determination where
the dissolved gases are removed from the water sample by boiling and
collected. The gases are then analyzed in an Orsat-type apparatus. Both
methods are intended for analysis of steam condensates but may be extended
to waters of similar character.
Analysis for dissolved hydrogen has become of interest in the nuclear
power field due to the use of hydrogen as a suppressor of water dissociation
in the primary systems of pressurized water nuclear power plants. The
concentrations of interest are much higher than in the case of boiler systems,
up to 25 mg/litre and higher. A method often used employs a gas transfer
apparatus for volumetric determination of the total dissolved gas evolved
from a water sample under vacuum. A Blacet-Leighton apparatus is used to
determine the proportion of hydrogen in a sample of the gas by measuring
the reduction in gas sample volume after removing the hydrogen by
exposure to heated copper oxide.
Iron
The presence of iron in water is objectionable because it contributes to
problems of taste, discoloration, turbidity, and deposits. For many uses,
such as high-pressure boiler feedwater, process waters for fabric dyeing and
paper making, photographic film manufacture, icemaking, and food proc-
essing, the water must be almost entirely free from iron.
The occurrence of iron in water varies widely depending on source of the
supply. The iron content of some deep wells and industrially contaminated
streams can be very high, while other sources may be iron-free. Iron, if
present (uncomplexed) in natural waters, normally is found in the form of
ferrous salts; however, upon contact with air it is readily oxidized to ferric
salts, which precipitate to give characteristic rust-colored deposits. Ferric
iron in many instances will remain in suspension in water, because of
occasional difficulty in coagulating ferric hydroxide. It can be removed by
energetic flocculation followed by filtration.
Corrosion is the chief source of the iron in boiler water. The iron often
creates serious problems in high-pressure steam-generating systems. Cor-
228 MANUAL ON WATER
rosion of metals in boiler and preboiler systems releases iron in ionic form,
and iron oxides which circulate with the water and may not be removed by
blowdown. Such oxides form adherent coatings and can lead to tube failures
by overheating, caustic gouging, and hydrogen damage. On surfaces of high
heat transfer, ferric iron complexes with sodium silicate, if present, to form a
very hard, strongly adherent deposit called acmite.
Cooling waters containing suspended or dissolved iron also can develop
serious fouling problems due to deposition of iron oxide or reaction
products of iron with treatment constituents such as chromate and phos-
phate. Dissolved or suspended ferric and ferrous iron also can foul ion
exchange resins used for water softening.
The analysis of iron in water usually is done spectrophotometrically or
colorimetrically, although a few titrimetric procedures exist for milligrams-
per-litre concentrations. ASTM Methods D 1068 are based upon the
intensity of an orange-red color produced by the reaction of ferrous iron
with orthophenanthroline or bathophenanthroline. Method A uses ortho-
phenathroline, and is designed for iron concentrations in excess of 0.05
mg/litre. It also is capable of distinguishing the distribution of iron between
the ferrous and ferric states. The proportion of soluble and suspended iron
can be obtained by difference from determinations made on filtered and
unfiltered samples. Method B uses the more sensitive color reagent batho-
phenanthroline, and is capable of detecting less than 1 jug/ litre of iron. It is
particularly applicable to high-purity water and steam condensates. Table 6
indicates areas of applicability of the three iron methods.
Good results also have been obtained with a 2-2' bipyridine [60], 2,2',2"
tripyridine, and with thioglycollic acid (mercaptoacetic acid) [61]. The
principle of the bipyridine method is the formation of a pink complex with
ferrous iron at a pH of 3 to 4. Tripyridine develops a reddish purple color
with iron over a wide pH range. The use of this reagent along with
ethylenediamine to complex heavy metals is the basis for a standard method
[4\. The reaction between thioglycollic acid and iron gives a relatively stable,
reddish purple color. 'A thiocyanate [62] method also has been used.
Comparative results on samples of low iron concentration using batho-
phenanthroline, orthophenanthroline, bipyridine, arid thiocyanate methods
have been reported [63], and Herre [64] made a detailed comparison of
thioglycollic acid, potassium thiocyanate, sulfosalicylic acid, ferron, and o-
phenanthroline. The bathophenanthroline method has been modified to
eliminate the extraction step, the lower limit being 5 /ug/litre iron [65]. Ethyl
alcohol is added to effect the solution of the iron-bathophenanthroline
complex in water. Grat-Cabanca [66] has studied interferences in the
bipyridine method with particular attention given to cobalt, nickel, and
manganese. He found that no interference from these ions resulted if the
ratio of the interfering ion to the iron concentration did not exceed 15,22, or
80, respectively.
Several colorimetric agents exist for iron and a number of new reagents
TABLE 6—Iron determinations.
Method D 1068 Precision Method A Precision Method B
Basis of measurement colorimetric, o-phenanthroline colorimetric, bathophenanthroline
Range concentrations of iron above 0.05 mg/ litre in indus- low concentrations of iron (200 yug/litre and under)
trial water and industrial waste water in industrial water
Application Industrial water and industrial waste water. For dis- industrial water with low iron concentrations. Total
solved and total iron iron
Summary of method the ferrous iron, solubilized with acid, forms an orange- for total iron the iron is reduced with hydroxylamine
red complex with 1,10-phenathroline. The intensity of hydrochloride and reacted with bathophenanthro-
the color produced is proportioned to the amount of line to form a red ferrous complex. The complex is
ferrous iron in the water. Ferric iron is converted to extracted with n-hexyl or isoamyl alcohol and color
ferrous with hydroxylamine hydrochloride for total intensity measured
and dissolved iron
Interferences interference if Cu, Co, Cr, Zn are ten times Fe concentra- if the pH is between 3.3 and 3.7, a I-mg/ litre concentra-
tion; Ni when in excess of 2 mg/ litre; Bi, Ag, and CN- tion of the following ions does not interfere: Cu, Mn,
must be absent; methods given for removal of inter- Al, Zn, Mg, Na, SiO2, NO3-, PO4=
ferences
Instrumentation or special Nessler tubes, filter photometer, or spectrophotometer Nessler tubes, filter photometer (green filter), or spec-
equipment capable of reading at 510 nm trophotometer capable of reading at 533 nm
230 MANUAL ON WATER
have been proposed. Several of these may be applied to specific problems,

but many have serious interferences.
In collecting iron samples, if the concentration of iron is less than 0.1
mg/ litre, it is necessary that the sample be collected under acid to assure full
recovery of iron. It is often convenient to run both iron and copper on the
acidified sample.
Load
Lead is practically nonexistent in natural waters; in cases where lead is
present, it is directly due to corrosion reactions or waste contamination.
Mine waters and wastes from metal plating industries are the most common
sources of contamination in raw-water supplies. Most instances of lead being
present in water may be traced to contact with lead piping or lead-bearing
jointing compounds used with threaded piping. The character of the water
determines its plumbo-solvency. Naturally occurring waters that are highly
acid or contain large amounts of carbon dioxide and are low in calcium and
magnesium bicarbonates are apt to dissolve significant amounts of lead.
Organic content in acidic waters also increases the plumbosolvency.
The lead content is primarily of interest in potable water supplies because
it has a cumulative poisoning effect in the human body. For industrial uses,
lead is not usually a significant problem.
Lead, especially in small amounts found in waters is usually determined
with dithizone [67]. The lead dithizonate precipitate is soluble in chloroform
or carbon tetrachloride and imparts a red color to the solution. A method
using a preliminary dithizone extraction and reextraction into an acidified
aqueous phase [4] isolates the lead for subsequent determination. Only
bismuth, thallium, and stannous tin interfere, and a procedure is proposed
for the elimination of bismuth and tin interference [4]. The method
determines from .0.01 to 0.05 mg of lead.
Before the introduction of dithizone, small amounts of lead were deter-
mined by comparing colloidal precipitates of lead sulfide with standards
similarly prepared. Electrolytic methods for small amounts of lead have also
been used, but they do not compare favorably with dithizone.
Polarigraphic methods for low concentrations of lead have been de-
veloped. Ullman [68] and co-workers devised a procedure using a quiet
mercury pool cathode cell to determine down to 0.01 mg/litre lead in waste
waters. The method is also capable of determining like concentrations of
cadmium, copper, nickel, and zinc on the same sample.
Manganese
Manganese occurs in soil and rocks as manganese dioxide and can be
dissolved in natural waters by the action of bacteria under anaerobic
conditions [69]. The presence of complexing agents such as carbon dioxide,
tannins, and lignins accelerate the rate of manganese concentration. This

dissolved manganese is readily oxidized to manganese dioxide by air, and is
objectionable because of the formation of a brown manganese dioxide
precipitate. If process water is not treated for removal of manganese, this
precipitation can take place in the system, giving rise to staining and other
objectionable characteristics, much the same as those of iron. The function
of manganese in boiler waters has not been thoroughly investigated, but it
may be anticipated to accelerate corrosive tendencies which may be present.
In cooling systems, manganese has been shown to cause corrosion reactions
in copper alloy tubing.
ASTM Method D 858 uses persulfate to oxidize manganese ions to
permanganate; the purple color produced is the basis for determining the
concentration. The method is capable of determining both soluble and
insoluble manganese. Provisions are made for elimination of interferences
from organic matter and halogens. The method also includes corrective
procedures to compensate for other color-producing ions. Similar methods
exist using perioxidate as the oxidizing agent.
A method [70] for manganese using tetrabase (4-4'-tetramethyldiamino-
triphenylmethane) can be useful. While not as accurate, and subject to more
interferences than the persulfate method, it has the advantage of greatly
increased sensitivity.
Formaldoxime has been used to produce a reddish-brown color with
manganese in alkaline media, and is the basis for colorimetric methods [77].
Interference by iron and copper is eliminated by complexing with cyanide in
the presence of tartrate, and by removing the excess cyanide with zinc. Zinc
and many other metals do not interfere.
Nickel
Nickel is very seldom found in natural waters, but has become significant
in industrial supply or process waters because of the increasing use of
stainless steels and other nickel-containing alloys in water systems. Aside
from the presence of nickel as an indicator of corrosion-erosion of such
alloys, this determination is of importance in water chemistry for high-
pressure boilers. Nickel behaves much like copper in this environment and
may be precipitated as the oxide or as metallic nickel. Its possible role in
corrosion of other metals has not yet been elucidated.
In ASTM Methods D 1886 nickel is determined with both sodium
diethyldithiocarbamate and dimethylglyoxime. Method A, using sodium
diethyldithiocarbamate, is subject to more interferences than Method B,
using dimethylglyoxime, but is more sensitive, being able to detect 0.001
mg/litre. The nickel diethyldithiocarbamate is extracted with carbon tetra-
chlorid,e in the presence of citrate, as in the copper salt. A correction for
copper is made by reading the absorbance of the sample at two wavelengths,
one the maximum for nickel and the other the maximum for copper.
232 MANUAL ON WATER
Method B depends on the formation of a wine-red complex of nickel with

ammoniacal dimethylglyoxime. While not as sensitive as Method A, it has
the advantage of fewer interferences and is a more rapid method, being able
to determine from 0.01 to 5.0 mg/litre nickel.
Heptoxime [4] has been used as a colorimetric reagent in determining
nickel. For use of this reagent, iron and copper are first removed by
precipitating with cupferron and extracting with chloroform. The nickel is
then separated from other interferences by extracting the nickel heptoxime
with chloroform. After separation, the nickel is re-extracted into the
aqueous phase with hydrochloric acid, where the color is developed with
heptoxime. Nickel in the range of 0.05 to 0.25 mg is determined.
A derivative of heptoxime, 4-isopropyl 1,2-cyclohexanedionedioxime, has
been reported [72] as the most sensitive of the dioxime-type reagents for the
spectrophotometric determination of nickel. The use of this reagent was
further improved by extraction of the nickel salt into xylene [73] so that
sensitivities of 0.005 mg/litre nickel could be obtained. Ratios of iron,
cobalt, and copper to nickel of 20, 2, and 8, respectively, can be tolerated.
Nitrate
Nitrate in low concentrations occurs in most natural waters and ranges in
concentration generally from 0 to 70 mg/litre, the extremes being rare. It is
mostly derived from nitrogenous organic matter of animal origin; decay of
vegetable matter in soil yields very little nitrate. For this reason, significant
concentrations of nitrate in water may be indicative of water sources
polluted at some earlier time with sewage or animal wastes. The strata
through which the water percolates can also be a source of significant
amounts of nitrate. Nitrate in water generally is of serious concern only
where the supply is used for potable purposes. Nitrate in concentrations
above 20 mg/litre has been reported as the cause of methemoglobinemia in
infants, a disease characterized by blood changes and cyanosis, in which the
hemoglobin apparently becomes incapable of transporting oxygen.
Nitrates find applicability in water treatment in corrosion-inhibitor
compounds for cooling systems; in low-pressure boilers, nitrate alone or in
conjunction with certain organic compounds gives effective control of
intergranular corrosion (embrittlement) of boiler metal.
ASTM Method D 992 is a rapid, simple, and precise colorimetric
procedure. A yellow color is developed in the sample by the reaction of
nitrate ion with concentrated sulfuric acid and a chloroform solution of
brucine alkaloid. An accuracy of 0.5 mg/litre is obtainable in the 0 to 50-
mg/litre range.
One method [74] purported to be very reliable is based on the formation of
4-nitroso-2,6-xylenol by reacting the nitrate in the sample with 2,6-xylenol in
a highly acidic medium, in the presence of ammonium chloride. The
concentration of the 4-nitroso-2,6-xylenol which is directly related to the
nitrate content is determined by reading the optical density at 304 nm.

Provisions are made for removal of interference from nitrate and ferrous
ions; several other ions investigated do not interfere. The method determines
0 to 50 mg/litre of nitrate-N with a sensitivity of 0.13 mg/litre.
Another method [75] that is simple and rapid is based upon the ultra-
violet absorbance of nitrate. The sample is filtered and acidified, and then
read at 220 nm and 275 nm. Nitrate absorbs at 220 nm but not at 275;
absorbance at 275 nm is traceable to organic interference and is used to
calculate the appropriate correction. Nitrite, chromate, and detergents
interfere; corrections for these can be applied, but they make the test more
time-consuming and less accurate.
Much-used colorimetric procedures based on the yellow color developed
in reacting phenoldisulfonic acid with nitrate ions have been found by some
investigators to be less reliable than the brucine alkaloid method, although
much work has been done to improve this method. Besides being subject to
interference from chloride, it has been shown \76\ to give low results in the
presence of ammonium ions. The sample can be treated to eliminate these
interferences, but special precautions must be taken to avoid loss of nitrate.
If both chloride and ammonium interferences are present, further sample
treatment complications arise. In addition, this approach is not satisfactory
for samples containing organic matter.
Nitrite
Nitrite in natural water is generally regarded as evidence of sewage
pollution. Its presence indicates that nitrogeneous organic matter from
sewage or animal wastes is undergoing active oxidation or nitrification and
that the process is not complete. The occurrence of nitrite is also possible if
nitrate-containing water supplies come in contact with certain reducing
materials.
Nitrites are occasionally used for corrosion control in closed cooling
systems. Discharge of nitrite-treated cooling water into surface supplies may
create pollution problems, or be mistakenly interpreted as indicative of
sewage dischargers. When sampling waters for nitrite analysis, the bottle
should be sterile, and the determination should be made as soon as possible
after collection of the sample.
The ASTM Test for Nitrite Ion in Undustrial Water (D 1254) provides
two procedures for determining nitrite ion. The most accurate method is
colorimetric and is based upon the diazotization of sulfanilic acid by the
nitrite ion, followed by coupling with alpha-naphthylamine. The reaction
takes place at a pH of 2.0 to 2.5 to produce a reddish purple color which can
be measured photometrically at 520 nm or by visual comparison. Its
application is to samples that contain less than 0.5 mg/litre nitrite and are
not highly polluted. This range can be extended by dilution. A thorough
discussion of interferences is also included.
234 MANUAL ON WATER
The alternative method is a titrimetric procedure for samples with a nitrite

content of 30 to 100 mg/litre. It is primarily applicable to systems having a
high nitrite content, particularly those where nitrite is added as a corrosion
inhibitor. The method is based on the reduction of permanganate by nitrite
ions. A known volume of acidified standard permanganate solution is
reacted with the sample, the excess permanganate reacted with a known
excess of standard sodium oxalate, and the excess oxalate back-titrated with
standard permanganate. Reducing agents obviously interfere. Other diazo-
tization-coupling procedures have used sulfanilamide and Af-(l-naphthyl)
ethylenediamine hydrochloride [77] and sulfanilic acid and TV-(l-naphthyl)
ethylenediamine hydrochloride [78}.
Another procedure [79] which determined nitrite plus total nitrate and
nitrite is very similar to the brucine alkaloid test for nitrate. It was found that
at sulfuric acid concentration below 25 percent, only nitrite reacted with the
brucine alkaloid to give the yellow color. Thus, a method was devised based
on the brucine color development in approximately 17 percent sulfuric acid
followed by further development in 50 percent acid, which gave nitrite values
directly and nitrate by difference. The method is fast and convenient, giving
results in the range from 0 to 1 mg/litre nitrogen with 2 percent accuracy.
Nitrogen
Dissolved nitrogen is not ordinarily of interest in industrial water. In
pressurized water nuclear power reactors, the concentration of dissolved
nitrogen in the primary water is sometimes important because of the
possibility of radiation-induced reactions of nitrogen with dissolved oxygen
to produce nitric acid, or with dissolved hydrogen to produce ammonia.
These reactions can produce wide variations in the pH of the primary water.
Nitrogen has been determined with the gas transfer and Blacet-Leighton
apparatus. Dissolved oxygen and hydrogen are removed from the total gas,
and the remainder reported as inert gas, most of which is nitrogen.
Oxygen
Oxygen is dissolved in varying amounts in most waters. It is essential to
the life of fish, and other aquatic organisms. In pure water at room
temperature, oxygen dissolves to concentration levels of approximately 8.5
mg/ litre. The solubility decreases as the dissolved solids and temperature
increase, but increases with increasing pressure. In industrial waters, dis-
solved oxygen is primarily of interest because of its relation to corrosion
reactions. Minute amounts of oxygen in boiler water and feedwater can
cause severe corrosion. Experience has shown that the dissolved oxygen
content should be below 0.03 mg/litre, and preferably below 0.005 mg/litre
in feedwater for high-pressure boilers.
In other waters, dissolved oxygen may not be the direct cause of
corrosion, but it will greatly enhance other corrosive tendencies. An example

is the corrosion of copper and copper alloys by ammonia. The corrosivity of
carbon dioxide is also greatly enhanced by the presence of oxygen.
The methods in ASTM D 888 are applicable to industrial waters and have
a wide range of usefulness. The basic approach common to precision
Methods B and C, and to routine Methods D and E, all depends on the
quantitative oxidation of manganous hydroxide by dissolved oxygen and the
subsequent quantitative release of iodine by the reaction of the resulting
manganic compound with potassium iodide. Special sampling equipment is
required and the foregoing reactions are carried out before the sample is
exposed to the air.
The differences in the four methods—B, C, D, and E—relate to the
manner in which the resultant iodine is titrated. A potentiometric titration
with dilute sodium thiosulfate is the most sensitive and accurate. The less-
precise methods use starch to detect the end point, and different concentra-
tions of sodium thiosulfate for different concentration ranges of dissolved
oxygen.
Some colorimetric procedures have also been devised for determining
dissolved oxygen. Among these are precision Method A of ASTM D 888,
which requires only one reagent, an indigocarmine dye [80, 81], and is used to
determine dissolved oxygen at from 0 to 50 /zg/litre. Another [82] uses leuco
methylene blue, which is oxidized by the dissolved oxygen to methylene
blue. Concentrations down to 2 /ug/litre can be detected.
Four methods for determining dissolved oxygen in industrial waste water
are incorporated in ASTM D 1589 and are designed to cope with various
types of interferences. Included are the Alsterberg (azide), Rideal-Stewart
(permanganate), the Pomeroy-Kirschman-Alsterberg modification, and a
polarographic method. ASTM Methods D 1589 discuss the appropriate
application of each approach.
Instrumental methods for continuous monitoring and recording of dis-
solved oxygen are available. The required equipment is relatively expensive,
but can be justified where the need for continuous measurement is great and
where qualified personnel are available for maintenance of the equipment.
Phosphate
Phosphate occurs in certain rocks and sands, municipal sewage, animal
manure, and in soil, and consequently may be expected in water. Aside from
naturally occurring phosphates, additional phosphates eventually are added
to water supplies through agricultural runoff, and the widespread use of
phosphate-built detergents. In groundwaters, phosphate is found in minute
quantities, the amount dependent upon the original pollution of the water
and strata through which it has percolated. This suggests the utility of
phosphates as additives to underground waters, to conveniently trace their
flow.
236 MANUAL ON WATER
Phosphates and polyphosphates are used extensively in the treatment of

boiler water. Metaphosphates are used as stabilizer additives in water supply
systems, and also as corrosion inhibitors and scale preventives in once-
through and recirculating cooling systems. A film of iron metaphosphate is
deposited on the metal surfaces of dynamic systems, inhibiting corrosion. In
cooling systems, reversion of metaphosphates and polyphosphates to ortho-
phosphate often occurs rapidly, for several reasons. This leads to precipita-
tion of calcium phosphate, and the formation of fouling deposits.
The phosphate ion is important in treatment of feedwaters and boiler
waters. Phosphates are used almost universally as scale preventing chemicals
in boilers operating above 2758 kPa (400 psi). Phosphate is preferable to
soda ash in this pressure range because it is stable, while the carbonate partly
decomposes to produce hydrate alkalinity and releases additional carbon
dioxide into the steam. Calcium hardness entering a phosphate-treated
boiler is precipitated as an alkaline calcium phosphate (hydroxyapatite)
when proper hydrate alkalinity is maintained. Incoming mangesium hard-
ness precipitates as magnesium hydroxide. Both these precipitates are soft,
only slightly adherent sludges when compared with other forms of precipita-
tion. Even in boilers where the incoming calcium and magnesium from
feedwater are practically zero, phosphate excess often is maintained in the
boiler water as a guard against unexpected contamination, or hardness
pickup from condenser tube leakage.
Phosphates usually are injected directly into boiler drums. If phosphate
treatment must be added to the feedwater ahead of the boiler, meta-
phosphate usually is selected, and feed methods carefully controlled to
prevent hydrolysis of polyphosphate to the ortho form in the preboiler
system. At the lower temperatures encountered in feedlines and econo-
mizers, and in the absence of optimum alkalinity conditions, calcium
phosphate can form as an adherent scale, instead of the soft sludge
developed in the boiler at the much higher temperatures.
It can be seen that, depending upon the application, analysis of ortho-
phosphate and determination of the ratio of orthophosphate to meta- or
polyphosphate are of considerable importance. ASTM Methods D 515
provide four approaches to the determination of phosphate. The application
and suitability of the various methods are indicated in Table 7. Only Routine
Method B for phosphate does not provide for determining phosphorus
present in the more complex phosphate species such as tripolyphosphate,
metaphosphates, and pyrophosphate. When it is essential that all of the
phosphorus be determined, conversion of the complex phosphates to
orthophosphate is made by boiling an acidified sample as described in
Precision Method A and in Routine Methods A and C. A more rapid
procedure for converting polyphosphates [83] has been reported and
adopted in Routine Method A. The difference between the orthophosphate
concentration before and after such conversion is a measure of poly-
phosphate phosphorus.
TABLE 7— Phosphate methods.
ASTM Method
D 515 Gravimetric Method Routine Method A Routine Method B Routine Method C
Basis Of measurement gravimetric volumetric colorimetric colorimetric
Test for total PO4^ ortho- and total PO4S ortho-PO4s ortho- and total PO4S
Range 10 mg or more of PO+F ~~* 0.1 mg/ litre or more of 2 to 25 nig/ litre 0.1 to 1.5 mg/ litre (can be
PO4F extended)
Precision 3 percent for 10 mg; 1 percent Approximately 2 percent 2 percent difference between varies with quantity deter-
for 50 mg duplicate determinations mined. Generally good
Application all industrial water all industrial water boiler water and boiler feed- relatively uncontaminated
water; suitable for routine waters
analysis
Interferences phosphite: converts to PO4=; same as gravimetric method strong color; High concentra- above 0.5 mg/ litre iron, 25
As and| V: precipitate with tion of ferric iron mg/ litre SiO2, 1000 mg/-
PO4=;must be removed HC1, litre Cl~
H2SO4, HF, and NH 4 + salts oxidizing agents bleach blue
thereof: retard precipitation color
of ammonium phospho-
molybdate
Summary of method successive separations of o- o-phosphate reacts with am- measurement of the yellow o-phosphate is reacted with
phosphate are made by pre- monium molybdate in acid color of phospho-vanado- ammonium molybdate to
cipitating as ammonium medium to form a molyb- molybdic acid which is de- form phosphomolybdate,
phosphomolybdate and then dophosphate, which is re- veloped in the sample which in turn is reduced
as magnesium ammonium duced to molybdenum blue with SnCl2 solution to
phosphate; ignited and complex w i t h amino- molybdenum blue
weighed as magnesium pyro- naphthol-sulfonic acid.
phosphate Color intensity is propor-
tional to phosphate con-
centration. Bismuth salt
added to sulfuric acid re-
agent intensifies blue color
fourfold
Instrumentation analytical balance none spectrophotometer or filter visual color comparison;
photometer for range 400 spectrophotometer or filter
to 420 nm photometer at 650 nm
238 MANUAL ON WATER
Frequently it is desirable to estimate the concentration of a particular

species. Tripolyphosphate and pyrophosphoric acid may be determined by a
combined titration and gravimetric determination of pyrophosphate. Ortho-
and metaphosphates do not interfere with this method. Both Bell [84] and
Jones [55] have reported methods for determination of metaphosphates in
the presence of the other phosphate species.
Silica
Silica concentration can be a very important consideration in some
industrial installations such as steam generation and cooling-water systems.
The silica is brought into these systems in the makeup (raw or partially
treated) water, or by airborne dusts. Natural (raw) waters may contain more
than 100 mg/litre of silica in solution. The silica and other materials in the
makeup water will concentrate in the boiler or cooling water system due to
evaporation, and form deposits when solubility limits are exceeded. It is
therefore important to determine that the silica concentration in the boiler or
cooling water increases in proportion to the concentrations of other
constituents present in the makeup water. If the silica in the boiler or cooling
water is less than expected, the formation of siliceous deposits in the system
may be indicated.
Silica is of special importance in steam generation, because under certain
conditions it can form troublesome scales of silica and complex silicates in
the boiler, or siliceous deposits in turbines. Because of the very low heat-
transfer characteristics of these materials, extremely thin layers can cause
metal overheating, corrosion, and failure of boiler tubes, as well as reduced
heat transfer. The silica in high-pressure boilers can be carried over with the
steam. As the boiler pressure increases above 2758 kPa (400 psi), the
tendency for silica to be selectively carried into the steam increases; with
constant pH and boiler pressures, carry-over is proportional to the amount
of silica in the boiler water. As pressure rises, the ratio of silica in steam to
silica in boiler water increases rapidly. This silica deposits hard, glassy scales
on turbine blades, resulting in pluggage of small passages and imbalance of
the rotor. Investigations have shown that the key to minimizing silica scales
is to keep the silica in the boiler water at a sufficiently low level to maintain
silica in the steam at not more than 0.02 mg/litre [86\.
Silica can also cause formation of deposits in cooling-water systems, but
here it usually is considered of minor importance unless the concentration is
beyond the solubility limit (usually 150 mg/litre or higher), or if the
magnesium content and pH value are high.
In other industrial applications, silica may be added as part of the
treatment program (example: corrosion control in potable water distribution
systems). Silica analysis is frequently used to maintain proper treatment
levels in systems of this type.
Procedures for the determination of silica are included in ASTM Methods
D 859. Table 8 gives a summary of the important aspects of the methods,
and will aid in the selection of the proper procedure applicable to specific
problems.
Colorimetric Method B of Method D 859 has been suggested for the
determination of silica in high-pressure boilers, where silica, even in very low
concentrations, is of high importance. Special precautions [87] must be
taken during sampling to avoid loss of silica to the sampling equipment. In
sampling steam, errors are said to result if the pressure of the steam is
reduced ahead of the condenser.
Sulfides
Sulfide is present in water generally as the result of industrial waste
contamination or decomposition of sulfate by anaerobic bacterial action.
Sulfide can be very damaging to equipment and industrial processes. In
relatively high concentrations, it will attack normally resistant alloys such as
stainless steel and brass and cause rapid deterioration. Sulfide can be
substantially removed by acidification followed by aeration. Complete
removal can be accomplished by subsequent chlorination.
Hydrogen sulfide is a toxic gas with a characteristic offensive odor and
taste, and is fairly soluble in water. In water it reacts as an acid and can be
actively corrosive. Under certain conditions in the presence of oxygen, it
may even be oxidized to sulfuric acid. In addition, hydrogen sulfide reduces
and precipitates constituents of many cooling-water treatment chemicals,
causing unwanted deposits. Because of these objectionable characteristics, it
is desirable to remove it from process and potable waters.
Hydrogen sulfide is found occasionally in surface waters, resulting from
miscellaneous pollution or anaerobic decomposition of organic compounds.
It occurs most commonly in well waters located near "sour gas" oil fields, or
as a result of anaerobic bacterial reducing action on inorganic sulfur
compounds such as sulfate. It may also be produced in soils by bacterial
action, and is a consideration in the corrosion of buried metal piping.
Hydrogen sulfide may be produced in sodium sulfite-treated boiler waters
with operating temperatures of 274° C (525° F) and above. On high-
temperature surfaces (superheaters, etc.), the sulfite breaks down to sodium
sulfide and sodium tetrathionate. The sodium sulfide is then hydrolyzed to
hydrogen sulfide and passes into the steam, where it is very corrosive.
Quantitatively, and semiquantitatively, odors indicating hydrogen sulfide
in water can be confirmed with lead acetate paper, which is quickly darkened
by exposure to the vapor from a slightly acidified sample. In ASTM Method
D 1255, hydrogen sulfide is determined as part of the total and dissolved
sulfides, and the concentration can be calculated from the concentration of
dissolved, sulfides and the pH of the sample.
It is best to analyze sulfide-containing samples immediately. When this is
not possible, the sulfide content should be fixed by adding a suitable
TABLE 8—Silica methods.
ASTM Designation
D 859 Precision Method A Precision Method B Routine Method C Routine Method D
Test for total silica (colloidal and crys- crystalloidal silica crystalloidal silica crystalloidal silica
talloidal). When used with
Method B, C, or D, colloidal
silica can be determined by
difference.
Range as low as 1 mg/ litre below 1 000 Mg/ litre below 10 mg/ litre below 10 mg/ litre
Precision 0.4 mg/ litre 1.3 jug/ litre + 3 percent of 0.1 mg/ litre 0.1 mg/ litre
concentration determined
Type of water raw water, cooling water, boiler steam condensate, ion ex- raw water, cooling water, boiler raw water, cooling water, boiler
water, waste water change bed, nuclear reactor water, waste water water, waste water
Interference no interference no interfering material present turbidity, color (tannin, Cu+ same as C, except tannin does
jon), chromate ion inter- not interfere
ferences
Applicable to need for total silica content system containing very low rapid control analysis where rapid control analysis where
silica no color is present due to color is present due to
tannins or other organics. tannins
Instrumentation none spectrophotometer 815 nm or same as B or use matched same as C
filter photometer 640-700 Nessler tubes or Duboseq-
nm or 815 nm type color comparator
Theory silicon compounds dissolved or reactions of soluble silica with (a) utilizes yellow silicomolyb- in the presence of a reducing
suspended in water are concen- molybdate ion, blue color de- date color complex for com- agent such as sodium sulfite,
trated and precipitated as veloped with l-amino-2- parison or (b) alternative, to soluble silica reacts with mo-
partially dehydrated silica by naphthol-4-sulfonic acid. increase sensitivity and lybdate ion to form a blue
evaporation with either per- The color is read photo- eliminate color interference, complex. The color is read
chloric or hydrochloric acid, metrically and compared l-amino-2-naphthol-sulfonic photometrically and com-
followed by HF treatment with standards acid is added providing blue pared with standards
color for comparison
measurements
quantity of zinc acetate or cadmium chloride solution to precipitate zinc or

cadmium sulfide.
Frequently, because of the presence of sewage or interfering components,
hydrogen sulfide is separated by evolution in a stream of inert gas and
collected in a zinc acetate or cadmium chloride solution before making a
determination. Tests have shown that the segregation of hydrogen sulfide by
inert gas transfer is not quantitative and that the error is large when the
concentration is low.
ASTM Method D 1255 determines total and dissolved sulfides. This is a
colorimetric procedure based on the fact thatp-aminodimethylaniline, ferric
chloride, and the sulfide ion react in acid solution to produce methylene
blue. The concentration range of the method is 0.1 to 20 mg/ litre of sulfide,
and the accuracy is ±0.1 mg/litre for clear, colorless solutions and ±15
percent for colored and turbid solutions.
Widely used titrimetric procedures for sulfide depend upon the reducing
action of sulfide on a standard iodine solution in an acidified medium. The
excess iodine is back-titrated with standard sodium thiosulfate, using starch
as an internal indicator. Other reducing agents interfere.
An amperometric back-titration has been suggested by Goldman [88] and
is an adaption of the phenylarsenoxide titration for chlorine (see ASTM
Methods D 1253 for chlorine). Two 10-ml aliquots of standard iodine
solution are added and titrated successively with 0.00564 N phenylarsen-
oxide solution on the same sample at a pH of 4. The difference between the
two titrant volumes is equivalent to the sulfide present. From 0 to 0.2
mg/litre sulfide may be determined in this manner. It is stated that the range
can be extended by adjusting sample volumes and reagent concentrations,
and that color and turbidity should cause no interference.
Sulfate
Sulfate ions are relatively abundant in natural waters, especially in hard
waters. In industrial waters, sulfate is of interest mainly where evaporative
processes introduce potential scaling problems and where control of total
dissolved solids is a major consideration. Sulfate often is a constituent of
boiler deposits and heat exchanger scales, softening and other treatment
procedures being required for deposit control. The determination of sulfate
for regulation of allowable concentrations in boiler water and in cooling
towers (where sulfates often are increased by addition of sulfuric acid for
alkalinity reduction and pH control) can be useful. The determination of
sulfate is also useful in the estimation of total dissolved solids.
Sulfate concentration in potable water supplies also deserves considera-
tion. High amounts of sulfate impart a faintly bitter taste to the water, and
cause laxative effects, especially on children. While the human body usually
can adjust to relatively high sulfate waters, certain environmental conditions
may indicate their use to be inadvisable.
242 MANUAL ON WATER
ASTM Methods D 516 provide three tests for determining the sulfate
concentration of water samples. These methods and their applicability to
various types of water are listed in Table 9.
There are several other titrimetric procedures for determining sulfate. In
the benzidine procedure, an addition product of benzidine and sulfate is
precipitated from an acidified solution, washed, redissolved in hot water,
and titrated with standard sodium hydroxide. This method, while not highly
accurate, is especially useful for determining sulfate in the presence of
relatively high amounts of sulfite. In another procedure [89,90], the sulfate is
precipitated with barium chloride, filtered, and the barium sulfate re-
dissolved in a known volume of 0.05 M EDTA. The EDTA is then back-
titrated with standard barium chloride solution using a suitable metal
phthalein as indicator. Several other variations of this approach exist.
Sulfite
The occurrence of sulfite salts in natural waters is rare, because sulfite is so
readily oxidized by air. Because of this ease of oxidation, sulfite finds
widespread application as an oxygen scavenger in boiler waters to prevent
corrosion. At high temperature and boiler pressure, sulfite can break down
to yield sulfur dioxide, and, in the presence of alkali, hydrogen sulfide to the
steam. In such instances, the sulfite concentration must be carefully con-
trolled. It is desirable to limit the use of sulfite to as low a concentration as
practicable for complete oxygen removal, including necessary residuals.
Sulfur dioxide is formed when sulfur-containing compounds are burned
or when certain metallic sulfides are strongly heated in air. It occurs
naturally in the gases emanating from volcanoes and fumaroles and is
present in the air surrounding manufacturing areas or cities, because of the
combustion of fuels and other sulfur-containing materials. It is readily
soluble in water, where it becomes sulfurous acid or salts thereof, and has
been found in hail, snow, and especially frost in the neighborhood of cities
and towns. Sulfurous acid has also been found in numerous mineral waters
and in springs near active volcanoes. In other naturally occurring waters,
both dissolved sulfur dioxide and sulfite, because they are readily oxidized
by the oxygen in the air, are unlikely to occur except in very special cases.
Sulfur dioxide has many uses in manufacturing, but in water and waste
treatment its use is confined mostly to dechlorination procedures and to
chromate-reduction waste treatment applications. In dechlorination, sulfur
dioxide reacts with chlorine to form sulfuric acid and hydrochloric acid,
both of which are neutralized by the natural alkalinity of the water being
treated.
Total sulfur dioxide in water includes free sulfur dioxide, bisulfite, and
sulfite ions. Like carbon dioxide, the relative concentrations of the three
ionic species depend upon the pH of the solution. The determination of
sulfur dioxide usually is based upon its reducing properties in solution, the
same approach used for determining sulfite ions.
TABLE 9—Sulfate.
ASTM Method D 516 Precision Method Routine Method A Routine Method B
f
Basis of measurement gravimetric turbidimetric volumetric
Range 20 to 100 mg/ litre SO4'=(can be 10 to 100 mg/ litre SO4= 5 to 1000 mg/ litre SO4=
extended)
Precision 1 percent of amount present as good as 5 percent or 2 mg/ litre, which single operator — 0.5 mg/litre
ever is greater. Depends on inter-
ferences and skill of analyst.
Application all types of industrial water and rapid routine testing of industrial rapid determination of SO4= over wide
industrial waste water waters range in industrial water for routine
Summary of method sulfate ion is precipitated and sulfate ion is precipitated as barium or control tests
weighed as barium sulfate after sulfate and a suspension made sulfate ion is titrated in alcoholic solu-
removal of silica and other in- Glycerin and sodium chloride are addedtion with standard barium chloride
soluble matter. to stabilize suspension and minimizeusing thorin as the indicator. End
interferences point change: yellow to pink, rela-
Interferences SOf? may oxidize and precipitate insoluble material — must be removed tively sharp
S= \ SO4'= dark color cations: removed by ion exchange
water quality affects suspension of above 2 mg/litre F~
barium sulfate 50 mg/litre NO;T
2 mg/litre PO 4 = /can be elimi-
SOi", CrO4=, S = Jnated
1000 mg/litre Cl at low SO4= concen-
tration
Instrumentation or special analytical balance photometer suitable for measuring microburet
equipment between 350 and 425 nm ion exchange column
cell path of 40 nm magnetic stirrer helpful
244 MANUAL ON WATER
ASTM Methods D 1339 list three volumetric methods for the deter-
mination of sulfite in industrial waters. Two precision methods, A and B, fix
the sulfite in the sample by adding hydrochloric acid, potassium iodide, and
potassium iodate before it is exposed to the air. The iodine chloride formed
reacts with the sulfite present, and the excess is then back-titrated with
sodium thiosulfate.
Method A uses a special sampling apparatus to exclude air and allow
addition of the fixing reagents before the sample is exposed to the air. After
fixing, the sample is drained into a beaker and titrated using an electrometric
dead-stop apparatus. The technique allows determination of 0.0 to 6.0
mg/litre sulfite as Na2SO;i.
Method B is a simplification of Method A. The sample is collected in a
manner to exclude air, fixed immediately upon opening, and titrated to a
starch indicator end point. It is designed for 6 mg/litre Na2SO? and up.
Routine Method C is a direct titration of the sample with potassium
periodate after addition of hydrochloric acid and potassium iodide. It is
designed for control tests where good precision is not required and where the
sulfite concentration exceeds 3 mg/ litre.
Sodium and Potassium

Sodium and potassium are found in practically all waters. The ubiquitous
sodium is of importance when salinity or total dissolved solids are a
consideration in the use of the water. Sodium also is important in boiler
waters, where it constitutes a large portion of the cations present. At one
time, there was a trend toward the use of potassium salts for hideout control
in boiler water treatment, but this trend has slackened and potassium now is
of lesser importance. Sodium concentrations in boilers are easily removed by
blowdown, and sodium salt deposition rarely is encountered, except in areas
of water starvation or under hideout conditions.
The analysis for sodium sometimes can be used as an indication of the
purity of water; for example, in steam condensates, the sodium concentra-
tion can be used to indicate water carry-over from boiler system into the
steam. In demineralizer effluents, sodium determinations will indicate the
completeness of cation exchange.
Sodium and potassium can be determined gravimetrically or by a flame
photometer. The flame photometer (ASTM Methods D 1428) provides by
far the simplest and most rapid method for potassium and sodium in low
concentrations. Method A, using either direct-measuring or internal-stan-
dard flame photometers, is especially recommended for concentrations of
sodium and potassium between 1 and 100 mg/ litre. Method B covers the
range from 0 to 1 mg/litre, using the direct-intensity method with either type
of instrument, and is applicable to high-purity waters. Special precautions
are required to avoid exposing the instrument to dust, drafts, and unclean air
when working in these low concentrations. Method C is designed for
analysis of deposits and is satisfactory for sodium contents of 0.1 percent or

more.
Glass electrodes sensitive to sodium ions have been used to determine
sodium concentrations electrometrically, with direct-reading instruments.
The sodium electrode also is sensitive to potassium and hydrogen, but much
less so, and their effect can be minimized. Bower [91] used such an electrode
to determine sodium in irrigation waters. The effect of hydrogen ions was
eliminated by buffering, and that of potassium ions by precipitating with
tetraphenylborate. Very good agreements with flame photometer deter-
minations were found.
Zinc
Zinc is found in small quantities in most industrial waters, resulting from
corrosion of galvanized iron and brass in condensing, cooling, and dis-
tributing systems. A determination for zinc may be made as an indication of
deterioration of particular equipment, for identification of the source of this
constituent in water-formed deposits, to determine the concentration of zinc
in systems using zinc salts as corrosion inhibitors, or to determine zinc as a
pollutant in waste water.
The nephelometric procedure [92], in which interfering substances are
removed by hydrogen sulfide and turbidity is then developed with ferro-
cyanide, has been used successfully. The most suitable range for direct
comparison with standards is 1 to 10 mg/litre. Since the zinc content is
concentrated during the procedure, the lower limit in the original sample is
about 0.05 mg/litre and the optimum comparison level about 0.2 mg/litre.
A turbidimetric method using sodium diethyldithiocarbamate also has
been reported. [93, 94].
Zinc may be determined with dithizone (diphenylthiocarbazone). The
reaction product is red when extracted into carbon tetrachloride, and this is
used to measure the zinc content photometrically. There are several pro-
cedures that use various complexing agents to eliminate interferences.
Among these, bis (2-hydroxyethyl) dithiocarbamate is purported [95] to be
the best. Zinc in the amount of 0.01 mg/litre can be determined.
A direct method using zincon to develop a blue color with zinc, in which
heavy metals including the zinc are complexed with cyanide, has been
reported [96]. The zinc is released selectively with chloral hydrate and reacts
directly with the zincon reagent. This procedure is the basis for ASTM
Methods D 1691.
Methods of Reporting Results

The reporting of results is an important part of every analysis. Just as poor
sampling can render an analysis valueless, poor reporting can hamper
interpretation. Refer to ASTM Method for Reporting Results of Analysis of
246 MANUAL ON WATER
Water (D 596) and ASTM Method for Reporting Results of Examination

and Analysis of Water-Formed Deposits (D 933). Results should be re-
ported in a form that is amenable to comparison, and explanatory references
may be necessary to ensure completeness or to detail the meaning of the
data. For example, odor thresholds should be accompanied by a record of
the temperature of analysis. If an analysis for organic extractables is run, it
is necessary to report the solvent used, such as hexane-extractables or chl-
oroform-extractables. Similarly, when analyzing for phosphate, it is possible
to determine total phosphate, orthophosphate, and polyphosphate. When
reporting, it is necessary to indicate the results properly. Any departures in
analytical procedure that were necessitated must be recorded and reported.
If some constituents were not measured directly, this should be noted; for
example, dissolved solids (by difference).
ASTM Method D 596 presents recommendations as to the nature, extent,
and form in which results of analysis should be presented (see Methods D
1192, D 860, D 3370, D 1066, and D 887). It can serve as a guide to accurate
reporting. It also contains a table of conversion factors for converting
milligrams per litre to equivalents per million and vice versa.
Analytical results must be reported in a manner that will not imply a
fictitious significance. If numerical values are improperly reported, they may
lead to serious misinterpretation. The reported value should contain but one
digital value that is not absolutely precise. This digit is the last one to the
right. A weight of 4.7165 g implies a precise measurement accurate to
±0.000«, whereas 4.7 g would be the reported value for weighing the same
object on a single-beam balance accurate to ±0.« g. The value n is used to
designate the least significant figure. The plus-or-minus sign describes a
variance inhererit in the measuring process. The usual way of describing this
variability is by the root-mean-square of the variance, known as the standard
deviation.
In making calculations, the result will not have significance greater than
the least significant value used in the computation. For example, 6.250 times
1.1 equals 6.8750, but must be rounded to 6.9. The method of rounding
dictates that, if values of 5 or greater are dropped, the preceding digit is
raised by one. Values of 4 or less are dropped without any change in the
preceding digit; for example, 1.008 times 4.3 equals 4.3344, which is rounded
to 4.3 in practical application.
Zeros are never added to values to the right of the decimal point if the
precision is not warranted. A temptation often exists to balance the columns
of numbers in tabulating results. For example, assume a water analysis:
pH 6.83
Aluminum as Al, mg/ litre 0.25
Nitrate as NO3, mg/litre 6.5
These values tell the reader something about the methods that were used.
The pH reading of 6.83 indicates the use of a precise potentiometric
technique. A less precise instrument would give a result reported as 6.8,

while a color comparison procedure would probably be estimated as 7. The
accuracy of the aluminum and nitrate methods is 0.01 to 0.5, respectively.
Adding a zero to 6.5 to show 6.50 would be improper.
Methods of Checking Analysis

Before any analysis is finally reported, it should be thoroughly checked for
possible errors. In general, errors inherent to the analytical procedure are
avoided by running replicate samples. However, other errors can occur
which will not be obvious. For example, mislabeling, recording erroneous
aliquots, getting the sample sequence out of order, errors in taking or
recording instrument readings, or misreading calibration curves are all
mistakes which have plagued analysis occasionally. Procedures for dis-
covering such errors are essential.
As mentioned earlier in this chapter, an important part of every analysis is
a description of the physical appearance of the sample. Even though this
description must be brief and limited in information, it can be used as the
first check on the analysis. There must be a correlation between appearance
and the reported results. For example, if 15 mg/litre iron is found, the
physical description generally should indicate presence of a precipitate or
magnetic particles. If neither is present, and the pH is not acidic, an error is
indicated. The intensity of yellow color of a chromate-treated cooling water
can serve as a check on the chromate analysis. If a sample is reported to have
significant amounts of oil present, the description should have noted an oily
film, cloudy appearance, or a large amount of suspended solids. Other
correlations will become obvious to the analyst upon consideration.
When a complete water analysis has been run, an anion-cation check
should be made. Proper balance can be checked by totaling separately the
determined concentrations in equivalents per million of all anions and
cations. Deviation from a balance indicates an error (see ASTM Method D
596).
When it is not practicable to determine all cations (for example, during
field investigations), it is possible to determine remaining cations by
difference, using readily obtained total anions (expressed in equivalents per
million) as the basis. Total calcium and magnesium (expressed in epm)
subtracted from total anions (in epm) will closely approximate sodium and
potassium concentration (also in epm) in the sample. As an alternative,
missing cations can be computed, using hypothetical combinations of
determined anions and cations, and compared with total dissolved solids as
determined by conductimetric or gravimetric procedures. With this method,
calcium (in epm) is combined with epm of anions in the sample, in the order:
carbonate (or bicarbonate), sulfate, chloride, and nitrate, until all calcium
has been combined. Then magnesium (in epm) is combined with equivalent
quantities of remaining anions, in the same order. Any remaining uncom-
bined anions (alkalinity, sulfate, chloride, or nitrate) after all calcium and
248 MANUAL ON WATER
magnesium have been combined may be assumed to exist as sodium or

potassium salts. When equivalents of the hypothetical combinations are
converted to milligrams per litre, the total of all salts so computed should
agree closely enough with determined total dissolved solids to provide a
reasonable check for errors.
The relationship between the various constituents can also serve as a
check. For example, carbonate could not be present in a sample with pH at
4.0 (acid to methyl orange indicator).
Special attention should be given to samples that represent an entire
system such as a boiler system or a cooling system. For example, if samples
of cooling water makeup and circulating water from an evaporative cooling
tower are received, there should be a logical relationship between the two.
The cooling water should be a concentrated solution of the makeup. If the
tower water has been treated with sulfuric acid to reduce alkalinity and
prevent CaCO3 deposition, the hardness and sulfate concentrations in the
circulating water should increase by an appropriate concentration factor.
The calcium concentration may vary slightly with alkalinity changes, but the
magnesium does not; thus, magnesium hardness can be used to calculate the
concentration factor. The chloride content should be similarly concentrated.
By using these concentration factors, a check can be made on the amount of
sulfuric acid used to reduce the alkalinity to the observed value, and
therefore the observed sulfate content. Serious deviation of determined
results from these computed cross-checks indicates possible error in analysis
or variation in system control. When checking any series, similar procedures
can be used. Look for constituents in one sample that should appear all
through the series. Also, concentration or dilution factors should appear in
certain other samples. Cross-checking can reveal inconsistencies and pos-
sible errors which otherwise might not be apparent.
Previous analyses on samples that are run periodically serve as an
excellent check on current results. Large deviations or sudden changes may
indicate a system upset, but also could be the result of analytical error. Such
unexpected results should be checked carefully before reporting.
As one can see, each sample has its own unique origin and set of physical
characteristics. The history and observed properties can assist the analyst in
selecting the proper laboratory procedures for analysis. Once the analysis is
complete, a review of the origin of the sample and its physical appearance
along with cross-checking of the reported results, both within the analysis
and through any series, provides a final check for detection of possible
errors.
pH
The pH of water is a measure of its hydrogen ion activity. It is used to
indicate the acidity or basicity of aqueous solutions, which includes natural
waters, and is the result of the dissociation properties of the solutes. Both
important and useful, pH measurements have widespread application. These

measurements are used in many programs for effective water treatment,
purification, and corrosion control.
ASTM Method D 1293 covers the determination of pH in industrial
waters and industrial waste waters; it is an electrometric procedure using a
glass electrode. The method, along with its Appendix, gives a comprehensive
discussion of the apparatus used, electrode treatment, preparation and use of
reference buffer solutions, standardization, and procedures for pH deter-
mination of batch samples and flowing samples using flow cells.
Electrometric methods of measuring pH are by far the most accurate, but
other methods for approximating pH can be used where precise control is
not necessary. In the routine procedures for acidity and alkalinity in Method
D 1067, indicator dyes are used to show the end points of the titrations to
reference pH levels. The indicators are useful because they change color in
various ranges of hydrogen ion concentrations. There is a large number of
weakly acidic (or basic) organic dyestuffs that undergo such color changes.
The most suitable and accurate indicators are those which change color
relatively sharply within a narrow range of pH. Such dyes make possible
relatively rapid and inexpensive methods for estimating pH. The prepared
sample is read by matching colors with a graduated set of permanent
standards mounted in a slide or roulette wheel, or prepared in the form of
ampoules. Standards are prepared from buffer solutions, or they may be
colored glass or solutions.
Errors in this approach arise from personal differences in color percep-
tion, temperature variations, differences in total salts concentration, chlo-
rine, and specific indicator properties (buffer effect of the indicator).
Dye-impregnated papers also exist for approximating pH levels.
Interpretation of Analytical Results

Any collection of analytical data is meaningless unless an informed and
logical interpretation can be made. Interpretation in many cases is both the
beginning and the end of an analysis. In setting up an analytical program,
the interpretations to be made will influence the choice of determinations.
For example, to determine whether a water treatment program is func-
tioning effectively, not only must the results of the analysis be interpreted in
terms of the system, but also the constituents to be analyzed and the tests to
be run must be chosen with the nature of the interpretation required in mind.
Several factors will influence the interpretation. The type of water and the
system, the number and location of sampling points, the frequency of
sampling and testing, and the sensitivity of the methods employed are
important considerations. A single sample of river water could hardly be
used to interpret the nature of the river as a supply source. No information
would be obtained as to variations at different depths, variations with
respect to location or seasonal changes, or occurrence of periodic con-
250 MANUAL ON WATER
lamination. On the other hand, such an analysis might help determine what
type of analysis is needed, and what sample treatment or precautions are
necessary for future analyses. In the first case, an interpretation that is too
extensive may lead to an erroneous picture and future trouble. As noted,
proper interpretation of a single sample can serve as a useful, although
limited, tool in guiding future work.
The amount of time elapsed between sample collection and the actual
analysis is another important factor to consider when interpreting data.
Chemical and physical changes can take place after a sample is drawn, and
such possibilities must be considered.
Choosing significant data is an important part of interpreting results. In a
well-planned program of analysis, only the significant constituents will be
analyzed. Therefore, in interpreting the results, only the significance of the
concentrations reported and their interrelation remain to be translated into
meaning. Past analysis also will have an influence on interpreting a current
analysis.
An example will help illustrate the interpretation of significant data.
Suppose heat exchangers in an evaporative, open recirculating cooling
system are to be protected from corrosion by controlled deposition of a thin,
impervious layer of calcium carbonate. If the water supply is amenable to
this approach, such deposition can function as an inexpensive but effective
corrosion control procedure. However, close regulation of certain important
constituents must be maintained. If the circulating water is too acidic
(relatively speaking; not necessarily in the free-acid range), corrosive con-
ditions will prevail and the protective scale will be removed, or its formation
prevented. On the other hand, conditions too favorable for scaling will
permit excessive buildup of calcium carbonate deposition. An equilibrium
condition is the desired state. To maintain this equilibrium (in actual
practice, very slightly scaling) several factors must be known and regulated.
These are the calcium concentration, the methyl-orange alkalinity, total
dissolved solids, temperature, and pH. By utilizing Langelier Saturation
Index data (available in the form of nomographs, charts, and tables),
appropriate calculations can be made to determine the scaling or corrosive
nature of the cooling water. Adjustments in scaling or corrosive tendencies
can then be made by altering the alkalinity and pH of the circulating water.
Similar analyses and interpretations must be made on the makeup water
added to the cooling system. In maintaining such a corrosion control system,
frequent interpretations must be made about the significance of changes in
concentration of the various constituents. Each change in concentration
must be considered in relationship to the other constituents. An increase of
20 mg/litre in calcium content of makeup water may or may not signal the
need for control adjustment. If the alkalinity or total dissolved solids also
change appropriately, no adjustment in chemical feed may be necessary. The
importance of proper interpretation, which in this example involves a special
calculation, is apparent. The various constituents are interdependent, and
corrosion control can be maintained successfully only after due considera-

tion of all the influencing factors.
References
[2] "Study and Interpretation of the Chemical Characteristics of Natural Water," U. S.
Geological Survey, Water-Supply Paper 1473, pp. 92-95.
[3] Larson, t. E. and Henley, Laurel, Analytical Chemistry, ANCHA, Vol. 27, 1955, p. 851.
[4] Standard Methods for the Examination of Water and Waste Water, American Public
Health Association, New York, N. Y., 14th ed., 1975.
[5] Lamar, W. L., Analytical Chemistry, ANCHA, Vol. 21, 1949, p. 276.
[6] Palin, A. T., Water and Sewage Works, WSWOA, Vol. 104, 1957, p. 492:
[7] Staats, F. C. in Symposium on Continuous Analysis of Industrial Water and Industrial
Waste Water, ASTM STP130, American Society for Testing and Materials, 1952, p. 33.
[8] Partridge, E. P., Discussion, Proceedings, American Society for Testing and Materials,
ASTEA, Vol. 41, 1941, p. 1297.
[9] Parker, R. O. and Ziobro, R. J. in Symposium on Steam Quality, ASTM STP 192,
American Society for Testing and Materials, 1956, p. 27.
[10] Rivers, H. M., Trautman, W. H., and Gibble, G. W., Transactions, American Society
of Mechanical Engineers, TASMA, Vol. 72, 1950, p. 511.
[//] Sisson, A. B., Straub, F. G., and Lane, R. W. in Symposium on Steam Quality, ASTM
STP 192, American Society for Testing and Materials, 1956, p. 37.
[12] Place, P. B., Discussion, Proceedings, American Society for Testing and Materials,
ASTEA, Vol. 41, 1941, p. 1311.
[13] Gurney, W. B. in Symposium on Steam Quality, ASTM STP 192, American Society for
Testing and Materials, 1956, p. 3.
[14] Rosenthal, Robert, in Symposium on Continuous Analysis of Industrial Water and
Industrial Waste Water, ASTM STP 130, American Society for Testing and Materials,
1952, p. 12.
[15] Nagy, Z. and Polyik, E. N., Hidrologiai Kozlony, HIDRA, Vol. 39, 1957, pp. 243-4;
Chemical Abstracts, CHABA, Vol. 56, 1962, p. 15299 c.
[16] Will F., Analytical Chemistry, ANCHA, Vol. 33, 1961, pp. 1360-1362.
[17] Shull, K. E., Journal, American Water Works Association, JAWWA, Vol. 52, 1960,
pp. 779-785.
[18] Packham, R. F., Proceedings, Society for Water Treatment and Examination, PSWTA,
Vol. 7, Part 2, 1958, pp. 102-126.
[79] Giebler, G., Zeitschrift Fuer Analytische Chemie, ZACFA, Vol. 184, 1961, pp. 401-411.
[20] Knight, A. G., Proceedings, Society for Water Treatment and Examination, PSWTA,
Vol. 9, 1960, pp. 72-80.
[27] Crowther, A. B. and Large R. S., Analyst, ANALA, Vol. 81, 1956, p. 64.
[22] McKinney, D. S., Proceedings, American Society for Testing and Materials, ASTEA,
Vol. 39, 1939, p. 1191.
[23] Nozaki, T., Nippon Kagaku Zasshi, NPKZA, Vol. 80, 1959, pp. 1278-1281.
[24] Freediani, K. A. and Babler, B. J., Industrial and Engineering Chemistry, Analytical
Edition, IENAA, Vol. 11, 1939, p. 487.
[25] Kahler, H. L., Industrial and Engineering Chemistry, Analytical Edition, IENAA, Vol.
12, 1940, p. 266.
[26] Dzhaparidze, E. K., Gigiena I Sanitariya, GISAA, Vol. 24, No. 6, 1959, p. 79.
[27] Debras-Guedon, J. and Voinovich, I., Bulletin de la Societe Francaise de Ceramique,
BFCEA, No. 42, 1959, pp. 43-54.
[28] Nzumasa, Y., Nasu, Y., and Seo, T., Nippon Kagaku Zasshi, NPKZA, Vol. 81, 1960,
pp. 430-434.
[29] Chow, T. J. and Thompson, T. G., Analytical Chemistry, ANCHA, Vol. 27, 1955, p. 18.
[30] Rains, T. C., Zittel, K. E., and Ferguson, M., Analytical Chemistry, ANCHA, Vol. 34,
1962, pp. 778-781.
[31] Nakajima, S., Hinata, Y., and Amano, Y., National Technical Report, NTROA, Vol. 3,
1957, pp. 222-229.
252 MANUAL ON WATER
[32] Kawakami, H., Anken Kiyo, Vol. 10, 1958, pp. 194-199.
[33] Skougstad, M. W., U. S. Geological Survey, Water Supply Paper 1496-8, 1961, pp.
19-31.
[34] Dye, J. F., Journal, American Water Works Association, JAWWA, Vol. 50, 1958,
pp. 800-820.
[55] Kinney, D. S. and Amorosi, A. M., Industrial and Engineering Chemistry, Analytical
[36] Fishman, M. J. and Skougstad, M. W., Analytical Chemistry, ANCHA, Vol. 35, 1963,
pp. 146-149.
[37] Goldman, E. and Byles, D., Journal, American Water Works Association, JAWWA,
Vol. 51, 1959, pp. 1051-1053.
[38] Lei, C. U. and Chou, K. S., Hua Hsuek Skik Chieh, HUAKA, 1959, pp. 127-128.
[39] Drosdoff, N. and Nearpass, D. C., Industrial and Engineering Chemistry, Analytical
[40] Hamaguchi, H. and Hashimoto, J., Nippon Kaguku Zasshi, NPKZA, Vol. 79, 1958,
pp. 2119-2122.
[41] Ogata, H. and Hiroi, K., Bunseki Kagaku, BNSKA, Vol. 8, 1959, pp. 21-25.
[42] Kerr, J. R. W., Analyst, ANALA, Vol. 85, 1960, pp. 867-870.
[43] Winland, G. and Meckenstock, K., Zeitschrift Fuer Analytische Chemie, ZACFA,
Vol. 176, 1960, pp. 96-110.
[44] Akiyama, T. and Yabuuchi, T., Kyoto Yakka Daigaku Gakuho, KYDGA, Vol. 7, 1959,
pp. 57-59.
[45] Graham, D. W., Journal, American Water Works Association, JAWWA, Vol. 35, 1943,
p. 159.
[46] Pieters, H. A. J., Hanssen, W. J., and Geurts, J. J., Analytical Chimica Acta, ACACA,
Vol. 2, 1948, p. 382 (in English).
[47] Goodman, B. L., Water and Sewage Works, WSWOA, Vol. 108, 1961, p. 80.
[48] Kravtsova, N. M., Trudy Komissii Po Analitichesloi Khimii, Akademiya Nauk S. S.
S. R., Institut Geokhimii i Analiticheskoi Khimii, TKAKA, Vol. 8, 1958, pp. 161-168.
[49] Polyanskii, V. N., Sbornik Trudov Moskovskii Vechernii Metallurgicheskii Institut,
No. 2, 1957, pp. 253-354; Referativnyi Zhurnal, Khimiia, Abstract, RZKHA, No. 11,
016, 1968.
[50] Besev, A. I. and Ivaryitin, M. I., Vestnik Moskovskogo Universiteta, Seriya Matematiki,
Mekhanki, Astronomii, Fiziki i Khimii, VMUMA, Vol. 13, No. 2, 1958, pp. 177-181.
[5/] Nemodruk, A. A. and Stasynchenko, V. V., Zhurnal Analiticheskoi Khimmi, ZAKHA,
Vol. 16, 1961, p. 407-411.
[52] Bowles, P. F. and Nicks, P. F., Analyst, ANALA, Vol. 86, 1961, pp. 483-485.
[53] Abbott, D. C. and Harris, J. R., Analyst, ANALA, Vol. 87, 1962, pp. 497-499.
[54] "Methods for Analysis," Journal, Association of Official Agricultural Chemists, JOACA,
6th ed., 1945, p. 638.
[55] ASTM Method E 53, Chemical Analysis of Copper (Electrolytic Determination of
Copper), Annual Book of ASTM Standards, Part 32, American Society for Testing and
Materials, 1969.
[56] Bellach, E. and Schoubae, P. J., Analytical Chemistry, ANCHA, Vol. 30, 1958, p. 2032.
[57] Yamamura, S. S., Wade, M. A., and Sikus, J. H., Analytical Chemistry, ANCHA,
Vol. 34, 1962, pp. 1308-1312.
[55] "Power Station Chemistry," Publication H-ll, Edison Electric Institute, 1960.
[59] Rummel, J. K. in Symposium on Continuous Analysis of Industrial Water and Industrial
Waste Water, ASTM STP130, American Society for Testing and Materials, 1952, p. 42.
[60] Moss, N. L. and Mellon, M. G., Industrial and Engineering Chemistry, Analytical
[61] Swank, H. W. and Mellon, M. G., Industrial and Engineering Chemistry, Analytical
Standard Methods for the Examination of Water and Waste Water, American Public
Health Association, New York, N. Y., 12th ed., 1965, pp. 154-163, 482-485.
[62] Woods, J. T. and Mellon, M. G., Industrial and Engineering Chemistry, Analytical
[63] Andres, R. F., Proceedings, 14th Annual Water Conference, Engineering Society of
Western Pennsylvania, PEWPA, 1953, p. 106.
[64] Herre, E., Brennstoff- Wareme-Kraft, BRWKA, Vol. 13, 1961, pp. 456-459.
[65] Nakashima, F. and Sakai, K., Bunseki Kagaku, BNSKA, Vol. 10, 1961, pp. 89-93.
[66] Grat-Cabanca, M., Analytica Chimica Acta, ACACA, Vol. 19, 1958, p. 108 (in French).
[67] Sandell, E. B., Colorimetric Determination of Traces of Metals, Interscience, New York,
3rd ed., 1959, Chapter 23.
[68] Ullman, W. W., Pfeil, B. H., Porter, J. D., and Sanderson, W. W., Analytical Chemistry,
ANCHA, Vol. 34, 1962, pp. 213-216.
[69] Ingols, R. S. and Milroy, R. D., Journal, American Water Works Association, JAWWA,
Vol. 55, 1963, pp. 282-290.
[70] Gates, E. M. and Ellis, G. K., Journal of Biological Chemistry, JBCHA, Vol. 168, 1947,
p. 537.
[71] Sandell, E. B., Colorimetric Determination of Traces of Metals, Interscience, New York,
N. Y., 3rd ed., Vol. 3, 1959, p. 612.
[72] Hooker, D. F. and Banks, C. V., "Preparation, Properties, and Applications of Some
Substituted Alicyclic vic-Dioximes," Ames Laboratory, Iowa State College, ISC-597,
Ames, Iowa, March 1955.
[73] McDowell, B. L., Meyer, A. S., Jr., Feathers, R. E., Jr., and White, J. C., Analytical
Chemistry, ANCHA, Vol. 31, 1959, pp. 931-934.
[74] Montgomery, H. A. C., and Dymock, J. F., Analyst, ANALA, Vol. 87, 1962, pp.
374-378.
[75] Goldman, E. and Jacobs, R., Journal, American Water Works Association, JAWWA,
Vol. 53, 1961, pp. 187-191.
[76] Hora, J. B., and Webber, P. J., Analyst, ANALA, Vol. 85, 1960, pp. 567-569.
[77] Shinn, M. B., Industrial and Engineering Chemistry, Analytical Edition, IENAA, Vol.
13, 1941, p. 33.
[78] Montgomery, H. A. C. and Dymock, J. F., Analyst, ANALA, Vol. 86, 1961, pp. 414-416.
[79] Fisher, F. L., Ibert, E. R., and Beckman, H. F., Analytical Chemistry, ANCHA, Vol. 30,
1958, pp. 1972-1974.
[80] Buchoff, L. S., Ingber, N. M., and Brady, J. H., Analytical Chemistry, ANCHA, Vol. 27,
1955, p. 1401.
[81] Buchoff, L. S. and Ingber, N. M., U. S. Patent 2,967,092, 1961.
[82] Rothstein, V. P. and Shemyakin, V. N., Teploenergetika, TPLOA, Vol. 9, No. 2, 1962,
pp. 54-56; Chemical Abstracts, CHABA, Vol. 57, 1962, p. 3214.
[83] Robertson, R., Journal, American Water Works Association, JAWWA, Vol. 52, 1960,
pp. 483-491.
[84] Bell, R. N., Analytical Chemistry, ANCHA, Vol. 19, 1947, p. 97.
[85] Jones, L. T., Industrial and Engineering Chemistry, Analytical Edition, IENAA, Vol.
14, 1942, p. 536.
[86] Jacklin, C. and Brower, S. R., "Correlation of Silica Carryover and Solubility Studies,"
ASME Paper No. 51-A-91, American Society of Mechanical Engineers, ASMSA, 1951.
[87] Straub, F. G., Bulletin 364, University of Illinois, Urbana, 111., 1946.
[88] Goldman, E. J., Journal, American Water Works Association, JAWWA, Vol. 51, 1959,
pp. 935-936.
[89] Mukai, K. and Gota, K., Bunseki Kagaku, BNSKA, Vol. 6, 1957, pp. 732-735.
[90] Rumler, F., Herbolsheimer, R., and Wolf, G., Zeitschrift fuer Analytische Chemie, Vol.
166, 1959, pp. 23-24.
[97] Bower, C. A., Proceedings, Soil Science Society of America, SSSAA, Vol. 23, 1959,
pp. 29-31.
[92] Fairhall, L. T. and Richardson, J. R., Journal, American Chemical Society, JACTA,
Vol. 52, 1930, p. 938.
[93] Mayne, J. B. O. and Moordhof, G. H., Analyst, ANALA, Vol. 78, 1953, p. 625.
[94] Kress, K. E., Analytical Chemistry, ANCHA, Vol. 30, 1958, p. 432.
[95] Margerum, D. W. and Santacana, F., Analytical Chemistry, ANCHA, Vol. 32, 1960,
pp. 1157-1161.
[96] Platte, J. A. and Marcy, V. M., Analytical Chemistry, ANCHA, Vol. 31, 1959, pp.
1226-1228.
Chapter 12—Sensory Examination of

Water
It is paradoxical that complaints about water quality are usually based on

subjective properties, that is, odor, taste and appearance, but analytical
emphasis is usually chemical, physical, or biological. The reason for this
anomaly is that the latter properties are more readily standardized, and
values may be reported within relatively narrow numerical limits. Subjective
properties must be measured by persons. Since people vary in sensory acuity,
it is much more difficult to establish either/intensity or characterization
values for subjective properties. Although one often sees threshold odor or
taste data presented in absolute values, there should be cognizance of the
implied uncertainty of such values. Individuals vary in sensory ability from
day to day or within a day. Groups of individuals show even greater
tendency to vary. Consequently, subjective data are best described in a
statistical or probability manner. An average or median value plus the range
or associated confidence limits tell much about the distribution of a sensory
property.
Sensory testing of water is generally restricted to taste and odor, color and
appearance. Taste and odor are usually combined as a single term in water
technology, probably because consumer complaints on these properties are
often difficult to separate. Most recent published information actually
concerns odor and says very little about taste. Much more needs to be
learned about both properties. Color may be rated against fixed standards,
but often color or appearance judgments are recorded as part of the
analytical summary of water quality. In the following section, various
aspects of sensory testing will be examined.
Panelists
Any sensory testing program is doomed to failure if the persons involved
lack interest or are not properly trained. Greater flexibility exists for
selecting panel members in research than in field operations. Research
panelists may be screened for their sensitivity to taste and odor stimuli, and
the panel may be tailored to specific needs. This is not the case in water and
waste treatment plants or field survey crews. Field personnel are selected for
254
CHAPTER 12 ON SENSORY EXAMINATION OF WATER 255
abilities other than their sensory response. Often, very few persons, or
perhaps only one, are available for these measurements. In such circum-
stances, they must be trained to best utilize their sensory, skills. Complex
interactions of mixed stimuli and person-chemical interactions well be
operating. It is necessary to get relative response data to specific standard
stimuli from personnel concerned with sensory measurement.
Individuals without prior sensory panel testing experience are sometimes
preferred in research. They do not have prejudices about methods or
responses. All persons require adequate training. Persons vary considerably
in the time required for training before their responses to reference
substances become stabilized. Some persons report reproducible, uniform
threshold values almost immediately. Others gradually improve their sensi-
tivity over a period of training. A confidential record summarizing such
personal facts as age, physiocardiac and allergy history, smoking habits, and
calibration test results has been found to be helpful in long-range studies.
Individual results or comparative performance data among panelists should
not be discussed; comparison tends to discourage less-sensitive persons,
instills a false competitive undertone to panel testing, and may bias
subjective testing results.
Good morale and motivation must be maintained. One method which has
been successful is an informal newsletter. Newsletters summarize the ob-
jectives and overall results and describe how the results fit into the scheme of
water quality control. This has been very effective. Panelists have volun-
teered comments and useful suggestions, and there has been no difficulty in
keeping them interested. Temporary physical discomfort caused by allergies,
colds, etc., will affect sensory response. Panelists should be excused during
these periods. A change in personal habits may permanently affect sensory
acuity. Persons who stop smoking have been found to have increased odor
sensitivity to reference materials. This is quite obvious if calibration records
are maintained, but it is otherwise difficult to discern.
Persons show marked differences in response to various tastes and odors.
These differences cannot be reconciled or estimated by a single reference or
substance. Only long-term experience with several references will permit
reasonable evaluation of individual acuity. The differences between indi-
viduals must also be considered. The ultimate purpose of the sensory tests
will determine the panel make up. If only persons with average sensory
responses are used, the results will not be representative of the general
population and may be misleading in water quality control. Persons
representing an extreme of the general population are the ones who most
often register complaints. It may be necessary to use supersensitive persons if
quality control limits are sought. If a general response range is sought, then a
broader-response panel must be used. Obviously, the larger the panel, the
more it is likely to represent the general population.
In some cases, sensory studies are begun with a large number of panelists
and these are gradually reduced in number without any significant loss in
256 MANUAL ON WATER
discriminating power. In extensive tests [/],' eight persons excellently

represented a group of approximately 50 for olfactory response. The number
of panelists will be determined by balancing cost and time against the
precision desired. Panels of 15 to 30 persons are routinely used in food and
beverage product development or water research. In water quality control,
only a few are needed because concern is usually with threshold or absence
of taste and odor rather than suprathreshold conditions or sensory quality
profiles.
Two kinds of fatigue may be involved in sensory testing. Physical fatigue
is caused by overextending the sensory systems by too many tests and is
easily avoided by properly designed experiments. It is a good practice to run
only a few tests at one session for each individual. These sessions should be
randomized during the day. There is both a day-to-day and time-of-day
variability in sensory response. In some cases this may be unimportant, but it
exists and should be recognized. A second form of fatigue is physiological.
Certain chemical substances such as mercaptans paralyze the olfactory
senses. After such substances are inhaled, subsequent olfactory response is
partially or completely deadened.
Panelists should be cautioned that eating highly flavored foods or
excessive smoking prior to sensory testing may influence results. A good
practice is to ask panelists to refrain from smoking or eating for at least 30
min prior to the tests. Excess use of cosmetics will also affect results. It is
important to have panelists wash their hands in odor-free soap prior to
handling the glassware used in the tests. This prevents interference by
materials carried on the hands.
Testing Laboratory
Measurement of taste and odor should be made in a controlled environ-
ment. Very few laboratories have such facilities, but they are necessary if
reproducible, meaningful results are to be obtained. Background odors
interfere with the ability to distinguish the intensity and character of sensory
response. Excessive heat, humidity, or distracting activity also influence
responses. These extraneous conditions explain the discrepancy of taste and
odor results reported by various laboratories.
An odor-testing laboratory used in research on taste and odor in water
will be described as an example of the considerations involved. The
laboratory is divided into two areas devoted, respectively, to sample
preparation and sensory testing. The areas are separated to prevent bias in
the panelists' judgment which might arise from their observing sample
dilutions being prepared. Guessing tendency is reduced. Samples are passed
between the analysts and the panelists via sheetmetal pass throughs (Fig. 1).
The panel area has separate automatically controlled humidity temperature
'The italic numbers in brackets refer to the list of references appended to this chapter
CHAPTER 12 ON SENSORY EXAMJNATION OF WATER 257
FIG. 1 — Tester passing samples to panelist—Mellon Institute.
units. Safelights which may be fitted with various monochromatic filters are
provided over each booth (Fig. 2). This is important if the samples being
tested are colored. Color bias is eliminated by selecting a filter that will mask
the color. This prevents the panelists from estimating the sample concentra-
tion or differentiating the sample from odor-free blanks by color intensity.
The atmosphere is kept odor-free by activated-carbon filters on the exit
ducts of the dehumidifier and air conditioner and a separate wall-hung
carbon filter. Normally ambient conditons are 21° C (70° ± 2°F) and 40 to 45
percent relative humidity. Panelists making odor measurements in the
laboratory are shown in Fig. 3.
The sample preparation area contains a water bath for sample tempera-
ture adjustment. The bath, controlled at 40° ± 1°C, is kept in a hood to
minimize odor or water volatility in the working area. Great care is taken to
see that these areas are kept clean and odor-free. All primary dilutions of
odorants are made in another laboratory. No odorous materials are stored,
and both smoking and eating are prohibited in this area. Glassware for the
sensory tests is not used for any other purpose. It is thoroughly cleaned in
chromic acid and detergent with numerous tap and odor-free reagent water
rinses after each use. Reagent or dilution water is prepared by first distilling
tap water, then passing the distillate through fresh columns of activated
carbon. The carbon is changes regularly, and reagent water is always freshly
prepared prior to use. Odor tests are made with 500-ml wide-mouth
258 MANUAL ON WATER
FIG. 2—Taste and odor panel testing booth—Mellon Institute.
FIG. 3—Panelists making odor measurements—Mellon Institute.

Erlenmeyer flasks covered with watch glasses. These flasks were chosen
rather than the often-used, ground-glass-stoppered flasks because the latter
are difficult to free from trace odors and they do not provide as large an
opening for the nose during odor testing. Details of construction and
operating practices and precautions are given in the original reference [2],
Mathematical Evaluation
Statistical methods of testing are important in evaluating sensory testing
information. They provide a basis for experimental design and facilitate
interpretation of findings.
Analysis-of-variance procedures are generally used to test for the signifi-
cant elements contributing to experimental results. This technique is ex-
cellent if the basis and limitations are recognized. It assumes that the variate
(results) is normally (Gaussian) and continuously distributed. In odor
testing, the results are based on stepwise dilutions in discrete intervals, hence
the continuity assumption is not strictly valid. But if the dilution intervals
are not greater than that corresponding to the accuracy of the tests plus the
variations in personal response, this deviation is not too important. There
are other important assumptions in analysis-of-variance tests. Whether these
assumptions preclude valid use of the technique for data interpretation
depends on the intent. If the interest is only in evaluation of data on hand,
these assumptions are of no concern. But if statistical inferences are to be
drawn from the data and the findings are to be extended to other
applications, the limiting assumptions apply.
Certain limitations in subjective testing might occur in the scoring
systems, the time of judging, the analytical systems, etc. These invalidate
conventional analysis of variance. The limitations dictate the use of larger
samples and nonparametric techniques for evaluation of the results. The
results of organoleptic testing may be classified in two forms, (1) quantal
responses and (2) scalar responses, which represent continuous distributions.
Quantal responses are those in which the criterion is acceptability and the
answer is yes or no. There is no intermediate value or extension. Quality
characteristic ratings, on the other hand, can often be described by
continuous scales in which intermediate values are possible. Quantal or
preference ratings are used to select materials or characteristics that deviate
slightly from small sample lots. A good example would be the tests for odor
preference of several substances by the use of the method of ranked pairs.
Each substance is given a preference rating against every other substance.
The ratings are ranked and the ranks analyzed by a procedure based upon
maximum-likelihood statistics.
Many subjective qualities are measured by a system of ranking. Thus an
array of samples is arranged by the tester in ascending (or descending) order
for such things as intensity of taste or odor. Scoring is a similar procedure
except that samples are assigned some value from a prearranged scale. These
260 MANUAL ON WATER
scoring methods depend upon familiarity of the testers with the samples. The
smaller the differences between samples, the greater the panelists' training
must be to denote these differences and the greater the chance of low
precision. Ranking methods are fast and allow many samples to be tested in
a reasonable time.
The number of steps or grades in the scoring scale to be used must be
developed from experience and the precision of the test. Five- to ten-point
scales are usually used in sensory testing although fewer and more steps have
been used. The common 0 to 5 scale denotes "no odor" to "intense or
objectionable odor" for the respective extremes. At times, several qualities
may be rated and multiple scoring systems may be assigned. The food
industry uses such systems routinely. It is necessary that those conducting
the test indoctrinate the testers thoroughly, select a useful scale, make trial
runs, and refrain from implying a degree of accuracy or acceptability not
inherent in the system.
Rating scale results are incorporated in standard experimental, statistical
designs. The number of tests is determined by the number of samples, panel
size, and consideration of the type of error (I or II). Type I error rejects the
null hypothesis when it is true. Type II error accepts the null hypothesis
when in fact it is false. The null hypothesis is the basis of the test. Detailed
mathematical explanations are available [3]. Commonly, scoring scales are
prepared in which the panelists rate the acceptability on some prescribed
scale such as 1 for extremely poor to 9 for excellent.
Odor references may be rated by various scales. Sometimes, quality
characteristics are described as acrid, sour, sweet, etc., and in other cases
these are translated into a numerical or weighted value. It has been
recognized that the odor type is important in considering both odor
classification and odor threshold. Condequently, the "hedonic index" has
been recommended for mitigating the differences between these two systems.
Briefly, the geometric value of odor threshold is assigned to one scale and a
numerical value of the quality (characterization) to another scale. The
product of these has been suggested as an index (hedonic) for rating odor in
water. Although this is a gross simplification, it denotes the nature of the
problem and an attempt to facilitate communication between investigators
in the water field. An acceptable system would allow everyone to evaluate
odor values with the same degree of objectiveness.
The hedonic index [4] operates on the assumption of a continuum of
physiological preference and a superimposed scale of successive categories of
response (like and dislike). In characterization testing, it is desired to record
the "first impression," not a reasoned, intellectual rating. The hedonic scale
is a simple system which testers readily understand, but it is liable to bias and
suffers from any limitations of the test subjects and limited test conditions. It
is necessary that the proper statistical design and method of evaluation be
employed. Limitations are caused by interactions such as those between
persons and odor types. Another objection is the fact that strong odor types
often predominate over weak odor types.
If subjective test values are to be useful, they must account for possible
fractionation. In water work, it could mean that a given sample location, say
on a stream, could yield odor values which are not representative of any
other locations, even those nearby, because the odor components weather or
decompose. The hedonic index will require modification, or some new scale
will have to be developed to meet the needs of water investigators. Complex
odors are best studied by profile methods. It has been noted that, as a
complex odor sample is diluted with reagent water, different odors are
detected than the one associated with the original sample. Instrumental
methods supporting the odor test are the best means of studying the nature
of the components of the odor mixture.
Odor-profile evaluation promises to be an acceptable means of evaluating
odor changes along a waterstream containing multiple municipal and
industrial effluents.
A number of methods for conducting and evaluating subjective testing
have been developed. The proper choice depends on the complexity of the
tests, the discriminatory power desired, and economics. It should be
emphasized that the proper statistical design of experiments must be made
before tests are begun. It is usually too late to rectify a poorly planned
experiment after the data are obtained. There is an ever present need to
balance practical or applied efficiency against statistical efficiency. Little
purpose is served if complex statistical forms are used when the problem
does not merit such sophistication.
Probably two testing procedures are best known and most used: the
triangular and the paired comparison tests. The triangular test is made by
presenting three coded samples to the tester. One differs from the other two
and the tester must distinguish the different flask. The testing procedure is
one of the best available if subtle differences exist between samples. Such
differences are involved at threshold dilutions of odorous waters. The
ranking tests described elsewhere are preferred when large differences exist.
In paired testing, the unknown sample is presented to the tester who matches
the unknown with one of a series of randomized dulutions of a known
substance or mixture. An estimation of the unknown is derived from the
over-all results.
Other techniques are available for specific purposes. These include paired
comparisons and duo-trio methods. The duo-trio is a paired test with the
standard also presented to the tester but in a manner differing from the
triangle test. First a paired sample is identified and given to the tester and
later two unknowns in random order. The tester must pick the different
sample. Although the samples are tested against the standard, the prob-
ability is 0.5 rather than 0.33 as in the triangle test.
Series tests are often used in water treating plants. A series of dilutions is
prepared and testers select one or more odorous samples from the array. If
odor thresholds are to be measured, care must be exercised to avoid having
panelists first exposed to suprathreshold concentrations.
In its simplest form, the Weber-Fichner law defining response to sensory
262 MANUAL ON WATER
stimulation states that the sensation produced is approximately proportional

to the logarithm of the concentration. This fundamental rule provides a
guide to all sensory responses.
Organic Chemicals
Organic substances are the chief source of taste and odor. They are
contributed to waters by industrial and municipal wastes, agricultural
activities, decay of vegetation, algae-actinomycete reproduction, and other
natural man-made processes. Floods loosen decaying vegetation and cause
sudden increases in lignin and tannin concentrations. These substances
impart color; a few operators have been able to correlate color and quality.
The organic materials rarely occur as single components. Usually a number
are present simultaneously and their combined sensory effect may be quite
unlike that of any of the original materials. Research has generated many
synthetic organic chemicals that are not readily biodegradable. If these
persistent or refractory substances also cause taste and odor, then special
treatment procedures are dictated. Often the sensory effect attributed to
specific organics is caused by other substances.
There have been many reports of the sensory effect of common detergents.
Investigation showed no sensory effect at levels of detergent considerably
greater than those found in most waters (maximum 10 mg/litre). Taste
thresholds of 50 mg/litre were found, but even higher levels were necessary
for detection of odor. Usually perfumes in packaged detergent products are
detected rather than the detergent itself. The presence of detergent, especially
in potable supplies from wells, is considered an indicator of pollution. Often
it is detected through foaming before infiltration of other products is noted.
Hard detergents are essentially not biodegradable, and activated carbon
treatment for sensory quality control is not usually effective except at very
high carbon dosages. Adsorption on carbon is enhanced below pH 9.
Chlorination may lead to odorous chloroproducts. Introduction of soft or
biodegradable detergents has decreased the possibility of long-term persis-
tance of these substances. However, their presence is usually the first
indicator of contamination.
Regardless of the source of organic material, water treatment personnel
should monitor its presence on a regular basis. Analyses made during taste
and odor problems are not too informative unless a base line has been
established during trouble-free periods.
Organic chemicals can cause a bad taste in fish and other aquatic foods.
The chemicals become concentrated in the flesh.
Although there have been many attempts to correlate taste and odor with
specific chemical properties, no completely successful scheme has been
developed. As new analytical devices of greater sensitivity are evolved, each
in turn has hopefully been assigned to correlation of sensory properties with
chemical structure. Claims for instrumental-sensory correlation have never
been universal and are usually restricted to a minor group of substances.
Somf of these correlations may be valid, but their weakness lies in

dependence on limited physical or chemical property evaluation. The devices
identify a single chemical or chemical fragment (radical or ion). The human
nose measures a number of properties simultaneously, integrates them, and
produces a composite sensory effect. Consequently, the nose far surpasses the
most complex instrumentation for taste and odor monitoring.
The effect of phenol on taste and odor in water has been a controversial
subject for many years. In water technology, "phenol" is a term used to
describe a variety of hydroxy derivatives of benzene or its condensed nuclei.
The substances constituting phenols are actually defined by the analytical
test used. Consequently, varying concentrations of phenol may be reported
on the same sample depending on the method of test or the modification of a
given method. The chemicals reported as phenols exhibit individual taste
and odor effects if tested separately. A wide range of threshold values will
result.
The present U. S. Public Health Service recommended limit for phenol is
0.001 mg/litre in drinking water. This limit is based on taste and odor rather
than toxicological considerations [5]. Since phenols are found in many
industrial waste waters, there had been a long history of attempts to
demonstrate a correlation between phenol content and the taste and odor of
water. Because of the limitations just noted and the presence of other
substances in water that modify the effect of phenols, it is not surprising that
controversy has existed. Phenols are not only found in industrial and
municipal waste waters but are also produced by decomposition of vege-
tation. These naturally occurring phenols are periodically flushed out by
rainfall and cause typical taste and odor problems. Extensive studies [6] have
examined the relationship between phenols in industrial waste waters and
sensory quality. The results were very complex. No useful correlation was
found between phenol level and odor following incremental chlorination.
Although phenolic compounds are often a factor in sensory quality, it is
misleading to base taste and odor control solely on phenol content.
Inorganic Chemicals
Inorganic chemicals are less often associated with taste and odor in water
than are organic substances. They impart tastes rather than odors. Sulfides
are one of the most troublesome inorganic chemicals. The major water
supply in Florida, the Ocala limestone formation, contains concentrations of
1.0 to 4.0 mg/litre of dissolved sulfides (measured as hydrogen sulfide). The
odor persists because the water pH ranges from 7.2 to 7.8. Adjustment of pH
to 9.0 or higher removes odor and improves taste. The hydrogen sulfide
remains in the water as long as dissolved sulfides are present. Aeration,
chlorination, and ion exchange are all used in treating these waters.
Inorganic chemicals play an indirect role in biologically produced taste
and odors in water. Nitrogen and phosphorus compounds from agricultural
runoff and packaged detergents, for example, provide important biological
264 MANUAL ON WATER
nutrients. These stimulate algal and biological growth and may not, only
make water unpalatable but limit its use in domestic and industrial
operations.
Dissolved salts are a factor in water quality. The saline water desalting
research program is the major hope for many nations without suitable water
supplies. Inland chloride problems may be caused by discharge of industrial
wastes or deep-well brines. These obvious pollutants are usually controlled
before they reach taste threshold concentrations. Fluorides in the con-
centration used in water treatment are not usually considered a factor in
sensory quality.
Biological
Microscopic flora and fauna flourish or perish as a function of their
aqueous environment. Environmental factors include temperature, pH,
chemical composition, sunlight incidence, and dissolved oxygen level,
among others. Food and elemental nutrient requirements depend on the
specific organism. These microorganisms frequently consume complex
organic matter and produce intermediate substances as a consequence of
their metabolic processes. The metabolic intermediates and final products
may cause taste and odor problems. Algae, actinomycetes, protozoa,
rotifers, copepods, sulfur and iron bacteria, and Crustacea are among the
most important microorganisms. Biological action, both anaerobic and
aerobic, destroys or modifies odorous compounds. But odorous by-products
may form from materials that have little or no odor and these can modify the
character and intensity of specific odorants.
Biological self-purification of streams is an important factor in neutral-
izing man-made wastes. This self-purification capacity is an economically
significant phenomenon. It should not be exceeded, however, and it should
be equitably apportioned among municipalities and industry.
Taste and odor components of biological origin are very likely to appear
with seasonal frequency. Flushouts after low-flow periods of high biological
activity are a common difficulty. Where rivers freeze, anaerobic bottom
deposits are swept downstream during spring thaws. Thermal shocks that
destroy the natural stream biota can also induce tastes and odors. Decay of
vegetation is the leading cause of taste and odor in Western and South-
western states where high solar intensity acts upon shallow supplies. As
carbonaceous matter is metabolized, the concentration of by-product carbon
dioxide is increased; this in turn stimulates biological growth.
Algae are the most troublesome microorganisms, as far as tastes and
odors are concerned. If they do not cause the problem directly, they provide
food for other organisms which produce odorous metabolic products.
Decomposing algae release nonvolatile essential oils that are taste and odor
factors [7, 8]. Organic sulfur compounds have been identified as products of
bacterial decomposition of algae. Water use contributes fertilizing materials
for these organisms. Phosphate particularly promotes proliferation of algae.
Frequently, taste and odor problems occur long after algae blooms have
passed their peak; actinomycetes feeding on the algae are reported to be the
cause. This is confirmed by the observation that various waters containing
unlike algae populations will have similar odors after a period of heavy
growth. Pure algae cultures have been developed which yield earthy, grassy,
and fruity odors [P-77].
To minimize algae- and protozoa-created taste and odors, new reservoirs
should have the topsoil stripped and surrounding vegetation cut back.
Furthermore, domestic and industrial wastes which contribute nutrients
should be kept out of reservoirs. Since algae blooms occur at predictable
intervals, preventive measures are possible, for example, by treating in-
cipient blooms with an algicide.
Actinomycetes are organisms closely related to both bacteria and fungi.
They are colonial in behavior and are usually associated with soils, although
an aquatic group is a significant source of taste and odor in water. They
impart an earthy, musty, or woody odor to water and to fish flesh. An
earthy-smelling compound, geosmin [12-14], and related chemical com-
pounds have been isolated from actinomycetes. Mucidone, a chemically
different compound that produces a musty odor, has been isolated from
natural waters [75].
Quite often, tastes and odors have been attributed to algae when the real
cause was actinomycetes [76], These organisms grow in mud or silt,
especially if organic matter and nutrients are plentiful. They are quite
temperature-sensitive; 15°C represents a minimum for spore germination
and vegetative growth. Rising temperature increases growth, with no
apparent inhibition up to 38° C. They will grow in the dark in the laboratory,
but shallow waters up to depths of 28 cm (11 in.) enhance growth. Deep
bodies of water tend to discourage growth. Complex organic substances
(such as proteins) in water provide nutrients. Shallow-water reservoirs with
luxurious growths of grasses and rushes will harbor high concentrations of
actinomycetes in warm weather. Algae (planktonic and attached) serve the
same nutrient purpose. The degree of alkalinity affects growth; the greater
the alkalinity, the more favorable the environment. Growth is poor under
acidic conditions.
The metabolic by-products of actinomycetes, such as aromatic amines,
ketones, aldehydes, saturated fatty acids, and unsaturated aromatics have
been found to cause taste and odor in water. Minimum concentration in raw
water appears to be about 0.2 mg/litre actinomycetes. Isoamyl and isobutyl
amines, fatty acids, and unsaturated aromatics have been identified as
significant odorous metabolic by-products.
Streptomyces spores are very small and pass through common bacterial
filters. Sand filters in water-treating plants are therefore not effective in
removing them. Spores passing the filters can become attached to mineral
deposits in water distribution mains. There they can grow and impart taste
and odor not found in the water leaving the treatment plant.
Time or duration of odor production depends on temperature and source
266 MANUAL ON WATER
of nutrition. The first odor is usually fishy and runs two to ten days. The
next phase depends on the nutrient supply. If the primary stages of the
actinomycete growth have exhausted the nutrient, small sex cells are formed
as a zygote or seed. This ends the odor conditions. When sufficient nutrient
remains, the primary stage forms a motile zygote which develops into a large
secondary filament. These rapidly growing secondary filaments contribute a
grassy taste and odor to raw water which lasts two to ten days and then
changes to a pungent marshy odor. This odor may also last two to ten days,
and, if enough nutrient is available, will alter to a potato-bin or slightly
musty odor. If the nutrient is exhausted at any time, the odors soon cease.
This sequential development provides a clue to the occurence of recognizable
types of odor. It should serve as a guide to the waterworks operator to assure
action in the early stages of actinomycete activity.
Copper sulfate treatment of water for algae control is a preventive
measure widely used. Use of copper sulfate or copper citrate and other
chemicals for weed or algae control may upset the natural biological
balance. Cumulative build up of these compounds in fish caught for human
consumption poses questions about toxicological or palatability factors. If
copper sulfate levels are high, fish are eventually killed, thus creating further
biological imbalance which may result in increased growth and incidence of
algae.
Control of actinomycetes in reservoirs has not been successful. Various
chemicals have been tried with incomplete control and unfavorable eco-
nomics. It remains necessary to treat actinomycetes at the water plant,
because inflow of sewage and vegetal matter induces prolific growth in
reservoirs. The resulting odors have been a major problem in certain areas,
such as the Southwestern part of the United States, during the period April
to November, when temperature and light are conducive to actinomycete
propagation.
The taste and odor remaining in water is a function of the treatment
employed. Where actinomycetes are present, addition of as much as 10
mg/litre of chlorine may be necessary to reach the breakpoint. Threshold
odors rise with additions of chlorine until a free available chlorine residual is
present. The chloro-derivatives are especially odorous and not amenable to
complete oxidation even when a residual chlorine content is present. Large
additions of activated carbon will minimize odor but may not remove all of
it. The point of addition of the activated carbon is important. If chlorine or
chlorine dioxide are used, they will be preferentially absorbed by the
activated carbon. The carbon should therefore be added after these agents
have had enough time to become effective.
Taste
"Taste and odor" are used jointly in the vernacular of water technology.
Most of the reported studies concerning the taste and odor problem actually
emphasize or deal exclusively with odor. This arises from the inability to
separate the two sensations because taste is usually masked by olfactory
sensations. In some cases, taste tests are more sensitive than odor tests, but
very little has been published on taste alone. Four basic tastes are recog-
nized: salty, sweet, sour, and bitter. Variations of these are caused by
olfactory influences. Taste results from the presence of a substance in an
aqueous layer on the tongue, causing excitation of the receptor cells. As in
olfaction, different cells respond to different basic tastes. Taste is an overall
summation of these individual responses. Adaptation to taste, as in the case
of odor, occurs because the frequency of nerve impulses decreases under the
action of a steady stimulus. Taste sensitivity changes with age, as does the
spatial arrangement of the taste buds.
The same basic testing considerations important in odor testing apply to
taste testing. Background odors, temperature, and humidity should be
regulated; color bias should be recognized and controlled if necessary; and
the equipment used in the test should be free of foreign tastes. The taste test is
made by having the panelist(s) rinse the mouth with taste- and odor-free
water, expectorate the rinse, and repeat the procedure with the substance
being tested. The rinse water should not only be free of background taste but
be at sample temperature. Sample sizes should be reasonable, neither too
large nor too small. The sample size is controlled in some degree by the basic
taste. Sweet tastes usually require somewhat larger samples than bitter
tastes, because the latter cause sensation somewhat more rapidly. Usually,
taste testers make several tests per session without loss of precision, although
this has to be established in each situation. Threshold measurements are
based on the premise that concentrations in excess of these values are
unacceptable. This is true for odor but does not apply to mineral-induced
taste. Consumers prefer slightly mineralized water to distilled water. Con-
sequently, psychological scaling methods have been suggested as alternatives
to taste threshold measurements for establishing limiting standards for
mineralization.
An important consideration in taste testing is the possible hazard from
toxic substances. These may be the sensory factors in the water or the
reagents used in preparing the sample. Samples of water having residual
chlorine require dechlorination if the presence of organic tastes is to be
discerned. The dechlorination reagent of choice is sodium thiosulfate.
Sodium arsenite is often used in odor testing but should not be used in taste
testing because of the toxicological hazard.
Odor Terminology
The two most common descriptions of an odor are its intensity and its
characterization. Intensity is a measure of relative strength. Characterization
is the description of how it smells. For example, sweet and pungent are two
common characteristics. The threshold odor is the concentration of an odor
268 MANUAL ON WATER
which is just barely detectable to the individual. Suprathreshold odor

concentrations are those exceeding the threshold concentration. Since
people differ, often considerably, in their sensitivity to a given odor, the
actual threshold odor value for axgiven sample will vary. In fact, it is more
precise to report a range of threshold odor values together with the number
of observations rather than an average, because an average is usually based
on too few determinations to have application to the population as a
whole.
Odor changes are not usually tolerated by the water consumer. It should
be noted that the preceding statement did not say "the absence of odor."
Odor analyses are concerned with threshold or near-threshold odor numbers
or odor intensity index values. These are related terms. The threshold odor
number, TON, is the number of times a sample must be diluted for the odor
to be just detectable. For example, if 25 ml of sample in a standard 200-ml
testing volume is just detectable, the dilution factor or TON is 8. Such
numerical values are easily comprehended for samples of low-odor intensity,
such as potable or drinking waters. But some natural waters and some
industrial waste waters have intense odors. The threshold odor number
system gives extremely large numerical values for such waters. Toâvoid the
difficulty of interpreting results the odor intensity index, OH, system is used.
The OH is the number of times a sample must be diluted in half with odor-
free water to reach threshold odor concentration. In the example given in the
foregoing, the .OH would be 23 or 8 by this system. TON and OH values are
therefore related as follows:
TON = 2°"
There is another significance to the OH values. It has been shown [7] that
one OH value represents the order of reproducibility of the threshold odor
determination.
Odor Parameters
Any study of odor in water requires a reproducible method of test.
Extensive research [7] led to the development of a standard procedure,
ASTM Test for Odor in Industrial Water and Industrial Waste Water (D
1292) [77]. This procedure uses a three-flask array in panel testing. Two
flasks contain odor-free blanks and the third may be either a sample dilution
or a third blank. The test samples are prepared so that concentrations are
gradually increased from subthreshold level. Exposure to suprathreshold
concentrations usually interferes with subsequent threshold odor measure-
ment.
Reproducibility with several odorants has been found to be of the order of
approximately one-fold dilution or one OH unit. This means that dividing a
given odor sample in half by odor-free water in each step of the dilution
sequence is approximately equal to the accuracy of the panelists. Dilutions
of less than 50 percent are often used, but this is unjustified. The natural
variability for any person from day to day or even within a day may well
exceed this implied accuracy even under ideal environmental conditions.
Very important effects for person-chemical interaction have been demon-
strated. In other words, the kind of odorant and the choice of panelist are
highly interdependent with regard to the threshold odor result. This should
not be surprising. People vary in their ability to detect a given odor and they
also vary in their relative sensitivity to various odors. This contributes to the
range of personal preferences for many products from perfume to water. In
these studies, a variety of odor testing procedures was evaluated. It was
found that the environmental conditions were more important than the test
method if similar precautions were used.
The literature contains references to threshold odor tests made at a variety
of temperatures. Many plants use 60° C because this corresponds to hot-
water service in domestic applications. Others use lower temperatures,
particularly 40° C, which approximates body temperature. Statistical evalua-
tion of odor tests at 21°C, 40° C, and 60° C showed that (1) odor response
varied by 4 Oil values (16-fold change of concentration) between the least
and most sensitive person, and (2) between 21° and 60° C there was an
increase in overall sensitivity by the panel, corresponding to a decrease in
odorant concentration equivalent to 1.4 OH units. Although higher tempera-
tures do improve odorant volatility, there are other considerations. Panelists
preferred 40° C. They complained of losing initial odor sensations on second
inhalation and of steaming at 60° C. At 21°C, room temperature, they
regarded the odor sensation as being dead. The 40° C temperature has been
incorporated as a standard condition by ASTM. These tests also demon-
strated that person-chemical interaction and the overall variability of the test
procedure were approximately one Oil value.
Background odor may interfere with olfactory response. There has been
little published evidence to support this somewhat obvious generalization.
Food and beverage laboratories recognize this fact and do provide odor-free
atmospheres in their sensory testing laboratories. Few water laboratories
have such facilities. As part of the foregoing investigation, the background
odor effect was established for the standard odorants w-butanal and m-
cresol. Measurement of the threshold odor of each was made with the same
odor and with the differing odor in the atmosphere. The results showed that
there is a marked effect. The olfactory system is confused. If m-cresol was
the background odor, then detection of n-butanol in the threshold test was
enhanced but detection of m-cresol was hindered. Similarly, a background
of n-butanol suppressed detection of n-butanol and enhanced detection of
m-cresol in the threshold test. The only possible recommendation is that
background odor must be eliminated during odor testing.
270 MANUAL ON WATER
Odor Mixtures
Reports describing odors in water are often conflicting. This may be
because the complex effects of mixtures of odors are not usually recognized
in the water pollution field. Rarely are odors in natural or industrial waste
water the result of only one substance; more often they are a composite of a
number of odorants. When individual odorants mix, the result may be one
of additivity, synergism, or antagonism. If the substances combine to give an
odor intensity equivalent to the contribution of the individual odors, this is
additivity. For example, a mixture Of one part A and one part B of equal
odor intensity combine to produce an intensity equal to either 2A or 2B.
Synergism takes place when the combined odor intensity of the mixture is
greater than would be expected by simple addition of these separate
intensities. This occurs if A plus B produce an odor greater than 2A or 2B.
Similarly, antagonism is the suppression of combined odor effects so that A
plus B produces less odor than simple addition.
As if the odor mixture effect were not complicated enough, it is necessary
to consider that measurement involves persons of varying olfactory ability.
The previously cited examination of procedures and environmental factors
showed that there are significant person-to-person differences and person-
to-chemical interactions. The interaction refers to the varying sensitivity of a
given person to different kinds of odors. The effect of mixtures of odorants
in aqueous solution was studied. To do this, statistically-designed experi-
ments were devised so that the effects of person-to-person differences and of
person-to-chemical interactions could be isolated from the effect of odor
mixtures. Theoretical models were developed which were based on the
concepts of additivity, synergism, and antagonism [18]. These mathematical
models were developed for binary and for complex higher-ordered mixtures.
Several methods of data analysis were employed to eliminate any bias
possible from the choice of experimental procedure. Eight odorants of
differing sensory, physical, and chemical properties were selected: m-cresol
(phenolic); «-butyl mercaptan (sulfide); pyridine (amine); acrylonitrile (un-
satrated); n-amyl acetate (ester); n-butanol (alcohol); 2-4 dichlorophenol
(chlorinated phenol); and acetophenone (ketone). These were studied as
binary mixtures and as a blend of all eight. For the binary mixture study,
solutions were prepared of two components where one was 0, 25, 50, 75, or
100 percent of the total. Threshold odor measurements were made by
panelists using the ASTM Method D 1292 at 40° C. Synergism or antag-
onism of olfaction was the rule rather than the exception. It was also an
excellent demonstration of the importance of precise dilution preparation in
sensory testing. Often, very minor changes in relative concentration of one
component may change the resulting sensory response greatly.
An olfactory test of a complex mixture of eight odorants showed that
most of the 14 panelists either detected odor in the basic mix (containing !/g
threshold concentration of each of the eight odorants) or required further
CHAPTER 12. ON SENSORY EXAMINATION OF WATER 271
dilution to get to threshold concentration. Only a few required higher

concentrations than the basic mix.
Recent studies by others have also shown these complex odor effects.
These laboratory and field investigations [79,20] were approached in an
entirely different manner but reached the same conclusions. Some of the
same chemical odorants were used, thus facilitating comparison.
There is important practical meaning in these results for persons con-
cerned with water pollution control. Conceivably, several waste-water
streams, each with no perceptible odor, may mix in the receiving water body
to produce unacceptable tastes and odors. It is obvious that use of single-
odor threshold values as criteria for acceptable water quality is inadequate.
The combined effect of all contributors must be considered if meaningful
and realistic data are to be obtained.
Threshold Odors
It is very difficult to find information on the threshold odor concentration
of chemicals in aqueous solution. Available data are questionable because of
the lack of environmental and procedural control. A summary of threshold
odor values obtained by ASTM Method D 1292 at 40° C for various
chemicals is given in Table 1.
The wide range of odor threshold concentrations has practical signifi-
cance. This range, particularly at the lower end, is of more importance than
the average value. This is the segment of population most apt to recognize
impaired water quality and register complaints. For example, the average
phenol threshold is at 6 mg/ litre, but the most sensitive testers could
distinguish 17 /ug/litre. The lease sensitive testers required 17 mg/litre, a
thousand-fold difference! These are interesting values because the U. S.
Public Health Service Drinking Water Standards [5] specify a limit of 1
/zg/litre phenol. This limit, which is below the 17 /ug/ litre lower limit found
in these tests, is based on taste and odor induced by chlorination. Chlorina-
tion of phenolics may greatly alter and intensify the odor of the parent
substance.
Chlorination
Chlorination destroys bacteria and organisms such as algae, and directly
or indirectly destroys tastes and odors. There are situations, however, when
chloro derivatives are more obnoxious as taste compounds than the original
chemical. Chloro derivatives are handled by a variety of methods. For
example, addition of ammonia prior to chlorination often prevents forma-
tion of chlorine-reaction products. Ammonia-chlorine treatment prevents
creation of odors by chlorination but in itself does not remove odors.
The chlorine requirement of waters varies as a function of the complexity
272 MANUAL ON WATER
TABLE 1—Odor threshold concentrations for various chemicals".

Number of
Number of Observa- Threshold Odor Level,* mg/litre
Chemical Panelists tions Average Range
Acetic acid 9 9 24.3 5.07 to 81.2
Acetone 12 17 40.9 1.29 to 330
Acetophenone 17 154 0.17 0.0039 to 2.02
Acrylonitrile 16 104 18.6 0.0031 to 50.4
Allyl chloride' 10 10 14700 3 660 to 29 300
n-Amyl acetate 18 139 0.08 0.0017 to 0.86
Aniline 8 8 70.1 2.0 to 128
Benzene^ 13 18 31.3 0.84 to 53.6
n-Butanol 32 167 2.5 0.012 to 25.3
p-Chlorophenol 16 24 1.24 0.02 to 20.4
o-Cresol 13 21 0.65 0.016 to 4.1
m-Cresol 29 147 0.68 0.016 to 4.0
Dichloroisopropylether 8 8 0.32 0.017 to 1.1
2-4 Dichlorophenol 10 94 0.21 0.02 to 1.35
Dimethylamine 12 29 23.2 0.01 to 42.5
Ethylacrylate 9 9 0.0067 0.0018 to 0.0141
Formaldehyde 10 11 49.9 0.8 to 102
2-Mercaptoethanol 9 9 0.64 0.07 to 1.1
Mesitylened 13 19 0.027 0.00024 to 0.062
Methylamine 10 10 3.33 0.65 to 5.23
Methyl ethyl pyridine 16 20 0.05 0.0017 to 0.225
Methyl vinyl pyridine 8 8 0.04 0.015 to 0.12
/3-Naphthol* 14 20 1.29 0.01 to 11.4
Octyl alcohol** 10 10 0.13 0.0087 to 0.56
Phenol 12 20 5.9 0.016 to 16.7
Pyridine 13 130 0.82 0.007 to 7.7
Quinoline 11 17 0.71 0.016 to 4.3
Styrened 16 23 0.73 0.02 to 2.6
ThiophenoF 10 10 13.5 2.05 to 32.8
Trimethylamine 10 10 1.7 0.04 to 5.17
Xylene* 16 21 2.21 0.26 to 4.13
n-Butyl mercaptan 8 94 0.006 0.001 to 0.06
"Journal, American Water Works Association, Vol. 55, 1963, p. 913.
^Threshold values based upon pure substances.
^Threshold of a saturated aqueous solution. Solubility data not available.
^Dilutions started with saturated aqueous solution at room temperature; solubility data ob-
tained from literature for correction back to pure substances.
of their components. It may change suddenly. For example, during thermal

overturns of reservoirs, raw-water characteristics are subject to sudden
modification. It is important to note that although there may be a
relationship between organic content of water (as measured by chlorine
demand) and taste and odor in some cases, this is not a dependable
correlation. Many organic compounds do not exert a chlorine demand and,
conversely, chlorine demand is not indicative of absolute organic concentra-
tion. Chlorine reactions are dependent upon the concentration of reactants,
reaction time, pH, temperature, and chemical, physical, and biological
factors. Low chlorine-demand waters are usually acceptable over a longer
period of time after treatment because chlorine levels have been adequate to
maintain residual chlorine and not just initiate chemical reactions.
Chlorine dioxide has been used to treat phenolic contamination but it is

not applicable to all situations. It is useful, however, where sewage, algae,
and certain industrial wastes affect taste and odor quality. European plants
often use ozone for odor control, but economic considerations preclude its
use in the United States. Reservoirs tend to stratify and anaerobic zones are
created. The decomposition of organic matter in this zone leads to increased
taste and odor problems and an associated increase in chlorine demand.
Often, other problems occur simultaneously, such as requirement for
manganese removal. Oxidation, flocculation, pH adjustment, and chlorina-
tion must then be used^ for treatment of the water.
Chlorine reactions with phenolic compounds have been the basis of
numerous studies. The various chloro-phenolics differ considerably in taste
and odor potential, and in the treatment required. Most important is the fact
that a phenolic concentration far lower than its threshold level may produce
decided medicinal odors after chlorination.
Not all organic matter is destroyed in plant chlorination processes. Some
organic material becomes associated with colloidal particles of high electro-
phoretic charge which frequently escape coagulation and settling steps.
These materials are considered a principal source of post-treatment tastes
and odors.
A survey of water-treating plants showed that taste and odor problems
associated with dechlorination of samples originally containing a chlorine
residual were not uncommon. Before taste or odor tests can be made on such
waters, they must be dechlorinated. Unfortunately, the commonly used
dechlorination reagents, sodium thiosulfate and sodium arsenite, frequently
induce sensory effects. Dechlorination with thiosulfate often releases a
sulfurous odor. This is most noticeable when the water is near the threshold
odor level and no dilution with odor-free water is needed. Sodium arsenite is
less apt to produce odor, but it is not used in many laboratories because of
its acute toxicity. Although a variety of reagents has been tested for
dechlorination, none have been found to be better than sodium arsenite.
There is need to study this problem.
It was found that the ammonia nitrogen in most waters, less than 0.5
mg/litre, does not generally produce sufficient chloramine to modify
threshold -odor measurement significantly. If proteins and other nitrogen-
containing substances are present, this simple criterion based on ammonia
nitrogen may not apply.
Odor profile analyses from the raw-water source through the steps in
water treating plants have been found to be a very valuable aid in adjusting
treatment dosages and maintaining finished water quality.
Activated Carbon
The use of activated carbon for adsorption of taste and odors from
organic chemical contaminants is common. The dosage required is not
predictable from simple measurement of odor intensity or characterization
274 MANUAL ON WATER
of the raw water. It is a function of the physical-chemical properties of the

entire water-impurity system. Some organic chemicals are preferentially
adsorbed. In the absence of adequate adsorption capacity for all the organic
materials present in water, equilibrium conditions are created, controlled by
varying relative adsorption of individual components. Desorption of prev-
iously adsorbed materials may take place [27]. Adsorption in the presence of
suitable capacity follows the universally accepted Fruendlich adsorption
isotherms.
Although carbon is an efficient adsorbant for organic chemicals, it is
rarely used by itself. Economic considerations prompt its use in conjunction
with other treatment processes. The combined treatment is dictated by
maximizing efficiency at lowest cost. Carbon is most frequently added in
granular form as a slurry, with the point of application governed by the raw-
water quality. In some cases it has been added to raw water prior to
chlorination to minimize particularly obnoxious chloro-organic formation.
In other cases it has been added after the coagulation- process to remove
residual taste and odor components. One practice which merits critical
review is that of adding activated carbon in the flash mixers preceding
coagulation. The coagulant forms a matrix around the carbon, preventing
effective use of its adsorption capacity. Addition of the carbon in large
basins requires multipoint application, usually through distribution nozzles
at the basin surface. The carbon then settles through the water, which is
traveling in a horizontal direction. Another form of carbon treatment used
in some areas is the fixed activated-carbon tower, down which the water is
allowed to percolate.
Instrumental-Sensory Correlation
Recent developments in analytical instrumentation have expanded the
potential for instrumental-sensory correlation. Despite these advances, the
human nose is still a more sensitive detector of many organic materials than
the best man-made detectors. Particularly promising are the ionizatitm
detectors used with gas-liquid chromatography. The flame-ionization and
electron-capture detectors have been used to detect nanogram and picogram
concentrations of organic elements of complex tastes and odors. The
chromatographic and various infrared and other spectrograph techniques
may provide a key to greater understanding of sensory properties. Of
immediate practical use is the possibility of monitoring selected waste waters
or raw waters for specific organics known to be associated with taste and
odor in water problems. The early-warning feature of such a monitoring
program would be helpful to many industrial waste water and domestic
water treatment plant operators.
Correlations of instrumental and sensory responses have been made, but
the results are meager [22]. Typical are the characterizations of specific gas-
liquid chromatographic peaks separated from actinomycete cultures as
woody, musty, or earthy odors [23]. The changes in gas-liquid chromato-

graphic profiles of natural waters are being studied to provide an indication
of potential taste and odor problems and to indicate those organic con-
taminants meriting further characterization by spectrographic procedures
[24].
References
[7] Baker, R. A., Journal, Water Pollution Control Federation, JWPCA, Vol. 34, 1962, pp.
582-591.
[2] Baker, R. A., Water Pollution Control Federation, JWPFA, Vol. 35, No. 11, 1963, pp.
1396-1402.
[3] Anderson, R. L. and Bancroft, T. A., Statistical Theory in Research, McGraw-Hill, New
York, 1952.
[4] Peryam, D. R. and Pilgrim, F. J., Food Technology, Vol. 11, 1957, pp. 9-14.
[5] U. S. Public Health Service, Drinking Water Standards, PHS Publication No. 956, U. S.
Government Printing Office, Washington, D. C., 1962.
[6] Kinney, J. E., Journal, American Water Works Association, JAWWA, Vol. 52,1960, pp.
505-514.
[7] Silvey, J. K. G., Proceedings, 44th Texas Water Works and Sewage Short School,
PTWOA, Vol. 44, 1961, p. 126.
[8] Henley, D. E., Glaze, W. H., and Silvey, J. K. G., Environmental Science and Tech-
nology Vol. 3, 1969, pp. 268-271.
[9] Jenkins, D., Medskar, L. L., and Thomas, J. F., Environmental Science and Technology
Vol. 1, 1967, pp. 731-735.
[10] Silvey, J. K. G., and Roach, A. W., Journal, American Water Works Association,
JAWWA, Vol. 51, 1959, pp. 20-32.
[77] Roach, A. W. and Silvey J. K. G.; American Microscopical Society, Vol. 77, No. 1,1958,
pp. 36-47.
[72] Medskar, L. L., Jenkins, D., and Thomas, J. F., Environmental Science and Technology,
Vol. 2, 1968, pp. 461-464.
[13] Gerber, N. N., Biotechnology and Bioengineering, Vol. 9, 1967, pp. 321-327.
[14] Rosen, A. A., Safferman, R. S., Mashni, C. I., and Romano, A. H., Applied Micro-
biology, Vol. 16, 1968, pp. 178-179.
[75] Dougherty, J. D. and Morris, R. L., Journal, American Water Works Association,
JAWWA, Vol. 59, 1967, pp. 1320-1326.
[16] Silvey, J. K. G., "Actinomycetes in Water Distribution Systems," University of Illinois
Engineering Experimental Station, Circular 81, 1963.
[77] Annual Book of ASTM Standards, Part 23, 1968.
[18] Baker, R. A., Journal, Water Pollution Control Federation, JWPFA, Vol. 35, 1963, pp.
728-741.
[79] Rosen, A. A., Peter, J. B., and Middleton, F. M., Journal, Water Pollution Control
Federation, JWPFA, Vol. 34, 1962, pp. 7-14.
[20] Rosen, A. A., Skeel, R. T., and Ettinger, M. E.,-Journal, Water Pollution Control
Federation, JWPFA, Vol. 35, 1963, pp. 777-782.
[27] Baker, R. A., Journal, American Water Works Association, JAWWA, Vol. 56, pp. 92-98.
[22] Reviews of Correlations of Objective-Subjective Methods in the Study of Odors and
Taste, ASTM STP 451, American Society for Testing and Materials, 1969.
[23] Silvey, J. K. G., Glaze, W. R., Hendricks, A., Henley, D.. and Matlock, J., Journal,
American Water Works Association, JAWWA, Vol. 60, 1 '68, pp. 440-450.
[24] Baker, R. A. and Jennings, R., "A Case Study of Microorganic Contaminants in River
Water," Industrial Water Engineering, Vol. 7, 1970.
Chapter 13—Sampling and Identification

of Deposits in Steam and Water Systems
Pure water in an uncorrodible vessel would never leave a deposit.

Although this ideal case has been very nearly reached in some instances,
deposits are still found in most water or steam systems. The examination and
analysis of these deposits provide the person responsible for water treatment
with information about phenomena in the water system, and will frequently
indicate to him the means for correcting the condition that caused the
deposit.
This chapter covers the nature of deposits in steam and water systems, the
sampling of deposits for analysis, and methods for identification of the
constituents of the deposits. Quantitative chemical analysis is dealt with in
Chapter 14. Some of the instrumental methods discussed in this chapter are
not at present widely used in the examination of deposits, but brief
descriptions are included because there are instances where they can be used
to advantage. Deposits may be divided into three general classes: inorganic,
organic, and biological.
As shown in Fig. 1, inorganic deposits result from corrosion of containing
surfaces; from precipitation by chemical reaction between two or more
constituents of the water; or precipitation from physical causes, such as
change in solubility with pressure or temperature, by evaporation to dryness,
or by relief of supersaturation. Closely akin to these, but still somewhat
distinct, is the sedimentation of suspended matter.
Organic deposits generally result from precipitation of organic materials
naturally present in the water supply, contamination of condensate returns
by process materials, or precipitation of certain organic treatment chemicals.
Biological deposits may occur as a result of precipitation, due to changing
conditions, of biological matter present in the raw-water supply, or they may
be formed where they are found because of utilization of nutrients in the
water and a favorable environment for biological growth.
The number of possible techniques available to the modern analyst is so
large that only a bare outline can be given in this chapter. References in the
chapter will suggest where more detailed information can be found.
276
CHAPTER 13 ON SAMPLING AND IDENTIFICATION OF DEPOSITS 277
FIG. 1—Deposits on the waterside of a heat exchanger.
Classification of Deposits
Since most of the crystalline water-formed deposits have the same
composition and structure as hydrothermal minerals, it is both helpful and
logical to use the mineral scheme for classification. For diagnostic purposes,
one of the most important types of information is the identification of the
compounds in a deposit according to the mineral species present.
Basic Compounds
Some relatively soluble metallic salts form, with the hydroxide of that
metal, very insoluble basic double salts. These are most often found on
metallic surfaces as corrosion products where water carries a considerable
amount of oxygen. An alkaline condition is not required and the com-
pounds may be formed at pH values below 7.0. Copper can form basic
carbonates, chlorides, and sulfates. Atacamite (aCuQ2-3Cu(OH)2) is
frequently found as a hard, very adherent corrosion deposit in condenser
tubes. Zinc forms a basic carbonate, as 2ZnCO3; 3Zn(OH)2 and a basic
magnesium chloride, MgCl2-5Mg(OH)2-8H2O, may be formed in seawater.
278 MANUAL ON WATER
Carbonates
Carbonates of most of the alkaline earth and heavy metal elements are
insoluble. The respective bicarbonates are much more soluble. However,
heat and an increase of pH value shift the bicarbonate to the normal
carbonate form to produce the respective insoluble compounds. Aragonite
(TCaCOs) and calcite (/SCaCOs) are found in heat exchangers and softener
and boiler sludges. Of the two, calcite is the more stable and is the high-
temperature form. Dolomite (CaCOs. MgCOs) may also be found where the
calcium-magnesium ratio is favorable to its formation. Occasionally, modi-
fied forms of calcium carbonate such as ankerite (CaCO3.(F.e, Mg, Mn)
COs) and gaylussite (CaCO3.Na2CO3.5H2O) precipitate in low-pressure
boilers.
Siderite (FeCOs) may precipitate from waters containing ferrous bi-
carbonate or may form as a corrosion product in condensate containing
carbon dioxide. Normal carbonates of other heavy metals are rarely formed
in industrial water.
Hydrated Oxides
The class of hydrated oxides contains principally the trivalent corrosion
and precipitation products of iron and aluminum. The term "hydrated
oxide" is more one of custom than exactitude. The formula M2Os.H2O is
correct for composition, but MO(OH) is more correct for structure.
The "hydrated oxides" of iron and aluminum are interrelated in that
temperature and time cause the more highly hydrated forms to lose water
and undergo alterations of crystal structure. This relationship is shown in the
following diagram:
It is believed that the y monohyd rates (lepidocrocite and boehmite) in

industrial deposits are more likely formed by recrystallization from rapidly
precipitated amorphous gels. The a monohydrates (goethite and diaspore)
are more likely to have formed without the gel precursor. The 7 forms are
usually formed at lower temperatures than the a forms.
There is also a series of hydrated oxides of zinc, but the composition of
this series is indefinite. X-ray patterns show differences but are designated
only as ZnA, ZnB, and ZnE. The principal mode of formation is by the
corrosion of metallic zinc in hot water. ZnE is the common species.
Hydroxides
Simple hydroxides precipitate from solutions of high pH values, or under
special conditions hydrolyze in corrosion reactions. Brucite (Mg(OH)2) is
found in boiler and softener sludges but it can also form in environments of
relatively low pH values such as those found in lower stages of seawater
evaporators. Calcium hydroxide (Ca(OH)2) may scale out in lines carrying
lime slurries or in superheaters and steam lines by thermal hydrolysis of a
previously deposited calcium carbonate.
Ferrous hydroxide (Fe(OH)2) has been identified in corrosion pits, and
nickel hydroxide (Ni(OH)2> in exfoliated flakes from cupro-nickel stage
heaters.
Metals
Free metals in a deposit are usually the end product of a reducing
condition acting on a solution or an oxide of the metal. Dendritic copper is
present in many high-pressure boiler sludges and sedimentary deposits.
Copper lends weight to the mass of sludges, and part of its undesirability is
caused by the settling tendency it provides in areas of low water velocity.
Metallic iron may be found in some corrosion products either as an
inversion product of ferrous oxide at high temperatures or as a result of
isolation by the corrosion of surrounding grains. In the former case, FeaO4 is
also formed as follows:
4FeO — Fe + Fe3O4
Metal Oxides
Metal oxides are invariably the products of corrosion reactions, the
dehydration of hydrated oxides, or the decomposition of carbonates. The
oxides may be loose and porous or compact and adherent according to
conditions of formation. Oxygen for the formation of the oxide can be
provided by dissolved oxygen in the water or by bound oxygen from
molecular H2O. In the latter instance, hydrogen is produced.
Magnetite (FesCh) is formed under the greatest variety of conditions. FeO,
Fe3O4, and FeaOa can all be found in aqueous corrosion products of iron
280 MANUAL ON WATER
according to the amount of oxygen present. Fe3O4 and occasionally FeO are
the products in steam.
The two copper oxides, cuprite (CuzO) and tenorite (CuO), may both be
components of boiler sludge or corrosion films on copper, brass, or bronze.
High oxygen concentration and high pH values are favorable conditions for
theibrmation of tenorite, whereas cuprite is more likely the solid phase of a
lower pH and lower dissolved oxygen concentrations.
Corundum (aAhOa) has been found in high-pressure boilers and turbines
as an alteration of precipitated alumina. Zincite (ZnO) is a hot-water
corrosion product on metallic zinc.
Organic Salts
The mono- and di-hydrates of calcium oxalate (CaCzO^Ô and CaCzCV
IH'zb) are practically the only organic salts of sufficient insolubility to be
found in industrial deposits. Certain food liquors containing oxalic acid and
calcium cause severe oxalate scaling in evaporators.
Phosphates
The insoluble phosphates in industrial waters are formed by the reaction
of soluble calcium, magnesium, or iron with a soluble phosphate added as a
form of treatment. Most boilers today are treated to precipitate the calcium
purposely as hydroxyapatite (Caie(OH)2(PO4)6) sludge. Although it is usu-
ally fluid and soft, it can form a compact scale. It may also undergo an
alteration in areas of high heat-exchange rates and highly concentrated
boiler water to form a hard scale, wilkeite (Cai0O[(Si, P, S)O4]6). Sodium
calcium phosphate (NaCaPO*) forms under similar conditions.
Other calcium phosphates, whitlockite (/JCaaPzOs) and dicalcium phos-
phate (CaHPO4), have crystallized out in chemical lines feeding boiler
drums.
Magnesium hydroxyphosphate (MgsPO4 • Mg(OH>2) is a sludge com-
ponent of boiler water having a high phosphate and a low silica concentra-
tion.
A low-temperature ferric phosphate, vivianite, Fe3(PO4)2'8H2O, is
occasionally found in cooling systems. It frequently is mixed with iron
oxides in deposits having a glassy appearance.
Silica
Uncombined silica as quartz or cristobalite forms by the hydrothermal
alteration of solid silicates or by direct crystallization from dissolved silica.
Quartz (SiO2> is one of the hardest of the minerals, but crystal size is small in
water-formed deposits so that the extreme hardness of the natural mineral is
not imparted to the deposit. A unique property of solubility in high-
temperature steam provides a mechanism for its crystallization as a solid

when the steam temperature is reduced. Quartz formation in steam turbines
is caused by volatile transport of silica from the boiler water. Cristobalite is
another form of crystalline silica in steam turbine deposits. Uncombined
silica may also be a component of siliceous boiler scales.
Free silica in heat exchanger deposits results from the deposition of sand
or dirt suspended in the cooling water. The principal property of such
deposits depends upon the substance forming the matrix of the individual
grains of silica.
Simple Silicates
The silicates of single cationic elements are classified as simple silicates.
The calcium and magnesium silicates have incongruent solubilities and
contain water of hydration when formed hydrothermally. The composition
of the solids is dependent upon temperature and pressure as well as on the
relative concentrations of the dissolved reacting ions.
The hydrous magnesium silicates, serpentine (3MgO-2SiO2'2H2O), talc
(3MgO-4SiO2-H2O), and sepiolite (2MgO • 3SiO2 • 2H2O), are soft, rather
poorly crystallized compounds. Detection by instrumental means is not
always successful. They are more frequently components of soft sludges
rather than hard scales. The formation results from the alteration and
crystallization of an absorption complex of dissolved silica on magnesium
hydroxide. Serpentine forms under the most varied conditions and for this
reason is the most common of the tfiree forms.
The hydrous calcium silicates, gyrolite (2CaO3SiO2-H2O), xonotlite
(5CaO-5SiO2-H2O), and foshagite (5CaO3SiO2-3H2O), like the hydrous
magnesium silicates, crystallize poorly but unlike them may be hard, tough,
and highly insulating. The common form is xonotlite, with gyrolite and
foshagite being somewhat rare.
The anhydrous simple silicates, fayalite (2FeO • SiO2), mullite (3AbOr
2SiO2), olivine (2(Mg, Fe)O-SiO2), and forsterite (2MgO-SiO2), are
commonly high-temperature dehydration products of related hydrous forms.
Willemite (2ZnO • SiO2) is one of the few anhydrous simple silicates formed
hydrothermally.
The sodium silicates, Na2SiO3 and j3Na2Si2O5, are somewhat soluble in
high-temperature steam and therefore are components of turbine blade
deposits. Their water solubility is appreciable, so that they may be removed
by washing the turbine internally with wet steam.
Complex Silicates
Silicate compounds of more than one cationic element are classified as
complex silicates. These hydrothermally formed compounds are often
crystallized from environments of very high heat input and are usually
282 MANUAL ON WATER
hydrous. Regions of high heat input become highly concentrated in boiler

water salts that contain sodium silicate. Conditions favorable to reactions
with oxides of aluminum, iron, and calcium to form the respective sodium
silicate complexes exist in these regions. The sodium silicate complexes are
not easily corrected by treatment of the water, since they may form in water
with no hardness present.
All are very hard and adherent and resist dissolution by ordinary acid
cleaning procedures. They may form in place on the steam-generating
surface or by reaction with sodium silicate and accumulated iron oxides as in
the case with acmite (Na2O-Fe2O3-4SiO2). Analcite (Na2O • A12O3 • 4SiO2 •
2H2O) is the commonest aluminum analog and may be present is sludge
form as well as scale. Natrolite (NaaO AhOs-BSiCVZHaO) and noselite
(5Na2O3Al2C>3-6SiO2-2SO3) occur much less frequently than analcite.
Acmite has also been found in turbine blade deposits, suggesting that the
components of the reaction may be steam soluble.
Cancrinite (4Na2O-CaO-4Al 2 O 3 -2CO2-9SiO2-3H2O) forms when
calcium sludge is contained in the concentrated sodium silicate and sodium
aluminate reaction liquid. Pectolite (Na2O-4CaO6SiO2'H2O) is a boiler
scale under aluminum-free conditions.
Sulfates
Of all the scale formers, calcium sulfate is most readily recognizable
because of the large crystals which are formed. Anhydrite (CaSO4) crystal-
lizes on heat-exchange surfaces in the form of large needle-like crystals
which grow perpendicular to the heat-exchange surface. The individual
crystals can be seen by the naked eye. Hemi-hydrate (CaSCV VSIfcO) and
gypsum (CaSO-rlHhO) are lower-temperature forms of calcium sulfate, the
former more likely to be found on steam generating surfaces and the latter in
nonsteaming exchangers.
Celestite (SrSCX) and barite (BaSO4) are highly insoluble but rather rare
because of the infrequency of strontium and barium in industrial water.
All sulfate scales are difficult to remove as they do not respond to acid
cleaning.
Sodium sulfate occurs in superheater and turbine blade deposits as
thenardite (Na2SO4-V), metathenardite (Na2SO4-I), Na2SO4 III, and burk-
eite (2Na2SO4-Na2CO3).
Sulfldes
Sulfides of metals are quite insoluble but generally do not protect metals
on which they are formed. Sulfides are usually associated with corrosion in
the presence of hydrogen sulfide or with a reducing reaction on a sulfate.
Cuprous heat-exchanger tubing may contain chalcocite (Cu2S) or covellite
(CuS).
Boiler and superheater tubes have contained ferrous sulfides (FeS)

pyrrhotite and troilite, probably as high-temperature reduction products
from sulfur originally as sulfate or sulfite.
A list of water-formed compounds is contained in Table 1 of Chapter 14.
Sampling of Deposits
Deposits are sampled at least twice before being submitted to chemical or
physical tests. The field man first collects the gross sample from its point of
formation, and the technician then prepares this sample for final examina-
tion. Deposits are not homogeneous but may differ in composition from one
part of a system to another. It is therefore important that the field samples be
taken from the point of occurrence without physical or chemical alteration.
Sampling Adherent Deposits

The removal of a sample that adheres closely to the surface upon which it
has been deposited is often extremely difficult. This is true of many deposits
in boiler tubes, pipelines, and condensers. It is sometimes possible to remove
a short section of a pipe or condenser tube, and then to dislodge the deposit
by mechanical or thermal shock.
In most instances, a portion of the deposit must be removed without
damage to the surface to which it adheres. Removal of a hard adherent
deposit, such as analcite (Na2O-Al2O3-4SiO2-2H2O), anhydrite (CaSO4),
gehlenite (3CaO • AhCh • 2SiO2>, serpentine (3MgO • 2SiO2 • 2H2O), magnetite
(FeaCk), quartz (SiCh), or magnesium hydroxide (Mg(OH>2), will require
the use of a chisel, a hammer, or an excellent steel scraper. The use of
such tools to remove an extremely adherent deposit requires care. A
representative sample should be obtained, but damage to the equipment
must be avoided. It is often impossible to remove those samples of a deposit
that would be most desirable for solving the problem of deposition. Any
sample, in such cases, is better than none, even though it may not be the
most desirable one.
Hard, adherent deposits are sometimes encountered on the buckets,
nozzles, and diaphragms of steam turbines. A closely adherent deposit can
be removed from the buckets of a steam turbine by peeling or scraping
downward over the surface of the bucket with a sharp penknife, the cutting
edge of the knife being used with the point down. A small envelope, with the
sealing flap open and extending upward, is a very good receiving receptacle.
Squeezing the edges of the envelope will make the sealing flap concave. By
holding the top edge below the point of the knife in peeling, the loosened
deposit drops conveniently into the open envelope. Similar deposits often
found in heat exchangers or evaporators can be removed for sampling by
cracking.
284 MANUAL ON WATER
Loosely Adherent Deposits

Deposits that adhere loosely to the surface upon which they have formed
are readily removed. Many types of scrapers can be used: knife, spatula,
spoon, a thin piece of wood, sheet metal, or even cardboard. Deposits of this
kind are sludge in steam generators (boilers), cooling towers, cooling
systems, air-conditioning systems, feedwater heaters, water softeners, con-
densers, and heat exchangers, and also biological deposits in various water
systems.
Thin Films
Some deposits, particularly on rough or irregular surfaces, may be so thin
that scraping does not remove much of the material. If the deposit is not too
hard, a stiff bristle brush may be much more effective than scraping. If the
deposit is tightly adherent, and the surface is smooth, thermal or mechanical
shock will sometimes loosen it. Thermal shock may produce some alteration
or decomposition of the deposit; note should be made in the information
accompanying the sample when this method has been used.
Biological Slimes
It is frequently impossible to identify a biological slime definitely if it is
allowed to become dry before examination. If at all possible, therefore, the
deposit should be kept in its environmental water. In some cases, where gas
formation is active, as with sulfate-reducing bacteria, it may be necessary to
refrigerate the sample during shipment, or to add a bactericide. The latter
procedure of course kills the bacteria; it would make preparation of cultures
impossible. The discretion of the sampler must be relied upon in such
instances.
Selection of Samples
The selection of a sample or samples is usually somewhat limited, but
there are instances where a wide choice is possible. In cases where it is
possible to select the samples of a deposit, such selection is best guided by a
thorough consideration of the problem involved. Very often, the removal of
a number of samples will result in more informative analytical data than
would be obtained from one composite representing the entire mass of
deposit. A typical example is the sampling of deposits from a steam turbine.
Samples from each stage are more desirable than a composite sample from
all stages. Conversely, in the case of a tube failure in a steam generator, if
inspection shows other sections of the boiler to be in satisfactory condition, a
single sample from the affected area should suffice.
Deposits sometimes are encountered in layers. Usually it is possible in
such cases to select several samples so that the individual layers may be
examined in the laboratory, if desired.
Sample Quantity
The most desirable amount of deposit to be submitted as a sample is not
specific. The quantity of deposit that can be removed is often limited. In such
instances it is better to submit a single mixed sample of half a gram or less
than to collect no sample.
A deposit of 5 g is desirable for an extensive X-ray diffraction study,
although informative data may be obtained with as little as 0.1 g. A routine
chemical analysis usually is possible with a sample of 10 g, but an elaborate
laboratory investigation might require as much as 100 g.
Although there is a difference of opinion about the desired amount of
deposit to be submitted as a sample, it is definitely best to provide
considerably more sample than laboratory examination might require.
Labeling and Shipment of Samples

After collection in the field, the sample must be properly protected to
prevent change or contamination during shipment. A glass or plastic
container is adequate for samples of deposit only, and a clean wooden box
sealed against atmospheric moisture is suitable for samples that include the
underlying surface. Biological deposits must be collected and kept in contact
with the mother liquor, and be shipped in refrigerated containers.
The gross sample should always be accompanied by a detailed history.
This includes the name and description of the equipment from which the
sample was removed, the precise location, the appearance and extent of
the deposit prior to removal, the exact method used in removing the sample,
the temperature and pressure of the liquid or vapor phase from which the
deposit was formed, the chemical analysis of the water, and an account of
any abnormal or unusual conditions existing during formation of the
deposit. Further details are given in ASTM Field Sampling of Water-
Formed Deposits (D 887) [/].'
Laboratory Sampling
The final sampling is made by selecting a small representative portion
from the gross sample for examination by special instruments. If the sample
is received in contact with its original confining surface, the same pre-
cautions must be taken as described earlier in removing it from that surface.
If it is possible to make a separation of heterogeneous layers, it may be
advisable to do so, and to make separate examinations of these smaller select
286 MANUAL ON WATER
portions. The composition next to the confining surface may be different

from that which was in contact with the water. Separation of these two
different surfaces should be made to ascertain this possibility. After the final
separations have been made, the handling of the sample for each type of
instrument becomes part of a specialized technique for that particular
instrument. For example, portions on which index of refraction is required
should be crushed to final size, whereas the portions should be ground for X-
ray and spectrographic examination. Various other physical and chemical
treatments may be given to portions of the sample undergoing identification.
Methods of Identification
The examination of any deposit begins with visual observation and the
recording of the description in terms of size, shape, contour, color, odor,
hardness, and magnetism. The choice of subsequent methods for identifica-
tion is frequently suggested by the visual appearance. In the case of inorganic
deposits, the proper step, after visual examination, is to study them under
magnification. This often reveals characteristics not visible to the unaided
eye. The texture of various layers is more pronounced, and recognition of
individual crystals may be possible.
After crushing or grinding, but before other tests are run, it is desirable to
heat a small amount of the sample gently in a crucible and to note its
behavior. A pronounced darkening of the sample is frequently indicative of
organic matter. The odor as the sample is heated can be used to identify
many types of organic material, such as carbohydrate, protein, amines, oil,
coal, sulfur compounds other than sulfate, and a number of others. The
suspicion of a given type of organic compound from the odor on heating
may be verified by extracting the sample with a suitable solvent, and using
the microscope, X-ray diffraction, or absorption spectroscopy to help
confirm the initial suspicion. Table 1 indicates the various techniques
applicable to the detection and identification of the types of material listed,
and is intended as a reference to the instrumental descriptions that follow.
The identification of water-formed deposits is not a matter of using the
proper single method, but of coordinating a number of methods to make
each supplement the others [2,3]. No single method tells all there is to be
known about a sample. Each has a special field of application with its
inherent advantages and disadvantages. The applications and limitations of
commonly used methods of identification are discussed in the following so
that the proper instrument or combination of instruments may be applied to
the investigation of these deposits.
Biological Microscope
The term "biological microscope" does not refer to a particular type of
microscope but rather to the use to which the microscope is put, namely the
TABLE 1—Comparison of principal fields of application and limitations of the methods

used for identification.
Instrument Principal Application Limitations
Biological microscope identification of microorganisms number of organisms must be
in industrial water samples. sufficiently great to be ob-
Recognition of inorganic sus- servable in the microscopic
pensions and differentiation field.
from organic matter.
Chemical (petrographic)
microscope identification of individual crys- particle size must be not less
tals and amorphous clumps. than 10 /um. Crystals must be
Examination of microstruc- non-opaque. Sample must be
tural details. Qualitative micro- insoluble in immersion
chemical analysis. medium.
X-ray diffraction identification of crystalline compounds must not be
compounds. amorphous but particle size
may be less than 0.01 pm.
Spectroscope qualitative detection of cationic sample must be low in volatile
constituents, constituents.
Spectrograph quantitative estimation of minor same as for spectroscope.
cationic constituents.
examination of biological matter. The essential requirement is to magnify the

objects under observation so their distinctive form and features become
readily recognizable. The biological microscope operates functionally as
illustrated in Fig. 2. Light from a source such as a lamp is reflected in a mirror
FIG. 2—Functional diagram of biological microscope.

288 MANUAL ON WATER
located at the base of the instrument. The reflected beam is directed upward
through a condenser and through the sample, which has been mounted on a
glass slide and covered with a thin cover glass. The image of the micro-
organisms is magnified as the light continues upward through the lenses
of the objective and the eyepiece. The magnified image is either observed
directly at the eyepiece or photographed with a camera.
The living matter in biological deposits may consist of either single cell or
multicell species performing the functions of growth, reproduction, and
metabolism. Wide variations exist, however, in the conditions under which
these functions are performed. Indeed the environmental conditions are often
quite limited in this respect for any one kind of organism. For maintenance of
life some organisms require iron, some sulfur, some carbon, all with or
without oxygen, and they may be deprived of these essentials by changes in
industrial operations. Thus, in addition to the living organisms, the sample
may contain metabolic by-products and skeletons of dead organisms.
Observations of the material are made by enclosing a small drop of the
sample-bearing liquid between a glass slide and a cover glass. The outlines of
the objects under examination are observed under transmitted light. The
various shapes may be described as similar to rods, filaments, chains, stalks,
and capsules. Figure 3 shows some typical examples of magnified biological
material.
Identification of the specific species is made by recognition of charac-
teristic shapes and colors or by characteristic reaction to reagents. Some
organisms will adsorb specific dyes and be recognizable by this property.
Inorganic oxide sheaths may be dissolved from an organism by acids,
thereby exposing features not recognizable with the sheath in place. If the
sheath is iron oxide, the dissolved iron will color the liquid between the cover
glass and the slide, suggesting an iron bacterium.
Recognition by shape is sometimes difficult and, for confirmatory in-
formation, the organisms may be inoculated into a culture medium. After a
suitable incubation period the medium is tested for specific metabolic by-
products of the organism. For example, the formation of hydrogen sulfide in
a sulfate culture medium is confirmatory evidence of the presence of sulfate-
reducing bacteria.
Not all samples as received in the laboratory are suitable for biological
examination. When the number of organisms is so small that a single drop of
liquid might not contain any observable individuals of a suspected species,
the sample may be centrifuged to concentrate the organisms and the
observation made on the concentrate. On the other hand, samples may
contain so much clay or other suspended matter as to greatly interfere with
the examination. Such samples may be diluted to reduce the concentration
of impurities, or separation of biological from inorganic material must be
made before mounting the sample under the microscope.
The report of a microscopic examination usually indicates only those
recognized microorganisms for which definite proof of presence was estab-
(a) (No. 10) Cyclops large organism. Also _(<e) Gram positive bacillus,
shows (A) Synedra, (B) Ceratium, and (Q CO Gram positive cocci.
Nastoe. (g) Asterionella.
(b) Crenothrix with ferric oxide sheath in (h) Stephanodiscus.
place. (0 Crenothrix with iron removed and stained
(c) Spirogyra. with methylene blue.
(d) Anabena. (/) Ceratium.
FIG. 3—Typical biological shapes.

290 MANUAL ON WATER
lished. Quantitative values are assigned only when specific cultures can be
made.
A microscope with three objectives ( X 10, X43, and oil-immersion X 97),
two eyepieces ( X5 and X10), and an Abbe condenser can be purchased for
about $400. A more adequate assembly, including an illuminator and a
mechanical stage, can be bought for about $700.
The ASTM Test for Iron Bacteria in Industrial Water and Water-Formed
Deposits (Method D 932) [7]; ASTM Tests for Sulfate-Reducing Bacteria in
Industrial Water and Water-Formed Deposits (Method D 993) [7]; and
ASTM Identification of Types of Microorganisms and Microscopic Matter
in Industrial Water and Industrial Waste Water (Method D 1128) [7] provide
for identification of some bacteria and other microorganisms of industrial
importance. There are other sources \4-6\ of detailed information on the use
of the biological microscope.
The Chemical Microscope

The chemical (or petrographic) microscope is so named because of the
accessories and design which make it particularly suitable for the study of
both optical properties and microchemical reactions. Rocklike deposits lend
themselves most readily to direct examination by this microscope. Figure 4
illustrates the magnified structures of some typical water-formed deposits.
The functional operation illustrated by Fig. 5 is similar to that of the
biological microscope with the addition of a polarizer and analyzer. Light
from a source such as a lamp is reflected by the mirror and passes through
the polarizer and condenser. The polarized light passes through the sample,
which is supported by a glass slide mounted on a mechanical, graduated
stage. The image of the object is magnified as the light passes upward
through the objective, the analyzer, and the cross-hair eyepiece.
When ions in a solution react to form solid particles, the particles usually
consist of repeating geometrical arrangements of the atoms in a three-
dimensional network. This network of atoms or ions is called a crystal. If
conditions were such that only one solid phase could precipitate from a
solution, and the conditions of precipitation remained constant, the resulting
solid product would consist of one or more cyrstals of the same substance.
Ordinarily, however, it is possible for more than one substance to precipitate
from a water at the same time, and the conditions of precipitation are seldom
constant, so that we rarely have a single homogeneous deposit containing
only one crystal phase. Also, variations in the composition of the water and
changing conditions may give rise to a deposit consisting of layers of
material of varying composition. Berry, Allen, and Snow [7] of the U. S.
Steel Corp. have described a technique for examination of polished sections
by reflected light, using a metallurgical microscope. While their studies were
applied to refractories, the techniques are equally applicable to detailed
study of the layering in a coherent scale of sufficient thickness. The edge of a
deposit is ground approximately flat on a medium-grit polishing paper, after
(a) Striated calcium sulfate scale. (d) Calcium sulfate lienic hydrate deposits.
(b) Ferric oxide scale. (e) Magnetite scale.
(c) Magnesium phosphate scale.
FIG. 4—Typical water-formed deposits.
which the piece is impregnated by immersion in a molten resin, such as

glycol phthalate. The piece is removed while still hot, and allowed to cool in
air. This resin is fairly hard at room temperature, and allows wet polishing of
the surface of the scale without extensive tearing out of individual grains.
292 MANUAL ON WATER
FIG. 5—Functional diagram of chemical microscope.
Many of the epoxy resins, especially those of low viscosity, are also excellent
impregnating materials. Many crystal phases can be recognized by their
shape and brightness under reflected light, even though direct observation of
optical properties cannot be made.
In some cases, it is possible to make thin sections of a scale, so that the
structure may be examined by transmitted light under the microscope.
Observation of birefringence and crystal orientation may suggest the
compounds present in such a scale. Most boiler scales and sludges are not
sufficiently coherent, however, nor do they consist of sufficiently large
crystals for such techniques to be fruitful.
A crushed sample, preferably 100-mesh or finer, is suitable for two types
of microscopic examination. Many crystals and compounds in deposits from
water can be identified and their amounts estimated by examination of the
powder under a chemical microscope equipped with polarized light, or with
a petrographic microscope. Identification is accomplished by immersion of
the powder in oils of various refractive indexes; determination of the indexes
of the crystalline compounds is made by observing their relief against the oil
of known refractive index.
Refractive indexes are determined by crushing the sample to a size which
will free from the agglomerate as many single crystals as possible without at
the same time producing particles too small for clear resolution under the
microscope. Sludge particles are often too small to provide optical data.
Large-size particles are no deterrent because these can be crushed to
sufficiently small size for measurements.
Amorphous material and glasses, as well as crystallized compounds, have
optical characteristics that can be measured and may form a basis for
identification. For examination of this type of material, optical microscopy
is superior to X-ray diffraction.
Chemical microscopy is simply the technique of performing normal
qualitative chemical tests on a microscope slide. The advantage of the

microscope for this purpose is twofold. In the first place, observation of the
crystals formed by these qualitative tests may permit distinction between
several ions, all of which form precipitates with a given reagent. The second
advantage is that the tests are quite rapid, and may be carried out with a very
small volume of solution. This is of utmost importance where the size of the
sample is limited. Chamot and Mason's Handbook of Chemical Microscopy
[8] gives details for carrying out most of the common microchemical tests,
and also gives illustrations of many of the crystals formed by these reactions.
Tests for phosphate, ferrous and ferric iron, copper, zinc, nickel, calcium,
magnesium, manganese, sodium, sulfate, chloride, and a number of others
can be readily made by this method.
The results from observations with the chemical microscope include a
description of the appearance of the magnified gross structure, and a report
of the presence of specific compounds and elements as determined from their
refractive indices and reactions with spot-test reagents. The chemical
microscope is a standard instrument and differs from the biological micro-
scope in that it is equipped with a graduated rotating stage and polarizing
prisms. Normal cost is between $650 and $1400. Detailed procedures for
petrographic examination and tables of optical data are available in
standard texts [9-12].
X-Ray Diffraction
X-ray diffraction produces a pattern of lines on photographic film or
peaks on a recorder chart. The position of the lines on the film or the peaks
on the chart for any given type of X-radiation is determined by the distances
between layers of atoms in the crystalline components of the sample. These
interatomic distances can be calculated from the measured positions of the
lines or peaks, and used to identify the crystalline compounds of a sample.
The essential functioning of the X-ray diffraction unit is illustrated in Fig.
6. X-rays pass through a collimator tube which restricts the beam to a very
FIG. 6—Functional diagram of X-ray-diffraction unit.

294 MANUAL ON WATER
thin pencil of parallel rays. The collimating tube may be an integral part of
the camera. The pencil of X-rays is intercepted by the sample located on the
central axis of the camera. A large percentage of the beam penetrates the
sample and is caught and dissipated in the beam trap. A small fraction of the
primary beam is diffracted to the outer circumference of the camera and
strikes the photographic film at definite positions depending upon the angles
of diffraction. The magnitude of these angles of diffraction is determined by
the atomic arrangement within the crystal compound producing the diffrac-
tion. After sufficient exposure, the photographic film is developed and
patterns are produced, those shown in Fig. 7 being typical. These are
compared with patterns or data produced by standard compounds. When
the position and intensity of lines from the pattern of an unknown
compound match those from the pattern of the standard, the unknown
material is identical with that of the standard. Identification therefore
requires a series of standards or a set of authenticated diffraction data. The
best single source of crystal data is the "Card Index File of X-ray Diffraction
Data for Chemical Analysis," available from ASTM. Details of procedure,
with additional literature references, are given in ASTM Identification of
Crystalline Compounds in Water-Formed Deposits by X-Ray Diffraction
(Method D 934) [/].
The X-ray diffraction method of analysis is essentially one of identifica-
tion of crystalline compounds present in a deposit. It is sometimes referred
to as the "powder diffraction method," a name that indicates the most useful
form in which samples are prepared for analysis. Most crystalline, water-
formed deposits do not occur as powders but must be finely ground for
proper mounting and for crystal distribution. Sludges usually are very finely
divided but may cement or harden on drying and may also contain flakes or
larger particles which are truly part of the sample. The only limitations on
the type of sample that can be used for powder diffraction are that it be
crystalline, that it be susceptible to grinding, and that a sufficient amount be
available for manipulation.
Although the X-ray diffraction method is not usually considered to be a
micro-method, with special care an analysis can be made with less than 10
mg of sample. Exposure of a sample to X-rays may require from 1 to 4 h,
and samples subject to change in the atmosphere in that length of time may
be protected by enclosure in an airtight capsule during exposure.
In recent years, direct measurement techniques of X-ray diffraction have
been introduced, employing counter tubes for detection of X-ray intensity.
These instruments have distinct advantages. Ordinarily, they employ a
somewhat larger sample, so that better integration of the overall com-
position of a sample is possible. Further, although only one sample may be
run at a time, the method is considerably faster than photographic methods,
because under most circumstances sufficient of the pattern can be run in a
matter of 30 min to 1 h. For detection of the presence or absence of a certain
phase, it is necessary to scan only over an area where a line unique to the
FIG. 7— Typical X-ray-diffraction patterns.
296 MANUAL ON WATER
compound being sought is located. For instance, a time 2 or 3 min is

sufficient to determine the presence of quartz in a turbine deposit. The
pattern obtained is more nearly quantitative than that from film. Accurate
quantitative analysis, however, must depend on careful consideration of the
absorption characteristics of the sample, whether one is using the photo-
graphic or direct-measurement technique. Because of the fact that a large,
flat sample is used, direct-measurement techniques are somewhat prone to
preferred orientation unless the sample is prepared very carefully. This
property is sometimes an advantage in qualitative analysis, as it may
markedly increase the intensity of the strongest line, thus allowing detection
of smaller amounts. A good example of this is anhydrite, which can be
detected in a sludge in concentrations as low as 1 percent because of this
orientation, even though the anhydrite pattern is not exceptionally strong.
In the absence of substances in the sample that give high background, the
sensitivity of the direct measurement technique is about the same as that of
the photograph for those compounds which are well crystallized. With
poorly crystalline substances, the advantage is decidedly with the direct-
measurement, or diffractometer, technique because broad peaks whose
intensity is not greatly above background are much more easily detected by
variation in the background level on a chart than gradual differences in
blackening on a photographic film. Thus, the sensitivity for poorly crystal-
lized magnesium silicate in a boiler sludge is frequently appreciably greater
with a diffractometer than with photographic methods.
The method does not ordinarily detect low percentages of compounds in a
deposit. The minimum amounts that can be detected may vary from 1 to 30
percent, depending upon the efficiency of the compound as a diffractor and
the absorption characteristics of other materials in the mixture. For
example, 1 percent copper can be detected in magnesium hydroxide, whereas
from 5 to 10 percent copper must be present for detection in an iron oxide
sample. Copper is a good diffractor and magnesium hydroxide has low
absorption compared with that of the iron oxides. Poorly crystallized
compounds, such as frequently occur in the case of magnesium silicate, may
escape detection when present in amounts as great as 30 percent.
Solid solutions frequently can be detected if the solute has appreciably
changed the lattice dimensions of the solvent phase. Solids of the solid
solution type are common with such sodium salts as are deposited in
superheaters and drypipes.
Crystal size or degree of crystallization may be determined qualitatively
from the uniformity and breadth of the diffraction lines in a pattern. It is
frequently possible to differentiate between the poorly crystallized and the
well-crystallized species in a mixture.
In both of the foregoing techniques, the diffractometer has a marked
advantage over the photographic method. Photographic film is not entirely
dimensionally stable, and it is therefore sometimes difficult to detect very
slight solid solution. Also, in a camera that will allow reasonable exposure,
the lines on the film are relatively small, and it is not possible to detect very
slight variation in line width.
By scanning over the area of interest at slow speed with the diffractometer,
peaks of sharp lines can be located within 0.01 deg, thus permitting detection
of extremely slight solid solution. Further, by employing narrow slits and
fast time constants in the instrument, it is possible to achieve extremely high
resolution, of an order that would be almost physically impossible on a
photographic film. This allows at least partial separation of diffraction lines
of well-crystallized substances which are so slightly separated that they
might appear as a single line on a film. Of course, if the lines are exactly
superimposed, they cannot be separated even by high-resolution techniques.
There are, however, several distinct advantages to the photographic
method. First, a pattern of high quality can be obtained on a very small
amount of sample, such as 2 or 3 mg. Second, in a sample that gives a very
complex pattern, such as one containing a large number of crystal phases,
certain compounds found in deposits from water are prone to give somewhat
spotty lines if they are not ground sufficiently fine. These spotty lines stand
out quite well on the film, and permit separation of the pattern into two
distinct types of lines, which frequently aids identification. Quartz, calcium
sulfate, calcium carbonate, and analcite are compounds that often develop
rather large crystals and are particularly likely to give spotty lines. The third
advantage for film lies in the ease with which a number of X-ray patterns
may be compared by placing the films edge to edge on a viewing box. This is
done much more easily with films than with charts.
Because of its ability to give useful patterns with samples consisting of
very small crystals or of crystals that are opaque to light, the X-ray method is
generally more widely applicable to deposits than the petrographic micro-
scope. The microscope, however, can give useful information about the state
of aggregation, can detect amorphous materials which contribute little or
nothing to the X-ray pattern, and permits identification of a single crystal on
a whole slide of material, thus providing a sensitivity much greater than that
of the X-ray method, provided the crystal is of sufficient size. Where both
instruments are available, it is generally advisable to run the X-ray diffrac-
tion pattern first, thus giving a good average composition of the sample, and
to use the petrographic microscope to detect glasses and amorphous
materials—those substances present in too small amounts for X-ray de-
tection—and information about the state of agglomeration of the sample.
The report of an X-ray diffraction analysis should give the name and
formula for the crystalline species identified, together with an estimate of the
relative amounts of the identified phases present, and any remarks regarding
crystal size, solid solution, or other pertinent information that may have
been obtained.
The equipment used in the photographic method consists essentially of a
298 MANUAL ON WATER
source of constant high-voltage electrical current, an X-ray diffraction tube

in an appropriate mount, and a cylindrical camera. Suitable commercial
units are manufactured at a cost of $7000 or more.
The diffractometer is considerably more complex, and consists of a source
of highly stabilized high-voltage electrical current, the X-ray tube, a counter
tube, an amplifier, ratemeter and recording circuits, and an adjustable
goniometer for mounting the counter tube and the sample and providing for
rotation of the counter tube about a focusing circle upon which are the X-ray
tube, the sample, and the counter tube. Such equipment is available as
commercial units starting at about $14 000.
X-Ray Spectroscopy
Secondary X-rays are generated when X-rays of sufficiently short wave-
length are absorbed by matter. These secondary rays are characteristic of the
elements present in the irradiated material. The X-ray spectrograph is an
instrument for separating the X-radiations from the various elements of the
absorbing material, and for detecting and recording their wavelengths and
relative intensities. The apparatus used for such detection is not greatly
different from the X-ray diffractometer, and most of the components of one
instrument are used for the other.
The chief difference lies in the fact that the sample is the source of the X-
radiation in X-ray spectroscopy, rather than the X-ray tube, as in dif-
fraction. A single crystal, cut parallel to one of the principal planes, takes the
place of the powdered sample in the diffractometer technique, and diffracts
the various wavelengths of X-rays at different angles, thus allowing each
wavelength to be separately detected by the counter tube.
X-ray spectra are very simple, consisting at the most of about 15 lines, not
all of which are of significant intensity. These lines as a group vary in
wavelength in a regular manner with the atomic number of the element.
Complete X-ray spectra are therefore known for all elements. The wave-
lengths have been precisely calculated, and most of them have been very
carefully checked experimentally.
The smaller the atomic number of the element, the longer the wavelength
of the corresponding lines of the X-ray spectra. "L" series spectra are used in
analysis for the heavy elements, that is to say, those heavier than barium,
and the "R" series for the lighter elements. Elements of atomic number
smaller than titanium give rise to radiation of sufficiently long wavelengths
that their absorption in air is too great for good sensitivity. A helium
atmosphere is therefore maintained from the sample, past the analyzing
crystal, to the window of the counter tube. This has allowed detection of
elements as light as sulfur. Softer radiation than this is absorbed strongly in
the window of sealed counter tubes, and counting losses are consequently
high. This has been solved recently by the use of a flow counter tube
employing a thin organic film such as Mylar2 and Formvar2 as a window in

the tube, and, since this tube is not gastight, allowing the counting gas to
flow continuously through the tube. Elements as light as sodium have been
detected by such counters, and quite useful intensities have been obtained for
aluminum, silicon, phosphorus and, except for quite low concentrations, for
magnesium. This method is useful over a very wide range of concentrations,
from as low as a few parts per million in favorable cases up to 100 percent.
Some elements influence the excitation of others in a manner that would be
predicted from a knowledge of the laws of absorption of X-rays and the
periodic table of the elements.
Optical Spectroscopy
Emission spectroscopy is a method of analysis for elements that depends
on the fact that when atoms are excited by either thermal or electrical means,
light is given off as discrate wavelengths as the atoms return to their normal
state. If this light is defined by a slit, separated into its various wavelengths
by a prism or diffraction grating, brought to a focus by a lens system and
viewed with an eyepiece, recorded on film, or measured by photoelectric
means, the elements that have given rise to the spectral lines can be identified
by locating the characteristic wavelengths in published tables of spectra.
The common sources of excitation for spectrographic analysis are flame,
d-c arc, a-c arc, and high-voltage spark. Each has special uses, and the
advantage and limitations of each one will be discussed briefly.
The flame of a Bunsen or Meker burner is at a sufficiently high
temperature to excite the atoms of only the alkali and alkaline earth metals
and a few others, if the concentration is sufficiently high. Propane-air,
propane-oxygen, acetylene-air, oxygen-acetylene, and oxygenhydrogen
flames produce successively higher temperatures. The atoms of about 43
elements can be excited with acetylene-oxygen or hydrogen-oxygen flames.
The sensitivity for many of these elements is not very high. The flame is a
highly reproducible source, however, if fuel and' air or oxygen pressures are
carefully controlled; it is quite satisfactory for the analysis of solutions which
represent ideally homogeneous samples. The flame spectrophotometer is
actually a small direct-reading spectrograph.
The d-c arc, which is operated at a voltage from 50 to about 250, is an
extremely sensitive source because it is capable of producing very high
temperatures. The detection of a number of elements in parts-per-million
concentrations in a solid sample is quite possible with this energy source.
The reproducibility of this source is somewhat limited because of the
tendency of the arc to wander. It is therefore considered highly satisfactory
2
These are registered trademarks of E. I. du Pont de Nemours & Co., Inc., and Shawinigan
Products Corp., respectively.
300 MANUAL ON WATER
for qualitative analysis because of its sensitivity, but it is probably not the
best one for quantitative analysis.
The high-voltage a-c are, using a voltage of 1000 or more, is a steadier and
more reproducible source than the d-c arc and, except for extreme cases, has
adequate sensitivity for most of the elements. The chief drawback to this
source is the dangerously high voltage, which requires safety interlocks and
careful shielding for protection of the operator.
The high-voltage spark is a very reproducible source, using 10 000 to
50 000 V across the two electrodes, and a synchronous interrupterto extinguish
the spark and prevent it from becoming an arc. The discharge is predom-
inantly electrical in nature, rather than thermal, and in many cases it is cool
enough that one can handle the electrodes immediately after exposure. A
condenser and inductor are usually added to this circuit to produce some
heating effect. Although the reproducibility of the discharge is quite high,
because of the low temperature comparatively little material is consumed,
and sampling problems are sometimes major. The spark source, however, is
readily adapted to the analysis of solutions, where a completely homogeneous
sample is possible.
The simplest type of dispersing instrument is the visual spectroscope,
commonly called the Bunsen spectroscope. It consists of a focusing lens, a
slit, a prism, and a telescope on a movable arm for viewing the spectra
produced. In addition, a wavelength scale is usually sent through the prism
from a side tube, and brought to focus in the telescope just above the spectrum
of the sample to be studied. The spectroscope is limited to the elements
that have lines in the visible part of spectrum, and is generally considered
to be a qualitative instrument, although some estimate of amounts can
usually be made. The instrument is simple and it has many potential uses,
but it has the drawback of poor dispersion, which makes it difficult to resolve
some lines that are normally close together. In addition, simultaneous ob-
servation of all parts of the spectrum is not possible.
The spectrograph employs either a prism or a diffraction grating as the
dispersing device and, for general use, photographic film as the recording
medium. For routine production control, where the same elements are
normally being determined, it is entirely practical to replace the photo-
graphic film with a photoelectric device for recording the relative light
intensities of the various elements to be analyzed. Generally, however,
problem in industrial water treatment are sufficiently varied that the direct-
reading spectrograph would be of limited use.
Metals, which conduct electricity, can be used as their own electrodes, but
most of the materials encountered in industrial water treatment are non-
conductors. Therefore, the sample is usually placed in a small core in a
carbon or graphite electrode, which is generally made the lower electrode.
The upper electrode is usually a plain carbon or graphite rod, ground to a
blunt point. The cavity in the lower electrode may be formed with a center
post, to decrease wandering of the arc, and may also be undercut, the
narrow neck tending to increase the arc temperature by decreasing the con-
duction of heat down the electrode. Metal electrodes are occasionally used
for special cases, but graphite is quite satisfactory for general work.
Very satisfactory electrodes can be made for the a-c arc and high-voltage
spark by thoroughly mixing the powdered sample with graphite, often with
a buffer also, and pressing into a pellet under high pressure. The pellet is
placed in a metal holder, the other electrode being a tipped carbon rod.
This technique is not wholly satisfactory for use with the d-c arc, because
of the high temperature.
Fineness of the powdered sample is important. Since the sample usually
melts in the d-c arc, a 200-mesh sample is adequately fine. For the a-c arc
and high-voltage spark, however, where relatively little material is con-
sumed, the sample should be at least 325-mesh, or should be fused with a
flux to ensure uniformity. Alkali metal borates are satisfactory fluxes for
most samples. Solutions are atomized into the flame or spark. In the latter
case, a rotating graphite disk which dips into a container of the solution
can be used; the spark is generated at the side of the disk opposite the
solution container. One great advantage for the solution method is the
ease with which synthetic standards are prepared, and the simplicity of the
addition of an internal standard to the unknown sample to serve as a
reference. This procedure increases accuracy by allowing the determination
of concentration ratios rather than absolute values of concentration.
ASTM Committee E-2 on Emission Spectroscopy publishes a number of
methods in a single volume for determination of impurities in various
materials. Most of these, however, are not applicable to deposits from water
because of the lack of a uniform matrix. Possible methods have been pub-
lished elsewhere, and are currently being submitted to Committee E-2 for
inclusion in future publications. These so-called "Universal Methods" are
generally of somewhat lower accuracy than those for determining impurities
in rather pure materials, but are of much wider applicability. It seems
probable that many of these will be adaptable for use in deposits from water.
A complete spectroscope, with an adequate d-c arc source and various
accessory materials, can be purchased for slightly under a thousand dollars.
Spectrographic units, which include microphotometers for measuring the
photographic film or plate, start in the neighborhood of $13 500, and may
run as high as $45 000 (with computer interface and cathode ray tube
display, up to $160 000), depending upon the complexity of the equipment
desired.
The microphotometer is actually a dual instrument. It normally consists
of equipment for projecting a reference film, on which the location of the
sensitive lines of the various elements are indicated, beside that of the
unknown spectrum for qualitative analysis. In addition, a densitometer
portion permits determination of the photographic blackness, expressed as
percent transmittance, for any given line. A comparison of photographic
blackness with known samples is used to plot working curves for the
302 MANUAL ON WATER
various elements to be determined quantitatively, and the transmissions of

the lines on the unknowns are converted to concentrations from these
curves. The results of spectrographic analysis are recorded as elements,
without indication of the state of combination. A number of standard
reference books are available for all operations dealing with the spec-
trograph.
Electron Microscope
Many particles of sludge, as well as certain corrosion products, are too
finely divided to be studied with an optical microscope. These small par-
ticles can often be profitably studied by electron microscopy. The electron
microscope is similar to the optical microscope in many ways, but the entire
optical path is in high vacuum. Electrons are generated by a hot filament,
and are driven from the filament by a focusing cap at high negative
potential, located immediately behind the filament. The electron beam is
further focused by a series of electrostatic or electromagnetic lenses, corre-
ponding to the condenser in an optical microscope. The crystals, or other
matter making up the sample, absorb electrons according to their thickness
and the absorbing power of the atoms, and the shadow thus formed is
viewed on a fluorescent screen, or photographed. The resolving power of
the electron beam is very high, and magnification of the photographic plate
may be used to bring out more detail.
The sample for examination should be dispersed as much as possible, as
agglomerates of a large number of particles cannot be studied except at
the edges. A thin film of a resin such as Formvar or collodion is perpared
by placing a drop of solution of the resin on water and allowing the solvent
to evaporate. The film is then picked up on a platinum screen, and the
sample deposited on the film and allowed to dry.
A second technique that may be applicable is the replica technique,
normally used to study metal surfaces. In this method, the resin is applied
to the surface to be studied, allowed to dry and then stripped off, usually by
immersion in water. The film retains the contours of the surface to which
it was applied; thus it is thinner where there were elevations and thicker
where there were depressions. The film is removed from the surface of the
water on the platinum screen, as in the foregoing, and viewed with the
microscope. The replica may be shadowed with chromium metal evaporated
on the surface to improve contrast.
Much can be learned about the crystal shapes and the sizes of sludge
particles, the type of material which acts to bind them together, and ten-
dencies to twin, or branch, by study of electron micrographs. Most of these
phenomena occur in crystals that are under 1 Aim in size, and this makes
study with the optical microscope nearly impossible. Because certain unique
properties of a crystalline substance, such as refractive index, cannot be
measured with the electron microscope, this instrument should not be
considered a means of identifying a completely unknown substance. Rather,
identification of the phases should be completed prior to the electron

microscope study.
Molecular Absorption Spectroscopy

The identification of organic substances usually presents some problems
not common in the inorganic field. Most of the substances occurring in
water-formed deposits are either not crystalline, or they are sufficiently
complex mixtures that the crystal phases would not be identifiable by an X-
ray diffraction pattern. Many are soluble in the oils used for refractive index
determination, which limits the use of the microscope. Although methods
such as melting and boiling points, refractive index, elemental analysis, and
similar techniques can be used in some cases to identify pure compounds,
none of these is generally applicable to deposits, which usually are not pure.
Most organic compounds, and many inorganic as well, show sharp
absorption bands in the infrared region of the spectrum. These absorption
bands have their source in different modes of vibration and rotation within a
molecule. The most useful analytical wavelength range is from about 2 to 15
;um, or slightly longer in the case of the halogens. The following types of
linkage give bands in specific regions, which can be used to characterize the
particular group: C—H (aliphatic), C—H (aromatic), N—H, O—H (pheno-
lic), S—H, C—O, C=O (aldehyde), C=O (acid), C—C, C=C, C=C, C=
N, C—Cl, C—Br, C—I. Each specific absorption band will shift somewhat,
due to changes in adjoining groups, but the band will still be within the same
general region.
The source of infrared radiation may be a lamp whose filament operates at
a low temperature, thus giving more infrared and less visible light, or it may
be a Globar (see footnote 2) element. The temperature of the source must be
carefully controlled to prevent change in the distribution of spectral
intensity.
The light from the source is passed through the sample, then broken up
into its various wavelengths by a prism, usually of quartz or rock salt.
Alternatively, a diffraction grating may be used, but care to prevent
interference from overlapping orders must be taken. In the 2 to 15-/um range
commonly used for survey work, a rock salt prism is generally employed, as
the transmission of quartz drops off sharply beyond 3.5 /zm. Rock salt is also
the usual cell window for nonaqueous solution techniques.
After isolation of the desired wavelength band by a variable slit, the
radiation falls on a detecting and measuring device, most commonly a
thermocouple. Lead sulfide photocells are more sensitive detectors for
wavelengths up to about 5 Aim.
Scanning of the absorption spectrum is accomplished by rotation of the
prism rather than by movement of the slit. The output of the detecting device
is amplified and fed to a recorder.
A most useful survey device for completely unknown nonvolatile solid
samples is the potassium bromide pellet technique. The sample is ground
304 MANUAL ON WATER
briefly with powdered potassium bromide and the mixture is placed in an

evacuable die. A mechanical vacuum pump removes the entrapped air, and
the mixture is then pressed at about 103 425 kPa (15 000 psi). The absence of
air results in a solid pellet free of bubbles, which is used as the absorber. This
technique is not fully quantitative, but does tell where the absorption bands
lie.
More quantitative results are obtained by selecting solvents that do not
absorb in the region of interest and running the spectrum in these solvents.
Concentrations employed are of the order of 1 to 10 percent.
Gases also usually give characteristic infrared absorption bands, and these
bands are the basis for continuous monitoring of gas streams.
A few typical applications to water-formed deposits are the detection of
oil in turbine deposits and classification as to type, determination of the type
and degree of degradation of amines, and possible identification of the
source of stream pollutants.
Atomic Absorption Spectroscopy

Another type of absorption spectroscopy makes use of light absorption by
atoms rather than by molecules. When atoms are diffused by atomization
into a flame, these atoms absorb light of the same wavelength as is emitted at
high temperature by that same element. The amount of light absorbed by the
atomized atoms is proportional to the element concentration in the vapor of
the sample flame.
For example, nickel is to be determined in a sample of water. The water
sample is atomized into a wide vaporizing oxyacetylene flame. A beam of
light containing radiation emitted from a nickel source is passed through the
sample flame. The absorbance of the nickel radiation in passing through the
flame-containing nickel atoms is measured by a photometer. The amount of
nickel in the sample is determined by reference to a calibration curve
showing the relationship of absorbance to concentration.
Solutions in the micrograms-per-litre range of concentration may readily
be tested. Scales and solid deposits can also be analyzed with good
agreement with wet chemical analysis. The solid samples are first taken into
solution and the solutions are atomized into the flame in a manner similar to
that for water samples. The method has the advantage of speed, high
sensitivity, and relative freedom of interference of one element on another.
Complete atomic absorption spectrometers are priced to $3000 to $8000.
The foregoing descriptive material is intended to suggest some uses of
instrumental methods to aid in the identification of water-formed deposits.
Some of these methods are described in detail in various ASTM methods.
Others are described more fully in the references.
Summary
The primary object of all sampling procedures is to bring the original
specimen to the place of observation or analysis without alteration of
composition, structure, or physical properties. Methods of microscopy,

diffraction, and spectroscopy provide means for determining facts about the
chemical and physical nature of the sample which are not available by other
means. According to the type of information desired about the sample, there
is a choice of methods, each of which is highly specific. The fields of
principal application and limitations are compared in Table 1.
To indicate further the kind of information each method provides, a
typical analysis report for each method is given in Table 2 on two samples of
water-formed deposits.
TABLE 2—Typical report form for results by different methods of identification for two
different samples.
Sample A Sample B
CHEMICAL ANALYSIS
SiO2 19.7 SiO2 17.9
Fe203 26.6 Fe203 2.6
CaO 10.8 CaO 30.8
MgO 16.2 MgO 17.9
S03 0.1 SO3 0.1
P205 12.3 C02 24.2
Cu 5.4 Loss on ignition 30.5
Loss on ignition 8.5
MICROSCOPICAL ANALYSIS
Considerable amount of iron oxide present in Large amount of highly birefringent calcite
colored bands. Some hydroxyapatite and present. Magnesium silicate also present,
considerable amount of poorly crystalized Iron oxides present as a minor constituent,
compound, probably magnesium silicate.
Small quantities of dendrites of metallic
copper.
X-RAY DIFFRACTION ANALYSIS

Hematite, serpentine, hydroxyapatite, and Calcite and serpentine present in about equal
metallic copper all identified with probable amounts. Serpentine present in extremely
percentages in the order named. Particle size small particle size while the calcite was
for the serpentine and the hydroxyapatite present in two forms, one in crystals of quite
very small. large size and the other in crystals of very
small size.
SPECTROSCOPIC ANALYSIS
Constituents over 5 percent: Constituents over 5 percent:
Si, Fe, Ca, and Mg Si, Ca, and Mg
Constituents 0.5 to 5 percent: Constituents 0.5 to 5 percent:
Cu Fe
Constituents 0.05 to 0.5 percent: Constituents 0.05 to 0.5 percent:
Mn and Cr Zn, Al, and Mn
SPECTROGRAPHIC ANALYSIS
Mn 0.13 Zn 0.43
Cr 0.07 Al 0.17
Ni 0.01 Mn 0.08
Cu 5.29 Pb 0.02
306 MANUAL ON WATER
A typical report from the examination of water by the biological

microscope is as follows:
Sample A—Organisms of the short rod-shaped variety were observed in
clusters of 1 to 10 surrounded by a mucoid capsule. Considerable quantities
of rust-brown hydrous ferric oxide surrounded the capsules. The organism is
identified as an iron bacterium of the genus Siderocapsa.
Sample B—Sample contained a number of protozoa which moved quite
rapidly across the field of vision of the microscope. Living diatoms of
characteristic shape were present along with other isotropic, suspended,
inorganic materials. Algae of both the green filamentous and the unicellular
variety were present.
References
[/] Annual Book of ASTM Standards, Part 31.
[2] Symposium on the Identification of Water-Formed Deposits, Scales, and Corrosion
Deposits by Physico-Chemical Methods. Proceedings, ASTM,'Vol. 43, 1943, pp. 1269-
1310.
[3] Symposium on Identification of Water-Formed Deposits, ASTM STP 256, American
Society for Testing and Materials, 1959.
[4] Bergey, D. H., Manual of Determinative Bacteriology, Williams and Wilkins Co., Balti-
more, Md., 1938.
[5] Manual of Methods for Pure Culture Study of Bacteria, Biotech Publications, Geneva,
N. Y., 1946.
[6] Tanner, F. W., The Microbiology of Foods, 2nd ed., Garrard Press, Champaign, 111.,
1944.
[7] Berry, T. F., Allen, W. C., and Snow, R. B., Journal, American Ceramic Society,
JACSA, Vol. 33, 1950, p. 121.
[8] Chamot, E. M. and Mason, C. W., Handbook of Chemical Microscopy, 3rd ed., Wiley,
New York, 1958.
[9] Larsen, E. S., and Berman H., "The Microscopic Determination of the Non-Opaque
Minerals," Bulletin 848, U. S. Geological Survey, 1934.
[70] Winchell, A. N., The Microscopic Character of Artificial Inorganic Substances of Arti-
ficial Minerals, Wiley, New York, 1943.
[11] Walstrom, E. E., Optical Crystallography, John Wiley, New York, 1943.
[12] Winchell, A. N., Elements of Optical Mineralogy, Wiley, New York, 1929, 1933, 1937.
Chapter 14—Chemical Arraly^

of Deposits
Many of the difficulties that arise during the industrial use of water are
closely related to water-formed sludges, scales, and corrosion products.
These water-formed deposits are the accumulations of insoluble material
derived from the water, or are formed by the reaction of water with surfaces
in contact with it. The composition of such deposits can be used to diagnose
the reason(s) for deposition, to determine preventive treatment of either the
water or its environment, and to effect their removal. A thorough and
accurate analysis is an important beginning in this process.
The analysis of water-formed deposits must be versatile and compre-
hensive in order to handle the most complex sample. It must also be flexible
enough to handle the simplest samples easily and efficiently. The usual
quantitative analysis will yield the elemental composition of the major
constituents of such deposits. When single elements or simple compounds
predominate, such an analysis is probably sufficient. For more complex
deposits, however, other tools must be used to determine accurately the
combinations of these elements. These include emission spectroscopy, X-ray
diffraction, X-ray fluorescence, electron probe microanalysis, flame photo-
metry, atomic absorption, petrographic microscope, and the electron micro-
scope. A method for atomic absorption and spectrographic analysis_is
outlined in ASTM Testing Water-Formed Deposits (Method D 2331) [7]1; a
procedure for identification by X-ray diffraction is dealt with in ASTM Test
for Identification of Crystalline Compounds in Water-Formed Deposits by
X-Ray Diffraction (Method D 934) [7]; X-ray fluorescence is covered in
ASTM Analysis of Water-Formed Deposits by X-Ray Fluorescence (Method
D 2332) [/]; flame photometry in ASTM Test for Sodium and Potassium Ions
in Industrial Water and Water-Formed Deposits by Flame Photometry
(Methods D 1428) [7]; and microscopy in ASTM Examination of Water-
Formed Deposits by Chemical Microscopy (Method D 1245) [7]; These and
other instrumental techniques are also discussed in Chapter 13.
It is often possible to determine the source or cause of water-formed
deposits from the results of chemical and physical analyses. The treatment
307
308 MANUAL ON WATER
for removal or prevention of such deposits can thus be accomplished

scientifically instead of by trial and error.
Frequently the analyst must know the composition of the water involved if
he is to make an accurate diagnosis of industrial water deposit problems. In
addition, information on the type and composition of equipment through
which the water passes, on the type of water treatment used, and on
temperatures and pressure existing in the system, can be useful. With this
information and the results of his own analysis, the chemist usually is able to
suggest a cause and cure for the water-formed deposit when he reports its
chemical composition.
Care must be exercised in selecting a representative sample for analysis,
since the value of an analysis is related to the sampling technique employed.
The analysis represents only that material actually analyzed, and, thus, the
information gained depends upon precise knowledge as to where and how
the sample was obtained. Sampling practice is dealt with in Chapter 13 and
ASTM Field Sampling of Water-Formed Deposits (Method D 887) [7].
Preliminary Examination of Deposits

The method of preparing a sample for analysis depends to a great extent
on the type of water-formed deposit: scale, sludge, corrosion product, or
biological deposit.
Immediately after removing the deposit from the container, its char-
acteristics should be recorded for later comparison with observations made
during the field sampling. Photographs are a very good form of record.
Pertinent information includes quantity, color, dimensions, hardness, mag-
netism, odor, contour, structure, and consistency. It is especially important
to note whether the sample was received in place on its confining surface. In
removing deposits from confining surfaces, great care must be exercised to
avoid contaminating the sample with particles of the surface.
When significant differences in deposit structure are visible, such as
laminations or colored bands, it may be desirable to divide these portions
and analyze them separately. Separations can be made with a sharp blade,
preferably under a low-power microscope. A separate bottle should be used
for each portion.
The entire sample or a representative portion should be freed from water
and oil so that it may be properly pulverized. If the material is very wet, it is
best to decant off the excess liquid (mother liquor) or remove the solid
material by filtration, and first to air-dry the sample or to dry it at a low
temperature such as 35° C to prevent loss by spattering. If the air-dried
sample smears or agglomerates when tested for pulverization, the sample
should be extracted with organic solvents, such as chloroform, and redried
at 105°C. If important to the solution of the problem, the solvent may be
evaporated and this extractable matter examined.
CHAPTER 14 ON CHEMICAL ANALYSIS OF DEPOSITS 309
Drying at 105°C and organic extraction constitute the first treatments if

mother liquor is absent. The dried material finally should be ground in a
mortar and quartered down to a sample of 10 or 15 g. This sample should be
further ground to pass a U. S. Standard Series, 149-jum (No. 100) sieve and
then should be bottled and tumbled to ensure thorough mixing. The bottle
should be labeled with all information necessary to identify its contents
completely.
Preliminary tests on the sample will often reveal sufficient information for
the analyst to make a direct approach to the analysis rather than to follow
the scheme discussed later. A spectrograph may be used for a quick
determination of the elemental major, minor, or trace constituents.
X-ray diffraction equipment operated by skilled personnel can be ex-
tremely useful for identifying compounds. Such equipment is not always
accessible and as a substitute, the petrographic microscope is very useful for
the initial examination. A simpler microscope, although limited in its
applicability, is useful if a petrographic model is not available.
If sufficient information accompanied the sample of water-formed de-
posit, shrewd guesses can be made about its likely composition. Preliminary
tests for these suspected constituents may be accomplished by means of spot
tests [2] and by simple qualitative tests. For example, carbonates will react
with acid to liberate carbon dioxide; iron can be dissolved in acid and
precipitated with ammonium hydroxide; sulfide can be detected by heating
with acid and noting the odor of hydrogen sulfide; and oil can be detected by
extracting with chloroform, evaporating, and examining the residue. Stand-
ard analytical texts will include complete lists of such qualitative tests.
Analysis of Deposits
Any analytical scheme can be only a guide, and the analyst must use his
skill and ingenuity in carrying out or modifying such a scheme. It is
impossible to draw a blueprint that will fit rigidly all water-formed deposits
that may confront the analyst. The amount of sample available may dictate
that only a few tests be made. Preliminary examination may show that only
a few tests are needed for determination of the important constituents.
Frequently, direct determinations for single constituents can be made
without going through a sequential analysis. Copper can be selectively
extracted with ammoniacal persulfate and the amount of copper determined
photometrically from the color of the extract. Silica can be dissolved from
the sample by molten sodium hydroxide and the extracted silica determined
colorimetrically with the molybdate-blue procedure. These and other rela-
tively rapid determinations of one or more constituents are given in ASTM
Method D 2331.
The analysis scheme shown graphically in Fig. 1, and the separate steps
are discussed in the following paragraphs.
310 MANUAL ON WATER
ORIGINAL FIELD SAMPLE

Preliminary examination (appropriate aliquot)
Photographic record Visual inspection Optical microscopic Physical tests
examination (magnetism, etc.)
Separate mother liquor
Mother liquor
(for analysis see text)
Dry (105 C) and extract with organics (CHCh, etc.)
I
Exploratory tests (aliquots of redried (105 C) extract residue
Organic extract
(for analysis see text)
Spectrographic X-ray diffraction Microscopic (Petregraphic and chemical) Qualitative

ISM ASTH Method D 2331) (See ASTM Method 0 934) (See ASTM Method D 1245 ) (chemical test)
Grind and screen (No. 100 U. S. sieve) remaining extract residue
H2S by evolution 1
*j£rt Water soluble CO? by evolution Ha and K ($« \ea and
(for onolysi, 1 ASTM Method D 513 ASTM Method
m Clnp D1428)
Ignition loss
Dissolve-HN03, HCland HC >4 (to fumes) - dilute and filter
Si02 and inclusim
HF treatment
Fuse residue with NajCOs
1 and add to main filtrate _
ASTH Method D 859 but use HCIO* instead ofS0H 2 4

(S*. Kit for ottonwt* in cose HC104 it prohibited)
Dilute to 1 liter (main stock)

(Base aliquot sizes on preliminary tests)
$64 Ca.ll B,Ba Mn Cra'ndV Reserve fi Pb, Cu, Ni, Zn, Fe. Sn, Al, and PCU
As BaS04 (See stock
by ASTM Method nl?
U5U
n Colorimetric- Titrate Add peroxide to Neutralize (NH4OH),
Method D 516 > ally with pot- FeS0 4 compare wi h add HN03,and electrolyze
assium periodate KMn04and color standar
0-Phenanthroline Cu and Pb plates
AddH2 04and Dissolve HN03
fil :er and replate
Add NH4C1, n iultalize Cu and Pb
BaS04 Dilute to volu le and aliquot
(KH4OH). Fil « and dis-
Calculate card ppt. Add
NH4OH, neutr •taST Fe.ZnjandNi Acidif
« add ammonium
JL oxalic acid (s,Jt.), double
ppt, filter and
Add NH4OH and filter molyodate and filter
Acidify filtrate Dissolve ppt. (HC1) -

F : *• (HjSCU), add H7S, Fe by ASTM Method D 1068
°f° Add excess NH4OHand filter, and ignite
Ca (NH4bHP04a nd filter 1 1
1 Z
f Add NrÔH to titrates filer
MgzPzOr Boil off
and Discard filtrate Calculate Z n Discard filtrate
MnjPjO; H2S, add NrUOH and
dimethylglyoxime Ammonium Mixed
Dissolve HNOJ-Fume, and filter ^ phomolybdate
H2S04, and determine Mn hydroxides
by ASTM Method D 858 color. Ni(C4H NzOzfe Add NH4OH and Dissolve HC1,
Subtract Mn from total mai nesia mixture. chill, add 6*
| Calculate Discard filterUtt gnite, and weigh cupferron
Mg-by difference 'i filtrate '
M82P207
Discard
filtrate Calculate PO* (See A,TM Method D 515)
Fume esidue Fume filtrate
(HzS04 -HNOs) HiSOt-HHOs and dilute.
Add NfeOH and filter,
ignite, and weigh
Reduce 1) (HC1)
Titrate UOjand A1&
11
Calculate Discard
:n Al filtrate
FIG. 1—Outline of a typical scheme of analysis.
Treatment of Extracts
Mother liquor separated from a field sample frequently will include solids
that led to formation of the deposit. Soluble constituents such as chloride,
sulfate, nitrate, and metal ions can be detected by standard analytical
procedures such as the ASTM methods Ref /.
Extracted organic matter sometimes can be identified as lubricant, fuel Oil,

rust preventive, or paint residue. However, heat and pressure associated with
deposit formation often will decompose the original organic matter and yield
decomposition products that are useless in fixing the original source of
organic contamination. Such tests as saponification number, iodine number,
viscosity, and additive identification sometimes are useful. Infrared spectro-
photometric examination may also give useful information about the
organic constituents.
Water-Soluble Matter
This test should be made immediately ahead of the ignition-loss test, and
on the same sample, to remove hydrates and certain other components that
might confuse the ignition-loss value. If appreciable water-soluble matter is
present, it should be examined for the usual water-soluble ions: chlorides,
bromides, fluorides, sulfides, sulfates, phosphates, nitrates, silicates, and
metal ions. If any are found in significant amounts, they should be
determined quantitatively. Normally the water-soluble content of a water-
side deposit will be low, but certain deposits, like calcium hydroxide, can
cause difficulty even thpugh they are reasonably soluble.
Ignition Loss
Ignition loss will include such items as organic matter, carbon, sulfur
(sulfite and sulfide), carbon dioxide from carbonates, and, on rare occasions,
nitrogen oxides or ammonia. Certain hydrates not removed by the water
solubility test also may be decomposed. The results of the ignition-loss test
are difficult to interpret, for while some constituents are being driven away,
others are being oxidized to higher weight values. For example, reduced
forms of iron oxide and copper can increase in weight on ignition.
The principal merit of ignition is in helping to confirm the presence of
constituents indicated by other tests. If a large amount of free copper is
found in preliminary inspection, a proportionate increase in weight on
ignition will be expected. On the other hand, if a large amount of organic
copper is observed in the extraction test, then a large loss of weight on
ignition would be expected.
Often, heating to a lower temperature than the usual 900 to 1000°C will
provide useful information. Carbonaceous samples, for example, should be
heated to a maximum temperature of 500° C. The report of analysis should
give the ignition temperature and length of ignition time.
Hydrogen Sulfide by Evolution

A qualitative test for hydrogen sulfide, using sulfuric acid and lead-acetate
test paper, should be made first. If this test is positive, a standard evolution
method should be followed [3].
312 MANUAL ON WATER
Carbon Dioxide by Evolution

A qualitative test for carbon dioxide should be made first by adding
mineral acid and passing any evolved gas into a solution of either calcium
hydroxide or barium hydroxide. If this test is positive (cloud produced), the
quantitaive determination can be made either gravimetrically [4], using the
standard Ascarite absorption tower, or volumetrically as given in the ASTM
Test for Total Carbon Dioxide and Calculation of the Carbonate and
Bicarbonate Ions in Industrial Water (Method D 513) [/], using modifi-
cations of a method originally developed for carbon dioxide in industrial
waters. The modifications permitting determinations to be made on solid
samples are only minor, and the method is more rapid than the gravimetric
method.
Sodium and Potassium

Sodium and potassium salts usually will be found with the water-soluble
portions (mother liquor or water extract) but in some cases they may occur
in complexes which remain behind in the silica residue and return to the
master solution with the fusion thereof. The quantitative determination can
best be made using a direct measuring-type flame photometer. Method C in
ASTM Methods D 1428 provides the detailed procedure.
Treatment of the Main Analytical Sample

Solution of water-formed deposits is generally best effected with a mixture
of mineral acids, including hydrochloric, nitric, and perchloric. Complete
solution of certain silicate complexes may be impossible by this means, but
these will be broken down by treatment of the silicate residue with
hydrofluoric acid and subsequent treatment with sodium carbonate. If
laboratory regulations prohibit use of perchloric acid, initial solution can be
brought about with nitric and hydrochloric acid or with nitric-sulfuric acid.
In the first instance, repeated dehydration may be required to separate the
silica. In the second case, analysis for certain constituents will be affected by
the sulfuric acid. For example, it will add sulfate ion and precipitate barium.
Separate samples must be used for such constituents when this method of
solution is used.
If a large amount of phosphate present in the original sample is subjected
to fuming with nitric, hydrochloric, and perchloric acids, considerable
phosphate will be separated with the silica, and some of it may be lost as
phosphoric acid if too much sulfuric or perchloric acid is used in the
hydrofluoric acid treatment of the silica residue. The silica residue may also
include barium sulfate and tin, as well as sodium and potassium from silica
complexes. These will be returned to the original sample, however, with the
sodium carbonate fusion.
The final filtrate from the silica determination is divided into several
aliquots as shown in Fig. 1. The size of these aliquots should be based on

preliminary exploratory tests so as to yield the best volume for each
determination. It will be noted that perchloric acid is recommended in place
of sulfuric acid in the hydrofluoric acid treatment. This eliminates intro-
duction of sulfate ion.
The analyses of several aliquots of the analytical scheme are discussed in
the following in the order of their occurrence in Fig. 1.
Sulfate
Sulfate is determined by precipitation as barium sulfate as given in ASTM
Test for Sulfate in Industrial Water and Industrial Waste Water (Method D
516) [/]. If the iron content is high, preliminary precipitation and filtration
to remove excess iron will reduce to a minimum the contamination of the
barium sulfate precipitate.
Calcium, Magnesium, and Barium

Calcium and magnesium are determined in series, after removing iron and
aluminum, by adding ammonium chloride and neutralizing with ammonium
hydroxide. Other metals need not be removed because the double precipita-
tion of calcium will reduce occlusions to insignificant proportions, while the
phosphates of such ions as copper, nickel, and ziric are largely soluble in
ammoniacal solution, though they are insoluble in neutral solution.
The calcium is precipitated as calcium oxalate. This can be ignited and
determined as calcium oxide or dissolved in sulfuric acid and titrated with
standard permanganate. Magnesium is precipitated as magnesium am-
monium phosphate and ignited to magnesium pyrophosphate. This residue
will also include manganese pyrophosphate as well as traces of heavy metals.
The presence of manganese accounts for treatment of the residue and
determination of magnesium by difference. It also is practicable to calculate
the manganese pyrophosphate from the manganese determination. If ex-
treme accuracy in the magnesium determinations is desired, or if the original
deposit contains very large amounts of copper, nickel, zinc, and lead, the
analyst may wish to remove these interferences by treating the solution with
hydrogen sulfide, filtering off the sulfide precipitate, and boiling the filtrate
to remove hydrogen sulfide before following the scheme shown in Fig. 1.
Alternatively, the heavy metals usually present in water-formed deposits
may be separated by the mercury cathode.
Manganese
Since the primary solution is a perchloric acid medium, manganese can be
determined colorimetrically by the standard potassium periodate method
described in various texts.
314 MANUAL ON WATER
Chromium and Vanadium

An experienced technician can do these two determinations simulta-
neously by the indicated scheme. The chromate and vandanate ions oxidize a
part of the ferrous ion and the excess ferrous ion is titrated with potassium
permanganate to disappearance of the red o-phenanthroline color. The vana-
dium is then oxidized by the permanganate. Titration is complete when the
pink color from one drop of permanganate persists for 30 s. The perman-
ganate consumed in the latter part of the titration is equivalent to the
vanadium. The chromium is determined by difference from the residual
ferrous ion calculated to have been used to reduce the vanadium.
If the vanadium content is relatively high, greater than 0.5 percent, the o-
phenanthroline end point is difficult to determine. In this case, it is more
practical to titrate the chromium and vanadium completely with perman-
ganate in the same manner, then to reduce the vanadium with ferrous
sulfate, oxidize the residual ferrous sulfate (but not the vanadium) with
ammonium persulfate, and finally titrate the vanadium with permanganate.
It also is possible to determine vanadium by electrometric titration with
ferrous sulfate.
Fortunately, vanadium is a rather unusual waterside deposit constituent,
so that precise analyses for it generally are not required.
Titanium
The standard method of oxidation with peroxide and determining tita-
nium by comparison with color standards is satisfactory for most boiler
deposit work. Vanadium produces a similar color and will yield essentially a
quantitative error. If the vanadium content is know, it can be subtracted.
Again, the absence of vanadium from most water-formed deposits makes
this error relatively unimportant.
Lead, Copper, Nickel, Zinc, Iron, Tin, Aluminum, and Phosphate

This large group of constituents is handled together because primary
removal of copper and lead, followed by removal of phosphate, eliminates
pertinent interferences from the schemes for the remaining constituents.
Plating copper and lead from perchloric acid solution generally results in a
dark copper plate containing traces of lead. This can be overcome by
dissolving the plated metals in nitric acid and redepositing them.
It is practical, but unnecessary, to separate copper and lead with hydrogen
sulfide and then to dissolve the precipitate in nitric acid and plate the metals
in the same manner. This method also separates tin as a sulfide but it will
remain dissolved in the acid electrolyte after plating. The hydrogen sulfide
must be driven off by heating before completing the scheme.
CHAPTER 14 ON CHEMICAL ANALYSIS OF DEPOSITS 3$$
Zinc may be determined gravimetrically as the oxide or colorimeft&ally

by the use of Zincon.2
The nickel dimethylglyoxime residue can be measured either gravi-
metrically or colorimetrically by standard procedures.
The phosphorus is removed early in the sequence to prevent interference.
It can be determined gravimetrically or volumetrically as ammonium
phosphomolybdate or gravimetrically as magnesium pyrophosphate. All are
standard analytical procedures.
The cupferron reagent used to separate iron and tin from the mixed
hydroxides decomposes slowly and should be freshly prepared for use.
Stability of cupferron is sometimes enhanced by adding 50 mg of aceto-
phenetidide per litre of 6 percent aqueous solution. If a small bag containing
ammonium carbonate is suspended in a bottle containing the dry cupferron
reagent, its quality can be maintained for a longer period.
Iron cupferron precipitate is only slowly attacked in the cold by 2 N
hydrochloric acid (used to wash the precipitate prior to washing with
ammonium hydroxide), but hot acid decomposes it. For this reason the
precipitation must be done in the cold. Nitric acid should be absent when
cupferron is used as a precipitant. For this reason the mixed hydroxides are
dissolved in hydrochloric acid.
It is also practical to determine iron and titanium as mixed oxides by
ashing the cupferron precipitate. If the precipitate is bulky, it should be wet-
oxidized with nitric and sulfuric acids and then reprecipitated with am-
monium hydroxide. Direct ignition is satisfactory for small precipitates. The
iron oxide can be corrected for the titanium oxide determined earlier. Iron
may also be determined colorimetrically by o-phenanthroline, with which
titanium does not interfere.
Precautions must be taken in determining aluminum as an oxide, since
aluminum oxide is a desiccant and will increase in weight if not kept in the*
proper atmosphere. Large amounts of aluminum may be determined by
sodium fluoride-sulfuric acid titration, using phenolphthalein or a pH meter
for detection of the end point.
Expressing Results
The results of analysis are generally reported in terms of basic oxides and
acid anhydrides, as shown in ASTM Reporting Results of Examination and
Analysis of Water-Formed Deposits (Method D 933) [/]. This procedure
does not provide for reporting definite molecular combinations. The ex-
pression of analytical results in terms of basic oxides should not be taken as
implying that any of the oxides are present in the sample as such. This
method of reporting results is only an arbitrary procedure of long usage.
2
"ZINCON" is a trade name of The La Motte Chemical Products Company for their 2-
carboxy-2'-hydroxy-5'-sulfoformazylbenzene.
316 MANUAL ON WATER
When definite combinations are desired, it is better to rely upon micro-

scopic or X-ray diffraction evidence than on hypothesis. ASTM Method D
933 gives definite rules for reporting results by these techniques. When such
apparatus is not available, however, probable compounds can be calculated
from the analytical data.
Conclusions
The preceding information on analysis of deposits is not intended to be
exhaustive. The discussion and tables of compounds likely to be found in
water-formed deposits (Table 1 and Chapter 13) should be used as a guide by
the analyst in determining his course of procedure. There may be occasions
when preliminary qualitative tests indicate the presence of elements not
included in the scheme discussed herein. In such cases, the analyst will
perforce have to use his ingenuity and modify the procedures to effect the
necessary separations and choose methods which will adequately determine
the elements so indicated. Many of the methods discussed in Chapter 11 for
the analysis of industrail waters may be applicable to the determination of
certain constituents of water-formed deposits, and the analyst is advised to
study that chapter carefully.
TABLE 1—List of compounds reported to be found in water-formed deposits.

Acmite Na2OFe2O3-4SiO2
Ammonium bicarbonate NH4HCO3
Analcite Na2O-Al2O3-4SiO2-2H2O
Anhydrite CaSO4
Ankerite CaCO3(Fe, Mg, Mn)CO3
Aragonite CaCOs
Atacamite CuCl2-3Cu(OH)2
Azurite 2 CuCO3-Cu(OH)2
Barite BaSO4
Bayerite A12O3-3H2O
Bloedite Na2SO4-MgSO4-4H2O
Boehmite A12O3-H2O
Bornite Cu5FeS4
Brochantite CuSO4-3Cu(OH)2
Brucite Mg(OH)4
Bunsenite NiO
Burkeite Na2CO3-2Na2SO4
Calcium oxalate monohydrate CaC2O4-H2O
Calcium phosphate (dibasic) CaHPO4
Calcium pyrophosphate Ca2P2O7
Calcium sodium phosphate CaNaPO4
Calcium sulfate CaSO4
Calcite CaCO3
Calcium aluminate 3CaO-Al2O3-6H2O
Cancrinite 4Na2O-CaO-4Al2O3-2COr9SiO2-3H2O
Celestite SrSo4
Cementite Fe3C
Cemicite PbCO3
a Chalcocite aCu 2 S
TABLE I—Continued.
/3 Chalcocite )8Cu 2 S
Chalcopyrite CuFeS2
Copper Cu
Copper aluminate CuAlO2
Corundum a Al2O,
Covellite CuS
Cristoballite SiO2
Cuprite Cu2O
Delafossite Cu2O-Fe2O3
Dolomite CaCOrMgCO,
Fayalite 2FeO-Si02
Ferrous bicarbonate Fe(HC03)2
Ferrous sulfate monohydrate FeSO4-H2O
Ferrous sulfate quadrahydrate FeSO4-4H2O
5 Ferric Oxide 6 Fe2O3
/? Ferric oxide monohydrate 0 Fe2OvH20
Fluorite CaF2
Foshagite 5CaO-3SiO2-3H2O
Forsterite 2MgO-SiO2
Gaylussite CaOVNa2COr5H2O
Gehlenite 3CaO-Al2Or2SiO2
Gibbsite 7 A1203-3H2O
Glauconite K2(MgFe)2AU(Si40,0)3(OH)12
Goethite a Fe2O3-H2O
Gypsum CaSO4-2H2O
Gyrolite 2CaO-3SiO2-H2O
Halite NaCl
Hematite Fe203
Hemihydrate CaSO4-^H 2 0
Hemimorphite Zru(OH)2SiO7-H2O
Hydromagnesite 4MgO-3CO2-4H2O
Hydrotalcite MgCo3-5Mg(OH)2-2Al(OH)3-4H2O
Hydroxyapatite Caio(OH)2(PO4)6
Hydroczincite 2ZnC03-3Zn(OH)2
Iron Fe '
Lazurite 3Na2O-3Al2O3-6SiO2-2Na2S
Lead Pb
Lepidocrocite y Fe2O3-H2O
Libethenite 4CuO-P2O5-H2O
Maghemite 7 Fe2O3
Magnesia MgO
Magnesium chloride hydrate (basic) MgCl2-5Mg(OH)2-8H2O
Magnesium hydroxyphosphate Mg3(P04)2-Mg(OH)2
Magnesite MgC03
Magnetite Fe2O3
Malachite CuC03-Cu(OH)2
Meta halloysite Al2O3-2SiO2-xH2O
Meta thenardite Na2SO4I
Montmorillonite Al2O3-4SiO2-H2O-nH20
Mullite 3Al203-2SiO2
Muscovite KAl2(Si3Al)0,o(OH2F)2
Natrolite Na2O-Al2O3-3SiO2-2H2O
Nontronite H4(A1, Fe)2Si2O9
Noselite 4Na20-3Al2O3-6SiO2-S03
Oldhamite CaS
Olivine 2(Mg, Fe)O-SiO2
Para sepiolite 2MgO-3SiO2-2H20
(Continued)
318 MANUAL ON WATER
TABLE 1— Continued.
Paratacamite 7 CuCl: 3Cu(OH):
Pectolite Na:O-4CaO-6SiOrH:O
Periclase MgO
Portlandite Ca(OH):
Pyrrhotite FeS
Quartz SiO:
Sepiolite 2MgO3Si02-2H2O
Serpentine 3MgO2SiO2-2H2O
Siderite FeCO,
Smithsonite ZnCO,
Sodium disilicate p Na2Si2O5
Sodium metasilicate Na2SiO.
Sphalerite 0ZnS
Syngenite K2SO4-CaSO4-H2O
Talc 3MgO-4Si02-H2O
Tenorite CuO
Teschemacherite NH4HCO,
Thenardite Na2SO4(V)
Thermonatrite Na2CO,H2O
Troilite FeS
Trona 3Na20-4C02-5H20
Vermiculite (Mg, Fc)3(Al, Si)4O,o(OH)2-4H2O
Vivianite Fe3(PO4)2-8H2O
Wedellite CaC2O4-2H20
Wilkeite Ca,00[(Si, P, S)04]6
Willemite Zn2SiC>4
Witherite BaCd
Whitlockite 0 Ca3P2O8
Wollastonite /3 CaSiOi
Wustite FeO
Xonotlite 5CaO-5SiO2-H20
Zinc A ZnOxH 2 O
B ZnO-yH2O
E ZnO-zH2O
Zincite ZnO
Zincosite ZnSCX
References
[/] Annual Book of ASTM Standards, Part 23.
[2] Feigl, Fritz, Spot Tests in Inorganic^Ânalysis, 5th ed., Elsevier, New York, 1958.
[3] Furman, N. H., Ed., Standard Methods of Chemical Analysis, 6th ed., Van Nostrand,
New York, Vol. 1, 1962, p. 1031.
[4] Furman, N. H., Ed., Standard Methods of Chemical Analysis, 6th ed., Van Nostranu,
New York, Vol. 1, 1962, p. 282.
Chapter 15—Sediment in Streams and

Other Water Bodies
All streams and other water bodies contain sediment in variable quantity
and character. Methods for measurement of these sediments are affected by
the sediment mode (suspended or deposited), by the characteristics of the
water body, and by the spacial and temporal variations of the sediments.
The purpose of this chapter is to indicate some of the needs for sediment
measurements, describe some of the more common sediment environments
in which it may be necessary to make measurements, and provide general
references to the commonly used methods for making these measurements.
Many activities in the conservation, development, and utilization of land,
mineral, and water resources affect sediment movement and thereby may
create sediment-related problems. For example, a change in the runoff
regime from a drainage basin may concentrate or disperse sediment in the
stream channel and,, in turn, affect the flow capacity of the stream. Because
of the complex interrelationships that affect sedimentation phenomena, a
knowledge of climate, the physical attributes of drainage basins, the
hydraulic and hydrologic characteristics of streamflow, and the quantitative
and qualitative aspects of sediment is required to solve sediment problems.
One might think that the solution to most sediment problems would be to
stop landscape and channel erosion. This is not only impossible, but would
severely upset the natural balance so that a host of new problems would
emerge. Experience has shown that erosion should be carefully controlled to
fit the situation. Intensive sediment movement (perhaps 100 t/ha) from a
small construction site may be tolerated for a few weeks, especially if the
sediment is deposited before it reaches a stream; but a relatively small
amount of top soil (perhaps 1 t/ha/year) from a cropland with adsorbed
herbicides may be intolerable if it reaches a stream.
Experience gained over the years has led to a greater understanding and
appreciation of sediment problems; however, recent lawsuits attest to the
fact that man's activity has caused, and will continue to cause, detrimental
environmental changes resulting from the transport of fluvial sediment. Life
scientists have become increasingly aware of the effect of sediment on the
flora and fauna of streams and other water bodies. As an example, bottom
sediments contaminated by industrial or municipal wastes may be dispersed
319
320 MANUAL ON WATER
by dredging operations, and thereby made available to enter the food chain
through fish and other aquatic life. Numerous references can be found
showing that bottom sediments in rivers are capable of adsorbing pesticides,
trace metals, and other toxic chemical constituents.
Glymph [7]1 indicates that we have recently begun to appreciate the water-
quality implications of sediment as a pollutant in the same context as
industrial wastes, effluents from sewage treatment plants, and other kinds of
pollutants. Rutherford [2] has asked the pertinent question: How much
longer can the strained soil colloids near commercial, industrial, and other
waste-producing areas continue to extract and store pollutants through ionic
exchange before they become totally clogged, and how fast are these soils
becoming unwanted sediments in nearby water bodies? These aspects of
sedimentation require new levels of sophistication for determining sediment
yields and sediment properties in small basins to form the basis for cost-
effective sediment and pollution management programs.
The scope of water-related sediment problems varies widely in intensity
and in time and space. The most prevalent problems, however, range from
the hour-by-hour impacts on small streams of sediments that are eroded
from cropland or construction sites during rainstorms to continuous impacts
on large streams of sedimentation phenomena that affect aesthetics, stream-
channel andpreservoir capacity, and the disposition of sorbed pollutants.
Although visual inspections may be helpful, a solution to a sediment
problem usually requires scientific measurements that serve to define the
quantitative and qualitative character of sediments suspended in, trans-
ported by, and deposited from natural and man-made streams and water
bodies.
Modeling experts have made some progress, though sometimes with
conflicting results, in outlining predictive equations in sedimentation. How-
ever, it seems that there is only a remote possibility for development of a
complete set of such equations. Therefore, numerical simulation techniques
using current and pertinent basic data offer the most practical basis for
predicting specific fluvial sediment phenomena.
Because of the impact of an ever-changing environment on sediment
movement, it is expected that there will be an increasing need for sediment
measurements to obtain information on its quantity and characteristics
which can be correlated with these environmental changes. Sediment data
are also needed to help verify the effectiveness of sediment-control measures.
Most sediment measurements are made to satisfy one of three general
program objectives:
1. To obtain general information from a network of sediment-measuring
stations which provide unbiased comprehensive information about sediment
concentration and movement.
2. To obtain special information relative to specific problems so water
managers can better choose among alternative actions.
CHAPTER 15 ON SEDIMENT IN STREAMS AND OTHER WATER BODIES 321
3. To disclose and describe process relationships between water, sedi-

ment and environment.
It is appropriate to point out that when sediment data have potential for
multiple use, including extrapolation in time or space, they should be
cataloged and available to potential users. Therefore, this discussion is
intended, first, to provide a better understanding of the essential requisites for
valid sampling and data-collection programs in which sediment is one of the
constituents of interest, and second, to foster the use of suitable standardized
equipment for the purposes of improving the accuracy of sediment data and
providing data truly comparable from one source to another.
Characteristics of Sediment Environments

Land Surface
The sediment environment at the land surface includes all areas not in
stream channels or inundated by water and, thus, represents immensely
diverse conditions. The immediate source area for much of the sediment in
stream channels and other water bodies is from the land surface. Sedimenta-
tion, or, more specifically, the transport and deposition of sediment, is a
staged process. Particles detached and entrained in a given event may move
downslope on the land surface only a short distance, or may become part of
the stream load.
A well-developed soil mantle indicates that the long-term erosion phase of
the sedimentation process has been slower than the soil-forming processes.
This condition requires the presence of good vegetative cover on the land
surface. Generally, well-developed soil consists of fine panicles mixed with
relatively high proportions of organic material.
Disturbance of areas with well-ceveloped soils, such as row-crop cultiva-
tion or construction, usually causes a complex change in the sediment
environment. Disturbances of the surface environment which do not result
in increased runoff, or in significant concentration of flow, may only
minutely increase sediment yield. To the extent that the disturbance mixes
the soils, the character of the water-sediment mixture in runoff will be more
varied in particle size and chemical composition than would be expected for
undisturbed soils.
Disturbances which increase runoff or channelize water generally cause
the sediment yield to increase dramatically. The' range of particle sizes in
transport broadens, with an attendant change in the chemical composition
from the soil being eroded. New channels or gulliers, or both, form which
may permanently alter the rate and character of sediment movement from
the particular landscape.
On any catchment, the sediment-contributing area is usually less than 100
percent of the catchment. The sediment sources and the rate of yield change
with the amount of disturbance, and both tend to increase with an increase
in the intensity of storm events.
322 MANUAL ON WATER
Regardless of conditions on an existing land surface, erosion and entrain-

ment occur mainly in response to rainfall events. Rainfall events naturally
vary in magnitude, the larger events occurring less frequently. There may be
no sediment-producing rainfall events in any given year or there may be tens
or hundreds of such events. Typically, one would expect anywhere from a
few events to 30 or 40 annually. These discrete rainfall events, which may last
for only a few days or even a few hours in a year, collectively make up the
annual sediment yield from a land surface.
Streams
Although the land-surface environments are the basic sources of supply
for stream sediment, the stream channel itself may be a sediment source. As
noted, the processes of erosion and deposition on the landscape cause a
highly variable quantity of sediment to be available for movement by
streams. In discussing sediment available to streams, it is useful to classify
sediment as either fine or coarse.
Fine Sediment—Fine sediment (< 0.062 mm) is easily suspended by
natural stream turbulence, and hence travels through a stream system with
about the same velocity as the water. The quantity of fine sediment moving
at a given stream cross section depends on the amount of erosion in the basin
and the routing of the particles by the flow to this section. The quantity
moving is only indirectly dependent on the flow rate of the stream. The exact
concentration of fine sediment in transport during stormflow is unpre-
dictable with respect to time, and frequently increases by a factor of 100 to
1000 times that of "normal flow." Thus, sediment measurement techniques
require a program to obtain a sufficient number of observations to define the
large temporal variations in concentrations.
Coarse Sediment—Coarse sediment (> 0.062 mm) usually is found in
appreciable quantities in the bed of a stream and may be transported both as
suspended load and as bedload. The mode and rates of transport depend on
the properties of the fluid, the sediment, the hydraulics of flow, and the
characteristics of the channel geometry. Thus; it follows that coarse-
sediment discharge often can be roughly related to water discharge at a given
stream transect (cross section).
The movement of coarse particles consists of a series of steps interrupted
by periods of no motion, when the particle is a part of the bed material. A
particle moves and becomes a part of either the suspended load or the
bedload whenever lift and drag forces or impact from another moving
particle, or both, overcome the resisting forces and dislodge it from its
resting place. The magnitudes of the forces vary according to the fluid
properties, the velocity and turbulence of the flow, the physical character of
the particle, and the degree of exposure of the particle. The degree of
exposure depends largely on the size and shape of the particle relative to
other particles in the bed mixture, and to the position of the particle relative
to relief features on the bed such as permanent obstructions or bedforms, or

both. Because of these factors, even in steady flow, the discharge of coarse
sediment can be expected to fluctuate significantly with time at any point,
and even vary substantially across the stream at any instant.
In order to define the coarse-sediment discharge at a transect, measure-
ments must be made at a number of different lateral locations. Also, because
of the close relationship between coarse-sediment discharge and the flow
forces, at any given time, particles that move as bedload at one transect may
or may not move as suspended load at another transect. As a result, the
bedload portion of the total-sediment discharge observed at one transect is
not necessarily representative of that at some other transect.
Suspended-Sediment Discharge—All particles in suspension, both fine
and coarse, that have a density greater than that of water tend to settle. As a
result, because the upward and downward turbulent currents are necessarily
equal, there is a tendency for the flux of sediment through any horizontal
plane to be greater downward than upward. This tendency is naturally
counteracted by the establishment of a vertical concentration gradient of
such a steepness that the concentration of upward moving parcels of water-
sediment mixture are just sufficiently greater than the concentration of
downward moving parcels to produce a net upward flux that just balances
the flux due to settling. Because of their higher fall velocities, the coarser the
particle, the steeper the gradient. Usually, the concentration of particles
smaller than 0.062 mm will be uniform throughout the entire depth.
Due to the spatial variations in the coarse and fine suspended-sediment
concentrations and in flow velocities, the discharge of suspended sediment at
any given instant must be determined from
Q
*-L CUdA (1)
where
Qs = "instantaneous" suspended-sediment discharge through a section
of area A,
U = velocity of sediment particles through an elemental area dA, and
C - suspended-sediment concentration in the elemental spatial volume
Ut'dA, in which t' is one unit of time used to express U.
In the practical application of Eq 1, U is considered to be equal to the flow
velocity and C is considered to be constant during any given sampling period
associated with its determination. Integrating the Eq 1 over time yields the
quantity of suspended-sediment discharge during any specified duration, T.
That is
s=$TQsdt=ST$cudAdt
A (2)
where S is the quantity of suspended sediment discharged through area A

during the elapsed time T.
324 MANUAL ON WATER
Lakes, Reservoirs, and Ponds

Lakes, reservoirs, and ponds are relatively still bodies of water which serve
as sinks for sediments. Those with outlets usually trap much but not all of
the sediment carried to them. Factors affecting the sediment environment
include the trap efficiency, the presence of suspended material, and the
character of the bottom sediments. The major variables that determine the
character of the sediment in this type of environment are the source and
character of the inflow, the size and shape of the water body, the volume of
the water body relative to the rate of outflow, the age of the water body, and
the time of year.
Lakes or reservoirs with one or more major inlet streams develop deltas
composed of coarse materials deposited as foreset beds. As a delta develops,
the extension of the river channel tends to form topset beds on the delta.
Fine materials are mostly deposited far into the impoundment as bottomset
beds. Wide fluctuations in water levels, such as those occurring in flood-
control and irrigation reservoirs, significantly affect the deposition patterns
in many reservoirs. Most natural lake bottoms show the results of hundreds
or thousands of years of the sedimentation processes. Lake currents and the
character and timing of the sediment-laden inflows will be reflected in the
panicle size and deposition patterns of bottom sediments.
To provide reliable information, sampling programs must have well-
defined goals and must take into account the factors controlling the
distribution of sediment in the lake, reservoir, or pond.
Estuaries
Significant differences in the sediment environment are found in different
„ types of estuaries. In estuaries where the tidal volume of sea water flowing
into and out of the estuary is very large relative to the volume of river flow,
sufficient mixing usually occurs so that the vertical distribution of salinity is
uniform with depth and only a longitudinal salinity gradient exists. Under
these circumstances, the flow is not vastly different from rapidly varying
unsteady flow in rivers, and sedimentation processes are roughly similar to
those that occur under like flows in a river.
In the so-called partially mixed estuary, which is the most prevalent type
of major estuary along the Atlantic and Pacific coasts of North America, the
volume of freshwater inflow is sufficiently large relative to the volume of
seawater flow that a complex pattern of salinity (and therefore density)
gradients develops. This produces a net vertical circulation in which seaward
from a point on the channel bottom—termed the null point or the point of
zero net flow—there is a predominance of landward flow of relatively dense
saline water along the bottom, and a predominance of seaward flow of less
dense relatively freshwater near the surface, even though the body of the
flow alternates direction periodically in accordance with the tide. This
generally produces a characteristic feature termed a "turbidity maximum."

The turbidity maximum manifests itself as a bell-shaped longitudinal
suspended-sediment concentration distribution that progressively develops
in magnitude and then abates and translates landward and seaward with
each flood and ebb.
In addition to suspended-sediment distributions engendered by vertical
circulation patterns, larger estuaries typically develop horizontal circulation
patterns, due to the Coriolis effect, major ocean circulation patterns, and
sometimes prevailing wind patterns. These horizontal circulation patterns
frequently cause a predominence of flow in certain channels during the flood
period and a predominence of flow in other channels during the ebb. These
patterns may produce extreme lateral variations in suspended-sediment
concentration and disposition of bottom sediments.
As a result of the different hydraulic and sediment characteristics in
estuaries, sampling techniques and requirements vary from one estuary to
another. The common characteristic of all estuaries is that at any given time
the concentration, composition, and transport direction of the suspended
sediment is different in some degree from every other point in the estuary,
and that at every point in the estuary these quantities vary continuously
throughout time. Consequently, a measurement at a single point or cross
section cannot be used to represent either the transport of suspended
sediment in the entire estuary nor the transport through any other point or
cross section in the estuary. These conditions dictate the necessity to define
precise sampling goals and, likely, to limit the usability of a given set of
samples to the specific purpose for which it was collected.
Conduits and Outfalls

The physical and chemical composition of the sediments entering a stream
from conduits or outfalls could include a wide variety of fluvial sediments,
and, in the case of many industrial operations or municipal water treatment
plants, metals, other inorganic material, or organic material. In contrast to
natural streams, the distinguishing characteristics of flow in conduits and
outfalls is that it is the waste or effluent of industrial or municipal
processing, or the outflow of an artificial watershed such as a storm sewer
system. In either case the conditions that determine the flow, the type of
sediments, and their concentrations are controlled mostly by man's activi-
ties. Sampling programs and sampling methods will be effective only if the
components of sediment inputs and hydraulic conditions are understood and
considered.
Methods of Measurement
A wide number of samplers and procedures have been developed and are
commonly used for sampling sediment in streams and other water bodies. In
326 MANUAL ON WATER
general, the requisites for sampling sediment that is in motion are sub-
stantially different from those for sampling stationary deposited sediment.
Prior to documentation of specific instructions by ASTM, specifications of
samplers and sampling procedures are given in Chapter 3 of a National
Handbook of Recommended Methods for Water-Data Acquisition pre-
pared by a working group of representatives from Federal agencies in-
terested in aspects of sediment measurement. The chapter also includes
instructions for sample handling and preservation, laboratory analysis,
methods for making reservoir surveys, methods for documenting stream
channel changes, and quality control for the sediment measurement pro-
gram.
References
[/] Rutherford, G. K., "Anthropogenic Sources of Heavy Metals for Sediment Transport,"
Proceedings, Conference on Fluvial Transport of Sediment-Associated Nutrients and
Contaminants, Kitchener, Ontario, 1976 (in process).
[2] Glymph, L. M., 1975, "Evolving Emphasis in Sediment-Yield Predictions," Present and
Prospective Technology for Predicting Sediment Yields and Sources, U. S. Department
of Agriculture, ARS-S-40, Washington, D. C, pp. 1-4.
Chapter 16—Radioactive Nuclides

in Water
Before the summer of 1945, it could be fairly stated that the problem of
radioactive contamination of water did not exist. True, some radioactive
springs were known, but these aroused little interest, except for possible use
of the supposed healing properties of the water. The successful testing of an
atomic bomb in the New Mexico desert on July 16, 1945, changed this
situation. Fogging of some packages of photographic film was traced to
radioactive debris from this bomb. The radionuclides had got into the
process water and then into the wrapping paper. This is the first known
damage to a commercial product from radioactively contaminated water.
Radioactivity may be defined as the spontaneous emission of penetrating
energetic radiation from certain atomic nuclei. The early workers in the field
soon classified the radiation into three types, called alpha (a), beta (0), and
gamma (7) rays, on the basis of their ability to penetrate matter. They were
also able to establish their nature, finding that alpha and beta rays were
high-speed particles, while gamma rays were electromagnetic waves similar
to X-rays. The results of these early studies can be tabulated as shown in
Table 1.
The emission of these radiations is associated with a nuclear disinte^
gration, which is the rearrangement of the constituents of the nucleus into a
more stable configuration with an associated release of energy. The rate of
emission is characteristic of the radioactive material (also called radio-
nuclide), and is described by the time required for half of the unstable nuclei
TABLE 1 — Types of radiation: nature and penetration ability.

Penetration
Type Air Solids Nature
Alpha few centimetres few micrometres fast-moving helium nuclei
Beta metres several millimetres very rapid electrons
(+ and -)"
Gamma many metres several centimetres electromagnetic radiations
of very high energy
"All the beta radiation from naturally occurring materials consists of negative electrons; the
emission of positive electrons (positrons) by certain man-made materials was discovered in the
1930s.
327
328 MANUAL ON WATER
to complete their rearrangements and emissions: the half-life. After the lapse
of one half-life, since the number of unstable nuclei has been reduced to one-
half the original number, the emission rate (or activity) is also reduced to one
half of its initial value. The quantity of radioactive material is measured by
the rate of disintegration; that amount of a nuclide in which 3.7 X.1010 atoms
disintegrate per second is defined as a curie. Ordinarily, subdivisions of this
unit, milli-, micro-, or even micromicrocuries, are used. These are 10-*, 10~6,
and 10~12 curies respectively. The term picocurie, is currently used as a
synonym for micromicrocurie.
The energy released in a nuclear disintegration is large, usually of the
order of a million times as much as is involved in ordinary chemical changes.
When alpha particles or gamma rays are emitted, these carry off sub-
stantially all the energy release. In beta emission, however, the available
energy is divided between the beta particle and a neutrino, which is not
ordinarily observable. The division of the energy varies in individual
disintegrations, so that a particular beta particle may have an energy from
zero up to the maximum available, the average being about one third of the
total energy released. Consequently, beta radiation has a continuous energy
spectrum, while alpha and gamma radiations are characterized by line
spectra.
The unit of energy most used in radioactivity is the electron volt, the
energy gained by an electron in passing through a potential difference of 1 V.
It is equal to 1.6 x TO"12 ergs. Alpha particles have energies of a few million'
electron volts (MeV); beta and gamma radiations may have energies from
zero up to a few MeV.
The three types of radiations interact with matter in somewhat different
ways. Both of the charged particles (alpha and beta) react primarily by
electrostatic interaction with the electrons of the material, transferring
energy to each electron that they approach sufficiently closely. Quantita-
tively, the actions differ considerably, with the heavier, slower-moving alpha
particle interacting with many more electrons per unit path length. The
interaction of gamma radiation with matter is essentially statistical in nature,
with each individual gamma photon transferring no energy until it interacts,
usually with an electron, whereupon it transfers all or a large part of its
energy. The gamma radiation has no definite range, but is statistically
reduced in quantity as it passes through matter. This gamma interaction has
been compared to the process of firing machine gun bullets at random into a
forest, in which each bullet is unaffected until it strikes a tree, when it loses
all or part (if it ricochets) of its energy. The extent of the interaction of the
three types of radiation with matter is illustrated in Table 2.
The primary effect of the interaction of any one of the three types of
radiation with matter is the transfer of energy to electrons, resulting in the
production of ions or of excited molecules. Secondary changes in matter
include such phenomena as the darkening of photographic film, the
initiation of chemical changes, and biological effects. The extent of such
CHAPTER 16 ON RADIOACTIVE NUCLIDES IN WATER 329
TABLE 2 — Interaction of radiation with matter.

Energy Transfer,
Range, cm MeV /cm Tenth-Thickness,acm
Radiation; Energy, MeV Air Aluminum Air Aluminum Lead Aluminum

Alpha, 5 3.5 ~ 0.001 8 1.43 2800
Alpha, 7 5.9 0.0033 1.19 2100
Beta, 0.5 125 0.06 0.0040 8.3

Beta, 3.0 1320 0.64 0.0023 4.7
Gamma, 0.1 0.037 5.1

Gamma, 1.0 2.9 13.8
Thickness needed to reduce intensity to 10 percent of incident intensity.
effects is dependent on the dose, or amount of energy absorbed by the

material. The unit of energy absorption is the rad, which is equivalent to the
absorption of 100 ergs/g of material.
Biological systems in particular are affected by radiation. Destruction of
cells by large doses is the basis for radiation therapy for cancer. The effects
of excessive exposure in man vary from skin burns through nausea, anemia,
loss of resistance to infection, and even death. Smaller radiation doses may
cause more subtle damage, leading to cancer and to genetic mutations. While
information about the potential damage from low levels of radiation is
unfortunately inexact, "acceptable risk" levels are established by the Inter-
national Commission on Radiological Protection (ICRP) and the National
Committee on Radiation Protection and Measurements (NCRP).
Because of their ionizing, photographic and, especially, biological effects,
the quantities of radioactive materials in commercial products, in air, in
water, and in foods, must be rigidly controlled. For each radionuclide, a
concentration may be determined, based on its physical, chemical, and
metabolic characteristics, such that no organ of the body will receive more
than the "acceptable risk" exposure given above. The ICRP and NCRP also
establish these maximum permissible concentrations.
Naturally Occurring Radionuclides
Heavy Element Series

Although radioactive contamination has been a problem for only a short
time, radioactive materials have always been part of the environment. The
major sources of natural radioactivity are uranium and thorium ores. These
contain a great variety of radionuclides, differing in chemical nature, half-
life, and type and energy of radiations, but linked genetically into three
series, each stemming from an element of very long half-life. The three series
are listed in Table 3.
Some of these elements can be dissolved, in ratios depending on their
330 MANUAL ON WATER
TABLE 3—Natural radioactive series.

Atomic
Num- Atomic Radiation,
ber Element Weight Historical Name Half-Life Remarks
URANIUM- RADIUM SERIES
92 Uranium 238 uranium I a 4.5 X lO9 yr
90 Thorium 234 uranium X-l ft 24 days
91 Protactinium 234m uranium X-2 ft 1.2 min
91 Protactinium" 234 uranium Z ft 6.7 h
92 Uranium 234 uranium II a 3 X 105 yr
90 Thorium 230 ionium a 8 X 104 yr
88 Radium 226 radium a 1600 yr
86 Radon 222 radon a 3.8 days
84 Polonium 218 radium A <*(ft) 3 min 0 0.03 percent
82 Lead 214 radium B ft 27 min
85 Astatine" 218 a 2s
83 Bismuth 214 radium C ft(a) 20 min a 0.04 percent
84 Polonium 214 radium C' a <IO"6 s
81 Thallium" 210 radium C" ft 15 min
82 Lead 210 radium D ft ~20yr
83 Bismuth 210 radium E ft(a) 5 days a ~ 10"5 percent
84 Polonium 210 radium F a 138 days
81 Thallium" 206 ft 4 min
82 Lead 206 radium G stable
THORIUM SERIES
90 Thorium 232 thorium a 1.4X 10'°yr
88 Radium 228 mesothorium 1 ft 5.7 yr
89 Actinium 228 mesothorium 2 ft 6.1 h
90 Thorium 228 radiothorium a 1.9 yr
88 Radium 224 thorium X a 3.6 days
86 Radon 220 thoron a 1 min
84 Polonium 216 thorium A <*(ft) 0.16s ft 0.01 percent
82 Lead 212 thorium B ft 10.6 h
4
85 Astatine" 216 a 3 X 10~ s
83 Bismuth 212 thorium C ft(<*) 60 min a 33.7 percent
84 Polonium 212 thorium C' a 3 X 10"7 s
81 Thallium'' 208 thorium C" ft 3 min
82 Lead 208 thorium D stable
URANIUM-ACTINIUM SERIES
92 Uranium 235 actino-uranium a 7 X 108 yr
90 Thorium 231 uranium Y ft 26 h
91 Protactinium 231 protactinium a 3 X 104 yr
89 Actinium 227 actinium ft(<*) 22 yr a 1.2 percent
90 Thorium 227 radioactinium a 19 days
87 Francium' 223 actinium K ft 21 min
88 Radium 223 actinium X a 12 days
86 Radon 219 actinon a 4s
84 Polonium 215 actinium A a(ft) 2 X 10~3 s ft <10"3 percent
82 Lead 211 actinium B ft 36 min
85 Astatine" 215 a 10~4 s
83 Bismuth 211 actinium C <*(ft) 2 min 0 0.32 percent
81 Thallium 207 actinium C" ft 5 min
84 Polonium" 211 actinium C' a 25 s
82 Lead 207 actinium D stable
"Minor component, <1 percent.
fc
Minor component, 33.7 percent.
'Minor component, 1.2 percent.
individual solubilities, whenever water runs through formations containing

them. In addition, since element 86 (radon) is an inert gas, it can move
through the soil and so reach water not in contact with the radioactive
formation. Because of its relatively long half-life, radon-222 can move
through the ground for much greater distances than radon-220 or radon-219.
In radioactive springs, radon-222 and its immediate daughters, radium A
through radium C', are usually the most conspicuous components. The
radium-226 content of ordinary surface- or groundwaters will generally be
less than 0.5 pCi/litre, while the radon-222 content of the same water may be
as high as 100 pCi/litre. In some spring waters, radon contents may be in the
range of microcuries per litre, with radium levels as high as thousands of
picocuries per litre.
Each of the foregoing radioactive series owes its occurrence in nature to
the existance of a long-lived parent, uranium-238, thorium-232, and
uranium-235, respectively. All the other elements in the series have short
half-lives, geologically speaking, and would no longer be present except for
continual formation.
Other Long-Lived Materials

Radioactive series do not occur in other parts of the periodic system, and
the only radionuclides existing are those few species of geologically long
half-life. The most important of these is potassium-40, which comprises
0.012 percent of natural potassium, and thus contributes radioactivity to all
potassium-containing systems. For each milligram of potassium, there will
be almost two radioactive disintegrations per minute (dpm). Water itself will
therefore be radioactive in proportion to its potassium content, plus a
contribution from biological material. The second most important natural
activity of this type in water is rubidium-87, which forms almost 28 percent
of all natural rubidium. Because of the low abundance of rubidium in the
earth's crust, as compared to potassium, rubidium activity is relatively
unimportant in freshwaters. Both potassium and rubidium contribute
significantly to the radioactivity of seawater, as shown in Table 4. Additional
minor radioactive elements of long half-life are included in Table 4.
Cosmic-Ray-Produced Nuclides
A few radioactive materials of short half-life exist in nature from
continuous production by cosmic-ray bombardment of the atmosphere.
These nuclides are also listed in Table 4. The most important of these are
carbon-14 and hydrogen 3 (tritium). Carbon-14 is formed by the action of
cosmic-ray neutrons on atmospheric nitrogen, while tritium is formed
directly in the splitting of atoms struck by the high-energy cosmic ray
particles. Both of these are oxidized, to carbon dioxide and water, and enter
the biosphere and hydrosphere. None of the other radioactive nuclides
332 MANUAL ON WATER
TABLE 4—Other naturally occurring radionuclides.

Specific Activity,
Half-life, Abundance, dpm/g" of Total in Sea,
Nuclide years % element MCift
Hydrogen-3 12.3 varies varies 200
Beryllium-7 53 days varies varies
Carbon- 14 5730 varies Oto 15 270
Potassium-40 ~109 0.012 1800 460000
Vanadium-SO 4 X 1014 0.25 0.1

Rubidium-87 5 X 10'° 27.85 50000 8400
Indium-US 6 X 10U 95.7 11
Lanthanum- 138 1 X 10" 0.09 52
Cerium- 142 5 X 10" 11.07 0.12

Neodymium-144 5 X 1015 23.87 0.26
Samarium- 147 1 X 10" 15.1 8200
Lutetium-176 ~10'° 2.6 ~12000
Rhenium- 187 ~10'° 62.9 ~270000

Platinum- 190 6 X 10" 0.012 0.8
Platinum- 192 ~10" 0.78 ~0.03
Bismuth-209 over 1018 100 <0.004
°dpm = disintegrations per minute.
*One megacurie = 106 Ci.
formed by cosmic-ray bombardment exists in large enough quantities to be

of any real importance.
Radionuclides in Reactor Coolants and Moderators

This subject is discussed in Chapter 18.
Radionuclides from Chemical Processing of Ores, Reactor Fuels, and

Blanket Materials
Ores
The primary natural materials useful in nuclear energy are uranium and
thorium. Uranium-235 is the only natural fissionable material; uranium-238
and thorium-232 are the raw materials for manufacture of artificial fis-
sionable materials. The first steps in nuclear technology are thus the mining
and refining of these materials. Uranium and thorium are each the pro-
genitor of long series of radioactive elements, any or all of which may be
present in the ores.
Of all the daughter elements, those of relatively long half-life include
thorium-230 (ionium), radium-226, and lead-210 (radium D) in the uranium-
radium series, and radium-228 (mesothorium 1) and thorium-228 (radio-
thorium) in the thorium series. The long-lived protactinium-231 and
actinium-227 occur in uranium, but since they are derived from relatively
scarce uranium-235 (0.71 percent), they are less important. In removing the
desired uranium or thorium from ore, the other radioactive elements present
will be left behind. Depending upon the process, these elements may occur in
various chemical forms with different solubilities. Those of long half-life will
persist much longer than any conceivable control measures and their
treatment must be absolutely foolproof. The more soluble materials, of
course, may eventually reach and contaminate waters. Of all the nuclides
listed in the preceding paragraph, radium-226 is the most harmful in a water
environment by reason of its biochemical properties, its long half-life, and
the energetic radiations of it and its immediate daughter products. The
present maximum permissible concentration of this material in uncontrolled
waters is only 4 X 10~9 /iCi/ml (about 9 disintegrations/min/litre), or 4X
10~12g/litre. Obviously, extreme precautions are necessary to limit losses to
this small amount.
Reactor Fuel and Blanket Materials

A wide variety of radionuclides is liberated from fuel elements when these
elements are processed chemically. Conventional processing at the present
time involves complete dissolution of the elements, separation and purifica-
tion of the heavy metals, especially uranium and plutonium, and discharge
of the waste to storage. The radioactive elements present in fuel elements
include both fission products and neutron-activated elements.
Fission products of the heavy metals include all the elements from about
atomic number 30 (zinc) to about atomic number 64 (gadolinium). The
elements as first formed have excess neutrons and rapidly change by
radioactive decay into more stable elements emitting beta and gamma
radiation. After a reactor has been operating for some time, it will contain a
large variety of radionuclides of both short and long half-lives. If a fuel
element is then removed from the reactor and stored, the short-lived
materials will change into either stable or longer-lived isotopes, so that most
of the radioactivity will then be in the form of materials of long half-life.
Some of the more important of these materials are listed in Table 5.
All reactor fuels, and especially all reactor blanket materials (materials
placed around the reactor core to form additional fissionable material from
neutrons that would otherwise escape), will contain a number of artificial
heavy elements, also listed in Table 5.
On dissolution of a fuel element, all these fission products and heavy
elements are free from confinement, and pass into the liquid. The volatile
materials, such as inert gases and halogens, pass off as vapors. If released to
the atmosphere, they can settle out or be washed into surface waters by rain,
thereby becoming potential contaminants of the water environment. The
fission products remaining after processing constitute an intensely radio-
active aqueous waste, which at present is stored indefinitely. The product
334 MANUAL ON WATER
TABLE 5—Nuclides in reactor technology.

Maximum PermissibleConcentration(Occupational)
Water Air
Nuclide Type juCi/ml a Hg/ litre /aCi/ml" jug/ litre
5 6
Strontium-89 fission product 7 X 10" 3X 10" 2 X 10~" 7 X 10"'°
Strontium-90 fission product 8 X 10"7 4X 10"6 2 X 10~'° 1 X 10"9
Ruthenium- 106 fission product 1 X 10"4 3X 10~5 2 X 10"8 6 X 10"9
Iodine-131 fission product 6 X 10~5 5X 10~7 6 X 10"9 5 X 10""
Cerium- 144 fission product 1 X 10"4 3X 10"5 2 X 10"9 6 X 10"10
Radium-226 natural 4 X 10"8 4X 10"5 8 X 10"12 8 X 10"9
Thorium (natural) natural 5X 10~7 5000 3 X 10"" 0.3

Thorium-234 natural 2X 10~4 9 X 10~* 1 X 10"8 4 X 10"'°
Uranium (natural) natural 2X 10~* 3000 3 X 10"" 0.05
Uranium-233 artificial heavy 3X 10"6 0.3 3 X 10"" 3 X 10"*
Plutonium-239 artificial heavy 3X 10"6 0.05 2 X 10"12 3 X 10"8
Americium-241 artificial heavy 3X 10~6 9 X 10"4 4 X 10"" 1 X 10"8
"International Committee on Radiological Protection, British Journal of Radiology, Supple-

ment No. 6, Table C. VIII, Dec. 1954.
phase, chiefly heavy metals, is also strongly radioactive; it requires special

methods of handling to avoid contamination of the -environment. Partly
because of the hazard, and partly because of the very high value of the
products, control is normally so thorough that only negligible quantities of
product are lost as wastes.
The extraordinary precautions necessary in handling these materials may
be judged from the quantities considered permissible in an industrial
environment (Table 5); if discharged to an uncontrolled area, these values
should be reduced by a factor of 10.
Radionuclides in Reactor Coolants and Moderators

In addition to the naturally occurring radioactive species discussed in the
preceding section, a large number of artificially produced radionuclides are
known. Most of these originate in the fission of heavy atoms or as a result of
the bombardment of stable materials by nuclear particles, especially neu-
trons ("activation"). Since a nuclear reactor functions by the fission of
uranium or plutonium and is an extremely strong source of neutrons,
radioactive isotopes are formed in large quantities.
In most reactors, the fission chain reaction is principally propagated by
neutrons of low (thermal) energies. The primary high-energy fission neu-
trons are slowed to thermal energies (moderated) by repeated interactions
with light elements, of which hydrogen is the most effective. Water, the
cheapest hydrogenous material, is a very common moderator. Water also is
frequently used for heat transfer in reactors.
Natural Water
Water itself, when exposed to a neutron flux, does not yield any important
radioactivities. There is a rather intense radiation from nitrogen-16, a
product of the interaction of neutrons with oxygen-16, but this material has
such a short half-life (7) that any system suitable for containing the reactor
will be adequate to protect against it. But many materials present in natural
waters will yield radionuclides of relatively long half-life under neutron
irradiation. The potential hazard from these materials must be considered in
plant operations, and, if discharged, their concentrations must be kept below
values at which they might be detrimental to the environment.
The largest American installation using natural water as coolant is the
reactor complex at Hanford, Wash. The effect of the reactor operations on
the Columbia River has been studied extensively, including the radionuclide
content of the water, and of fish and other biological forms in the river.
Although a large number of radionuclides are formed in the water during its
passage through the reactors, most of these have such short half-lives that
they very soon become insignificant. Radioactive forms of manganese,
arsenic, copper, sodium, phosphorus, zinc, chromium, and iron are prom-
inent in the reactor cooling water, and many others are formed in lesser
amounts.
In addition to the radionuclides formed by neutron irradiation of stable
elements in the water, there is always the possibility of fission product
leakage from the fuel elements into the cooling water. While careful design
and painstaking workmanship confine such leakage to trivial amounts, some
fission products can be identified in the water. The most important of these
are iodine, strontium, barium, and the rare earths. Generally speaking,
fission products are of greater biological significance than the induced
activities, and very elaborate instrumentation is provided to detect leakage
before it becomes a problem.
A rather special case of potential formation of radionuclides in a natural
water is the operation of nuclear-powered vessels. Here, although the
coolant moderator is usually very carefully purified water in a closed system,
the designer must consider the possibility of reactor neutrons passing
through the hull to activate sodium and other elements in the water in which
the ship floats. Sufficient neutron shielding to prevent this is therefore
included in the reactor plant.
Treated Waters
In most water-moderated or water-cooled reactors, extremely pure de-
mineralized water is used. This may be ordinary (light) water or deuterium
oxide (heavy water). In many systems the water is continuously purified by
ion-exchange demineralizers.
There are several reasons for using such rigorously purified water in
336 MANUAL ON WATER
reactors. One advantage of pure water is that only a small amount of

radioactivity is formed in it. In absolutely pure water, only the unimportant
7-s isotope nitrogen-16 would be formed. In practice, this ideal is unattain-
able, but the purer the water, the closer it can be approached. Another
advantage is neutron economy. Not only is every radioactive atom in the
water a nuisance, but its formation has required one neutron. Such
unproductive consumption of neutrons is expensive in reactor technology.
In spite of extreme purity and the use of highly corrosion-resistant alloys,
a minute amount of corrosion occurs in every system. Continuous bom-
bardment by the large neutron flux inside the reactor core induces very
significant activity even in these tiny amounts of corrosion products. The
nature of the alloys involved determines the character of the induced
activities. The most important induced radionuclides in'stainless steel
systems include isotopes of chromium, manganese, iron, cobalt, and nickel.
A second source of radionuclides is material specifically added for
corrosion control. Practically all the common materials of this type, such as
chromate, phosphate, and silicate, become radioactive in the reactor. A
special case is the production of tritium by neutron bombardment of lithium,
the lithium being added as hydroxide for pH control. Tritium, a pure beta
emitter of low biological effect, is considered a less serious problem than the
radioactive sodium or potassium that would result if hydroxides of these
elements were used.
Just as in natural water systems, there could be some minute leakage of
fission products from the fuel elements into the treated coolant water. Here
also, the fission products are more hazardous than the induced activities,
and their concentration in the water must be rigidly controlled.
Treated water systems, because of the high cost of the water, are
invariably closed, recirculating systems. Although the contaminated primary
coolant is retained within the shielded enclosure, some bleed-off and makeup
is usually necessary. Uncontrolled discharge of the radioactive waste water
to the environment is not permissible; therefore, waste tanks are usually
provided in which the water (a) is stored for decay, (b) is monitored before
discharge, and (c) is fed into the environment at a controlled rate so that
permissible levels are not exceeded.
Canal Coolants
In addition to reactor coolant and moderator, water is used in the "canals"
in which fuel elements are stored prior to processing. These fuel elements,
containing an enormous quantity of radioisotopes, require considerable
cooling to dissipate the residual heat of the radioactive decay. The gamma
radiation from these elements is very intense, so that they must be heavily
shielded to permit plant personnel to work in the vicinity. Water about 7.6 m
(25 ft) deep serves both of these purposes satisfactorily. The fuel elements,
after a sufficient cooling period, are removed by remote-control tools.
CHAPTER 16 ON RADIOACTIVE NUCLIOES IN WATER 337
Canal waters become contaminated by transfer of the radioactive primary

coolant adhering to the fuel element when it is removed from the reactor; in
the case of a leaking element, there is alslo some contamination from fission
products. The total contamination from both of these sources is ordinarily
minor, but provision must be made for draining any contaminated canal
water to waste tanks and replacing it with freshwater.
Radionuclides from Fallout
Fission Products
The most widespread type of radioactive contamination at present is fall-
out from nuclear weapons testings These tests have produced a considerable
atmospheric content of fission products of intermediate and long half-lives,
which gradually settle to the earth. In addition to settling under their own
weight and as a result of atmospheric currents, large quantities of these
materials are brought down by rain and snow.
In the United States during October 1958, it was unusual to find a rain or
a surface water that contained no detectable fission-product activity.
Levels of activity in rain may range from a few to thousands of picocuries
per litre, while in surface waters they will generally be of the order of tens
of picocuries per litre. In the period during and shortly after a weapons test
series, rain activity may reach the order of hundreds of thousands and
surface water activities hundreds of picocuries per litre. These increases are
primarily due to fission products of short half-life, which do jiot persist in
the atmosphere. Levels in groundwaters are ordinarily much lower than
these, depending largely on the speed with which the groundwater is
replenished by rain and surface water.
When rain containing fission product activity falls on the ground, much
of the activity is transferred to the soil, mostly by adsorption on soil par-
ticles or on atmospheric particulates washed down with the rain. As a
result, the fraction of radioactive material reaching streams or other bodies
of water is relatively small, except where large water areas are exposed to
direct rainout. Although water contamination is less than it would be if
these mechanisms were not operative, this cannot be counted as a clear
gain since much of the radioactivity remaining in the soil becomes available
to plants, and so enters the food chain with the possibility of eventual
ingestion by humans. Where rainfall is collected in cisterns for drinking
water, much of the radioactivity is removed by adsorption on the collection
surfaces, and by settling out and adsorption on the walls of the cistern.
The fission products cover a wide range of chemical elements. Conse-
quently, their behavior in water is characterized by large variations. Some,
such as zirconium, being very insoluble in most aquatic environments,
settle out more or less rapidly with paniculate material, thus becoming
concentrated in silts and sludges. Others remain in solution for a much
338 MANUAL ON WATER
longer time and hence can be carried much farther in streams. Still others
enter into the biological cycle and become distributed in a manner de-
pending upon the biota present in each stream.
As a result of the processes described in the preceding paragraph, com-
bined with normal radioactive decay, the total activity of a body of water
may decrease rapidly after contamination. This decrease, however, does
not necessarily imply a proportionate decrease in the biological hazard of
this water, which depends strongly upon the radionuclide composition of
the contamination. Removal of even large amounts of a relatively non-
hazardous component does not have much effect on the potential biological
damage.
Induced Activities
In addition to the fission products which are formed in fairly definite
proportions in the actual nuclear explosion, radioactive materials are
formed by irradiation with the neutrons produced in the explosion. The
nature and quantity of these induced radioactive nuclides depend on the
exact circumstances of the explosion, which determine the exposure to
neutron fluxes. Thus, in a high-altitude burst, only those materials present
in the bomb and in the air could be irradiated, while in a burst close to the
ground, all the elements present in the soil would be potential sources of
induced activity. In a subsurface explosion at sea, such materials as the
sodium in the water would become radiocative under the resultant neutron
bombardment.
The amount of fission product radioactivity formed during an explosion
is so large that normally the induced activity would not be noticeable in
the early stages; however, as the fission products of short half-life disappear,
those induced activities of longer half-life would become relatively more
prominent. A number of these induced radionuclides are concentrated by
biological systems and in many cases they have detected in fish or other
specimens taken from the water.
Table 6 lists some of the more important fission products and induced
radioisotopes that have been found in fallout.
Kadiotracers in Water
The number of potential applications of radiotracer studies in aqueous

systems staggers the imagination; only a few can be selected as examples.
For the sake of simplicity, tracer studies in which the water itself is the
subject will be chosen.
Selection of Tracer Materials

The prime requirement of any tracer is that it shall actually trace the
process under investigation. To illustrate, marbles are entirely unsatisfactory
TABLE 6—Radionudides found in fallout.

Long-lived fission products
Strontium-89 Ruthenium-106
Strontium-90 Cesium-137
Yttrium-91 Cerium-141
Zirconium-95 Cerium-144
Niobium-95 Promethium-147
Ruthenium-103 Samarium-151
Short-lived fission products

Molybdenum-99 Iodine-133
Technetium-99m I odine-135
Tellurium-132 Barium-140
Iodine-131 Praseodymium-143
Iodine-132 Neodymium-147
Induced activities
Carbon-14 Cobalt-57
Calcium-45 Cobalt-58
Manganese-54 Cobalt-60
Iron-55 Zinc-65
Iron-59
in tracing stream flow, oranges are reasonably satisfactory (at least so far as
surface flow is concerned), and fluorescein dye in solution is usually suitable.
The same principles apply with radiotracers. In flow studies, presumably the
best radiotracer for water would be tritiated water; reasonably good results
could be expected with tracers which react wîth the stream environment to a
very small degree, while tracers that tend to precipitate out, enter into the
biology of the stream, or be removed by ion-exchange properties of the silts,
would give poorer, less conclusive results.
After it has been established that one (or several) tracers are suitable, a
number of other characteristics of the radioactive material become im-
portant. These characteristics include ease of detection and measurement,
suitable half-life (long enough for the experiment but not so long as to
interfere with subsequent experiments or use of the water), low radiological
hazard, availability, and low (or at least reasonable) cost.
Hydraulic Tracing
A fairly large number of radionuclides have been used in tracing flow.
Tritium would be ideal except for the difficulty of measuring its extremely
soft beta radiation. Excellent results have been obtained with bromine-82,
but there is the objection that the material has a half-life of only 36 h. It is
therefore most useful in studying small artificial systems, such as settling
basins and the like. Another halogen that has been used successfully in such
applications is iodine-131.
In longer studies, a nuclide of greater half-life usually is needed. lodine-
131 is sufficiently long-lived for many experiments, but the maximum
permissible concentration of this material in public waters is so low that its
340 MANUAL ON WATER
use is thereby restricted. One of the best nuclides for flow tracing in nature is
rubidium-86, which has a half-life of 19 days, emits readily detectable beta
and gamma radiation, and is relatively insensitive to precipitation or other
sedimentation. Scandium-46 has also been used under conditions such that it
would remain in solution or at least form fine, essentially nonsettling
particles. In following the movement of silts, scandium-46 and gold-198 have
proved suitable.
Biological Tracing
The discussion here will be confined to studies of the movement of
radioisotopes in the biota of a stream. This is particularly important in
understanding the fate of radionuclides introduced into a stream. The
outstanding opportunity for this type of study has been in the Columbia
River, where use of the river water as reactor coolant has resulted in the
presence of a wide spectrum of irradiated radioelements in the stream. Many
of these materials are concentrated strongly by algae and other plankton;
much of this concentration appears to be a surface or adsorption phenom-
enon. As one examines other forms higher in the food chain, the number of
elements concentrated (and often the degree of concentration) becomes less,
partly because of rejection of these elements in the diet or in digestion, partly
by disappearance of nuclides of short half-life. Thus, in the Columbia River,
the nuclides most generally present in fish are phosphorus-32 and strontium-
89, in contrast to the reactor effluent activities, which consist largely of
arsenic, manganese, and copper.
Some laboratory experiments have been performed on the uptake of
certain nuclides by fish and other aquatic forms. Most of the work has been
concentrated on fission products, particularly strontium and cesium. The
results of these studies, oversimplified, may be summarized by saying that
many elements are taken up by the more primitive feeders, such as algae and
mollusks. Forms feeding upon these primitive forms generally take up less of
the insoluble elements, except in the contents of the digestive tract. A
number of nonfission-product nuclides, particularly zinc, cobalt, and iron
isotopes, are taken up by mollusks, and surprisingly large quantities of zinc
have been found in fish.
Disposal of Water Containing Radioactive Nuclides

Low-Activity Wastes
It is difficult to make hard-and-fast statements about the disposal of water
containing radioactive materials. As is the case with other wastes, disposal
varies with the nature and quantity of the material, and with the ability of
the environment to accommodate the waste. Since the term "radioactive
material" covers such a wide variety of nuclides of different physical,
chemical, and biological properties, the disposal of wastes will inevitably be

a matter of individual determination in each case.
A large majority of users of radionuclides can dispose of their wastes with
only the most elementary precautions. The Atomic Energy Commission has
established regulations for the sewer disposal of radionuclides by its
licensees. The average research user of isotopes need only measure the
activity of his waste to make sure that its concentration when discharged will
be acceptable. Beyond this, he must only make sure that the activity is
prevented from depositing on plumbing, where it might become a hazard.
Even where an isotope user must dispose of quantities too large for direct
discharge, simple treatment methods often suffice, for example, storage of
materials of short half-life. In this case the waste is retained until radioactive
decay reduces the activity to a level that will permit safe discharge. Where
storage would be impractical, as in the case of high activity or long half-life,
the radionuclide may often be removed from the waste water by coagulation,
precipitation, or ion-exchange techniques. The concentrated nuclide in solid
form can then be stored if convenient, or disposed of by burial. The
decontaminated water, after its radioactivity has been shown to be at a safe
level, can be discharged.
Waste water containing radionuclides at somewhat higher levels, say of
the order of a few millicuries per gallon, is rather common in the nuclear
industry. Here it is considered "low level" even though its activity may be
much higher than would be considered "low" outside the industry. Such
wastes often originate in laboratory or decontamination operations, or in
operations involving the draining of a small volume of highly radioactive
material. Normally such wastes require treatment prior to discharge. The
treatment may be impoundment for decay, coagulation or precipitation, or
ion exchange. A somewhat more stringent treatment, suitable for use with
higher concentrations of radionuclides, is concentration by evaporation,
suitable precautions being taken to minimize carry-over of spray in the
condensate. This converts the waste into two parts, a condensate ready for
monitoring and discharge, and "bottoms" of high-level waste. Electrodialysis
with ion exchange membranes, which has been suggested as an alternative
treatment to evaporation, must still be considered an experimental method
although theory and preliminary investigations have been somewhat prom-
ising. Since these treatments will not affect all nuclides to the same extent,
the discharge of the treated water must be regulated in accordance with its
nuclide composition as well as its total radioactivity content.
Wastes may be discharged to the environment only after necessary
treatment, when required, and after monitoring to make sure that the
radioactivity actually is at an acceptable level. Discharge may be through
sewer systems, when convenient, or directly into streams or other bodies of
water. Discharge into the ground is often practiced in controlled areas. In
this case, the ion-exchange capacity of the soil often provides a form of
storage so that somewhat higher levels of activity can be discharged than
342 MANUAL ON WATER
would be acceptable for direct stream discharge. Adequate monitoring is

required where ground disposal is used, to insure that the groundwater will
not be adversely affected.
High-Activity Wastes
High-activity wastes may be defined as those containing radionuclides at a
concentration of the order of curies per gallon. Most of these wastes
originate in the chemical processing of irradiated reactor fuels. Their
treatment is made difficult by their chemical properties (acidity, salt content,
and so on) as well as by their radioactivity. Generally speaking, these wastes
cannot be discharged. They must be stored, and treatment and disposal is
only a way of making the storage as inexpensive as possible. The wastes may
be stored in tanks, or disposal may be concealed as in discharge to the ocean
or into geological formations, or they may be converted into a solid,
nonleachable form. In all these cases, however, there is reliance on the
material staying where it is put, and not being allowed to leak into the
environment.
The only reliable, tested method for large quantities of high-level wastes is
permanent tank storage. The waste is usually concentrated by evaporation to
the greatest extent possible, the condensate being discharged as a low-level
waste. The concentrated bottoms are then transferred to special tanks. These
tanks must be corrosion-resistant and must provide for removal of the heat
produced by radioactive disintegration. Although storage is stated to be
"permanent," this cannot yet be considered definitely established, because
the nuclear industry is less than thirty years old and retention for centuries is
required. It is definitely established, however, that this treatment is ex-
pensive.
A treatment successfully used on small quantities of highly radioactive
wastes is conversion into a solid such as concrete. One way of doing this is to
add enough cement to the waste that it will set. The cement is usually
allowed to set in drums, which are then buried either in the ground or in the
deep sea. A modification of this method which offers much promise, and
which has been tested on a small scale, is conversion of the waste into a
glassy or ceramic material. The waste is calcined to alumina or zirconia and
leached. The leached radionuclides may be taken up on clay be adsorption or
ion exchange, and the clay is then fired to change its crystal structure in such
a way that the radionuclides cannot be desorbed or leached. The ceramic
pellets could presumably be stored much more cheaply than tank storage,
because of the two factors of reduced volume and noncorrosive nature.
Another form of treatment that has been proved in the laboratory involves
separation of the long-lived nuclides strontium-90 and cesium-137 on
barium sulfate and potassium alum, respectively. These two nuclides may
then be stored rather cheaply because the volume has been markedly
reduced. An aqueous phase suitable for discharge is obtained by solvent
extraction of the residual fission products from the waste water. These
fission products present a much simpler storage problem because of their
shorter half-lives.
Ultimate Disposal
The term "disposal at sea" really has two meanings. In one, it is
contemplated that the material remain sealed indefinitely out of contact with
the seawater; in the other, the material is introduced into the deep sea but is
not sealed from it, and reliance is placed on the slowness of mixing in the sea
to limit the spread of radioactivity into the human environment. Unfor-
tunately, our present knowledge of the sea is not adequate to permit full
confidence in slow mixing. Rates of transport, and especially of biological
transport, are still very uncertain. Therefore, all present disposal at sea is
actually sealed storage. Sealed containers, usually concrete-filled steel
drums, designed to withstand the pressures and corrosion of the sea, are
discharged into depths of 1000 fathoms or more. Presumably the drum and
eventually the concrete will be corroded away, but it is hoped that this
process will be slow enough that even the longer-lived dangerous nuclides
will have decayed to insignificant levels.
Because of the cost of tank storage and of transportation for burial at sea,
there is much interest in underground disposal. This is different from
discharge into the soil as utilized for low-level wastes. The concept with high-
activity wastes is utilization of sealed geological formations as permanent
storage. A number of types of geological formations have been considered;
at present the most promising appear to be salt domes.
Sampling and Measurement of Radioactive Water
Sampling
In sampling water for determination of its radioactivity, whether this be
for gross or for radionuclide assay, the normal principles of sampling apply.
Since no analysis is better than the sample, it is necessary to make certain
that the sample drawn is actually representative of what is needed, both in
space and time. The possibility that the analyst may require a continuously
composited sample, even one composited proportionally to flow rate, should .
not be disregarded. Where adequate mixing is not assured, a number of
samples from different points and various depths may be taken, either for
individual analysis or for compositing. The principles involved in this aspect
of sampling will not be mentioned further, because they have been covered in
detail in Chapter 9.
Sampling for radioactivity determinations, however, does have one
requirement not commonly found in other sampling. This is the need to
prevent loss of the radioactive material to the sampler or container. When,
344 MANUAL ON WATER
as is often the case, the amount of nuclide is only of the order of 10~12 g, or
even less, normal adsorptive losses on container walls may amount to an
important percentage of the activity. Glass or plastic materials appear to give
somewhat less difficulty of this kind than do metals. Where necessary,
addition of carrier materials or chelating agents or acidification may be
employed to minimize adsorptive losses.
Often, no adsorptive losses are found in sampling streams or other natural
waters. It appears that in such waters the easily adsorbable materials are
already on the surfaces of whatever suspended solids may be present.
Therefore, treatment to prevent adsorption may result in undesired transfer
of radionuclides from the suspended to the dissolved phase.
Measurement
Only the mos*t general survey of the problems of measurement of
radioactivity in water is in order here. Specific details are the subject of
numerous papers and books. As it is often necessary to measure extremely
small amounts of radioactive material in water, the problem of sensitivity or
detectability becomes paramount.
Detectability is limited by the random nature of the radioactive disinte-
gration process. The result is a statistical uncertainty in the number of counts
observed, setting an inherent lower limit to the precision of the deter-
mination. At low enough levels, there is a question of whether there actually
were any counts in a given period beyond what might be due to the
background, that is, the counts caused by cosmic rays and by naturally
occurring radionuclides in and around the instrument. Three techniques are
available to get enough counts to permit a definite answer to this question:
start with a large sample, detect as many of the radiations as possible, and
suppress the background. It is not often convenient to collect or process
samples larger than a litre or so, and sample size is often further restricted by
solids content. This will be further considered in the discussion of methods
of sample preparation.
High efficiency of detection is chiefly a function of the instrument type,
although it is also affected by sample preparation. As far as the instrument is
concerned, three factors are paramount. First is geometry, or fraction of
radiation emitted toward the detector. This is primarily determined by
sample and detector size, and by the distance between them. Next comes
absorption, which refers to failure of part of the radiation to penetrate the
sample solids, the air or other material between the sample and detector, or
the wall of the detector itself. Third is the actual detector efficiency, or the
fraction of radiations entering the detector which causes a response. Highest
overall efficiency occurs when a sample is placed inside a detector able to
respond to each disintegration. Such 100 percent efficiency is customarily
attained only with gaseous samples, although it can be approached with
solid samples in special 4-pi counters.
Background suppression is most commonly attained by surrounding the

detector with lead or other heavy shielding to block gamma radiation from
reaching it. Lead is used in thicknesses up to about 100 mm (4 in.); iron up to
200 mm (8 in.). Even such quantities of shielding will not block cosmic rays;,
these can only be eliminated by adding additional detectors which inactivate
the counter momemtarily when a cosmic ray passes through them. For
lowest background levels, careful choice of the materials of the detectors and
shielding is important. Thus iron normally contains less radioactive material
than lead, and so is more suited for critical shielding applications.
Instrumentation Types
Three principles of detection are used in measurement: gas ionization,
scintillation, and photographic darkening. The last is omitted as not strictly
an instrumental technique, although it is very widely used in such applica-
tions as personnel monitoring by film badges and in autoradiography. Each
of the other two techniques may be used to provide an average, pointer-type
indication, or to count individual events.
A second classification of instruments is by the type of radiation they are
designed to measure. Since alpha particles are characterized by small
penetrating power and high energy dissipation, alpha detectors must be
made with extremely thin windows for the entrance of the radiation, or
(more usual) with provision for the sample to be introduced into the
detector. Only moderate amplification is provided to take advantage of the
high energy dissipation, permitting low response for other types of radiation,
including cosmic rays. In scintillation alpha detectors, discrimination against
other radiations is provided by using a very thin zinc sulfide scintillating
screen which gives only very small light pulses with beta and gamma
radiation.
In the detection of beta particles, window requirements are less stringent
than for alpha particles. The ordinary Geiger-Miiller tubes may have glass or
metal walls with a thickness of 30 mg/cm2, or mica windows from 1 to 4
mg/cm2 thick. For the highest efficiency, the sample may be introduced into
the detector, as for alpha particles. In fact, the same instrument may be used
to detect alpha particles or beta (plus alpha) particles merely by changing the
voltage so as to provide higher or lower amplification. Scintillation beta
detectors commonly have an organic scintillator, such as an anthracene
crystal. The scintillator molecules may be dispersed in plastic to make
detectors of varying sizes and shapes.
Gamma rays are characterized by large penetration through matter;
therefore the detectors should have as much mass as possible. High atomic
number increases the response. Ionization chambers may contain gas under
high pressure, while gamma Geiger tubes usually have cathodes of silver,
bismuth, or other heavy metal. Windows may be made heavy (300 mg/cm2
or more) to reduce response to beta particles. The detection of gamma
346 MANUAL ON WATER
radiation by scintillation counters has reached a high state of development.

Most detectors use sodium iodide crystals (thallium-activated). The heavy
salt provides a large mass, and gamma absorption is enhanced by the large
percentage of iodine, with its high atomic number.
Some less-common instruments of more specialized type may be men-
tioned briefly. The 4-pi counter is designed to permit detection, with 100
percent efficiency, of alpha or beta particles. Sample preparation require-
ments are so stringent that they limit use of the instrument to calibration,
rather than to routine work. In liquid scintillation counters, the sample is
dissolved in an organic solution of a scintillator, so that all the radiations can
produce scintillations. This instrument has been especially developed for the
special cases of measuring the very weak radiations from carbon-14 (0.15
MeV) and from tritium (0.018 MeV).
Sample Preparation
Sample preparation techniques are intimately connected with the type of
radiation and with the instrumentation available. Alpha particles travel only
a few micrometres in water and beta particles only a few millimetres; thus
any detector measuring these in unprocessed water will be observing only a
very thin film immediately surrounding it. Such an instrument will therefore
respond only to relatively high radionuclide contents. Gamma rays, on the
other hand, have a much larger range in water, and large unprocessed
samples can be used. Even in this case, the material at large distances from
the detector will have poorer geometrical efficiency.
For alpha and beta measurements, and for many gamma measurements,
separation of the radioactive material from the water is required. The most
common technique for this is evaporation of the water and transfer of the
residual solids to a dish or planchet suitable for the counter being used.
Removal of the water does not completely eliminate self-absorption loss.
Even with a sample as thin as 5 mg/cm2, an appreciable part of the beta
radiation and as much as 50 percent of the alpha radiation are lost.
Chemical separation is required where a specific radionuclide must be
determined in a mixture. This technique must also be used if the solids
content of the water is so high that simple evaporation is inapplicable.
Hazards
The present thinking on radiological hazards is best summarized by the

staff reports of the Federal Radiation Council. The most important points
are that the total accumulated radiation dose should be kept as low as
possible, and in any case shall not exceed 5 rems for each year beyond age
18, nor shall the dose in any 13-week period exceed 3 rems. Where exposure
is restricted to skin, extremities, bone or thyroid, somewhat higher doses are

permitted.
Hazards to workers with radioactive materials may be external and
internal. External hazards arise as a result of the worker being close to a
source of ionizing radiation. Measurement of radiation levels in the vicinity
of radioactive material is relatively simple in most cases. The primary
concern in this situation is with the highly penetrating gamma rays, since it is
a simple matter to shield out the beta particles. Alpha emitters are of no
concern as external sources, since this type of radiation is unable to
penetrate the nonliving exterior layer of the skin. The permissible dose from
the external sources of ionizing radiation is discussed in the National Bureau
of Standards Handbook 59, as amended April 15, 1957.
Several kinds of measuring devices are in general use as monitors of
radiation levels. One of the most useful is the film badge, routinely worn
during working hours. This is simply a holder for X-ray film. At the end of a
predetermined period, which may be a week or more, the films are removed
from the badge, developed, and the accumulated radiation dose determined
by comparison with films exposed to known doses. The film badge is widely
used because of its simplicity, reliability, and permanence as a record. The
film badge usually includes devices to permit identification of the type of
radiation; by varying the thickness and nature of the cdvering of the film it is
possible to distinguish beta particles, soft gamma rays, harder gamma rays,
and neutrons.
Survey meters are commonly used for more rapid indication of radiation
levels. One type (ionization chamber) indicates the radiation level by
measuring the ion current in a gas space between two electrodes. This
current is proportional to the ionization produced by the radiation. A
second type employs a Geiger tube and gives an indication proportional to
the counting rate. This type, while considerably more sensitive, is less
directly related to dose rate and is not operable in high radiation fields. Both
types may come with shutters thick enough to stop beta particles and thus
differentiate between beta and gamma radiation.
Two special forms of these instruments deserve special mention. One is the
dosimeter, which is a small ionization chamber that can be carried in a
pocket. Before use, the electrodes are charged to a definite voltage, and the
total accumulated dose can be read at any time by measuring the decline in
voltage. The other form is the alpha survey meter, usually an ionization
chamber or proportional chamber equipped with a very thin window
through which alpha particles can pass. The primary function of this
instrument is the measurement of alpha contamination of surfaces.
Three factors are paramount in the control of external radiation hazard:
shielding, distance, and time. Storage in suitable shields reduces radiation
from sources to an acceptable level. Well-designed shielding during actual
work with nuclides can minimize the radiation dosage. The radiation level
decreases as the square of the distance from a point source, and many
348 MANUAL ON WATER
operations can be carried out safely by remaining at a sufficient distance

from the source. Remote handling tools, simple or complicated, are
important in such cases. When it is impossible to reduce the radiation level in
an operation to a low level, the duration of exposure may be controlled,
either by performing the operation rapidly, or by using several sets of
workers to complete it.
Internal hazard through ingestion or inhalation of radioactive material is
more subtle. The most important factors in its prevention are good working
habits, particularly careful planning, and good housekeeping practices. Not
only should the quantities of nuclides used in an experiment be kept as small
as possible, but nuclides of low physiological hazard should be used
whenever there is a choice. All operations that may produce airborne
radionuclides should be carried out in an adequate hood. Good house-
keeping practices include working in trays, regular monitoring and cleanup
of surfaces, careful washing when leaving the laboratory, and prohibition of
eating, drinking, or smoking in the laboratory.
Monitoring of workers for internal contamination is most often done by
analysis of urine, or less often of feces. If a worker is aware that he has
ingested radioactive material, he should induce vomiting. Treatment beyond
this is strictly the function of the physician, who should be called immedi-
ately.
The Federal Radiation Council has divided the possible intake by humans
of four of the more hazardous radionuclides into the three ranges and
appropriate responses listed in Table 7. The upper levels of Range II
correspond to the radiation protection guides for population groups. The
maximum permissible body burdens for other radionuclides and the maxi-
mum permissible concentrations of radionuclides in water are given in
National Bureau of Standards Handbook 69. For drinking water supplies,
the Public Health Service Drinking Water Standards, 1962, states that they
shall be approved when radium-226 and strontium-90 concentrations do not
exceed 3 and 10 pCi/litre, respectively, or if the total human intake from all
sources is within the limits recommended by the Federal Radiation Council
for control action. In the known absence of strontium-90 and alpha emitters,
TABLE 7—Graded scale of action in response to transient intakes.

Range of Radionuclide Intake, pCi/day
Scale of Action Ra-226 1-131 Sr-90 Sr-89
Range I: Periodic confirmatory Oto20 0 to 10 Oto20 0 to 200
surveillance as necessary
Range II: Quantitative surveil- 2 to 30 10 to 100 2 to 200 200 to 2 000
lance and routine control
Range III: Evaluation, and ap- 20 to 200 100 to 1 000 200 to 2000 2 000 to 20 000
plication of additional
control measures as
necessary
the water supply is acceptable when the gross beta concentration does not
exceed 1000 pCi/litre.
Sources of Information
Because of the novelty of the subject of radioactivity to many of the readers of this Manual,
information on the literature of this field may be useful. It is not possible to make an exhaustive
compilation of sources; accordingly, the following materials have been listed in the belief that
they form a useful survey. Very possibly, equally good materials have been omitted through lack
of acquaintance with them.
Books
For the novice in the field, books undoubtedly give the best introduction to a subject. The
first three books listed are excellent surveys of radioactivity and atomic energy, at a level
comprehensible to the nonspecialist:
Friedlander, G., Kennedy, J. W., and Miller, J. M., Nuclear and Radiochemistry, 2nd ed.,
Wiley, New York, 1964.
Lapp, R. E. and Andrews, H. L., Nuclear Radiation Physics, 3rd ed., Prentice-Hall, New
York, 1963.
Glasstone, S., Sourcebook on Atomic Energy, 3rd ed., Van Nostrand, Princeton, N. J., 1967.
The next book can be recommended as a classic in the field, although somewhat old. It is
particularly strong with respect to naturally occurring radionuclides:
Rutherford, E. R., Chadwick, J., and Ellis, C. E., Radiations from Radioactive Substances,
Cambridge University Press, 1951.
The next group includes somewhat less general texts, each good in its field, which is generally
indicated by the title:
Murray, R. L., Introduction to Nuclear Engineering, 2nd ed., Prentice-Hall, New York, 1961.
Glasstone, S., Principles of Nuclear Reactor Engineering, Van Nostrand, Princeton, N. J., 1955.
McCullough, C. R., Safety Aspects of Nuclear Reactors, Van Nostrand, Princeton, N. J., 1957.
Sharpe, J., Nuclear Radiation Detectors, Methuen, London, England, and Wiley, New York,
1955.
Taylor, D., The Measurement of Radioisotopes, Methuen, London, England, and Wiley, New
York, 1951.
Smyth, H. D., "A General Account of the Development of Methods of Using Atomic Energy
for Military Purpose Under the Auspices of the United States government" (1940-1945), U. S.
Government Printing Office, Washington, D. C., 1945 (better known as the "Smyth Report").
Glasstone, S., Ed., The Effects of Nuclear Weapons, U. S. Government Printing Office,
Washington, D. C., 1957 (much information on fallout).
Coryell, C. D., and Sugarman, N., Eds., Radiochemical Studies: The Fission Products,
McGraw-Hill, New York, 1951.
Wahl, A. C., and Bonner, N. A., Radioactivity Applied to Chemistry, Wiley, New York, 1951.
Compendia
Intermediate between books and the periodical literature are reports of conferences, groups
of papers, and similar compendia. A few of these may be listed:
National Nuclear Energy Series—a record and summary of Manhattan District and early
Atomic Energy Commission research. McGraw-Hill, New York.
Proceedings, International Conference on the Peaseful Uses of Atomic Energy, the record of the
1955 Geneva Conference^6 volumes, United Nations, New York, 1956.
Proceedings, Second International Conference on the TeacefuJJUses of Atomic Energy, the
record of the 1958 Geneva Conference, 34 volumes, United Nations, "New York, 1959. (An
edited series of 12 volumes is available from Pergamon Press, New York.)
Progress in Nuclear Energy, nine series each covering a specific area. Probably will appear
annually. Pergamon Press, New York.
350 MANUAL ON WATER
A further group of useful publications of this general type include:

Semi-annual Reports to Congress, U. S. Atomic Energy Commission, U. S. Government
Printing Office, Washington, D. C. Each of the later reports is subtitled according to its
principal subject. These include:
Sixth Semi-Annual Report, "Atomic Energy and the Life Sciences."
Eighth Semi-Annual Report, "Control of Radiation Hazards in the Atomic Energy Program."
Thirteenth Semi-Annual Report, "Assuring Public Safety in Continental Weapons Tests."
Twenty-first Semi-Annual Report, "Radiation Safety and Major Activities in the Atomic
Energy Programs."
Twenty-fifth Semi-Annual Report, "Atomic Industrial Progress and Second World Con-
ference."
Annual Reports, United Kingdom Atomic Energy Authority, H. M. Stationery Office, London,
England.
"The Nature of Radioactive Fallout and Its Effects on Man," hearings before the Special
Subcommittee on Radiation of the Joint Committee on Atomic Energy, 85th Congress, U. S.
Government Printing Office, 1957.
"Industrial Radioactive Waste Disposal," hearings before the Special Radiation Subcommittee
of the Joint Committee on Atomic Energy, 86th Congress. U. S. Government Printing Office,
1959.
Reports of the National Academy of Sciences—National Research Council, Washington, D. C.:
The Biological Effects of Atomic Radiation*
"The Disposal of Radioactive Waste on Land"
Symposium on Radioactivity in Industrial Water and Industrial Waste Water, ASTM STP235,
American Society for Testing and Materials, 1959.
Recommendations of the National Committee on Radiation Protection and Measurements
(NCRP), published as Handbooks of the National Bureau of Standards. U. S. Government
Printing Office, Washington, D. C. The handbooks include:
Handbook 49, "Recommendations for Waste Disposal of Phosphorus-32 and Iodine-131 for
Medical Users," Nov., 1951.
Handbook 51, "Radiological Monitoring Methods and Instruments," April, 1952.
Handbook 52, "Maximum Permissible Amounts of Radioisotopes in the Human Body and
Maximum Permissible Concentrations in Air and Water," March, 1953.
Handbook 53, "Recommendations for the Disposal of Carbon-14 Wastes," Oct., 1953.
Handbook 54, "Protection Against Radiations from Radium, Cobalt-60, and Cesium-137,"
Sept., 1954.
Handbook 58, "Radioactive Waste Disposal in the Ocean," Aug., 1954.
Handbook 59, "Permissible Dose from External Sources of Ionizing Radiation," Sept., 1954
(amended April 15, 1957).
Handbook 69, "Maximum Permissible Body Burdens and Maximum Permissible Concentra-
tions of Radionuclides in Air and Water for Occupational Exposure," June, 1959.
Kinsman, S., et al, "Radiological Health Handbook," Office of Technical Services, U. S. De-
partment of Commerce, Washington, D. C.
"Glossary of Terms in Nuclear Science and Technology,' USA Standard N 1.1-1967, American
Society of Mechanical Engineers, United Engineering Center,. 345 E. 47th Street, New York,
N. Y. 10017, 1957.
Periodical Literature
The periodical literature in nuclear energy is so voluminous that only a brief glimpse into the
field can be given. In addition to the usual journal publications, a vast number of government
reports (including U. S. Atomic Energy Commission, U. K. Atomic Energy Authority, and
similar groups) are available. For this reason, the abstract literature in this field is especially
important. Fortunately, the first journal listed below does an outstanding job of covering the
report literature, in addition to providing adequate coverage of the journals. The other abstract
coverage listed is generally somewhat more specialized than Nuclear Science Abstracts.
Abstracts, Reviews, and Bibliographies
Nuclear Science Abstracts, NSAB-A
Public Health Engineering Abstracts
Journal, American Water Works Association, JAWWA, Abstract Section

Chemical Abstracts, CHABA, (outstanding coverage of journal literature)
Sewage and Industrial Wastes, SIWAA: Annual Review of Literature on Radioactive Wastes
Analytical Chemistry,ÂNCHA: Reviews (usually each April)
Water Pollution Abstracts
TID Bibliographies, U. S. AEC, especially:
TID-4563, "Special Sources of Information on Isotopes"
TID-3511, "Bibliography on Uses of Radioactive and Stable Isotopes in Industry"
TID-3050, "Bibliography of Selected AEC Reports of Interest to Industry"
AERE Bibliographies, United Kingdom Atomic Energy Authority, especially:
AERE-Lib (LI) "Sources of Information in Atomic Energy."
AERE-Inf/Bib % "A List of Reports and Published Papers by AERE Staff (with Supple-
ments).
"Publications on Atomic Energy," H. M. Stationery Office, London, England, 1956.
"Selected Readings on Atomic Energy," U. S. Govt. Printing Office, Washington, D. C., 1955.
General Journals
Nucleonics, NUCLA (particularly valuable for nuclear news)

International Journal of Applied Radiation and Isotopes, IJARA (applications of radio-
nuclides in industry, medicine, and science; radiation techniques; radiochemistry)
Science (contains technical articles of general interest)
Nature (generally similar to Science)
Analytical Techniques, Including Instrumentation:
Analytical Chemistry, ANCHA
Analytica Chimica Acta, ACACA
Analyst
Reviews of Scientific Instruments
Water Journals:
Journal, American Water Works Association, JAWWA

Sewage and Industrial Wastes, SIWAA
Water Pollution Research
Miscellaneous
Transactions, American Geophysical Union, TAGUA, (hydrology; meteorology; natural

activity)
Radiation Research, RAREA
Industrial and Engineering Chemistry, IECHA (especially radiation waste treatment)
Health Physics
Chapter 17—Nuclear Water

Technology
The first application of nuclear reactor systems to commercial power

generation was at the Shippingport Atomic Power Station, which began
operating in December 1957. This station operates as part of the Duquesne
Light Company grid, supplying power to the Pittsburgh, Pa. metropolitan
area.
Today, 21 years later, nuclear power stations satisfy a significant portion
of our electrical power requirements. As of December 1970, there were in the
United States 104 civilian power reactors operating or on order, having a
combined electrical generating capacity of 86794 M We (megawatt electrical).
Of these, stations generating a total of 8000 MWe were operational. The
total U. S, electrical generating capacity in operation at this same time was
approximately 341 000 MW. Forecasters indicate that, by the last decade of
this century, nuclear power will represent 40 percent of the approximately
1 400 000-MWe electrical generating capacity that will be installed by that
time and will generate over 50 percent of the electricity used.
Most of the power reactors in service or under construction today are
thermal reactors using various coolants and moderators. Great Britain
played a pioneering role in the development of gas-cooled (carbon dioxide)
and graphite-moderated reactors and is the largest manufacturer of gas-
cooled reactors. Canada has specialized in heavy water-cooled and mod-
erated reactors. However, the majority of plants in use or under construction
in the United States and elsewhere use light water for both coolant and
moderator. The two types of light-water reactors are pressurized water
reactors (PWR) and boiling water reactors (BWR).
The few fast reactors in operation or under construction today are
basically demonstration plants. While they cannot compete economically
with thermal reactors at the present time, they are considered to be the
reactor type of the future. Fast reactors use either materials with high mass
number (liquid metals like sodium, for example) or gases as a coolant.
In this chapter, boiling water reactors and pressurized water reactors using
natural water coolant, those most widely used for power generation, are
discussed.
352
CHAPTER 17 ON NUCLEAR WATER TECHNOLOGY 353
Pressurized Water Reactor Power Plant
A typical pressurized water reactor system (PWR) is shown in Fig. 1. Heat
generated by the fissioning fuel in the nuclear core is removed by water
(reactor coolant) circulating at high pressure through the primary circuit.
The water in the primary circuit both cools and moderates the reactor. The
primary water is kept under pressure to keep it from turning to steam.
The heat is transferred from the primary coolant to the secondary system
coolant (also light water, but unpressurized) in a shell-and-tube type heat
exchanger, or to a steam generator turning the secondary coolant to steam.
Typical operating conditions are summarized in Table 1. The steam
produced in the steam generator is at a lower pressure and temperature than
the primary coolant. Therefore, the secondary portion of the cycle is similar
to that of the moderate-pressure fossil-fueled plant.
Boiling Water Reactor Power Plant

In a BWR power plant (Fig. 2), the nuclear core is contained in a pressure
vessel. Light-water coolant is circulated through the core, but, since it is
unpressurized, is turned directly to steam. The steam is separated from the
bulk coolant and delivered directly to a turbine and then to a condenser. The
FIG. 1—Typical pressurized water reactor (1 psi = 6.895 kPa).

354 MANUAL ON WATER
TABLE 1—Typical operating conditions for PWR systems.

Primary System PWR BWR
Temperature 288 to 321°C (550 to 610°F) 285 to 313°C (545 to 596°F)
Pressure 13790to 15514kPa 6895 to 10 342 kPa
(2000 to 2250 psig) (1000 to 1500 psig)
Secondary System
Steam temperature 266 to 274°C (510 to 525° F) none
Pressure 4826 to 6205 kPa (700 to 900 psig) none
essential difference between the PWR and BWR is that the latter uses a
direct cycle, that is, no heat exchanger.
Typical operating conditions for the BWR are given in Table 1.
Reactor Components
A power reactor is composed of a core or group of fuel elements, control
rods, coolant, and moderator (Fig. 3). Heat energy is generated in the
reactor core, which contains an array of fuel assemblies. Various typical
PWR cores are rated from approximately 1500 to 3500 MWt (megawatt
thermal) with corresponding electrical outputs of 500 to 1200 MWe. An 800-
MWe core, for example, is approximately a right circular cylinder with an
equivalent diameter of 345 cm (136 in.), an active height of 348 cm (137 in.),
and is made up of 217 fuel assemblies. About 80 metric tons of uranium
oxide (UCh) are used in the reactor core. The core is completely open,
shrouded only at the outer periphery.
Each fuel assembly is made up of 176 fuel rods supported within a rigid
frame consisting of stainless steel upper and lower end fittings, Zircaloy fuel
rod spacer grids, and five Zircaloy axial guide tubes which will accept a
control element assembly (control rods). The overall length is about 394 cm
FIG. 2—Schematic arrangement of boiling water reactor power plant.

FIG. 3—PWR reactor assembly (1 in. - 2.54 cm).
(155 in.), and the cross section is approximately 8 in. (20 cm) square. Each
assembly weighs about 585 kg (1300 Ib) (Fig. 3).
Each fuel rod contains sintered uranium oxide pellets of approximately
10 mm (0.4-in.) diameter and approximately 15 mm (0.6-in.) long. The
Zircaloy-4 cladding tube is approximately 0.66 mm (0.026 in.) thick. This
material has good corrosion behavior in the reactor environment and has
low neutron absorption cross sections. The tube length accommodates the fuel
pellet column and a fuel and gas expansion space. A diametral gap between
the pellets and the cladding provides for differential expansion between the
cladding and fuel to limit clad strain.
Control rods are located in selected fuel assemblies and contain materials
356 MANUAL ON WATER
possessing high neutron absorption cross sections. Their relative position

(depth) in the core regulates the rate of power generation. Control rod
materials commonly used are boron as the carbide (IkC) or the alloy silver-
indium-cadmium (Ag-In-Cd) (Fig. 3).
The coolant, or core heat removing medium, is water. A typical 800-MWe
PWR coolant system flow is approximately 19 m 3 /s (300000 gal/min).
Coolant temperature at the core inlet is 288° C (550° F) and at the core exit
316°C (600° F). In a pressurized water reactor, the coolant water also serves
as the moderator.
Materials of Construction
The materials of construction listed in Table 2 are typical of modern water
reactors and consists of 300-series stainless steel primary system surfaces,
Zircaloy-clad fuel elements, and, in PWR's, Inconel-600 steam generator
tubing. With the possible exception of core materials, conventional con-
struction materials are used throughout the reactor coolant and steam-
condensate cycle.
These materials are common to both systems. For BWR's and those
PWR's that produce superheated steam, no copper alloys can be used for
condenser or feed water heater material; stainless steel only is employed.
This is to prevent thermally insulating copper deposits from forming on
fuel element surfaces (BWR) or steam generator (superheat section) tubing
(PWR).
TABLE 2—Materials exposed to coolant in a typical water reactor coolant system.

Reactor vessel
vessel cladding weld-deposited SS*
vessel internals 304 SS
instrument and control rod
drive nozzles Ni-Cr-Fe alloy 600"
Fuel cladding and guide springs Zircaloy-4
Control rod drives 304 SS and 410 SS
Pipe cladding-main loop pipes 304LSS
Surge and spray piping 316 SS
Steam generator
bottom head cladding weld-deposited SS*
tube sheet cladding weld-deposited Ni-Cr-Fe alloy-6000
tubes Ni-Cr-Fe alloy-6000
divider plate 410 SS
Pumps
casing 316 SS
internals 304 SS
Pressurizer
cladding Type 304 SS* and Ni-Cr-Fe alloy 600
heaters Ni-Cr-Fe alloy-600«
"Inconel-600.
Êquivalent to Type 304 SS as deposited.
Water Quality
High-purity water is needed in nuclear applications principally to pre-
clude fouling of heat-transfer surfaces, prevent corrosion, and preclude the
creation of radioactive materials through nuclear activation processes.
Corrosion can impair material integrity or increase corrosion product
concentrations, which result in activated residues on the primary system
surfaces, filters, and in waste systems.
Water has many uses in a nuclear plant, and the end use dictates the
quality requirements. The water used is largely recycled, but normal valve
or gland leakage may contain radioactivity and must be collected and
processed prior to reuse or discharge. Radioactive liquid waste processing
employs a wide variety of water purification techniques. Water is also used
for shielding in fuel storage pools, as a cooling fluid in reactor and auxiliary
cooling circuits, and as the working fluid.
As in any power cycle, long-term reliability is a necessary goal. To
achieve this goal, water quality control is established to opitimize the
environment in which the system materials will operate. The main objectives
of this control are:
1. To reduce the general corrosion of system materials.
2. To maintain material integrity of components and piping.
3. To assist in reactivity control of the reactor in PWR's.
The first two objectives are applicable to any power plant water treatment
program regardless of whether conventional or nuclear fuel is used. The
nuclear water technologist deals largely with the same problems as the
conventional power plant chemist.
There are many similarities in water quality requirements between PWR
and BWR power plants. In the primary system of a PWR or BWR, the
presence of the reactor neutron flux results in unique chemical conditions
not found in fossil-fueled plants. This flux has two major effects on water
chemistry control:
1. Impurities can be made radioactive.
2. Radiation can induce or cause normally nonoccurring chemical reac-
tions to take place.
Makeup Water Quality

To assure water quality in tfie reactor primary coolant, makeup water
must be maintained to equivalent specifications. The typical water quality
specification for reactor coolant makeup water is given in Table 3.
One example of a common impurity which can be introduced is the
chloride ion. The potential for stress corrosion of austenitic stainless steels
in the presence of chloride and oxygen is well documented. Since austenitic
stainless steel is used extensively in construction of PWR systems, it is
obvious chloride and oxygen must be rigidly controlled. The specifications
for chloride and oxygen are as follows:
358 MANUAL ON WATER
TABLE 3 — Reactor coolant makeup water chemistry.

Specifications Operating Limit
specific resistivity 0.5 megohm-cm
total solids 0.5 ppm
pH [25°C (77° F)] 6.0 to 8.06
dissolved N2 and Oa nondeaerated
halides
chloride 0 to 0.15 ppm Cl
fluoride 0 to 0.1 ppm F
mg/1
Chloride <0.15
Dissolved oxygen [temperature 66° C (1506F)] <0.10
In a PWR, hydrazine is employed during startup operations to ensure
that oxygen is not present to induce the stress corrosion reaction.
Oxygen reacts with hydrazine as follows:
The reaction rate is temperature dependent and is extremely rapid in the

range of 66° to 121° C (150° to 250° F) (Fig. 4). A competing decomposition
reaction also accelerates as the temperature increases. At approximately
204° C (400° F), hydrazine decomposes rapidly and chemical deoxygena-
tion of water is no longer practical (see Fig. 5). At power, oxygen can enter
the coolant system in aerated makeup water. This oxygen is immediately
scavenged by dissolved hydrogen in a reaction induced by the reactor
radiation.
PWR Water Coolant Chemistry
Water in the primary system of a PWR is employed for both cooling and
neutron moderation. It is basically high-purity water to which chemical
additions have been made to achieve specific properties. For example,
boric acid is used in the coolant of modern PWR's to control long-term
reactivity changes. Pressurized water reactors use control materials in three
different forms:
Control Reactivity Worth,
Method Purpose Material approximate %
Control power boron carbide 8.0
rods maneuvering silver-indium-
cadmium
Burnable power shaping, boron/aluminum 5.0
poison compensation
for fuel burnup
Soluble compensation for boric acid 10 to 15
poison fuel burnup,
shutdown
FIG. 4—Effect of temperature on the N2ff4;O2 reaction rate constant.
The control rods are adjusted according to the need to compensate for
short-term changes, thus maintaining the power level. The movement of
the rods is usually initiated by an automatic controller. Power maneuvering
is accomplished by control rod movement also. A certain number of the
control rods are always kept in the "full out" position to provide the means
for fast reactor shutdown if needed.
Burnable poisons are employed as fixed poison rods in the fuel assemblies.
They accomplish two purposes: (1) they compensate for fuel burnup, and
(2) they help shape the power distribution within the core. Early in core
life, when the new fuel contains more fissile material, the burnable poi-
sons absorb neutrons, thus holding the reactivity and power down. By the
end of cycle, when a large portion of the fissile material is gone due to fuel
burnup, the burnable poisions are not present. Through reactions induced
by neutron absorption, they are converted into neutral materials with small
neutron absorption cross sections. Also in the beginning of the cycle, vhen
new fuel is introduced into the core, the local variations in power density are
relatively high. Judicious placement of the burnable poison rods makes more
360 MANUAL ON WATER
FIG. 5—Effect of temperature on the decomposition of fl2^4 for a reaction time of 0.8 min
with initial ^2^4 of 0.5 ppm in carbon steel and stainless steel tubing.
uniform power generation possible. Since power output is limited by the

hottest portion of the core, uniform power distribution results in higher
power putput for the same core size.
Water reactors are characterized by large reactivity changes over a core
lifetime, including both long-term changes due to fuel burnup and shorter-
term changes associated with power maneuvering. When control rods are
used to compensate for these changes, their changing pattern of insertion
introduces substantial and time-varying distortions of the spatial power
distribution in the reactor core. Unless the control rod system is designed
and operated with great care, these distortions may correspond to highly
nonuniform power density distributions. The adoption of "chemical shim"
control (reactivity control by boric acid dissolved in the reactor water) for
PWR's has largely bypassed this problem, and so has been a source of major
improvement in power distribution. In principle, with chemical shim, there is
no need for control rods in the core during power operation, although they
must be held in the withdrawn position as a means of rapid shutdown.
In practice, however, the adjustment of the boric acid concentration—
which is done by a "feed-and-bleed" system—is cumbersome and ex-
pensive if used for short-term reactivity adjustments. The system is used
more beneficially for compensating the excess reactivity incorporated for
fuel burnup. This amounts to 10 to 15 percent reactivity for a core employing
a three-batch refueling scheme. Since the early PWR plants have been
intended for base-load operation, the other control requirements during
power operation are small, and these reactors will operate almost as "rod-
less" reactors.
The limitations of chemical shim control are quite apparent. In addition to

the problem of short—term reactivity adjustment, they include a practical
limit on total capability for reactivity compensation. The limit is imposed by
the positive component of the moderator temperature coefficient of re-
activity, which increases approximately in proportion to the boric acid
concentration, and which may cause undersirable dynamic characteristics if
it becomes large enough to outweigh the negative component and yield a net
positive coefficient. The use of chemical shim is expected to remain an im-
portant feature of the PWR technology, but the importance of control rods
is expected to increase with time. In order to control core reactivity, boric
acid is controlled much like any additive or treatment chemical in power
operations. Boric acid dissolved in the primary coolant controls only the
slowly changing reactivity effects. The mechanism by which this occurs is the
absorption of neutrons by the B'° isotope of natural boron. Natural boron
contains approximately 20 atom percent B'°, which has a very thermal
neutron cross section of approximately 3840 barns. The reaction is a follows:
1. B'° (n, o) Li7
2. B'o («, 2a) H^
The second reaction, although substantially less probable, does produce
tritium activity, which must be analyzed and accounted for in waste hand-
ling.
Reaction 1 results in Li7 production of about 0.1 ppm/day at a boron
concentration of 1070 ppm (B10 200 ppm) in a plant having rated power of
2500 M Wt. The effect of alkali metal input is to partially neutralize the boric
acid and increase the coolant pH (water quality is discussed below).
In a pressurized water reactor, water has been shown to undergo various
equilibrium reactions which are normally unexpected for the thermal con-
ditions of the coolant. These reactions generally can be summarized as
follows:
3. Water decomposition and formation
4. Ammonia synthesis and decomposition
5. Nitric acid synthesis
6. Nitric acid reduction
The kinetics of these reactions are quite rapid during reactor operation, and
many intermediate products are postulated before formation of the end
products. Since these reactions are dependent on the reactor flux, they are
first order with respect to the reactor power level and the concentrations of
reactants.
It is apparent from Reactions 3 through 6 that the presence of free hydro-
gen can shift the equilibrium of these reactions to preclude the excess
decomposition of water or formation of nitric acid. For this reason, PWR's
operate with an excess of hydrogen (added as hydrogen gas).
362 MANUAL ON WATER
Hydrogen concentration control in the current generation of PWR's is

accomplished by the chemical and volume control system (CVCS). This
system, unique to pressurized water reactors, allows direct and continuous
control of water quality in the primary coolant on a completely automatic
basis.
The presence of boric acid fixes the lower pH specification of 4.5. Boric
acid is slightly ionized at reactor operating temperatures so that it does not
greatly influence the fluid pH. Therefore, the addition of very small amounts
of strong or weak base constituents greatly influences the pH at high
temperature (Fig. 6).
There are several benefits arising from operation at adjusted pH. First, the
general corrosion of system materials is reduced, and second, core reactivity
has been reported to be affected by changes in coolant pH in large com-
mercial PWR's.
In general, the metal corrosion rates from typical PWR materials are low,
and tests have shown that they stabilize after 200 days. Since an increase in
pH reduces the corrosion rate, the coolant chemistry is adjusted to obtain
this effect.
FIG. 6—pH of various solutions (with and without boric acid) versus temperature.
PWR Operating Chemistry

The effect of radiation-induced reactions on PWR coolant chemistry is
illustrated in Fig. 7. In this figure, the coolant is high-purity light water (that
is, with no alkalinity or chemical shim additives). The pH of the water,
measured at 24° C, ranges between 7 to 8 and is due to the ammonia which
has formed by Reaction 2 of the foregoing. Hydrogen gas has been added to
the coolant. At Shippingport and other power reactors, hydrogen losses
have been shown to be due to various system water losses and pressurizer
vent valve steam phase leakage.
PWR plants generally use nondeaerated, but otherwise high-purity,
makeup water. This introduces some oxygen and nitrogen into the reactor
coolant. The dissolved oxygen is reacted (Reaction 3 given earlier) with
excess hydrogen to form water. The nitrogen and hydrogen react to form
ammonia (Reaction 4). The actual concentration of ammonia due to
synthesis is determined by the reaction rate and rate of removal of ammonia
by ion exchange resins which continuously purify a portion of the coolant.
The effect of these radiation-induced reactions on coolant chemistry when
air (oxygen and nitrogen) is present in excess of the dissolved hydrogen
FIG. 7—Radiation-induced changes in high-purity, light-water PWR coolant.

364 MANUAL ON WATER
concentration is shown graphically in Fig. 8. The reactor power was

increased slowly during the test. Ammonia, which was present in the coolant
from the previous operation, was neutralized as nitric acid was formed. As
the nitrate ion concentration increased, the coolant became acidic, and
chromate (Cr+*) ions were detected.
The concentration of chromate ions increased during the period, indica-
ting chemical attack on the system. After hydrogen was added, the reactions
were reversed—that is, the nitrate and chromate ion concentrations de-
creased—and the coolant pH increased to approximately the initial con-
dition.
Reactivity changes have been reported to be caused apparently by changes
in coolant pH. These effects have been reported in Saxton and other nuclear
power stations. The magnitude of change has been between 0.2 and 0.9
percent &KJK. This result is of interest since it could stretch core operating
lifetime by about 200 megawatt days/metric ton. The reactivity insertions
experienced have not been abrupt, requiring about 6 to 24 h for completion.
Figure 9 shows a typical reactivity response measured at Saxton. In this test
sequence, a cation exchanger was placed into service to remove lithium
FIG. 8—Plant start-up with air in loop (low power).

FIG. 9—Change of reactivity with coolant pH at Saxton—20 MWt—no boron.
(approximately 0.9 to <0.01 ppm), which had been used to maintain a high
pH condition in the coolant. No boron was present during this test. Coolant
average temperature began to decrease coincidentally with lithium removal
(pH decrease). After the pH stabilized at approximately 5.75 (low pH), T
average continued to decrease for 24 h. Following this operation, lithium
hydroxide was added, and coolant T average began to rise and returned to
the predicted burnup rate after 24 h.
Figure It) shows the correlation of experimental data from field studies at
Yankee. The data points represent operation over a wide variety of boric
acid concentrations and include essentially zero boron.
Early speculations on the reason for pH-induced reactivity swings were
that neutron poisons from either defective silver-indium-cadmium control
rods or soluble boric acid were removed from core surfaces as a result of
alkali additions to the coolant. Later observations indicated that (1) the
magnitude of the effect was independent of boric acid concentration; (2)
there was no evidence of the appearance or disappearance of neutron
poisons during the reactivity change; (3) there was no effect at zero power;
and (4) it was independent of the choice of alkali but related by magnitude
and direction of pH change.
Tests now indicate that a more satisfactory hypothesis would involve the
effect of coolant chemistry on the structure or distribution, or both, of
corrosion product films on clad fuel surfaces. Changes in the properties of
these deposits affect (1) heat-transfer characteristics, which in turn affect fuel
temperature and, hence, reactivity due to the fuel Doppler effect; (2) the
extent and nature of local boiling, thereby changing the local and core
average voids, which would affect the void coefficient of reactivity; and (3)
366 MANUAL ON WATER
FIG. 10—pH coefficient of reactivity versus coolant pH (277° Q (530° F).
system hydraulic properties with associated changes in flow, that is, core AP,
and temperature.
BWR Water Chemistry

The following two aspects of BWR chemistry make effects that occur in
both the feedwater and primary coolant systems unique relative to fossil and
PWR systems and significantly affect corrosion product chemistry:
1. No chemical additions are made to the feedwater or primary coolant to
control pH or oxygen concentration.
2. The formation of hydrogen and oxygen, which occurs in the primary
system because of the radiolytic decomposition of water, has a significant
effect on the behavior of materials in the primary system and parts of me
feedwater system.
The use of chemicals for coolant chemistry control raised significant
questions regarding radiolytic decomposition of these chemicals, which
would occur in the primary system and could lead to significant radioactivity
and corrosion problems. For these reasons, and also because of the
successful operation of the BWR in the absence of chemical additives, no
consideration is given to changing feedwater or primary system chemistry by
the addition of chemicals normally used in fossil or PWR systems.
The radiolytic decomposition of water in the primary coolant, which
generates 0.0014 to 0.0028 standard cubic centimetres per second/megawatt
thermal (0.03 to 0.06 standard cubic feet per minute/megawatt thermal) of
the stoichiometric mixture of hydrogen and oxygen, controls the oxygen and
hydrogen concentrations in the reactor water, steam, and turbine and

feedwater heater drains. Oxygen concentration in the primary coolant varies
from 100 to 300 ppb. Oxygen concentration in the steam, which approxi-
mately corresponds to that predicted by steam-water equilibrium in the
primary system, varies from 10 to 30 ppb. Turbine and feedwater heater
drains vary in oxygen concentration both with the rate of radiolytic
decomposition in the primary coolant and also with temperature and
pressure. Feedwater oxygen concentration varies from 10 to 30 ppb,
depending on the effectiveness of condensate deaeration.
To serve as a basis for comparison with the plant operating data presented
in the following, the General Electric BWR water quality requirements for
feedwater and primary coolant are presented in Table 4. The limits in the
table define the quality of the reactor water and feedwater for plant design
and operational control when operating at rated pressure. Allowance for
operation outside these limits for specific time periods is made for start-up
and nonroutine occurrences.
BWR Feedwater Chemistry

In a typical BWR feedwater system, as shown in Fig. 2, all heater drains
are cascaded to the hot well for treatment by the full-flow condensate
deepbed demineralizer or filter-demineralizer. Corrosion products in the
condensate result from corrosion or erosion of steam and turbine piping,
turbine components, extraction steam piping, and feedwater heater shell,
tubing, and drain lines. With the exception of a few plants still operating
with copper-alloy feedwater heater tubing, the corrosion products resulting
in the condensate from the aforementioned sources are limited primarily to
soluble and insoluble compounds of iron. Low concentrations of corrosion
products also result from the corrosion of the condenser tubing, which may
be either a copper alloy or stainless steel. Upon passing through the full-flow
condensate treatment equipment, ionic species are almost quantitively
removed; insoluble species are removed with a variable efficiency, depending
on their chemical and physical nature.
TABLE 4—BWR water quality specifications.

Feedwater Primary Coolant
Corrosion product concentration, ppb
total <30 NS"
copper < 2 NS*
Conductivity, nmho/cmQ250C < 0.1 <1.0
PH 6.5 to 7.5 5.6 to 8.6
Oxygen Concentration, ppb <14/> NS°
Chloride Concentration, ppb **< <200
NS = not specified.
/>
<200 ppb oxygen in forward pumped plants.
'Sufficiently low to meet primary coolant specification.
368 MANUAL ON WATER
Coolant Purification Systems

Both BWR and PWR power plants use sidestream purification systems to
remove impurities from reactor coolant. In a PWR, the system serves several
functions.
The Chemical and Volume Control System (CVCS) (Fig. 11) is essentially
a sidestream loop on the primary coolant system (PCS). It permits con-
tinuous bleeding and feeding of water from the PCS, through purifying ion
exchangers and filters, and into a gas-pressurized water reservoir called the
volume control tank. Purified water from this tank is charged back into the
reactor at the same rate as the letdown flow, thereby effecting reactor
coolant volume stabilization. The simplicity of the bleed-and-feed process
permits the addition of concentrated boric acid to increase boron con-
centrations or the replacement of the reactor coolant with makeup water free
of boric acid to reduce boron concentrations. The hydrogen concentration in
the coolant is controlled by maintaining a hydrogen overpressure on the
volume control tank.
Hydrogen concentration control in the current generation of PWR's is
accomplished by the CVCS. This system, unique to pressurized water
reactors, allows direct and continuous control of water quality in the
primary coolant on a completely automatic basis.
Auxiliary Cooling Water

As in any power plant, a closed circulating cooling wajter system is
employed to remove heat from components and equipment. In nuclear
plants, this system is used to cool control rod drive mechanisms, waste gas
compression aftercoolers, coolant letdown system, nonregenerative heat
exchangers, etc.
In general, these systems employ high-purity makeup water and use the
standard chromate corrosion inhibitor treatment. Variations in this ap-
proach are used, but, to date, the majority use the chromate treatment
shown in Table 5. While the discharge of chromates to the environment is
strictly limited, leakage from these systems is extremely small and has not
been considered a problem. The alternative treatments, using organic
compound inhibitor systems or simple pH adjustment, are in some limited
use and are considered effective.
Activation of the metal ion additives in these systems has not been
experienced, since the neutron flux gamma dosages are small and residence
time in even these low flux levels averages only about a few seconds.
Radioactivity
Nuclear steam supply systems employing pressurized water reactors are
indirect-cycle systems. The primary coolant, which picks up radioactivity
FIG. 11—Chemical and volume control system flow diagram.
370 MANUAL ON WATER
TABLE 5—Operating chemistry.

BWR PWR
Chlorides 1.0 ppm chlorides 0. 1 5 ppm
pH 9.0 to 9.7 ppm fluorides 0.1 ppm
Chromate PH 8.0 to 8.5 ppm
(as CrO4) 500 ppm chromate 175 to 500 ppm
from a number of sources, is separate from the water which serves as

working fluid for the steam turbine. The steam generator serves as a barrier
between the two fluids.
Because of its closed-cycle design, control of radioactivity in a PWR is
easily accomplished. Holdup in the system provides for decay of short-lived
radioisotopes. Other fission products activities are continuously and ef-
ficiently removed from the water by the ion exchangers in the CVCS. BWR
power plants are not of the closed-cycle design; however, off-gas systems are
designed to reduce radioactive discharges to allowable, safe levels.
Fission Products
Fission products from two basic sources may be present in the reactor
coolant. The first is uranium contamination on the surface or in the metal
matrix of fuel cladding and core structural materials. This is essentially the
background or base activity and has been present in all reactors. A second
source of fission products arises when the cladding, which encapsulates the
fuel, is penetrated or has minor leakage. This condition has been observed in
reactors and does not impede safe continued operation of the plant.
The system design provides for operation of the core with a significant
fraction of the fuel elements being defective. Routine monitoring of fission
products provides the operator with trends in coolant activity and a history
of the relationship between various activity species. The information on
activity and ratios of activity of specific nuclides is useful for assessing
changes in the source of fission products.
One method widely used for routine monitoring of fission products is the
radiochemical analysis of 1-131 and 1-133. The relationship between these
isotopes is a function of the release mechanism; that is, recoil from surface or
a cladding interstitial uranium atom produces a different ratio than diffusion
from a defective fuel rod. Both mechanisms may be operating and releasing
fission products to the coolant simultaneously. However, the ratio between
the two iodine isotopes would reveal which release mechanism was pre-
dominant.
If fuel cladding failure is suspected, confirmation can be obtained by
observing the behavior of short half-life fission products during start-up
power transients. This effect, called "water-logging," is the result of water
entering the defective fuel rod during shutdown, and ejection of that water
and its contained activity upon resumption of power operation.
This mechanism introduces high concentrations of short-lived fission

products into the coolant at a rate much faster than would diffusion from the
defected area. This diffusion mechanism would require a comparatively long
time, and the short-lived fission products would be decayed. Figure 12 shows
the peaking of 1-131 at Connecticut Yankee during a start-up with defective
fuel cladding.
Activation of Water
The isotopes N-13, N-16, N-17, and F-18 result from neutron activation of
oxygen isotopes occurring naturally in the light-water moderator. The
activity levels are a function of the reactor flux and system volume:
Nuclide Half-Life Reaction Target Source

N-16 7.13 s 0'6(w,p)N-16 water
N-17 4.14 s 017 (w,p)N-17 water
F-18 109.8 min 018 (p,n)F-18 water
N-13 9.99 min O16 (p,«)N-13 water
FIG. 12—Connecticut Yankee reactor coolant 1-131 concentrator versus power level, MWe.
372 MANUAL ON WATER
The level of N-16 activity determines the secondary shielding require-

ments for the reactor plant. This isotope (T\ 2 - 7.13 s) emits beta particles
having maximum energy of 3.32, 4.39, and 10.49 MeV, and gamma rays
of 6.18 and 7.1 MeV. The isotope N-17, which emits delayed neutrons,
must also be considered when designing plant shielding and should be
considered in the design of sample line delay times (transit time from reactor
to sample point) to protect personnel during sampling operation.
A typical decay curve of radioactive reactor coolant contaminant activities
in a PWR is shown in Fig. 13. The gamma activity, after a degassed sample
has decayed about 15 min, is due principally to N-13 (Ti/2 = 9.99 min) and
F-18 (7*,;2 = 109.8 min). The samples were taken with the reactor opera-
ting at steady-state full power. Normally, the fission product contribution
to gross gamma activity is small.
FIG. 13—Decay of degassed reactor coolant (1 in. = 2.54 cm).

Table 6 lists the major nonvolatile activities present is the Shippingport

plant reactor coolant and gives the approximate percent of total gamma
activity one hour after sampling. The activity is diminished very rapidly by
a factor of 10 000 in a few hours until a long half-life residual of corrosion
product activity remains. Volatile activities are the noble gas fission products
Xenon and Krypton and Argon-41. Argon-40 enters the system with air
dissolved in makeup water and becomes activated to Argon-41.
TABLE 6—Shippingport Atomic Power Station nonvolatile gamma activity. Core I Seed 3.
Percent of Total
Activity at Sampling Time. Gamma Radiation
Nuclide Half-Life dpm/ml" juc/ml /> At 1 h
N" 9.99 min 1.82X 104 8.19X 10"' 0.3
W"7 24 h 2.22 X 10' 9.00 X 10 4 2.9
Na :4
15 h I.92X 10" 8.64 X 10~4 1.8
p,s
109.8 h 2.8 X 104 I.26X 10": 21.3
Cs"s 32 min 1.03X 104 4.68 X 10~- 3.1
Mn 56 2.58 h 5.65 X 10' 2.54 X 10~- 4.5
Mn M 313davs 3.43 X I0: 1.09X 10"4 0.2
CuM 12.8 h ' 6.4 X 104 2.88 X 10": 39.2
I1JI 8 days 4.2 X 10' I.89X 10~5
I1" 21 h 1.03X 10' 4.67 X 10"4 14.8
I"4 52 min l.OOX 104 4.5 X 10"'
,,,5
6.7 h 1.92X 10: 8.64 X 10
Percent of gross gamma activity at I h accounted for 88.1

" dpm = disintegrations per minute.
h
/uc = microcuries.
Activation of Impurities
The concentration of impurities in reactor coolant is generally kept
extremely low by the strict control of makeup water quality, and by internal
purification. However, the small quantities of impurities present do become
activated and these can add to the system activities. The two most common
isotopes are:
Nuclide Half-Life Nuclear Reaction Target Source
Na24 15 h Na23 (n, y)Na24 impurity in water
Ar41 1.83 h Ar40 (n, y)Ar** air in water
Activation of Coolant Additives

The major chemical constituent in a chemical shim coolant is boric acid
for reactivity control. The nuclear reaction which provides for control
produces a stable isotope Li7. A second reaction produces tritium:
Nuclide Half-Life Reaction Source of Nuclide

H3 12.33 years B10 (n,2a) H3 boron in coolant
374 MANUAL ON WATER
The chemical additives employed to increase the pH of the primary coolant

(LiOH, KOH, NH4OH) are used in very low concentration (10-4M) so
that their presence is not a major factor, even if activated. K41 is the only
isotope which is easily activated:
Potassium hydroxide K4l(n, y)K42Tl/6 = 12.4 h
Those plants which employ lithium hydroxide use lithium with the Li7
isotope only to avoid the generation of tritium from the Li6 isotope:
Lithium hydroxide Li6 (n, a)H3TI/6 = 12.33 years
Corrosion Product Activation

Some of the corrosion products generated by the primary system mater-
ials are released to the coolant and are transported to the core surfaces.
This causes the metallic constituents to become activated, resulting in
radioactive oxide deposits. Formation of corrosion products in a PWR
results in a concentration of about 0.050 mg/litre in the circulating coolant.
Generally, the corrosion products correspond in composition to the
system materials. In a modern PWR, the available surface areas in contact
with reactor coolant and their respective corrosion rates are as follows:
Corrosion Rate Area, Annual Metal
Material mg/D2-month m2 (ft2) % Release, g/year
Stainless 2 2 340 (26 000) 10 5 860
steel 4 13 860 (154 000) 60 68 500
Inconel-600 small 4 860 (54 OOP) 30 small
Zircaloy
21 060 (234 000) 100
Approximately 74 kg (165 Ib) per year of metal or 92 kg (205 Ib) of metal

oxide are released.
The corrosion products are a function of system materials, and in the
current generation of PWR's they are metal oxides having the following
approximate composition:
Element Weight Percent
Fe Iron 10
Chromium 30
Nickel 60
These corrosion and wear products are released from the surface of the plant
piping and components and transported by the circulating coolant. A small
percentage of these oxides is removed from the circulating coolant by the ion
exchanger in the CVCS. The metal oxides we are concerned with are those
which are deposited on the core where they become activated by core
radiation, the degree of activation depending upon the residence time on the
core. Initially, on a clean core surface, some of the corrosion products which
strike the surface adhere to the zirconium oxide surface. As more oxide
particles strike the surface, the deposit begins to build up. This continues
unless the shear forces on the oxide particles, due to the coolant flowing past
the particles, equal or exceed the forces causing the corrosion products to
adhere to the surface. Therefore, the erodable or loose layer will be limited
by the localized increase in coolant velocity through the core, the amount of
corrosion products released to the coolant being a function of the change in
velocity.
These activated metal oxides are then transported by the circulating
coolant. Again, a small percentage will be removed by the CVCS ion
exchanger, but some will redeposit on the plant surfaces as well as on the
core. When they deposit on the plant surfaces, a radiation level increase on
these surfaces will result and will be reflected in the radiation levels of the
primary system. Subsequent operation will result in release of deposits back
to the coolant which can again redeposit on the core to become further
activitated. The nuclides which have the greatest contribution to these
radiation levels are as follows:
Nuclide Half-Life Nuclear Reaction Target Source
51 50 51
Cr 27.8 days Cr («, 7) Cr steel and Inconel
Mn54 313 days Fe54 (n, p) Mn54 steel and Inconel
Fe59 45 days Fe58 (n, y) Fe59 steel and Inconel
Co58 71 days Ni58 (n, p) Co58 steel and Inconel
Co60 5.26 years Co59 (n, 7) Co60 stellite, steel
Cu64 12.8 h Cu63 (n, y) Cu64 17-4 PH steel
Zr95 65 days Zr94 (n, y) Zr95 Zircaloy
W181 130 days W180 (n, 7) W181 steel
W187 24 h W1M> (n, 7) W187 steel
The deposits of irradiated corrosion products found on components have

not generally been a serious impairment to equipment maintenance. It does
require that maintenance be planned and that contamination and personnel
exposure be considered in this planning.
PWR Steam Generator Chemistry

Water chemistry of the working fluid of a PWR secondary cycle is similar
to that employed in conventional moderate pressure boiler operations of
4826 to 6895 kPa (700 to 1000 psi). PWR systems use two types of steam
generators: (1) the saturated-steam U-tube design, and (2) the straight tube
generator which produces superheated steam. Cross sectional views of these
generators are shown in Fig. 14.
376 MANUAL ON WATER
FIG. 14—5/eom generatorfor an 800-M We PWR nuclear steam supply system (two required)
(1 ft- 30.48 cm; 1 in. = 2.54 'cm).
Water conditions for the two types of units reflect necessary differences in
water conditioning and auxiliary water treatment systems and equipment to
ensure efficient long-term operation.
One of the unique conditions of PWR operation is that reactor coolant at
a higher pressure than the secondary water may leak into the boiler if tubing
integrity is breached. This can cause the introduction of primary system fluid
containing boric acid and radioactive materials into the steam-condensate
cycle. The special considerations concerning this situation are discussed
later.
U-Tube Steam Generators

The U-tube steam generator is very similar in operating characteristics to a
conventional, fossil-fueled, moderate-pressure, saturated-steam generator.
The units operate at approximately 268 to 288° C (515° to 550° F) and
approximately 5344 to 6895 kPa (775 to 1000 psi) throttle pressure. Water
conditioning specifications and auxiliary treatment equipment are also
similar to drum-type boilers operating at equivalent thermal conditions.
The unit pressure vessel is carbon steel and the boiler tubing is 19-mm (3/4
in.) diameter iron-chromium-nickel (alloy 600) having 40 to 50 mil wall
thickness. Chemistry adjustment to the boiler and feedwater is done (1) to
minimize corrosion of materials of construction, and (2) to ensure long-term
tubing integrity. The selection of additives and control limits recognizes that
the normal running, transient, and upset conditions may occur during plant
operation and that may result in potentially deleterious conditions to the
system.
Water conditions typically maintained with the steam generator are listed
in Table 7. These conditions recognize that occasional deviations from
normal may occur due to incursions of raw (cooling) water from surface
condenser tube leakage.
The steam generator is protected from corrosion or fouling of heat-
transfer surfaces, and steam purity is ensured by the control of chemical
additives and blowdown. Specific corrective measures largely depend on the
quality of raw water (whether brackish or freshwater, etc.) and the size or
rate of the leak.
An operational philosophy with respect to in-leakage is outlined in Table
8 and is based on the objectives stated in the preceding paragraph.
When "solids" chemical treatment is employed, there is no need for
condensate polishing equipment, since feedwater contaminants will con-
centrate in the generator or react with phosphates to form precipitates which
can be removed by blowdown.
Feedwater
The steam generator acts as a concentrating device during operation.
Soluble feedwater impurities which enter the steam generator become
concentrated in the water since the steam leaving the drum does not carry off
any appreciable amount. The recommended feedwater quality values given
in Table 9 cover a number of dissolved and suspended constituents and are
based on external or corrosion sources. For example, any contribution to
TABLE 7—Steam generator chemistry limits.

Impurity Normal Abnormal
Total dissolved solids 500 ppm max 1000 ppm max
Chloride 75 ppm max 250 ppm max
Total silica 10 ppm max 10 ppm max
Suspended solids 50 ppm max 100 ppm max
Dissolved oxygen 0.01 ppm max 0.01 ppm max
Chemical Additives for

Corrosion Control
Phosphate 15 to 25 ppm PO4 15 to 40 ppm PO4
[9.8 to 10.2 pH [9.8 to 10.5 pH
at 25° C 77° F] at 25° C (77° F)]
julfite 5 toJO pjnn SO, 5 to 10 ppm SO,
TABLE 8 — Criteria for maintaining steam generator water chemistry.
Feedwater Conditions Basic Steam Geneiator
(Ingress Impurity Only) Operational Limitations Water Chemisty Chemistry Control
Total dissolved solids none TDS<500 ppm normal
<0.5 ppm (normal) pH 9.8 to 10.2
PO4 15 to 25 ppm
Cl 75 ppm max
Total dissolved solids LIMITED TDS<1000 ppm immediately start chemical injection pumps to
0.5 - 2.0 ppm schedule inspection and pH 9.8 to 10.5 maintain required phosphate and pH conditions.
(abnormal) repair of condenser as PO4 15 to 40 ppm Increase blowdown rate to limit total dissolved
soon as system load re- Cl 250 ppm max solids concentration.
quirements permit.
Total dissolved solids EMERGENCY TDSOOOO ppm immediately start chemical injection pumps to main-
<2.0 ppm (excessive) (a) immediately reduce load pH 9.8 to 10.5 tain required phosphate and pH conditions. Do
as necessary to permit isola- PO4 15 to 40 ppm not continue to operate system if pH cannot be
tion of damaged conden- Cl 250 ppm max maintained above 8.0, or total dissolved solids
ser section. below 1000 ppm.
(6) prepare for orderly shut- if system goes into an orderly shutdown operation,
down of the system if feed- prepare to establish chemical conditions for wet
water TDS concentration layup of steam generator and auxiliary equipment
cannot be quickly reduced during condenser repair.
below 2.0 ppm.
feedwater impurity concentration due to steam generator carry-over is
allowed in addition to the contribution from an external source. Thus, if 0.2
percent moisture carry-over occurs with 500 ppm dissolved solids in the
steam generator, then 1 ppm solids would be present in the feedwater due to
carry-over, and the limit on total dissolved solids (TDS) for normal
operation would then be 1.5 ppm instead of 0.5 ppm.
The TDS limit is specified to minimize steam generator fouling. Since
TDS can, in general, be measured quickly by conductivity instrumentation,
this category of impurities serves as a primary indicator of cycle contamina-
tion. The normal chloride concentration limit assures that long-term opera-
tion will not result in chloride-induced corrosion problems within the
secondary cycle. The silica concentration limit assures that silica carry-over
to the turbine by the vapor or steam solubility phenomena does not occur to
any appreciable extent with respect to turbine fouling. The oxygen and pH
specifications should limit the pickup of iron and copper from the feedwater
cycle to those levels necessary for long-term steam generator cleanliness.
Secondary cycle operation is considered normal as long as feedwater TDS
due to external sources is less than 0.5 ppm. Coupled with this is the
requirement that pH value, oxygen level, and copper/ iron levels are within
the values given in Table 9.
Operation with feedwater TDS concentration between 0.5 and 2.0 ppm
should be limited, and condenser repairs should be scheduled as soon as
practical. Values of feedwater solids concentration due to external source
contribution above 2 ppm are considered emergency levels. The unit load
should be immediately reduced, the damaged condenser section isolated, and
necessary repair work scheduled. The steam generator-turbine cycle should
be shut down if feedwater concentration cannot be quickly reduced to values
below 2 ppm.
TABLE 9 — Feedwater chemistry limits.

Impurity Normal Abnormal
Total dissolved solids 0.5 ppm max 0.5 to 2.0 ppm
(excessive is
greater than
2.0 ppm)
Chloride 0.3 ppm max 1.2 ppm max
Total silica 0.02 ppm max 0.02 ppm max
Copper and total iron 0.05 ppm max 0.05 ppm max
Dissolved oxygen 0.01 ppm max 0.01 ppm max
Chemical Additives for

Corrosion Control
Feedwater pH [25° C (77° F)]
(a) Feedwater system
containing copper
alloys of construction 8.8 to 9.2 8.8 to 9.2
(b) Copper free feedwater 8.8 to 9.2
system 9.2 to 9.4 9.2 to 9.4
Hydrazine 0-02 ppm max 0.02 ppm max
380 MANUAL ON WATER
Once-Through Steam Generator (OTSG)

The water quality specifications for the secondary system feedwater are
given in Table 10. These specifications utilized conventional superheat
experience. The OTSG requires high-purity water quality, and the steam
plant contains condensate polishing equipment to maintain this high water
quality. Several potential water chemistry problems were investigated to
establish water chemistry specifications. They included the following:
1. The behavior of dissolved and suspended solids in the pressure/-
temperature region of interest.
2. The effects of boric acid leaks in the steam and condensate cycle.
3. The removal of deposits.
Dissolved and Suspended Solids

Because of the lower operating pressure and temperature involved in the
nuclear systems [6378 kPa and 299° C (925 psia and 570° F) full load],
extension of data relating to the behavior of dissolved salts in this new
environment was required. These data were obtained with a test boiler at the
B&W Alliance Research Center. A portion of these tests was reported
earlier. The data obtained were for the most commonly encountered
materials, and those whose past behavior had varied from the most to the
least soluble in high-pressure steam. The salts chosen were sodium chloride,
sodium hydroxide, and sodium sulfate. Careful consideration was given to
sampling procedures for superheated steam to obtain reliable data. The
sampling procedures were established by earlier research.
It was found that sodium hydroxide and sodium chloride had solubilities
as high as 75 ppb at 6378 kPa (925 psia). If these compounds are present in
the feedwater in concentrations below 75 ppb, they will dissolve completely
in the steam and pass through the steam generator. Since the turbine
TABLE 10—Secondary feedwater specification.

Total solids (dissolved and suspended) 50 ppb max
Dissolved Oxygen 7 ppb max"
Total silica (as SiOa) 20 ppm max
Total iron (as Fe) 10 ppb max
Total copper (as Cu) 2 ppb max
pH at 25° C (77° F) (adjusted with ammonia) 9.3 to 9,5
h
Total hardness .
Lead ff
"Hydrazine is used to scavenge oxygen.
''No specification is listed owing to control analysis limitations. However, care should be
taken to eliminate hardness constituents due to possible steam generator deposition problems.
'Lead contamination of the feedwater should be avoided in view of reported problems with
nickel-chromium-iron alloy in oxygenated water containing lead. In specifying zero for lead,
it is the intent that the level should be kept below the lowest values detectable by our acceptable
referee methods.
operates wet after the first stage, these compounds will pass through the
turbine and have no effect on performance of the system.
Tests showed that Na2SO4 had low solubility (5 ppb); therefore, if it is
present in concentrations higher than 5 ppm, it may form deposits. However,
tests showed that soluble salts are easily removed by rinsing with feedwater-
quality water.
Primary-To-Secondary Leakage
In a PWR or closed-cycle power plant, reactor water can leak into the
boiler water or working fluid if steam generator tube integrity is breached.
Should this occur, primary water containing boric acid and radioactivity will
enter the secondary system.
U-Tube Steam Generator

The concentration levels that result within the secondary cycle will depend
on the primary coolant leakage rate, the boric acid and activity values in the
primary coolant, steam generator blowdown rate, and employment of
condensate demineralization.
The chemistry specifications outlined in Table 7 limit total dissolved solids
to 1000 ppm during abnormal operation. This limit requires steam generator
blowdown in the event of boric acid introduction. Since activity will be
present in the steam generator, the blowdown fluid must be processed in the
waste management system.
The room-temperature measurement of steam generator water pH will be
depressed upon the introduction of boric acid, so that operation within the
specified ranges given in Table 8 will not be possible. This is not adverse
during operation, because boric acid does not significantly depress the pH of
the higher-temperature steam generator water. As a control limit, however,
the room-temperature pH measurement during operation with boric acid
contamination must be maintained above 8.0. The upper limit of phosphate
additive given in Table 7 for abnormal operation is 40 ppm. This concentra-
tion is sufficient to buffer the maximum amount of boric acid permissible
within the 1000-ppm TDS limitation.
The volatile nature of boric acid results in carry-over of boric acid with the
steam to the condenser hot well. The amount of carry-over has been
determined to be about 8 percent for the secondary operating conditions of a
U-tube steam generator. The boric acid volatility (or solubility) in the steam
is dependent only on the steam pressure-temperature value. For every 1000
ppm of boric acid in the steam generator water, it is expected, therefore, that
the steam will contain about 8 ppm.
Deposition of boric acid on turbine stages has not presented a problem in
any operating plant to date. Some light deposition of boric acid on the last
382 MANUAL ON WATER
stages of low-pressure turbines has been observed, but these deposits have
not affected turbine capability.
The feedwater will become contaminated with boric acid because of the
volatility phenomenon. For instance, about 60 ppm of boric acid will be
expected to be feedwater with 800 ppm boric acid in the generator, if
condensate demineralization is not utilized. Feedwater impurity limits must,
therefore, be raised to accomodate the presence of boric acid. Because of the
lower amount of boric acid in this part of the cycle, a maximum level of 1.0
ppm ammonia should be utilized to maintain feedwater pH values above 8.0.
Additionally, since ammonia is volatile, all portions of the feedwater system,
including those sections that bypass the normal ammonia injection points,
will equilibriate to the same pH level.
The carry-over of the biologically significant radioactive iodine has been
shown to be influenced by boiler water chemistry. When coordinated
phosphate/sulfite treatment has been used, no vaporization beyond normal
entrainment in moisture has been observed.
The benefit derived is a significantly lower release of iodine from the
secondary system off-gas. This is because the only form of iodine that is
volatile is the elemental (I2) form. In the presence of reducing sulfites in the
steam generator, any iodine introduced via primary-to-secondary leakage
will react to form nonvolatile iodides:
The alkaline pH as a result of the phosphates in the steam generator results

in hydrolysis of elemental iodine to nonvolatile iodides and iodates:
These reactions prevent the escape of iodine activity from the steam
generator as volatile iodine vapors in the steam, and limit iodine activity
release to that associated with moisture carry-over.
Once-Through Steam Generator

The special considerations of superheat operations have not, at the time of
this writing, been observed in plant scale operations. However, extensive
testing has been done and is presented here.
In the test boiler, a simulation was made of a reactor coolant to secondary
system leak. One accelerated test was equivalent to a 0.0009 m 3 /s (15
gal/min) leak on a full-size unit at hot operating conditions. The simulated
reactor coolant in that test contained 13 000 ppm H3BO3. The test boiler
condensate was not polished during the test period. Boric acid was very
volatile (^1800 ppm H3BO3) at these pressure-temperature conditions [6378
kPa, 299° C (925 psia, 570° F)] and did not concentrate in the OTSG.
The amount of boric acid that would be present in the secondary system
feedwater in a full-size unit depends mainly on the leak rate, the concentra-
tion of H3BO3 in the reactor coolant, and the HjBO3 removel efficiency of
the condensate demineralizers. We have made extensive calculations, as-
suming leak rates from 0.000003 to 0.003 m 3 /s (0.05 to 50.0 gal/min), reactor
coolant H3BO3 concentrations from 0.006 to 0.82 m 3 /s (100 to 13 000
gal/min, and 0, 50, and 100 percent removal of H3BO3 by the condensate
polishing demineralizers. Table 11 lists some of these data.
The feedwater reference pH at 25° C (77° F) decreases with increasing
boric acid in the secondary system. However, boric acid at the concentra-
tions studied has little effect on the pH at elevated temperatures. In most
cases, the pH [25° C (77° F)] does not decrease below 9.4 [25° C (77° F)] as
long as the demineralizer is removing at least 50 percent of the H3BO3
leaking into the secondary system. Also, in some cases, the addition of a
small amount of ammonia will compensate for a pH [25°C (77°F)] decrease
below 9.3.
Earlier research on the behavior of both high-flow-rate deep-bed and
Powdex condensate polishers was extended to study their behavior in the
event of a boric acid leak. Table 12 shows the ultimate capacities obtained in
laboratory tests using 1.0, 10.0, or 100.0 ppm H3BO3 and —1.5 ppm NH 3
[equivalent to a pH (room temperature) of 9.5 at normal conditions] in the
influent. The results show that both types of systems can remove low
concentrations of boric acid that would be present in the secondary system if
a tube leak occurred. The presence of boric acid does not significantly affect
the capability of deep-bed demineralizers for removing sodium and chloride.
The condensate polishing system will continue to protect the OTSG and
turbine from condenser leak contaminants during a simultaneous OTSG
tube leak.
Summary
In this brief review, we have attempted to present the salient features of
the water chemistry control required for reliable operation of water reactors.
The 20 years of development and more than 10 years of operational
experience with the control principles described in this paper have led to an
accumulation of the technology which supports reliable plant operations.
Bibliography
Allen, A. O., The Radiation Chemistry of Water and Aqueous Solutions, Van Nostrand, New
York, 1961.
Bartnoff, S., Weisman, J., and Layman, W. H., Transactions, American Nuclear Society,
Vol. 7, No. 2, Nov. 1964.
Bergmann, C. A., "Effect of Low Cobalt Impurity Stainless Steel on Coolant Activity of
PWR's," American Chemical Society, April 1964.
Byrnes, D. E., "Some Physicochemical .Studies of Boric Acid Solutions at High Tempera-
tures," WCAP-3713, Sept. 1962.
TABLE 11—Reactor coolant to secondary-system boric acid leak calculations?
Reactor
Leak Coolant pH@ pH@ pH@ pH@ pH@ pH@
Rate, H3B03, H3B03fc 25° C 140°C 188°C 237° C 278° C 304° C NH^
f
gal/min* ppm ppm (77°F)r (285° F)< (370° F) (160°F)< (532° F)< (579° F)< ppm
0.05 13000 0.08 9.5 7.1 6.61 6.26 6.09 6.06 0
0.05 100 0.0006 9.5 7.1 6.61 6.26 6.09 6.06 0
0.5 13000 0.8 9.43 7.1 6.62 6.26 6.09 6.06 0
0.5 100 0.006 9.5 7.1 6.62 6.26 6.09 6.06 0
5.0 13000 8.1 9.09 7.08 6.61 6.26 6.09 6.06 1.7
5.0 100 0.06 9.5 7.09 6.61 6.26 6.09 6.06 0
50.0 13000 81.4 8.1 7.00 6.61 6.29 6.14 6.14 24
50.0 100 0.63 9.46 7.12 6.64 6.29 6.15 6.14 0
7.0* 5.71 5.36 5.21 5.36 5.43
9.S> 7.1 6.61 6.26 6.09 6.06
"'Assuming 50 percent removal of HjBOj by condensate polishing demineralizers. Lithium is 2.0 ppm Li in reactor coolant at all times. Ammonia is 1.5
ppm NHU in condensate and feedwater. The calculated values are for equilibrium conditions using 100 percent condensate polishing and 50 percent steam
flow to the feedwater heaters.
''Boric acid concentration at the demineralizer influent, shell side of the feedwater heater, and in the steam generator.
'The pH with H3BO3 and 1.5 ppm NH 3 present.
<AThe amount of NHj necessary to increase the pH to 9.3 @25°C (77°F).
''The pH of pure water at various temperatures.
'The pH of 1.5 ppm NHj at various temperatures.
®l gal/ min = 0.00006 m3s.
TABLE 12—Condensate polishing demineralizer tests.

DEEP-BED TESTS"
Influent, ppm Effluent, ppm Ultimate Capacity
HsBOj NaCl Na Cl Ib HiBO^/ft'OH Fora Anion
1.0 0.5
10.0 2.5
100.0 4.5
1.0 <0.1 <0.001 <0.05 0.7
10.0 <0.1 <0.001 <0.05 2.4
100.0 <0.1 <0.001 <0.05 4.1
POWDEX TESTS*
Influent, ppm Ultimate Capacity
HiBOj Ib HjBOi/lb dry anion (OH)
1.0 0.03
10.0 0.14
100.0 0.24
-All tests used NH4R-ROH resin in a 2:1 ratio. Flowrate was 50 gal/min/ft 2 and temperature
25°C (75°F).
6
All tests used NH4R-ROH resin in a 1:1 ratio and dosage of 0.15 lb/ft 2 . Flowrate was 163
litres/min/m 2 (4 gal/min/ft 2 ) and temperature 25°C (75°F).
Cohen, P., Water Coolant Technology of Power Reactors, American Nuclear Society
Monograph, Gordon and Breach Science Publishers, New York, 1969.
Cohen, P. and Graves, H. W., Jr., Nucleonics, Vol. 22, No. 5, May 1964.
Connally, J., "Closed Cycle Reactor Plant Operating Experience at Yankee Rowe," National
Association of Corrosion Engineers, Houston, Tex. 1969.
Copson, H. R. and Berry, W. E., Corrosion, Vol. 16, 1960, pp. 79t-85t.
Copson, H. R. and Berry, W. E., Corrosion, Vol. 18, 1962, pp. 21t-26t.
Dickinson, N. L. et al, Proceedings, American Power Conference, Vol. 19, 1957, pp. 692-706.
Eichenberg, et al, "Effects of Irradiation and Bulk UO2," WAPD-183, Oct. 1957.
Graves, R. H., "Chemistry and Waste Management at the Conneticut Yankee Atomic Power
Plant," National Association of Corrosion Engineers, Houston^ Tex., 1969.
Hicks, J. H., et al, "Nuclear Steam Supply System Water Chemistry," B&W TPI-11,
Babcock & Wilcox.
Lechnick, W., Martucci, J. A., and Scherer, R. D., "Shippingport Operations," DLCS 36403,
Operational Chemistry, Part 5, Chapter 1, 1963.
Martucci, J. A. and Reckman, B. J., Jr., "Water Technology for PWR Nuclear Power
Plants," Nov. 1970.
Meek, M. E., "The Calculated pH of Aqueous Boric Acid Solutions as a Function of
Temperature and Added Base Content," WCAP-3269-51, March 1951.
Miller, D. A. and Bryant, P. E. C., "Corrosion and Coolant Chemistry Interactions in
Pressurized Water Reactors," National Association of Corrosion Engineers, March 1970.
Picone, L. F., "The In-Pile Test of Chemical Shim," WCAP-3729, June 1963.
Rodden, C. J., "Analysis of Essential Nuclear Reactor Materials," Division of Technical
Information, U. S. Atomic Energy Commission, 1964.
Sawochka and Pearl, "Corrosion Product Chemistry of a Boiling Water Reactor," NEDO
13129, Class I, Aug. 1970.
Stein, W., "Chemistry at the Indian Point Nuclear Power Station," National Association of
Corrosion Engineers, Houston, Tex., 1969.
Tash, J. A., Martucci, J. A., and Dunn, D. N., "Shippingport Operations," DLCS 36401,
Operational Chemistry, Part 4, Chapter 3.
Van Rooyen, D., Copson, H. R., and Berry, W. E., "Corrosion Behavior of Nickel-
386 MANUAL ON WATER
Chromium-Iron Alloy 600 in Borated Pressurized Water Reactor Environments," National

Association of Corrosion Engineers, Houston, Tex., 1969.
Ward, C. T., Mathis, D. L., and Staehle, R. W., Corrosion, Vol. 25, No. 9, 1969.
WCAP-3696-8, "Pressurized Water Reactor pH-Reactivity Effect, Final Report," Oct. 1968.
Weisman, J. and Bartnoff, S., "The Saxton Chemical Shim Experiment," WCAP-3269-24,
July 1965.
Wells, A. F., Structural Inorganic Chemistry, 3rd ed., Oxford University Press, Oxford,
England, 1962.
Whyte, D. D., "A Long Term Corrosion Test of Reactor Structural Materials in a Chemical
Shim Solution, WCAP-3737, March 1964.
Whyte, D. D. and Taylor, G. R., Corrosion and Corrosion Product Behavior in Chemical
Shim Solutions," WCAP-3725, Dec. 1964.
Chapter 18—Practices for Measurement

of Radioactivity
This chapter provides a review of the accepted counting practices cur-

rently used in radiochemical analyses. It is divided into four sections:
General Information, Alpha Counting, Beta Counting, Counting. The
General section contains information applicable to all types of radioactive
measurements while each of the other sections are specific for a particular
type of radiation.
Summary
This chapter is a compilation of the various counting techniques em-

ployed in the measurement of radioactivity [7,2].' The important variables
that effect the accuracy or percision of counting data will also be presented.
Because a wide variety of instruments and techniques is available for
radiochemical laboratories, the types of instruments and techniques to be
selected will be determined by the information desired. In a simple tracer
application using a single radioactive isotope having favorable properties
of high purity, energy, and ample activity, a simple detection will probably
be sufficient, and techniques may offer no problems other than those
related to reproducibility. The other extreme would be a laboratory re-
quiring quantitative identification of a variety of radionuclides, prepara-
tion of standards, or studies of the characteristic radiation from radio-
nuclides. For the latter, a variety of specialized instruments would be
required. Most radiochemical laboratories require a level of information
between the two extremes.
A basic requirement for accurate measurements is the use of accurate
standards for instrument calibration. With the present availability of good
standards, only the highly diverse radiochemistry laboratories require
instrumentation suitable for producing their own radioactive standards.
However, it is advisable to compare each new standard received against the
previous standard.
Thus, the typical laboratory may be equipped with proportional or
387
388 MANUAL ON WATER
Geiger-Mueller counters for beta counting, sodium iodide or germanium

detectors, or both, in conjunction with multichannel analyzers for gamma
spectrometry, and scintillation counters suitable for alpha- or beta-emitting
radionuclides.
General
Experimental Design
In order to properly design valid experimental procedures, careful con-
sideration must be given to the following:
(a) The radionuclide to be determined.
(b) The relative activity levels of interferences.
(c) The type and energy of the radiation.
(d) The original sample matrix.
(e) The required accuracy.
Having considered the foregoing, it is now possible to make the following
decisions:
(a) In what chemical or physical form must the sample be for radioassay?
(b) What chemical purification steps are necessary?
(c) What type of detector is required?
(d) Is energy spectrometry required?
(e) How long must the sample be counted in order to obtain statistically
valid data?
(/) Must isotopic composition be determined?
(g) What size sample is required?
For example, gamma-ray measurements can usually be performed with
little or no sample preparation whereas both alpha and beta counting will
always require chemical processing. If low levels of radiation are to be
determined', very large sample and complex counting equipment may be
necessary.
More detailed discussions of the problems and interferences are included
in the sections for each particular type of radiation to be measured.
Apparatus
The apparatus required for the measurement of radioactivity consists, in
general, of the detector and associated electronic equipment. The latter
usually includes a stable power supply, preamplifiers, a device to store or
display, or both, the electrical pulses generated by the detector, and one or
more devices to record information.
Some detectors and high-gain amplifiers are temperature senstive; there-
fore, changes in pulse amplitude can occur as the ambient room temperature
varies. For this reason, it is necessary to provide temperature-controlled air
conditioning in the counting room.
CHAPTER 18 ON PRACTICES FOR MEASUREMENT OF RADIOACTIVITY 389
Instrumentation should never be located in a chemical laboratory where

corrosive vapors will cause rapid deterioration and failure.
Power Supply—Detector and electronic responses are a function of the
applied voltage; therefore, it is essential that only a very stable, low-noise
electrical supply be used or that suitable stabilization be included in the
system.
Shielding—The purpose of shielding is to reduce the background count
rate of a measuremnt system. Shielding reduces background by absorbing
some of the components of cosmic radiation and some of the radiations
emitted from material in the surroundings. Ideally, the material used for
shielding should itself be free of any radioactive material that might con-
tribute to the background. In practice, this is difficult to achieve as most
construction materials contain at least some naturally radioactive species
(such as potassium-40, on members of the uranium and thorium series). The
thickness of the shielding material should be such that it will absorb most of
the soft components of cosmic radiation. That will reduce cosmic-ray back-
ground by about 25 percent. Shielding of beta- or gamma-ray detectors with
anticoincidence systems can further reduce the cosmic-ray or compton scat-
tering background for very low level counting.
Detectors have a certain background counting rate from naturally occur-
ring radionuclides and cosmic radiation from the surroundings, and from
the radioactivity in the detector itself. The background counting rate will
depend on the amounts of these types of radiation and on the sensitivity of
the detector to the radiations.
In alpha counting, low backgrounds are readily achieved since the short
range of alpha particles in most materials makes effective shielding easy.
Furthermore, alpha detectors are quite insensitive to the electromagnetic
components of cosmic and other environmental radiation.
Care of Instruments—The requirements for and advantages of operating
all counting equipment under conditions as constant and reproducible as
possible have been pointed out earlier in this section. The same philosophy
suggests the desirability of leaving all counting equipment constantly
powered. This imples leaving the line voltage on the electrical components at
all times. The advantage to be gained by this practice is the elimination of the
start-up surge voltage which causes rapid aging and the instability which
occurs during the time the instrument is coming up to temperature.
A regularly scheduled and implemented program of maintenance is
helpful, if not required, to obtain satisfactory results. The maintenance
program should include not only checking the necessary operating con-
ditions and characteristics of the components, but also regular cleaning of the
equipment.
Sample and Detector Holders—In order to quantify counting data, it is
necessary that all samples be presented to the detector in the same "geo-
metry." This means that the samples and standards should be prepared for
counting in the same way so that the distance between source and detector
390 MANUAL ON WATER
remains as constant as possible. In practice, this usually means that the

detector and sample are in stationary positions. Another configuration often
used is to have the detector in a fixed position. Another configuration often
used is to have the detector in a fixed position, within the shield, and beneath
it a shelf-like arrangement for the reproducible positioning of the sample at
several distances from the detector.
Special Instrumentation
This section describes some radiation detection instruments and auxiliary
equipment that may be required for special application in the measurement
of radioactivity in water.
4-Tr Counter—The 4-?r counter is a detector designed for the measure-
ment of the absolute disintegration rate of a radioactive source by counting
the source under conditions that approach a geometry of 4-7r steradians. Its
most prevalent use is for the absolute measurement of beta emitters [3,4],
and for this purpose a gas-flow proportional counter similar to that in Fig. 1
is common. It consists of two hemispherical or cylindrical chambers whose
walls form the cathode, and with a looped-wire anode in each chamber. The
source is mounted on a thin supporting film between the two halves, and the
counts recorded in each half are summed. An argon (90 percent)-methane
(10 percent) gas mixture can be used; however, pure methane gives flatter
and longer plateaus and is preferred for the most accurate work. The
disadvantage is that considerably higher voltages, about 3000 V, rather than
the 2000 V suitable for argon-methane, are necessary. As with all gas-filled
proportional counters, very pure gas is necessary for very high detector
efficiency. The absence of electronegative gases that attach electrons is
particularly important since the negative pulse due to electrons is counted in
FIG. 1 — The 4ir-counting chamber (I in. - 2.54 cm).

this detector. Commercial Chemically Pure (CP) gases are ordinarily

satisfactory, but they should be dried for best results. A high-voltage power
supply for the detector, an amplifier, discriminator, and a sealer complete
the system.
To convert counting rate to disintegration rate, the principal corrections
required are for self-absorption in the source and for absorption in the
support film. The support film should be as thin as practicable to minimize
absorption of beta particles emitted in the downward direction. Mylar or
other polyester film with a thickness of about 0.9 mg/cm 2 is readily available
and easily handled. However, it is too thick for accurate work with the
lower-energy beta emitters. For this purpose thin films (= 5 to 10 Atg/cm2)
are prepared by spreading on water a solution of a polymer in an organic
solvent. VYNS [3], Formvar [4], and Tygon [5] plastics have been used for
this purpose.
Since it is a part of the chamber cathode, the film must be made
electrically conducting by covering it with a thin layer (2 to 5 ^tg/cm 2 ) of gold
or palladium by vacuum evaporation. The absorption loss of beta particles
in the film must be known. Published values can be used, if necessary, but
for accurate work an absorption curve using very thin absorbers should be
taken [3]. The "sandwich" method, in which the film absorption is calculated
from the decrease in counting rate that occurs when the source surface is
covered with a film of the same thickness as the backing film, is suitable for
the higher beta energies.
The source itself must be very thin and deposited uniformly on the support
to obtain negligible self-absorption. Various techniques have been used for
spreading the source, for example, the evaporation of 63Ni-dimethylglyoxime
onto the support film [3]; the addition of a trifluoroethylene resin suspension
[5], collodial silica, or insulin to the film as spreading agents; and hydrolysis
[4\. Self-absorption in the source or mount can be measured by 4-vr beta-
gamma coincidence counting [6,7]. The 4~ir beta counter is placed next to a
sodium iodide scintillation crystal, or a portion of the chamber wall is
replaced by a sodium iodide crystal, and the absolute disintegration rate is
evaluated by coincidence counting [8,9]. By adding a suitable beta-gamma
tracer, the method has been used for pure beta as well as beta-gamma
emitters [10]. Accurate standardization of pure low-energy beta emitters (for
example, 63Ni) is difficult, and the original literature should be consulted by
those inexperienced with this technique.
Photon (gamma and strong X-ray) scintillation counters with geometries
approaching 4-rr steradians can be constructed from sodium iodide (thal-
lium) crystals in either of two ways. A well crystal, that is, a cylinderical
crystal with a small axial hole, covered with a second crystal will provide
nearly 4~ir geometry for small sources, as will two solid crystals placed very
close together with a small source between them. The counts from both
crystals are summed as in the gas-flow counter. The deviation from 4-7r
geometry can be calculated from the physical dimensions. For absolute
392 MANUAL ON WATER
gamma-ray counting, the efficiency of the crystal for the gamma energy
being measured and the absorption in the crystal cover must be taken into
account.
The liquid scintillation counter is also essentially a 4-rr counter for beta
particles, since nearly all the radiations are emitted into and interact with the
detecting medium.
Low-Geometry Counters—This type of instrument is particularly useful
for the absolute counting of alpha particles. The alpha emitter, in the form of
a very thin solid source, is placed at a distance from the detector such that
only a small fraction ( < 1 percent) of the alpha particles are emitted in a
direction to enter the counter. This solid angle is obtained from the physical
measurements of the instrument. The space between the source and detector
is evacuated to eliminate the loss of alpha particles by absorption in air. The
detector can be any counter that is 100 percent efficient for all alpha particles
that enter the sensitive volume—a gas-flow proportional counter with a
window that is thin (about 1 mg/cm2) covering. The advantages of this
instrument for absolute alpha counting are (1) that the effect of absorption
of alpha particles in the source itself is kept to a minimum since only
particles that travel the minimum distance in the source enter the detector
(particles that have longer paths in the source are emitted at the wrong
angle), and (2) that back-scattered alpha particles (those emitted into the
source backing and reflected back up through the source) lose sufficient
energy so they cannot enter the detector. One such instrument is described in
Reference 12.
Internal Gas Counters—The internal gas counter, so named because the
radioactive material, in the gaseous state, is placed inside a counting
chamber and thus becomes part of the counting gas itself, is useful for high-
efficiency counting of weak beta and X-ray emitters. The radiations do not
have to penetrate a counter window or solid source before entering the
sensitive volume of a detector. The counter may be an ionization chamber,
or it may be operated in the Geiger or proportional mode. Most present-day
instruments are of the latter type, and they generally take the form of a metal
or metal-coated glass cylinder as cathode with a thin anode wire running
coaxially through it and insulated from the cylinder ends. A wire through the
wall makes electrical contact to the cathode. The counter has a tube opening
through which it may be connected to a gas-handling system for filling. The
purity of the gas is important for efficient and reproducible counting,
particularly in the proportional mode.
In a modification of the internal gas counter, scintillation counting has
been used in place of gas-ionization counting. The inner walls of the
chamber are coated with a scintillation material and the radioactive gas
introduced. An optical window is made a part of the chamber, and the
counting done by placing this window on a multiplier phototube to detect
the scintillations. This system is particularly useful for counting radon gas
with zinc sulfide as the scintillator. Additional details on internal gas
counting may be found in Reference [13].
Spectrometers and Energy-Dependent Detectors—The availability of

energy-dependent detectors (whose output signal is proportional to the
energy of the radiation detected) that are easy to operate and maintain and
have good resolution makes it possible to measure not only the total activity
of a radioactive sample but the energy spectrum of the nuclear radiations
emitted. Nuclear spectrometry is most useful for alpha particles, electro-
magnetic radiation (gamma and X-rays), and conversion electrons, since
these radiations are emitted with discrete energies. Beta spectra have more
limited use since beta particles are emitted from a nucleus with a continuous
energy distribution up to a characteristic maximum (E-max) making a
spectrum containing several different beta emitters difficult to resolve into its
components. The advantages of spectrometric over total activity measure-
ments of radioactive sources are: increased selectivity, sensitivity, and
accuracy, because nuclide identification is more certain; interference from
other radioactive nuclides in the sample is diminished or eliminated, and
counter backgrounds are reduced since only a small portion of the total
energy region is used for each radiation. The detectors for alpha spectra are
gridded ion-chambers and silicon semiconductor detectors. These are de-
scribed in ASTM Recommended Practice for Alpha Spectrometry (D 3084)
[7]. Gridded ion-chambers are no longer available commercially and must be
constructed by the user. A variety of semiconductors can be purchased, and
these detectors have essentially replaced ion-chambers for alpha spec-
trometry, although the chambers have the advantage of high efficiency
(nearly 50 percent) for large area sources.
The principal detectors used for gamma-ray spectrometry are thallium-
activated sodium iodide scintillation crystals, Nal (Tl), and lithium-drifted
germanium semiconductors, Ge(Li). For X-rays very-low-energy gamma
rays, lithium-drifted silicon semiconductors, Si(Li), intrinsic germanium, and
gas-filled thin-window (about 1 mg/cm2) proportional counters are used.
Sodium iodide is hygroscopic, so the crystal must be hermetically sealed,
and the entire crystal-phototube package must be light-tight. The complete
spectrometer also requires a high voltage power supply for the phototube
(usually operated at 800 to 1000 V), a preamplifier, linear amplifier, pulse-
height analyzer, and output recorder. The crystal is packaged in aluminum
or stainless steel. The portion of the cover through which gamma-rays enter
is normally thinner than the rest of the package in order to reduce low-
energy photon attenuation. Sodium iodide crystals are available in a large
range of sizes and shapes, from 25- by 25- mm cylinders to hemispheres and
cylinders at least 305 mm in diameter. Information on the types of crystal
packages and mountings that can be used is available from the manu-
facturers.
Germanium and silicon detectors are junction-type semiconductor devices
(see ASTM Test for Gamma Spectrometry of Water (D 2459) [/]) in which a
large sensitive region has been produced by drifting metallic lithium into
germanium or silicon under the influence of an electric field at an elevated
temperature (100 to 400° C). The crystal functions as a "solid ion chamber"
394 MANUAL ON WATER
when a high voltage is applied. Because of the high mobility of the small
lithium atom, the detector must be kept cold to prevent the diffusion of
lithium out of the crystal. In addition, in order to obtain high resolution, the
detector must be operated at low temperatures to reduce thermal noise. At
room temperature, sufficient free electrons will be present in the crystal to
obscure the measurement of gamma and X-rays (but not of alpha particles).
Consequently, the detectors are operated and kept at liquid nitrogen
temperatures by a cryostat consisting of a metallic cold-finger immersed in a
Dewar flask containing liquid nitrogen. If the detector is allowed to warm to
room temperature for a short time, its resolution will deteriorate and in an
hour or so it will lose sufficient lithium so it cannot function as a detector
without redrifting. The detector is kept hermetically sealed in a vacuum to
prevent impurities from condensing on the surface and lowering its resis-
tance, and to reduce beat transfer from the room to the crystal. Aluminum is
the usual covering, and a molecular sieve or vac-ion pump is incorporated
into the system to maintain the vacuum. The electronic components required
to obtain spectra are similar to that for sodium iodide crystals, except that
because smaller pulses must be measured, high-quality electronics are
needed. The complete system includes a high-voltage bias supply for the
detector (up to 5000 V for large depletion volumes), a preamplifier, amplifier
(usually charge-sensitive), biased amplifier (if needed), pulse height analyzer,
and recording device.
A gamma-ray entering either a Nal(Tl) crystal or a semiconductor
detector may lose all or part of its energy in the detector. In the former case,
through multiple Compton interactions or the photoelectric effect, a full
energy peak is obtained. Otherwise, only part of the energy will be observed
and a Compton continuum spectrum is seen. An alternative process for high-
energy gamma-rays (< 1.02 MeV) is pair production, in which an electron-
position pair is produced, and gamma-ray peaks are observed at 0.511 MeV
intervals below the full energy peak. The two most important operating
characteristics of gamma detectors are efficiency and resolution. The "peak-
to-Compton" or "peak-to-valley" ratio is frequently given in the literature
and is related to both efficiency and resolution. These parameters should be
specified by the manufacturer and the conditions under which they were
measured should be given.
The resolution of sodium iodide crystals is usually specified in terms of the
width of the full-energy gamma-ray peak at half its maximum, the "full
width at half maximum" (FWHM). This is shown graphically in the gamma-
ray spectrum in Fig. 2. The resolution improves with increasing energy and
the standard for comparison is usually the 0.662 MeV gamma ray emitted in
the decay of 137Cs. Good sodium iodide crystals have resolutions in the range
of 6.5 to 7 percent for 137Cs. Detection efficiency for the same geometry and
window thickness is a function of several parameters, and much published
information of efficiencies for various energies, detector sizes, source to
detector distances, and other variables is available [14], The efficiency for
FIG. 2—Pulse height or energy spectrum of Cs-137.
gamma-ray detection may be expressed in various ways. Of primary interest

in spectrometry is the full peak efficiency—the fraction of incident gamma
rays that give full-energy peak for particular source-detector configuration.
For a 102-mm-thick Nal(Tl) crystal, with the source on the surface (zero
distance), this fraction is about 0.24 for the 0.66-MeV gamma ray of 137Cs
and about 0.14 for the 1.33-MeV gamma ray of ^Co. The peak-to-valley or
peak-to-Compton ratio is the ratio of counts at the maximum height of the
full-energy peak to the counts at minimum of the Compton continum (Fig.
2). A high ratio indicates narrow peaks, that is, good resolution, for that
particular efficiency. The Compton spectrum does not give useful informa-
tion in gamma-ray spectrometry and can be considered as "noise." The ratio
varies with energy and is frequently given for the 1.33-MeV peak of 60Co. It
increases as the crystal size increases and, after passing through a minimum,
increases as the source-to-detector distance increases, since a larger fraction
of the gamma-rays passes through the full depth of the crystal. A peak-to-
valley ratio of 12:1 for a crystal is very good. This ratio can be increased by
anticoincidence shielding to cancel Compton events as described earlier. The
efficiency of silicon for gamma rays is considerably less than sodium iodide
because of its lower atomic number (the efficiency for photoelectric absorp-
tion of gamma-rays is proportional to Z5) and lower density (the density of
sodium iodide is 3.67 g/cm3 and of silicon 2.4 g/cm3), so it is more efficient
than sodium iodide for the same size detector. For a 1-MeV gamma-ray, the
total absorption coefficient is about 2 mm-' for sodium iodide, 1.5 mm-1 for
396 MANUAL ON WATER
silicon, and 3 mm-1 for germanium. However, germanium detectors are not
yet available in sizes approaching that of sodium iodide. The efficiency of a
germanium (lithium) detector is generally expressed by comparison with that
of a 76- by 76-mm cylindrical Na(Tl) detector. Comparison is made between
the full-energy peak efficiencies for the 1.33-MeV gamma ray of ^Co when
the source is 250 mm from the detector. A germanium detector with a
volume of 35 cm3 has an efficiency about 5 percent that of a 76- by 76-mm
Nal(Tl) crystal. Larger Ge(Li) detectors are available with relative ef-
ficiencies of 25 to 30 percent.
There are limitations in the efficiency of the light production and
collection processes in the sodium-iodide-photomultiplier system that make
its resolution inferior to that of semiconductor detectors. One important
factor is that about 500 eV are required to produce an electron at the
photocathode in a sodium iodide detector system, while the average energy
to produce the analogous electron-hole pair in silicon is only 3.5 eV and in
germanium 2.8 eV. The resolution of semiconductor detectors does not
change greatly with energy. Presently available germanium detectors have
resolutions of 1.5 to 2.8 keV at 1.33 MeV and are from 3 to 30 percent
efficient compared with a 76 by 76 mm Nal(Tl) detector. This greater
resolution makes this detector the one of choice for gamma-ray spectrometry
and cancels to some extent the higher efficiency available from sodium
iodide. Since the pulses from a single photopeak are spread over a much
smaller energy range in germanium than in sodium iodide, the background
under the peak is much less. This means that for small sources of moderately
energetic gamma rays, germanium is more sensitive than sodium iodide. This
is indicated in Table 1, where the efficiencies and backgrounds of a 76- by 76-
mm sodium iodide crystal and a 35-cm3 (5.5 percent efficiency) germanium
detector are compared.
Spectra of beta particles and conversion electrons can be obtained with
sodium iodide and semiconductor detectors sufficiently thick (a few centi-
metres) to absorb the particles completely. One disadvantage of sodium
iodide and cooled lithium-drifted semiconductors is their relatively thick
entrance windows. Other semiconductor detectors, particularly the silicon
surface barrier type, have thin entrance windows and can be used for beta
particles at room temperature or better at 0°C. The 5- to 10-keV resolution
for 600-keV electrons is better than the 12- to 30-keV resolution for 5-MeV
alpha particles.
Good spectra of low-energy beta particles, conversion electrons, and X-
rays can be obtained with a gas-flow proportional counter provided that a
linear preamplifer is used. The resolution is intermediate between sodium
iodide and germanium. To reduce back-scattering, the chamber should be
made of low-Z material. A counter constructed of a cylinder of graphite-
impregnated plastic, lucite ends, and a thin coaxial center wire gives good
spectra for such radiations [14]. A hole is cut into the outer wall and covered
with aluminized Mylar film to provide a thin entrance window. Argon (90
percent)-methane (10 percent) is a suitable counting gas.
TABLE 1—Comparative performance of Sodium iodide (thalium) and germanium (lithium)

gamma-ray detectors.
Background, Counting Detection Limit,
Photon Energy cps, Efficiency, % photons/s
Detector Under Peak A B A B
0.14 MeV
Nal 24 26 18 2.4 3.5
Ge 0.7 12 4 0.92 2.8
0.66 MeV
Nal 20 14 9 4.1 6.3
Ge 0.11 1.3 0.68 3.6 6.8
1.33 MeV
Nal 8 5.8 3.8 6.2 9.5
Ge 0.055 0.75 0.38 4.5 8.9
NOTES—
Nal: 76 mm diameter by 76 mm high cylindrical detector.
Ge(Li): 35 cm3 active volume—5.5 percent efficiency.
A = small source placed on detector.
B = 57 by 57 by 57-mm-thick source place on detector.
Counting efficiency: percent of photons emitted from the source that give a full-energy peak.
Shielding: 152 mm of iron, 3.2 mm lead.
Counting time: One 30 000 s count for both source and background.
Detection limit: The number of photons emitted from the source whose net count equals
twice the counting error, or
where N is the total number of counts recorded when the source is measured, and NB is the
total number of counts recorded when the background is measured.
Organic scintillators, such as anthracene and polystyrene polymerized

with scintillating compounds, are also useful for beta spectrometry. They are
packaged with a phototube in a manner similar to sodium iodide crystals.
Liquid scintillation mixtures also give beta spectra, and the ouput of a
commercial liquid scintillation counter can be fed into a multichannel pulse-
height analyzer to obtain a spectrum [2]. A spectrum of 210Pb 210Bi-210Po
in Fig. 3 shows the resolution obtainable by liquid scintillation counting of
aqueous samples in a dioxane-based solution. The 210Bi curve is from a beta
particle, and the 2l°Po peak is from an alpha particle. Organic scintillators
are preferable to sodium iodide for beta spectrometry because less back-
scattering occurs.
The output pulses of any energy-dependent detector, after linear amplifi-
cation, must be sorted out according to energy to obtain the spectrum of
incident radiation. The high-resolution available in detectors requires ana-
lyzers with hundreds of channels to realize their full resolving power. The
amplified pulse is digitized by an analog-to-digital converter (ADC), and the
resulting number for a particular pulse is recorded in a pulse counter whose
location is determined by digital circuitry. This makes it possible to use a
digital computer to count and store in its memory the number of pulses in
each channel. This conversion and storage is relatively slow, and the
FIG. 3—Spectrum of™Pb ™Bi ™Po.
analyzer is blocked from processing a second pulse until the previous

processing is completed. The time required to process a pulse increases with
channel number. The instruments now available are sufficiently fast for
almost all water measurement purposes. Some loss of pulse information is
acceptable, as the analyzers measure and record "live time" fairly accurately.
Thus, the counting time recorded by the analyzer will be the actual time it
was in a condition to receive detector pulses, and not the elapsed time. To
maintain good accuracy, the activity of the sample should be adjusted to give
live times of 90 percent or more. A computer may be permanently combined
with the ADC to operate only in the pulse-height analyzer mode ("hard
wired"), or a separate and larger computer is "soft wired" to the ADC and
can also be programmed for operations other than pulse height analysis,
such as data reduction and spectrum resolution. In either case, this type of
analyzer makes possible automatic digital readout on printers, paper, and
magnetic tape, automatic spectrum plotting, cathode-ray tube curve and
digital presentation, and internal data reduction.
All multichannel pulse-height analyzers currently available are trans-
istorized, and are fairly reliable instruments and relatively easy to operate.
Their maintenance and repair is, however, a specialized skill similar to other
computer repair. In comparing analyzers, some of the important specifica-
tions to consider are number of channels, count capacity, stability, live-time
accuracy, linearity, type of pulse input acceptable, and ADC speed. The
minimum number of channels useful for sodium iodide gamma-spec-
trometry is 128; germanium (lithium) detectors should be used with at least a
1000-channel analyzer, and alpha and beta spectra can profitably use 100 to
400 channels, depending on the energy range to be covered. Analyzers with
4096 channels are fairly common, and larger analyzers are available for
special purposes.
Semiconductor detectors require low-noise charge-sensitive amplifiers.
Because of their excellent resolution, semiconductor detectors are often used
with a biased amplifier following the main amplifier to isolate a portion of
the spectrum for analysis. This makes is possible to use smaller analyzers
than would otherwise be necessary.
Anticoincidence Counters—Substantial background reduction can be
achieved in beta and gamma counters by surrounding or covering the sample
detector with another detector also sensitive to beta or gamma radiation,
and connecting them electronically so that any pulse appearing in both
detectors is cancelled and not recorded as a count. This is usually referred to
as anticoincidence shielding, and is recommended for obtaining very low
backgrounds. This type of counter was used for many years in directional
studies of cosmic rays, and was first applied to reducing the background of
beta counters by Libby in his study of natural 14C. The thick metal
shielding(lead, iron, or mercury) ordinarily used to reduce cosmic-ray and
gamma-ray background must also be present, and is placed outside the
anticoincidence shielding. Gas-filled beta detectors are generally shielded by
400 MANUAL ON WATER
gas-filled detectors, and such anticoincidence shielding is effective primarily

against the paniculate component of cosmic rays. The anticoincidence
shielding for beta counters may consist of a number of long Geiger tubes
("cosmic-ray counters") surrounding the sample detector or a large (about
15.24 cm square) gas-flow detector, with several anode wires so that the
entire area of the counter is sensitive, placed just above the sample detector.
For counting solid beta sources, the sample detector has a diameter of 2.54
to 5.08 cm. Surrounding these counters on all six sides, there is frequently a
layer of high-purity copper to absorb gamma rays emitted from the
outermost shielding, and 102 to 152 mm of lead or iron on all six sides. This
is the form usually taken by the commercially available anticoincidence
shielded beta counters. Plastic or inorganic scintillators could also be used as
the anticoincidence shielding.
Anticoincidence shielding of gamma-ray detectors operates in a similar
way, and is particularly useful in reducing the Compton continum back-
ground of gamma rays [75]. Gamma rays that undergo Compton scattering
and produce a pulse in both the detector and the anticoincidence shield are
cancelled electronically. Ideally, only those gamma rays that are completely
absorbed in the sample detector itself produce a count which is recorded
with the total energy of the gamma ray (full-energy peak). There are second-
order effects that prevent complete elimination of Compton scattering, but
the improvement is substantial. The anticoincidence shield can be a large
sodium iodide or plastic scintillator suitably attached to phototubes. They
usually have a large annular hole into which the sample detector, a smaller
sodium iodide detector, or germanium iodide detector is placed [76,77].
Coincidence Counters—In coincidence counting, two or more radiation
detectors are used together to measure the same sample, and only those
nuclear events, or counts, that occur simultaneously in all detectors are
recorded. The coincidence counting technique finds considerable application
in studying radioactive decay schemes, but, in the measurement of radio-
activity, the principal uses are for the standardization of radioactive sources
and for counter background reduction.
Coincident counting is a very powerful method for absolute disin-
tegration rate measurement [8,18]. Both alpha and beta emitters can be
standardized if their decay schemes are such at /3-7, y-y, |3-j3, «-£, or «-X-
ray coincidence occur in their decay. Gamma-gamma coincidence counting
with two sodium iodide crystals, and the source placed between them, is an
excellent method of reducing the background from Compton-scattered
events. Its use is limited, of course, to counting nuclides that emit two
photons in cascade (which are essentially simultaneous), either directly as in
^Co, or by annihilation of positrons as in 65Zn, or by immediate emission of
a gamma ray following electron capture decay. If the crystals are operated
with single-channel pulse-height analyzers to limit the events recorded from
each crystal to one of the full-energy peaks of the photons being emitted,
then essentially only those photons will be counted. Noncoincident pulses of

any energy in either one of the crystals will be cancelled, including cosmic-
ray photons in the background and degraded or Compton scattered photons
from higher-energy gamma rays in the sample. Thus, the method reduces
interference from other gamma emitters in the sample. If, instead of single-
channel analyzers, two multichannel analyzers are used to record the
complete spectrum from each crystal, singly and in coincidence, then the
complete coincident gamma-ray spectrum can be obtained with one mea-
surement. The efficiency for coincidence counting is low since it is the
product of the individual efficiencies in each crystal, but the sensitivity is
generally improved because of the large background reduction [79], This
technique is often referred to as two-parameter or multidimensional gamma-
ray spectrometry.
Additional background improvement is obtained if the two crystals are
surrounded by a large annular sodium iodide or plastic scintillation crystal
connected in anticoincidence with the two inner crystals. In this case, a
gamma ray that gives a pulse but is not completely absorbed in one of the
two inner crystals and also gives a pulse in the surrounding crystal is
cancelled electronically. [16,19] This provides additional reduction in the
Compton scattering background. Lithium-drifted germanium detectors may
be used in place of the inner sodium iodide crystals for improved resolution
and sensitivities [77]. All of the equipment described in this section is
available commercially.
Sampling
Collect the sample in accordance with the applicable part of the following:
ASTM Sampling Steam (Method D 1066)
ASTM Sampling Water (Method D 3370) (Particularly Section 14.3)
Sample an appropriate volume depending on the expected concentration
of radioactivity in the water. For precise measurements without long
counting times, it is advisable to count an aliquot which contains at least 1
nCi of radioactivity.
Chemical treatment of samples to prevent biological or algal growth is not
recommended and should be avoided unless essential. When necessary, the
reagents used should be selected to avoid chemical interaction with the
radioactive species in the sample. Samples should be analyzed promptly.
Chemical treatment of samples to retain radioactive species in solution
may be employed, but the specific treatment should be carefully selected.
The use of oxidizing acids such as HNO? is not recommended when iodide is
present, since it may be oxidized to iodine and lost or be absorbed on plastic
containers if used. In some cases, extreme chemical treatment may be used to
keep a particular chemical species in solution; examples are strongly alkaline
conditions to hold molybdenum and ruthenium in solution, or acid con-
ditions with fluoride ion to keep zirconium in solution. The addition of an
402 MANUAL ON WATER
acid such as hydrogen chloride is generally desirable to reduce hydrolysis

and the loss of activity on container walls. Frequently, samples will contain
insoluble material. In such cases, the sample should be treated by one of the
following methods:
1. Filter the insoluble material and analyze both the filtrate and insoluble
matter on the filter. During filtration, some material may be sorbed onto the
filter and assumed to be insoluble when in fact it is soluble.
2. Centrifuge the sample and analyze both phases. The insoluble phase
should be washed with distilled water to remove all soluble material without
dissolving the insoluble fraction.
In either of the foregoing separations, when the total activity is required,
the insoluble matter may be dissolved and recombined with the soluble
fraction. When radioactivity is left on the walls of the sample container, it
should be desorbed and added to the sample.
Composite samples may be made by mixing aliquots of successive samples
collected by an automatic sampler. Analysis of such composite samples
yields average results only and loses information on short-term effects.
Instrument Operation and Control

These following procedures assure that counting equipment is functioning
properly and remains in calibration.
Establishing Counter Characteristics—The first step in instrument control
is to establish the operating characteristics of the system. The efficiency for
counting the nuclide of interest is carefully measured under the conditions to
be employed. Counting conditions are selected; that is, gain, discriminator
setting, and voltage are optimized for the radionuclide of interest. The
operating voltage should be set so that any change in counting rate is
minimized for a given voltage fluctuation. The discriminators should be
adjusted to exclude noise and unwanted interferences from the nuclides
being counted. Adjustment should be made to optimize the signal-to-noise
ratio. When the counting conditions have been selected, known interferences
should be monitored to determine such things as the effect of counting betas,
in an alpha counter, or alpha pulses in a beta proportional counter.
Daily performance checks should be made and a log maintained for each
instrument. This log should include the count for a standard and a
background. When the counting rate differs statistically from the expected
performance, additional counting should be done to determine if the counter
is malfunctioning. High background can indicate either an instrumental
problem or counter contamination.
Certified standards are available from several suppliers. Most solution
standards have the pH controlled and carrier added to assure that hydrolysis
or sorption, or both, do not change the concentration of the solution. When
dilutions are made, it is important to maintain the stability of the solution by
diluting with proper matrix. The standard should be stored in a container
which minimizes evaporation by loss either through the walls or out of the
stopper.
If a planchet is prepared as a standard, it should be placed in a suitable
container for storage which will prevent the surface containing the activity
from being contacted. A recommended practice is the preparation of two
standards, using one and storing the second for periodic checks to see that
the working standard has not been altered.
Counter Control Charts—The daily standard counts made on any counter
must be evaluated on a statistical basis and the best way to do this is to
maintain a control chart on each counter. A control chart is a graph showing
the number of counts recorded in a fixed counting period against day of the
year. The initial entry, N is determined from the average of several
measurements over a period of days. Then the error bands of x and
± 2±3S
S x are entered (Sx- \/N) and lines are drawn on the chart which allow for
decay of the standard over the year. For a long-lived source, such as cesium-
137, which has a half life of 30 years, about 2.5 percent of the activity is lost
in one year. An example of a control chart is shown in Fig. 4.
FIG. 4—Typical counter control chart.
The result of the standard count is entered in the control chart and the
following action is taken:
If the result is inside the ± 2SX band, the counter is considered to be in
control.
If the result lies outside the ± 2SX band but inside the ± 35V band, the
counter is considered to be in control, but this result should be flagged.
If the result lies outside the ± 35V band, the counter is considered out of
control. Corrective action is needed if repeated counts remain outside the
± 3SX band.
For alpha and gamma spectrometry it is also important to monitor for
system resolution.
404 MANUAL ON WATER
In addition to the control charts made on each counter, all pertinent in-
formation about the system should be kept in a log book or permanent file:
Counter Logs
a and j3 Counters 7 Spectrometers
Standard counts Standard counts
Background counts Standard resolution
System changes Background counts
Control charts Control charts
Counting Statistics
Each nuclear disintegration [5,20,27] is a completely random and inde-
pendent process. Established methods of statistical analysis are available to
describe the random phenomenon of nuclear disintegration. The total
number of particles counted in a time period can be shown to deviate from
the average according to the expression, N ±\fN~, where N is the number of
counts in the counting period t. Similar behavior is found in the background
count rate for any counter, and the recorded background will deviate from
the average according to the expression NB ±-\fNB, where NB is the number
of counts recorded in the counting period tB. The net count rate found
is then
Net count rate =
where Sx is the standard deviation of the net count rate and is given by
or for the net count rate with its standard deviation we have
Net count rate =
In practice, many counts will "be made which deviate from the average by
more or less than the standard deviation since 5V by definition would give
those limits X - Sx and X + Sv which would include about 68 percent of
all observed values. In terms of the population of observations between
limits, we can express the change of probability of a result being between the
limits X - K S{ and X + KS, as "confidence levels" where K is simply a
multiplier for Sv. In evaluating the effect of K, we can tabulate K and the
confidence level.
Confidence level, percent K

68.3 1.0
90 1.645
95 1.96
99 2.575
99.73 3.0
The 95 percent "confidence limit" is frequently used and has been gen-
erally accepted, since errors outside this band are considered statistically
significant. In practice, K is often rounded off to 2.0 and the term 2SX, or
2a, error is often used. In some cases the 3SV, or 3a, error is used and gives
the 99.7 percent confidence limits.
Precision—This is a measure of the reproducibility of a measurement.
There are a number of items which affect the precision of radioactivity
measurements; the more important of these are the following:
Position: For point sources, the observed radioactivity is proportional to
the reciprocal of the square of the distance of the source from the detector.
All sources being directly compared must be measured at the same source
to detector distance.
Radiation Scattering: Changes in the sample support and backing mater-
ial can seriously affect the precision of radiation measurements, particularly
that of beta radiation. Standards and samples should have the same backing
material. The use of a sample support rack constructed of a material with
low atomic number will reduce the effect of side-scattering. Scattering from
backing material can be minimized by use of a thin, low-atomic-number
material.
Background: This radioactivity must be measured with no sample near
the detector and subtracted from all measurements of gross sample activity.
This requirement becomes more critical as the net sample activity becomes
small with respect to the background. Routine periodic measurements of
the background must be performed to check for possible detector con-
tamination or malfunction.
Absorption: Alpha and beta radiation is partly absorbed by the sample
406 MANUAL ON WATER
itself being counted and by all substances separating the sample from the
detector. This effect is usually small for gamma radiation, but beta and alpha
radiation are seriously affected. If sources of the same atomic number and
mass are compared on the same counter in the same geometry, absorption
will be constant under these conditions but should not be ignored. Since
the sample mass may vary significantly, calibration curves must be con-
structed to correct for changes in self-absorption.
Quenching: This is any process which reduces the photon output in a
scintillation system; this in turn reduces the measured count rate. Quenching
can be caused by such things as sample color and chemicals in the sample.
The need to correct for this phenomenon can be avoided if samples of the
same color and same chemical composition are compared. If this is not
possible, most instrument manufacturers and texts describe methods for
construction of calibration curves to correct for this phenomenon [22-24].
Overall Uncertainty in a Determination—Measurement results should be
reported with the estimated overall measurement uncertainties shown. There
are two approaches to this. One is a rigid propagation of uncertainties,
which is not sound in practice since the individual components are not well
characterized, and the second is a combination of known uncertainties. In
combining uncertainties, use the relationship
where
S = overall uncertainty of measurement (RSD),
S\ = random counting uncertainty (RSD),
5*2 = other random uncertainties (RSD),
and the work can be divided into two main classes.

Gamma Spectrometry: Assuming that the integrity of the sample is
known, the sampling and treatment effects are at a minimum and then only
two significant terms are present. These are accuracy of calibration and
precision of counting, and we have:
where
Sy spec = overall error for the measurement (relative standard deviation
or RSD),
SV = calibration uncertainty (RSD), and
Sa = counting uncertainty (RSD).
In general, Sa is only important at low count rates, that is, environmental

and discharge samples.
Separative Work: The sample treatment introduces uncertainty into the

measurement. One estimate of great merit is to say that
where
Sx = overall uncertainty for measurement (RSD),
SM = intrinsic precision of method (RSD),
Sa = counting uncertainty (RSD), and
Sc = calibration uncertainty (RSD).
The intrinsic precision may be determined by doing a single operator

precision (SOP) test at three or four activity levels.
SOP Test Error Due to Counting
Level, dpm a % (RSD) a % (RSD)

10 8 C, SMl
100 4 C2 SM2
10 000 2 C3 SW3
(NOTE: The use of ASTM Methods which have been round-robin tested will give you a method
where Sx is known.)
Minimum Detectable Activity—The minimum detectable activity (MDA)

is a statistical measure of the sensitivity of a counting determination. In
the analysis of environmental samples or discharge samples, the sensitivity
obviously becomes an important and often critical item. For gross counting,
the minimum detectable activity may be arbitrarily defined as the net count,
which is equal to twice the standard deviation for the net count, that is,
25.r.
Where complex Nal(Tl) spectra are being analyzed and the components
being sought are not completely separable by instrument resolution, the
detection limit should be considered as three times the standard deviation.
Alpha and Ge(Li) gamma spectrometry generally have well resolved peaks
so that a two-standard-deviation uncertainty may be used as a limit of
detection.
In computing S.v, the sample and background count rates should be
considered, that is
If the sample and background counting times are equal (/ = IB), we can use a
slightly different form
408 MANUAL ON WATER
counts
The total count form is simpler to use in estimation of MDA's, but care
must be taken in converting from total counts back to concentration in
the sample. An example of the calculation of MDA's follows for ASTM
Test for Radioactive Barium in Water (D 2038), which covers the deter-
mination of barium-140 in water:
By beta counting, assuming 80 percent chemical yield and counting at 40
percent efficiency with 1-cpm background. For 100-min count periods we
have
counts
and at MDS we know that ns~-~ nb
MDA = 2nb = 2 X 100

MDA = 2 200 = 28.2 counts
in disintegrations per minute (dpm) terms
MDA
or for 1 litre sample, MDA = 0.4 pCi/litre.

By gamma counting of the sample as received, we can count the 1.6-MeV
peak from lanthanium-140 using either sodium iodide or germanium lithium
detectors. These two cases are:
Sodium iodide, 4.9 percent efficiency and a background of 6.4 cpm.
Assuming 100-min counts.
MDA =
MDA= 107 counts
in dpm terms
MDA = = 21.9 dpm
or for a 1 litre sample, MDA = 9.9 pCi/ litre.
Germanium-lithium, 0.85 percent efficiency and a background of 0.12

cpm. Assuming 100-min counts
MDA =
MDA = 8.8 counts

in dpm terms
MDA= = 11.5dpm
or for 1 litre sample, MDA = 5.2 pCi/litre.
An independent assessment of this method by the Environmental Pro-

tection Agency (EPA) [25] suggested sensitivities of 1 pCi/litre for a 1-litre
sample by beta counting and 10 pCi/litre for a 3.5-litre sample by gamma
counting. The value of the germanium lithium detector is obvious in this
case—good resolution is available with a good sensitivity—and in com-
paring these MDA's to the maximum permissible concentration (MPC) for
barium-140 in water, we find
MPC = 3 X 104 pCi/litre soluble

2 X 104 pCi/litre insoluble
[ Beta MDA =0.4 pCi/litre"J

Nal MDA = 9.8 pCi/litre I all less than 1 percent of the
Ge(Li) MDA = 7.8 pCi/litreJ most restrictive MPC
Calculations and Symbols

To calculate the amount of activity of a substance X, the following
general method should be used
where
N = number of counts accumulated,
/ = sample counting period, min,
NB = number of background counts accumulated,
tB = background counting period, min,
C = net counts per minute per millilitre, cpm/ml
V = volume of sample, millilitres, and
Y = fractional recovery of species (unity in methods where complete
recovery is assumed).
The disintegration rate (£>) can be calculated as follows
where
D = disintegrations per minute per millilitre, dpm/ ml
410 MANUAL ON WATER
E = fractional efficiency of counter, cpm/dpm, and

R = reagent blank correction (cpm) measured on an actual blank sample.
In gamma spectrometry we have
E= GP
where
G = fractional abundance of gamma ray concerned and
P = photopeak detection efficiency (fraction).
In beta counting
where
fj ~ detector efficiency factor for a source of given beta energy,
fa = absorber factor for total absorber thickness (detector window + air
space + absorber + cover) for a source of given beta energy,
fs= self-absorption and backscatter factor for a given precipitate thick-
ness and given beta energy, and
fm = number of charged particles per disintegration.
The concentration of Substance A may be calculated as follows:
where
A = concentration in microcuries per millilitre (juCi/ml).
For decay correction, use the expression
D = D° exp (-0.693 t/ 7)
where
D° = disintegration rate at time zero, that is, reference time
t = elapsed time between measurement and reference time, and
T = half-life of nuclide.
Parent-daughter relationships are commonly shown by use of subscripts

1, 2, etc. for count rates, disintegration rates, or values of t and T.
Alpha Counting
This section covers the measurement of the alpha particle radioactivity
of water. It is applicable to alpha emitters having energies above 3 MeV
at activity levels above 0.5 picocuries per sample.
Summary—Alpha particles are characterized by intense loss of energy
in passing through matter. This intense loss of energy is used in differ-
entiating alpha radioactivity from other types through the dense ionization
or intense scintillation it produces. This high rate of loss of energy in
passing through matter, however, also makes sample preparation conditions
for alpha counting more stringent than is necessary for other types of
radiation.
Alpha Detectors—Alpha radioactivity is normally measured by one of
several types of detectors in combination with suitable electronic com-
ponents. The detector devices most used are ionization chambers, pro-
portional counters, silicon semiconductor detectors, and scintillation
counters. The associated electronic components in all cases would include
high-voltage power supplies, preamplifiers, amplifiers, sealers, and re-
cording devices.
Detection Technique—In all of these systems, the initial event is con-
verted to an electrical pulse which is amplified to a voltage sufficient to
operate the sealer mechanism where provision is made for recording each
pulse. The number of pulses per unit of time is directly related to the disin-
tegration rate of the test sample. The efficiency of the system can be deter-
mined by counting standards prepared in the same manner as the samples.
An arbitrary efficiency factor can be defined in terms of a different radio-
nuclide such as natural uranium, polonium-210, plutonium-239, or ameri-
cium-241.
Measurement Variables—The measured alpha counting rate from a
sample will depend on a number of variables. The most important of these
variables are geometry, source diameter, self-absorption, absorption in
air and detector window, coincidence losses, and backscatter. These are
discussed in detail in the literature [4,5] and in many cases can be measured
or corrected for by holding conditions constant during the counting of
samples and standards. These effects may be described by the following
relationship
where
cpm = recorded counts per minute, corrected for background, and

dpm = alpha disintegrations per minute.
Gr = point source geometry, which is the solid angle subtended by the
412 MANUAL ON WATER
sensitive area of the detector. The effect of this variable is

eliminated by maintaining a constant geometry for both standard
and sample counts.
/ = backscatter factor, or ratio of cpm with sample backing to cpm
without backing. Not important in alpha counting since back-
scatter is small. In samples mounted on copper or stainless steel
planchets for counting 2ir geometry, backscatter may be taken
as 2 percent (backscatter factor equals 1.02) without any serious
error. In samples mounted on platinum, backscatter may be taken
as 4 percent.
faw = factor to correct for losses due to absorption in air or window
of external counters.
fd = factor to correct for dispersion of the source from a point con-
figuration. The effect of this variable is eliminated by preparing
the standard in the same configuration as the samples.
/„.„ = factor to correct for the absorption and scattering of alpha par-
ticles in the sample and its mount. This was covered further in
an earlier section.
/ = factor to correct for losses due to the resolving time of the
detector and its associated electronics; / =(1 - ni) where n is
observed counting rate, cpm, and t is resolving time, min.
Alpha counters have low backgrounds and high efficiencies. However,
some counters are easily contaminated internally and care should be taken
to avoid contamination. Silicon detectors operated in vacuum may become
contaminated due to recoil from sources. Recoil contamination can be
eliminated by maintaining an air absorber of 12 /ug/cm/ 2 between source
and detector [6].
Interferences—Some alpha counters are sensitive to beta radiation with
a degree of efficiency depending on the detector [4,5~\. In these cases,
electronic discrimination is often used to eliminate the smaller pulses due
to beta particles.
Apparatus
lonization Chambers—Alpha particles entering the sensitive region of an
ionization chamber produce dense ionization of the counting gas. The
electrons are collected at the anode, thus developing a voltage pulse. Gridded
ion chambers are used for alpha spectrometry, but have been replaced by
solid-state detectors. The major advantage of the ion chamber is its ability to
measure large area samples with essentially a 2?r geometry. The chambers
are operated at an overpressure with gas mixtures such as 10 percent
methane-90 percent argon or 2 percent ethylene-98 percent argon. The peak
resolution of such a detector is about 50-keV FWHM; refer to ASTM
Recommended Practice for Alpha Spectrometry (D 3084).
Proportional Counters—Alpha particles entering the sensitive region of a
proportional counter produce ionization of the counting gas. In this case, the
electrons are accelerated toward the anode, producing secondary ionization
and developing a large voltage pulse by gas amplification. Proportional
counters are usually operated in the "limited proportional" range, where the
total ionization is proportional to the primary ionization produced by the
alpha particle.
Proportional detectors are generally constructed of stainless steel or
aluminum; see Fig. 5. No additional shielding is required for alpha
proportional counting. The counter should be capable of accepting mounts
up to 51 mm in diameter. Proportional detectors are available in two types,
either with or without a window between the sample and the counting
chamber. The manufacturer's specifications for either type should include
performance estimates of background count rate, length and slope of the
voltage plateau, and efficiency of counting a specified electrodeposited
standard source along with the type of gas used in the tests. For a window
flow counter, the window thickness in milligrams per square centimetre
should also be specified. With windowless low counter, the sample and
sample mount should be made of an electrical conductor in order to avoid
erratic behavior, due to static charge buildup.
Alpha emitters are counted with proportional instruments in 2?r, or 50
percent, configuration. TWO-TT geometry is obtained by placing the sample
on a flat planchet inside the detector. Half the alpha particles are emitted
downward into the planchet, of which approximately 2 percent are back-
scattered in the upward direction. The other half are emitted upward into the
gas volume.
Typical parameters for the alpha windowless flow counter are: back-
ground count rate—10 counts per hour; length of voltage plateau— 300 V;
slope of voltage plateau—1 percent per 100 V for an electrodeposited source.
For a window flow counter, typical values are: window thickness—1
mg/cm2; background count rate—10 counts per hour; length of voltage
plateau—300 V; slope of voltage plateau—1 percent per 100 V for an
FIG. 5—Typical chamber geometry for proportional detector.

414 MANUAL ON WATER
electrodeposited source; and efficiency—35 to 40 percent for an electro-

deposited source. Gases commonly used in both types of alpha proportional
counters are 10 percent methane-90 percent argon, pure methane, or pure
argon.
Scintillation Counters—In scintillation counter, the alpha particle trans-
fers energy to a scintillator, such as zinc sulfide (silver activated). The
transfer of energy to the scintillator results in the production of light at a
wavelength characteristic to the scintillator, and with an intensity propor-
tional to the energy transmitted from the alpha particle. The scintillator
medium is placed in close proximity to the cathode of a multiplier
phototube, light photons from the scintillator strike the photocathode, and
electrons are emitted. The photoelectrons are amplified by the multiplier
phototube and a voltage pulse is produced at the anode.
The counter size is limited by the multiplier phototube size, a diameter of
51 mm being the most common. Two types of systems may be employed. In
the first, the phosphor is optically coupled to the multiplier phototube and
either covered with a thin (< 1 mg per square centimetre) opaque window or
enclosed in a light-proof sample changer. With the sample placed as close as
possible to the scintillator, efficiencies approaching 40 percent may be
obtained. The second system employs a bare multiplier phototube housed in
a light-proof assembly. The sample is mounted in contact with a disposable
zinc sulfide disk and placed on the phototube for counting. This system give
efficiencies approaching 50 percent, a slightly lower background, and less
chance of counter contamination.
A major advantage of alpha scintillation counting is that the sample need
not be conducting. For a 51-mm multiplier phototube with the phosphor
coupled to the tube, typical values obtained are: background count rate—20
counts per hour, and efficiency for an electrodeposited standard source—35
to 40 percent. With a disposable phosphor mounted on the sample, typical
values are: background count rate—10 counts per hour, and efficiency for an
electrodeposited standard source—45 to 50 percent. For both systems,
voltage plateau length is 150 V with a slope of 5 percent per 100 V.
Liquid scintillation counting of alpha emitters with a commercially
available instrument overcomes many of the problems inherent in other
techniques [8,11-13]. Typical background counting rates range from 440 to
700 counts per hour. Sample preparation involves mixing the sample aliquot
with a suitable liquid scintillator solution or gel phosphor before counting.
Planchet preparation is eliminated, volatile components are retained, and
the completely enclosed sample cannot contaminate the counting chamber.
The sample is uniformly distributed in the scintillator so there is no self-
absorption, resulting in a counting efficiency of almost 100 percent. Because
of the high alpha energies, considerable chemical quenching effects can be
tolerated before counting efficiency is reduced. Coincidence losses are small
in liquid scintillation counting at count rates up to 106 cpm. For samples
which contain both alpha and high-energy beta emitters, difficulties do arise
in distinguishing between the two. The problem is primarily due to the broad
continuum of beta energy distribution up to the maximum energy, and the
poor resolution of liquid scintillation spectrometers. This problem is aggra-
vated because the light yield per MeV of alpha particles in most liquid
scintillators is approximately tenfold lower than a beta particle of equivalent
energy, putting the pulses from alphas and high-energy betas in the same
region. Correction for beta activity may be made by certain mathematical or
graphical techniques. It is preferable to separate the alpha emitter from the
bulk of the beta activity by chemistry.
Semiconductor Detectors—The semiconductor detectors used for alpha
counting are essentially solid-state ionization chambers. The ionization of
the gas in an ionization chamber by alpha particles produces electron-ion
pairs while in a semiconductor detector, electron-hole pairs are produced.
The liberated charge is collected by an electric field and amplified by a
charge-sensitive amplifier. In general, silicon surface barrier detectors are
used for alpha counting. These detectors are /i-type base material upon
which gold is evaporated to make a contact. The semiconductor material
must have a high resistivity since the background is a function of the leakage
current. This leakage current is present in an electric field since the starting
material is a semiconductor, not an insulator. The reversed bias which is
applied reduces the leakage current, and a "depletion layer" of free-charge
carriers is created. This layer is very thin and the leakage current extremely
low; therefore, the interactions of photons with the detector will have a
neglible effect. Since the detector shows a linear response with energy, any
interactions of beta particles with the detector can be eliminated by
electronic discrimination. The semiconductor is of special interest in alpha
counting, where spectrometric measurements may be made, since the
average energy required to produce an electron-hole pair in silicon is 3.5
±0.1 eV, compared with the 25 to 30 eV to produce an ion pair in a gridded
ionization chamber. Consequently, silicon detectors provide much improved
resolution and also normally have lower background count rates.
The detector size is generally less than 25 mm (1 in.) in diameter, since the
resolution decreases and cost increases with detector size. For best results,
the sample should be electrodeposited to make an essentially massless
source. The detector is operated in a vacuum chamber. Typical backgrounds
range from 0.3 to 0.6 counts per hour. Additional specifications for detectors
may be found in ASTM Method D 3084.
Absorption—The range of alpha particles is a few centimetres in air and a
few thousandths of a centimetre in solids. Even fairly short path lengths in
air and passage through thin windows will result in the absorption of some
fraction of the alpha radiation.
Calibration of alpha counting equipment for specific nuclide measurement
should be carried out using a standard of similar alpha energy measured
under exactly the same conditions as the sample to be counted.
When the gross alpha activity of a sample is to be measured [ASTM Test
416 MANUAL ON WATER
for Alpha Particle Radioactivity of Water (D 1943)], the counting equip-

ment should be calibrated using a standard. The standard should contain the
same weight and distribution of solids as the sample and be mounted in an
identical manner. If the samples contain variable amounts of solids or
carrier, a calibration curve should be prepared relating weight of solids
present to counting efficiency. The efficiency factor (E) is expressed as a
fraction of the disintegration rate of the standard.
Calibrated standards of Pu-239, Am-241, Po-210, Th-228, Ra-226, U-233,
U-235, and U (Nat.) are readily available either from the National Bureau of
Standards or commercial organizations. Calibrated standards of other
alpha-emitting radionuclides may be procured from the aforementioned
suppliers upon special request.
Source Preparation—The sample whose alpha activity is to be determined
must be appropriately mounted for the final measurement. This can be
accomplished in a number of ways, some of which are dependent upon the
counting method used. In all cases, however, the sample must be mounted in
a uniform and reproducible manner.
The most straightforward and commonly used method of preparing a
sample for alpha counting is by evaporation onto a suitable planchet.
Evaporated sources are usually counted in end-window or windowless
proportional counters or in zinc-sulfide scintillation counters. Producing a
thin, uniform, and reproducible source by this technique is difficult, and
considerable care is necessary. Evaporations must be carried out slowly to
avoid spattering, and are usually performed under infrared lamps.
Mounting of solid-free samples resulting from radiochemical separations
of specific nuclides is best accomplished by evaporating the sample in the
center of a planchet. Define an area in the center of the planchet and keep
the sample within this area during evaporation. This can usually be
accomplished by carefully evaporating the sample in small portions. Sam-
ples which contain solids present the problem that the material does not
evaporate uniformly but tends to deposit in crystals and aggregates,
producing uneven deposits consisting of rings, ridges, or localized amounts
of solids. This will be the case when water samples are prepared for alpha
counting without chemical separation of specific nuclides. These samples are
best mounted by allowing the sample to evaporate on the entire surface of
the planchet. This will produce a more uniform source and minimize the self-
absorption effects. Surfactants can be used to spread the sample during
evaporation. Dissolution of the solids on the planchet with water and
reevaporation may be necessary to obtain a uniformly distributed source.
Samples which contain carriers can be mounted using a precipitation
technique. The amounts of carrier used should be kept small (< 10 mg) so as
to minimize self-absorption. The carrier is precipitated by suitable chemical
procedures and the precipitate is mounted for counting by either filtration
through a filter or by evaporation of a slurry of the precipitate onto a
planchet. The final surface density of such a deposit should be less than 1
mg/cm2. These samples should be counted in an end-window proportional

counter or in a zinc scintillation counter because they are nonconductors.
Filtration of a precipitate onto a filter produces a more uniform and
reproducible source than the evaporation technique. When evaporating a
slurry of a precipitate onto a planchet, care must be used to obtain a uniform
and reproducible source. However, this is usually more easily obtained with
a slurry than when evaporating a solution.
Electrodeposition is the most effective and widely used technique for
preparation of a thin, uniform, and reproducible source for alpha counting.
Electroplated sources can be counted in any alpha counting equipment ex-
cept, of course, a liquid scintillation counter. The chemical separation of
specific nuclides from the sample is required prior to electrodeposition. The
alpha-emitting element is electrodeposited onto a flat metal disk. Platinum,
stainless steel, nickel, and tantalum have been used for this purpose. Informa-
tion on electrodeposition of alpha emitters has been published [26-28]. The
electrodeposition technique produces the most suitable sources for alpha
spectrometry. One disadvantage of this technique is that, for some samples,
quantitative deposition is not always obtained and a recovery correction
factor must be applied. Electrodeposition followed by alpha spectrometry is
particularly useful for alpha analysis in which yield measurements are deter-
mined by addition of a known amount of an isotope of the same element to be
measured (see ASTM Method D 3084).
The use of liquid scintillation counting avoids some of the difficulties
encountered in preparing samples for solid source counting, but has the
disadvantage of blindly counting all alphas as does any gross method. Self-
absorption problems are eliminated and counting efficiencies are 90 to 100
percent. Samples are prepared for liquid scintillation counting by dissolving
the sample in a scintillation mixture [29,30] or extracting the alpha-emitting
element into a scintillation mixture containing the extractant [37].
Calculations—This method is useful for comparing activities of a group of
samples as in a tracer experiment. The following equation may be used to
calculate the results.
Activity C =
where:

/ = sample counting period, min,
N B = number of background counts accumulated,
te = background counting period, min,
C = net counts per minute per millilitre (tpm/ml),
V = volume of sample, millilitres, and
Y = recovery of species (unity in methods where complete recovery
is assumed).
418 MANUAL ON WATER
The disintegration rate or concentration in microcuries may be calculated

as follows:
Disintegration rate
D = disintegrations per minute per millilitre (dpm/ml), and

E = efficiency of counter (fraction) (cpm/dpm).
Concentration A =
where:
A = concentration in microcuries per millilitre (/uCi/ml).
Results may also be reported in terms of equivalent natural uranium

activity, employing an efficiency determined by use of a natural uranium
reference standard.
r
Alpha activity equivalent to natural uranium, dpm/ml = -pr-
t>u
where
EM = efficiency of counter for natural uranium (fraction).
To properly evaluate the result, the uncertainty associated with the con-
centration should be calculated according to the recommended procedure
given earlier.
Beta Counting
This section covers the measurement of the beta particle radioactivity of
water. It also covers the general techniques used to prepare and measure
the activity resulting from radiochemical separation of specific nuclides or
groups of nuclides in water samples. It is applicable to beta emitters with
activity levels above 10 picocuries per sample for most counting systems.
For samples of lower activity, see ASTM Recommended Practice for Low-
Level Activity Measurements (D 3085) [2]. The method is not applicable
to samples containing radionuclides that are volatile under conditions of the
analysis. General information on radioactivity and measurement of radia-
tion may be found in the literature [9,11}.
Summary—Beta radioactivity may be measured by one of several types
of instruments which provide a detector and a combined amplifier, power
supply, and sealer. The most widely used detectors are proportional or
Geiger-Muller counters, but scintillation systems offer certain advantages.
Among the gas ionization-type detectors, the proportional-type counter
is preferable because of the shorter resolving time and greater stability of the
instrument. For perparing solid sources from water samples for beta activity
measurement, the test sample is reduced to the minimum weight of solid
material having measurable beta activity by precipitation, ion exchange, or
evaporation techniques. For measuring solid sources resulting from in-
dividual radiochemical separation procedures, the precipitate is appro-
priately mounted for counting.
Beta particles entering the sensitive region of the detector produce
ionization or scintillation photons which are converted into an electrical
pulse suitable for counting. The number of pulses per unit time is directly
related to the disintegration rate of the sample by an overall efficiency factor.
This factor combines the effects of sample-to-detector geometry, sample self-
shielding, backscatter, absorption in air and in the detector window (if any),
and detector efficiency. Because most of these individual components in the
overall beta-particle detection efficiency factor vary with beta energy, the
situation can become complex when a mixture of beta emitters is present in
the sample. The overall detection efficiency factor may be empirically
determined with prepared standards of composition identical to that of the
sample specimen, or an arbitrary efficiency factor can be defined in terms of
a single standard such as cesium-137 or other nuclide. Gross counts can
provide only a very limited amount of information and therefore should be
used only for screening purposes or to indicate trends.
Liquid scintillation counting avoids many sources of error associated with
counting solid beta sources, such as (1) self-absorption, (2) backscattering,
(3) loss of activity during evaporation due to volatilization or spattering, and
(4) variable detection efficiency over a wide beta-energy range. In addition to
the greatly improved accuracy offered by liquid scintillation counting,
sample preparation time and counting times are significantly shorter.
Sample preparation involves only adding a sample aliquot to the scintillator
or gel phosphor. Because every radioactive atom is essentially surrounded by
detector molecules, the probability of detection is quite high even for low-
energy beta particles. Radionuclides having maximum beta energies of 200
keV or more are detected with essentially 100 percent efficiency. Liquid
scintillation can, at times, be disadvantageous due to chemiluminescence,
phosphorescence, quenching, or the typically higher backgrounds.
Organic scintillators, such as p-terphenyl plus a wave shifter, in a plastic
monomer are polymerized to form sheet material of any desired thickness.
The plastic phosphor counting system [10] has its widest use as a beta
particle detector for separated, solid samples rather than for beta spectro-
metry applications.
This plastic beta scintillator phosphor is mounted directly on the sample
and is discarded after counting. The phosphor-sample sandwich is placed in
direct contact with the photomultiplier tube yielding essentially a 2n
configuration. Since the output pulse of the detector system is energy
dependent, the counting efficiency for a given phosphor thickness of 0.0254
cm yields the highest counting efficiency with the lowest background.
Solid samples, precipitates from radiochemical separations containing 3
420 MANUAL ON WATER
to 5 mg/cm2 of stable carrier, are measured in such a system. For Y-90, a

solid sample of this type would have a counting efficiency of 45 to 50 percent.
A plastic scintillator phosphor system with a 25-cm photomultiplier tube
shielded with 12.7 mm of lead has background in the order of 2.5 cpm. For
very low backgrounds, about 0.25 cpm, the photomultiplier tube and sample
assembly are fitted into a well-type hollow anode Geiger tube operated in
anticoincidence. The entire assembly is then place in a heavy shield.
The system has many advantages, but reduction of background is
probably most important. The reduction occurs since the scintillation does
not see the surrounding mechanical components of the counter. The
additional advantage of keeping the counter itself free from contamination
by enclosing the phosphor-sample sandwich is also important.
A note of caution is advisable at this point. Any beta particle detection
system, whether internal gas counters or scintillation counters, will detect
alpha particles. It is not possible to electronically discrimate against all the
alpha pulses.
If a sample is suspected of containing alpha activity, a separate alpha
measurement must be made to determine the alpha contribution to the beta
measurement.
Solid Source Counting—The observed count rate for a solid source is the
interaction of many variables. The most important of these are the effects of
geometry, backscatter, radiation, source diameter, self-scatter, and self-
absorption, absorption in air and the detector window (for external coun-
ters), and coincidence counting losses. These effects have been discussed
[77,20] and in many cases can be reduced or corrected for by counting test
standards and samples under identical conditions. For absolute measure-
ments of a single, specific nuclide, appropriate correction factors must be
applied.
These effects may be described by the following relation
where
cpm = recorded counts per minute, corrected for background, and
dpm = disintegrations per minute yielding beta particles.
Gp = source-to-detector geometry, which is the solid angle subtended
by the sensitive area of the detector. This is corrected for most
easily by maintaining constant counting configuration and geo-
metry for standards and sample measurements.
fhs = backscatter factor, or ratio of cpm with backing to cpm without
backing. This is the phenomenon in which particles originally
emitted away from the detector are scattered back toward it by
reflection off the source backing material. Its value ranged from 1
to almost 2, depending on the thickness and atomic number of the
backing material. The backscatter error may be minimized by
using identical backing material for counting standard and samples.
fm. = factor to correct for losses due to absorption in the air and window
of external counters. It is equal to the ratio of the actual counting
rate to that which would be obtained if there were no absorption
by the air and window between the source and sensitive volume of
the detector. It is related to the absorption coefficient and density
of the absorber by the approximate equation, af w= exp (-JLUC),
where /x — absorption coefficient, in square centimetres per milli-
gram, and x = absorber density in milligrams per square centi-
metre. In practice, it can be accounted for, together with the
geometric factor, by maintaining the geometry of the counting
configuration constant for standards and samples.
fh= factor to correct a disk source counting rate to the counting rate
of the same activity as a point source on the same axis of the
system. This can be corrected for by preparing and counting the
standard in the same configuration as the samples.
fssg = factor to correct for the absorption and scatter of beta particles
within the material accompanying the radioactive element.
f(. = factor to correct the counting rate for instrument resolving time
losses and defined by the simplified equation,/.= (l-nt), where
n = observed counts per minute and / = instrument resolving
time in minutes. Generally, the sample size or source-to-detector
distance is selected to preclude counting losses. Such losses are
negligible for count rates of less than 80 000 counts per minute if
proportional detectors are used. Information on the effect of
instrument resolving time on the sample count rate as well as
methods for determining the resolving time of the counting
systems may be found in the literature [77].
Liquid Scintillation Counting—The observed count rate for a liquid

scintillation sample is directly related to the beta (plus conversion electron)
and positron emission rate in most cases. The important exceptions are (1)
beta emitters whose maximum energy is below 200 keV, and (2) counting
systems wherein quenching decreases the expected photon yield, thereby
decreasing the overall detection efficiency significantly below 100 percent.
Low-energy beta emitters such as tritium or carbon-14 can be measured
accurately only when the appropriate detection efficiency factor has been
determined with a known amount of the same radionuclide counted under
identical conditions. Quenching losses are greatest at low beta energies.
Quenching may be evaluated by comparison to known quench standards of
the same radionuclide, using the channel ratio technique, or with other
techniques as described in the manufacturer's instructions. For absolute
measurements of low-energy beta emitters, appropriate correction factors
must be applied
422 MANUAL ON WATER
where
dpm = disintegrations per minute, yielding beta particles;
cpm = recorded counts per minute, corrected for background;
fe* = detection efficiency factor; observed cpm/dpm ratio for known
standard of some radionuclide counted under identical conditions
with no quenching; and
fq = quench correction factor, observed ratio of cpm-quenched/cpm-
unquenched for known standard of same radionuclide with equiv-
alent quenching.
In tracer studies or tests requiring only measurements in which data are

expressed relative to a defined standard, the individual correction factors
cancel whenever (1) sample composition, (2) sample weight, and (3) counting
configuration and geometry remain constant during the standardization and
tests.
The limit of sensitivity for both Geiger-Muller and proportional counters
is a function of the background counting rate. Massive shielding or anti-
coincidence detectors and circuitry, or both, are generally used to reduce the
background counting rate to increase the sensitivity [11]. For a more com-
plete discussion of this, see ASTM Method D 3085 [7].
Interferences
Solid Sources—Material interposed between the test sample and the
detector, as well as increasing mass in the radioactive sample itself, produces
significant losses in sample counting rates. Since the absorption of beta
particles in the sample solids increases with increasing mass and varies
inversely with the maximum beta energy, sample residue must remain
constant between related test samples and should duplicate the residue of the
evaporated standard.
Most beta-particle counters are somewhat sensitive to alpha, gamma, and
X-ray radiations, with the degree of efficiency dependent upon the type of
detector [77]. The effect of the interfering radiations on the beta counting
rate is more easily evaluated with external-type counters where appropriate
absorbers can be placed between the sample and detector to evaluate the
effects of interfering radiation.
(NOTE: The detection efficiency may be assumed to be unity for beta emitters having
maximum energies of 200 keV or greater.)
Liquid Scintillation—Liquid scintillation samples are subject to inter-
ference by substances which quench or enhance the scintillation process.
This includes any chemiluminescence which changes the photon yield.
Substances such as oxidants, organohalides, ketones, and aldeyhydes are to
be particularly avoided. A second type of quenching is color quenching, in
which a colored solution impairs the light collection efficiency for the
photons produced.
Apparatus
Beta Particle Counter—The end-window Geiger-Muller tube and the
internal or external proportional gas-flow chambers are the two most
prevalent types of detectors. Other types of detectors include scintillators
and solid-state detectors. The material used in the construction of the
detector and its surroundings should contain a minimal level of detectable
radioactivity. If the detector is of the window-type, the window thickness
may be used in calculating beta ray attenuation; however, direct calibration
of the entire counting system with standards is recommended. The manu-
facturer should provide all settings and data required for reliable and
accurate operation of the instrument. Detectors requiring external posi-
tioning of the test sample shall include a support of low-density material
(aluminum or plastic) which assures a reproducible geometry between the
sample and the detector. Because different sample-to-detector geometries are
convenient for differing sample activity levels, the sample support may
provide several fixed positions ranging from 0.5 to 10 cm from the detector.
Liquid Scintillation—Liquid scintillation counting systems use an organic
phosphor as the primary detector. This organic phosphor is combined with
the sample in an appropriate solvent which achieves a uniform dispersion. A
second organic phosphor often is included in the liquid scintillation "cock-
tail" as a wavelength shifter. The wavelength shifter efficiently absorbs the
photons of the primary phosphor and re-emits them at a longer wavelength
more compatible with the multiplier phototube. Liquid scintillation counting
systems use either a single multiplier phototube or two multiplier phototubes
in coincidence. The coincidence counting arrangement is less likely to accept
a spurious noise pulse which occurs in a single phototube, and thus provides
lower background. The requirement that both multiplier phototubes re-
spond to each has a slight effect on the overall detection efficiency of betas
with E-max >200 keV; however, system response to beta £-max <200 keV
will be significant. The need to minimize detectable radioactivity in the
detector and its surrounding is likewise important in liquid scintillation
counting. To achieve this scintillation grade, organic phosphors and solvents
are prepared from low carbon-14 materials such as petroleum. The counting
vials are of low-potassium glass or plastic to mimimize counts due to
potassium-40. Liquid scintillation provides a fixed geometry from a given-
size counting vial and liquid volume.
Detector Shield—The detector assembly should be surrounded by an
external radiation shield of massive metal equivalent to approximately 51
cm of lead and lined with 3.2-mm-thick aluminium. The material of
construction should be of minimal detectable radioactivity. The shield
should have a door or port for inserting or removing specimens. Detectors
having other than completely opaque windows are light-sensitive. The design
of the shield and its openings should eliminate direct light paths to the
detector window; beveling of door and opening generally is satisfactory.
Liquid scintillation counting systems must provide an interlock which
424 MANUAL ON WATER
protects the photocathode of the multiplier phototube from light when the
sample counting chamber is opened.
Associated Electronic Equipment—The high-voltage power supply ampli-
fier, sealer, and mechanical register normally are contained in a single
chassis. The power supply and amplifier sections are matched with the type
of detector to produce satisfactory operating characteristics and to provide
sufficient range in adjustments to maintain stable conditions. The sealer
should have capacity for storing and visually displaying at least 9 x 105
counts. The instrument should have an adjustable input sensitivity matched
to that of the detector, and variable high-voltage power supply. (An
adjustable power supply and meter are unnecessary for liquid scintillation
systems.) Counting chambers of Geiger-Muller and proportional counters
contain a suitable counting gas and an electrode. Counting rates which
exceed 10 000 cpm must be corrected for dead-time loss when using a Geiger-
Muller tube. As the applied voltage to the electrode is increased, the
counting chamber exhibits responses which are characteristic of a particular
voltage region. At low voltages of the order of 100 V, there is no
multiplication of the ionization caused by a charged particle. At voltages
approaching 1000 V, there is appreciable amplification of any ionization
within the counting chamber; however, the size of the output pulse is
proportional to the amount of initial ionization. When operated in this
voltage region, the device is known as a proportional counter. Usually there
is a region at least 100 V wide, known as a plateau, wherein the count rate of
a standard is relatively unaffected. The operating voltage for proportional
counters is selected to approximate the middle of this plateau in order to
maintain stable responses during small voltage shifts. The plateau region is
determined by counting a given source at voltage settings which differ by 25
or 50 V. The number of counts at each setting is recorded, and the resultant
counts versus voltage are plotted as shown in Fig. 6. Voltage plateau curves
are to be remeasured periodically to ensure continued instrument stability,
or whenever an instrument malfunction is indicated. If the voltage is
increased beyond the proportional region into the 1500 to 2000 V region, the
pulse size increases and the dependence on the initial ionization intensity
FIG. 6—Counting rate as afunction of applied voltagefor a proportional counter exposed to a

source emitting both a and /3 particles.
disappears. This is the beginning of the Geiger counting region, where a

single ion pair produces the same large pulse as an intense initial ionization.
Alpha Interference—Alpha particle interference can be substantial and
must be considered with any type of beta counter. One technique involves
placing a thin absorber between the solid source and the detector. The
absorber diameter should exceed that of the detector window. The absorber
should be placed against the window to minimize beta particle scatter. Any
absorber which stops alpha particles also will somewhat attenuate low-
energy beta particles. For example, an aluminum absorber of 7 mg/cm 2 will
absorb 48 percent of beta particles of 0.35-MeV maximum energy. The alpha
particle absorber is not recommended for use with internal beta particle
detectors, especially when either the composition or activity ratios of the
radioelements or radioactivity level might vary significantly between sam-
ples. Chemical separation of the alpha and beta particle emitters produces a
higher degree of accuracy for internal detector measurements. Published
information on beta particle absorption [//] should be used as a guide for
use of an absorber. In liquid scintillation spectra, the alpha component
appears as a peak on the beta continuum and thus provides a basis for
resolving the two [32].
Self-Absorption and Backscatter—Radioisotopes emit radiation uniformly
in all directions. If the radioisotope is intimately mixed with the detector, as
in the case with liquid scintillation counting, essentially all beta particles
interact with the detector. With solid source counting, however, the source
and detector usually are related by a solid angle of 2-ir or less. Even so, the
number of beta particles reaching the detector can be significantly affected
by other factors.
When a radioactive source decays by beta emission, the beta particles
(electrons) are emitted with a distribution of energies ranging from zero to
definite maximum value. The lower energies are absorbed by the sample
itself and this self-absorption is particularly large for certain nuclides, such
as carbon-14 and sulfur-35, in which all the beta particles are of low energy.
It is therefore, desirable to make solid samples as thin as possible.
Self-absorption corrections can be made, if necessary. A series of counts is
made on a number of weighed samples of homogeneous material, with the
material spread uniformly over the sample mount. The data are corrected for
background, and graphed on a semi-log plot, with the logarithm of the
counting rate per milligram of sample plotted against the number of
milligrams. The resulting curve can be extrapolated to zero sample weight to
give the counting rate per milligram corrected for self-absorption [5],
Samples consisting of the final product from radiochemical separations
generally are of the same approximate weight from determination to
determination because they usually contain the same amount of carrier, and
the self-absorption is essentially constant. Corrections can be made by
running a standard amount of activity of the nuclide through the procedure,
with sample activities expressed in terms of this reference standard.
426 MANUAL ON WATER
Beta particles are easily deflected in passing through matter. Some beta
particles emitted in a direction away from the detector can be deflected by
the source-backing material and scattered toward the detector. This phen-
omenon leads to an increased number of beta particles reaching the detector
over that expressed from geometric considerations alone. Backscatter
increases rapidly with increasing thickness and atomic number of the
backing material. Backscatter effects may be held constant by maintaining
sample size and mounts constant for relative measurements. If backscatter
corrections must be made, these corrections can be determined by counting a
sample first on thin film mount, and then recounting when backed by the
desired backing material. Backscatter does have the effect of increasing the
number of particles which reach the detector, thereby increasing the
counting efficiency.
Calibration and Standardization

Calibration and Standardization for General Measurements—Solid
sources. Place a known amount of cesium-137 or other activity (approx-
imately 5 X 10r4 to 5 X 10~3 microcuries, with the smaller amount more
appropriate for Geiger-Muller detectors) in a volume of water containing
salts equivalent to those of the test samples and prepare for counting as
directed. Throughout the experiment, the evaporation, mounting, counting,
and the density of the plate solid of this reference must be identical with
those of the test samples. Count for length of time necessary to produce the
desired statistical reliability. The combined efficiency factor then is ex-
pressed as a percentage of the disintegration rate of the reference standard.
(CAUTION: This factor is inaccurate for beta particles whose energies differ
appreciably from those of cesium-137; nevertheless, count rates for solid
beta sources of undefined energy are often reported relative to cesium-137
standard. The fact that the 662-keV gamma ray of barium-137 is signif-
icantly converted accounts for the fact that about 1.10 electrons leave the
source for every disintegration. For this reason, cesium-137 standards
should be calibrated in terms of combined beta plus electron emission rate.)
Liquid Scintillation—Add a known amount of cesium-137 (approxi-
mately 10~2 microcuries to the liquid scintillation "cocktail." The volume and
composition of the cesium solution and the "cocktail" should simulate
those normally used. For samples without serious quenching problems, the
counting efficiency should approach 100 percent. (Note that cesium-137
standards must be calibrated in terms of beta plus conversion electron
emission rate.)
Calibration and Standardization for Tracer Experiments—Add a known
quantity of activity of a reference solution of the tracer (approximately 5 X
10~3 microcuries) to a nonradioactive standard test sample and process.
Source Preparation
Solid Sources—The sample whose activity is being determined must be
appropriately mounted for the final measurement. The exact form of the
mount is somewhat dependent upon the particular instrument for counting.

In general, most counters will accept either flat plate or dish mounts.
Water samples can be evaporated on either flat disks or dishes, with the
dishes being preferred for high solids samples. The dishes should have a flat
or concentrically ringed bottom whose diameter is no greater than that of the
detector window, preferably having 3.1-mm-high sidewalls with the angle
between the dish bottom and side equal to or greater than 120 deg to reduce
sidewall scattering. Sample dishes with verticle sidewalls may be used, but
the exact positioning of these dishes relative to the detector is very
important. This factor becomes critical for dishes having the same diameter
as the detector. Dishes having sidewalls more than 3.1 mm in height are not
recommended. Dishes should be of a material that will not corrode under the
plating conditions and should be of uniform surface density, preferably great
enough to reach backscatter saturation [77].
Samples resulting from radiochemical separations are customarily in the
form of a precipitate on a filter. These can be mounted on plates, dishes, or
other suitable backing. The filter paper should be fixed to the backing with a
double-backed adhesive strip or a liquid adhesive, and should be covered to
prevent sample loss and to prevent contamination of the counting equip-
ment. Such covering can be accomplished with a cover of transparent plastic
tape or film, or by spraying the precipitate with a collodion. Nonconducting
coverings should not be used in internal flow counters.
Liquid Sources—Liquid samples of sufficiently high activity level may be
added directly to an appropriate amount of liquid scintillator solution. For
aqueous samples the liquid scintillator solution must contain a solvent such
as dioxane, which is miscible with water, or a surfactant to maintain a single
phase with immiscible solvents. Low-activity liquid samples may be con-
centrated by evaporation before addition to the liquid scintillator solution.
Liquid samples are sometimes concentrated or separated, or both, by
extraction into a suitable organic solvent. This solvent often is suitable for
direct addition to the liquid scintillator solution. Caution is advised to avoid
solvents which cause substantial quenching.
Radioactive Decay and Decay Curves

Decay curves are useful for identifying radionuclides by their character-
istic half-lives and in determining the presence and number of interfering
radioactive substances. When two or more beta-emitting isotopes are present
in a source, an analysis of the decay curve is required to resolve them since
the semi-log plot of the disintegration rate versus time is a curve whose slope
changes as the isotopic composition changes. After sufficient time, the
longest-lived activity predominates. The half-life of this long-lived com-
ponent can be determined from extrapolation of the straight-line portion of
the curve to /0, the original count rate for that radionuclide, as shown in Fig.
7. In this graph, the extrapolated line of the 4.6 day half-life component
subtracted from the original curve yields a curve that represents all other
428 MANUAL ON WATER
FIG. 7—Decay curve.
components; this residual curve may be processed again in the same way to
reduce (in principle) any complex decay curve into its component parts. In
actual practice, uncertainties in the observed data place a practical limit to a
three-component system except in unusual cases, and often a two-component
system may not be satisfactorily resolved because of similar decay constants.
Since the residual curve in Fig. 7 (1 day half-life) does not deviate from the
straight line, and the original curve is concave except in the final portion, a
two-component system on independently-decaying activities is indicated. In
accumulating decay data, care must be taken to avoid disturbing the source
and to locate the sample in the same geometrical position for each
measurement. If the instrument is calibrated for the specific characteristics
of the source and for the energies of the radiations measured, the count rate
shown at the ordinate intercept of the extrapolated straight lines at t0 will be
equal to the disintegration rate at /0. However, this technique requires
specific beta-energy knowledge for each component of the sample, as well as
counting instrument standardization at these same energies under identical
conditions. In practice, gamma spectrometry is preferred; see ASTM Test
for Gamma Spectrometry of Water (D 2459) [7] for resolving mixtures of all
radionuclides except pure beta emitters. An additional complication in
decay curve interpretation is the growth of radioactive daughters. See
standard Refs 11 and 20 for a discussion of this.
Beta Particle Maximum Energy

When a radioactive source decays by beta emission, the beta particles
(electrons) are emitted with a distribution of energies ranging from zero to a
definite maximum value. The maximum energy of the electrons is character-
istics of a particular nuclide and is the "beta energy" shown in nuclear data
tables. Determining the approximate maximum energy of the beta particles
from a radioactive source thus aids in radionuclide identification.
The combined effects of a continuous beta spectrum and scattering
produces an approximately exponential absorption law for a beta particle
source of a given maximum energy. In interposing varying thickness of
aluminum between the source and the detector and plotting the count rate
versus the thickness of aluminum absorber in mg/cm 2 on semi-log paper, an
empirical absorption curve is produced. The plotted data can be compared
with a standard absorption curve determined for a particular nuclide to
check for other radionuclides. If the sample and absorber are as close as
possible to the detector and only one radionuclide is present, the absorption
curve is nearly a straight line until it tails into the constant background count
rate of the instrument, which in most cases includes the gamma rays from the
source. Visual inspection of the point at which the beta activity is not
detectable above the total background gives an approximate thickness value,
although this is usually low. Better results can be obtained by subtracting the
background from the total absorption curve, as shown in Fig. 8, to obtain
the 220 mg/cm2 component. The point at which the straight-line extra-
polation intersects the count rate error of the background is the absorber
thickness value. The maximum beta energy can then be obtained from the
range energy relation curve, Fig. 9. If the lower-energy region of the curve
deviates from the straight-line extrapolation to zero absorber thickness, a
430 MANUAL ON WATER
FIG. 8—Absorption curve.
second component is indicated. The values on the extrapolated line sub-

tracted from the second curve produced the 40 mg/cm2 component. The sum
of the count rate for both components at zero absorber thickness is the true
count rate.
The best absorption curve is obtained with the absorber against the
detector and the source positioned for minimum clearance of the thickest
CHAPTER 18 ON PRACTICES FOR MEASUREMENT OR RADIOACTIVITY 431
FIG. 9—Empirical range-energy relationship for electrons absorbed in aluminum. Experi-

mental values by several observers on monoenergetic electrons are shown. For monoenerg'etfc
electrons, the range coordinate refers to the extrapolated range. For continuous 0-ray spectra,
the energy coordinate refers to the end-point energy E . . . , and the range coordinate becomes
• the maximum range. The smooth curve represents the empirical relationship, developed by Kat:
and Penfeld, Reviews of Modern Physics. Vol. 24, 1952, p. 28.
absorber to be used. The source must not be moved after the first count rate,
which is obtained with no absorber between the source and the detector. The
activity of the source should be close to the maximum allowed by resolving
time of the instrument. Absorption coefficients given in the literature and
half-thickness data utilize the initial portions of the absorption curve and are
not as accurate as the total absorption determination. More accurate
methods for determination of beta particles range, from absorption curves,
are the comparison method of Feather [77] and Harley et al [33]
Many liquid scintillation counting systems provide an output suitable for
accumulation and storage of beta spectra of known maximum energy. To
obtain the most accurate beta maximum energy measurement, the pulse
height-energy relationship of the system must be calibrated [34]. With the
aid of this relationship, the pulse height distribution of the sample may
be converted to a Curie plot [35] and extrapolated to obtain maximum beta
energy. Excellent agreement with beta-energy spectrometry values has been
obtained with this technique [36]. A Curie plot is shown in Fig. 10.
Calculations—Results may be expressed in observed counts per minute
per millimetre for comparing relative activities of a group of samples as in
tracer experiments
432 MANUAL ON WATER
FIG. 10— Curie plot of the /3- spectrum of "4 In. Note especially that N means N(n), the
number of fi rays in a momentum interval A, of constant size. The horizontal coordinate is the
kinetic energy E, which corresponds to the midpoint of the momentum interval n + An. When
spectral data give a straight line, such as this one, the N(n) is in agreement with the Fermi
momentum distribution. The intercept of this straight line, on the energy axis, gives the
disintegration energy Emax (E*), if the rest mass of the neutrino is zero. (From Lawson and
Cork, Physical Review, Vol. 57, 1940/7. 982.
where
t = sample counting period, min,
NB — number of background counts accumulated,
K = volume of initial sample, ml, and
C = net counts per minute per millilitre (cpm/ml).
Results also may be expressed in terms of equivalent cesium-137 activity,

using the empirical efficiency determined for the reference standard
where
C = net counts per minute per millilitre (cpm/ml), and
£i37Cs = efficiency for counting Cs-137 beta particles plus conversion
electrons (count rate + emission rate).
If it is known that only one nuclide is present, its disintegration rate may
be determined by use of the efficiency factor measured with a reference
standard of the same nuclide. The results may be calculated as follows
where
B = background in counts per minute (cpm),
D = disintegration rate per minute per millilitre (dpm/ml),
E = efficiency of counter for specific nuclide (counts + disintegrations),
V = millilitre of test specimen,
Y = recovery of species (unity in cases where complete recovery is
assumed), and
t = count period, min.
If it is desired that the activity be expressed in units of microcuries, the

following expression is used:
where
A = concentration, microcuries per millilitre,
D = disintegration rate per minute per millilitre (dpm/ml), and
2.22 X 106 = conversion factor, disintegration per minute per microcurie.
Gamma Counting
This section covers the measurement of gross gamma radioactivity of
water. Since gamma radiation is a penetrating form of radiation, it can be
used for samples of any form and geometry as long as standards of the same
form are available and are counted at the same geometry to calibrate the
detector. Because of this penetrating nature, small variations in sample
density or sample thickness are usually not significant and gamma counting
is the preferred method in general radiochemical work. When a standard
cannot be obtained in the matrix being counted, a correction for the
different absorption in the matrices must be made.
Since different nuclides emit distinct and constant spectra of gamma
radiation, the use of an energy-discriminating system provides identification
and measurement of all the components present in a mixture of radio-
activity, and this technique is covered in ASTM Recommended Practice for
Gamma-Ray Spectrometry (D 2459) [7]. Gamma counting and gamma
spectrometry are applicable to levels of about 10 picocuries or above.
General information on gamma ray detectors and gamma counting is
covered in the literature [11,16].
434 MANUAL ON WATER
Summary
Gamma counting is generally carried out using solid detectors since a gas-
filled detector will not provide adequate stopping power for energetic
gammas. In solids such as sodium iodide (thallium) or cesium iodine, the
gammas interact by excitation of atoms, and energy is transferred to orbital
electrons and then released as light photons when the orbits are refilled. The
scintillations are easily detected and amplified into useable electrical pulses
by a multiplier phototube. The Nal (Tl) detector is the recommended
detector for gross gamma counting.
In semiconductor detectors such as Si (Li) and Ge (Li) the gamma photons
produce electron-hole pairs and the electrons are collected by an applied
electrical field. A charge-sensitive preamplifier is used to detect the change
transferred and produce a useable electrical pulse. The semiconductor
detectors are widely used in gamma spectrometry (see ASTM Method D
2459-72) and are not reviewed in this practice.
The output pulses from the multiplier phototube or preamplifier are
directly proportional to the amount of energy deposited, which could either
be total and be included in the photopeak or fractional and be included in
the continuum or escape peaks, in the detector, by the incident photon. The
pulses may be counted using a sealer or analyzed by pulse height to produce
a gamma ray spectrum.
Gamma photons interact with the detector by three distinct processes. The
photoelectric effect results in complete absorption of the photon energy and
produces the full energy or photopeak shown in Fig. 11 results. The third
interaction is pair production, which occurs at energies above 1.02 MeV and
results in disappearance of the photon when an electron positron pair is
produced. The electron and positron give up their kinetic energy to the
detector and the resulting electron joins the electron population of the
detector; the positron, however, is annihilated in combining with an electron
and produces two gamma photons of 0.511 MeV each. One or both of the
0.511 MeV photons may escape from the detector without interacting and
the "single escape" and "double escape" peaks shown in Fig. 12 result. The
Comptons from a higher energy photon always present an interference
problem in the counting of gamma photons, and appropriate corrections
must be made for this effect. Pair production can also be considered as an
interference since the excape peaks may have an energy equal to the lower-
energy gamma of interest.
The change of absorption coefficient with gamma energy results in a wide
variation of detection efficiency. The detection efficiency falls rapidly as
gamma energy increases, for a fixed size of detector. Two other important
effects are seen as a result of variation of absorption coefficient; firstly, low-
energy photons may be absorbed in massive samples, such as large bottles of
water, and erroneous results obtained. A similar absorption effect is seen in
germanium (lithium) systems where the can around the detector acts as an
absorber for very-low-energy gammas and the efficiency passes through a
FIG. II—Compton contenuum.
maximum usually around 100 keV. The second results is that for low-energy
gammas a thin detector may be as efficient as a much thicker one since the
low-energy gammas are easily stopped in the thin detector.
Because of this variation in efficiency and the possible interferences from
other activities, gross gamma counting is only reliable when used to compare
standards and samples of the same nuclide. The use of gross gamma
monitoring systems should be avoided when possible, and, in all cases,
proper allowance must be made for the lack of accuracy.
Interferences
The natural background persent at all locations is detected very efficiently
by gamma detectors and presents a significant interference which must be
reduced by the use of shielding. Low-radioactivity-level lead or steel should
be used and, for a 76- by 76-mm sodium iodide detector, about 102 mm of
436 MANUAL ON WATER
Part of 24Na Gamma

Spectrum 76mm x 76mm NalfTi)
FIG. 12—Single and double escape peaks.
lead or 15.24 cm of steel shielding produces an acceptable background for

most work. Details of shield designs are given by Heath [37]. Lead shields are
a source of X-rays when high activity samples are counted, and this X-ray
emission can be reduced in amount and energy by using a graded liner such
as 1.6 mm of cadmium and 0.4 mm of copper—the copper being nearer the
detector. The cadmium strongly absorbs the lead X-rays and in turn emits its
own X-rays at a much lower energy, and these are strongly absorbed by the
copper. Any residual copper X-rays are usually below the energy level of
interest.
The "Compton" and "pair production" effects can be very significant
interferences and must be corrected for.
At high count rates, "random sum peaking" may occur. Two absorptions
may occur within the resolving time of the detector and are summed and
seen as one pulse. For a detector of resolving time / and a count rate of A
counts per unit time, the time window available for summing is 2A t (since
the count summed could occur as early as / before or as late as / after the
other count) and the probability of another count at any time is simply A.
Therefore, the sum count rate will be 2A2t in unit time. Random summing
is strongly dependent on the count rate A, and if summing occurs it can be
reduced by increasing the sample to detector distance.
Well counters which have very high efficiencies are prone to summing
since for a given source strength the count rate is higher than for a detector
of lower efficiency. For moderate and high source strengths, the tradeoff is a
poor one and the well counter is best suited for low level work where its high
efficiency is an important advantage.
Cascade summing may occur when nuclides which decay by a gamma
cascade are counted. Cobalt-60 is an example; 1.17 MeV and 1.33 MeV from
the same decay may enter the detector and be absorbed, giving a 2.50 MeV
sum peak. Cascade summing may be reduced by increasing the source to
detector distance.
The resolution of a gamma detector is the effective limit to its utility even
when complex data reduction methods are used. A typical 76- by 76-cm
sodium iodide (thallium) detector will give full widths at half maximum
(FWHM) of about 60 keV at 662 keV gamma energy and about 90 keV at
1.33 MeV gamma energy.
Apparatus
Sodium Iodide Detector Assembly—A 76- by 76-mm sodium iodide de-
tector is activated with about 0.1 percent thallium iodide, cylindrical, with or
without an inner sample well, optically coupled to a multiplier phototube and
hermetically sealed in a light tight container. The sodium iodide (thallium)
crystal should contain less than 5 ppm of potassium and be free of other
radioactive materials. In order to establish freedom from radioactive mater-
ials, the manufacturer shall supply a gamma spectrum of the background
of the detector between 0.08 and 3.0 MeV. The resolution of the detector
for the cesium-137 gamma at 0.662 MeV should be less than 60 keV FWHM
or less than 9 percent.
Shielding for the detector shall be constructed as needed from low-
radioactivity-level lead or steel. A thickness equivalent to 10.16 cm of lead
usually provides adequate shielding. If the shield is made of lead; a graded
liner should be used unless the shield-to-detector distance is more than 254
mm.
High-voltage Power Supply—500 to 2000 V dc regulated to 0.1 percent
with a ripple of not more than 0.01 percent.
Preamplifier: Linear Amplifier System—to amplify the output from the
multiplier phototube to a maximum output of 10 V.
Analyzer with Sealer and Timer—A single-channel discrimination system
will accept all or any part of the output from the amplifier and pass it to the
sealer. Any pulses lying outside the preset limits are rejected. The lower limit
is usually referred to as the "threshold" and the difference between the two
limits is the "window."
Sample mounts and containers may consist of any reproducible geometry
container which is commerically available. Other considerations are cost,
ease of use, disposal, and effective containment of radioactivity for pro-
tection of the workplace and personnel from contamination.
438 MANUAL ON WATER
Geometry Control System—A system of shelves or supports for the

various sample containers which allows the user to place a sample in any of
several preset and reproducible geometries.
Beta Absorber—A beta absorber of 3 mm (H in.) to 6 mm (14 in.) of
aluminum, beryllium, or lucite should completely cover the upper face of the
detector to prevent betas from reaching the detector.
Energy Efficiency Relationship

Because of the rapid fall-off in gamma absorption as gamma energy rises,
the detection efficiency shows a similar effect and Fig. 13 shows the ef-
ficiency versus gamma energy plot for a 76- by 76-mm sodium iodide
(thallium) detector. The portion of the curve at low energy shows that, as the
absorption coefficient increases, geometry becomes the limiting factor. The
maximum efficiency seen is well below 50 percent due to the presence of the
beta absorber and the containment of the detector. The 76- by 76-mm sodium
iodide (thallium) detector is the most widely used size, and a large amount of
data are available in the open literature on the use of this size and results
obtained. Heath [2] has written a comprehensive review and supplied many
gamma-ray spectra in both graphical and digital form.
Other sizes of detector may be used. However, the following should be
noted.
FIG. 13—Efficiency versus gamma energy.

Smaller detectors such as 38 by 38 mm will give efficiencies which are low

and fall off more rapidly as gamma energy increases. Small or thin de-
tectors are useful for the measurement of low-energy gammas since they
are less responsive to high-energy ones, and the interference from Compton
effects is reduced.
Larger detectors will give higher efficiencies and less fall-off as gamma
energy increases. Larger detectors are useful for situations where the highest
attainable efficiency is desired and for the assembly of complete absorption
detectors. The increase in efficiency is accompanied by an increased back-
ground count rate and an increase in the probability of summing in the
detector.
Well detectors will give very high efficiencies, up to about 80 percent for
low- and moderate-energy gammas. The well detector is useful for low levels
of activity, and the background of a well detector is essentially the same as
that of a plain cylindrical detector of the same overall dimensions. Summing
becomes a definite problem at high activities since both random and cascade
summing result from the high efficiencies and the high geometry of the de-
tector.
Calibration and Standardization

The system should be calibrated for energy or gain by using at least two
sources of different gamma energies. Cesium-137 at 662 keV and cobalt-60 at
1.17 and 1.33 keV are generally adequate. These sources are used to establish
two parameters: first, the required MeV or keV per volt of amplifier output,
and second, the condition that zero MeV is equal to zero volts output.
Subsequent efficiency calibrations are carried out with the same gain
settings.
Efficiency calibrations are obtained by counting known amounts of single
nuclides in the geometry and matrix which will be used for the unknown
samples. Standards are available from the National Bureau of Standards
and other sources. Extreme care should be taken in the preparation of
subsamples of standards and whenever possible gravimetric dilution and
division techniques should be used. The threshold and window levels used in
volts are obtained using the energy calibration data obtained previously.
In comparing a standard and samples of the same nuclide, accurate results
are possible. When an arbitrary standard is used to estimate the activity of a
mixture of nuclides, the data are not accurate since the real detection
efficiency varies with energy. Because of this gross gamma count, data
obtained on mixtures of nuclides are only useful as an indication of ap-
proximate activity. Such data can be used for sample screening to select
aliquot sizes for gamma spectrometry, comparison of process or waste
samples with each other, and as a general screening for liquid wastes.
Data obtained using an arbitrary standard for a mixture of nuclides
should be reported as being based on a "calibration using nuclide-X" or as
440 MANUAL ON WATER
being "relative to nuclide-X." Data obtained by using this technique should

not be reported as dpm or Ci units without qualification.
Source Preparation
Samples must be prepared in an identical manner to the standards and
counted in the same geometry under identical conditions. For gamma
counting, samples may be liquid, solid, or gaseous form in suitable con-
tainers.
The evaporation of liquid samples to dryness before counting is not
necessary. However, samples which have been evaporated to dryness for
gross beta counting can also be gamma counted. Evaporation may also be
used as a sample concentration technique, especially with pure water such as
evaporator condensates where large samples may be evaporated to small
volume and counted with a higher efficiency.
Gaseous samples can be collected in any type of gas container, but care
must be taken to ensure that all the gas containers of a particular type have a
constant wall thickness. Wall thickness variation can be a significant prob-
lem in counting radioactive xenons which have gammas of 80 keV and 250
keV energy.
Calculations
Results may be expressed in observed counts per minute per millilitre for
comparing relative activities of a group of samples as in tracer experiments
where
t = sample counting period, min,
NB = number of background counts accumulated,
V = volume of initial sample, ml, and
C = net counts per minute per millilitre (cpm/ml).
Results also may be expressed in terms of equivalent cesium-137 activity,

using the empirical efficiency determined for the reference standard
where
C = net counter per minute per millilitre (cpm/ml), and
£"i37Cs = efficiency for counting Cs-137 beta particles plus conversion

electrons (cpm/dpm).
If it is known that only one nuclide is present, its disintegration rate may
be determined by use of the efficiency factor measured with a reference
standard of the same nuclide. The results may be calculated as follows
where
N' = number of counts accumulated,
B = background in counts per minute, (cpm),
D = disintegrations per minute per millilitre (dpm/ml),
E = efficiency of counter for specific nuclide (cpm/ dpm),
V = millilitres of test specimen,
Y = recovery of species (unity in cases where complete recovery is
assumed), and
/ = count period, min.
If it is desired that the activity be expressed in units of microcuries, the

following expression is used
where
A = concentration, microcuries per millilitre,
D = disintegration rate per minute per millilitre (dpm/ml), and
2.22 X 106 = conversion factor, disintegrations per minute per microcurie.
References
[5] Pate, B. D. and Yaffe, L., Canadian Journal of Chemistry, Vol. 33, 1955, pp. 15, 610,929,
1656.
[4] Blanchard, R., Kahn, B., and Birkoff, R. D., Health Physics, Vol. 2, 1960, p. 246.
[J] Hallden, N. A. and Fisenne, I. M., "International Journal of Applied Radiation and
Isotopes, VoL 14, 1963, p. 529.
\6\ Merritt, J. S., Taylor, J. G. V., and Campion, P. J., Canadian Journal of Chemistry, Vol.
37, 1956, p. 1109.
[7] Gunnick, R., Colby, L. J., and Cobble, J. W., Analytical Chemistry, Vol. 31, 1959, p. 796.
[8] Friedlander, G., Kennedy, J. W., and Miller, J. M., Nuclear and Radiochemistry, 2nd
ed., Wiley, New York, 1964.
[9] Price, W. J., Nuclear Radiation Detection, 2nd ed., McGraw-Hill, New York, 1964.
[10] Campion, P. J., Taylor, J. G. V., and Merritt, J. S., "'International Journal of Applied
Radiation and Isotopes, Vol. 8, 1960, p. 8.
[77] Friedlander, G., Kennedy, J. W., and Miller, J., Nuclear and Radiochemistry, 2nd ed.,
Wiley, New York, 1964; Overman, R. T. and Clark, H. M., Radioisotope Techniques,
McGraw-Hill, New York, 1960; Price, W. J., Nuclear Radiation Detection, McGraw-
442 MANUAL ON WATER
Hill, New York, 1964; Flynn, K. F., Glendenin, L. E., and Pradi in Organic Scintillators
and Liquid Scintillation Counting, D. L. Horrocks and Chim-Tzu Peng, Eds., Academic
Press, New York, 1971, p. 687.
[72] Curtis, M. L., Heyd, J. W., Olt, R. G., and Eichelberger, J. F., Nucleonics, Vol. 13, May
1955, p. 38.
[73] Watt, D. E. and Ramsden, O., High Sensitivity Counting Techniques, MacMillan, New
York, 1964.
[14] Crouthamel, C. E., Adams, F., and Dams, R., Applied Gamma-Ray Spectrometery, 2nd
ed., Pergamon Press, New York, 1970.
[75] Nielson, J. M. in Gamma-Ray Spectrometry in Physical Methods of Chemistry, A.
Weissberger and B. W., Rossiter, Eds. Vol. 1, Part 3 D, Chapter 10, Wiley, New York,
1972.
[16] Perkins, R. W., Nuclear Instruments and Methods, Vol. 33, 1965, p. 71; Wogman, N. A.
and Perkins, R. W., Nuclear Instruments and Methods, 1968.
[77] Cooper, J. A., Ranticelit, L. A., Perkins, R. W., Haller, W. A., and Jackson, A. L., "An
Anti-Coincidence Shielded Ge(Li) Gamma-Ray Spectrometer and Its Application to
Neutron Activation Analyses," Report BNWL-SA-2009, Pacific Northwest Laboratory,
Richland, Wash., 1968.
[18] "Metrology of Radionuclides," Proceedings, Symposium, International Atomic Energy
Agency, Vienna, 14-16 Oct. 1959.
[79] Nielsen, J. M. and Kornberg, H. A. in Radioisotope Sample Measurement Techniques
in Medicine and Biology, Proceedings, Symposium, International Atomic Energy
Agency, Vienna, 24-28 May 1965, p. 3.
[20] Lapp, R. E. and Andrews, H. L., Nuclear Radiation Physics, Prentice Hall, New York,
1948; 2nd ed., Englewood Cliffs, N. J., 1954.
[27] Jarrett, Allen A., "Statistical Methods used in the Measurement of Radioactivity with
Some Useful Graphs and Nomographs," Atomic Energy Commission U-262.
[22] Bell, C. G. and Hayes, F. N., Liquid Scintillation Counting, Pergamon Press, New York,
1958.
[23] Birks, J. B., The Theory and Practice of Scintillation Counting, Pergamon Press,
New York, 1964.
[24] The Current Status of Liquid Scintillation Counting, E. Bransome, Jr., Ed., Greene
and Straiten, New York, 1970.
[25] "Environmental Radioactivity Surveillance Guide," ORP/SID 72-2, U. S. Environ-
mental Protection Agency, 1972.
[26] Sill, C. W. and Olson, D. G., Analytical Chemistry, Vol. 42, 1970, 1956.
[27] Puphal, K. W. and Olson, D. R., Analytical Chemistry, Vol. 44, Feb. 1972.
[2«] Mitchell, R. F., Analytical Chemistry, Vol. 32, 1960, pp. 326-328.
[29] Bell, C. G., and Hayes, F. N., Liquid Scintillation Counting, Pergamon Press, New York,
1958.
[30] Moghissi, A. A. in Organic Scintillators and Liquid Scintillation Counting, Academic
Press, New York, 1971 (in preparation).
[37] Horrocks, D. L., Review of Scientific Instruments, Vol. 55, 1964, p. 334.
[32] Bogen, D. C. and Welford, G. A., Proceedings, International Symposium of Rapid
Methods for Measuring Radioactivity in the Environment, LAEA-SM-148/3, 1971, p.
383.
[33] Harley, J. H., Hallden, N. A., and Fisenne, I. M., Nucleonics Vol. 20, 1962, p. 59.
[34] Flynn, K. F., Glendenin, E., Steinberg, E. P., and Wright, P. M., Nuclear instruments
and Methods, Vol. 27, 1964, p. 13.
[35] "Tables for the Analyses of )8 Spectra," National Bureau of Standards Applied Mathema-
tics Series Reports No. 13, U. S. Government Printing Office, Washington, D. C., 1952.
[36] Flynn, K. F. and Glendenin, L. E., Physical Review, Vol. 146, 1959, p. 744.
[37] Heath, R. L., "Scintillation Spectrometry Gamma-Ray Spectrum Catalog," Government
Report Nos. IDO 16880 and ANCR-1000.
Appendix—Reference Tables
and Curves
Table of Atomic Weights
The following values apply to elements as they exist in materials of terrestrial origin and
to certain artificial elements. They are based on the assigned relative atomic mass of carbon-12
and when used with regard to the footnotes they are considered reliable to ± 1 in the last digit,
or ±3 if that digit is a subscript. These values are accepted by the International Union of
Pure and Applied Chemistry, Division of Inorganic Chemistry Commission on Atomic
Weights, and are the 1971 data revised in 1973.
Name Symbol Atomic No. Atomic Weight

Actinium Ac 89
Aluminum Al 13 26.98154
Americium Am 95
Antimony Sb 51 121.75
Argon Ar 18 3994sb.c.d.g
Arsenic As 33 74.9216'
Astatine At 85
Barium Ba 56 137.34
Berkelium Bk 97
Beryllium Be 4 9.01218*
Bismuth Bi 83 208.9804"
Boron B 5 10.81c:d:e
Bromine Br 35 79.904C
Cadmium Cd 48 112.40
Cesium Cs 55 132.9054"
Calcium Ca 20 40.08*
Californium Cf 98
Carbon C 6 12.01 lb'd
Cerium Ce 58 140.12
Chlorine Cl 17 35.453C
Chromium Cr 24 51.996C
Cobalt Co 27 58.9332'
Copper Cu 29 63.54*c'd
Curium Cm 96
Dysprosium Dy 66 162.5o
(Continued)
443
444 MANUAL ON WATER

Einsteinium Es 99
Erbium Er 68 167.2*
Europium Eu 63 151.96
Fermium Fm 100
Fluorine F 9 18.99840"
Francium Fr 87
Gadolinium Gd 64 157.25
Gallium Ga 31 69.72
Germanium Ge 32 72.59
Gold Au 79 196.9665"
Hafnium Hf 72 178.4,
Helium He 2 4.00260b-°
Holmium Ho 67 164.9304"
-Hydrogen H 1 1.0079M
Indium In 49 1 14.82
Iodine I 53 126.9045"
Iridium Ir 77 192.2:
Iron Fe 26 55.847
Krypton Kr 36 83.80
Lanthanum La 57 138.9055h
Lawrencium Lr 103
Lead Pb 82 207.2d'g
Lithium Li 3 6.94,c-d-e-«
Lutetium Lu 71 174.97
Magnesium Mg 12 24.305c-g
Manganese Mn 25 54.9380"
Mendelevium Md 101
Mercury Hg 80 200.5,
Molybdenum Mo 42 95.94
Neodymium Nd 60 144.24
Neon Ne 10 20. 17,*
Nejrtunium Np 93 237.0482f
Nickel Ni 28 58.70
Niobium Nb 41 92.9064"
Nitrogen N 7 14.0067b-e
Nobelium No 102
Osmium Os 76 190.2
Oxygen O 8 15.9994b-c-d
Palladium Pd 46 106.4
Phosphorus P 15 30.97376"
Platinum Pt 78 195.0,
Plutonium Pu 94
Polonium Po 84
Potassium K 19 39.098
Praseodymium Pr 59 140.9077"
Promethium Pm 61
Protactinium Pa 91 231.0359f
Radium Ra 88 226.0254flg
Radon Rn 86
Rhenium Re 75 186.207
Rhodium Rh 45 102.9055"
Rubidium Rb 37 85.467,'
Ruthenium Ru 44 101.0-
Samarium Sm 62 150.4
Scandium Sc 21 44.9559"
Selenium Se 34 78.9*
Silicon Si 14 28.086d
(Continued)
APPENDIX—REFERENCE TABLES AND CURVES 445

Silver Ag 47 I07.868C
Sodium Na 11 22.98977a
Strontium Sr 38 87.62"
Sulfur S 16 32.06a
Tantalum Ta 73 180. 947.,"
Technetium Tc 43
Tellurium Te 52 127.6o
Terbium Tb 65 158.9254"
Thallium Tl 81 204.37
Thorium Th 90 232.038 l f
Thulium Tm 69 168.9342"
Tin Sn 50 118.6,
Titanium Ti 22 47.9,,
Tungsten W 74 183.8,
Uranium U 92 238.029b'c-e
Vanadium V 23 50.941 4 b'e
Wolfram W 74 183.8,
Xenon Xc 54 131.30
Yitterbium Yb 70 173.04
Yttrium Y 39 88.9059"
Zinc Zn 30 65.38
Zirconium Zr 40 91.22
"Mononuclide element.
Êlement with one predominant isotope (about 99 to 100
percent abundance).
Êlement for which the atomic weight is based on meas-
urements calibrated by comparing with synthetic mixtures
of known isotopic composition.
'/Element for which variation in isotopic abundance in
terrestrial samples limits precision of atomic weight given.
''Element for which there is the possibility of large varia-
tions in atomic weight due to inadvertent or undisclosed
artificial isotopic separation in commercially available mate-
rials.
/Most commonly available long-lived isotope.
£Jn some geological specimens this element has a highly
anomalous isotopic composition, corresponding to an atomic
weight significantly different from that given.
Viscosity of water.
Temperature, Viscosity,
°C Pascal-second
10 0.0013077
15 0.0011404
20 0.0010050
25 0.0008937
30 0.0008007
35 0.0007225
446 MANUAL ON WATER
Density of water."
Density of Water, kg/m3 at T deg C
Temperature
Deg°C 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
0 999. 8396 8463 8528 8591 8653 8713 8771 8827 8882 8934
1 8985 9034 9082 9127 9171 9213 9254 9293 9330 9365
2 9398 9430 9460 9489 9516 9541 9564 9586 9606 9625
3 9642 9657 9670 9682 9692' 9701 9708 9713 9717 9719
4 9720 9719 9716 9712 9706 9699 9690 9680 9668 9654
5 9639 9622 9604 9584 9563 9540 9516 9490 9462 9433
6 9403 9371 9338 9303 9266 9229 9189 9148 9106 9062
7 9017 8970 8922 8873 8822 8769 8715 8660 8603 8545
8 8485 8424 8362 8298 8233 8166 8098 8029 7958 7886
9 7812 7737 7661 7583 7504 7424 7342 7259 7174 7088
10 7001 6913 6823 6732 6639 6545 6450 6354 6256 6157
11 6056 5955 5852 5747 5642 5535 5427 5317 5206 5094
12 4981 4866 4751 4634 4515 4396 4275 4153 4029 3905
13 3779 3652 3523 3394 3263 3131 2998 2863 2728 2591
14 2453 2313 2173 2031 1888 1744 1599 1452 1305 1156
15 1006 0855 0702 0549 0394 0238 0081 9923* 9763* 9603*
16 998. 9441 9278 9114 8949 8783 8615 8447 8277 8106 7934
17 7761 7587 7412 7235 7058 6879 6699 6518 6336 6153
18 5969 5784 5597 5410 5221 5031 4841 4649 4456 4262
19 4067 3870 3673 3475 3275 3075 2874 2671 2467 2263
20 2057 1850 1642 1433 1223 1012 0800 0587 0373 0158
21 997. 9942 9725 9506 9287 9067 8846 8623 8400 8176 7950
22 9724 9496 7268 7039 6808 6577 6344 6111 5877 5641
23 5405 5167 4929 4690 4449 4208 3966 3723 3478 3233
24 2987 2740 2492 2243 1993 1741 1490 1237 0983 0728
25 0472 0215 9958* 9699* 9439* 9179* 8917* 8655* 8391* 8127*
26 996. 7862 7596 7329 7061 6792 6522 6251 5979 5706 5433
27 5158 4883 4607 4329 4051 3772 3492 3211 2929 2647
28 2363 2079 1793 1507 1220 0932 0643 0353 0062 9770*
29 995. 9478 9184 8890 8595 8299 8002 7704 7405 7106 6805
30 6504 6202 5899 5595 5290 4984 4678 4370 4062 3753
31 3443 3132 2820 2508 2195 1880 1565 1249 0932 0615
32 0296 9977 9657 9336 9014 8691 8368 8044 7718 7392
33 994. 7066 6738 6409 6080 5750 5419 5087 4755 4421 4087
34 3752 3416 3079 2742 2403 2064 1724 1383 1042 0700
35 0356 0012 9668* 9322* 8976* 8628* 8280* 7932* 7582* 7232*
36 993. 6880 6529 6176 5822 5468 5113 4757 4400 4043 3684
37 3325 2966 2605 2244 1881 1518 1155 0790 0425 0059
38 992. 9692 9324 8956 8587 8217 7846 7475 7103 6730 6356
39 5982 5606 5230 4854 4476 4098 3719 3339 2959 2577
40 2195 1813 1429 1045 0660 0274 9888* 9500* 9112* 8724*
41 991. 8334 7944 7553 7161 6769 6376 5982 5587 5192 4796
_42 4399
"ASTM Designation E-500.
Chemical conversion factors: residue to pertinent

element or radical.
Weighed Sought Factor
AbO., Al 0.529
BaSO4 Ba 0.588
BaSCX SO4 0.412
CaO Ca 0.715
CdS S 0.778
CO: CO, 1.364
CuO Cu 0.799
Fe20, Fe 0.699
KMnO 4 Mn 0.348
Mg2P:O7 Mg 0.219
Mg:P:O7 PO4 0.853
Mn:P;O7 Mn 0.387
(NH4),PO4- 12MoO? P 0.0022
Ni-glyoxime Ni 0.203
PbO: Pb 0.866
SnO: Sn 0.788
TiO^ Ti 0.60
ZnO Zn 0.803
Dissolved oxygen (saturation values).

Temperature, °C mg/ litre
10 11.33
15 10.15
20 9.17
25 8.38
30 7.63
35 7.1
40 6.6
45 6.1
50 5.6
Ionic equivalent conductance at 25° C.

Ion Asa
K* 73.52
Na* 50.11
H+ 349.82
.NH4+ 73.4
'/* Ca" 59.5
to Mg" 53.06
Cb 76.34
NO3- 71.44
to SO4~ 79.8
OH- 198
HCO3- 44.48
to C02- 83
HS- 72
HSO3- 71
to SOr 80
H2PO4- 29
to HPO4- 60
'/3 PO*- 78
"Xo = Ionic equivalent conductance at infinite dilution.
448 MANUAL ON WATER
Dissolved oxygen from air in contact with water—solubility of oxygen as a function of

temperature and pressure.
Temperature conversion formulas.

Deg C = 5/9 deg F - 32
DegF = 9/5degC+32
Bicarbonate, carbonate, and hydroxide relationships.

Bicarbonate Carbonate Hydroxide
P=0 M 0 0
P<KM M-2P 2P 0
P=KM 0 2P 0
P>KM 0 2(M-P) 2P- M
P- M 0 0 M
M = Methyl orange alkalinity.

P = Phenolphthalein alkalinity.
Common conversion factors.

Multiply By To Obtain
Acres 43,560 square feet
Acre-feet 325,851 gallons
Centimetres 0.394 inches
Centimetres 0.01 metres
Centimetres 10 millimetres
Cubic feet 1728 cubic inches
Cubic feet 7.48 gallons
Cubic feet 28.32 litres
(Continued)
Common conversion factors—Continued.

Cubic feet per minute 62.43 pounds of water per minute
Cubic feet per second 0.646 million gallons per day
Cubic feet per second 448.83 gallons per minute
Cubic inches 16.39 cubic centimetres
Cubic metres per day 0.000264 millon gallons per day
Cubic metres per day 0.1834 gallons per minute
Feet 30.48 centimetres
Feet of water 0.8826 inches of mercury
Feet of water 62.43 pounds per square foot
Feet of water 0.4335 pounds per square inch
Gallons 3785 cubic centimetres
Gallons 0.1337 cubic feet
Gallons 231 cubic inches
Gallons 3.785 litres
Gallons, Imperial 1.2 gallons, U. S.
Gallons of water 8.345 pounds of water
Grains 0.0648 grams
Grains per gallon 17.12 parts per million
Grains per gallon 142.86 pounds per million gallons
Grams 15.43 grains
Grams 0.001 kilograms
Grams 1000 milligrams
Grams 0.0021 pounds
Grams per litre 58.42 grains per gallon
Grams per litre 8.345 pounds per 1000 gallons
Grams per litre 1000 parts per million
Inches 2.54 centimetres
Inches of mercury 1.133 feet of water
Inches of mercury 0.4912 pounds per square inch
Inches of water 0.0736 inches of .mercury
Inches of water 0.036 pounds per square inch
Kilograms 2.205 pounds
Kilograms 1000 grams
Kilometers 3281 feet
Kilometers 0.6214 miles
Litres 1000 cubic centimetres
Litres 0.0353 cubic feet
Litres 61.02 cubic inches
Litres 0.2642 gallons
Metres 100 centimetres
Metres 3.281 feet
Metres 39.37 inches
Milligrams 0.001 grams
Millilitres 0.001 litres
Milligrams per litre 1 parts per million
Million gallons per day 1.547 cubic feet per second
Ounces 437.5 grains
Ounces 28.35 grams
Parts per million 0.0584 grains per gallon
Parts per million 8.345 pounds per million gallons
Pounds 7000 grains
Pounds 453.6 grams
Pounds per cubic foot 0.016 grams per cubic centimetre
Pounds per square inch 2.31 feet of water
Pounds per square inch 2.036 inches of mercury
Square miles 640 acres
(Continued)
450 MANUAL ON WATER
Common conversion factors—Continued.

Tons, long 2240 pounds
Tons, metric 2205 pounds
Tons, short 2000 pounds
Percentage of silica in steam versus boiler drum pressure at selected boiler water pH.
Interconversion factors from milligrams per litre to equivalents per million of the
common ions.
Multiplier Multiplier
mg/ litre epm to mg/ litre epm to
Ion toepm mg/ litre Ion to epm mg/ litre -
AT" 0.1112 8.99 Cn" 0.03843 26.02
&r 0.014S6 68.68 0.03333 30.00
or
C03=
0.04990 20.04 F 0.05263 19.00
Cu" 0.03148 31.77 HCOf 0.01639 61.02
Fe" 0.03581 27.92 H2por 0.01031 96.99
Fe"* 0.05372 18.62 HS' 0.03024 33.07
It 0.9921 1.008 HPO4= 0.02084 47.99
K+ 0.02558 39.10 HS04" 0.01030 97.07
Mg" 0.08224 12.16 HSO3" 0.01233 81.07
Mn" 0.03640 27.47 r 0.00788 126.91
Mn~ 0.07281 13.735 NO3" 0.01613 62.01
+
Na 0.04350 22.99 N02" 0.02174 46.01
+
NH 0.05543 18.04 OH" 0.05880 17.01
Pb" 0.00965 103.60 P04= 0.03159 31.66
Sr" 0.02282 43.81 S= 0.06237 16.03
Zn" 0.03059 32.69 SiO4F 0.02172 46.04
=
Br~ 0.01251 79.92 S04 0.02082 48.03
cr 0.02820 35.46 SCh =
0.02498 40.03
Analyses of typical public water supplies in the United States.
(From U. S. Geological Survey Water Supply Paper 1812)
(milligrams per litre)
Water supply," 1 2 3 4 5 6 7 8
Date of collection 4/17/62 8/24/61 7/21/61 5/16/61 3/16/61 1/8/62 7/11/61 8/25/61
Silica (SiO2), mg/ litre 3.1 1.3 4.5 2.0 5.7 7.1 2.7 4.8
Iron (Fe), mg/ litre 0.04 0.05 0.02 0.06 0.06 0.12 0.03 0.00
Manganese (Mn), mg/ litre 0.00 0.09 0.00 0.00 0.00 0.00
Calcium (Ca), mg/ litre 4.5 32 20 6.9 27 1.0 3.2 40
Magnesium (Mg), mg/ litre 0.4 11 7.8 1.0 3.9 0.6 0.6 8.8
Sodium (Na), mg/ litre 2.4 4.0 18 1.8 4.7 i.l 2.9 8.2
Potassium (K), mg/ litre 0.8 0.8 2.8 0.5 2.8 0.4 0.4 1.9
Bicarbonate (HCOa), mg/ litre 8 128 36 13 32 8 11 106
Carbonate (CO3), mg/ litre 0 0 0 0 0 0 0.3 0
Sulfate (SO4), mg/ litre 7.0 20 46 9.0 43 0.8 1.5 45
Chloride (Cl), mg/ litre 3.5 8.0 32 3.8 16 2.2 3.6 15
Fluoride (F), mg/ litre 0.0 1.0 0.2 0.0 0.8 0.0 0.1 0.9
Nitrate (NO3), mg/ litre 0.0 0.4 1.4 0.9 3.6 0.2 0.0 1.2
Dissolved solids, mg/ litre 31 149 187 41 130 22 27 198
Hardness as CaCOs, mg/ litre 13 125 82 21 84 5 10 137
Noncarbonate hardness as CaCo3,
mg/ litre 6 20 52 11 58 0 1 50
: Specific conductance (micromhos

at 25°Q 47 261 284 62 211 18 34 313
PH 6.4 7.5 7.9 6.5 7.2 6.4 9.1 7.9
Color 4 2 10 3 2 5 1 2
Turbidity 1 1 1
"Water supplies are identified as follows:

1. Boston, Mass. Norumbega Reservoir (finished). 5. Philadelphia, Pa. Delaware River (finished).
2. Chicago, 111. Lake Michigan (finished). 6. Portland, Ore. Gull Run River (finished).
3. New Orleans, La. Mississippi River (finished). 7. San Francisco, Calif. Hetch-Hetchy supply (finished).
4. New York, N. Y. Catskill and Delaware River supplies (finished). 8. Washington, D. C. Potomac River (finished).
Analyses by U. S. Geological Survey of typical surface waters of the United States,
(milligrams per litre)
1 2 3 4
Surface waters" max min avg max min avg max min avg max min avg
Oct. 1961 Oct. 1961 Oct. Dec. 13 Mar. 29 Oct. 1961
Jan. 1 to May 1 to to Sept. Oct. 1 to Apr. 1 to to Sept. Dec. 1 to Mar. 1 to 1961 to to 28, to Apr. to Sept.
Date of Collection 10, 1962 10, 1962 1962 31, 1961 30, 1962 1962 9, 1961 31, 1962 Sept 1962 1961 5, 1962 1962
Silica (SiO2), mg/ litre 8.7 2.1 4.4 9.0 7.6 9.5 6.4 5.6 6;2 16 15 15
Iron (Fe), mg/ litre 0.20 0.00 0.01 0.03 0.10 0.07
Manganese (Mn), mg/ litre 0.14 0.00 0.03
Calcium (Ca), mg/ litre 20 8.2 16 8.7 7.5 7.5 46 29 33 83 35 61
Magnesium (Mg), mg/ litre 8.0 2.6 6.0 2.5 1.8 2.4 14 7.1 8.3 24 14 17
Sodium (Na), mg/ litre 10 2.2 7.3 9.0 5.9 7.4 26 7.6 10 64 13 40
Potassium (K), mg/ litre 2.5 1.2 1.9 2.9 1.6 2.1 3.0 2.0 2.0 9.2 4.0 7.6
Bicarbonate (HCO3), mg/ litre' 66 18 51 41 31 36 86 73 80 301 146 195
Carbonate (COs), mg/ litre 0 0 0 0 0 0 0 0
Sulfate (SOa), mg/ litre 30 17 26 7.8 6.4 6.9 103 42 49 178 45 124
Chloride (Cl), mg/ litre 9.0 3.2 7.5 7.6 4.1 5.4 36 10 14 21 8.2 14
Fluoride (F), mg/ litre 0.6 0.0 0.2 0.1 0.1 0.1 0.4 0.1 0.2 0.5 0.3 0.4
Nitrate (NOs), mg/ litre 5.7 2.6 4.1 4.0 1.7 2.3 5.8 4.2 4.4 4.6 3.0 4.0
Dissolved solids, mg/ litre 125 46 104 81 59 69 296 154 175 554 220 394
Hardness as CaCOs, mg/ litre 83 32 65 32 26 28 173 102 116 307 145 223
Noncarbonate hardness as
CaCOs, mg/ litre 31 17 24 0 1 0 102 42 50 61 26 66
Specific conductance
(micromhos at 25° C) 198 86 168 113 84 96 475 249 288 867 346 607
pH 7.8 6.5 6.9 7.7 8.2 6.9 7.6 7.3
Color 8 2 4 15 32 25 20 5 11
5 6 7 8
Surface waters" max min avg max min avg max min avg max min avg
Apr. 1 Oct. 1961 Nov. 1 Oct. 1961 May 8 Oct. 1961 Mar. 9 June 1 Oct. 1961
Date of Collection Sept. 1 to to 30, to Sept. Mar. 1 to to 30, to Sept. Sept. 27 to to 23, to Sept. to 28, to 30, to Sept.
30, 1962 1962 1962 31, 1962 1961 1962 30, 1962 1962 1962 1962 1962 1962
Silica (SiO2), mg/ litre 8.7 7.4 9.5 11 9.5 11 19 18 18 18 6.5 13
Iron (Fe), mg/ litre 0.16 0.00 0.03 0.01 0.00 0.00
Manganese (Mn), mg/ litre
Calcium (Ca), mg/ litre 46 32 40 48 39 43 230 55 77 22 18 20
Magnesium (Mg), mg/ litre 14 8.0 10 18 18 19 51 12 20 6.4 3.9 4.1
Sodium (Na), mg/ litre 26 11 18 37 34 34 212 28 68 10 3.6 6.2
Potassium (K), mg/ litre 3.1 2.1 2.7 - 3.9 3.8 3.7 4.3 3.2 4.1 1.8 0.8 1.7
Bicarbonate (HCO3), mg/ litre 160 101 128 193 172 177 239 170 195 101 61 76
Carbonate (COs), mg/ litre 0 0 0 0 0 0 0 0
Sulfate (SCk), mg/ litre 47 34 41 37 29 33 730 77 156 20 11 15
Chloride (Cl), mg/ litre 36 11 19 54 51 57 200 20 58 5.0 1.0 3.0
Fluoride (F), mg/ litre 0.2 0.0 0.2 0.3 0.2 0.3 0.5 0.3 0.4 0.3 0.1 0.2
Nitrate (NOs), mg/ litre 1.8 1.7 2.2 1.8 0.5 1.6 8.9 2.3 3.6 1.5 0.4 1.0
Dissolved solids, mg/ litre 279 157 218 330 268 293 1,629 314 531 140 69 99
Hardness as CaCOa, mg/ litre 171 113 140 194 172 186 785 185 271 88 54 67
Noncarbonate hardness as
CaCOa, mg/ litre 40 30 35 36 30 40 585 46 111 5 4 4
Specific conductance
(micromhos at 25° C) 434 275 352 540 464 516 2,180 477 803 221 125 161
pH 7.6 7.2 7.5 7.2 7.3 7.9 7.5 7.7 7.8 7.7
Color 10 5 18
"Surface waters are identified as follows:
1. Delaware River at Trenton, N. J. (Water-Supply Paer-1941). 5. Mississippi River at St. Francisville, La. (WSP-1944).
2. Roanoke River at Jamesville, N. C. (WSP-1941). 6. Colorado River at Austin, Tex. (WSP-1944).
3. Ohio River at Lock and Dam 53, near Grand Chain, 111. (WSP-1942). 7. Colorado River near Grand Canyon, Ariz. (WSP-1945).
4. Missouri River at Nebraska City, Nebr. (WSP-1943). 8. Columbia River near The Dalles, Ore. (WSP-1945).
Temperature conversion___«~-
tables (-100 TO 1000 deg)
C F C F C F C F
-73.3 -100 -148 5.0 41 105.8 33.3 92~~ 197.6 211 530 986
Interpolation -67.8 -90 -130 5.6 42 107.6 33.9 93 199.4 282 540 1004
Factors -62.2 -80 -112 6.1 43 109.4 34.4 94 201.2 288 550 1022
-56.7 -70 -94 6.7 44 111.2 35.0 95 203.0 293 560 1040
C F -51.1 -60 -76 7.2 45 113.0 35.6 96 204.8 299 570 1058
-45^6 -50 -58 7.8 46 114.8 36.1 97 206.6 304 580 1076
0.56 1 K8 -40 -40 -40 8.3 47 116.6 36.7 98 208.4 310 590 1094
-34.4 -30 -22 8.9 48 118.4 37.2 99 210.2 316 600 1112
-28.9 -20 -4 9.4 49 120.2 37.8 100 212 321 610 1130
1.11 2 3.6 -23.3 -10 14 10.0 50 122.0 43.0 110 230 327 620 1148
-17.8 0 32 10.6 51 123.8 49.0 120 248 332 630 1166
-17.2 1 33.8 11.1 52 125.6 54.0 130 266 338 640 1184
1.67 3 5.4 -16.7 2 35.6 11.7 53 127.4 60.0 140 284 343 650 1202
-16.1 3 37.4 12.2 54 129.2 66.0 150 302 349 660 1220
-15.6 4 39.2 12.8 55 131.0 71.0 160 320 354 670 1238
2.22 4 7.2 -15.0 5 41.0 13.3 56 132.8 77.0 170 338 360 680 1256
-14.4 6 42.8 13.9 57 134.6 82.0 180 356 366 690 1274
-13.9 7 44.6 14.4 58 136.4 88.0 190 374 371 700 1292
2.78 5 9.0 -13.3 8 46.4 15.0 59 138.2 93.0 200 392 377 710 1310
-12.8 9 48.2 15.6 60 140.0 99.0 210 410 382 720 1328
100.0 212 413
-12.2 10 50.0 16.1 61 141.8 104.0 220 428 388 730 1346
3.33 6 10.8 -11.7 11 51.8 16.7 62 143.6 110.0 230 446 393 740 1364
-11.1 12 53.6 17.2 63 145.4 116.0 240 464 399 750 1382
-10.6 13 55.4 17.8 64 147.2 121 250 482 404 760 1400
3.89 7 12.6 -10.0 14 57.2 18.3 65 149.0 127 260 500 410 770 1418
-9.4 15 59.0 18.9 66 150.8 132 270 518 416 780 1436
-8.9 16 60.8 19.4 67 152.6 138 280 536 421 790 1454
4.44 8 14.4 -8.3 17 62.6 20.0 68 154.4 143 290 554 427 800 1472
-7.8 18 64.4 20.6 69 156.2 149 300 572 432 810 1490
-7.2 19 66.2 21.1 70 158.0 154 310 590 438 820 1508
5.00 9 16.2 -6.7 20 68.0 21.7 71 159.8 160 320 608 443 830 1526
-6.1 21 69.8 22.2 72 161.6 166 330 626 449 840 1544
-5.6 22 71.6 22.8 73 163.4 171 340 644 454 850 1562
S.S6 10 18.0 -5.0 23 73.4 23.3 74 165.2 177 350 662 460 860 1580
-4.4 24 75.2 23.9 75 167.0 182 360 680 466 870 1598
-3.9 25 77.0 24.4 76 168.8 188 370 698 471 880 1616
To convert from any Celsius -3.3 26 78.8 25.0 77 170.6 193 380 716 477 890 1634
temperature to the -2.8 27 80.6 25.6 78 172.4 199 390 734 482 900 1652
Fahrenheit equivalent, -2.2 28 82.4 26.1 79 174.2 204 400 752 488 910 1670
select the temperature C -1.7 29 84.2 26.7 80 176.0 210 410 770 493 920 1688
from the center (boldface) -1.1 30 86.0 27.2 81 177.8 216 420 788 499 930 1706
columns, then read -0.6 31 87.8 27.8 82 179.6 221 430 806 504 940 1724
temperature F to the 0 32 89.6 28.3 83 181.4 227 440 824 510 950 1742
right. To convert from F 0.6 33 91.4 28.9 84 183.2 232 450 842 516 960 1760
to C, select the tempera- 1.1 34 93.2 29.4 85 185.0 238 460 860 521 970 1778
ture F from the center 1.7 35 95.0 30.0 86 186.8 243 470 878 527 980 1796
(boldface) columns, and 2.2 36 96.8 30.6 87 188.6 249 480 896 532 990 1814
read equivalent tempera- 2.8 37 98.6 31.1 88 190.4 254 490 914 538 1000 1832
ture C to the left. Use 3.3 38 100.4 31.7 89 192.2 260 500 932
factors above for inter- 3.9 39 102.2 32.2 90 194.0 266 510 950
polations.
4.4 40 104.0 32.8 91 ,195.8 271 520 968
Volume conversion factors.
\ to-
from \ cm3 litres m3 in.3 ft3 yd3 Fluid oz Fluid pt Fluid qt gaUU.S.) gal (U.K.)
cmf 1 0.001 1 X 10~6 0.06102 353 X 10"5 1.3X 10"6 0.03381 0.00211 0.00106 2.64 X 10"4 2.2 X 10"4
litre 1000 1 0.001 61.02 0.03532 0.00131 33.81 2.113 1.057 0.2642 0.220
m3 1 X 106 1000 1 6.1 X 104 35.31 1.308 3.38 X 104 2113 1057 264.2 220
in3 16.39 0.01639 1.64X 10' 1 5.79 X 104 2.14X 10'5 0.5541 0.03463 0.01732 0.00433 0.00360
ft 3 283 X 104 28.32 0.02832 1728 1 0.03704 957.5 59.84 29.92 7.481 6^.229
yd3 7.65 X 10s 764.5 0.7646 4.67 X 104 27 1 259 X 104 1616 809.7 202.0 108.2
fluid oz 29.57 0.02957 2.96 X 10"5 1.805 0.00104 3.87 X 10~5 1 0.0625 0.03125 0.00781 0.00651
fluid pt 473.2 0.4732 4.73 X 10~4 28.88 0.01671 6.19X 10'4 16 1 0.50 0.1250 0.1041
fluid qt 946.4 0.9463 9.46 X 10~4 57.75 0.03342 0.00124 32 2 1 0.25 0.2082
gal (U.S.) 3785 3.785 0.00379 231.0 0.1337 0.00495 128 8 4 1 0.8327
gal (U.K.) 4546 4.546 0.00455 277.4 0.1605 0.00595 153.7 9.608 4.804 1.201 1
Maximum boiler water silica concentration versus drum pressure at selected boiler water pH.
(Based on maintaining 0.02 mg/litre SiO2 maximum in steam.)
Flow rate conversion factors.

from
to-> litre/s gal/min ft 3 /s ft3min m 3 /s m3/min
litre/ s 1 15.85 0.03532 2.119 0.001 0.06
gal/min 0.06309 1 0.00223 0.1337 0.000063 0.00379
ft 3 /S 28.32 448.8 1 60 0.02832 1.699
ft 3 /min 0.4719 7.481 0.01667 1 0.000472 0.02832
m 3 /s 1000 15852 35.31 2118.6 1 60
m3/min 16.667 264.2 0.5885 35.31 0.01667 1
Index
A Sources, 196
Alpha radioactivity, 290-291, 190,
Abbe condenser, 290 411-417
Absorption, 75, 76 Counting, 411
Band, 303 Rays, 289
Characteristics, 294 Alumina, 183
Spectroscopy, 298-306 Aluminum, 141, 191, 210, 314
Acid mine drainage, 105 Amines, 120, 121, 188, 218
Acidity, 16, 68, 69, 200 Ammonia, 40-41, 176, 16, 185, 211
Chemistry, 200 Nitrogen, 212
Sources, natural, 68, 200 Ammonium, 185, 211, 212
Actinomycetes, 265 Analysis
Control, 266 Checking results, 247, 288
Activated carbon, 273 Deposits, 309
Adsorption, 188 Ground waters, 8
Activated sludge, 79-82 Interpretation of results, 249, 250
Activation, 208 Programs, 158, 160-166
Corrosion products, 374 Public waters, 314-315
Of water, 371 Reporting results, 284, 246
Water additives, 373 Scheme, 310
Water impurities, 372 Selecting methods, 159
Activity, induced, 208, 338 Surface waters, 8, 314, 315
Adsorption, 75, 76 Systematic, 279
Agglomeration, 49 Units of expression, 312, 313, 246
Air contact affects, 184, 185 Water, scheme for, 162-166
Air coolers, 93 Analytical programs
Alarm systems, 159, 160 Factors in selection, 178, 183
Algae control, 73, 266 Methods, checking analysis, 247,
Alkalinity, 15,189, 196,197,199, 201 248
Accuracy, 201 Nature and scope, 178
Chemistry, 197 Planning, 181
Concentration range, 201 Primary considerations in
Methods, 189, 20r planning, 181
pH relationship, 199 Reporting results, 246
459
460 MANUAL ON WATER
Sampling, 187 Deposits, 284

Anthracite filter, 48 Growths, 73
Apparatus, special, 183-187 Methods, 67
Aquarium, balanced, 98, 99 Microscope, 286
Aquatic life, 86, 89 Odor, 264
Life Advisory Committee, 105 Pollution, 39, 42, 73, 93
Aquifer, 96, 100 Systems, 329
Aromatic, 271 Tracing, 340
Assimilation, 104 Treatment, 79-82, 97, 98
Atomic absorption spectroscopy, Biota, 112-115,98,99
272, 277, 193 Birefringence, 291
Atomic weights, 443, 444-445 Bisulfite, 242
Autumn, 112 Blanket materials, 333
Autotrophs, 102 Blowdown, 52
Automatic monitoring, 145 Boiler water, 59
Sampling, 126-128
B Silica content, 450
Boiling water reactor (BWR) power
Background, diffraction, 265 plant, 353, 365-369, 372-374,
Odor, 150, 256 376-383
Radioactivity, 301, 331 Boilers, process, 19-23
Bacteria, 13, 58-60, 85, 88, 89, 101 Bromide, 191, 216
Iron, 208 Bromine, 203
Pathogenic, 73, 103 For sterilization, 72
Sulfate-reducing, 209 Buffer capacity, 13, 14
Bacterial cells, 209 Bunsen spectroscope, 300
Bacteriophage, 103
Barium, 191, 213, 313 c
Radioactive, 335
Basic compounds, 277 Calcium, 44, 204, 217, 313
Oxides, 278 Calibration, 186, 415, 426, 439
Taste, 267 Canal coolants, 336
Basicity, 210, 15, 68, 69 Carbon dioxide, 15, 70, 214, 312
Bass, 113, 114 Removal, 70
Beta radioactivity, 190, 327,418-432 Carbon-14, 331, 339
Counting, 423 Carbonate, 58, 59, 214, 215, 278
Beverage industry, 44 Carbonate-bicarbonate-hydroxide,
Bicarbonate, 58, 59, 214 448
Calculation, 215 Carryover, 53
Bicarbonate-carbonate-hydroxide, Cascade, 90
448 Cathodic protection, 77
Biochemical oxygen demand (BOD), Centrifugation, 49
51 Checking analyses, method of, 247,
Biological and other tests, 209 248
Corrosion, 78, 79 Chemical
INDEX 461
Conversion factors, 447 Control, water treatment process,

Industry, 20, 26 125
Microscope, 287, 290 Conversion factors, 447, 448, 449,
Oxygen demand, 195 454, 455, 456, 457
Treatment, 57, 70 Formulas, temperature, 448
Chloramines, 72 Conveying, 23
Chloride, 15, 219, 192 Coolants
Chlorination, 60, 71, 271 Canal, 336
Chlorine, 66, 202 Chemistry, 358, 362, 368
Combined, 72, 220 Cooling, 84, 85, 92, 93
Concentration, 72, 192, 220 Tower, 20, 52, 93
Demand, 202, 220 Water, 20, 84, 85, 92, 93, 42
Dioxide, 72 Copper, 192, 222, 314
Free, 72, 220 Corrosion, 44, 40, 77, 78, 79
Removing gases, 71 Analyzer, 174
Requirement, 202 Barriers, 77
Residual, 72, 202 Biological, 78
Total available, 220 Inhibition, 77, 94
Chromate, 195 Products, 94, 307
Chromium, 192, 221, 314 Corrosiveness, 16, 40, 41, 94, 174
Cleanliness, 186 Corrosivity, 189
Climate, 112, 113 Cosmic ray products, 331
Coagulation, 74, 75, 49, 189 Coulometric instruments, 173, 174
Cobalt, radioactive, 339 Counters, radioactivity, 392, 400,
Color, 16, 43, 44, 170, 173, 202 402, 403, 411,412,413,423,
Colorimetric analyzer, 202 433, 434
Combustion, wet, 66, 67 Alpha, 419
Comparison tests, odor, 261 Beta, 423
Duo-trio, 261 Gamma, 433, 434
Paired, 261 Criteria, water quality, 105
Series, 261 Crustaceae, 29, 103
Triangular, 261 Crystal growth deactivators, 76
Concentration, biological, 340 Crystalline compounds, indentifica-
Condensate, 21 tion of, 292-298
Condenser, Abbe, 290 Crystallization freezing, 56
Conductance Curie, 328
Ionic equivalent, 447 Cyanide, 225
Specific, 168
Conductivity, 168, 203 D
Electrical, 168, 203
Conservation, 97 Dall flow tube, 134
Constant differential meter, 136 Data handling, 142
Consumption, 84, 93 Transmission, 143
Consumptive use, 6 Deaeration, feedwater, 50, 51
Contaminants, 8-16 Vacuum, 51
462 MANUAL ON WATER
Dechlorination, 71, 273 Displacement meter, 139

Decomposition Disposal of nuclides, 341-343
Aerobic, 79-82 At sea, 343
Anaerobic, 72 Underground, 343
Zone, 104 Dissolved gases, 15
Degasification, 50, 70 Dissolved oxygen, 15, 44, 70, 447,
Degradation zone, 104 448
Deionizer, mixed-bed, 62 Matter, 9, 10, 14, 43, 44
Delay time, 231,237-239 Saturation values, 447, 448
Demineralized water, 57-66 Solubility versus temperature, 447,
Demineralizer, 57, 61-66 448
Density, 169, 446 Distillation, 54, 55
Density of water, 169, 446 Compression, 54
Deposits, 93, 94 Drinking water standards, EPA, 180,
Analysis of X-ray fluorescence, 263, 348
307 Drug industry, 27
Examination, 283 Duo-trio comparison tests, odor, 261
Formation by impurities in steam,
42, 128, 450, 457 £
Identification, 283-286
Sampling, 283 Ebullition, 50
Desalination, 5, 55 Effective size, 48
Demonstration plants, 5, 56, 57 Effects, physiological, 31, 71, 102,
Distillation, 5, 54, 56 103
Electrodialysis, 5, 56 Mineral acid, 100
Freezing, 5, 57 Effluent quality, 147, 148
Detectability, 344, 407, 408 System, 161
Detectors, radiation, 344, 389, 393- Electrical conductivity, 168, 203
399,411,437 Electrode, glass, 171, 190, 249
Devices, flow measurement, 124, Specific ion, 172, 192
127, 129 Electrodialysis, 5, 56, 79
Dew point, 108 Electronics industry, 32, 33
Dialysis, 79 Electron volt, 328
Diatomaceous earth, 49 Electroplating, 25
Dichromate, 195 Electrostatic interaction, 328
Die-away, 112, 113 Emission spectroscopy, 268, 277
Differential head meter, 136 Emulsion breaking, 61
Diffractometer, 298 Energy, 108, 110
Diffractor, 298 Transfer, 329
Dilution, 51 Entrainment, 53
Discharge, controlled, 5 Environmental Protection Agency,
Ocean, 55 (EPA), 145, 180, 263, 348
Underground, 35 Epilimnion, 112
Disinfection, 71, 72, 73 Equilibrium, dynamic, 97, 98, 100,
Dispersing instrument, 300 101
INDEX 463
Equivalent conductance, 447 Flow measurement, 134-143

Esthetics, (aesthetics), 115, 116 Dall How tube, 134
Evaporation, 19, 54, 55, 84, 92, 93 Devices, 124, 127, 129, 133, 136-
Of radioactive wastes, 341 143
Evapotranspiration, 6 Differential pressure flowmeter,
Exchange capacity, 61, 62, 67 136
Extracts, 188, 189, 310 Flumes and weirs, 136, 137
Extraction, organic, 188, 189 Open channel
Extractable matter, 188 By Parshall flume, 136
By weirs, 137
F Points, 135
Primary units, 140
Fallout, 337, 338 Selection, 140
Fatigue Types, 140
Physical (sensory testing), 256 Transmission, 143
Physiological (sensory testing), Systems, 127
256 Venturi meter tube, 139
Fauna, 112, 115 Flow rate, conversion factors, 327
Feedwater heaters, 50, 51 Flowmeter, inertial, 140
Ferric sulfate, 74, 75 Fluoride, 192, 226
Ferrobacillus ferrooxidans, 105, 208 Flumes, 136, 137
Filter Fog, 20, 93
Calcite, 68 Food industry, 44
Cloth, 48 Supply, 114
Contact, 68, 71 Fungi, 16
Metal gauze, 48
Precoated, 49 G
Pressure, 48
Rotary drum, 48 Galvanic cell, 244
Slow sand, 48 Gamma photon, 329
Trickling, 80, 81 Counting, 433-434
Filtration, 48, 79 Radioactivity, 328, 329, 433-440
Fish, 113, 114 Rays, 328
Fish toxicity, 209 Gases, dissolved, 15
Fishing, 113, 114 Gauze, metal, 48
Fission, 337 Generation of power, 21-23, 42
Products, 337, 370 Geometry, radioactive, 392
Flame photometry, 277, 282 Geothermal gradient, 109
Floats, 136 Glass electrode, 190, 299
Floe, 74, 75 Ground water, 86, 87, 109
Flood storage, 110
Flora, 107, 113 H
Flotation, 49
Flow, 134 Habitat, 112-115
Flowing samples, 125 Halogens, 66, 202, 203, 220
464 MANUAL ON WATER
Hardness, 9-13, 44, 57-60, 171, 203, Index, refractive, 169

204 Inductor, 300
Carbonate, 9 Industrial pollution, 38, 40, 86
Noncarbonate, 9 Industry, 38-45, 86
Soap, 203 Beverage, 43, 44
Hazards Chemical, 43
Biological, 73, 103 Drug, 44
Radioactive, 347 Food, 4, 44
Head area meter, 136 Lumber, 23
Heat, 19,20,86, 107-117 Mining, 23
Heat exchangers Paper, 25, 43
Corrosion of, 40 Petroleum, 43
Deposits on, 40 Power,~2I, 26
Heat of fusion, 17 Steel, 44
Process boilers, 19 Sugar, 24, 44
Transfer, 19, 20 Textiles, 26, 43
Vaporization, 19 Inertial flowmeter, 140
Heavy elements, 330 Infrared, 108, 281
Hedonic index, 261 Injection wells, 86, 87
Heterotrophs, 101 Inorganic chemicals, odor, 263
Hydrated oxides, 278 Insolation, 109
Hydraulic clutch, 27 Instrument housing, 149-160
Power, 98, 27 Instrumental methods, 145, 146, 167
Tracing, 339 Instrumental-sensory correlation,
Turbines, 27 225
Hydrazine, 185, 192 Interaction, electrostatic, 328
Hydroelectric power, 110 Interconversion factors (mg/liter to
Hydrogen, 192 epm), 450
Cycle, 62 International Commission on Radio-
Dissolved, 192 logical Protection, 329
Ion, 10, 248 Interpretation of analytical results,
Sulfide, 40, 180, 16, 311 249
Hydroponics, 59 Inversion, 100, 112
Hydrozides, 192, 214, 279 Iodide, 191, 216
Hydroxide-bicarbonate-carbonate, Iodine, number, 191, 216, 311
448 Radioactive, 339
Hypolimnion, 100, 112 Ion exchange, 61-66
Exclusion, 65
I Retardation, 65
Ionic equivalent conductance, 447
Ignition loss, 311 lonization detector, 345, 412
Impurities, 8-11, 35-37 Iron, 14, 44, 60, 193, 227, 314
Affecting power industry, 21, 22, Bacteria, 208
23, 35-37 Ferrous, 105
In steam, 42 Iron-59, 339
Incineration, 73, 74 Irrigation, 86, 87
INDEX 465
Isothermal, 112 Constant differential, 131, 137

Differential head, 130, 136
J
Displacement, 131, 137
Functional, 136
Jackson turbidimeter, 190 Gentile flow tube, 136
Head-area, 136
K Inertial, 140
Orifice plates, 136
Kjeldahl nitrogen, 212
Pitot tube, 137
Pumps, 137
L Rotameter, 137
Laboratory practices, 245-248 Velocity, 136
Lagoons, 111 Methods
Lakes, 110,226 Checking analyses, 247
Land subsidence, 86, 87 Reporting results, 246
Lattice dimensions, 266 Microbiologic flora, 98, 191
Lead, 192, 230 Microorganisms, 16, 39, 42, 98, 191
Lime-soda softener, 57 Microscope
Lumber industry, 23 Biological, 286
Chemical, 287, 290
M Electron, 302
Eyepieces, 287
Magnesium, 44, 209, 217,313 Mechanical stage, 287
Maintenance, monitoring instru- Objectives, 287
ments, 175-176 Petrographic, 290-293
Makeup, 21, 357 Mine drainage, 105
Manganese, 14,44, 60, 193, 230, 231, Mining industry, 23
313 Misapplication of instruments, 149,
Radioactive, 339 150
Zeolite, 67 Mixed-bed deionizer, 62
Mathematical evaluation, odor tests, Moderator, 334
259 Molds, 16, 101
.Matter Molecular absorption spectroscopy,
Extractable, 188 271
Flotant, 49 Monitoring, 145-176
Oxidizable organic, 51, 188, 195, Monitors, radiation, 345
220 Multiple recycle reuse system, 89
Solvent extractable, 188 Municipal pollution, 39, 111
Suspended and dissolved, 9,10,14,
43, 44, 206, 207 N
Mercury, 44
Metabolic byproducts, 288 National Committee on Radiation
Metal oxides, 279 Protection, 329
Metals, 279 Natural contamination, 38, 329
Meters Natural purification, 98
Classes of, 130-131, 136-137 Radioactive series, 330
466 MANUAL ON WATER
Nephelometry, 170, 190, 207 Statistics, 259-261

Nesslerization, 211 Synergism, 270
Neutralization, 68, 69 Terminology, 268
Neutron, 334 Testing laboratory, 268
Emitting fission products, 337 Threshold, 204, 260, 268
Moderator, 334 Triangular comparison tests, 261
Nickel, 193, 231, 314 Ohio River Valley Water Sanitation
Nitrate, 19, 212, 232 Commission (ORSANCO),
Nitrite, 18, 212, 233 86, 161
Nitrobacter, 102 Oil production by recovery, 24
Nitrogen, 191, 234 Oils, 189
Ammonia, 191, 212 Once-through-steam generator
Kjeldahl, 212 (OTSG), 380, 392, 383
Organic, 213 On-site reuse, 84-96
Total, 212 Open channel flow measurement,
Nitrosomonas, 102 136
Noncarbonate hardness, 9 Weirs, 137
Nozzle, 124 Optical properties, 288-293
Nuclear disintegration, 327 Spectroscopy, 299
Power reactors, 352-383 Ores, 23, 332
Nuclides, 327 Organic matter, 16
Tables, 330, 33, 334, 339 In deposits, extracted, 311
Nitrogen, 213
O Salts, 280
Substances, odor, 262
Objectives, waste water sampling, Organoleptic testing, 254-275
131 Mathematical evaluation, 259
Odor, 44, 66, 73, 76, 189, 204, 254- Orthophosphate, 193, 237
275 Outline
Additivity, 270 Principal sample groups, 184
Antagonism, 270 Typical scheme of deposit anal-
Background, 256, 269 y&es, 310
Biological, 264 Outputs, monitoring systems, 159-
Characterization, 267 166
Intensity, 267 Oxidation
Index, 268 Aerobic, 36, 66, 79-82, 111
Mixtures, 270 Catalytic, 66, 67
Organic substances, 262 Pond, 66
Paired comparison tests, 261 Oxidation-reduction (REDOX)
Panelists, 254, 255 potential, 172, 190, 205
Parameters, 268, 269 Oxygen, 193, 234
Profile evaluation, 256, 261 Demand, 188, 195
Quality, 260 Chemical, 195
Series tests, 261 Total, 188
Supra threshold, 261 Dissolved, 111, 193,234
INDEX 467
Saturation values, 447, 448 Municipal, 39

Scavengers for, 68 Organic, 38
Removal, 40, 41, 57, 58 Physical, 39
Ozone, 66, 67 Thermal, 43, 92, 93, 107-117
Types, 38
p Pond, spray, 66
Potassium, 14, 15, 143, 244, 312
Paired comparison tests, odor, 261 Potential, oxidation-reduction, 172,
Panelists, odor, 254, 255 190, 205
Paper industry, 24, 25, 43 Power
Parshall flume, 136, 137 Generation of, 21-23, 42, 98
Paniculate matter, 206, 207 Hydroelectric, 110
Petroleum industry, 43 Industry, 21, 26
pH, 10, 16, 39, 171, 190, 249 Steam-electric, 27, 98, 375, 376,
Phenol, 219, 189 380
Phenolic compounds, 189, 273 Practices, laboratory, 245-248
Phosphates, 59, 66, 193, 235, 314 Precipitation, 3, 57
Complex, 237 Precision, 201, 245-248, 250, 404-
In deposits, 280 406
Meta, 236 Preservation of samples, 132
Ortho, 237 Pressurized water reactor (PWR),
Pyro, 237 353, 357-364, 369, 375
Tripoly, 236 Priming, 53
Phosphorus, 193 Principal sample groups, outline of,
Photographic detectors, 345 184
Photometry, flame, 307, 312 Process control, 163
Photon, gamma, 329 Instrumental, 163-166
Photosynthesis, 114 Integration, reuse systems, 90
Physical fatigue Uses, 21, 23, 25, 26, 43, 44
Methods of measurement, 231, Processing, water for, 6, 21-24, 43
235, 239 Product, water in, 26, 27
Taste and odor testing, 256 Profile evaluation, odor, 256, 261
Treatment, 46-57 Program of analysis, 177-188
Physiological fatigue (sensory Properties, physical, 18, 189, 190,
testing), 256 202, 203, 205, 206, 445, 446
Phytoplankton, 116 Thermodynamic, 18
Pilot tubes, 137 Protozoa, 103
Planning a reuse system, 95 Purification, natural, 98
Pollution, 38, 39,99, 100 Pyrogen, 31
Biological, 39, 42, 73, 93
Chemical, 39, 43
Defined, 39 Q
Effects on industry, 38-45
Heat, 107-117 Quality criteria, 105
Industrial, 38 Effluent, 147, 148
468 MANUAL ON WATER
Monitoring, stream and river, 147, Recovery zone, 104

148 Recycle, 84-96
Problems in reuse, 92 Recreation, 115
Corrosion and mineral de- Refractive index, 169
posits, 94 Rem, 347
Heat dissipation, 112, 113 Reporting results, analytical, 246
Organic deposits, 93 Reservoirs, 4, 6, 7, 8, 110
Quantal response, 259 Response to sensory stimulation
(Weber-Fichner law), 261
R Results, reporting, 246
Return flow, 72-74, 81
Rad, 329 Reuse, future outlook on industrial,
Radiation, natural, 108, 109 81
Hazards, 347 Profitable aspects, 80
Infrared, 108, 281 Quality problems, 79, 80
Monitors, 345 Reuse of return flow, 86
Ultraviolet, 72 Reuse systems, on-site, 84-96
Types, 327, 328 Cascade system, 90
Radioactivitation methods, 190 Kaisers' Fontana mill, 90
Sampling, 130 Multiple recycle, 89
Storage, 341, Planning, 95
Treatment, 341, Process integration system, 90
Wastes, disposal, 341, 542 Sewage, 87
Radioactivity, 130, 190, 208 Simple recycle, 89
Measurement, 387-440 Tandem, 88
Radionuclides, 327, 329 Reuse systems, use of industrial
Natural, 330-332 waste, 88
Of radium, 329, 330, 332 Rivers, 108, 112
Radiotracers, 338 Rotameter, 137
Radium, 331 Rotifera, 103
Radium-226, 331 Runoff, 3, 4
Radon, 331
Rain, 108 S
Rainout, 337, 338
Ranking, sensory testing, 259 Safety, laboratory, 245, 248, 347
Raw material, water as a, 26 Saline water, 5, 55, 56, 57, 74
Reactor, components, 354 Salt, 113
Reactor Concentration, 136
Coolant, 334 Sample
Fuels, 333 Preparation, odor, 257
Materials of construction, 356 Preservation, 132
Readout systems, 159, 160 Radioactive, 130, 343
Reagent water, 34 Sampler, automatic, 119
Recharge, 86, 87 Foerst, 130
Recovery, oil production by, 24 Kemerrer type, 130
INDEX 469
Sampling, 118-133 Indicator, 142

At atmospheric pressure, 119-123 Recorder, 142
At elevated temperature, 126-127 Sampling, 143, 401
Boiler water, 126-128 Totalizer, 142
Composite, 118 Totalizer-indicator-recorder
Considerations, instrumental combinations, 143
monitoring, 149-160 Sedimentation, 46
Continuous, 119 Sediments, 319-326
Control of water treatment Self-purification, 97
processes, 125 Sensory stimulation, response to,
Deep, 130 264
Deposits, 283-286 Sensory testing, 254-275
Equipment, 132, 133, 134 Separation, oil-water, 49, 50
Instantaneous, 118 Phase, 49, 50
Manual, 118 Sequestration, 65, 66
Grap, spot, or snap, 118 Sewage, 38, 39, 79-82, 87, 101, 111
Method of, 118 Sewer disposal of radionuclides, 341
Nozzle, 124 Shielding, 347, 389, 423,
Points, location, 132, 151 Significance of reuse, 84
Precautions, 133 Silica, 11, 14,v 165, 194, 238, 280,450
Preparation, 346 In boiler water, 450
Procedures, 131 In deposits, 280
Process control, 125 In steam, 450
Radioactive water and wastes, Removal, 63
130, 344 Silicates, complex, 281, 282
Steam, equipment for, 126-128 Simple, 281
Stations, 124 Silver iodide, 73
Subatmospheric pressure, 128, Skin samples, 124
129 Sludge, 277
Superatmospheric pressure, 123 Activated, 79-82
Waste water, 131 Blanket, 61, 62
Equipment for, 133, 134 Soap hardness, 203
Water cooling coil, 127 Sodium, 14, 15, 193, 244, 312
Water-formed deposits, 276, 283- Iodide detector, 437
286 Softener, lime-soda, 57
Saponiflcation number, 311 Softening, 57, 53, 57, 59
Scalar responses (sensory testing), Solid solutions, 296
259 Solids, dissolved, 207
Scale formation, 307 Solubles constituents of deposits,
Schmutzdecke, 79 281
Scintillation detector, 345, 414 Solutions, solid, 296
Secondary instrumentation Special apparatus for the laboratory
By transmission type (mechanical, methods presented, 183-187
electrical, pneumatic), 143 Specific gravity, 20
By type information desired, 143 Specific ion electrode, 172
470 MANUAL ON WATER
Spectrograph, 307 Deposits, 282, 318

Spectrographic analysis, 307 Reducing bacteria, 209
Spectrometer, 393 Sulfide, 239, 241, 282
Spectrophotometric examination, Sulfite, 242
infrared, 281 Sulfur dioxide, 242, 244
Spectroscope, Bunsen, 300 Sulfuritic material, 105
Spectroscopy, absorption, 298-306 Summer, 112
Atomic absorption, 193, 272, 277 Suprathreshold odor, 256
Emission, 268, 277 Surface tension, 240, 241, 170
Optical, 299 Surface water, 3, 4, 7, 8
X-ray, 307 Surfactant, 77
Spectrum, 108 Surveilance systems, 160
Spores, 98 Suspended and dissolved matter,
Spot tests, 308 206, 207
Spray irrigation, 87 Suspended solids, 206, 207
Ponds, 87
Stripping, 50, 51 T
Spring, 112
Springs, 109 Tandem reuse, 88
Statistical uncertainty, 344, 406 Taste, 103, 147, 16, 44, 266, 267
Statistics, counting, 404, 405 Temperature, 235, 239, 242, 244,
Steam 113, 169
Deposit forming impurities, 42, Conversion tables, 48, 454, 455
128, 450, 457 Effects of, 113, 119
Flashing, 52 Tests
Generators, 375, 376, 380 Qualitative, 308
Purity, 128 Spot, 308
Sampling, 126-128 Textiles, 26, 43
Water for, 42 Thermal loading, 108-117
Steel industry, 44 Patterns, 112
Sterilization, 71, 72, 73 Thermocline, 100, 112
Storage of radioactivity, 342 Thermodynamic properties, 18, 19
Straining, 49 Thorium, 291, 294
Stratification, 110 Threshold, 255, 256, 268, 271
Stream and river quality monitor- Number, 268
ing, 147, 148 Odor concentrations, 268
Streams, 3, 108, 112, 147, 148 Odors, 271
Streptomyces, 205 Treatment, 75, 76
Stripping, 50, 51 Quality, 260
Strontium, 191, 213, 194 Terminology, 268
Radioactive, 339 Time constants, 297
Strontium-90, 339 Tin, 314
Subjective properties, 255 Titanium, 314
Sugar industry, 24, 44 Titrator, 174
Sulfate, 15, 241, 242 Toxicity, 209, 190
INDEX 471
Diatoms, 190 Water

Fresh water fish, 209 Analyses, typical 10, 452, 453
Treatment, biological, 79-82, 97, 98 Analysis, 177-253
Chemical, 57-73 Scheme, 184-194
Physical, 46-56 Units of expression, 247, 248,
Radioactive wastes, 340-343 447, 448, 449
Triangular comparison tests, 261 As raw material, 26
Trickling filter, 80 Boiler, 12
Tritium, 331 Sampling, 126-128
Trout, 112, 113, 114 Consumptive use, 6
Turbidimeter, Jackson, 207 Cooling, 20, 42, 84, 85, 92, 93
Turbidity, 16, 170, 190, 207 Demineralized, 57-66
Absolute, 170 Demand, 6
Turbines, hydraulic, 27 Cooling, 6
Domestic, 6
Industrial, 6
U Irrigation, 6
Ultimate disposal of radioactive Public, 6
wastes, 341-343 Rural, 6
Ultrapure water, 28-37 Density of, 446
Ultraviolet, 108 Formed deposits, 316-318,
Ultraviolet sterilization, 72 Examination of, 307, 308, 309
Units, 247, 448, 449, 457 Sampling of, 307
Uranium, 330-333 Formed sludges, 279, 280-284,
Use, 84-92 307
of industrial water, 18 Ground, 86, 87, 109
U-tube stream generator (UTSG), Hardness, 9-13
376, 377, 381 Irrigation, 86, 87
Ocean, 5, 55-59
Processing, 6, 21, 24, 43
V
Quality, 8, 9, 92-94, 105,112, 113,
Vanadate, 314 147, 148
Vanadium, 314 Quality criteria, 105
Vapor, water, 97 Quality factors, 8-16
Velocity meter, 136 Reagent, 34
Venturi tube meter, 139 Recharge, 86, 87
Viruses, 98 Recreational use, 115
Viscosity of water, 445 Requirements, industrial, 4, 21-
Volume conversion factors, 456 26, 43, 44, 86
Resources, 3, 4, 6
Reusable supplies, 84, 89
W Runoff, 3, 4
Washing, 24 Saline, 5, 17
Waste, radioactive, treatment of, Sampling, 118-133
340-342 Standards, 180, 263, 221
472 MANUAL ON WATER
Storage in reservoirs, 7 Secondary, 307

Supply, 3, 4 Spectroscopy, 307
Surface, 3, 4, 7, 8,
Vapor, 97
Viscosity of, 445 Y
Weber-Fichner law, 261
Weirs, 136, 137 Yeasts, 98
Wells, 86, 87
Wind, 108, 112 Z
Windage loss, 52
Winter, 112 Zeolites, 61, 62, 67
Withdrawals, 6, 84, 85 Zinc, 245, 314, 315
Zones, 103, 104
X Decomposition, 104
X-ray diffraction, 293, 294, 307 Degradation, 104
Fluorescence, 307 Recovery, 104
LEHICH UNIVERSITY LIBRARIES
3 =1151 Q O D L 4 D 2 7 2 3

STP 442a-1978

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

STP 442a-1978

Uploaded by

Copyright:

Available Formats

MANUAL ON WATER

ASTM SPECIAL TECHNICAL PUBLICATION 442A

List price $28.50

AMERICAN SOCIETY FOR TESTING AND MATERIALS

Printed in Tallahassee. Fla.

This manual is intended as brief reference source of information on water.

Chapter 1—Water Sources and Supply

Water is one of the most important engineering materials used by

Precipitation—The Basic Source

Runoff—Source of Surface Supplies

Saline Water—Prospective Future Source

TABLE 1 Saline water demonstration plants.

TABLE 2— Summary of reservoirs showing usable storage.

Connecticut 413 500 135

Illinois 203 700 66.4

Louisiana 1 307 600 426

Minnesota 3 822 500 I 246

Nevada 673 300 219

North Carolina 3 112 600 1 014

Pennsylvania 1 739 500 567

South Carolina 3 759 800 1 225

Texas 18 852 900 6 143

West Virginia 932 800 304

Total 278 121 200 90 626

Quality Considerations in Industrial Supplies

General Quality Factors

specific industrial applications. Hardness also is an important factor.

Bicarbonate (HCOa) 25 92 128 140 179 78 332 254

Dissolved solids 59 129 222 657 410 87 284 318

Other Quality Parameters

Arizona 4 Phoenix 250

^District of Columbia 18 Washington 115

Florida 19 Jacksonville 280

Illinois 26 Chicago 140

Indiana 28 Evansville 155

Iowa 33 Des Moines 95

Kansas 34 Kansas City 250

Kentucky 37 Louisville 110

Michigan 45 Detroit 100

Missouri 51 Kansas City 85

Nebraska 53 Lincoln 180

New Jersey 55 Jersey City 40

New Mexico 58 Albuquerque 99

New York 59 Albany 52

North Carolina 65 Charlotte 24

Ohio 67 Akron 130

Oklahoma 74 Oklahoma City 130

Rhode Island 80 Providence 28

Utah 93 Salt Lake City 180

mg/litre and averages from 5 to 15 mg/litre. Under certain conditions, silica

Potassium is ordinarily present in freshwater in low concentrations, seldom

Hydrogen sulfide may range from 0.0 to several mg/litre in groundwaters

Reuse of water is being practiced widely in industry. The practice will

Chapter 2—Uses of Industrial Water

Some Thermodynamic Properties of Water

temperature rise would be twice as great or twice the volume would be

Basic Water Uses

TABLE 1 — Some industrial process heat hollers.

cost of modern circulating hot-water systems. However, many factors are

FIG. 1—Paper mill steam system.

artificially generated hydraulic power for special applications. The most

Role of Water in Industrial Processes

Conveying—Just as our watercourses convey a natural burden of sus-