AMERICAN SOCIETY OF CIVIL ENGINEERS

Founded November 5 , 1852

TRANSACTIONS
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

Paper No. 3179

DIFFUSERS FOR DISPOSAL OF SEWAGE IN SEA WATER

By AM Rawn,1 Hon. M. ASCE, F. R. Bowerman,2 M. ASCE, and

Norman H. Brooks,3 M. ASCE

With Discussion by Mess r s. J. M. Jordaan, Jr; C. H. Lawrance; and AM Rawn,

F. R. Bowerman, and Norman H. Brooks

SYNOPSIS

F or large sewerage systems the eff ectiveness of sewage disposal by dilu-

tion in the ocean depends on effi cient dispersal of sewage into the r eceiving
waters. By the us e of lar ge multiport diffus ers on the ends of their outfalls,
the Los Angel es County Sanitation Districts have, in r ecent years, gr eatly im-
proved the characteristics of the ocean water near the disposal site despite
steadily incr easing discharges of primary efflu ent.
This paper includes a history of the ocean outfall system operated at Whites
Point, Ca lif. , by the Los Angel es County Sanitation Districts; an outline of cer-
tain t echniques developed to predict dilutions of sewage effl eunt discharged in
s ea water; a de s cription of the use of multiport diffusers for improving initial
dilution; and a proc edur e for succ essful hydraulic design of large diffus ers .

INTRODUCTION

The disposal of sewage efflue nt to bodies of water for dilution and final dis-
posal r equires an effectiv e mingling of the effluent with the water body to
Note .-Published essentially as printed here, in March, 1960, in the Journal of the
Sanitary Engineering Division , as Proceedings Paper 2424 . Pos itions and titles given
ar e those in effect when the paper or discussion was approved for public ati on in Trans -
actions.
1 Engrg. Cons., Retired Chf. Engr . and Genl. Mgr. , Los Angel es County Sanitation
Dists., Los Angel es , Calif.
2 Asst. Chf. Engr., Los Angeles County Sanitation Dists., Los Angeles , Calif.
3 Assoc. Prof. of Civ. Engrg. , California of Tech., Pasadena Cal if.; Cons. to Los
Angel es County Sanitation Dists., Los Angeles , Calif.

344

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 345

achieve the following purposes: ( 1) the effective oxidation of the fine suspen-
soids and dissolved organic compounds; (2) the reduction of bacterial popu-
lation; (3) the prevention of odor nuisance; and (4) the removal of particles
by sedimentation with natural detrital bottom materials.
The disposal of sewage effluents into water bodies involves hydrodynamic
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

relationships relating to the dispersal of one fluid into another; increased com-
plexity results in the ocean from the substantial differences in fluid densities
which cause the effluent to rise toward the surface of the salt-water body. For
reasons to be cited subsequently, there are significant advantage in dispersing
sewage effluent near the floor of the ocean as a means of increasing initial di-
lution. Some of the concepts embodied herein are applicable to the disposal
of sewage effluent to freshwater bodies; however, disposal to the ocean is the
primary subject of this presentation.
Where the favorable exchange of water through tidal or current movements
exists, primary sewage treatment will usually afford adequate protection to the
receiving waters, often without the necessity to chlorinate. Since primary
treatment involves sedimentation processes which remove floatable constituents
of sewage as scum from the surface of the sedimentation tanks while at the
same time removing readily settleable solids from the bottom of the tanks, it
can be seen that the primary treatment process eliminates the common com-
plainst voiced against sewage disposal by dilution, to wit, the building; of bottom
sludge beds or the fouling of surface waters with floatable debris. No such
similar protection is afforded by fine screening plants since these do not sepa-
rate material according to their specific gravity, but only as to size. Floating
rafts of fine sewage debris will be the almost inevitable result of the disposal
of finely screened effluent into a body of ocean water.
The security of recreational ocean waters with respect to public health may
usually be guaranteed through disinfection by artificial means such as chlorin-
ation, or by designing the outfall so as to prevent sewage effluents from reach-
ing bathing areas in dangerous concentration. Where sufficient oxygenated
diluting water is available to achieve initial dilutions of, say, 50 or more, there
appears to be no need for more reduction of organic content of the sewage ef-
fluent than that provided by primary treatment, in order to prevent odors or
other septic nuisances. Such dilutions are often found available where primary
sewage effluent is proposed for discharge to ocean bodies.

OUTFALL SYSTEM OF LOS ANGELES COUNTY SANITATION DISTRICTS

The performance of a particular outfall or system of outfalls must be evalu-

ated in terms of the physical circumstances extant at the site of the installa-
tion. No direct comparisons can, or should, be drawn between ocean outfalls
unless consideration is given to the similarities and differences in the systems
so compared.
Site Characteristics. -The Los Angeles County Sanitation Districts' outfalls
extend off Whites Point, a small promontory on a peninsula of land extending
into the Pacific Ocean southwestward from Los Angeles. The outfall site was
chosen subsequent to work done by District engineers in the early 1920's; how-
ever, 6 miles of tunnelling, under a range of low mountains was required to
reach the shoreline, so that the first ocean outfall was not commenced until
1935 and not completed until 1937. In 1958 the outfall system was disposing

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 346 SEWAGE DIFFUSERS

of the sewage from an area of almost 600 sq miles with a population of more
than 2,500,000. Characteristics of the outfall site are as follows:
1. The rocky Palos Verdes Peninsula, comprised largely of shales and
schists, juts about 5 miles into the Catalina Channel from the Southern Cali-
fornia coastline, terminating at its most offshore face in a cliffwhich rises a
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

distance of 80 ft or more abruptly from the ocean's edge.

2. The bottom, offshore from the peninsula, extends seaward at a slope of
about 1 in 50 as a wave-washed ledge of rock, becoming sandy at a water depth
of about 30 ft or 40 ft, with detrital muds commencing at a water depth of about
80 ft to 90 ft, remaining as mud bottom further offshore into the Catalina Chan-
nel.
3. Strong littoral currents sweep the promontory, running predominantly
parallel to the shoreline, with occasional weak currents of short duration mov-
ing toward shore.
4. Weak winds, usually from the west, cause movement of surface water
toward shore, but have as their principal effect the establishment of a mass
transport of water such as to induce upwelling of cold bottom water near the
shore. The favorable effect of this latter phenomenon will be investigated sub-
sequently.
The Outfall System.-At Whites Point, the Los Angeles County Sanitation
Districts have constructed three ocean outfalls; one built in 1937, 60 in. in di-
ameter and extending 5,000 ft from shore to a depth of 112 ft; a second, built
parallel to the first, in 1947, having an internal diameter of 72 in., originally
terminating the same distance from shore, but later lengthened to 6, 800 ft from
shore in 1953, at which time a diffusion structure was also constructed on the
line; and a third, 90 in. in diameter, completed in 1956, extending 8,000 ft from
shore to a depth of 212 ft, terminating in an even more efficient diffusion struc-
ture than that placed on the 72-in. outfall. The hydraulic capacity of the three
pipes is about 500 mgd. The outfall system is designed for disposal of primary
effluent, depending on natural dispersion and oxidation in the receiving waters.
The outfalls are shown in Fig. 1 and 2.
The three ocean outfalls have been built as trenched conduits through the
rock section of the ocean's floor, becoming semiflexible, articulated pipelines
laid on the surface of the sand and mud at greater ocean depths.
All three lines have been constructed from cast reinforced concrete pipe
with meehanite cast iron tongue-and-groove or ball-and-socket joints. Exam-
ination of the 60-in. pipe in 1957 showed that the cast iron had corroded at the
rate ol about 10 mils annually, so that penetration of corrosion (that is, graph-
itization) of about 0.2 in. has occurred during the approximately 20 yr of sub-
mersion. No significant loss in strength has yet been occasioned, although
continued graphitization at the same rate may cause joint failure in another 40
yr to 50 yr. The most recently placed outfall has sacrificial magnesium anodes
attached to the cast iron end rings. Concrete cores bored from the 60-in. di-
ameter outfall show that the crushing strength of the concrete exceeds that
measured more than 20 yr ago.
Efflu ent Disposal Without Diffus ers. -In 1933, the Districts commenced a
routine observation of the ocean environment and measurement of bacterial
populations. An intensified survey program was commenced in 1949, where,
in addition to visual observations and bacterial analyses, the following factors
were measured:

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 347

I j, 18~/ 8 CD:~?l•~~~~wh~~~~~~;::;;~~:r:d,I
~M~h, ;r::::; \ · ~ / /\ / /\ 1P'
✓
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

h /\ /\ ,\
60" I.D. R.C. Pipe

Drilled diffuse,
ports locoted
11
6 above
spring line

11
DIFFUSER FOR 60 OUTFALL
see bulkhead detail

rn
A
~ ~-="'F'------""F--"'"l"''-----="f"---~•Surface
50'
"'" "~-;;:--f - - - j f - - - j 100'
·-si.--+------1150'
.
A Section A·A
72" OUTFALL
BULKHEAD DETAIL DIFFUSER

216'

TYPICAL DIFFUSER SECTION

11
DIFFUSER FOR 72 OUT_FALL

FIG. 1.-OCEAN OUTFALL DIFFUSERS OF THE LOS ANGELES COUNTY SANITATION

DISTRICTS

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

·-,---,..----t----t------+-- SURFACE
,.,.
c..:,
ex,

, ,.1 ,,,,.1,,-
~l~l , SEE BULKHEAD
,I_;.,.----
I A?;,
DETAIL

UJ
M
~
Q
M
Note:
Each Leg ldentic
8
>rj
>rj
g
M
PIPE :x,
UJ

SE CTION B-B B_J

BULKHEAD DETAIL TYPICAL DIFFUSER SECTION

F IG . 2. -DIFFUSER FOR 90-IN. OUTFALL OF THE LOS ANGELES COUNTY SANITATION DISTRICTS

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 349

1. Ocean currents in the vicinity of the point of discharge at the surface

and at a depth of 10 ft;
2. Wind velocity; and
3. Water temperature at the surface and at various depths at two points,
one 5,000 ft and the other 7, 500 ft offshore along the line of extension of the out-
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

falls.
The level of bacterial activity is seen in Fig. 3 to be directly related to the
amount of sewage effluent; in Fig. 3 the bacterial concentrations are based on
the percentage of bacterial counts (coliform M.P.N.-as defined in "Standard
Methods for the Examination of Water and Waste Water," American Public
Health Association, 11th Edition) exceeding 10 per ml. California regulations
establish for bathing waters a limit of 10 coliform organisms per ml as the
level beyond which public health and safety may be impaired, qualified by the
stipulation that the frequency of counts greater than 10 coliform per ml may
reach, but not exceed, 20% of the number of samples taken, based upon twenty
consecutive samplings. It is presumed_ that the occurrence of coliform counts
exceeding 10 per ml at a greater frequency than 20% of the number of readings
is sufficient to establish some credence in the data. By 1952, bacterial levels
were approaching the point where improvements were indicated in the means
of disposal at Whites Point (Fig. 3) and it was indicated that further thought
should be given to improving the mixing of sewage with sea water at the point
of discharge and to exploring design modifications which might improve con-
ditions at the outfall site. The sewage field resulting from the discharge from
the then open- ended pipes was objectionabl e for the following reasons:
a. Surface layers of poorly diffused sewage sometimes traveled with shal-
low, wind-driven ocean currents, which move predominantly shoreward;
b. Mixtures of sewage and sea water, often exceeding 2% sewage, were quite
stable and maintained identity as "sewage fields" for extended periods of time
and for several miles of travel;
c. Shoreline coliform counts at times exceeded allowable limits;
d. Visible sewage fields created ar eas of discolored water seen from the
shore; and
e. Poorly diffused sewage effluent created odor problems, both in and near
the floating field.
Effect of Ocean Temperature Gradients. -Experience at Whites Point showed
coliform concentrations at the shore in the vicinity of the outfalls to be higher
in winter than in summer. The difference was due, in part, to stronger shore-
ward winds occurring during winter months, but in the main, seemed to be in-
fluenced greatly by the stronger ocean-temperature gradients in the summer.
Sewage effluent discharged from a deep ocean outfall mixes first with cold
bottom water. It was often observed that sewage-sea water mixtures reached
the ocean surface some 10 °F colder than nearby surface water. At times, the
colder mixture of sewage and sea water was sufficiently more dense than sur-
face ocean water that, on reaching the ocean surface, it would r esubmerge and
spread laterally below the surface, a phenomenon readily demonstrated by
dumping fluorescein dye into the rising column and observing it spread later-
ally and disappear under nearby surface water. Since the phenomenon just
described appeared at times coincident with lower bacterial counts at shore
sampling stations, both effects being observed during the summer and fall, it

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 350 SEWAGE DIFFUSE RS

240
2 20
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

200 :!icz'.'
18 0
"' ":, ' -
- :::::
160 LL
SEWAGE -: LL
,__
140 "
l!; O ,__
FLOW 120
'- -'-
:,
MILLIONS 100 ""~ t= ~e-..J..J 't=-
IL ~
OF 80
60
:.
t--
GALLONS :,
40 " 0 '-
DAILY 'i
20
~~ ' - ~o' -
0
...,.., "',.., "',.., ....,.., .,"' ,..,OI .,0 ., ... ...,.., ...... .,... "'... ....... .,... ...
N 0 ,.., ., .... a,
'°!!? "''!' "'"''!' "'"''!' '!'"' "''!'
OI
;:; N
!!? !!? !!? !!? !!? !!? !!? '!? !!? !!? !!? !!? '!? !!? !!? !!? '°!!? "'
!!? !!?
100
\ ,,,,,-- 8 0
,,,,,,.,., "lo =~
20
0
"'
a:

Jt~ 100
<t
w
>-
a:
CD
"''
~
80 m
<t
0 I 2 3 4 60 3: 0
\:'.:;:_(_1_0 00 n .) 40
(.')
20 z
0 1i:
::::,
.,. 100 0
0
80
60 0
w
40 0
20 z
w
a..
"'::::,
"'
(.')
~
"7.0 60 ...J
40 a..
20 :::.
<t
0 "'
80
"7.0 6 0
'-
I-"' :, ';
40 ~ v
20
0
...,.., ....,.., .,,.., ,..,OI .,- ... ,..,...~ ...., 1,"' "'... ....... .,... ... ,.., .,
!!?
~
"'
~
!!? ~ ~ ~
0
1, ~
N
OI
~ OI
OI0 -
"' "'
N
"' ~ "'
"' '!' "'
"' ....
"' "'
'!' '!' "'
~
- ~ !!? "' ~ !!? !!? ~ '!'

FIG. 3.-ANNUAL PERCENTAGE OF COLIFORM COUNTS EXCEEDING 10 PER ML

AT SHORE STATIONS

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 351

was deemed advisable to measure more carefully some of the physical proper-
ties of the ocean environment in search of a solution.
To the weekly measurement of temperatures at two specific locations off-
shore was added the measurement of surface and subsurface temperatures at
1, 000-ft intervals from shore along the line of extension of the outfalls to a
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

distance of 10,000 ft seaward. Temperatures were measured with a highly

sensitive thermistor mounted on a sounding cable and the data plotted on pro-
file sheets ( Fig. 4). The presence of cold bottom water at shallower depths
during the summer and fall is accounted for by upwelling of bottom water along
the Southern California coast, resulting from the prevailing westerly winds.
Warm surface waters are quite naturally explained during the summer and fall
as a result of the greater insulation. The winter-spring regime is character-
ized by very little change in water temperature from surface to bottom in near-
shore water, in part because of the absence of upwelling, but also because the
shallower waters are more completely mixed by the stronger wave action. The
deep-water diffusers, built by the Sanitation Districts at the ends of their ocean
outfalls, have been designed to invade the area of colder bottom water in order
to create submerged fields of dilute sewage effluent.
Studi es of Diffusers. -Small-scale model studies confirmed theoretical con-
clusions that a multiple-port diffuser system was superior to other systems
tested. Preliminary calculations demonstrated that a multiple-port diffuser
could lie on a seaward sloping ocean floor and flow full through all ports at all
conditions of flow without costly penalty in hydraulic head loss.
In July 1953, with the consent and cooperation of the East Bay Municipal
Utility District, the writers studied the operation of the Oakland Outfall Dif-
fuser, located in the ship channel between Oakland and Yerba Buena Island in
San Francisco Bay. The average depth to the top of the horizontal, single-
barreled, multiport diffuser is about 40 ft. Digested sludge is discharged with
the sewage effluent for dispersion in the bay. Because of high turbidity, the
diver inspected by feel, with zero visability. Inshore sections of the diffuser
evidenced about 1/ 4 in. of "greasy slime" around the inside of ports, with the
deposit lessening toward the offshore end of the diffuser, indicating that most
of the flow had been passing through the innermost ports. The buildup of sedi-
ment, grease, and slime was very limited, causing no concer n over cl ogging of
the pipe, even though the pipe had been in service for the discharge of primary
effluent and digested sewage sludge for about 1 yr. Rec ent observations on the
Sanitation Districts' outfall diffuser have verified these earlier findings.
Construction of Diffus ers . -Based on theoretical analysis and the operating
structure at Oakland, diffusion structures were designed and placed on the Dis-
tricts' existing 60-in. outfall and 72-in. outfall. The 72-in. outfall was dis-
charging a flowof about 100 mgd 5,000 ft seawardand it was deemed advisable
to convey that quantity of sewage farther from shore and to greater depth. An
additional 1,500 ft of 72-in. diameter pipe was added to the outfall, extending
the structure to 6,500 ft seaward in 150 ft of depth. Fig. 1 shows the design of
the 60-in. diffuser and the 72-in. diffuser.
A single-barreled diffuser was deemed adequate for the 60-in. outfall, dis-
charging 50 mgd at 100-ft depth. The original three-outlet cast-ir on diffuser
placed at the seaward end of the 60-in. outfall was closed with steel plates,
leaving only a small opening at the top and bottom of the central port to provide
some self cleansing. Commencing with the first inshore pipe section from the

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 352 SEWAGE DIFFUSE RS

TH OU SANDS OF FEE T
0 2 3 4 5 6 7 8 9 10
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

~ '-- 55° ....... 0

'--""'"
"'~,..,.,_----
,,.,,,,__
...__

.._,. ._
"-....

"--
"'
~

/
/
'\.

' '\
'--
-55°

-
..----- 50
~~
'" /
/ "- ...__
- --- ---- I-
w
w

-
,,~
u.
''""--1lb,..,
"'
_,,,. I 00 !!':
~
·~' ,
--To,__,
/' :,:
I-
a.
ILLUSTRATING ""flh:: -.__ II.I
C
WINTER-SPRING REGIME I 50

TH OUSAN_DS OF
·~

FEET
" .,,.
-~
,
~o,- 200

0 2 3 4 7 8 9

50
I-
w
w
u.
10 0 !!':
:,:
I-
a.
w
ILLUSTRATING C
150
SUMMER - FALL REGIME

200

FIG. 4.-OCEAN-TEMPERATURE PROFILES AT WHITES POINT

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 353

cast iron diffuser, 9-in. diameter holes were bored slightly above the spring
line on opposite sides of the pipe, two in each 18-ft pipe section ( Fig. 1). The
additional head loss for the diffusers did not exceed 2 ft at peak flow. The
operational experience gained from the two diffusers during 1954 and 1955 pro-
vided confirming data for the multiport diffuser to be used on the next proposed
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

outfall.
In November 1956, the Districts completed and placed in service the third
of its ocean outfalls, 90-in. inner diameter, extending seaward 8,000 ft to a
depth of 212 ft; Fig. 2 shows details for this diffuser. The total cost for con-
structionof the 90-in. outfall and its diffusion structure was $1,686,000; it has
a design capacity, based on dispersing efficiency, of about 150 mgd mean flow.
A summary of the physical characteristics of the three outfalls and diffusers
is shown in Table 1.
Inspection and Maintenance of Diffus ers. - Following a year of discharge of
the primary effluent thru the 60-in. diffuser and the 72-in. diffuser, a physical
inspection was made of the condition of the ports to determine if sedimentation
was occurring in the main barrel of the pipe or if the ports were clogging. In-
spection showed all ports to be flowing; side ports in the terminal pipe section
were approximately half filled with" soap-like" deposits which partially blocked
the flow, but were readily dislodged. The bottom half of the terminal pipe sec-
tion was filled with a deposit which tapered so rapidly that it was not discernible
in the next pipe section. There was no flow from the bottom slot in the end
gate; however, sewage effluent was found to be flowing from the hole in the up-
per end of the gate. The observed deposits were found only inthe last pipe
section and it was not"deemed necessary to clean the diffuser at that time.
After about 4 yr of operation, a careful inspection (by free-swimming divers
using self-contained underwater breathing apparatus) showed that the deposition
of settleable material had reached the third port from the terminal of the legs
of the 72-in. line and the diffuser was cleaned. Using [Link] divers,
the bulkheads which closed the ends of the legs were jacked partially open and
sufficient sewage effluent passed thru the diffuser to eject the settled solids
out onto the ocean floor. The floatable material blown horn the pipe was col-
lected at the ocean surface to prevent its drifting onshore. No further action
was necessary to satisfactorily scour the debris out of the diffuser. The dif-
fuser on the 90-in. outfall was similarly cleaned without incident; a routine
program of flushing every 2 or 3 yr appears to be a simple, relatively inex-
pensive means of cleaning the diffusers.
Evaluation. -The experience gained from almost 5 yr of operation of mani-
fold diffusers located at considerable depth in an ocean environment having
regular and rapid change of water has clearly demonstrated certain ameliorat-
ing effects resulting from such diffusers when compared with previous poorly
diffused discharge. One very desirable effect has been a reduction in the bac-
terial population in near-shore waters. The most dramatic changes occurred
in the area immediately above the point of discharge. Whereas, well-defined
"boils" were readily distinguishable before adequate diffusion was initiated, it
has become difficult to detect the location of the sewage discharge by other than
chemical or bacteriological means, except on infrequent occasions during the
winter when discolored sea water appears over the diffusers during periods of
high flow. Oceanographic studies made prior to the use of diffusers could
readily depend on the sewage bubbles as points of reference; with the diffusers
installed, it has become necessary to establish off-shore locations by use of

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 354 SEWAGE DIFFUSERS

sextant, as only in this manner can one be assured of sampling from an area
directly over the point of discharge, or at some predetermined distance from
the outlet.

TABLE 1.-SUMMARY OF DIFFUSER CHARACTERISTICS

Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

Outfall
1 2 3

Outfall inside diameter, in

inches 60 72 90
Year outfall placed in service 1937 1947 1956
Year diffuser placed in service 1954a 1953 1956
Length of ocean outfall exclud-
ing diffuser, in feet 4,500 6 ,400 7,900
Total length of diffuser, in feet 384 648 2,400
Average depth of discharge (be-
low mean sea level), in feet 108 155 203
Number and diameter of ports 26 at 9 in.b 40 at 9 in.b 2 at 15 in .
18 at 8.4 in. 18 at 8 in. 32 at 7.5 in.

68 at 6 ,5 in,

Shape of ports circular circular circular

(sharp- (bell- (bell-
edged) mouthed) mouthed)
Total 2ort area
Ratio: 0.94 0.85 0,63
Outfall cross section
Spacing of ports, in feet 2 every 2 every 24
18 ft 24 ft
Design peak flow, mgd 90 150 260
Calculated additional head loss
due to diffuserc at design
peak flow, in feet 1.9 2.1 6.1
Design friction factor in dif- 0.026 O.030 (legs) 0.024
fuser, f (Darcey-Wiesbach) 0 .026 (stem)
Manning roughness n, equiva-
lent to f - values above 0.0155 0.0155 0.015

b a Diffuser constructed by bulkheading end of outfall and drilling holes in main pipe.
Includes 2 slots in each end bulkhead , each slot being equivalent to one 9 in. diameter
port. c Includes pipe friction in diffuser but not density head due to changes in depth.

In addition to the disappearance of the sewage bubbles, the sewage field or

"sleek" has become so indistinct as to be indiscernible to other than a prac-
ticed eye . Density differences are reduced to the point that wave patterns
move through the dilute sewage-sea water without discontinuity; reflective and
refractive properties are so little changed as to make it almost impossible to
see the sewage field.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 355

Prior to the installation of diffusers, the area immediately over the points
of discharge smelled faintly, but distinctly, of sewage and industrial wastes
and, at times, the odor was wind-borne for as much as 1 mile. With diffusion,
the problem of odors has ceased except during calms when a faint odor may be
detected directly over the discharges. With even the slightest breeze this odor
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

is so quickly dissipated as to be indiscernible a distance of 1,000 ft. The vi-

cinity of the sewer outlets has, for many years, been a favorite fishing grounds
for a number of commercial sports-fishing boats that depend on the attraction
of the sewage effluent for an assurance of a fair day's catch of bottom fish.
Since the use of diffusers, these fishing craft have been noted to station them-
selves at times directly over the diffusers.
Benthic studies in the vicinity of Whites Point disclose that ocean plant life
has ceased for a distance of about 1/ 2 mile in all directions from the termini
of the diffusers, possibly caused by toxic effects of industrial wastes, lessened
salinity, or a decrease in transmitted sunlight (necessary for plant growth) re-
sulting from increased turbidity in the immediate vicinity of the sewage dis-
charge. Since the operation of the diffusers, the surface turbidity immediately
over the diffusers has decreased; on occasion, the turbidity has actually been
less than that of the surrounding ocean waters. The explanation for this ap-
parent anomaly seems to lie in the fact that much of the diluting water is drawn
from the bottom layers of ocean water; turbidity from sewage effluent, added
to the very clear bottom waters, apparently results in a less [Link] mixture
than that of the surface waters which usually contain plankton.
The effect on the concentration of coliforms in shore waters is not so easily
demonstrated as is the effect on ocean water in the immediate vicinity of the
point of discharge. Of advantage is the fact that submerged sewage fields gen-
erated by the diffusers are subjected to shorewardcurrents which are weaker
than those at the surface, allowing more time for natural reduction of coliform
concentrations. Additionally, much higher initial dilutions have been achieved
with diffusers, although this advantage is partially lost because the dilution
rate occurring in dilute sewage fields formed by diffusers is slower than that
occurring in the narrow field originating from an open pipe. For these reasons
the benefits to be derived from diffusion will depend on a great many factors
other than the simple relationship between the sewage concentrations in dif-
fused or undiffused fields.
The general pattern of improvement in density of coliforms has met expec-
tations (Fig. 3) . It is not possible to make direct comparisons between nu-
merical results since many factors obscure the pattern of coliform distribu-
tion resulting from the discharge of sewage effluent through the diffusers: to
wit, the continuing increase in discharged sewage effluent, overlapping effects
of discharges from local systems, and the disturbance of the ocean floor with
depth charges in the construction of outfalls.
Averaged dataon the concentrationof sewage effluent in the ocean immedi-
ately overlying the points of discharge have been compared in Table 2 for the
periods prior to and subsequent to the use of each diffuser. The diffuser on
the 90-in. diameter outfall has produced a sewage field averaging about 1 part
sewage effluent to 400 parts of sea water during the winter-spring regime
when the sewage fielcl is at the surface; when the field is submerged surface
samples taken directly over the diffuser yield negative results, both for chlo-
ride changes and for the presence of coliform bacteria; no measurements have
as yet been taken of the dilution obtained in the submerged fields.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 356 SEWAGE DIFFUSERS

Summary. -The outfall system of the Los Angeles County Sanitation Dis-
tricts has provided a case history spanning more than 20 yr of experience in
the disposal of a primary sewage effluent to a favorable ocean environment.
Since 1953, the installation and operation of deep-water multiport diffusers
has provided comparative experience indicating major reductions in odor, dis-
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

coloration, and turbidity, all being significant improvements in the aesthetic

acceptability of the discharge, as well as measurable reductions in bacterial
populations. Inspection and ma}ntenance of the diffusers have been accom-
plished in depths to 212 ft without difficulty.

DILUTION IN A RISING JET OF SEWAGE

To analyze horizontal discharge from an ocean outfall, assume that sewage

effluent is discharged near the ocean floor from the end of a single pipe or

TABLE 2.-IMPROVEMENT IN INITIAL DILUTION BY USE

OF DIFFUSERS AT WHITES POINT

Dilutiona
Outfall Without diffuser With diffuser

60-inch 40 170
72-inchb 30 300
90-inch C 400

a Ba s ed on average of maximum sewage concentrations at ocean surface over point

of discharge, determined by ammonia concentration measurements. b 72-in. outfall
was extended from depth of 110 ft to 155 ft at time diffuser was constructed. c Never
operated without diffuser.

port. Most important are determination of the rate of dilution of the sewage
as it rises toward the surface of the salt water and the shape, density, and other
characteristics of the jet or "rising column" of sewage. Discharge from a
vertical or an inclined port or outlet will not be discussed herein because un-
der such circumstances the effluent is given vertical momentum initially, thus
rising to the surface faster and in a shorter path without other compensating
influences.
Except as indicated the ocean current is assumed to be nil, and ocean salin-
ity and temperature are assumed uniform.
Description of the Probl em .-Whenthe sewage flows into sea water from the
end of a pipe, it is immediately subjected to a buoyant force proportional to the
difference in density between the sewage and the surrounding sea water. This
force deflects the sewage jet, or stream, toward the surface and accelerates
the sewage upward. The relative motion between the sewage jet and the sea
water develops shear stresses. Turbulence is generated and mixing takes
place first around the peripheryof the jetor risingcolumnand finally through-
out the whole column. As the mixing progresses, density differences are de-
creased and the vertical force driving the sewage effluent toward the surface

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 357

gradually decreases. Since the mixing is greatest around the periphery, that
volume is decelerated most rapidly, and velocity gradients are developed across
the column causing shear and turbulent diffusion throughout the entire column.
From energy considerations, it can be shown that the upward velocity of the
column as a whole gradually decreases as it rises, except possibly for a short
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

distance at the start of its rise where differential densities are sufficient to
accelerate the motion of the sewage effluent.
There is a complex interrelationship between the buoyant gravity force and
the turbulent shear force. Both forces depend on the nature of the flow pattern,
yet the flow pattern is determined by the forces acting; added to this dilemma
are the uncertainties of the turbulent mixing process.
The appearance of the rising column is shown in Fig. 5 which is a photo-
graph of a small jet in a laboratory tank. It is not surprising that there should
be some difficulty in a direct physical analysis. The flow is inherently fluc-
tuating due to its turbulent nature. Where a column of rising sewage effluent
breaks the surface of the ocean, the so-called "bubble" shifts back and forth
at random, making it impossible to identify any fixed point as the center; di-
luted sewage effluent arrives at the surface in numerous small "gusts." Divers
report that large shifting eddies comprise the periphery of the column. At
best, only time averages of the actual phenomenon can be considered.
The flow from an ocean outfall is essentially that of-a submerged jet with
buoyancy, with the flow field limited by the water surface and the ocean bot-
tom. For a submerged jet without density difference in an infinite or semi-
infinite fluid field, there are no external forces, such as gravity, affecting the
flow; the only force is the internal shear. Consequently, considering the whole
region of flow, the total momentum flux of the flow is constant. On the other
hand, when external gravity forces deflect the jet into a curvilinear path, the
momentum flux is no longer constant, and there is no simple way in which the
buoyant force can be fitted into the equations for momentum flux. Likewise,
the elementary principles of continuity and conservation of energy cannot be
applied to the rising column. The volume of upward flow is constantly increas-
ing due to the dilution, and energy is dissipated in turbulence.
A comprehensive experimental study of the hydrodynamics of submerged
jets with buoyancy is needed. Some outstanding studies of diffusion of sub-
merged jets without gravity effects have been published; especially noteworthy
is the work of M. L. Albertson, Y. B. Dai, R. A. Jensen, and Hunter Rouse,
whose analysis4 is based on reasonable theoretical hypothesis and experimen-
tally determined constants. It may be possible to extend their analysis to the
more general case with a gravity force. Also of interest is the ingenious the-
oretical solution by G. Abraham5 for the case of vertical discharge with buoy-
ancy in an infinite medium.
Without the benefit of comprehensive research, one recourse is to resort
to model experiments and empirical curve fitting. A. M. Rawn and H. K. Pal-
mer6 did this for both horizontal and vertical jets. The empirical formula de-

4 ".D iffusion of Submerged Jets," by M. L. Albe rtson, Y. B. Dai, R. A. J ensen, and

Hunter Rouse, Transactions, ASCE, Vol. 115, 1950, p. 639.
5 "Diffusion in Submerged Circular Vertica l Turbulent Wate r Jets with Density Dif-
ference Between the Jet and the Surrounding Water," by G. Abraham, Tech. Rept. 138-
2, Hydr. Engrg. Lab. Wave Research Projs., Univ. of Calif., Berkeley, August, 1959.
6 "Predetermining the Extent of a Sewage Field in Sea Wate r," by A M Rawn and
H. K. Palme r, Transactions, ASCE, Vol. 94, 1930, p. 1036.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 358 SEWAGE DIFFUSERS
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

FIG. 5.-RISING COLUMN OF DYED FRESH WATER IN A SMALL LABORATORY

TANK OF SALT WATER

FIG. 6.-SKETCH OF SINGLE RISING COLUMN SHOWING NOMENCLATURE AND

COORDINATE SYSTEM

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 359

rived from their experiments for dilution at the surface applies only to one
particular set of units because the equation is dimensionall y nonhomogene ous.
With a somewhat different approach, utilizing dimensional analysis, it is pos-
sible to develop relationships in terms of dimensionles s parameters. Further-
more, without considering some of the familiar dimensionles s hydraulic num-
bers, questions arise as to the range of conditions over which the empirical
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

formula is valid and as to the degree of similarity of flow patterns in model

experiments to those of prototype cases. In addition, where dimensionles s
numbers are used, direct comparisons with other hydraulic phenomena are
facilitated.
In the following four sections, the theory of hydraulic models is applied to
the ocean outfall problem, and the original Rawn and Palmer6 data for hori-
zontal discharge are reanalyzed in this light. A new solution for the dilution
is presented in graphical form.
lJimensional Analysis. -In this analysis, the dilution is defined as the ratio
of the volume of the total mixture of sewage and sea water to the volume of the
sewage fraction; in other words, S = 1/ p, where p is the fraction of sewage in
a sample. The dilution S is not necessarily the dilution at the ocean surface
but is considered a variable depending on y, the height above the center of the
outlet, as defined by the schematic diagram in Fig. 6. When y = y 0 , the total
depth from center of the outlet to the water surface, then S = S0 the dilution at
the top of the rising column. Furthermore , the dilution will also vary across
the jet, becoming greater toward the periphery; hereinafter, S will denote the
minimum dilution, which occurs along the axis of the jet. An additional com-
plication is that turbulent fluctuations will cause the dilution at any one point
to vary considerably with time, but since these fluctuations are not of especial
importance, the analysis will deal only with the time-average dilutions.
In the simplified case with no temperature gradients or currents, the dilu-
tion S0 in the center of the rising column at the surface may be considered a
function of the following independent variables:
Symbol Definition Dimensions

Yo Total depth from center of outlet to surface L

D Initial diam eter of jet (same as diameter of pipe, L
port, or nozzle if there is no jet contraction)
V Nominal jet velocity= 4Q/1rD2, where Q = L/ T
discharge
g' Apparent acceleration due to gravity (Eq. 1) L/T2
II Kinematic viscosity of the sewage L2/ T
The small difference between the kinematic viscosities of sewage and sea water
is inconsequent ial; hence, only one viscosity is included in the dimensional
analysis. Moreover, the mass densities of sewage and receiving water are
considered the same as far as inertia of the fluids is concerned; the density
difference is important only in the buoyancy calculations.
The apparent acceleration due to gravity, g', is the acceleration caused by
the buoyant force, and is related to the ordinary gravitation constant, g, by the
relation
g' = g ~s . . . . . . . . . . . . . . . • . . . . . (1)

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 360 SEWAGE DIFFUSERS

in which s is the specific gravity of sewage and .6s deno tes the difference in
specific gravities of sea water and sewage.
Provided the discharge pipe is raised slightly above the ocean floor at the
outlet end, as is usual, the actual distance from the center of the outlet to the
bottom appears to be an unimportant variable in most practical cases, because
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

the jet quickly turns upward away from the ocean bottom .
Including S0 , which is dimensionless, there are six p ertinent variables, in-
volving two fundamental units, length and time . Therefore , by the Pi Theorem,
there are 6 - 2 = 4 independent dimensionless numb ers which can be formed
from these six variables . It is most convenient to choose them as follows:
(a) So
(b) Yo/D
V
(c) F= ( Froude number) . . . . . .. (2)
JgiD
and
VD
(d) R= (Reynolds number) . . . . . . . (3)
V

Thus S0 may be considered a function of the other thre e parameters,

.. . (4)
where F and R are the familiar Froude and Reyno lds numb e rs in hydraulics.
The Reynolds number used is for the jet at its origin, and characterizes the
r elative importance of viscosity. Fo r R > 2,000, the jet flow will usually be
turbulent, but the turbulence is probably not fully developed until R. reaches
10,000 or 20,000 . As R incr eas es still further, the kinematic nature of the flow
may be expected to change little.
The Froude number gives the r elative influence of gravity on the jet. For
a free liquid jet discharging horizontally into air,

F = _ V_ .. . . . . . . . . . . . . . . (5)
Igo
For small values of f', the jet is deflecte d sharply by gravity, whereas for large
values of F, the jet has more momentum and the traj ect or y turns downward
more gradually . For a submerged jet the Froude number, .as defined in Eq. 2,
has the same significance as it does for the case of the liquid jet into air, as -
suming here, as in the discha r ge into air, that friction is a secondary influence
in det e rmining the shape of the flow pattern near the point of discharge. For
the submerged jet in the laboratory tank shown in Fig . 5, the Froude number
F = 2.2 and the depth to diamete r ratio y 0 / D = 15.3 . This condition could cor -
r es pond to a prototype discharge of 139 mdg from an 84-in. diameter outfall
at a depth of 108 ft, using Froude-numbe r similarity.
With dimensional analysis, the number of independent quantities has been
r educ e d to three, each of which has a particular phy sical significance. How-
eve r, the actual functional relations hip between S and the three variables y 0 / D,
f' , and R must be determined expe rim entally.
Hydraulic Similitude and th e Use of an Hydraulic Model.-H the three pa-
r amet e rs are identical for two distinct cases, then S0 must also be identical

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 361

for the two cases. The fundamental requirement of any model study is to make
the relevant dimensionless numbers the same in the model as in the prototype.
Geometric similarity is certainly necessary, but far from sufficient. How-
ever, it is virtually impossible in any model study to achieve similarity for
both Froude and Reynolds numbers simultaneously; 7 this is found to be the
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

case in the ocean outfall problem under consideration.

For a submerged jet with no buoyancy, Albertson, et al4 made the~asic as-
sumption that the turbulent flow pattern is dynamically similar for all Reyno lds
numbers, provided of course that the jet flow is fully turbulent. Their experi-
mental results appear to justify the assumption. Since the rising column of
sewage is a submerged jet with buoyancy, it is basically a combined diffusion
and gravitational phenomenon, with the Reynolds number characterizing the
turbulent diffusion and Froude number, the buoyancy effect. Therefore, it will
also be assumed that the flow pattern for the rising column of sewage is inde-
pendent of the Reynolds number in the full y turbulent range. To test this as-
sumption, the original data of Hawn and Palmer6 were re-examined to deter-
mine if any relationship between dilution and R could be detected.
Re -examination of Rawn-Palmer Data for R eynolds Number Effect.-The
original Rawn-Palmer6 study included 388 experimental observations of the
dilution at the surface ( So) for horizontal jets of fresh water into sea water.
The nozzle diameter, D, ranged from 1/ 4 in. to 2 in. nominal diameter; vari-
ous depths (Yo) up to 13.25 ft were used. (The experiments were made from
a raft in Leis Angeles Harbor, and thus simulated an ocean outfall situation,
except that the discharge nozzle was not near the bottom as for a real outfall,
but was cantilevered down from the raft. However, as previously indicated,
this is not considered an important difference.) For each of these runs the
values of y 0 / D, F, and R were calculated from the original unpublished basic
data. Careful analysis of these data by multiple plotting revealed no appre-
ciable influence of Reynolds number which ranged from 5,000 to 40,000.
Reynolds numbers for actual outfalls may easily be of the order of 105 or
106. For example, a flow at 10 ft per sec from a 1-ft diameter port gives a
Reynolds number of approximately 106. As the Reynolds number becomes
large, its importance as a similarity number markedly decreases for many
hydraulic phenomena. It seems r easonable, then, to hypothesize that if So is
independent of R in the r ange of 5,000 to 40,000 (sometimes a rather critical
range) it remains so for larger R -values as well.
Consequently, for the simple case of a rising column in a homogeneous
ocean, S0 was considered a function of only two variables, namely
S0 = f( y0 / D, F) . ... ... . .. ..... .. . (6)
Model studies may be performed on the basis of geometric similarity and
Froude's law alone, provided that the flow pattern in the model is fully turbu-
l ent .
Graphical Solution for the Dilution . -A graphical solution for S0 was readily
formulated, assuming that R is no longer an important variable . Smoothed
curves were fitted to the data with the results shown in Figs. 7 and 8. Since
both sets of curves represent the same relationship between the variables S0 ,
y0 / D, and F, values of S may be obtained from either figure.

7 Engineering Hydr aulics , Edited by H. Rous e, Chapte r II "Hydr aulic Similitude" by

J. E . Warnock, John Wiley and Sons, Inc., 1950.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 362 SEWAGE DIFFUSERS
With r e sults presented in this manner, some inter esting observations are
possible. For any given horizontal outfall the y0 / D r atio is fixed, and F is di-
re ctly propo rtional t o V or Q. When y0 / D is about 40 it appears from Fig . 7
that as F inc reases fr om 2 up to 5 t her e is a progressive decr ease in S . When
0
F incr eases beyond 5, S0 r em ains a lmost consta nt up to F = 10, and then sta rts
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

to i ncr ease rapidly as F goes still higher.

120

100
- l..-
- - - c.--
-'-- 80

---
--
60 -
.i.- L---
80

60
~
i-
L---
- L---
L-- l...-
L---
c..--

--
- l..-

-~ - ,..___
r-- 50

~ L-,----
--- 1...--- L---

--
--.....,
L-- L--

---
1--.
1--.
i.,,J i.,,J 50 '!,0 - t'----.
~
::::i
C)
i:::
::::i
C) 40
~ ---
"?,0___.. 1..-"'
~
i.---
1..,.-- r--.
)'...
i'..40 _

~
C)
Q:i
"I:
1--.
"I:

~ 30
13/-
!ft----
1-

i--
- +--
-i-----....
i----... .....,
I'-,.. "'~ I"-. ::,
.::
~
;::::

~ r---..... ~[Link]'
(:)
~ 11/ r--- t-- ,,
~
~ :-... t---~
')'.._
I'-- I~ ~

--
C;)
~ 20
~ ~t---.....
,, i~ ,..___ r---.....t,.._
"- '25
~
~,~ .........._
,6 t---..... ' t-,.. "\
r--
r-- ....._t---.._
'r--... t--.. '20

10 r---. . . 'I'----, ,.._

J'-..._
I'---
)'...._
I~
""' '-15

2 3 4 5 6 8 10
' 8

20 30 40
Q
F = FROUDE No. =
fo 2 Yg 1 D

FIG. 7.-DILUTION AT TOP OF RISING COLUMN AS A FUNCTION OF y / D AND F FOR

HORIZONTAL DISCHARGE (BASED ON RA WN-PALMER DATA)0

When y 0 / D is large ( say a round 80 to 100) , then it takes r elatively large r

Froude numbers to produce the same effects on S as those just described for
0
y 0 / D = 40 . The depth becomes a more important factor than F. In other
weirds, if the column must ris e through 100 pipe diameters of depth, increasing
the strength of the jetting action- and the horizontal displacement of the jet -
for F-values up to 30 or 40, do es not r esult in an inc reas e in S (or a down-
0
turn in the constant S0 curves in Fig. 7 ) .

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 36 3

The effect on S0 of changing the s ever al var iabl es in Fig. 7 is s hown by the
s t r aight lines in F ig . 9. If only the depth y0 is increas ed keeping a ll other var -
iabl es constant, t hen the change in dilution is found by fo llowing a v ertical li ne.
On the other hand, an incr eas e in Q fo r a given outfall (fixed Yo and D) is r e - .
fl eeted only in an i ncr eas e in F in dir ect propor t ion with no change in y 0 / D. A
furth er pos s ibility is to dec r eas e t he s ize of t he outlet, while m aintaining the
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

120 32
/ 16 ~
100 z; 8 ~
fl V i
80 4- '<I
/ §
~~ _/

~
~ IV /
2 ~,,
60 V i..
V
......
......
~ 50
/ / -1"' /
V
~ /. I / I/
~ ~
::::> V./ I,, V
::::> C) 40
C)
...... /r/ l/1
sq:
~ ...... 1,,,-: ff'/ V l
C)
~ 30
~ .V V
Ct, V
sq: /
~
J

~ ~
~
~
~
~
20
/ V
J/ ~ v / V
J
J
I
,, / V/ 7 I
I~V I
zvV
~I~ J

V I NOTE: 0 .__..__
~ = FROUDE = ;0 2 Vg 1 o_
10
/
I/ .'
1,/ ' F No.

JV 32
8 v,s
8 10 20 30 40 50 60 80 100

Sa= DILUTION AT TOP-OF RISING COLUMN

FIG. 8. - DILUTION AS A FUNCTION OF y 0 /D FOR CONSTANT F FOR HORI ZONTAL

DISC [Link] (BASED ON RAWN- P ALMER DATA)

s am e depth and s a m e discha r ge, t he r eby increasing V; since F ex D-2.5, and

y0 / D ex (D- 1 ), t he r e sult is a line with s lop e 2/ 5 t o 1 a s shown. Another pos -
sibility is that a numb er of dis cha r ge outlets ( N) a r e pr ovide d i nstead of only
one, and t hat t he outlet diamet er D is decreas ed enough t o leave th e discharge
velocity and aggr egate dis cha r g e a r ea constant. Since in this cas e F oc n-O. 5,
the changed condit ions follow t he line with 2 t o 1 s l op e .

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 364 SEWAGE DIFFUSER S

Certain comparison s can be drawn from Fig. 9 to demonstrat e the effec-

tiveness of the several possible methods for increasing initial dilution of sew-
age effluent in sea water. For exampl e, consider a discharge of 125 mgd
through the open end of a 6-ft diameter outfall, lying at 108 ft of depth to pipe
centerline; values for the dimen'sionl ess variables are y / D = 18 and F = 3.0.
0
These are plotted as point A in Fig. 9. At that point either Fig. 7 or 9 shows
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

that the dilution at the top of the rising column will be S = 10. Points B, C,
0

120

100 ....
~ :i: ~ WUMBER OF OUTLETS FOR COWSTANT Q
<ti ~ ( ASSUMING NO INTERFERENCE J
80 - - - i';'i "'
Cl:
~
~ c:i
~ -:,.·
60
h.. h.. c:,
4J 4J
~ ~
50
::::> ::::>
C)
C)
h.. 40
~
C)
"t
h..
ct)
"t
~ 30
li..
C)
~ ~
~ '.::!
c:i c:i 20
,, Q, V INCREASING
~lc:i Y., D CONSTANT

8+-----------+---+---+---+-------+-
------◄
2 3 4 5 6 8 10 20 30 40
Q
F = FROUDE No. =
f 0
2
Vq 1 o
FIG. 9.-DIFFERE NT METHODS FOR INCREASING DILUTION

and D in Fig. 9 illustrate three alternative methods for increasing S to 40.

0
They are summarize d in Table 3.
The first possibility (point B) would involve extending the outfall offshore
to a depth of 294 ft; unless increased distance from shore is needed for other
reasons, this method of increasing the dilution is so costly as to be infeasible.
The third possibility (point D) would probably be infeasible in most cases since
increasing thevelocity of discharge from6.8 ft per sec to44 ft per sec results
in an increase in velocity head loss from 0. 7 ft to 31 ft. The use of 10 or 11
separate discharge ports adequately spaced to avoid interferenc e appears to

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 365

be the most feasible alternative (point C). Little additional energy is re-
quired, and a diffusion structure can be built at a fraction of the cost required
to greatly lengthen or deepen an outfall. Such a line of reasoning led to the
decision by engineers of the Sanitation Districts to improve the ocean dispos-
al facilities by means of multiple-outlet diffusers.
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

Other Investigations .-It has sometimes been suggested that the flow pat-
tern surrounding a submerged ocean outfall is similar to convection over a
heat source. Many investigations have been made of convection due to point or
line sources of heat, one of the more recent such being a study by B. R. Morton,
G. I. Taylor, and J. S. Turner.8 When there is a source of heat in a fluid, the
fluid adjacent to the source is heated, becomes lighter than the surroundings,
and starts rising, thereby inducing convection currents. A rising column of
fluid is formed over the heat source although no fluid has been injected into the

TABLE 3.-EXAMPLES OF ALTERNATIVE METHODS OF INCREASING S0

Yo, in D, in Number of
Point YolD Method
feet feet Outlets

B 49 294 6 1 Increase depth of discharge

to 294 ft.
C 58 108 1.86 10.4 Use 10.4 outlets (that is, 10
(theor.) or 11) of D = 1.86 ft in-
stead of one of D = 6 ft.
D 46 108 2.35 1 Increase discharge velocity
by factor of 6.5, using
D = 2.35 ft

system. The energy driving this type of flow is entirely derived from the heat
or buoyancy. On the other hand, in the case of a submerged discharge from an
ocean outfall, the energy available to the flow for entrainment and mixing is of
two types: (1) potential energy due to the submergence of sewage in sea water
( equivalent to t:.ss y O in hydraulic head), and (2) kinetic energy of the jet. Con-
sequently, the analogy is good only in case the kinetic energy of the discharge
is smallcompared with the potential energy. The methodof applicationof con-
vection theory to the ocean outfall problem will be considered in detail in a
later report, where density stratification of the receiving water will a lso be
introduced as a variaole.
Another possibility for analysis of a submerged jet with buoyancy is to as-
sume that the dilution increases with L, the curvilinear length along the axis
of the rising column ( Fig. 6), at the same rate that dilutions in an ordinary
submerged jet increase with distance along its axis . In effect, one is simply
straightening out the rising column, and neglecting the effect of the difference
in density. In essence, then, this approximation is the opposite of the one dis-
cussed above; the kinetic energy is now large compared with the potential en-
ergy due to density differences.
8 "Turbulent Gravitational Convection from Maintained and Instantaneous Sources,"
by B . R . Morton, G. I. Taylor, and J. S. Turner, Proceedings, Roy. Soc . A., Vol. 234,
Jan-Ma r., 1956, p. 1.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 366 SEWAGE DIFFUSERS

The second assumption has been used implicitly by E. K. Rice9 (who dis-
cussed only vertical discharges) and by P. Cooley and S. L. Harris.1° The
latter conducted an outstanding experimental investigation under the direction
of C. M. White on the use of submerged jets to prevent or destroy density
stratification in a reservoir. Although they were dealing with relatively small
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

density differences, and therefore high Froude numbers, the same principles
apply.
Erman A. Pearsonll has given a comprehensive summary of these and other
investigations, and a list of references.
Interference between Jets. -Where there are many ports discharging from
a diffuser, there will be some interference between the flow patterns estab-
lished by the individual jets. If the diffuser ports are arranged in a long line,
the flow of diluting water toward the diffuser will be essentially normal to the
whole diffuser and thus not radial toward each individual r-ising plume of ef-
fluent, Such a restriction should tend to reduce the dilution somewhat below
that which would be obtained from a single port.
Another type of interference is the actual contact of one rising column with
another. Rawn and Palmer6 observed that the diameter of the column at the
top is approximately L 0 / 3, where Lo is the curvilinear length as shown in
Fig. 6. Hence, when adjacent rising columns have travelled a distance L equal
to about three times the port spacing, the columns may be expected to merge
and further dilution will be substantially curtailed.
Stratification. -If the receiving water is stratified due to variations in tem-
perature and/ or salinity, the flow pattern may be considerably altered. In
some instances where warm surface water overlies colder bottom water, a
rising column of sewage may not reach the surface, as discussed previously.
The sewage effluent discharged from a diffuser mixes with relatively large
quantities of cold bottom water, sometimes permitting the mixture to become
dense enough that it cannot penetrate a warm surface layer; thus, one of the
important advantages of a diffuser may be the creation of a submerged sewage
field, which might not occur with discharge from an open-ended outfall.
The problem of predicting possible submergence in the presence of density
stratification is complex and cannot be discussed here except briefly. As a
first approximation, one could assume an ocean consisting of only two distinct
laye rs-a cold bottom layer and a warm surface layer. For the rise of the
sewage in the cold water up to the level of the warm-water interface, the graph-
ical relations in Figs. 7 and 8 can be used to determine an approximate value
of the dilution S at that level. If s is the specific gravity of sewage and sw and
sc the specific gravities of the warm and cold layers of sea water respectively,
then the mixture will stay submerged below the warm surface layer if
(S-l)sc+s
S > SW ••••• • •••• • ••••• (7)

9 "Discharge from Submerged Outfalls,"by E. K. Rice, thesis presented tothe Univ.

of California, in Berkeley, Calif., in 1949, in partial fulfilment of the requirements for
the degree of Master of Science.
10 "The Prevention of Stratification in Reservoirs,• by P. Cooley and S. L. Harris,
Journal of the Institute of Water Engineers, London, Vol. 8, No. 7, November, 1954, p.
517.
11 "An Investigation of the Efficacy of Submarine Outfall Disposal of Sewage and
Sludge ," by Erman A. Pearson, State Water Pollution Control Bel., Publication No. 14,
Sac ramento, Calif., 1956.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 367

In addition, the density of the mixture should be enough greater than that of the
overlying layer to overcome the residual inertia of the rising column.
The relation between dilution and temperature difference may be conven-
iently illustrated by an example, assuming the following data:
Temperature, Specific
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

OF Gravity
Effluent 70 0.9987
Sea water ( dilutent), salinity = 33,600 ppm 50 1.0258
Consider a column of sewage which rises from a depth of 150 ft to 80 ft, and
in so doing becomes diluted with 50 parts of sea water at 50 ° F. Its specific
gravity will then be about 1.0253, which is equivalent to plain sea water at 55°
(same salinity, 33,600 ppm). If a thermocline exists above 80 ft, with water
significantly warmer than 55 °F, then a submerged sewage field may be antici-
pated. This is commonly the case in the summer and fall months along the
southern California coast.
Even if the temperature gradients are not strong enough to prevent the sew-
age from surfacing, the sewage may still plunge under the surface after leaving
the vicinity of the bubble, forming a field which is mostly submerged. This
phenomenon was observed in the vicinity of the Whites ·Point outfalls before the
construction of diffusers. Apparently the sewage was heavy enough to sink, but
vertical momentum of the large plumes rising from the 5 and 6 ft open-ended
pipes could not be overcome by the small density difference.
When the sewage field does stay at the surface, the temperature gradients
are still beneficial in making the sewage field thicker, and hence more dilute.
Since the edges of a rising jet become diluted most rapidly, this peripheral
flow is stopped at a lower level even though the less dilute central core breaks
through to the surface. Sampling in the vicinity of the diffuser for the 90-inch
outfall at Whites Point indicates that the sewage field frequently extends 50 ft
to 100 ft down from the surface when it is not submerged.
Ocean Currents. -Strictly speaking, all of the foregoing material applies to
the case of no ocean current. While the aspirator action of the discharging
jets from a diffuser is responsible for the intimate mixing of sewage and sea
water, nonetheless the complete success of ocean disposal depends on ocean
currents for gross exchanges of water. Without these currerits the continuous
discharge of s ewage into the same ocean-water mass would gradually build up
an intolerable concentration of sewage, and the sewage discharging from the
diffuser would be remixed with older sewage.
If the currents are strong, the dilutions achieved in the ocean may consider-
ably exceed those which may be calculated on the basis of single-port dis-
charges into a calm ocean by Fig. 7. While the initial mixing results from the
jet action, whether or not there is a current, the natural turbulence in the ocean
may further increase the dilution if there is diluting water available. When a
strong current sweeps past a diffuser, the large supply of clean water not only
satisfies the demands of the jet entrainment process, but also provides s ea
water betwe en the small individual diluted sewage streams generated by each
port discharge. However, in this cas e the se small individual fields soon mix
with the intermediate clean water to form a rather large more or less homo-
geneous sewage field of a siz e commensurate with the siz e of the diffuser. It
is this large field which, in the presence of a strong curr ent, will be more di-

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 368 SEWAGE DIFFUSERS

lute even though the initial dilution developed in the rising column itself may
not be greatly different from that in a calm ocean.
Considering the total flow of water in a current stream, one may calculate
by the continuity principle an average dilution which may be applied to the over-
all sewage field generated by a diffuser. If Q is the sewage discharge, U is
the current velocity, h is the thickness of the sewage field generated, and b is
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

the width of the stream intercepted ( that is, the projection of the diffuser nor-
mal to the current) then the continuity equation simply states that
Sa Q = U b h . . . • . . . . . . . . . . . . . . . (8)
Unfortunately there appears to be no reliable way to predict the initial field
thickness h in Eq. 8 other than by experience. Probably h will tend to decrease
as the current velocity increases (and vice versa), thus somewhat offsetting
the increase in Sa with increasing values of U. Nonetheless, unless the density
stratification in the ocean is very strong, the field will gradually become thick-
er as it moves down-current as a result of vertical mixing along the bottom
surface of the field and also along the top surface if the field is submerged.
Hence an effective value of h can be taken at some distance downstream from
the diffuser to allow for some vertical mixing. Values of 1/ 4 or 1/ 3 of the
ocean depth are not considered unreasonable.
As an example of the use of Eq. 8, consider an onshore current of 0.8 ft per
sec (0.5 knot) passing the diffuser for the 90-in. outfall at Whites Point (Fig.
2):
U 0.8 ft per sec
b 2 x 1200 x sin 63 ° 30' = 2150 ft
h 60 ft (assumed, but preferably measured)
Q 180 mgd (average flow) = 280 cfs
S = 0.8x2150x60 = 370
a 280
It is important to note that this dilution is an average for the entire field,
whereas the dilutions represented by Fig. 7 are minimum values, occurring at
the center of the individual rising columns, before a large homogeneous field is
formed downcurrent.
At this point one may well ask whether the dilution should be calculated in
this way or by Fig. 7. If the average dilution for the field as a whole ( Sa) is
larger than S0 achieved in each rising column with no current, it probably in-
dicates availability of more diluting water than that immediately taken up by
the aspirator action. There will be local areas with higher than average con-
centration in the "bubbles," but as the current carries the field downstream,
the natural turbulence in the ocean will cause internal mixing utilizing all the
available diluting water and thus promoting uniformity of the field. Hence, in
this case the effective dilution may be considered as the average Sa.
On the other hand, if S0 by Fig. 7 is larger than Sa by Eq. 8, then the cur-
rent does not make additional water available over that drawn in by the initial
jet mixing. The actual dilutions achieved over a diffuser will probably be less
than S0 by Fig. 7 because of mutual interference between the rising columns
(unless the port spacing is very large).
Apart from the beneficial effects of improving the initial dilution at the out-
fall site, the currents are of course an adverse factor when they carry the sew-

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 369

age field rapidly back to shore. An analysis of the behavior of sewage fields,
once formed, is beyond the scope of the present paper, but the problems are
briefly described in the following section.
Initial Dilution in Relation to Ultimate Disposal. -The discussion heretofore
has dealt almost exclusively with the determination of the initial dilution of
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

sewage with sea water at or near the point of discharge, whether there be sin-
gle outlets or many. Although it is highly desirable to get as large an initial
dilution as possible, that alone does not assure safe disposal. As indicated
previously, it is also necessary that the site have favorable currents, and that
the sewage effluent is discharged at a distance from shore sufficient to allow
for further reduction of bacterial concentration before the diluted sewage
reaches the shoreline.
That dilution alone cannot reduce the concentration of coliform bacteria to
the California standard for salt-water bathing (10 per ml) is apparent in that
the primary sewage effluent may have concentrations of the order of 106 per
ml; reduction of concentration by five orders of magnitude is necessary from
the time the sewage is discharged into the ocean until it reaches the surf zone
in order that shoreline waters are not to exceed 10 coliforms per ml. It is
doubtful if any reasonable diffusion system, assisted. by further mixing in the
ocean as the sewage field travels toward shore, could consistently produce
physical dilutions of more than 1,000 to 1. Thus a reduction of bacterial count
by two more orders of magnitude must come about through processes such as
mortality and sedimentation in the ocean.
In the disposal process the engineer can control only the initial dilution
through the choice of outfall site and type of diffuser. Removal of coliforms
by natural processes is beyond his direct control; therefore a diffuser which
produces a high initial dilution yields greater assurance of good results. More-
over, with a diffuser all other requirements for sewage disposal can be met
including turbidity, toxicity, dissolved oxygen levels, odors, and slicks.
An analysis of what takes place in the sewage field after it leaves the vicin-
ity of the diffuser depenqs on an understanding of the mechanism of turbulent
diffusion in the ocean as well as sedimentation and mortality of bacteria. This
paper is not broad enough in scope to discuss these factors, but other investi-
gations12, 13, 14, 15, 16 along these lines (primarily in connection with the de-
sign of a new ocean outfall for the city of Los Angeles in 1955-56) have been
performed.
Summary. -The original Rawn and Palmer6 experimental data for sub-
merged jets with buoyancy have been reanalyzed to produce Figs. 7 and 8 which
give the dilution in the center of rising column of sewage at the ocean surface

12 "Studies on Coliform Bacteria Discharged from the Hyperion Outfall," by S. C.

Rittenberg, Final Bacteriological Report to Hyperion Engineers by Allan Hancock Foun-
dation, Univ. of So. Calif., Los Angeles, Calif.
13 "Sewage Disposal in Santa Monica Bay, California," by C. G. Gunnerson, Proceed-
ings, ASCE, Vol. 84, No. SAl, February, 1958. ---
---i_4 "Ocean Outfall Design," Hyperion Engineers (a joint venture), Los Angeles , Octo-
ber 15, 1957.
15 "Methods of Analysis of the P erformance of Ocean Outfall Diffusers with Appli-
cation to the Proposed Hyperion Outfall," by N. H. Brooks, Report to Hyperion Engi-
neers, Pasadena, Calif., April 5, 1956.
16 "Diffusion of Sewage Effluent in an Ocean Current," by N. H. Brooks, Proceed-
ings , First Internatl, Conf. on Waste Disposal in the Marine Environment, Univ. of Calif,,
Be rkeley, Calif., July 23, 1959,

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 370 SEWAGE DIFFUSERS

as a function of two dimensionless numbers: the ratio of depth to jet diameter

and the Froude number for the discharge. This new relationship is considered
to have more reliability than the earlier empirical equations, inasmuch as the
principles of hydraulic models and dimensional analysis have been utilized in
developing .Figs. 7 and 8.
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

The most practical way to achieve an increase in initial dilution appears to

be to increase the number of discharge jets by means of a relatively long mani-
fold or diffuser. If the receiving water is stably stratified, such as when warm
water overlies cold, the small jets of sewage from a diffuser may never ap-
pear at the surface, a most desirable objective.

HYDRAULIC DESIGN OF DIFFUSERS

For a given ocean outfall, improvement of dispersal of sewage effluent is

accomplished by use of outlet diffusion at the end of the outfall sewer. If the
sewage is discharged at a single port or "en masse," its dispersion and dilu-
tion will be slower than if it is discharged over a large area through a number
of ports. In fact, without the use of multiple-outlet diffusers, other conditions
being equal, much longer outfalls into deeper water are necessary to provide
the same degree of dispersion and consequent shore protection.
An effective and simple type of diffuser is one which distributes the outflow
through many ports over a large area with minimum head loss and interference
between rising columns. Diffusers which discharge several radial jets from a
single small manifold chamber have been used, but are not as effective because
of the interference between rising columns. The following material presumes
a diffuser consisting of one long pipe, or several branching ones, with discharge
ports at intervals along the pipes.
Basi c Hydrauli c R equirements. -The successful design of a long diffuser
with a large number of outlet ports requires careful hydraulic calculations.
The basic hydraulic requirements of such a diffuser are outlined herewith.
Flow Distribution. -The division of the outflow between the various ports
should be fairly uniform. If the diffuser is laid on a sloping sea bottom, it will
be impossible to achieve uniform distribution between ports for all rates of
flow . In such cases, it is advisable to make the distribution fairly uniform at
low or medium flow, and let the deeper ports discharge more than the average
port discharge during high rates of flow. To allow substantially less than av-
erage discharge from deeper ports is considered unsafe from the point of view
of possible clogging of the deeper part of the diffuser .
Velocity in Diffuser .--The flow velocity in aHparts of the diffuser should be
high enough to prevent gross deposition of sludges or grease. For settled sew-
age, velocities of 2 fps to 3 fps at peak flow are adequate (but borderline) since
these will tend to scour material settled during low flow periods. If deposition
takes place in any part of the diffuser over an extended period of time, the
cross section of the pipe or outlet may become so constricted that locally the
velocity will be reduced, a cycle that would accelerate the deposition proc ess.
The final result may be complete clogging of the terminal ports and failure of
the diffuser to completely perform its dispersal function.
Ease in Cleaning.--Even carefully designed diffusers will require occasion-
al cleaning to remove accumulated grease, slimes, and grit as cited previously
under the heading "Outfall System of the Los Angeles County Sanitation Dis-

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 371

tricts: Inspection and Maintenance of Diffusers." These accumulations tend to

increase the apparent friction factor (mainly by decreasing cross-sectional
area), thereby reducing flow from offshore ports and increasing flow from in-
shore ports. Cleaning can be accomplished by flushing or pulling a ball through
the line .
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

Prevention of Sea Water Intrusion.--All ports should flow full in order to

prevent the intrusion of sea water into the pipe. Sea water entering the pipe
will be stagnant and will tend to trap grit and other settleable matter. Such
deposits reduce the hydraulic capacity of the diffuser, thereby limiting its use-
fulness for future years when higher flows might be expected.
Total Head Loss. --If effluent pumping is necessary or the available gravity
head is limited, the total head loss in any proposed diffuser should be kept
reasonably small.
Port Design. -The outlet ports may quite satisfactorily be circular holes in
the side of the pipe without nozzles or tubes or other projecting fittings. For
optimum dilution the jets should discharge horizontally, with no initial upward
componentof velocity. The inside of the hole should preferably bebellmouthed
to minimize clogging and to provide a discharge coefficient which will remain
constant over a period of years.
Hydraulic Analysis. -The hydraulic analysis of a multi-port diffuser is basi-
cally a problem in manifold flow and is somewhat complex. A discourse on the
subject is not within the purview of this paper, but may be found elsewhere.17, 18
The purpose of the following analysis is primarily to illustrate one manner in
which a diffuser can be designed, and to demonstrate some basic principles
peculiar to the design of ocean outfall diffusers.
Gravity Effects.-Several gravity effects are important in diffuser flow.
The effects of density difference on the rising column has previously been dis-
cussed under the heading "Dilution in a Rising Jet of Sewage: Dimensional Anal-
ysis" where it was stated that the shape of the discharge jet near the port is
governed only by the Froude number F as defined by Eq. 2, assuming ideal flow
conditions upstream of the port. According to Rouse, 19 for a circular orifice
in a large tank, the Froude number should be greater than 0.59 in order for
the orifice to flow full. For a rounded port, it is reasonable to take F > 1 as
the criterion for flowing full. With every port in the diffuser flowing full, there
is no way in which the sea water may re-enter the pipe, once initially expelled,
and the diffuser will continue to remain full of sewage effluent.
In making hydraulic calculations, the pertinent pressure at any point in the
diffuser is the pressure differential between the fluid inside the diffuser and
the sea water outside at the level of the port. Working in reverse order from
the deepest or farthest point of a diffuser backward, the decrease in depth
tends to increase the pressure differential between the sewage inside the dif-
fuser and the sea water outside, in spite of the fact that the pressure of both
sewage and sea water decreases. Henceforth, the use of the terms "pressure"
and "pressure head" will herein refer to the pressure differential. The change

17 "Mechanics of Manifold Flow," by J. S. -M,eNown, Transactions, ASCE, Vol. 119,

1954, p. 1111.
18 "Appl!cation of Conformal Mapping to Divided Flow," by J. S. McNown and E. Y.
Hus, Proceedings of the First Midwestern Conf. on Fluid Dynamics, J. W. Edwards,
Ann Arbor, Mich., 1951, p. 143.
19 "Elementary Mechanics of Fluids," by H. Rouse, John Wiley and Sons, Inc., 1946,
p. 105.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 372 SEWAGE DIFFUSERS

of this pressure head due to a change of elevation of Li.z will be equal to Li.s Li.z.
s
To the hydraulic engineer designing sewage diffusers for the first time,
everything will seem upside down. Indeed, it is quite feasible to visualize the
diffuser flow pattern as being inverted, with the vertical scale reduced by the
factor Li.s / s and with the diffuser discharging water into air with the same
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

Froude numbers. The analogy should also help to make clear why it is some-
times difficult to achieve uniformity of discharge from a diffuser built on a
sloping ocean floor.
Characteristics of Flow from a Single P ort.-The hydraulic analysis of a
diffuser is essentially a step-wise process starting at the extreme outer end.
The ports are assumed to be far enough apart so that the flow in the vicinity
of any one port is independent of the rest of the diffuser flow. This is a rea-
sonable assumption, as the port spacing in an adequate diffuser would be at
least 10 port-diameters, and probably considerably more. The discharge from
each port is figured separately in turn, and added to the quantity of flow car-
ried by the diffuser pipe downstream. Between consecutive ports, the effective
pressure head is increased by the amount of the friction loss and the density
head ( [Link] Li.z) . The key to the problem is the analysis of lateral discharge
from a port in the side of a pipe.
The rate of discharge, q, from an orifice or port in the side of a pipe is ex-
pressed by
q = en av'2g""E .. ... ... ... ....... (9)
in which Cn is the discharge coefficient, a denotes the cross -s ectional area of
port, and E is the total head in the main flow at the port. The total head, E,
includes the pressure head in the main pipe relative to the ocean at the location
of the port plus the velocity head of the main flow. In the neighborhood of the
discharge port, it is assumed that there is no energy loss for the main flow in
passing the port. In other words, there is perfect pressure r ecovery compen-
sating for reduction in velocity head in the main flow because of the lateral
discharge. McNownl 7 has shown this to be a good assumption.
The discharge coefficient, Cn, is not a constant along the diffuser, but de-
creases as the velocity head (V2 / 2 g) becomes a larger part of the total ener-
gy (E). By an extension of a theoretical analysis of branching flow by McNown
and Hsu18 it has been found that Cn can be expressed as a function of the r atio

r/E
2
as shown in Fig. 10. Theoretically the curve in Fig. 10 applies only to
small discharges from small holes with diameter less than one -fourth of the
main pipe diameters, as is always the case with diffusers. For small q, the
velocities upstream and downstream of the port are approximately equal ( that

is, V n "" V n- 1) and either Vn or V n-1 may be used to calculate the ratio
although Vn-1 is the more convenient. Furthermore, it is presumed that the
r:/E,
discharge [Link] for negligible velocity of approach ( r:/E-o) is 0. 61 for
a sharp-edged port and 0.91 for a rounded port. These values are reasonable
for discharge from a tank or reservoir.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

1.001 i i I .00
1111
1 1 \11/1~ I I I I I I
f-
Z
090 I
: - • ~to p
" ~o
"-
.._ O[y7:
II
1 - - - - - J ,__ _ ___ _
n/II "1 I I 1~ro
' _
W I t---.~ , i Vn Vn-1
- I ,~r--.. 1 -- -
U 080 "- - -- - - - - -J- 1o so
LL . I ~ I I,--,.
LL I "i-.. qn = CD an -y2gEn --l-l-
w I I ~ I ' • '

0 I I , 1 1 1 1 1
U 0 .70 I I " 1 1 1 1 0.70 Ul
t"1
w
(9
II
I
I
I
I
"
\
I\ ~
Q
0:: 0 .60 SHARP-EDGED I \ 0.60 t"1
<I: I -r--:~R;
I I r--._ I 8
>xj
u I r-------...._ I r\ >xj

I I -----,.__ I\ I I I I 1050 C:
Cf)
0 0 .50 I I " I\ . Ul
t"1

C
I ~- ~ :::0
Ul

:::I I I I 1111111
0 .01 0.015 0.02 0.03 0 .0 4 0.06 0.0 8 0.1
i
0.15
I J3J'lJ[Link]
0.2 0.3 0.4 0.6 0 ,8 1.0
2
Vn -1 /4 _
vi /,.
2g / En - 2g/ En
c,,,
-CJ
FIG. 10.-DISCHARGE COEFF ICIE NT F OR A SMALL PORT ON THE SIDE OF A PIPE c,,,

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 374 SEWAGE DIFFUSERS

Calculation Procedure.-The whole calculation procedure as used in the de-

sign of a diffuser may be formulated mathematically as follows:
Let: D diameter of pipe;
dn diameter of nth port, counting n from offshore end;
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

an area of nth port;

vn mean pipe velocity between nth port ·and (n + l)th port (see
Fig. 10 );
.6Vn V n - V n-1 = increment of velocity due to discharge from nth
port ( or group of ports) ;
hn t, Pn/y = difference in pressure head between the inside and
the outside of the diffuser pipe just upstream of nth port ( ex-
pressed in feet of sewage);
Vn2
hn + ~ = total head at nth port (same either side by as-
sumption above) ;
discharge coefficient for ports (see Fig. 10);
discharge from the nth port;
head loss due to friction between (n + 1) and nth port;
distance between (n + 1 ) and nth port;
Darcy friction factor;
change in elevation between (n+l) and nth port (measured to
center of port; positive when ( n+ 1) port is not as deep as the
nth port);
s specific gravity of sewage;
DS difference in specific gravity between sea water and sewage
First it is necessary to select E1; then q for the first port is:

ql = Cn a1 ✓ 2gE1 = Cnid1 2 ~ . . . . . . (10)

Next, one finds the pipe velocity
ql
. . . . . . . . . . . . . . . (11)
~D 2
4
and velocity head V1 2 / 2 g .
Proceeding to port No. 2, one finds E2 by

Ez = E 1 + hn + .6s .6z 1 . . . . . . . . . . . . . . . (12)

5
y2
1
The ratio g / E2 is calculated for use in. Fig. 10 to find Cn. Then
2
.. . ;(13)

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 375

and
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

This procedure is continued step by step back up the diffuser using the general
relations:
qn = CD an ,/ 2 g En . . . . . . . . . . . . . . ( 15)

CD = ¢(v;~12 / En) . . . . . . . . . . . . . . . (16)

. . . . . . . . . . . . . . . . . . . (17)

Vn=Vn-l+t:,.V n············· . . . . (18)

and
Ln Vn2
hfn = f D 2g· ............... . (20)

A tabular form was devised to facilitate the calculations as illustrated by Ta-

ble 4. If the port discharges and pipe velocities change slowly, it is expedient
to make the stepwise calculations for small groups of ports, as was done in
Table 4. In this case, Eq. 1 7 is changed to read

•... (21)

wherein m is the number of ports considered in a group.

By the nature of the calculations, it is apparent that one cannot decide on a
particular total flow before starting the calculations. It is necessary to esti-
mate the flow from the end port ( q1) which will correspond to the desired total
flow.
Selection of Port Sizes and Spacing and Pipe Sizes.-During the process of
the calculation, the designer is at liberty to change the pipe size, the port size,
and/ or the port spacing. To keep the velocity high enough at the end of the dif-
fuser, it is sometimes necessary to reduce the size of the pipe in one or more
steps from beginning to end of the diffuser. The size of the discharge ports
may be varied in order to keep the discharge uniform from port to port. The
spacing between ports is rather inflexible, inasmuch as practical considerations
dictate that the spacing be either equivalent to the length of a pipe section or
multiple or simple fraction thereof. The entire design process inevitably re-
quires some trial and error arrangements in order to get one arrangement
which is satisfactory at various total rates of flow.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 376 SEWAGE DIFFUSERS

For a diffuser which is laid on a zero slope, the relative distribution of flow
would be the same at all rates of discharge. This is because all the head terms
are proportional to the square of the velocity. In that case, where there are
no differential elevations, one calculation would suffice for all rates of flow.
For example, to double the rate of flow, one would need only to quadruple all
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

the heads and double all the velocities and discharges.

It is essential that the end of the diffuser pipes be bulkheaded, otherwise the
flow will not be forced out of the discharge ports near the end of the diffuser,
and an excess of flow will be discharged through the open-ended pipe. The
bulkheads should be removable for flushing the line.
In the process of makirig the hydraulic calculations it was found that a good
rule of thumb was to assure that the sum of all the port areas is less than the
cross-s ectional area of the outfall pipe. In fact, it was quite apparent that it is
impossible to make a diffuser flow full if the aggregate jet areas excee ds the
pipe cross-section, area, since that would mean that the average velocity of
discharge would have to be less than the velocity of flow in the pipe and that
the flow would have to be decelerated before being discharged from the ports.
This is not physically possible; consequently some of the ports would not flow
full, or not flow at all, and the diffuse r would not work properly. It is probable
that the principal reason for the difficulties of some of the early multiple-outlet
diffusers, such as the Deer Island Outfall4 in Boston, Mass. was the failure to
observe that simple criterion.
Application to Outfalls of the Los Angeles County Sanitation Districts. -The
foregoing calculation procedures is illustrated by a set of design calculations
for the diffuser for the 90-in. outfall (see Fig. 2) . Table 4 lists hydraulic cal-
culations for a low flow of 79 mgd, a condition s elected as an example because
the density head terms 6.s [Link] have a significant effect on the results. In the
s
following material, the several steps in the hydraulic design of the diffuser for
the 90-in. outfall will be reviewed.
Initially, it was necessary to select the location, orientation, and length of
the two diffuser legs from oceanographic and economic considerations, includ-
ing initial dilutions obtainable, prevailing currents, temperature profiles, and
costs of alternatives. It was decide d to make two diffuser legs 1,200 ft long
with an included angle of 127 °, the bisector of the angle being approximately
normal to the coastline. In this way, the width of the band of current inter-
cepted ( or the width of sewage field formed) was largest for a shoreward cur-
rent, and least for a longshore current. No ports were provided in the main
trunk, as was do ne in the 72-in. outfall, since the width of the sewage field
would ther eby only be increased for long shore curr ents which are not critical.
Furthermore, the grade of the pipe is much steeper on the main trunk making
a balanced hydraulic design more difficult to achieve if ports were to be placed
in the main pipe.
The two diffuser legs consist of 60-in. reinforced concrete pipe with one
port in each 24-ft s ection, facing alternate directions . The two legs were laid
out with as nearly identical profiles as possible, to ensure symmetry and equal
distribution of the flow.
From the detailed hydraulic calculations, the designer is able to select rea-
sonable port and pipe diameters. In the case in question it was found that the
diffuser pipes could satisfactorily be of a singl e diameter throughout-60 in.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

. OUTFALL
TABLE 4. -SAMPLE HYDRAULIC CALCULATI ONS FOR DIFFUSER DESIGN, 90 - IN
DIFFUSER, FLOW OF 79 MGD

~
Remarks
0 t_ 0

. u
0
u u
.,. .9 M,. t:i. s/s = 0.026,

~r>
~

~-.s ..._; ~. . .: ....--~

--' -o A= 1ro 2/4 = 19.63 sq u,
"
"'1f :§" i-Z~
~ §~"-=
~
N 0 · 11
-i o· "- ~ 0. :;f ~.....: m = number of ports in
"'§ 8 :::: ~8 8 ii; SI >_ >
~-- ~ u~ o.. "'0 group in Col. 1
:l"a
20 cu

."- i ~
". ~ "1u ~
"
'C
.5 ~: ~., -c•
~:N
I

$ "'. .,,- s ug " "' C: II

~ ·- +
E
~
,!:

c •
"'.
.~
Q.C
~f ~
::" i"
]- z e -1 .,
"~"'* "'"'
N>~ N
~ bO
] .: ....
"u
i~ i
0 >:: s •
>
"'Ea:
u •
r::
0" ."'.
B$ u,"~8
ls ~
~Ox
u
.s ~
"'~a: ~---
~p...
~ .S
~

~.
-c·~

"' 0

~.s.5 v-,"
W Ul
30.~-- s -;;~
0
u "t: n .s- ~.e 0"' :g 'O II M .,

~ 1 Cl~"'.5 0E-<"1"
u
ti"'
~ i z
~

"
0
0. Q .s ~ ~~ i:i:.S E-<"' g~ g, ~ .s .s ii: .s > ·-"
ol
"' o..r
( 1) (2J I (3l (4) (5) (6) (7) (8) (9) (lDl I (11J (12) (13) I (15J (16J I (11J (18)
04J (19) (/l

1 0.199 0 0.91 4.00 0.204 0.20 0.00065 0 .024 24 0.000 + 0.43 I0.210 Slope of pipe = 1.8% for flrst
0.011
trj

~
0 115.0 1 1.227 15.00 48 0,000+ 0.86 0.022
0.232 504 ft
2-3 24 7 .5 0.307 M 0 .210 0.003 0.91 1.03 0.105 0.31 0.0015
4-5 72 11 II ti 0.232 0,006 0.91 [Link] 0.110 0.42 0.0027 " 0 .001 0 .255 f = 0.024 is equivalent to
6-7 120 0.255 [Link] 0.905 1.13 0.115 0.53 0.0044 II Q,QQl 0.278 Manning n = 0.015 C)
1.17 0.119 0.65 0.0066 II Q,QQ2 0.302 trj
8-9 168 0.278 O.Q16 0.90
216 0 .302 0.022 0.89 1.21 0.123 0.77 0.0093 " 0 .002 0.326
10-11
12- 13 264 0.326 0.029 0.885 1.25 0, 127 0 .90 0.013 " 0.003 0 .351 S:l
1-rj
0.351 0.036 0.875 1.28 0 . 130 1.03 0.016 II Q,QQ4 0.377
14 - 15 312 1-rj
0,377 0.043 0.87 1.32 0.134 1.16 0.021 " 0.005 0.404
16-17
18- 19
360
408 6.5 I 0,230 0.404 0.052 0.86 1.01
1.04
0.103
0.106
1.26
1.37
0.025
0.029
" 0 .006
0.007
0.432
0.461
g
20-21 456 0.432 0.057 0.855 II
trj
0.461 0.063 0.85 1.07 0.109 1.48 0.034 11
0.008 o.34 I 0.009 0.478 Slope of pipe = 0. 7% for
22-23 504
1.07 0.109 1.59 0.039 II Q,QQ9 0.496 remaini ng 696 ft. ::0
24-25 552 0.478 0.071 0.84 (/l
0.496 0.079 0.835 [Link] 0.110 1:70 0.045 II Q,QlQ 0.515
26-27 600
0.087 0.825 1.09 0.111 1.81 0.051 " 0 .012 0.536
28-29 648 0.515
0.095 0.82 1.11 0.113 1.92 0, 057 " 0 .013 0.558
30-31 696 0.536
0.558 0.102 0.81 1.12 0.114 2.03 0.064 11
0.0 15 0.582
32-33 744 0,608
0.582 0.110 0.80 1.13 0.115 2. 15 0.072 " 0.017
34-35 792 0.635
0.795 1.14 0.116 2.27 0.080 II Q.0 18
36-37 840 0.608 0.118
0.785 1.16 0.118 2.39 0,089 II Q,Q2Q 0.664
38-39 888 0.635 0.126
0.134 0.775 1.17 0.119 2.51 0.098 " 0.022 0.695
40 -41 936 0.664
0.77 1.19 0.121 2,63 0 .107 " 0.025 0,729
42-43 984 0.695 0.141
0.147 0 .765 1.21 0 .123 2.75 0.117 " 0.027 0.765
44-45 1032 0.729
1080 0.765 0.153 0,755 1.22 0.124 2.87 0.128 " 0.029 , 0.803
46-47
1128 0.803 0 .159 0.75 1.24 0,127 3.00 0. 140 11
0.032 0.844
48 - 49
50-51 1176 0.844 0.166 0.745 1.27 0.129 3 .13 0.152 24 0 .018 0.11 I 0.004 0.866 Wy_e is 1200 ft. from end

Sum 13.95 Q=61.5 each 1200 I o.31 14 I o.36 QTOTAL = 123cfs = 79 mgd
log

a When a group of ports is used, d is tances are to the centroid of the group. c:..,
b Aggregate port area for both legs is 27 .9 ft. or 63% of cross section of 90-inch outfall
2 pipe. -::,
-::,

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 378 SEWAGE DIFFUSERS

After some trial and error calculations, a port size of 7.5 in. was chosen for
the outermost 14 sections, and 6. 5 in. for the other 34 sections in each leg.
Quadruple-sized ports ( 15-in. diameter) were placed in the center of each of
the end bulkheads in order to maintain favorable pipe velocity and to make use
of the better than average diffusion opportunity at the extreme borders of the
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

sewage field formed by the diffuser.

In Table 4, the ti. Zn-values are based on an average measured grade of 1. 8 %
for the outermost 500 ft and 0. 7% for the inner 700 ft, thus accounting for a
total drop of about 14 ft. With a 2.6 % density difference, this drop makes a
total head differential of O. 36 ft of fresh water head; that is, even under hy-
drostatic conditions the pressure head difference between inside and outside
of the pipe would be O. 36 ft greater at the wye than at the end. From the top of
Col. 6 and the bottom of Col. 18 in Table 4 it may be noted that the total head
increases from 0.20 ft to 0. 87 ft as one works backward from the end of the
wye. Of this increase of 0.67 ft, 0.36 ft is due to change in elevation and only
0.31 ft to friction. Consequently, this set of calculations demonstrates the im-
portance of considering the differential elevation at low flow. At a flow of 240
mgd, the computed total head was found to rise from 3.25 ft to 6.67 ft, an in-
crease of 3.42 ft; of this amount 3.06 ft is for friction loss, with only 0.36 ft
for change in elevation, as before.
The effect of the sloping bottom on the distribution of flow between the ports
is apparent in Fig. 11 on which are plotted the distributions for a high flow of
240 mgd and a low flow of 79 mgd. The port sizes have b een selected in a way
which makes the port discharges practically uniform at the low flow, but at
high flow, the offshore ports discharge on the average about 25% morethan the
inshore ports. Going shoreward from the end of the diffuser (that is, in the
order in which the calculations are made), the flow from each port gradually
increases for low flow primarily because of the increase in elevation. On the
other hand, at high flow, the situation is reversed; in spite of the increasing
elevation and accumulated pipe friction losses, the port discharge decreases
very gradually because the discharge coefficient decreas e s fast er than the
head increases. It is seen that there are four factors which tend t o change the
port discharge: change in elevation, pipe friction, change in Cn, and change in
port diameter. For a successful diffuser these four factors should be kept as
nearly in balance as possible.
As indicated previously, an hydraulic design yielding discharge distributions
such as those in Fig. 11 is safe inasmuch as there is virtually no danger of
gradual choking of the offshore end of the diffus er. If the friction factor f be-
came considerably higher than the design value of 0. 024 ( which is equivalent
to Manning n = 0.015), the discharges from offshore ports would be decreased
while the discharges from nearer-shore ports would be increased. Conse-
quently, as the diffuser is designed, an increas e in f would only tend to make
the distribution of discharge more uniform at high rates of flow .
The velocity distribution within the diffuser pipe may b e read from Table 4,
Col. 11. At low flow (79 mgd) , the velocity in the pipe has a minimum of 0.20
fps at the very end, and maximum of 3.12 fps just downstream from the wye.
At the high flow (240 mgd), the corresponding velocity figures are 0.82 and
9.46 fps, respectively. Experience has shown that cleaning is necessary every
few years, especially in the last few sections of pipe where the velocity is the
lowest. This has been easily accomplished by opening the end bulkheads and
flushing with as much flow as possible by diverting effluent from the other out-
falls. Hydraulic calculations for the flushing operation can also be made by
the proce dure outlined previously. With the bulkhead opened, a large part of

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

4I
Jl4l UD
4

..... r- -
-.
..: 3I
Q • 240 1,1. G~-,- ,- ,-
'
r- , r
,- r- , r

3
-2.- I

ILi
<!)
a:
<(
::c en
(.) txJ

~
(/) 2
I2
0
l-
0
a: txJ
0
a. s:i
ILi
,-r ,.-
,, Q ,79 1,1 . G.D .
i-r-1 I Q. 79 1,1 .G.D. - '2l
'2l
...J
<!)
z I
g
txJ
(/)
::0
en

1v4n~: ,~,kFP~~! !,i;I ,,~, 1,1,' 'k' 1,i;I 1,i;I 1,i;I !,kl 1,0 qi 't'~
I
'
t t t t ~ _..,
~ t t , , lj' f , , , ... ~
0

11
,
~~ l,i;I 1,i;I
, ,
'1' , ,
1,i;I ,~,, ,~,I
7f PORTS~24' --
'1' 1,i;Hll
1
0
'f 'f

- - - - - - - 6 t " P O R T S @ 2 4 ' -- - - - ---<)

J I I J I I J I I I I
0 I 2 3 4 5 6 7 8 9 10 II 12
DISTANCE FROM OFFSHORE END - {Hundreds of feet)

FIG. 11.-CALCULATED DISTRIBUTION OF DISCHARGE ALONG EACH LEG OF DIFFUSER FOR c.:,
--1
90-IN. OCEAN OUTFALL <J:>

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 380 SEWAGE DIFFUSERS

the flow discharges from the open end and pressure within the manifold is re-
duced; near the end, the pressure differential will be so small that the ports
will discharge very little, while further back up the line the friction loss and
density head will put the pipe again under positive differential pressure result-
ing in substantial port discharges, although much less than during ordinary
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

operation at the same total flow.

The lowest values of the Froude number under the low flow condition is 3.2
(by Eq. 2) at the port in the end bulkhead. Since this is considerably greater
than 1, the diffuser must operate full of sewage at all times, with no encroach-
ment of the sea water in the pipe.
The arrangement of the columns in Table 4 is general and could be applied
to any diffuser problem. Full columns have been allotted to D and f (Cols. 5
and 13), even though they are constant throughout this example, to indicate that
it is simple to make changes in D and f if desired. If the pipe diameter D is
changed, it is necessary to change V also ( Col. 11) and add an energy loss for
the transition.
The diffusers for the 72-in. and 60-in. outfalls were designed earlier by a
similar procedure, which differed only in some details. The main character-
istics of all three diffusers are summarized in Table 2 and the layouts are
shown in Figs. 1 and 2. It may be noted that the ports for the 72-in. outfall dif-
fuser are rounded, while those for the 60-in. are not. In the form er, the holes
were cast in new pipe, whereas in the latter the holes were drilled into the ex-
isting pipe while in place on the ocean floor. Hydraulically speaking, it makes
little difference whether the ports are rounded or not; one simply has to use
a larger sharp-edged hole to be equivalent to the smaller, rounded port. How-
ever, for a sharp-edged port through the wall of a heavy concrete pipe, there
is some doubt as to whether it will function as an orifice or a short tube, and
thus it is more difficult to predict the discharge coefficient.
The design calculations for the 72-in. outfall diffuser were checked by some
crude field measurements of port discharge velocity using a specially rigged
Price current meter held in front of the ports by a diver with the meter read
from the boat. In Fig. 12 measured port discharges for a total flow of 53 mgd
are compared with the calculated discharges for flows of 42, 89, and 138 mgd.
The agreement is seen to be reasonable except for the first port on the up-
stream leg which may well be an error in measurement.
Summary . -A diffuser can be designed by calculating the port discharges
one at a time starting with the offshore end. For discharge through lateral
ports in a pipe, the discharge coefficient is a variable function of the ratio of
the velocity head to the total head in the main pipe as shown by Fig. 10. Bal-
anced distribution of discharge among the ports can be secured by varying the
port diameter. A necessary requirement in selecting port sizes is to keep the
sum of port areas less than the cross-sectional area of the outfall.
In Table 4 the procedure for design is illustrated for the diffuser for the
90-in. diameter outfall of the Los Angeles County Sanitation Districts. Although
field confirmation of the hydraulic design is ,' difficult, limited observations con-
firm the theory; the three outfall diffusers in use at Whites Point have given
satisfactory service for several years.

CONCLUSIONS

1. The efficiency of disposal of primary sewage effluent to the ocean can be

greatly improved by a good dispersal system.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

-;
'+--'
<..i
EACH 48" DOWNSTREAM LEG - 9" PORTS 72" UPSTREAM LEG -8" PORTS TOTAL FLOW
(f)
f-
0::
0 B HIGH "LOW
0..
LL
0 7 W 214 c.f.s.
HIGl-i FLO 138 m.g.d.
0::
~ 6
>-
1, UPSTREAM LEG (/.l
tr!

~
I 0 123456789
MID FLOW
~
.~
A..'? " 72" PIPE - 8" PORTS
w
5 "'-- ---+- - - + ---+-
o M
ID FLOW 138 c.f.s.
89 m.g.a.
~ ., Cl
~Q. q,
tr!
~
If)
4
0
0::
LL
<t- ---+---+- --+---+- --+----+-- --+--
82 c.f.s.
--+---< / 53 m.g.d. ~~ l"' .,1" 8
>tj
.,►
I MEASURED FLOW R~o f\0'11 65 c. f.s. >tj
3
w ,,._e, 11.s\l :, KEY PLAN
<.!)
4 z m.g.a. g
~ 2· f- LOW F:OW LOW FLOn
tr!
I ~
(/.l
<..)
(f) I- . o - CA LCULATED FLOW
0
• - MEASURED FLOW
01 2 3 4 5 6 7 8 9 10 I 2 3 4 5 6 7 8 9
Note:
PORTS NUMBER PORTS ARE NUMBERED IN PAIRS

FIG. 12.-PREDICTED AND MEASURED DISCHARGES FROM 72- IN. OCEAN OUTFALL DIFFUSER

c:,.,

....
0:,

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 382 SEWAGE DIFFUSERS

2. Increased dispersal of dilution of the sewage in sea water can most easily
be achieved by a large multiple-port manifold or diffuser. By discharging the
sewage flow in a large number of small jets instead of a single jet from an
open-ended pipe, larger dilutions of the sewage in sea water will be achieved
in the rising column.
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

3. In the presence of stable thermal gradients in the ocean the rising plume
of sewage may never reach the surface. This will occur when the sewage mixes
with sufficient cold bottom water to make the mixture density greater than that
of the overlying surface layer of warmer sea water.
4. The possibility of achieving a submerged sewage field when thermal
gradients exist is greatly enhanced by the use of a large multiport diffuser.
5. Ocean currents are necessary for satisfactory disposal to effect ex-
changes of the water masses.
6. Strong ocean currents passing over a diffuser may also give greatly in-
creased dilution in the sewage field formed by the flow from the diffuser. The
arms of the diffuser should be so oriented as to intercept a wide band of the
current when its direction is unfavorable from the point of view of shoreline
pollution.
7. The Los Angeles County Sanitation Districts have installed three large
multiple-port diffusers which have performed satisfactorily in accordance with
the principles enunciated herein. (For layout, see Figs. 1 and 2.)
8. The operation of the diffuser has resulted in reduced coliform bacteria
populations along the shoreline in spite of the steadily increasing flow of sew-
age. (See Fig. 3)
9. There has been great improvement in the appearance in the water sur-
rounding the outfall and a large increase in physical dilution. (See Table 2)
10. Two of the three diffusers have been cleaned without difficulty after 5
and 2 yr of service respectively; the third diffuser did not require cleaning,
as it has not been in constant use.
11. The original Rawn and Palmer data6 for dilution in a rising column
from a horizontal discharge have been reanalyzed, resulting in the graphical
presentations of Figs. 7 and 8. The dilution at the top of the rising column is
now expressed as a function of the Froude number F and the ratio of depth to
jet diameter, y0 / D.
12. A study of these relationships shows that the most feasible way to in-
crease the dilution is to reduce the jet diameter D by use of a large number of
ports, without necessarily increasing the velocity of discharge.
13. Large diffuser manifolds should be so designed that (a) the distribution
of flow between ports is fairly uniform; (b) the velocities within the diffuser
pipes are sufficient to avoid undue deposition; (c) the structure is easily clean-
able; ( d) the system operates full of sewage without sea-water intrusion; ( e)
the total additional head loss is reasonable; and (f) the ports themselves are
of a simple design.
14. For a diffuser constructed of large pipes with small holes along the
sides the hydraulic analysis involves. a variable coefficient of discharge, de-
pending on the ratio of the velocity head to total head within the diffuser (Fig.
10). Calculations of the head-discharge characteristics for the system must
start at the downstream end and proceed upstream.
15. A successful diffuser usually has an aggregate port area which is less
than the outfall cross-section area.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 SEWAGE DIFFUSERS 383

The scope of this paper has been limited to a discussion of initial dilution
and hydraulic design of diffusers together with verification and illustrations
taken from the history of the outfall system of the Los Angeles County Sanita-
tion Districts. No attempt has been made to discuss in detail the complex
phenomena which take place within the sewage field once it is formed. None-
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

theless, considerations of factors such as coliform mortality and sedimenta-

tion, and diffusion by natural turbulence in the ocean should be considered in
any comprehensive outfall design. These subjects have been treated else-
where.12, 13, 14, 15, 16

APPENDIX.-NO T ATION

Text:
an = area of nth port;
b initial width of sewage field;
c0 = discharge coefficient for ports (see Fig. 10);
D diameter of sewage jet at point of discharge; or, diameter of diffuser
pipe;
dn diameter of nth port, counting n from offshore end;
V 2
En hn + ng = total head at nth port (same either side by assumption);
2
V
F - - - = Froude number;
v'g,"D
f = Darcy friction factor;
g = acceleration due to gravity;
g' = g t.s
s
= apparent acceleration due to gravity;
h initial thickness of sewage field;
= difference in pressure head between the inside and the outside of the
diffuser pipe just upstream of nth port (expressed infeetof sewage);
= head loss due to friction between (n + 1) and nth port;
= distance along axis of rising column from point of discharge to water
surface (see Fig. 6);
= distance between (n + 1) and nth port;
fraction of sewage (by volume) in a sewage-sea water mixture;
= rate of sewage discharge;

qn = discharge from the nth port;

VD
R = - - = Reynolds number;
11
s 1/p = dilution ratio at a point on the axis of a rising plume of sewage;

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384

 384 JORDAAN ON SEWAGE DIFFUSERS

dilution S at water surface ( y = y O) ;

= U b h/ Q = average dilution in sewage field;
specific gravity of sewage;
t..s difference in specific gravities of sea water and sewage;
Downloaded from [Link] by Texas A&M University on 03/23/25. Copyright ASCE. For personal use only; all rights reserved.

u velocity of ocean current;

V = jet velocity;
= mean pipe velocity between nth port and (n + l)th port;
= Vn - Vn-1 = increment of velocity due to discharge from nth port (or
group of ports) ;
y height above center of outlet;
t otal depth fr om center of outlet to water surface;
change in elevation between (n + 1) and nth port (measured to center
of port; positive when (n+ 1) port is not as deep as the nth port); and
II kinematic viscosity of s ewage .

DISCUSSION

J. M. JORDAAN JR.20_For the sake of academic interest the writer would

like to ask whether a deflector on each diffusor outlet, so arranged as to eject
the jets downward against the ocean floor, would not effect immediate and more
effective mixing of sewage effluents with the colder bottom layers of the sea
water. The upward rising plume is described by the authors as dipping down
again at the surface, presumably becau·s e it is heavier than the surface layers,
although its residual momentum first carries the effluent plume to the surface.
Would it not therefore be possible to effect the mixing almost entirely by jet
action so that the plume, mixed with colder, heavier water will tend to stay
down?
Directing the jets vertically down will no doubt give rise to scour but they
could be directed at, say, 45° down, against a protective slab covering the im-
pact area, The jet would strike the bottom and spread out similarly to the way
in which a downward directed jet of steam and vapour from a steam pipe, after
striking the floor, wells up in a highly turbulent roller. A somewhat higher
manifold pressure and smaller ports would be required to produce a high ef-
flux velocity, leading to better manifold flow distribution as well. The toroidal
vortex system would rise much slow er around each outlet nozzle after strik-
ing the bottom and due to the central "hole" would have a greater contact zone
over which turbulent mixing can occur. The effluent will also stay in contact
with the bottom longer where the orbital motion of the long period surface waves
sets up a turbulent zone of a f ew feet in depth increasing the mixing opportuni-
20 Head, Hydr. Sect. , Natl. Mech. Engrg. Research Inst., S. A. Council for Scientific
and Industrial Res earch, Pretoria, South Africa.

Transactions of the American Society of Civil Engineers, 1961, 126(3): 344-384