NUTRIENT REQUIREMENTS AND BIOLOGICAL SOLIDS

ACCUMULATION IN ANAEROBIC DIGESTION

RICHARD E. SPEECE* and PERRY L. MCCARTY

Massachusetts Institute of Technology, Cambridge, Massachusetts, U.S.A.

ANAEROBIC digestion as known by the sanitary engineer is a unique and hardy process
and has proved to be a powerful tool in the stabilization of strong organic wastes.
Practically any type of organic matter can be decomposed to methane and carbon
dioxide as the principal end products with very little in the way of toxic by-products
being produced. As BUSWELL3 has stated, methane fermentation can be carried on
in a mixed or enriched culture and hence it is possible to maintain the process on a
large scale continuously for apparently an indefinite period.
Primarily, in the past, anaerobic digestion has been applied to treatment of com-
plex insoluble wastes, such as sewage sludge. While feeding such heterogenous wastes,
the biological solids accumulation has been obscured due to the presence of signifi-
cant amounts of organic matter which are very resistant to biological stabilization.
In conventional activated sludge and trickling filter processes, the biological solids
accumulation is more apparent since the settleable solids are removed from the
waste stream prior to introduction in the aeration tank or on the filter. Thus, only
soluble and colloidal organic matter is biologically treated and the synthesis resulting
during treatment is approximated by the wasted sludge from the secondary treatment
process.
In evaluating a treatment process, it is important to know the degree of biological
synthesis as three undesirable effects can result from synthesis of organic wastes into
biological solids: (1) provision must be made for additional treatment of these solids,
(2) synthesis converts simple soluble compounds into more complex, more difficult
to degrade biological solids, and (3) synthesis of cellular protoplasm requires a
source of nitrogen which increases treatment costs for nitrogen deficient wastes.
The synthesis relationships have been established for many classes of substrates
undergoing aerobic stabilization, since it has been possible to feed pure substrates
along with the normal inorganic salts and achieve satisfactory aerobic treatment.
However, unsatisfactory operation has usually been experienced in the anaerobic
digestion of pure and relatively simple substrates when these were the sole organic
source. The difficulty experienced in the pure culture isolation of methane bacteria
has also retarded the development of fundamental knowledge concerning the anaerobic
digestion process.
The conversion of simple organic compounds to methane and carbon dioxide by
the methane bacteria is a vital key to the successful operation of the process. Con-
sequently, if this link fails the whole process is endangered. Unsatisfactory digestion
has usually been attributed to failure of the methane bacteria to process the simple
organic compounds. Therefore, it is important to study the process and determine
* Present location—University of Illinois, Urbana, Illinois, U.S.A.
u 305
306 R. E. SPEECE and P. L. MCCARTY

the growth requirements and synthesis relationships in order to more fully utilize the
excellent potentialities of anaerobic digestion as an organic waste treatment method
and also to place process design on a stronger scientific basis.

SYNTHESIS RELATIONSHIPS
7
HEUKELEKIAN et al and WESTON and ECKENFELDER19 have used a growth equation
of the general form shown in Equation (1), in connection with aerobic systems.
A =cF-kM (1)
where A =sludge accumulation, no./day
c ^constant, fraction of substrate synthesized
F - l o a d , no. B.O.D. or no. C.O.D. utilized/day
k —constant, endogenous respiration rate/day
M=volatile solids present in system, no.
This equation has proven to be a reliable expedient in engineering design for esti-
mating sludge production from a biological treatment process. HEUKELEKIAN et al.1
established the following equation from laboratory data of activated sludge processes
treating domestic sewage :
Λ=0·5^-0·055Μ
Retaining the same nomenclature as above, the load, F, was based on B.O.D. utilized.
WESTON and ECKENFELDER19 analyzed data of other investigators and plotted
AjM vs. F\M values to show the relationship between B.O.D. removed and activated
sludge produced. They thought a definite ratio should exist between B.O.D. removed,
the fraction oxidized and the fraction synthesized for any given set of conditions,
but that in practice it is difficult to determine these relationships exactly because
cellular material is oxidized concurrently with that of the waste.
MONOD 1 1 said that bacterial growth experiments generally give cell yields propor-
tional to the amount of substrate available. With a particular organism and substrate,
the yield was constant for all growth rates and substrate concentrations, and varied
only slightly with temperature. Therefore, the same basic principles which control
synthesis in aerobic processes would be expected to govern anaerobic processes.
17
STEWART et al applied basic principles in the formulation of a rational approach
to the design of continuous flow anaerobic systems and evaluated a growth equation
for a particular anaerobic system. Experimentally, they used a substrate composed
of Bacto-Tryptone, dextrose, and Bacto-Beef extract and reported the growth equa-
tion for this particular substrate undergoing anaerobic digestion to be:
^-0-18F-0-0247M
They based the load, F, on substrate C.O.D. utilized. These investigators stated that
other processes with different substrates and organism cultures will probably have
different process reaction constants.

GROWTH R E Q U I R E M E N T S IN ANAEROBIC D I G E S T I O N
BARKER studied the methane fermentation of acetate extensively and made the
1

following statement :
The methane bacteria have not been studied as extensively as most other groups of bacteria
of comparable scientific and practical importance. The reason is readily apparent. In order to
study the biology and biochemistry of bacteria most effectively it is necessary to use pure
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 307
cultures. Unfortunately, with methane bacteria this elementary but basic requirement has been
difficult and in many instances impossible to achieve.
This gives an insight into the primitive state of knowledge concerning methane
bacteria, in spite of the fact that they play such a strategic role and are so widely
utilized in biological waste treatment processes. A number of investigators have
reported results which indicate the need of a growth-promoting substance for methane
bacteria.
HEUKELEKIAN and HEINEMANN6 conducting experiments on the seeding of digesters,
concluded that the chief value of digested sludge lies in the production of favorable
environmental conditions and not in the numbers of bacteria added. This observation
may also indicate the presence of growth-promoting substances in the digested sludge.
STANDER15 studied the possibility of treating winery waste by anaerobic digestion.
He found it necessary to periodically add sludge from a digester fed raw sewage
sludge. Without these periodic additions, the winery waste could not be digested at
high loading rates. This indicates some necessary growth-promoting substance was
present in the digested sludge but not in the pure winery waste.
MCCARTY and VATH 10 reported that digestion of acetic acid alone could not pro-
ceed at high rates if only the normal inorganic salts were present. They found high
rates of acetate digestion required the presence of some substance contained in the
supernatant liquor solids from a domestic sewage sludge digester.
In this study unsatisfactory digestion developed with substrates in all categories as
the original seed sludge was purged from each digester. The common denominator
of difficulty being a gradual increase in the volatile acid concentration which
decreased only slightly even when feeding was stopped. Thus, conditions affecting
fermentation of acetate affected digestion of practically all the organic wastes.

PRELIMINARY STUDY ON STIMULATION OF ACETATE DIGESTION

In view of the reports of other investigators and the difficulty encountered in this
study, it was necessary to carry out a preliminary study to determine some pure
compounds which would promote satisfactory acetate digestion in the absence of the
heterogenous sludge used for initiating digestion. It was hoped that such compounds
could then be included with the daily feed to the digesters where biological solids
accumulation was being studied, enabling satisfactory digestion, and yielding reliable
data for each substrate.
A procedure similar to that used by MCCARTY and VATH 10 was followed in this
preliminary study. An attempt was made to operate the digesters under conditions
such that only the lack of stimulatory compounds limited the rate of acetate utiliza-
tion. It was endeavored to maintain all other conditions as nearly ideal as possible.
Continuous gas recirculation insured that the digester contents were uniformly mixed.
A sufficiently low buffer concentration was maintained to avoid ion toxicity. The
temperature was closely controlled at 35±0-5°C to prevent the microorganisms from
being thermally shocked, and the substrate was continuously fed at approximately
the rate at which it was being utilized, while maintaining a reserve volatile acid
concentration between 1000 and 1500 mg/1. There were occasions, however, when
both the volatile acids concentration and the normal inorganic salts were found to
be limiting the maximum rate of acetate utilization. Whenever such a deficit occurred,
it was remedied as soon as possible.
308 R. E. SPEECE and P. L. MCCARTY

The inorganic salts were initially maintained at about the following concentrations
within the digester:
NaHC0 3 5000 mg/1.
(NH 4 ) 2 H P 0 4 230 mg/1.
MgS04 10 mg/1.
NH4CI 280 mg/1.
KC1 40 mg/1.
MgCl 2 50 mg/1.
These concentrations, excepting that of NaHC0 3 , were subsequently tripled, since
they were found to be limiting acetate utilization at increased levels of stimulation.
Sufficient Cambridge, Massachusetts tap water, containing most trace elements
required for bacterial growth was added to the substrate and inorganic salts solution
to bring the total volume to 200 ml., which, when fed to the 61. digesters, gave a 30-day
theoretical detention time. A constant volume was maintained within the digesters
by daily wasting of about 200 ml. of the mixed contents.

o
FEED PUMP
GAS PUMP

AUXILIARY
EXCESS FEED ^ J o J |
GAS
TUBE ■L^1
GAS RECIRCULATION

k
J FEED BOTTLE

CONDENSATE TRAP
DIGESTER

4*i^ TUBE
WITHDRAWAL

FIG. 1. Continuous feed digester system.

The schematic layout of the digesters is shown in FIG. 1. Initially the digesters
were seeded with screened, digested primary sludge obtained from the Nut Island
Sewage Treatment Plant in Boston, Massachusetts. Operation of the digesters in-
volved analysis of a sample of the wasted contents for volatile acids concentration,
calculation of the feed dosage of acetate which would satisfy the acetate utilization
rate and maintain the volatile acids concentration between 1000 and 1500 mg/1. Gas
production was metered and recorded to provide a check on acetate utilization.
Periodically, solids concentration and ammonia and organic nitrogen were determined
on the effluent. Continuous operation was carried on until less than 1 per cent of the
original seed sludge remained. The accompanying acetate utilization rates had
decreased to about 300 mg/1. day.
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 309
Under these latter conditions, actual assay of pure chemical compounds for
stimulation potential in acetate digestion was commenced. Some of the compounds
chosen were reported to be growth requirements for certain types of bacteria. These
compounds were then added to the acetate digesters and their stimulation potential
assayed by the change in rate of acetate utilization.

RESULTS OF PRELIMINARY STUDY ON STIMULATION OF

ACETATE DIGESTION
A number of pure compounds were assayed for stimulation potential. Those
compounds which yielded positive stimulation are listed below along with the increase
in acetate utilization rates which resulted.
Compound Concentration Rate increase
Thiamine 20 mg/1. 1400 mg/1. day
Proline 50 mg/1. 500 mg/1. day
Glycine 40 mg/1. 1000 mg/1. day
Benzimidazole 15 mg/1. 800 mg/1. day
C0CI2 20 mg/1. 2400 mg/1. day
FeCl 3 500 mg/1. 2300 mg/1. day
Glycine, proline, benzimidazole and cobalt implicated vitamin B12 as a stimulatory
compound. Glycine appears as an intermediary product in the synthesis of a portion
of the vitamin B12 molecule, porphyrin. Cobalt is the particular ion attached at the
center of porphyrin for vitamin B12. Individual units of porphyrin are similar to the
proline molecule. Also, vitamin B12 contains a benzimidazole unit as part of its
structure. However, addition of vitamin B12 with an intrinsic factor of mannitol
failed to stimulate acetate digestion.
Since the intermediate metabolic steps in the splitting of the acetate molecule to
yield methane and carbon dioxide, have not yet been defined, it is premature to
attempt to explain the stimulatory function of these compounds. Combinations of the
above stimulatory compounds were found to increase acetate digestion rates from
700 to 6500 mg/1. day in a digester purged of essentially all the original seed sludge.
Sufficient information was obtained to fulfill the requirements for this particular
investigation on biological synthesis and nitrogen and phosphorous requirements.
However, additional studies will yet be required to evaluate and confirm the need
for any or all of the above stimulants and to determine the maximum extent of
stimulation that can be achieved by their presence.

PROCEDURE
Biological solids accumulation
The biological solids accumulation and consequent nitrogen and phosphorous
requirements were determined for the anaerobic digestion of a number of organic
compounds. The category and substrates chosen were :
Fats : Acetate
Octanoate
Oleate
Protein : Glycine
Leucine
Nutrient Broth
310 R. E. SPEECE and P. L. MCCARTY

Carbohydrates : Glucose
Starch
Cellulose
These categories of organic compounds are commonly encountered in biological
waste treatment processes. The substrates in each category were chosen because it
was considered each group would closely represent all the various compounds likely
to be encountered in that particular category. Fatty acids were chosen to represent
the fats category since the glyceride unit (fat) is hydrolyzed to yield glycerol and fatty
acids. Acetate, octanoate, and oleate represent short, medium, and long chain fatty
acids, respectively. The protein class presented a problem, since it is difficult to
choose "representative" compounds due to the varied and complex nature of the
molecules encountered in this group. However, since proteins are composed of amino
acids, it was decided to choose two amino acids and a common complex proteinaceous
material. Glycine, leucine, and nutrient broth were thus chosen. No precipitation of
substrate was found to occur in the digesters receiving any of these three compounds.
This was important, for precipitation of proteinaceous substrate would have obscured
cellular organic nitrogen data by addition of non-cellular solids containing organic
nitrogen. For the carbohydrate substrates, glucose, starch, and cellulose were chosen,
since they represent a gradation in size and complexity of molecule and are commonly
encountered.

20-LITER
RESERVOIR

FIG. 2. Batch feed digester system.

In an attempt to simulate field conditions of operation, a series of digesters was

set up as shown in FIG. 2. The apparatus was enclosed in a constant temperature
room maintained at 35±0-5°C to correspond to the mesophilic range of digestion
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 311
most commonly encountered in practice. Also, a closed atmosphere of gases produced
in the digestion process was maintained to eliminate the toxic effect of oxygen. With
this apparatus, it was possible under rigid anaerobic conditions to measure the gas
production volume at atmospheric pressure, waste a sample of the digester contents,
expel the gas accumulation, and introduce the feed solution. The loading rate at
which all substrates were fed was 0-04 no./ft3 day (670 mg C.O.D./1. day), to simulate
conventional digestion loading, and the contents were thoroughly mixed once a day.
All substrates were fed on at least two sludge retention times with a minimum of 5
and a maximum of 30 days.
The digesters were initially seeded with primary digested sludge from the Nut
Island Sewage Treatment Plant in Boston, Massachusetts. This sludge was first passed
through i in. mesh screen to remove coarse material and paper. One-fourth liter was
then transferred to each digester and 500 ml. of warm tap water was added to dilute
to the required 750 ml. It was necessary that the digester be purged of this original
seed sludge so that analyses of the digester contents would reflect only biological
solids. The purging operation was accomplished by daily withdrawal of a sample of
the uniformly mixed digestion contents and replacing this with the pure substrate and
necessary inorganic salts. By this procedure, less than 1 per cent of the original seed
sludge remained after 4 detention periods. After this time, all digesters were inverted
and the last vestiges of sediment were allowed to settle in the digester mouth and
were then washed out by expelling a portion of the digester contents. Therefore, after
that time, analyses of the contents of a digester reflected only biological activity
resulting from utilization of the particular substrate fed to that digester.
All digesters were operated until equilibrium conditions prevailed as indicated by
consistent organic nitrogen concentrations in the effluent and gas production records.
Final data was recorded over an interval of equilibrium operation of not less than
10 days and up to 45 days, to observe any fluctuations that might have been en-
countered over extended periods. No significant variations were noted during these
extended periods of equilibrium operation.
The sludge retention time was the same as the hydraulic detention time in this
study, since uniformly mixed samples of the digester contents were expelled daily.
Thus, the sludge retention time was controlled by varying the quantity of dilution
water added to the daily feed. A fraction of the mixed digester contents was expelled,
corresponding to the feed volume, prior to the daily feeding operation to maintain a
constant volume.
The daily feed to each digester consisted of 500 mg C.O.D. of the particular sub-
strate, and Cambridge, Massachusetts tap water was added to dilute the feed to the
required volume. Based on the study of stimulation of acetate digestion, the following
concentrations of inorganic salts were added to the digesters operating on fatty acid
and protein substrates, while it was found that these concentrations had to be tripled
in the carbohydrate digesters due to the increased synthesis which occurred :
230 mg/1. (NH 4 ) 2 HP0 4
10 mg/1. MgS0 4
280 mg/1. NH4C1
40 mg/1. KC1
50 mg/1. MgCl2
100 mg/1. FeCl3
312 R. E. SPEECE and P. L. MCCARTY

Of the list of stimulants from the acetate stimulation study, it was found necessary
only to add ferric chloride along with the normal inorganic salts to obtain satisfactory
digestion at the chosen loading rate after essentially all of the seed sludge had been
purged. The buffer capacity was maintained in the digesters receiving fatty acids and
carbohydrate substrates by adding sodium bicarbonate in the feed at a concentration
of approximately 6000 mg/1. This addition was not required in the proteinaceous
substrate digesters as sufficient buffer was produced of the ammonium bicarbonate
end product. Here, it was necessary to neutralize with hydrochloric acid the excess
ammonium bicarbonate alkalinity created in the digestion of the proteinaceous sub-
strates operating on the longer sludge retention times. Otherwise, the resulting
increase in pH led to ammonium toxicity which inhibited digestion, as reported by
MCCARTY and MCKINNEY. 9
The following analyses were made regularly on samples of the digester contents :
volatile acids concentration, C.O.D. of the mixed contents, C.O.D. of the solids-free
liquor, organic nitrogen of the mixed contents (for the protein substrates the organic
nitrogen was determined on the centrifuged and washed solids), ammonia nitrogen,
alkalinity, pH, suspended and volatile suspended solids. The gas from all digesters
was analysed for methane, carbon dioxide, and nitrogen content.
Microchemical analyses were run on the centrifuged and washed biological solids,
to determine the empirical chemical cell formulation in terms of carbon, hydrogen,
oxygen, nitrogen, and phosphorous. This was performed in the Microchemical
Analyses Laboratory of the Massachusetts Institute of Technology.
Organic nitrogen served as the basis for nitrogen requirement calculations. Phos-
phorous requirements were determined by microchemical analysis of the effluent
solids. Exceptions in microchemical analysis determinations were the cellulose and
oleate digesters which contained insoluble forms of the substrate and thus digester
solids would not be representative of bacterial cells, and were excluded from the
microchemical analyses.

RESULTS
Biological solids accumulation
The operating characteristics were determined for the anaerobic digestion of the
chosen substrates for at least two different sludge retention times, with the exception
of oleate. In this case reliable data could be collected for operation on only a 30-day
retention time. Also, due to unsatisfactory digestion with leucine and cellulose at 5
days sludge retention time, operation was switched to 7-5 days where successful
digestion was maintained.
TABLE 1 summarizes the operating data from all the runs. The substrate and the
sludge retention time on which the digesters were operated are listed in Columns 1
and 2. Column 3 records the number of days for which the digesters were operated
after equilibrium conditions had been reached, and during which final data was
recorded. The gas production was recorded daily and the gas was periodically
analysed to determine its composition. The methane fraction of total gas production
was then used to derive the actual methane production recorded in Column 4.
Uniformly mixed samples of the digester contents were drawn daily prior to
feeding. For the digesters receiving fatty acid and carbohydrate substrates, the mixed
effluents were analyzed for organic nitrogen. Since no organic nitrogen was contained
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 313
in these feeds, the organic nitrogen analysis reflected cellular nitrogen and enzymes
secreted into the solution. On several occasions, organic nitrogen analyses were run
on the centrifuged and washed solids. The results were essentially the same as on the
mixed liquor and indicated that enzyme concentrations in solution were negligible.
For the digesters receiving protein substrates, it was necessary to centrifuge and wash
the effluent solids in order to insure that the organic nitrogen analyses included no
traces of undecomposed substrate, which also contained organic nitrogen. This
would have obscured the actual organic nitrogen concentration due to biological
solids. Cellular organic nitrogen concentrations in the respective digesters are shown
in Column 5.

TABLE 1. SUMMARY OF DIGESTER OPERATION

Net
Sludge Days Daily N- synthesis
reten- operated CH 4 Cell-N reqm't Volatile of Removal C.O.D.
Substrate tion at product cone. mg N/g solids substrate efficiency balance
time equili- ml. mg/1. C.O.D. mg/1. C.O.D. /o /o
(days) brium utilized
/o

Acetate 5 45 143 12 5-3 60 67 111

Acetate 10 45 186 23 3-7 — 4-2 93 106
Acetate 15 45 205 36 3-7 — 4-2 98 110
Acetate 20 45 202 40 30 — 3-4 99 106
Acetate 25 45 206 46 2-8 — 3-2 99 108
Acetate 30 45 198 54 2-7 — 3-1 99 104
Octanoate 5 25 115 11 6-1 104 6-9 65 101
Octanoate 30 25 171 53 2-8 492 3-2 96 94
Oleate 30 25 175 51 2-7 — 31 95 94
Glycine 5 20 137 11 40 78 4-5 83 108
Glycine 30 13 156 55 2-9 468 3-3 93 88
Leucine 7-5 10 105 20 5-6 180 6-4 70 88
Leucine 30 17 156 54 2-8 400 3-2 95 86
Nutrient
broth 5 17 130 22 7-5 190 8-5 87 90
Nutrient
broth 30 17 126 104 5-8 700 6-6 89 82
Glucose 5 45 77 81 37 620 42 65 112
Glucose 10 45 114 130 22 990 25 86 99
Glucose 20 40 135 216 17 1970 19 93 96
Glucose 30 30 117 215 12 2080 14 89 84
Starch 5 10 49 60 36 500 41 50 101
Starch 30 10 152 293 15 2200 17 97 98
Cellulose 7-5 15 90 79 23 — 26 68 92
Cellulose 30 21 137 162 9 — 10 94 76

The nitrogen requirements as shown in Column 6 were calculated as follows :

^ . Cellular Organic Nitrogen in mg/1.
Nitrogen Requirement=——,— : — :—— — ;
(Sludge Retention Time in Days) (Loading Rate in gm
C.O.D./l. day) (Efficiency)
314 R. E. SPEECE and P. L. MCCARTY

For the acetate digester operating on a 5-day sludge retention time, a sample calcula-
tion is given :
12mg/l.
Nitrogen Requirement
(5 days) (0-670 gm C.O.D./l. day) (0-67)
=5-3 mg N/gm C.O.D. utilized
Column 7 lists the average volatile solids concentration. This value was not
obtained for the digesters operating on acetate. The feeding of oleate and cellulose
gave rise to the presence of insoluble, non-cellular organic matter within the digesters,
therefore volatile solids analyses of the contents were not determined. The net syn-
thesis in Column 8 was calculated as follows :
Net Synthesis =Nitrogen Requirement x 11-4
The factor 11 -4 was found to be the proportion of C.O.D. to cellular nitrogen in this
study as will be shown later.

10 12
PERCENT WASTED DAILY

h-1 1 1 f +-■-
30 25 20 15 10 7.5
SLUDGE RETENTION TIME (DAYS)

FIG. 3. Nitrogen requirements.

Removal efficiencies shown in Column 10 were based upon the C.O.D. con-
centration in the feed and on the average C.O.D. of the solids-free effluent. One
exception to this was the acetate substrate, where efficiencies were based upon volatile
acid concentrations in the feed and in the effluent. Corrections were made to the
C.O.D. of the effluents from the oleate and cellulose digesters. With these substrates,
the C.O.D. of the mixed effluent was determined, and reduced by a factor of 11-4
times the cellular nitrogen concentration to correct for the inclusion of bacterial
cells with particulate undecomposed substrate.
As a check on the reliability of the data, a balance was made on the C.O.D. input
to the system and the C.O.D. output from the system. The feed was the only source
of C.O.D. input, while methane and the effluent sample were the two sources of
C.O.D. output. The C.O.D. of the mixed effluent sample was determined, and the
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 315
total daily gas production and methane composition recorded. The volume of methane
at 35°C which is equivalent to 1 g of C.O.D. is theoretically calculated to be 395 ml.
The C.O.D. balance was made for the entire final data period and is recorded in
Column 11.
FIGURE 3 is a plot of nitrogen requirements as shown in TABLE 1 Column 6 vs.
sludge retention time. This reveals graphically the difference in biological growth
which can be expected during the digestion of various types of organic wastes, since
nitrogen requirements were directly proportional to net synthesis of substrate into
biological solids.

TABLE 2. CHEMICAL FORMULATION OF BIOLOGICAL SOLIDS

Sludge
Retention Empirical formula
^nhctratf» Time Ash N/P C.O.D./N
j u u o i i aiw
(days) C H O N P /o Wt. basis

Acetate 5 4-9 11-3 3-5 56 120

Acetate 10 5-1 10-4 3-1 — 35 — 12-4
Acetate 15 4-9 9-4 2-9 — 37 — 11-6
Acetate 20 4-3 8-3 2-2 — 32 — 10-4
Acetate 25 5-4 9-4 3-4 — 28 — 12-2
Acetate 30 5-0 110 2-7 003* 24 7 12-9
Octanoate 5 4-6 6-9 2-4 005 9 9 101
Octanoate 30 4-7 7-7 2-1 004 5 12 11-1
Oleate 30
Glycine 5 4-8 7-6 3-0 006 8 8 10-2
Glycine 30 4-9 9-0 30 011 11 4 11-2
Leucine 7-5 50 8-8 3-2 005 9 9 11-1
Leucine 30 4-1 7-6 3-3 005 5 9 80
Nutrient broth 5 4-1 6-8 2-2 004 10 12 9-1
Nutrient broth 30 3-8 7-0 2-8 005 16 9 7-8
Glucose 5 4-8 8-3 2-5 017 20 3 11-6
Glucose 10 5-1 8-5 2-5 014 14 3 12-0
Glucose 20 5-9 101 3-2 018 11 2 13-8
Glucose 30 4-9 10-2 3-1 0-21 26 2 11-8
Starch 5 5-0 8-7 2-9 009 6 5 11-5
Starch 30 5-3 9-1 2-5 006 6 8 12-6
Cellulose 7-5
Cellulose 30

* Phosphorous value determined on solids from another acetate digester utilizing 4800 (mg C.O.D.)/
1. day.

Chemical formulation of bacterial protoplasm

Microchemical analyses of the biological solids responsible for the anaerobic
digestion of all substrates, except the insoluble oleate and cellulose, were determined.
TABLE 2 shows the chemical formulation of these biological solids for the various
substrates and their respective sludge retention times. The N/P ratio in the biological
solids, based on weight, is given. Also, the C.O.D./N ratio is listed, as calculated
theoretically from the empirical formula.
316 R. E. SPEECE and P. L. MCCARTY

The average of all the empirical formulations in TABLE 2 yielded the formula
C4.8H8.8O2.9. For simplicity, the subscripts were rounded off to the nearest whole
number to yield the general empirical formula of anaerobic biological solids as found
in this study:
C5H9O3N
This compares closely with empirical formulations of biological solids produced
aerobically. SYMONS and MCKINNEY 1 8 reported a formula of C5H8O2N, while
HOOVER and PORGES8 reported C5H7O2N.

TABLE 3. EVALUATION OF GROWTH EQUATION CONSTANTS

Volatile A/F
Sludge Cellular Biological solids Removal F mg solids M/F
retention nitrogen volatile accumu- effi- C.O.D. accumu- mg solids
Substrate time (mg/1.) solids lation ciency utilized lation per mg
(days) Mn (mg/1.) per day /o per day per mg C.O.D.
M=9AxMn (mg) (mg/1.) COD. utilized/1.
A utilized/1.

Acetate 5 12 113 23 67 450 0051 0-25

Acetate 10 23 218 22 93 620 0036 0-35
Acetate 15 36 339 23 98 650 0035 0-52
Acetate 20 40 376 19 99 660 0029 0-57
Acetate 25 46 433 17 99 660 0 026 0-66
Acetate 30 53 500 17 99 660 0026 0-76
Octanoate 5 11 103 21 65 430 0049 0-24
Octanoate 30 54 508 17 96 640 0 027 0-79
Oleate 30 53 500 17 95 640 0 027 0-78
Glycine 5 11 103 21 83 560 0-037 0-18
Glycine 30 55 517 17 93 620 0-027 0-83
Leucine 7-5 20 188 25 70 470 0053 0-40
Leucine 30 54 508 17 95 640 0-027 0-79
Nutrient
broth 5 22 207 41 87 580 0071 0-36
Nutrient
broth 30 104 975 32 89 600 0-053 1-63
Glucose 5 81 761 152 65 430 0-35 1-77
Glucose 10 131 1230 123 86 580 0-21 2-12
Glucose 20 216 2030 102 93 620 016 3-27
Glucose 30 215 2020 67 89 600 Oil 3-37
Starch 5 60 565 113 50 330 0-34 1-71
Starch 30 293 2750 92 97 650 0-14 4-22

The C.O.D./N ratio for the general empirical formula, C5H9O3N, is 11-4 while
the cell weight/N ratio is 9-4. These ratios were verified experimentally in the course
of this study by data taken on volatile solids, C.O.D., and cellular nitrogen of the
biological solids. The N/P ratio appears to be approximately 7. All of the respective
values averaged together for the general empirical formula fell within ± 2 standard
deviations of the mean. There appears to be no statistical difference between the
general formula and any of the individual formulas.
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 317
Evaluation of growth equation constants
Since the biological solids and cellular nitrogen are interdependent, as shown
experimentally in this study and also stated by BENNETT and WILLIAMS,2 the growth

1 1
O ACETATE
A OCTANOATE

o 0 GLYCINE
o O LEUCINE
c = 0.054
k - 0.038
0
O
o 0
o

I .2 .3 .4 .5 .6 .7
M/F
FIG. 4. Evaluation of amino and fatty acid growth constants.

1 1
D GLUCOSE

\^ O STARCH

x c » 0.46

A/F 24 \
D

ΓΎ^
u v\ . <p
D
N

0 .6 1.2 1.8 24 3.0 3.6 4.2 4.8

M/F

FIG. 5. Evaluation of glucose and starch growth constants.

equation constants, c and k, can be evaluated on the basis of either net synthesis of
volatile solids or nitrogen requirements. It was decided to use the latter basis, since
these data were available for all substrates. Therefore, the cellular nitrogen con-
318 R. E. SPEECE and P. L. MCCARTY

centrations were multiplied by 9-4, the ratio between volatile solids and cellular
nitrogen as determined in this study, to obtain representative volatile solids
concentrations.
The data were grouped as follows: the amino and fatty acids; glucose and starch;
and nutrient broth by itself. This grouping was chosen, since the nitrogen require-
ments, as shown in FIG. 3, indicated this was a logical division. The constants were
not evaluated for cellulose alone since the daily wasting schedule of operation was
7-5 and 30 days sludge retention time (13 and 3 per cent). This did not represent a very
wide spread of points and would have magnified any error involved.

~ \ _
O NUTRIENT BROTH

C «= 0 . 0 7 6

k = 0.014

0 .2 .4 .6 .8 10 1.2 1.4 1.6 1.8

M/F

FIG. 6. Evaluation of nutrient broth growth constants.

The growth equation constants were evaluated as shown in TABLE 3, and FIGS. 4,
5 and 6. The data are plotted in the form A\F=c—kM\F, which is the straight line
equation form (y=mx+b) of the growth equation. The values A, F and M were
measured experimentally. The constant c is the j-axis intercept and k is the slope of
such a plot.
The line of best fit for the groups of amino and fatty acids, and glucose and starch
was determined by the method of least squares. Since only two points of data were
available for nutrient broth, the line of best fit could only connect these points:
The growth equations were evaluated to be :
Amino and fatty acids: A=0-05AF— 0-038M
Glucose and starch : A =0-46F —0-088M
Nutrient broth : A =0-016F— 0-014M
The loading rate, i% in these equations was based upon C.O.D. reduction between
the influent and effluent filtrate of the respective digesters.

DISCUSSION
Biological solids accumulation
The results of this phase of the investigation have elucidated the biological response
which occurs in the anaerobic digestion of each of the three classes of substrates
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 319
studied. The most significant observation from this phase of the research was the
discovery of the exceptionally high net synthesis and correspondingly reduced methane
production from carbohydrates which occurs at short sludge retention times. Thirty
to forty per cent of the carbohydrate C.O.D. was converted to biological solids
C.O.D. during anaerobic digestion at short sludge retention times as compared to
6-8 per cent conversion for substrates in the fats and proteins class under similar
conditions. This observation has not been reported hitherto and merits a note of
caution from those responsible for process design of waste treatment facilities.
The wastes most commonly encountered in anaerobic digestion have contained
excess nitrogen and biologically resistant volatile solids. Also, sludge retention times
have usually been 15-60 days. These conditions, plus the fact that net synthesis rates
for substrates in the fats and proteins class have been shown to be quite low, have
combined to mask the actual biological solids accumulation which occurs in anaerobic
digestion.
The discovery of the stimulatory effect of iron on acetate utilization by methane
bacteria proved to be a turning point in the progress of this study. Inclusion of the
stipulated quantities of ferric chloride with the normal inorganic salts accomplished
a universal reduction in the high concentration of volatile acids which had accumu-
lated in the digesters. Acetate utilization was stimulated by the addition of iron
filings as well.
BUSWELL and SOLLO4 found the presence of an inert surface within the digester
aided digestion. However, this stimulation phenomena cannot be completely ex-
plained on the sole basis of an inert surface being introduced to the digester. The
cellulose digesters provide a case in point where abundant surface area was present
and still unsatisfactory digestion resulted as the original seed sludge was purged.
The cellulose was fed in the form of shredded filter paper to which the bacteria could
attach since it was not being decomposed readily. However, inclusion of ferric
chloride in the daily feed to the cellulose digesters resulted in a doubling of the gas
production from approximately 150 to 300 ml./I. of digester/day. The volatile acids
concentration which had reached and steadily remained near 2000 mg/1. was reduced
to about 600 mg/1. in 10 days and eventually dropped to 150 mg/1. later on.
FIGURE 3 reveals a very marked decrease in nitrogen requirements for carbo-
hydrates as the sludge retention time was increased, due to conservation of nutrients
as a result of endogenous respiration. Nutrient broth exhibited a higher rate of net
synthesis than the other two substrates in the protein class.
Many investigators5»14'16 have operated digesters on short hydraulic detention
times and long sludge retention times by retaining the biological solids in the digester
through their recovery from the effluent and recycle back to the digester. Two
advantages are thus achieved :
1. Short hydraulic detention times result in smaller digestion tank volumes which
require less capital investment.
2. Long sludge retention times result in lower net synthesis, higher gas production
and higher purification rates with the lower capital and operating costs involved
therewith.

The constant B.O.D. and the B.O.D. reduction stages

The anaerobic digestion process is often considered to take place in two stages.
320 R. E. SPEECE and P. L. MCCARTY

In the first stage, a multitude of facultative and anaerobic bacteria break complex
organics into simple organic compounds such as fatty acids, alcohols, aldehydes, and
ketones. In the second stage, these simple organics are then in turn converted to
methane and carbon dioxide. The first stage of decomposition is brought about
through hydrolysis and also oxidation and reduction of the complex organics. Since
no inorganic oxidizing agent, such as oxygen is added and also no reduced organics
are removed from the waste stream, this first stage of change is brought about with
no reduction in C.O.D., and hence essentially no reduction in B.O.D. Minor excep-
tions to this would be when hydrogen gas is produced or when sulfates were used
during the first stage as hydrogen acceptors.
The real C.O.D. and B.O.D. reduction in anaerobic digestion occurs during the
second stage of methane fermentation where the reduction is directly proportional
to the methane gas production. This is true since no inorganic oxidant such as
molecular oxygen is added to the system. Therefore, the only way a decrease in
oxygen demand can be brought about is through the removal of a reduced com-
pound from the system. The unique characteristic of the B.O.D. reduction stage of
digestion is that carbon dioxide, which is produced in the process, can serve as the
hydrogen acceptor, being reduced to methane.
Conventionally, the essence of biological treatment of organic wastes is reduction
in the oxygen demand. No such reduction is accomplished in the constant B.O.D.
stage of digestion. Yet, this is the stage which appears to exhibit the highest rates of
net synthesis, with certain wastes, such as carbohydrates. Also, with carbohydrates,
the decreased methane production is due to the high fraction of substrate synthesized
in this constant B.O.D. stage.
The bacteria responsible for the constant B.O.D. stage are limited in the extent
to which they can break down organic compounds. For instance, these bacteria can
break down carbohydrates to simple acids and alcohols, but are unable to carry the
degradation further under anaerobic conditions. This step is performed by the
methane bacteria. Therefore, direct feeding of these simple organic compounds
results in the virtual absence of this constant B.O.D. stage population of bacteria,
since no food source is available to sustain them.
In the B.O.D. reduction stage, a population of methane bacteria, which is apparently
proportional to the B.O.D. of the waste, degrades the simpler organic compounds.
Since all substrates in this study were fed at the same loading rate, based on C.O.D.,
all digesters on the same sludge retention time should therefore have contained
comparable populations of methane bacteria.
Attention is now drawn to FIG. 3 again. The rate of net synthesis for anaerobic
decomposition of acetate is practically the same as for octanoate, oleate, glycine,
and leucine. From this observation, the hypothesis is drawn that the bacterial popula-
tion of the B.O.D. reduction or methane fermentation stage of digestion is responsible
for the major if not essentially the total decomposition of all these substrates.
Next, carbohydrate digestion would likewise require a population in the B.O.D.
reduction stage comparable to acetate digestion since the loading rates were equal.
The hypothesis is then drawn that the difference between the net synthesis rate for
carbohydrates and that for acetate, for each particular detention time was due to the
bacterial population responsible for the constant B.O.D. stage of digestion. Con-
sidering this to be true, growth equations were evaluated for the two stages of carbo-
 Nutrient Requirements and Biological Solids Accumulation in Anaerobic Digestion 321
hydrate digestion assuming the second stage of this decomposition would be the
same as for acetate. The results of this evaluation gave the following equations:
First Stage (constant B.O.D.):
^1=0·406^-0·100Μι
Second Stage (B.O.D. reduction):
v4 2 =0-054F-0-038M 2
An interesting observation of endogenous respiration was made from the glucose
digesters. After final data were recorded, all of the units were left standing with no
feeding for nine weeks. The volatile solids concentrations in the units were analyzed
at the end of the final data period, six weeks later, and nine weeks later. TABLE 4
shows the results of the analyses at these intervals. Thus, after 60 days with no feeding,
from 75 to 81 per cent of the volatile solids were still present. This represents an
endogenous respiration rate of less than 0-005/day as compared to 0-088/day,
observed for the case when feed was being administered.

TABLE 4. ENDOGENOUS RESPIRATION OF BIO-

LOGICAL SOLIDS IN GLUCOSE DIGESTERS AFTER
STOPPAGE OF FEED

Volatile solids concentration mg/1.

weeks Sludge retention time—days

5 10 20 30

0 620 990 1970 2080

6 540 900 1540 1580
9 500 800 1540 1560

It is hypothesized that just as the bacterial population in the constant B.O.D.

stage appeared not to be able to derive energy from the amino and fatty acid group
due to the lack of a hydrogen acceptor, so also, in the absence of a reducible organic
substrate, endogenous metabolism is considerably diminished. The bacteria appear
not to be able to attack their own protoplasm without a hydrogen acceptor, other
than CO2, present.

CONCLUSIONS
From this study on the biological nutrient requirements and growth in anaerobic
digestion, the following conclusions have been drawn :
1. From the results of this study, it is possible to predict for a given substrate, the
nitrogen requirements, biological solids accumulation, net synthesis, and methane
production which will occur during anaerobic digestion for various sludge retention
times.
2. Two stages exist in the anaerobic digestion of complex substrates, one in which
the B.O.D. remains constant and another in which the B.O.D. is reduced due to
production of methane.
x
322 R. E. SPEECE and P. L. MCCARTY

3. In the case of carbohydrates the major nutrient requirement and major bio-
logical solids accumulation result from the first or constant B.O.D. stage of digestion.
4. The growth equation for anaerobic digestion of glucose and starch was:
v4=0-46F-0-088M.
5. The growth equation for anaerobic digestion of amino and fatty acids was:
A -0-054 i^-0-038M.
6. The growth equation for anaerobic digestion of nutrient broth was: .4=0-076
F-0-014M.
7. The nitrogen requirements for all substrates were equal to AJ9-4.
8. The phosphorous requirements for all substrates were approximately one-
seventh of the nitrogen requirements.
9. Exceptionally high rates of acetate digestion, 6500 mg/1. day (0-36 no./ft3 day),
can be achieved by addition of combinations of iron, cobalt, thiamine, and com-
ponents of vitamin B12 to digesters purged of the original seeding material.
10. The addition of inorganic salts alone with the pure organic substrates studied
in this investigation enabled satisfactory digestion.
This investigation was supported by PHS Research Grant WP-192 from the
National Institute of Health, United States Public Health Service.

REFERENCES
1
BARKER H. A., 1956. Bacterial Fermentation, Wiley, New York.
2
BENNETT E. O. and WILLIAMS R. P., 1957. Appl. Microbiol 5, 14-16.
3
BUSWELL A. M., 1957. Sewage andindustr. Wastes 29, 717-721.
4
BUSWELL A. M. and SOLLO F. W., 1948. Sewage Works J. 20, 687-694.
5
DUVALL A. J., STEFFES A. M. and ANDERSON J. J., 1956. Wastes Eng. 27, 516, 530, 532.
6
HEUKELEKIAN H. and HEINEMANN B., 1939. Sewage and Industr. Wastes 11, 436-444.
7
HEUKELEKIAN H., ORFORD H. E. and MANGANELLI R., 1951. Sewage and Industr. Wastes 23,
945-958.
8
HOOVER S. R. and PORGES N., 1952. Sewage and Industr. Wastes 24, 306-312.
9
MCCARTY P. L. and MCKINNEY R. E., 1961. Water Pollution Control Fed. 33, 399-415.
10
MCCARTY P. L. and VATH C. A. Int. J. Air and Water Pollution, April 1962.
11
MONOD J., 1942. Recherches sur la Croissance des Cultures Bactériennes, Herman and Cie, Paris.
12
PLACAK D. R. and RUCHHOFT C. C , 1947. Sewage andindustr. Wastes 19, 423-440.
13
RUCHHOFT C. C , KACHMAR J. F. and PLACAK D. R., 1940. Sewage andindustr. Wastes 12,485-503.
14
SCHROEPFER G. J., FÜLLEN W. J., JOHNSON A. S., ZIEMKE N. R. and ANDERSON J. J., 1955. Sewage
and Industr. Wastes 27, 460-486.
15
STANDER G. J., 1950. Inst, of Sewage Purif., 438-447.
16
STEFFEN A. J. and BEDKER M., 1960. Pub. Works 91, 100-102, 7.
17
STEWART M. J., PEARSON E. A. and HIRAMOTO E. M., 1959. Proc. Fourteenth Ind. Waste Conf.,
Purdue University, 309-339.
18
SYMONS J. M. and MCKINNEY R. E., 1958. Sewage andindustr. Wastes 30, 874-890.
19
WESTON R. F. and ECKENFELDER W. W., 1955. Sewage andindustr. Wastes 27, 802-820.
 Discussion 323

DISCUSSION

H. HEUKELEKIAN: The authors are congratulated for the attempt to evaluate and
furnish much needed information regarding the quantity of sludge produced in the
anaerobic digestion of various soluble organic compounds. The following comments
are in order:
(a) The amount of growth produced per unit amount of substrate utilized in terms
of C.O.D. as given in TABLE 3 varies from 2-6-7-1 per cent for acetate, octanoate,
oleate, glycine, leucine and nutrient broth while for glucose these values vary from
11 to 35 per cent and starch from 14 to 34 per cent. The yields for the fatty acids and
amino-acids and nutrient broth are reasonable values for anaerobic conditions.
Question is raised regarding the yields obtained with glucose and starch. It should be
noted that the yields for all the substrates used including glucose and starch decrease
with increasing retention time. Glucose with a sludge retention time of 5 days gave
a sludge accumulation of 35 per cent on the basis of C.O.D. utilized. This value is as
high as has been obtained for the substrate under aerobic conditions. Information
in the literature indicates that the biological growths produced from the aerobic
decomposition of glucose varies from 30 to 60 per cent on the basis of C.O.D. utilized.
Work done in our laboratory and presented at Purdue Industrial Waste Conference
in 1962 gave yields of 38 per cent on the basis of C.O.D. utilized in the aerobic
treatment by activated sludge at a sludge age of 3-3 days. Since under anaerobic
conditions the decomposition proceeds only part way with much diminished free
energy change, the cell yields should be lower than under aerobic conditions. The
authors have observed that the exceptionally high synthesis is accompanied with
reduced methane production from carbohydrates at short detention periods. The fact
that some methane was produced indicates that not all the chemical energy of glucose
was completely utilized and hence the synthesis should have been lower than under
aerobic conditions when the energy of glucose is converted to CO2 and H2O. This
latter statement might be contested on the basis that part of the energy of the substrate
is assimilated without oxidation and oxidized later when the exogenous substrate
becomes exhausted. However, the results obtained by McWhorter and Heukelekian
(see paper in Proceedings of the Water Pollution Research Conference in London) do
not support this argument since the cell yields after the exhaustion of exogenous food
did not increase.
It should also be pointed out that the synthesis values reported apply to wastes
containing only soluble carbohydrates. Lower values would be obtained with wastes
containing soluble nitrogenous organic materials and fats. Still lower values should
be obtained with wastes containing suspended solids. Lowest values should be
obtained with sewage sludge with small amounts of soluble carbohydrates and
relatively high concentrations of fatty materials and nitrogenous organic matter.
There is also considerable concentration of organic matter which is not fermentable.
Therefore, the net gain by synthesis is greatly counterbalanced by the loss of organic
matter by conversion to carbon dioxide and methane and accordingly a 50 per cent
reduction of organic matter is obtained.
(b) A few comments are in order regarding the authors' concept of constant B.O.D.
324 Discussion
and the B.O.D. reduction stages in anaerobic digestion. This corresponds to the
concept of two-phase digestion heretofore designated as liquefaction and gasification.
The liquefaction phase corresponds to the author's constant B.O.D. stage in which
complex organic materials are solubilized by a group of organisms through a series of
transformations by hydrolysis and oxidation-reduction reactions into lower volatile
acids without reduction in B.O.D. This is followed by conversion of these volatile
acids to methane and carbon dioxide by a specialized group of methane bacteria.
The actual B.O.D. reduction takes place during this phase. The authors state that the
greatest rate of net synthesis takes place during the constant B.O.D. stage or during
liquefaction phase with wastes containing carbohydrates coincident with decreased
methane fermentation. It is difficult to explain this conclusion since even such a
simple carbohydrate as glucose must first be transformed extracellularly before it can
enter the cell and be metabolized. Whether this is the case or whether the glucose is
transformed at the surface or inside the cell the amount of energy liberated during
the breakdown is small and hence one should not expect a high rate of synthesis
during this stage. It is realized that the amount of energy liberated during the
methane fermentation stage is also low and hence the amount of growth produced
should also be low.

A. M. BUSWELL {University of Florida, U.S.A.): The relative importance of sludge

digestion is emphasized if we divide waste treatment into two steps : (a) removal of
pollutional material and (b) stabilization of the material removed. We then note that
about 20 per cent of the material removed is stabilized by aerobic treatment leaving
80 per cent to be stabilized by anaerobic digestion. The most excellent paper now under
discussion together with a paper by the same group presented at the Purdue Conference
in May of this year goes far to explain the importance of seeding and the effect which
seeding produces on the successful initiation and operation of the digestion process.
Fully fifty years ago Dr. Karl Imhoff had found that for the successful starting of
fermentation in the sludge compartment of an Imhoff tank (new or recently cleaned)
it was necessary to add sludge from an active tank to an amount equal to 25 per cent
of the digestion compartment volume. The purpose of the "starter" was supposed
to be the inoculation of the raw sludge with methane producing bacteria. Later our
group in Urbana, Illinois, found that if fibrous material such as paper fiber, straw, or
cornstalks were provided, successful starting could be accomplished with a relatively
small amount of sludge or even with overflow liquor. This led to the conclusion that
the sludge used by Imhoff served a second purpose in addition to bacterial inoculation,
namely to provide a surface or nidus on which the bacteria could rest. For work
with pure soluble compounds where a carbon balance was to be made, it was found
that shredded asbestos made a satisfactory nidus.
Now the present work shows that seed sludge serves a third purpose, namely to
supply certain "accessory" nutrients. This constitutes a major breakthrough.
It should be noted without prejudice that the present paper is based entirely on
data obtained with acetic acid. It has been shown that the breakdown of acetic acid
to give methane is dicarboxylation while the production of methane from other fatty
acids appears to follow a much more complicated path. If propionic acid, for example,
were to be decarboxylated it would yield ethane, but there is extensive evidence that
ethane is never formed in sludge digestion.
 Discussion 325
With respect to propionic acid, it is worthy of note that four papers within the last
year have reported that propionic acid digests much more slowly than either acetic
or butyric. And in two cases of "stuck" tanks where the volatile acids ran 10,000 ppm,
Chromatographie analyses showed the preponderance of propionic. Experiments were
recently reported from the Phelps laboratory in Gainesville which indicated a much
higher tolerance for acetic than propionic acid on the part of the sludge organisms.

D. E. BLOODGOOD {Purdue University, U.S.A.): In the anaerobic digestion of

organic matter it has been found in research at Purdue University, Lafayette, Indiana,
U.S.A., that the presence of ammonium bicarbonate in the system is necessary to
react with the volatile acids produced in the first stages of the process so that the
environment is never unfavorable to the organisms that produce methane.
In two field situations "stuck" digestion tanks have been started by the addition
of anhydrous ammonia to the digestion tanks in a quantity slightly in excess of the
amount needed to react with the volatile acids that are present. It is reasonable to
believe that the ammonia reacts to form ammonium acetate, is available for cell
synthesis and forms ammonium bicarbonate with carbon dioxide. The concentration
of the natural buffer in a digestion tank is a function of the amount of protein in the
organic feed, the rate of break-down of this material to ammonia and the retention
period in the digestion tank.

F. J. AGARDY, R. D. COLE and E. A. PEARSON (San Jose State College, San Jose,
California, U.S.A.): The authors have presented a number of interesting conclusions
as a result of their experimentation, with the most significant factor being the apparent
stimulation of acetate utilization in the presence of specific nutrients. Extensive
studies carried out over the years concerning the digestion of domestic sludge, a
substrate believed to contain essentially all required nutrients in excess, have courted
a "nearsighted" attitude regarding the use of supplemental nutrients to aid in the
anaerobic decomposition of specific organic substrates.
However, the authors did encounter problems in attempting to relate substrate
reduction to organism concentrations. Only when the substrate, intermediate and
product employed were soluble could the conventional volatile suspended solids
criteria be used to characterize biological masses. When the presence of insolubles
negated this procedure, measurements based on the organic nitrogen of centrifuged
and washed solids were used. Subsequently, a conversion factor was employed to
convert to a volatile solids base. In either case, a quantitative measure of the organic
catalyst, as appears in the following equations, was attempted.
bacteria
substrate ► > > ^products (1)
or, more accurately,
enzymes
substrate > ► ► »-products (2)
Admittedly, the measurement of the bacterial content of a digestor is a difficult
task and the quantitative measurements of the total enzyme concentration in a
heterogeneous system is at present impossible. However, if kinetics and growth studies
 326 Discussion
are to have any validity, some accurate measure of these system parameters is neces-
sary. Therefore, as a base against which to make such studies we have selected the
most stable constituent of a bacterial cell, namely, the deoxyribonucleic acid (D.N.A.)
content.1'2

Kinetics
Kinetics studies carried out by Agardy, Cole, and Pearson 3 during the past two
years have substantiated both the feasibility and the accuracy of D.N.A. evaluations
as a measure of the bacterial cell population. FIGURE 1 shows the relationship between

FIG. 1.

rate of substrate utilization and D.N.A. content as expressed by an equation modeled

after Michaelis:
Ky
M= (3)
Y+y
where: M i s the rate of substrate utilization, fC.O.D. removed/#D.N.A.-day in
reactor
y is the reactor substrate concentration, #C.O.D./#reactor contents
K is the maximum theoretical rate of substrate removal at infinite substrate
concentration
Y is the substrate concentration at 1/2 AT
The growth relationship for the same system is shown in FIG. 2 with the equation
taking the form:

-=cM-k (4)

where: - is the time required to replace the total system organism (D.N.A.), d a y 1
c is a constant, fraction of substrate synthesized
k is a constant, endogenous respiration rate, day - 1
The substrate employed was a dextrose-beef extract-tryptone mixture dissolved in
 Discussion 327
distilled water. The system was continuously fed and fixed, and the reaction
temperature was 100° F. The resulting constants were evaluated:
#=18-28 #C.O.D./#D.N.A.-day
7=6700 mg/1. C.O.D. in system
c =00051 #D.N.A./#C.O.D.
k =0-026/day
It was further found that in this soluble substrate system, the ratio of D.N.A. to
the dry weight of cells (as measured by volatile suspended solids) was 4-6 per cent
with a correlation coefficient, r=0-988. This ratio was in close agreement with values
reported by Vendrely4 (4-1 per cent) and Webb and Levy5 (4-3 per cent) for pure
culture bacterial cells.
A two-month field study on prototype digesters at three sewage treatment plants
confirmed the applicability of D.N.A. analyses as a measure of the biological popula-
tion of the systems. Thirteen D.N.A. analyses coupled with routine T.S., Y.S. and

0-09

0-08

O07
0-06

~ 0-05

0-04

O03
0-02

OOI

0 5 10 15 20 25
M
FIG. 2.

performance analyses yielded a reasonable degree of reproducibility, and average

values of D.N.A./V.S. were 3-5, 3-8, and 3-7 per cent for the three different plants.
The results of this study will be published shortly in a separate paper.
Although the analytical procedures whereby D.N.A. is extracted and evaluated
require further study as regards the effect of interfering substances, there is no doubt
that a more rational basis for kinetics and growth studies is now available and should
be further exploited.

Kinetic comparison
Kinetic aspects of continuous flow steady-state biological systems have not been
considered adequately in the design of laboratory experiments or in the design or
performance predictions of biological systems. Consideration of the classic growth
equation, \\T=cM—k by the authors along with the rational Michaelis model
expression for the substrate removal rate as a function of substrate concentration
328 Discussion

[M=Kyl(Y+y)] and evaluation of the constants, c, k, # a n d Y permits prediction of

performance for any combination of substrate and cell concentration as well as
substrate and organism residence times. The value of such predictions in the selection
and design of appropriate biological systems as well as the reliable prediction of
performance is obvious.

TABLE 1—COMPARISON OF BIOLOGICAL PROCESS KINETIC CONSTANTS

- = cM—k and M=
T Y+y

c k K Y
Process Substrate #v.s.s./ day - 1 #B.O.D./ ppm— References
#B.O.D. #V.S.S.-day B.O.D.

Settled sewage 0-50 0055 Heukelekian 6

Sew. and Ind. Waste
mixture 0-5-0-6 1 Helmers et al.7
Aerobic "") Skim milk 0-48 0045 51 100 Gram 8
activated V- Glucose 0-422 0087 30 355 Stack 9
sludge J Spent sulfide liquor 0-55 Eckenfelder and
O'Connor 1 0
Kraft (pulp and 0-76 Eckenfelder and
paper) work O'Connor 10
Refinery 0-70 Eckenfelder and
O'Connor 10
Brewery 0-93 Eckenfelder and
O'Connor 1 0
Pharmaceutical 0-77 Eckenfelder and
O'Connor 1 0
Synthetic fiber 0-38 Eckenfelder and
O'Connor 10

Anaerobic Ί Beef extract, dextrose,

continuous [_ peptone 0182 0025 0-374 57006 Stewart et al.11
flow f Glucose and starch 0-462 0088 Speece et al.1'2
fermenters J Amino and fatty
acids 0-0542 0038 Speece et al.12
Nutrient broth 0076 2 0014 Speece et al.12
Tryptone, dextrose,
beef extract 000513 0026 18-35 67006 Agardy et al.s
Tryptone, dextrose,
beef extract 0-1142 0030 0-7894 5030 Agardy et al.3

Notes: 1. # S . S . / # B . O . D . 3. # V . S . S . / # D . N . A . 5. # C . O . D . / f D.N.A.-day
2. f V . S . S . / # C . O . D . 4. f C . O . D . / f V.S.S.-day 6. ppm C.O.D.

TABLE 1 presents a summary tabulation of both aerobic and anaerobic process

constants for the growth and substrate removal functions from all available literature
as reported by the researchers or computed from their data. It is of interest to com-
pare the values of " c " and "&" particularly for aerobic and anaerobic systems with
similar substrates. For example, Stack's data 9 indicate a conversion efficiency of
0-42 #Y.S.S./#C.O.D. for aerobic activated sludge with a glucose substrate, while the
authors report a value of 0-46#V.S.S./#C.O.D. for an anerobic system with a glucose
 Discussion 329
and starch substrate. Moreover, the endogenous rate constants were practically
identical. While it is recognized that the values of c are reported on different systems,
nevertheless it is indicative of comparable synthesis for the two systems. Whether this
is true needs further exploration.
It is only by means of development of the process constants c, k, K and Y from
laboratory or pilot plant data that an engineer can accurately select the most economic
process for a given substrate and performance requirement, and predict performance
for a variety of operating conditions.

Nutrient stimulation
A number of their points are worthy of comment, the first of which is concerned
with the "discovery" that iron is a required nutrient in the fermentation of acetate.
Barker13 states that the fermentation is equivalent to a decarboxylation, not requiring
CO2 as a hydrogen acceptor. Dixon and Webb14 report that decarboxylative enzymes
as a rule require divalent or trivalent cations as a co-factor. Furthermore, Gale 15
states that formic hydrogenlyase, the enzyme which anaerobically breaks down
formate to hydrogen and carbon dioxide, will not function in an iron-deficient
medium. Similarly, butyric fermentations proceed more favorably in the presence of
iron.16 In light of the above, it would seem that the authors confirmed rather than
discovered that iron deficiencies might well curtail acetate utilization.

Material balance
A second point of interest centers on the employment of methane production as a
check on the system C.O.D. utilization.
A mass balance based on influent-effluent C.O.D. can be checked theoretically by
evaluating the equivalent C.O.D. of methane produced, provided no other oxidizable
gas such as H2S or H2 is a by-product of the fermentation. Admittedly, there are
analytical errors in measurement of the above-mentioned gases which would prevent
a 100 per cent C.O.D. materials balance. But one other factor will perhaps have a
greater effect, namely, the non-oxidation of acetate in the dichromate C.O.D.
test.17'18 Only if silver sulfate is used as a catalyst will acetate and other straight chain
acids be oxidized.19
The following equation illustrates the bacterial decomposition of organic matter:
bacteria (lowest oxidation state)
organic matter *C02+CH4 t + organic matter
(low oxidation state) (higher oxidation state) j
and a C.O.D. balance takes the form:
C.O.D. (in)-C.O.D. (out)-C.O.D. (CH4)
If the conventional C.O.D. test is used and silver sulfate is not added, then the
presence of acetate or other straight chain acids in either the influent or effluent will
result in an incorrect C.O.D. balance which will not agree with the C.O.D. equivalent
of the methane produced. The authors apparently recognized this problem when
acetate was the substrate and therefore employed volatile acids as the basis for the
materials balance, with the reaction equation being :
CH3-COOH *C02+CH 4
330 Discussion
However, in the case of the apparent 100 per cent removal of acetate, it is difficult to
see how the computed methane C.O.D. results in a materials balance in excess of
100 per cent. Is it possible that the "classic" concept of water entering into the
reaction is applicable here ?

G. J. STANDER: The two advantages quoted by the authors on page 18 of their paper
are of considerable economic importance in large-scale application. The addition of
nutrients is a matter which should be carefully studied with the object of operating the
process in such a way as to limit their use. The answer for this seems to be found in the
development of a matured sludge which could be achieved by a long residence time
of the sludge. The results of our pilot plant studies on wine distillery wastes and large-
scale anaerobic digestion of glucose-starch wastes (Water and Waste Treatment
Journal, May 1962) bring this latter point out very clearly. The establishment of a
matured sludge seem to be the building up of a healthy balance between exogenous
(cell cynthesis) and endogenous (cell lysis, i.e. breakdown) enzymic reactions in the
breakdown of organic substrate. The process of lysis in bacterial reactions is known to
liberate nutrient elements and food material in suitable form for assimilation in the
exogenous phase. It may be of interest to note that this process is explained in the
extended aeration (endogenous respiration) process. It seems that the supply of nutri-
ents requires an endogenous phase in the anaerobic digestion system and this could be
achieved by operating the system under conditions of high hydraulic loadings and
long sludge retention periods. We are carrying out our present basic research using the
concept of a balance between exogenous and endogenous reactions as a working
hypothesis. Consequently I cannot go into details here.
The achievement of short hydraulic retention times and long sludge retention is not
a problem in laboratory scale work but presents a major issue in large-scale applica-
tion. The retention of sludge sufficiently long to establish maturity in an active fer-
menting mass could not be achieved in our pilot experiments initially, in consequence
of which we could not confirm laboratory results. As reported in our aforementioned
publication we eventually succeeded in overcoming this difficulty bypassing the organic
waste through the fermenting mass using the Dorr Oliver clarigester on the reverse
flow principle. In this way sludge was constantly returned to the digesting mass. At a
daily loading of 50-60 x 103 gallons per day and a hydraulic loading of 2-5 to 3-5 days
the process has been in operation now for over 3 years without special nutrient addi-
tion. Sludge concentration for optimum operation is 50 to 60 g/1. Beyond this de-
sludging becomes necessary. Every 3 months about 20,000 gallons of a gelatinous
sludge are voided from the system without adverse effects. Incidentally I would like
to mention that in attempts to find a use for this material vitamin B12 was found to be
present in the dewatered sludge cake in a concentration three times that produced in
the commercial process for vitamin B12. Contrary to the authors' observations on
page 6, vitamin B12 seems to be a product of anaerobic digestion rather than a require-
ment.

SPEECE and MCCARTY : The authors appreciate the comments and contributions of
the discussors. Agardy and co-workers have utilized a clever and useful technique
for determining biological mass. Their technique of determining the D.N. A. content
and relating this to biological mass, enables the biological mass to be determined in
 Discussion 331
systems where volatile solids and organic nitrogen would be meaningless. However,
use of D.N.A. as a basis for calculating the growth equation in this study would not
have altered the results obtained.
Basically, the authors disagree with their assertion that comparable synthesis would
occur under aerobic and anaerobic conditions for a given substrate. The available
energy under anaerobic conditions is less than under aerobic conditions, since much
energy is released in the form of methane. However, synthesis is more probably
related to the relative amount of adenosine triphosphate (A.T.P.) formed by the cells
in the breakdown of the substrate. Unfortunately, the metabolic pathway is not
defined for the anaerobic transformation. Literature is cited to justify their statement.
However, the literature also contains data to indicate synthesis of glucose may be
75-80 per cent. We believe that the reduced synthesis under anaerobic conditions is
one of the advantages of anaerobic digestion.
The authors would agree with Agardy et al, in that they were a bit presumptive
to use the word "discovery" in reference to its stimulation potential for acetate
digestion. The references cited are logical. However, they are more "logical" in
hindsight than in foresight. About 30 compounds which appeared to be "logical"
stimulants from accounts in the literature were assayed with no success.
Concerning the reference of Agardy et al. to the discrepancy in C.O.D. balance,
silver sulfate was used as catalyst in all C.O.D. determinations to insure oxidation of
short chain fatty acids. It is difficult to see how the effect of water entering into a
reaction could affect the C.O.D. balance.
Dr. Weinberger has commented on the high concentrations of iron which produced
stimulation. The authors would point out that high concentration of iron can possibly
result in ion toxicity in an unacclimated digester.
The authors are also in agreement with Dr. Weinberger that C.O.D. is not a good
parameter upon which to base energy. However, due to our present state of knowledge,
it appears to be one of the more universal determination of energy. Dr. Weinberger
has expressed a desire that the authors express loading rate on the basis of organisms
rather than digester volume. This information is given in the last column of TABLE 3.
However, it will vary with different operating conditions, as shown, and, therefore,
is not as significant a basis for the expression of loading rate.
A question is raised by Dr. Weinberger about the absence of volatile solids data
for the acetate digestion run. This determination was not made during the data
period. However, it was experimentally verified to correlate with organic nitrogen
in a later study.
A significant point has been raised by Dr. Weinberger regarding the constant
B.O.D. stage. The authors are referring to the C.O.D. in solution plus that in sus-
pension in the waste stream.
Since, as he has pointed out, a fraction of the soluble waste is synthesized into
particulate cellular material and is removed from solution. Therefore, it must be
emphasized that the constant B.O.D. stage must take into consideration both the
C.O.D. in solution and the C.O.D. in suspension.
Dr. Weinberger has cited the reference to the possibility of nitrogen fixation in
anaerobic digestion. The authors noted an immediate response in the digester activity
when the ammonia nitrogen in solution was exhausted. The digestion rate decreased
due to the incapacity of cells to form protoplasm in the absence of a nitrogen source.
332 Discussion

Practically speaking, the authors believe that nitrogen fixation would not provide
significant amounts of nitrogen for synthesis in anaerobic digestion.
Dr. Buswell has noted the stimulatory effect of the addition of solid surface to an
anaerobic digester receiving a soluble substrate. The authors, however, did not
observe stimulation of digestion after additions of asbestos, bentonite or cellulose.
Dr. Buswell has raised a question concerning the authors' hypothesis that it appears
the bacterial population of the B.O.D. reduction stage of digestion is responsible for
the major if not essentially the total decomposition of the fatty and amino-acids.
He has cited instances to show that different methane bacteria are required to break
down propionic acid than those which utilize acetic acid. The authors would stress
that they are referring to mass of methane bacteria and not type of methane bacteria
in making the hypothesis.
The authors were pleased to hear of Dr. Standers's work along these lines and the
excellent results he has obtained. His work has been of interest to them since the
early part of this study.
The question has been raised by Dr. Heukelekian as to how it can be possible to
obtain 50 per cent reduction of volatile matter in the anaerobic digestion of sewage
sludge, if such high rate of synthesis prevails. Two reasons would be forwarded.
First, sludge retention times of ten to thirty days are maintained and second, the
volatile matter destroyed would not consist of 100 per cent carbohydrate content.
Both of these factors would effect lower net synthesis, and thus, 50 per cent reduction
of volatile matter would be feasible for anaerobic digestion of sewage sludge.

REFERENCES
1
LITWACK G., 1960. Experimental Biochemistry.
2
DAVIDSON J. N., 1960. The Biochemistry of Nucleic Acids. John Wiley, New York.
3
AGARDY F . J., COLE R. D . and PEARSON E. A. Activity parameters in anaerobic fermentation
systems (in preparation).
4
VENDRELY R., 1947. Un Symposium sur les Proteins. Masson et Cie. Paris.
5
WEBB J. M. and LEVY H. B., 1955. A sensitive method for the determination of deoxyribonucleic
acid in tissues and microorganisms, J. Biol. Chem. 213, 107.
6
HEUKELEKIAN H., ORFORD H. E. and MANGANELLI R., 1951. Factors affecting the quantity of
sludge production in the activated sludge process, Sew. and Ind. Wastes, 23, 8.
7
HELMERS E. H., FRAME J. D., GREENBERG A. E. and SAWYER C. H., 1952. Nutritional requirements
in the biological stabilization of industrial wastes. III. Treatment with supplementary
nutrients, Sew. and Ind. Wastes, 24, 4.
8
GRAM A. L., 1956. Reaction kinetics of aerobic biological processes, Report No. 2, T.E.R. Series
90, San. Eng. Res. Lab., Dept. of Eng., University of California.
9
STACK V. T. and CONWAY R. A., 1959. Design data for completely-mixed activated sludge treatment,
Sew. and Ind. Wastes, 31, 10.
10
ECKENFELDER W. W. and O'CONNOR D . J., 1961. Biological Waste Treatment. Pergamon Press.
11
STEWART M. J., PEARSON E. A. and HIRAMOTO E. M., 1959. Reaction kinetics of continuous flow
anaerobic fermentation processes, Proceedings of the 14th Industrial Waste Conference, Purdue
University, 309.
12
SPEECE R. E. and MCCARTY P. L., 1962. Nutrient requirements and biological solids accumulation
in anaerobic digestion, Int. Conf. on Water Pollution Research, London, September 3-7.
13
BARKER H. A., 1956. Bacterial Fermentations. John Wiley. New York.
14
DIXON M. and WEBB E., 1958. Enzymes. Academic Press, New York.
15
GALE E. F., 1947. The Chemical Activities of Bacteria. London University Tutorial Press, Ltd.
 Discussion 333
16
THIMANN K. V., 1955. The Life of Bacteria. MacMillan, New York.
17
NEISH A. C , 1952. Analytical methods for bacterial fermentations, National Research Council of
Canada, Saskatoon.
18
MOORE W. A., KRONER R. C. and RICHHOFT C. C , 1949. Dichromate reflux method for determina-
tion of oxygen consumed, Analytical Chemistry 21, 8, 1953.
19
MOORE W. A., LUDZACK F. J. and RICHHOFT C. C , 1951. Determination of oxygen-consumed
values of organic wastes—a comparative study of methods, Analytical Chemistry 23, 9, 1927.