WATER POLLUTION CONTROL RESEARCH SERIES • ORD- 17O5OFIMO5/7O
"OPTIMIZING LIPID
BIOSTABILIZATION"
U.S. DEPARTMENT OF THE INTERIOR • FEDERAL, WATER QUALITY ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the results
and progress in the control and abatement of pollution of our
Nation's waters. They provide a central source of information on
the research, development, and demonstration activities of the
Federal Water Quality Administration, Department of the Interior,
through in-house research and grants and contracts with Federal,
State, and local agencies, research institutions, and industrial
organizations.
Water Pollution Control Research Reports will be distributed to
requesters as supplies permit. Requests should be sent to the
Planning and Resources Office, Office of Research and Development,
Federal Water Quality Administration, Department of the Interior,
Washington, D. C. 20242.
-------
OPTIMIZING LIPID BIOSTABILIZATION
by
William Garner
Midwest Research Institute
Kansas City, Missouri 64110
for the
FEDERAL WATER QUALITY ADMINISTRATION
DEPARTMENT OF THE INTERIOR
Program #17050 FIM
Contract #14-12-198
FWQA Project Officer, R. C. Brenner
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
May, 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price 60 cents
-------
PWQA Review Notice
This report has been reviewed by the Federal
Water Quality Administration and approved for
publication. Approval does not signify that
the contents necessarily reflect the views
and policies of the Federal Water Quality
Administration, nor does mention of trade
names or commercial products constitute
endorsement or recommendation for use.
-------
TABLE OF CONTENTS
Page
Table of Contents ii
Abstract ill
Acknowledgements iv
Introduction 1
Experiment I - Performance of Digesters on Standard Loading and
Mixing Regimen 5
Experiment II - Effect of High-Shear Homogenization on the
Digestion of Purina Dog Chow 9
Experiment III - Effect of Homogenization on Limed Digesters
Heavily Loaded With Cottonseed Oil 12
Experiment IV - Effect of Homogenization on Soda-Ash Buffered
Digesters When Dog Food Loading is Augmented With Cottonseed
Oil 17
Experiment V - Effect of Homogenization on Soda-Ash Buffered
Digesters as Cottonseed Oil Gradually Replaces Purina
Dog Chow as a Feed 20
Experiment VI - The Effect of Motor Oil on the Digestion of
Purina Dog Chow and Cottonseed Oil 23
Discussion 26
Conclusions and Recommendations 30
Appendix A - Apparatus and Methods 32
Appendix B - Analytical Data 49
Bibliography 55
ii
-------
ABSTRACT
Laboratory-scale anaerobic digestion studies were carried out
to determine the effect of high-shear mixing (homogenization) on the deg-
radation of lipids . The studies showed that the intensity of mixing must
be carefully tailored to the rate and type of feed if benefits are to be
realized.
Dog food, cottonseed oil and motor oil were fed to the digesters
daily as slug loads. When properly operated, the 15 replicates digesters
gave reproducible results with good precision.
High-shear mixing has little effect on degradation of dog food
and was deleterious to lime-buffered solutions heavily loaded with cotton-
seed oil. When soda ash was used as a buffering agent, heavy loads of oil
caused the digester to "go sour" regardless of mixing system, although
homogenization gave somewhat better results. Under some conditions, homo-
genization led to serious foaming in the digesters . Kinetic data and
chemical analysis of drawdown samples confirmed the hypothesis that homo-
genization or soda ash buffering can accelerate the hydrolysis of a heavy
load of fat to a point where the saponif ication products overload the
methane fermentation or affect it by surface effects. Motor oil was not
readily digested, but did not appear to influence the digestion of cotton-
seed oil or dog food.
Other observations were: more rapid gasification occurs as the
feed rate is increased; lime and. soda ash cannot be used interchangeably
as buffers; feed COD cannot be used to predict methane yield; and the tem-
perature response of a system greatly exceeds Q^Q = 2.
This report was submitted in fulfillment of Program No. 17050 FIM,
Contract No. 14-12-198, between the Federal Water Quality Administration and
Midwest Research Institute.
iii
-------
ACKNOWLEDGEMENTS
This report was written by Dr. William Garner, who was also the
project leader. Other Midwest Research Institute personnel who contributed
to the project were Messrs. Eugene Sallee, Tom Shearon, John Morrison,
John Maberry, Earl White, Aubrey Cloud, John Unrau, Merrill Nissen,
Paul Garrett, Jack Honts, Mrs. Jeanne Robertson and Dr. Walter Langston.
Dr. Ross E. McKinney, Parker Professor of Civil Engineering, University of
Kansas, and Dr. Raymond C. Loehr, Professor of Civil and Agricultural
Engineering, Cornell University, were consultants to the research program.
Dr. J. Earl Barney II, former Head of the Analytical and Environmental
Sciences Section, provided technical supervision for the project.
iv
-------
INTRODUCTION
Wastewater must be renovated before it is returned to the envi-
ronment or reused. At present, domestic wastewater is processed in the
United States in order to remove pathogens and reduce the organic matter
which, in serving as growth media for common soil and aquatic bacteria,
would deplete the receiving stream of dissolved oxygen. The two unit
processes of biosorption and settling are employed for this purpose. Set-
tled organic matter consists of raw wastes as well a^ the flocculated
bacteria that have adsorbed and absorbed the non-settleable organic matter
in the stream. This settled organic matter is highly putrescible, and it
must be stabilized before it is spread on agricultural land, used as land-
fill, or disposed of at sea.
Sludge can be stabilized by chemical oxidation in the processes
of incineration or wet oxidation. It can also be stabilized by microbio-
logical oxidation and by the process of methane fermentation (or anaerobic
digestion). This latter process has been included in the design of most
of the larger plants that have been constructed in the United States during
the past several decades.!./ Digestion has many attractive features, but
the trade-offs are such that abandonment of this process in favor of oxi-
dation processes is gaining favor. Table I presents the advantages and
disadvantages of anaerobic digestion.
The wide variety of natural substances of biological origin that
are insoluble in water and soluble in such non-polar solvents as ether,
chloroform, and benzene are classified by the technologist as "lipids" or
"lipides," which is derived from the Greek HTTOS for fat. "Grease" is a
classification that is designated by the feel of the substance and the
ability of the material to make transluscent paper transparent. The term
"fat" implies a biological source, but a chemist would restrict the term
to describe the variety of liquids and solids that are formed naturally
by the esterification of long chain carboxylic acids and glycerol.
Loehr and Kukar=-' have shown that the complete spectrum of natural
lipids is to be found in wastewater. Hunter and Heukelekiaoi' place the
quantity of grease in Highland Park, N. J., sewage at 25%. A variety of
industrial wastes, especially from packing houses, dairies, and canning
plants, would increase the level of fat, while petroleum products, espe-
cially automotive motor drain oil and shop floor scrubbings, would con-
tribute a nonglyceride fraction.
-------
TABLE I
ADVANTAGES AND DISADVANTAGES IN ANAEROBIC
DIGESTION OF MUNICIPAL WASTE SLUDGES
Advantages
Approximately 2/3 of the BOD
in wastewater stream can be
processed without cost of
aeration
Can process hydrophilic sludges
of low solids content
Digested sludge more filterable
than fresh sludge
Low ORP inhibits corrosion
Some aerobically bioresistant
materials are digested
Produced methane can provide
power for plant
Disadvantages
Equipment is more expensive than
aeration equipment
Malodor from H2S and other
anaerobic metabolities
Process variables of pH, feed rate,
N&P nutrition, temperature and
inorganic concentration must be
carefully controlled
Air-oxidized H2S is highly cor-
rosive
Difficult to initiate fermentation
Grease is concentrated in digesters
as scum layers
Oxygen is lethal to process
Produces a strong supernatant liquor
which constitutes a heavy organic
and nutrient loading to main bio-
logical treatment process
A biochemical analysis of lipid digestion anticipates two sources
of difficulty. First, normally occurring wild bacteria will not be able
to metabolize certain synthetic molecular structures and some petroleum
derivatives. There is no definitive study in the sanitary engineering
literature on the relative rates of degradation of the various molecular
species of hydrocarbons although Davis^t/ presents sufficient evidence to
establish that different rates do indeed exist. Some of Davis' own re-
search^.' indicates that the normal paraffins are more rapidly oxidized by
some aerobic microorganisms. Beerstecher,2/ on the other hand, states
that the branched chain hydrocarbons are more susceptible to oxidation.
-------
Jeris and McCartyZ/ have shown that microbial mechanisms are
available in a methane fermenter to convert palmitic acid and water to
methane and carbon dioxide. The same order of thermodynamic potential
exists for decane:
CH3 - (CH2)s - CH3 + 6H20 > 7CH4 + 3C02 (AF - -37 kcal/mole)
Davisi' cites an abundance of evidence that the sulfate-reducing
bacteria can oxidize hydrocarbons. Anaerobic breakdown of hydrocarbons by
other bacteria has not been investigated, except for a report by Chouteau
et al.— that Pseudomonas aeruginosa could dehydrogenate n-heptane.
The second criterion for lipid digestion is the actual contact of
the microorganism and the surface of the lipid. While in the metabolism
of soluble materials, the rate of nutrient utilization is a function of the
concentration of nutrient, the rate of metabolism in a heterogeneous sys-
tem of lipid-water-bacteria is related to the actual interfacial area rather
than bulk amounts of material. Desnuelle—' has shown that surface area is
rate-limiting in the lipase-catalyzed hydrolysis of triglycerides. This
concept is easily extrapolated to the problem of scum-layer formation in
a sludge digester. Lipids, even though degradable, separate and float to
the top of the digester tank, forming a scum layer. Once this material has
separated, it is unavailable as a medium for bacterial growth unless me-
chanical energy is introduced to redisperse the lipids in the water.
McKinneylx' has recognized surface area as a possible rate-limiting factor
and suggested a high-speed dispersing nozzle in the sludge recycle system
as a solution to the grease scum problem.
The research described in this report was designed to evaluate
this redispersal of scum layers as the rate-limiting process in digestion
of a typical lipid that is recognized as biodegradable. The usual anaerobic
digestion system was modified to permit high-shear homogenization of any
lipids that separated as a supernatant layer. The protocol for the experi-
ments was designed to answer the following questions:
1. Does high-shear homogenization have any observable effect on
an established digestion pattern where lipids are present only in very
small concentrations?
2. Does high-shear homogenization have any effect on the diges-
tion process when high concentrations of a biodegradable lipid (cottonseed
oil) are present?
-------
3. Does the presence of a non-biodegradable lipid (motor oil)
influence the effect of high-shear homogenization on the biodegradation of
a biodegradable lipid (cottonseed oil)?
The above questions have been answered by the following series
of experiments:
Experiment I - Performance of 15 replicate laboratory digesters
on a standard loading and mixing regimen of Purina Dog Chow. (Question 1)
Experiment II - Effect of high-shear homogenization on the diges-
tion of Purina Dog Chow. (Question 1)
Experiment III - Effect of homogenization on limed digesters
heavily loaded with cottonseed oil. (Question 2)
Experiment IV - Effect of homogenization on soda-ash-buffered
digesters when dog food loading is augmented with cottonseed oil. (Ques-
tion 2)
Experiment V - Effect of homogenization on soda-ash-buffered
digesters when cottonseed oil gradually replaces Purina Dog Chow as a
feed. (Question 2)
Experiment VI - Effect of motor oil on the digestion of Purina
Dog Chow and cottonseed oil. (Question 3)
The data from these studies also have been analyzed for any
other possible information that might be useful for improving the anaerobic
digestion of wastes.
-------
EXPERIMENT I
PERFORMANCE OF DIGESTERS ON STANDARD
LOADING AND MIXING REGIMEN
The data for the baseline studies must be derived from several
different time periods. The overall experimental design was such that the
digesters were "fed" to a level that produced failure. Thus, after each
experiment that involved an overload "feed," there was a period of re-
adaptation or re-acclimation to establish the bacterial population. This
was followed by a period of several days' operation at a constant feed
level.
Initially the digesters were filled to the 1 gal. (2,785 ml)
level. The daily drawdown was 175 ml; the detention time, 21.6 days. The
digesters were inoculated with screened digester sludge and fed at a rate
of 1.32 g/^/day of dog food for 17 days. The feed level was then increased
incrementally to 3.96 g/^/day. During the period 10-9-68 to 10-15-68, all
15 digesters were stirred by head gas recirculation only. This period can
be compared to similar "blank" periods of 11-5-68 to 11-11-68, 11-21-68 to
12-5-68, and 3-9-69 to 3-14-69. The average gas production of the 15 di-
gesters during the period 10-9-68 to 10-15-68 is plotted as Curve B in
Figure 2, p. 10. This curve is quite linear.
The data from each of the four "blank" periods are presented in
Table II. The most significant information to be derived from the data are
the variability of the digesters in each period. This is presented as the
standard deviation divided by the average. Thus the four periods can be
compared. The value of 31.47. for the standard deviation for the period
10-9-68 to 10-15-68 is well within the average for microbiological systems
while the other three sets of data show that the variability during these
"blank" periods was very low for such a system.Ii/ One approach to the
overall variability of the system is to average the four values of the
standard deviations after weighting for the time period. This overall
standard deviation was 12.47». During the entire period of 10-24-68 to
1-7-69, the average gas production was 277 I with a standard deviation of
only 6.67.. The literature does not have comparable values because most
previous studies with anaerobic digesters have been performed with a single
unit.
During the nine-day period, 11-12-68 to 11-20-68, inclusive, the
digesters were operated without head gas recirculation or liquid recircula-
tion. The digesters were stirred only by occasional shaking. During this
period, the 15 digesters produced an average of 33.75 SL of gas from a
loading rate of 3.96 g/je/day. The gas yield was 0.250 4/g feed, but the
-------
TABLE II
PERFORMANCE OF DIGESTERS DURING "BIANK" PERIODS
Cumulative Gas Production in Liters
Digester
1A
IB
1C
2A
2B
2C
3A
3B
3C
4A
4B
4C
5A
5B
5C
Average
Standard deviation
as % of average
Feed rate g/4/day
Gas yield //g feed
Average methane con-
tent as %
10-9-68 to
10-15-68
(6 days)
17.55
30.70
22.74
29.76
30.73
45.45
22.82
15.93
19.11
30.00
39.14
30.01
45.51
3 9. .64
28.56
11-5-68 to
11-11-68
(6 days)
30.52
30.89
29.72
29.81
30.63
30.82
25.30
28.86
30.10
28.14
27.02
28.34
23.36
26.99
28.14
11-21-68 to
12-5-68
(15 days)
66.54
65.07
67.32
70.09
72.05
73.11
58.94
70.48
67.37
57.73
62.64
65.29
52.44
63.19
63.68
3-9-69 to
3-14-69
(5 days)
16.88
17.34
19.42
16.51
18.58
19.73
17.58
17.71
17.94
17.85
16.97
15.79
16.22
16.29
18.27
29.85
28.58
65.06
17.54
31.4
3.96
0.331
7.7
3.96
0.318
8.6
3.96
0.289
6.6
5.56
0.292
54
54
49
-------
nine-day period includes one day during which the bath temperature dropped
to 20°.C. The decrease in gas yield of this period when compared with the
preceding and succeeding periods does not appear to be significant.
The baseline level of feed was to have been established at 75%
of the feed level that would cause one-third of the digesters to malfunc-
tion either by becoming sour or passing through the feedstock undigested.
It was not possible to produce biochemical malfunction as the feed level
of dog food could be raised to only 20 g/gal. before the liquid pumping
system became clogged. This is comparable to a loading of 0.28 Ib volatile
solid/ft3/day (4.45
The data from this build-up period, 12-9-68 to 12-16-68, are
presented in Figure 1. When computer based curve fitting techniques were
applied to these data by Dr. James J. Downs, a linear equation,
GAS PRODUCTION = -0.032 + 0.072 FEED RATE
was obtained that accounted for 52% of the variance in the data. The in-
clusion of the temperature data gave the equation:
GAS PRODUCTION = -23.3 + 0.023 FEED RATE +4.1 In T°K
that accounted for 78% of the variance. The negative constants are dis-
turbing until it is realized that these data represent only a small portion
of the full range of possible relationships that would have the shape of
a typical growth curve. This analysis of these data leads to the inference
that digestion rate per unit waste load would increase with loading rate
at some levels of loading.
-------
37
I 35
34
0.33
>
» 6 0.31
X
u
a£
«/> f\
go
°2 0.29
»i
85 Ik
5*
i °-27
o
n
-------
EXPERIMENT II
EFFECT OF HIGH-SHEAR HOMOGENIZATION ON
THE DIGESTION OF PURINA. DOG CHOW
The digesters were operated during the seven-day period, 10-16-68
to 10-23-68, with a constant dog food loading. The 15 digesters were di-
vided into five groups of three:*
Group 1 - gas-stirred only;
Group 2 - Gas plus liquid stirring with spring-loaded ball
valve operating at 50 psig;
Group 3 - Gas plus liquid stirring plus homogenization with
relief valve set at 100 psig;
Group 4 - Gas plus liquid stirring plus homogenization with
relief valve set at 300 psig; and
Group 5 - Gas plus liquid stirring with relief valve set at
900 psig.
The digesters were filled to 3,785 ml and were maintained on a
daily feed of 15 g (3.96 g/f/day), 0.23 Ib volatile solids/ft3/day. The
detention time was 21.6 days. The pH was determined each day before
feeding and the system was buffered with lime (Ca(OH)2), and sodium
bicarbonate (NaHC03) so that 1 g of lime was added with the feed for each
0.10 pH unit below 7.00 to a maximum of 5 g and 0.1 g of NaHC03 was added
with the feed for each 0.10 pH unit below 6.50. Head gas was recycled at
the rate of 100 ml/min. Liquid recirculation rate was 100 ml/hr.
The results of the experiment are presented in Figure 2. There
is a greater variability in this group of data than in the data from the
later studies. Part of this is the result of averaging three rather than
five data. There appears, however, to be no significant effect in the
digestion process due to the homogenization process. Group 2 had the
greatest rate of gas production of the five groups but the rate is only a
few percent greater than the previous six-day "blank" period. Group 3 is
definitely lower in the early part of the study, but the line joining the
data of the last four days has the same slope as lines 1, 4, and 5.
In later experiments, the digesters were divided into three groups of
five. The coding of the digesters, however, as 1A, IB, 1C, 2A, 2B,
etc., was maintained throughout the entire research and has been kept
for this report.
-------
LOADING RATE » 3.96 g/fl/day PURINA DOG CHOW
35
1 = GAS STIRRING ONLY
2= GAS STIRRING PLUS LOW PRESSURE LIQUID RECIRCULATION
3 = GAS STIRRING PLUS LIQUID HOMOGENIZATION AT 100 LB/IN2
4= GAS STIRRING PLUS LIQUID HOMOGENIZATION AT 300 LB/IN2
5= GAS STIRRING PLUS LIQUID HOMOGENIZATION AT 900 LB/IN2
B = PREVIOUS SIX DAYS AVERAGE OF FIFTEEN DIGESTERS
STIRRED BY HEAD GAS RECIRCULATION ONLY
30
(/>
oc
I 25
Z
O
8
20
t/>
3 15
!S!
U 10
J
7
3 4
DAYS
Figure 2 - Effect of Homogenization on Digestion
10
-------
The experiment was discontinued after the seventh day because of foaming
problems. The problem of foaming plagued the entire research program.
This was the only pronounced effect on the digestion process due to high
shear homogenization when lipid levels were low.
11
-------
EXPERIMENT III
EFFECT OF HOMOGENIZATION ON LIMED DIGESTERS
HEAVILY LOADED WITH COTTONSEED OIL
After Experiment II, the group of 15 digesters was divided into
three groups of five:
Group A - Gas-stirred only;
Group B - Gas- and liquid-stirred with spring-loaded ball valve
operating at 50 psig; and
Group C - Gas- and liquid-stirred plus homogenization with relief
valve set at 300 psig.
Prior to Experiment III, the digesters were all functioning
satisfactorily except for plugging of the diaphragm pumps when the feed
level was raised to 20 g/gal./day. The digester contents had been pooled
and redistributed. This slight exposure to the atmosphere had caused
problems in the past, so the digesters were brought up to a feed level of
3.96 g/i/day dog food over a period of a few days. The digesters were all
performing satisfactorily and uniformly before the addition of cottonseed
oil, as can be seen from the data presented in Table III.
TABLE III
PERFORMANCE OF DIGESTERS PRIOR TO ADDING COTTONSEED OIL
Date
26 December
27 December
28 December
29 December
30 December
31 December
1 January
2 January
Average Cumulative Gas
Production in Liters
Group A
3.84
4.32
6.42
8.52
11.75
15.70
21.00
25.60
Group B
4.21
5.36
7.73
9.63
12.02
14.81
18.75
22.30
Group C
4.18
4.74
7.24
9.29
12.39
15.38
18.95
24.32
Remarks
3.96 g/£/day Purina Dog Chow
Digester contents pooled
Raw sludge only added
1.32 g/^/day Purina Dog Chow
2.64 g/^/day Purina Dog Chow
3.96 g/^/day Purina Dog Chow
3.96'g/jg/day Purina Dog Chow
3.96 g/4/day Purina Dog Chow
In Experiment III, the digester volume was 3,785 ml; the deten-
tion time, 21.6 days. Head gas was recycled at the rate of 100 tnl/min;
liquid recirculation was at the rate of 100 ml/hr. Lime was added with the
feed as a buffer to each individual digester at the rate of 1 g for each
12
-------
0.10 pH unit below 7 until 1-11-69, after which 0.1 g of NaHCOa was added
with the feed for each 0.10 pH unit below 7. The loading protocol (see
Table IV) was such that the dog food was gradually replaced by cottonseed
oil based on the approximate COD equivalence of 1 ml oil = 3 g dog food.
TABLE IV
PROTOCOL FOR DIGESTER LOADING IN EXPERIMENT III
Date
3 January
4 January
5 January
6 January
7 January
8 January
9 January
10 January
11 January
12 January
13 January
14 January
15 January
16 January
17 January
Dog Food
(g COD/A/day)
3.73
0.00
2
2.
3
3
3
49
49
73
73
73
2.99
2.24
2.24
.24
.49
.49
.49
2.
1.
1.
1.
Cottonseed Oil
(g COD/l/day)
0.70
0.70
0.70
0.70
0.70
0.70
.40
.10
.80
.80
.80
.50
.50
Total
(g COD/A/day)
1.49
1.
2,
2,
2,
2.
3.
3.
3.50
3.50
4.43
0.70
,19
,19
4.43
4.43
6.13
.09
.04
.04
,04
,99
.99
.99
3.
3.
5.
5.
5.
5,
4.
4.
4.
4.99
The data on gas production are presented in Figure 3. These data,
as well as analytical data presented in the Appendix, show'that the three
groups of digesters can be distinguished on every basis of comparison.
Furthermore, Group A, which was the standard to which the other groups
were to be compared, showed a unique behavior.
The following significant phenomena were observed during the
15-day period:
1. Group A produced more gas than Group B, which produced more
gas than Group C.
2. The methane content of gas produced by Group A was higher
than that produced by Group C as seen in Table V:
13
-------
100
90
to
Q
O
70
60
50
a.
uo 40
O
30
20
10
0
A = GAS STIRRED ONLY
B = GAS + LIQUID STIRRED (50 PSIG) C
C= GAS + LIQUID STIRRED + HOMOGENIZATION (300 PSIG) /
I I I I
I I I I
I
6 78 9 10 11 12 13 14 15 16 17
JANUARY
Figure 3 - Performance of Digesters Fed Dog Food and Cottonseed Oil Buffered with Ca(OH)2
-------
TABLE V
PERCENT METHANE IN DIGESTER GAS
1-10-69 1-17-69
Group A 57.4 63.4
Group B 41.8 50.8
Group C 39.8 40.2
3. The data from the neutral lipid extraction, when examined
for a single digester over a period of time, appear to reflect an adaptation
of the bacterial population to the added cottonseed oil. There also seems
to be a cyclic phenomenon in the individual lipid levels. When examined
in relation to time, the average lipid levels in the various groups that
are presented in Table VI imply the development of a bacterial flora more
capable of hydrolyzing the neutral oil. In the case of Group C, however,
the initial step of the breakdown of the oil to gas is too rapid, leading
to a build-up of acid intermediates. Methane digestion is inhibited by
the pH drop, and the slight increase in lipid levels of Group C that occurs
between 1-13-69 and 1-16-69 may also be the result of pH drop affecting the
hydrolyzing microorganisms.
TABLE VI
AVERAGE LIPID LEVEL IN EXPERIMENT III (g/I)
Group A Group B Group C Total Added*
1.21 0.81 0.76 2.91
0.74 0.52 0.62 7.40
0.59 0.40 0.96 11.36
Total added cottonseed oil from 1-3-69 to given date in grains per liter.
15
-------
The soluble COD and MLVSS data also confirm the premise of dif-
ferential adaptation. The MLVSS, which would measure both undigested dog
food and bacterial mass, decreased as the dog food load decreased, but
also decreased with the amount of shear. Soluble COD in Group A decreased
from 1-7-69 to 1-14-69 as the dog food load was decreased but Group C
maintained the same level. The decrease of MLVSS in the Group C digesters
could be the consequence of the formation of calcium soaps, these coating
the dog food particles and making them coalesce at the bottom of the
digester.
The malfunctioning of the digesters in Groups B and C can be
discerned as early as 1-8-69 when the pH values are examined. All pH's
were above 6.80, but Group A had values all above 7.00. Since lime was
administered to each individual digester at the rate of 1.0 g of Ca(OH>2
for each pH unit below 7.0, the Group C digesters were most heavily limed.
The lime maintained the pH in the normal operating range until 1-14-69
when all five digesters in this group had samples with pH readings 6.75
and below.
16
-------
EXPERIMENT IV
EFFECT OF HOMOGENIZATION ON SODA.-ASH BUFFERED DIGESTERS WHEN
DOG FOOD LOADING IS AUGMENTED WITH COTTONSEED OIL
Prior to beginning this experiment, the liquid volume in all the
digesters was reduced to 2,160 ml to allow the foam room for expansion.
The detention time remained at 21.6 days. Head gas was recycled at the
rate of 100 ml/min and liquid was recycled at the rate of 50 ml/hr. The
contents of the Group A digesters were pooled and diluted with fresh sludge
and this mixture was distributed evenly among all 15 digesters.
The digesters were then brought incrementally up to a daily dog
food feed level of 5.55 g/jfc (0.32 lb/ft3). Soda ash, Na2C03, was used to
buffer at the rate of 0.2 g added with the feed to each digester for each
0.10 pH and below 7.00.
During the break-in period from 2-6-69 to 2-12-69, the 15 digesters
produced 29.63 a of gas with standard deviation = 2.42 or 11.57. of the mean.
The dog food level was then maintained at 5.55 g/£/day, while an additional
1 ml of cottonseed oil was added from 2-14-69 to 2-17-69 and 2 ml was added
from 2-18-69 to 2-20-69. At the end of the experiment, the three groups
of digesters could not be differentiated on the basis of cumulative gas
production. These cumulative gas values, obtained with all pumps operating,
were 23.71 I, 23.98 it and 24.68 I, respectively, for Groups A, B, and C.
These data do not differentiate the groups. When daily gas production is
plotted, as in Figure 4, some slight differences in the three groups may
be discerned.
Table VII shows the results of lipid analyses during the period
of Experiment IV. As in the case of Experiment III, where lime was used
to buffer, the lipid level appears to be inversely related to the amount of
shear.
TABLE VII
AVERAGE LIPID LEVEL IN EXPERIMENT VI
Date
2-10-69
2-13-69
2-17-69
2-20-69
2-24-69
Group A
1.12
0.89
2.15
2.65
1.46
Group B
0.81
0.75
1.60
1.83
1.22
Group C
0.69
0.56
1.22
1.84
0.705
Total Added*
0.00
0.00
1.69
4.22
4.22
* Total added cottonseed oil from 2-10-69 to given date.
17
-------
5.00
4.00
. 3.00
z
o
O
2.00
1.00
O • GAS STIRRING ONLY
A • GAS + LIQUID STIRRING (50 PSIG)
D • GAS + LIQUID STIRRING +
HOMOGENIZATION (300 PSIG)
J0.844g/f/doy
[0.422 g/0/doy cottonseed oil feed
5.55 g/J/doy Purina Dog Chow
T3 U1516 17 18 1?"
20
FEBRUARY
Figure 4 - Digester Performance with Soda-Ash Buffering
18
-------
However, none of the three groups of digesters have adapted well
to the slugs of cottonseed oil. The premise that the functioning of the
hydrolyzing bacteria is affected by pH as well as the methanogenic bacteria
is reinforced here. The Group A digesters, which initially appear to re-
spond to the added oil, were the most rapid to fail. This can be seen in
the volume of gas produced as shown in Figure 4 and in the methane content
of the gas as well. On 2-14-69, all three groups of digesters were produc-
ing 53% methane while by 2-21-69 the average methane content of the gas
produced by the three groups was 47, 51, and 527,, for Groups A, B, and C,
respectively. The results of this experiment indicate that homogenization
had a slight positive action in overcoming the malfunction caused by heavy
oil loading accompanied by soda-ash buffering.
19
-------
EXPERIMENT V
EFFECT OF HOMOGENIZATION ON SODA-ASH BUFFERED DIGESTERS AS
COTTONSEED OIL GRADUALLY REPLACES
PURINA DOG CHOW AS A FEED
This experiment was similar to Experiment IV except that the
level of dog food loading was decreased while the level of cottonseed oil
was increased. The volume of the liquid in the digesters was 2,160 ml;
the detention time, 21.6 days. Soda ash was used to buffer the individual
digesters at the rate of 0.2 g added with the feed for each 0.10 pH unit
below 7.00. Protocol for loading as well as gas production and pH data are
presented in Table VIII.
The major conclusion of this experiment is that heavy loads of
cottonseed oil are deleterious to the performance of soda-ash buffered
digesters. All three groups of digesters showed malfunctions as the per-
centage of the COD load that was due to cottonseed oil was increased. The
three groups of digesters showed some differences in performance and, if
rated by the usual criteria, the digesters subjected to gas and liquid
stirring (Group B) gave a slightly better performance than the other two
groups. The slight difference (4%) in the cumulative gas production be-
tween Groups B and C is probably not significant. If the difference in
performance of the liquid stirred digesters was due to shear, it appears
that the spring-loaded ball valve, operating at 50 psig, could provide
adequate shear, if, indeed, the observed differences were due to shear.
The pH data from this experiment are presented in Figure 5.
Groups B and C appear to sour a day later than Group A. The MLVSS data
also appear to differentiate Group A from the other two groups as can be
seen in Table IX. The soluble COD values for 4-1-69 that were 24.2, 17.0,
and 15.8 g/4 for Groups A, B, and C, respectively, also appear to dif-
ferentiate.
TABLE IX
MIXED LIQUOR VOLATILE SUSPENDED SOLIDS
IN EXPERIMENT V (g/4)
Date Group A Group B Group C
3-18-69 22.1 23.1 17.8
3-25-69 14.2 16.2 12.5
4-1-69 16.0 16.1 13.5
20
-------
N)
TABLE VIII
EFFECT OF HOMOGENIZATION ON THE DIGESTION OF COTTONSEED OIL
WHEN FEED RATE IS KEPT CONSTANT IN COD
Feed Rate in Grams
(COD/ i/day)
Date Purina Dog Chow
3-15-69
3-16-69
3-17-69
3-18-69
3-19-69
3-20-69
3-21-69
3-22-69
3-23-69
3-24-69
3-25-69
3-26-69
3-27-69
3-28-69
3-29-69
3-30-69
3-31-69
4-1-69
4-2-69
4-3-69
5.23
5.23
5.23
5.23
5.23
4.58
4.58
4.58
3.94
3.94
3.94
3.29
3.29
3.29
2.62
2.62
2.62
3.29
3.29
3.29
Cumulative Gas Production
in Liters
Cottonseed Oil Group A
0
0
0
0
0
0.61
0.61
0.61
1.22
1.22
1.22
1.83
1.83
1.83
2.44
2.44
2.44
3.05
3.05
3.05
3.73
8.14
11.45
15.93
17.93
21.06
23.85
27.23
31.17
33.14
35.80
37.97
39.92
42.01
43.24
45.08
46.11
47.48
48.31
49.16
Group B
3.80
8.16
11.95
16.14
19.75
23.11
26.30
29.78
33.87
36.23
39.40
42.02
44.50
46.91
48.23
49.90
51.62
53.49
54.41
55.90
Group C
4.19
8.49
12.08
15.33
19.25
22.58
25.67
29.09
32.98
34.83
37.86
40.39
42.52
44.89
45.86
47.74
49.30
51.12
52.24
54.07
Average
pH in 15
Digesters
6.95
6.97
7.06
6.93
6.93
6.91
6.87
6.97
6.96
6.92
6.84
6.85
6.80
6.80
6.66
6.57
6.51
6.47
6.46
6.36
pH Range in
14 Digesters*
6.80-7.30
6.85-7.30
6.95-7.25
6.85-7.20
6.85-7.25
6.70-7.15
6.70-7.10
6.85-7.05
6.90-7.10
6.80-7.15
6.65-7.10
6.70-7.05
6.60-7.00
6.60-7.05
6.50-6.80
6.40-6.90
6.25-7.05
6.30-6.90
6.30-6.90
6.30-6.75
Digester 1A was malfunctioning during most of this study, and its pH has not been included in the
range.
-------
GROUP A - GAS STIRRING ONLY
4
7
6 8|
7
8 6
7
6 8
7
8 «
7
I I I I I I I
I I I
J !
j I i
j i
IA
1C
2A
GROUP B - GAS + LIQUID STIRRING (50 PSIG)
6
7
6 e
7
8 6
7
6 8
7
8 6
J _ I _ I.. I _ I _ I . I _ I _ I _ I _ I _ I _ I - 1 - 1 _ I _ I _ I _ L
_ _ . _ _
GROUP C - GAS + LIQUID STIRRING + HOMOGENIZATION (300 PSIG)
K
3A
3.
3C
iiriiiTiii ITIIII i i r
I I I I I I I 1 1 I 1 I I 1 I I 1 I I
5A
5C
15 14 17 18 19 20 21 22 23 24 25 26 27 » » 30 31 I 2 3 4
MARCH
APRIL
Figure 5 - Variation of pH in Individual Digesters During Experiment V
22
-------
EXPERIMENT VI
THE EFFECT OF MOTOR OIL ON THE DIGESTION OF
PURINA DOG CHCW AND COTTONSEED OIL
The evaluation of the effect of a non-biodegradable lipid on the
digestion process was carried out with the digesters filled with 2,160 ml
of mixed liquor. The detention time was 21.6 days. There were some dif-
ficulties with the break-in period and the Group C digesters were not yet
in full balance when the experiment began. The loading protocol is shown on
Figure 6 along with the performance data. It is apparent that motor oil
had little effect on the digestion of the dog food and cottonseed oil, as
neither the gas production rate, the methane content of the gas, nor the
pH was affected by its addition.
Since the poorer performance of Group C digesters could have been
due to the action of the homogenizers, the liquid pumps were turned off
for three days to evaluate this idea. Gas-volume data from each of the
three groups of five digesters are averaged and presented below as ratios
of the averages with Group A averages the common denominators for all
three groups of ratios.
Condition
1
2
3
4
Group A
1.00
1.00
1.00
1.00
Group B
1.06
0.99
1.00
0.97
Group C
0.68
0.81
0.89
0.70
where Group A = gas recirculation only.
B = gas -I- liquid recirculation (50 psig) .
C = gas + liquid recirculation + homogenization (300 psig).
Condition 1 existed on the fifth day (5-4-69) after operation
at a daily feed rate of 9 g of dog food and 1 ml of cottonseed oil with
both liquid and gas pumps operating.
Condition 2 existed on the fourth day (5-9-69) after 1/2 ml of
motor oil was fed daily in addition to the dog food and cottonseed oil.
All pumps were operating.
23
-------
F
E
E
D
GAS STIRRING ONLY
I 0.462 g/ fl cottonseed oi I
1 0.231 g/1 motor oil
9/0 Purina Dog Chow
/ /
/ / / Y 7 7~
/ /
/ 7 7 / / /
5 '
GROUP A = GAS STIRRING ONLY
§3
8 5
o
GROUP B = GAS + LIQUID STIRRING (50 PSIG)
I
3
2
1
0
I
I
I
GROUP C = GAS + LIQUID STIRRING + HOMOGENIZATION
I I I I I I I I I I I
(300 PSIG)
29 30
APRIL
1 2
MAY
3456 7 8 9 10 11 12 13
Figure 6 - Effect of Motor Oil on Lipid Digestion
14
15
16
-------
Condition 3 existed at the end of three days (5-12-69) with only
the gas pumps operating. The feed rate was the same as in condition 2.
Condition 4 existed after the liquid recirculation pumps had
been turned back on and had been operating for four more days (5-16-69).
The feed rate was the same as in condition 2.
The results of the above analysis seem to again implicate high
shear homogenization as a negative factor in lipid digestion.
25
-------
DISCUSSION
The Effect of Shear on the Digestion of Lipids
The results of the experimental program described in this report
present a strong caution against the use of high-shear homogenization as a
remedy for scum layer formation in full-scale anaerobic digestion. Ex-
periment III produced data to show that lipid digestion in lime-buffered
digesters can be slowed by a low-shear homogenization, while lipid digestion
is upset by high-shear homogenization equivalent to that produced by a
Waring Blendor. Experiments IV and V present data that show a slight
process improvement due to high-shear homogenization when soda ash is used
as a buffer in a system heavily loaded with lipid. This reversal, however,
occurs when the digesters were malfunctioning for another reason.
The anaerobic digestion process like all biological processes is
described by the term "steady-state dynamics." The hydraulic analogy is a
watershed in which flow into a portion of the system is matched by flow
out. If flows are not matched, the result is either flooding or draining
of an impoundment. A simplified description of the digestion process is
SLUDGE
(Insoluble Organic Matter)
R! Extracellular Enzymes
(Soluble Organic Matter)
R2 Acid Bacteria
Volatile Acids
R.
•3
Methanogenic Bacteria
CH4 + C02
In a well-operating digester, the rates of reaction in units of
material altered per unit time must be equal, R^ * R£ • R^. The ability
of the organisms to multiply and adapt to increased food levels, however,
varies greatly, with —ij4 —2. jt —2. . All of the technical data available
dt dt dt
26
-------
indicate that -^. > —2. . It appears that in the case of grease, —-- >
dt dt dt
dR2 > dR3 when adequate surface is created between the grease and the
dt dt
digester liquor. The soluble organic matter, in this case fatty acid salts
of calcium or sodium or soaps, can have two modes of actions. It can ac-
celerate R2 so that the volatile acids build up, lower the pH, and inhibit
methane fermentation. The result is a "stuck" digester. The other action
of the soaps is to directly inhibit methane fermentation by affecting the
bacterial surface. This, too, would lead to a "stuck" digester.
Soda ash, which supplanted lime as a buffer for the purpose of
reducing foaming, proved harmful because it, also, accelerated hydrolysis
too much. The data are consistent with, but do not prove, the following
hypothesis: Sodium soaps (sodium salts of long chain fatty acids) would
inhibit methane fermentation. Sodium soaps are much more soluble than
calcium soaps, but they are still very sparingly soluble. When sodium
soaps reach a certain level of concentration in water, the soaps form
emulsions. Most of the soap is present as the oil phase of an emulsion,
a portion acting as its own emulsifying agent. High-shear now functions
to provide interface between the soap micelle and bacteria that degrade
the soap to simpler metabolites that can be fermented to methane. McCartylr.'
has suggested the use of calcium chloride to reverse sodium oleate poisoning
in the same series of articles that advocate the use of sodium bicarbonate.
The results of the study, nonetheless, should not be interpreted
as condemning the incorporation of surface mixers in the design of digestion
tanks. There is a report—' of the installation of four 10 hp Lightning
mixers on a 100,000 ft3 tank in the town of Tonawanda, New York, with the
very rewarding result of eliminating scum formation. Since most of the
scum from the settling tanks is burned, this tank is not subjected to a
heavy grease load. The results of this research led to the prediction that
the operator of the Tonawanda Plant would be in serious trouble should the
situation arise when the grease burners are inoperative at the same time a
heavy slug of grease is fed to the digesters as a shock load. While me-
chanical mixing in addition to that caused by rising gas bubbles is es-
sential for high-rate digester performance, the hydrolysis of heavy grease
loads could proceed too rapidly as a result of creating grease-digester
liquor interface. In such a situation there would be a build-up of soaps,
either calcium soaps from lime, or sodium soaps from soda ash, that would
affect the methane bacteria and cause digester upset. Foam production,
which is especially severe when lime is used as a buffering agent, may also
negate any process advantage due to high-shear mixing.
27
-------
Ability of Digesters to Degrade Heavy Loads of Grease
It is reported by BasuM/ that heavy grease loads cannot be di-
gested. In Experiment III, during the last three days of fermentation,
the Group A digesters were loaded with a feed stock that was at least 43%
fat by weight and 70% fat on the basis of COD. This feed produced the most
rapid gasification and the highest methane content gas of any sample of
feed stock used in the study. High shear that was equivalent to that of
a Waring Blendor eventually caused a matched group of digesters, Group C,
to go sour. Even the lower level emulsification of the Group B digesters
caused malfunction. The level of gas stirring in Group A, however, was
such that about two volumes of gas equal to the liquid volume in the di-
gester were pumped to the bottom of the digester each hour. This type of
stirring cannot be quantitated, but the digester contents were undergoing
a lively bubbling due to gas recirculation.
Optimum Operating Conditions for Sludge Digesters
Loading rate: Manual of Practice 16—' differentiates "high
rate" as opposed to "standard rate" digestion on the basis of loading,
0.15 to 0.40 Ib volatile solids/ft3/day compared to 0.04 to 0.10 Ib. vola-
tile solids/ft^/day. The article goes on, however, to consider mixing as
an essential feature of the high rate system. In this study, the digesters
were able to accommodate 0.29 Ib volatile solids/ft-^/day of the highly
putrescible dog food. Since the experimental design was such that most of
the time constant feed levels were maintained, not much kinetic data can
be derived from the experiments to establish a digestion rate. Still,
during the periods of startup, the digesters responded in one day to any
load increase that was not an overload. The overload rate was not reached
by slow buildup at any time, so a maximum load cannot be set for the ex-
perimental system either with a dog food or a combination dog food and
cottonseed oil feed.
Temperature: The extreme temperature sensitivity of the process
can be seen in Figure 1. The of ten-published!^.' diagram of performance vs
temperature of the digester places the optimum temperature of the mesophilic
bacteria at 38°C. Occasional shifts in temperature in the digesters used
in this study showed that activity increased with temperature from 34° to
41°C. In contrast to our observations, Lyman, McDonnell, and KrupiJ./ cite
experience with full-scale digesters where a 5°C variation had no effect on
gas production.
pH; The normal process range recommended for pH is 6.80 to 7.20.
The results of the pH measurements in this study imply that the range may
be much more narrow, as any drop below pH 7.00 seemed to portend process
difficulties.
28
-------
Microbial ecology. Several hundred microbial smears were ex-
amined with the view to characterizing the dominant organisms in the di-
gestion process. No pattern developed that could be discerned by the
simple test except that the tetrad packets of cocci reported by CooksonJJL/
appear to increase during cottonseed oil feeding.
Buffering agent; This research program has also produced results
that show that soda ash and lime cannot be used interchangeably as buffering
agents. Unless overmixed, lime buffered digesters can tolerate heavy loads
of biodegradable oil, while under the same conditions soda-ash buffered di-
gesters go sour. Filbert—' has analyzed digester startup problems. He
quotes both McCarty and McKinney as advocating the use of sodium bicarbonate
as a buffer. Soda ash at a pH below 8.3 would be converted from sodium
carbonate to sodium bicarbonate.
Statistical Basis for Plant Design
This research program is innovative in the use of replicate di-
gesters and elementary statistical techniques for evaluation of the data.
Few other recently published research papers report the use of more than
a single digester to establish the design principles and parameters that
are used for the construction of full-scale plants. The study of the ef-
fects of radioactivity on digestion by Grune, Church, and KaplanZO./ used
statistical analysis to evaluate data from a bank of 18 digesters, but the
data from that study have little value for design. While the overall re-
search program was conducted with impressive precision, many individual
data varied widely from the mean. A design parameter without a measure of
variability is valueless. With this in mind, the difficulties encountered
in practice with anaerobic digestion are easily understood.
29
-------
CONCLUSIONS AND RECOMMENDATIONS
The program described in this report was initiated because sound
biochemical rationale predicted that the process of digesting waste fats
and oils would be accelerated by creating more favorable conditions for
bacterial hydrolysis of these water-insoluble materials. The results of
the study have not precluded the use of surface mixers or other devices to
incorporate scum layers into the main mass of digesting sludge. What has
been shown, however, is that if such a process modification is effected,
the intensity of the turbulence must be adjusted to the nature of the feed-
stock.
All waste treatment processes must be highly adaptable to the
variable quantity and composition of the feedstock. The anaerobic di-
gestion process is particularly susceptible to upset when any of the process
variables are altered. There is no question that the process is far from
a state of maximum optimization. There is, however, the question of the
possibility of further optimization while retaining the necessary degree
of adaptability to feedstock. Real time feedback control techniques that
have been adaptable to other chemical processes cannot be used for diges-
tion. Few such techniques are available for characterization of the feed-
stock and we do not yet fully understand the kinetics of methane fermenta-
tion.
Current ideas in sanitary engineering design are that the dis-
advantages of digestion outweigh the advantages. (See Table I.) Still
the majority of sludge in this country is probably stabilized by anaerobic
digestion and operators of these digesters are continually faced with the
problems of grease overload and scum layer formation.
A continuation of the research program described in this report
would surely yield better design parameters for digester mixing. Such a
program should include laboratory studies with replicate 1-gal. digesters.
Since gas stirring of the type employed in the present study creates a
fair degree of turbulence, the baseline process should be digesters that
are stirred only by the effervesce of the gas or briefly stirred before
sampling. With the quiescent digesters, the effect of increased turbulence
on the performance of the digesters in terms of conversion of volatile solids
to gas, and stability to shock loads of grease would be evaluated. Natural
municipal sludge would be used as a feedstock with authentic sewage grease
used as the lipid for shock loading.
Simultaneously with a laboratory program, operators in a selected
region, possibly the Missouri River Basin District (because of its many
animal-processing facilities), should be asked about their experiences with
malfunction due to grease overload.
30
-------
Information from these two programs will provide design criteria
for mixers that will help cure "scum" layers, and can be operated in a
manner that will not cause upset by a too rapid saponification of a shock
load of triglyceride.
31
-------
APPENDIX A
APPARATUS AND METHODS
32
-------
APPARATUS
A schematic drawing of a single digestion unit is shown in
Figure 7, while Figure 8 shows the actual components. A later modification
was the addition of a 1-1 foam trap between the digester and the gas pump.
Fifteen such digesters were assembled into a unit that included a circulating
water bath. The bath was maintained at 36°C ± 1° during most of the ex-
perimental period. The complete assembly is shown in Figure 9.
The most important single aspect of the project was the homogeni-
zation units because the entire program has been based on the premise that
scum-layer digestion might be accelerated as the material is homogenized
into the main body of the aqueous fermenting mass. Homogenization was to
be effected by pumping the lipid and aqueous phases together through an
orifice under pressure.
A Hoke No. 6528L4B pressure relief valve when fed with a Milton
Roy mRoy-110-A diaphragm pump has been shown to incorporate a considerable
amount of vegetable oil into water and produces a satisfactory emulsion
when the valve is set at less than 500 Ib. This finding overcame the
greatest potential engineering bottleneck in the project.
A Hoke pressure relief valve was constructed of brass and Teflon.
Although the brass was a possible source of copper contamination which
would have inhibited bacterial growth, this deleterious effect was elim-
inated by maintaining an adequate level of sulfide ion in the digester
liquor as proposed by Lawrence and McCarty.££/ Another potential problem
was a continual change in orifice geometry, due to normal wear or abrasion
by grit. A four-day trial with an oil and water system showed that no
unusual wear damage to the orifice occurred under these conditions. When
the system was used on a sludge that had not been settled, some abrasion
of the valve orifice was observed. It was necessary to replace these
relief valves several times during the study.
It was not possible to measure the homogenizing pressure with a
gauge while the system was in operation. The hydraulic "hammer" from the
diaphragm pump would have rapidly destroyed a gauge. An auxiliary hy-
draulic system with a pressure ballast tank and a snubber on the test gauge
was used to set and periodically check pump and relief valve pressures.
Gas recirculation was effected with a small Presso-Vac pump in
which the inlet stream was reduced by means of an orifice which consists
of glass capillary tubing in which was inserted wire. Frequent cleaning
and replacement of the diaphragm and Viton valves were necessary for con-
tinuous operation.
33
-------
OJ
-p-
* No. 4 stopper
** Na 13 stopper
* Tefbn Gasket
Fill
Portx
250 ml
Leveling
Hydraulic
Hoke Pressure
Relief
Bottle
Diaphragm
Pump mRoy
R-110-A
„.
fester
Liquor
105?
NaCI
atpH3
I—J
J Neptune
Gas Pump
5 gal
Carboys
Calibration
Scale
Figure 7 - Schematic Diagram of Digestion Apparatus
-------
.
u
Figure 8 - Photograph of Digestion Apparatus
-------
Figure 9 - Fifteen Digester Unit Assembly
-------
Gas was collected by displacement of an acidified 1070 NaCl solution from
one polyethylene carboy into another carboy which was under a slight pres-
sure. The larger the volume of gas produced, the larger was the pressure
on that gas due to hydraulic head. Appropriate conversion factors for
pressure and temperatures were used in calculating the amount of gas that
was produced.
Gas samples were collected from the system by inserting a glass,
gas-collection bulb in the vent line.
Liquid samples were obtained from the pressurized system by
opening the sample port. Stainless-steel tubing was used to connect the
liquid and gas systems to the glass bottle which served as the digester.
Nylon Swagelok fittings were employed where nec-essary. The liquid exhaust
tube was 1/4 in. O.D. which passed through a 3/8 in. O.D. tube placed in
the stopper. A Nylon Swagelok 3/8 in. to 1/4 in. union has been bored
through, and the 1/4 in. end ferrules were replaced with Teflon ferrules.
In this manner a gas-tight seal was maintained while the exhaust tube
could be placed to skim the surface or sample the main body of liquid.
Material was introduced into the digester through the liquid
and gas return line. Thus, there was no holdup.
37
-------
ANALYTICAL METHODS
CONVERSION OF DIGESTER GAS VOLUME DATA TO
STANDARD CONDITIONS
Several variables were considered in the conversion of the daily
gas readings to standard conditions. These are:
1. The time span over which the volume change was observed,
2. Reservoir temperature,
3. Barometric pressure,
4. Head of liquid in the reservoir,
5. Variation in diameter among the reservoirs, and
6. Vapor pressure of the brine solution.
Gas was collected in a polyethylene bottle that was filled with
a 10% brine solution acidified to pH 3. Liquid in this reservoir was dis-
placed by the gas produced during digestion into another bottle that was
placed superior to the first, as shown in Figure 10. Thus, the gas in
the reservoir was under a slight pressure at all times. The datum that
was recorded by the operator was the volume of liquid remaining in the
lower reservoir. Since digester gas could be vented or C02 added during
the liquid sampling and feeding operations, this datum was recorded both
at the beginning of the day, and after all the daily operations were com-
pleted. An adjustment for this variable time period had to be made. The
equation for calculating this adjustment is:
va = ^ (v1 - v2) (i)
where Va = gas produced during elapsed 24 hr
V| = previous afternoon reading of liquid content of the lower
reservoir in liters
V£ = morning reading
t = elapsed time in hours between Vi and V?.
38
-------
VENT
Figure 10 - Diagram of Location of Gas Reservoirs
39
-------
The equation for calculating the adjustment for reservoir tem-
perature is:
Vb = Va x 273'2 , (2)
273.2 + T
where T = temperature in °C, and Vfe is temperature and time corrected
gas volume.
If a table of adjustment factors is calculated, Eq. (2) can
be simplified to
Vb = Va x F (3)
A set of factors is shown in Table X.
TABLE X
FACTORS FOR TEMPERATURE ADJUSTMENT OF GAS VOLUMES
Temperature
l£ H. Factor
20 68 0.932
21 70 0.929
22 72 0.926
23 73 0.922
24 75 0.919
25 77 0.916
26 79 0.913
27 81 0.910
28 82 0.907
29 84 0.904
30 86 0.901
The adjustment for pressure must recognize the barometric pres-
sure, the liquid pressure head, and the vapor pressure of the brine solu-
tion in the reservoirs. The equation for this adjustment is:
(4)
40
-------
where Vc = gas volume at STP,
B = barometric reading,
H = liquid head,
W = vapor pressure,
each as millimeters of mercury.
In Figure 10,
h = height of liquid exerting pressure,
d = 635 mm, the distance one reservoir was mounted above the
other,
and
R1 and R" = height of liquid in the respective reservoirs.
These are related as follows:
h = d + R" - R1 (5)
but R" + R1 = C (6)
so h = d + c - 2R1 (7)
"C" was actually the height of the liquid when it was all in the lower
carboy and this amounted to 360 mm. Since the lower carboys were calibrated
in liters, the datum that was recorded daily by the operator is liters of
liquid. R1 is related to VL (the volume of the lower carboy) by the
empirical relationship
R1 = 18.4 VL (8)
where R' was in units of millimeters height per liter contained.
The variation in diameter among the 30 carboys that were used as
reservoirs did not exceed a maximum of 2% of the average. Since the error
in reading the volume of liquid in the carboy exceeded the variation in
diameter, the latter variation, while recognized, was not significant.
41
-------
Since we were interested in the pressure change over the time,
t , an average value for R' was obtained by use of the relation,
2R1 = 18.4 (V1 + V2) (9)
where V^ and ¥2 are liquid volumes in the lower carboy at times t ,
and t2 , respectively.
To convert the height, H , into pressure head in millimeters of
mercury, H , we multipled by the specific gravity of the brine which is
1.07, and divided by the specific gravity of mercury which is 13.8. The
correction for change in density due to temperature was too small to con-
sider.
H = 1^2211 (10)
13.8
By combining the various expressions, we obtain
H = i^|- [635 + 360 - 18.4 (VT + V2)] (11)
The third pressure correction, the vapor pressure of the brine,
is 18 mm Hg at 25°C. The variation of the vapor pressure with temperature
in the operating range is too small to be considered. Thus, W * 18 in
Eq. (4).
A general expression for standardizing the gas volume can now
be obtained by combining Eqs . (1), (3), (4), and (11).
360 . lg<4 (v + v
13.8 1 - 76Q
which reduces to
(Vj - V2) 0.0316F r -i
Vc = —i ^ [59.15 + B - 1.43 (Vx + V2)J (13)
where Vc is liters of gas at STP that would be produced in a 24-hr period.
The actual calculation was performed by an IBM 360 computer.
42
-------
CHEMICAL OXYGEN DEMAND (COD) DETERMINATIONS
COD determinations were performed as follows:
There was added in sequence to a sample flask, a 1-ml sample of
MLSS or MLSS filtrate, 20 ml distilled H20, 25 ml of 0.250 N K2Cr207, 30
ml of H2S04, doped with 22 g silver sulfate, AgS04, per 9-lb bottle of
acid, and 0.4 g HgS04. The mixture was refluxed for 2 hr and cooled to room
temperature. (Cooling is essential if the indicator is not to be oxidized.)
Two drops of ferroin indicator were added and the solution titrated to a
red-brown end-point with 0.1 N Fe(NH^)2(80^)2.
Seven replicate COD determinations on a MLSS filtrate gave re-
sults with a standard deviation of 470.
A small portion of the mixed liquor appeared to resist complete
digestion. A trace of white oil appears during the digestion which re-
fluxes as a steam distillate. When the mixture is cooled, it solidifies
and floats on the top of the aqueous layer.
During the experimental period, COD was determined in both the
mixed liquor and the mixed liquor filtrate. Data from the latter deter-
mination are presented in Appendix B as soluble COD.
SOLID DETERMINATIONS
The mixed liquor volatile suspended solid (MLVSS) was determined
by a slight modification of the Standard MethodsjJL/ procedures. Two
changes were made. The usual asbestos pad in the Gooch crucibles was pre-
coated with fine sand (Fisher S-151) and 1 ml of a 2% solution of Rohm and
Hass1 C-7 cationic polyelectrolyte was added to the 50-ml mixed liquor
sample prior to filtration.
Number 3 Gooch crucibles fitted with "F" covers were used. MLSS
was determined by heating at 110°C and "Ash" was determined by firing at
600°C in a Blue M-M-25A-1A muffle furnace. In each case, the crucibles
were allowed to cool in a desiccator loaded with Drierite. Heating was
carried out for at least 16 hr and it was demonstrated that "constant
weight" was achieved with both MLSS and ash in this period. Since the
amount of inorganic buffer added along with the feed was highly variable,
the MLSS and ash are not significant variables. The data from the MLSS-
ash are the MLVSS which are presented in Appendix B.
43
-------
The added polyelectrolytc introduces an error. Since more than
1.0 mg of polyelectrolyte was needed to produce a satisfactory precipitate
(20 mg was used), it was assumed that most of this added material remained
absorbed on the MLSS. This would cause a maximum error of the order of
+ 1.57, with the majority of samples that were examined. Some portion of
the polyelectrolyte could also affect the COD determination on the filtrate
very slightly.
NEUTRAL LIPID
Neutral lipid was determined by a modification of the method of
Loehr and Rohlich.22/ Fifty milliltters of the MLSS was placed in a 250-ml
separatory funnel equipped with a Teflon stopcock. One hundred milliliters
of acetone was added and the mixture shaken vigorously. Twenty-five milli-
liters of chloroform was then added and the mixture was shaken again. An
additional 75 ml of chloroform was thcr added and the mixture was shaken
gently. Two layers were allowed to separate. The lower layer was removed
and filtered through Whatman No. 1 paper. One hundred milliliters of the
filtrate was evaporated to a small volume in the hood and this small volume
was then transferred to a weighed 50-ml beaker. This smaller beaker was
then placed in the oven for at least 16 hr at 110°C before cooling and
weighing. Replicate lipid extractions showed excellent precision. Re-
covery of portions of cottonseed oil added to a composite sludge was 53.5,
55.5, and 56.57, for three determinations. The weight of lipid recovered
from the extract was multiplied by 36.5 (20/0.55) to give "LIPID" in grams
per liter, which is reported in Appendix B.
METHANE CONTENT OF DIGESTER GAS
The digester gas was analyzed by gas chromatography. A Perkin-
Elmer Model 154B instrument equipped with thermister detectors was used.
The separation column was packed with 80 to 100 mesh Porapak Q which gave
excellent separation of C02, CH^ and ^. The 4 ft x 0.25 in. column was
maintained at 40°C. Helium was the carrier gas. The relative detector
sensitivity as determined with analytical grade reagent gases was 1.37 for
the CH4/C02 ratio of peak heights.
Figure 11 shows the system that was used to introduce the sample
into the instrument. The 250-ml glass gas sampling bulb was equipped with
glass stopcocks that were sealed with silicone grease. The Teflon gas
sampling valve (large black knob) permitted the evacuation of the sample
bulb as well as the stainless-steel sample loop that can be seen behind
44
-------
Figure 11 - Sampling Apparatus for Gas Chromatograph
45
-------
the Teflon valve. A sample loop with a volume of 0.474 ml was used. It
is critical that the sample loop not be too large. When we attempted to
calibrate the column with a 0.9 ml loop, the column was overloaded with
pure C02- With the smaller loop, the calibration curve showed complete
linearity with all mixtures of C02 and
The sample system could be completely evacuated. A trace of air
was introduced into the sample from the stopcock extension when the evac-
uated bulb was attached to the gas reservoir exhaust line. This air did
not interfere. The column did not separate 02 and N2 . This was not
necessary since, prior to an experiment, the gas reservoirs and digesters
were thoroughly flushed with C02 • The gas reservoir samples were saturated
with H20 which did not produce any response with the fractometer under
the prevalent conditions.
Peak height above the baseline was measured on the readout chart
for CH4 and C02 •
CH peak x 100
"
CH4 peak + 1.37 C02 pealc
HYDROGEN ION CONCENTRATION
A Corning Model-7 pH meter equipped with standard size glass
and calomel electrodes was used to determine the hydrogen ion concentra-
tion of the sample. The electrodes were calibrated daily with pH 7 buffer.
A 100- to 175-ml sample that was freshly withdrawn from the digester was
stirred with a Teflon coated magnet that was rotated by a Mag Mix while
pH was being determined. The pH varied continuously with time, but con-
trary to the caution of Standard Methods that the pH would rise due to
loss of C0£, the pH rapidly decreased for a few seconds then decreased
continuously but much more slowly. The pH at the apparent point of sta-
bilization was recorded.
MICROBIOLOGY
Samples of the digester contents were taken for microblal assay
two to three times each week. These were diluted with one or two volumes
of water and spread on glass slides that were fixed by warming. The smears
were then stained with Gram's stain and observed with an oil immersion lens
at a magnification of 980.
46
-------
FEEDSTOCK
The digesters were fed Purina Dog Chow that was ground with a
Waring Blendor until it passed through a 20-mesh sieve. This material
readily wetted and could be transferred with a jet of water. The Purina
Dog Chow gave the following analysis:
Carbon = 48.90%
Hydrogen = 6.25%
Nitrogen = 4.49%
Moisture - 8.52%
Ash = 8.03%
Lipid = 7.87%
COD = 0,943 g/g solid
Maximum
Biodegradable = 84.45%
The cottonseed oil that was used as a typical biodegradable lipid
gave the elemental analysis:
Carbon » 77.71%
Hydrogen - 11.57%
Nitrogen = 00.00%
Oxygen = 10.72% (by difference)
The calculated COD from these results was 2.891 g/g oil. This
is slightly higher than the value of 2.785 that can be calculated for
pure glycerol trioleate. Based on a density of 0.9140 g/ml at 25°, the
oil has a value of 2.64 g COD/ml oil. Thus, in the reported experiments,
0.357 ml of cottonseed oil was the COD equivalent of 1 g of dog food.
The motor oil that was used as a typical non-biodegradable motor
oil was Philube Gear.Oil - all purpose 80 - MIL-L-2105 which is distributed
by Phillips Petroleum. It was not analyzed.
Tapwater originating in the Missouri River was used as liquid
makeup.
Sludge specimens were obtained most frequently from the first
stage digester of the two-stage digestion system of the southern Johnson
County (Kansas) Treatment Plant.
Ca(OH)2> NaHCO^ and Na2COo were used in various combinations as
buffering agents at various time periods during the study. Exact details
are given in the sections on Experimental Protocol.
47
-------
EVALUATION OF HOMOGENIZERS
The combination of diaphragm pump and pressure relief valve was
chosen to simulate the performance of the type of homogenizer that is
widely used to homogenize milk and a multitude of other polyphasic mate-
rials. A typical commercially available homogenizer is manufactured by
the Manton-Gaulin Company. The mRoy pump and Hoke pressure relief valve
were picked because they were adequate for the job, cheap in relation to
available commercial laboratory scale units, and not needed for long-term
service.
The homogenizing system was checked for two requirements. First,
the homogenizer had to produce a reasonably stable emulsion. A permanent
emulsion was not required as the surface of the digester was continually
skimmed and any surface lipid film (scum) would be re-emulsified rapidly.
The liquid pumps were set to circulate about 3 to 5% of the digester
volume per hour. The other requirement was that the shear that produced
the emulsion was not so high that it would rupture a significant number
of bacteria.
The ability of the system to emulsify cottonseed oil and dis-
tilled water was quantified by two rather subjective methods. The emul-
sions produced by the pump-relief valve system were examined by phase
microscopy of a hanging drop. The smaller micelles (globules of oil)
could not be counted because of the Brownian movement. Thus the absence
of large micelles was the criterion of emulsification. The second criterion
was "creaming time," the period required for oil to coalesce and form an
observable layer on the surface of the liquid.
Oil:water mixtures, 1:9 and 1:49, were examined in the system
with the relief pressure settings of 100, 200, 400, and 800 psig. All
pressure settings produced emulsions. This is a relevant fact as the
diaphragm pump is equipped with a spring-loaded ball valve that requires
50 psig pressure to open. Therefore, the digesters which were fitted
with pumps designed for liquid recirculation only, had some small func-
tion as a homogenizer. The pressure setting of 200 to 400 psig produced
an emulsion that was comparable to emulsions produced by a Waring Blendor
and also by treatment of the mixture with 20 khz sonic energy. The 800
psig setting caused heating of the sample without much observable benefit.
A sample of sludge was circulated through the pump-relief valve
system for 10 cycles. The sludge was examined at 970 magnification by
phase microscopy and by stained smears. No rupture of cells nor decrease
in population could be observed.
48
-------
APPENDIX B
ANALYTICAL DATA
49
-------
TABLE XI
RESULTS OF EXTRACTION OF NEUTRAL LIPID
1-16-69
1A 0.473 0.633 0.844 0.440 1.525 0.917 0.967
IB 0.618 0.720 1.361 0.571 1.252 0.774 0.710
1C 1.345 0.644 0.997 0.760 1.212 0.804 0.775
2A 0.655 0.676 1.205 0.830 1.117 0.808 0.589
2B 0.582 0.898 1.227 1.127 0.957 0.760 0.564
2C 0.655 0.473 1.398 1.062 0.688 0.601 0.382
3A 0.473 -- 1.270 1.247 0.495 0.866 0.524
3B -- 0.847 1.514 0.716 0.604 0.389 0.840
3C 0.618 0.916 1.154 0.513 0.950 0.484 0.494
4A 0.473 1.029 1.121 0.422 0.976 0.264 0.884
4B 0.510 1.222 1.121 0.695 0.772 0.204 0.884
4C 0.582 0.804 1.147 0.589 0.491 0.324 1.116
5A 0.364 1.534 1.325 0.640 0.772 0.633 0.909
5B 0.587 0.796 1.412 1.062 0.612 0.316 0.877
5C -- 0.665 0.830 1.069 0.299 0.524 1.040
(grams/liter)
1-26-68
0.473
0.618
1.345
0.655
0.582
0.655
0.473
--
0.618
0.473
0.510
0.582
0.364
0.587
~ ™
12-27-68
0.
0.
0.
0.
0.
0.
0.
0.
1.
1.
0.
1.
0.
0.
2-10-69
1.
1.
1.
1.
0.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
274
110
198
106
890
085
688
990
579
717
968
874
535
568
491
633
720
644
676
898
473
--
847
916
029
222
804
534
796
665
1-3-69 1-6-69
0.
1.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
0.
2-13-69
0.
0.
0.
0.
1.
0.
0.
0.
1.
0.
0.
0.
0.
0.
0.
855
921
688
855
136
895
855
444
147
430
579
612
597
510
622
844
361
997
205
227
398
270
514
154
121
121
147
325
412
830
2-17-69
1.740
1.915
2.319
2.115
2.646
1.445
1.951
2.341
1.410
0.870
1.023
1.227
1.219
1.390
1.241
0.440
0.571
0.760
0.830
1.127
1.062
1.247
0.716
0.513
0.422
0.695
0.589
0.640
1.062
1.069
1-9-69
1.
1.
1.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
525
252
212
117
957
688
495
604
950
976
772
491
772
612
299
2-20-69
2
2
2
2
3
1
1
2
2
1
1
1
1
2
1
.311
.799
.097
.592
.437
.860
.219
.282
.584
.168
.303
.598
.765
.821
.700
1
-13-69
0.917
0.774
0.804
0.808
0.760
0.601
0.866
0.389
0.484
0.264
0.204
0.324
0.633
0.316
0.524
2-24-69
0.
2.
1.
1.
1.
1.
1.
1.
0.
0.
0.
0.
0.
0.
0.
899
148
096
259
875
467
558
736
761
575
466
528
946
892
692
Dieestor 2-10-69 2-13-69 2-17-69 2-20-69 2-24-69 2-27-69
1A 1.274 0.855 1.740 2.311 0.899 0.943
IB 1.110 0.921 1.915 2.799 2.148 1.270
1C 1.198 0.688 2.319 2.097 1.096 0.914
2A 1.106 0.855 2.115 2.592 1.259 0.979
2B 0.890 1.136 2.646 3.437 1.875 1.117
2C 1.085 0.895 1.445 1.860 1.467 1.147
3A 0.688 0.855 1.951 1.219 1.558 1.372
3B 0.990 0.444 2.341 2.282 1.736 1.147
3C 0.579 1.147 1.410 2.584 0.761 0.681
4A 0.717 0.430 0.870 1.168 0.575 0.815
4B 0.968 0.579 1.023 1.303 0.466 0.797
4C 0.874 0.612 1.227 1.598 0.528 0.848
5A 0.535 0.597 1.219 1.765 0.946 0.859
5B 0.568 0.510 1.390 2.821 0.892 0.830
5C 0.491 0.622 1.241 1.700 0.692 1.041
50
-------
Digester
1A
IB
1C
2A
2B
2C
3A
3B
3C
4A
4B
4C
5A
5B
5C
Digester
1A
IB
1C
2A
2B
2C
3A
3B
3C
4A
4B
4C
5A
5B
5C
3-3-69
0.655
0.560
0.542
0.560
0.651
0.778
0.793
0.716
0.331
0.371
0.120
0.360
0.462
0.371
0.469
3-27-69
1.916
1.207
1.604
1.873
1.113
1.545
1.789
1.120
1.124
1.291
1.353
0.855
2.043
1.873
1.887
TABLE XI
3-10-69
0.345
0.455
0.578
0.695
0.836
0.575
0.691
0.484
0.444
0.444
0.527
0.469
0.433
0.495
0.447
3
(Concluded)
3-17-69
0.545
0.724
0.713
0.644
0.112
0.935
0.858
0.705
0.331
0.356
0.585
0.465
0.658
0.422
0.367
-31-69
2.105
1.960
1.993
1.491
1.356
1.429
0.825
0.978
0.393
0.342
0.469
0.484
1.156
0.680
0.735
Extraction of acidified samples.
3-20-69 3-24-69
1.000 1.310
0.898 1.385
1.062 1.156
1.276 0.884
0.796 1.094
1.047 1.178
1.342 0.935
1.149 0.898
0.404 0.225
0.465 0.422
0.607 0.509
0.531 0.458
0.538 0.589
0.538 0.502
1.058 0.716
4-3-69 4-7-69*
3.618 8.425
2.912 10.970
2.865 5.600
1.923 9.970
2.920 11.991
2.876 11.450
1.734 7.923
1.628 5.868
0.284 4.945
0.756 3.607
0.363 6.272
0.458 5.770
1.289 8.850
1.607 7.577
0.636 5.530
51
-------
TABLE XII
MIXED LIQUOR VOLATILE SUSPENDED SOLIDS
Digester
1A 19.5 17.3 12.8 12.8 7.6 -- 13.2
IB -- 21.2 12.9 6.2 6.7 -- 16.0
1C -- 20.4 13.1 4.6 6.0 16.2 14.2
15.2
2B 17.7 18.6 12.9 5.6 -- 16.8 13.8
2C 19.0 5.3 8.4 3,1 4.4 12.3 14.6
3A 15.9 5.8 12.4 5.9 3.6 8.6 11.2
3B 18.7 5.0 7.6 3.3 7.1 14.2 18.9
3C 20.0 5.2 8.4 3.1 1.1 10.4 8.2
4A -- 4.9 6.6 2.2 1.2 -- 16.3
4B 17.4 6.1 8.9 4.1 3.8 12.0 15.
4C 16.2 9.7 8.0 3.0 4.0 13.0 17.3
5A 14.0 5.7 7.3 4.4 2.4 12.7 19.7
5B 14.8 6.7 8.0 4.3 1.8 12.8 13.3
5C 13.2 12.4 7.4 4.2 1.9 13.8 14.1
(grams /liter)
12-3-68 12-10-68
19.5
--
--
16.9
17.7
19.0
15.9
18.7
20.0
--
17.4
16.2
14.0
14.8
13.2
3-4-69
13.9
14.2
10.1
14.6
14.4
17.0
16.1
15.9
12.7
14.5
5.0
14.5
14.0
14.1
12.3
17.3
21.2
20.4
19.8
18.6
5.3
5.8
5.0
5.2
4.9
6.1
9.7
5.7
6.7
12.4
12-30-68
12.8
12.9
13.1
13.0
12.9
8.4
12.4
7.6
8.4
6.6
8.9
8.0
7.3
8.0
7.4
3-11-69
13.5
20.1
18.3
20.1
20.0
17.7
16.9
16.6
16.2
15.8
16.9
13.5
14.8
13.8
14.4
1-7-69 1
12.8
6.2
4.6
4.3
5.6
3»1
5.9
3.3
3.1
2.2
4.1
3.0
4.4
4.3
4.2
3-18-69
22.0
23.8
19.4
20.8
24.6
10.6
18.6
15.2
8.9
17.6
15.1
15.4
17.2
17.0
15.4
-14-69
7.6
6.7
6.0
--
--
4.4
3.6
7.1
1.1
1.2
3.8
4.0
2.4
1.8
1.9
2-18-e
--
16.2
14.9
16.8
12.3
8.6
14.2
10.4
--
12.0
13.0
12.7
12.8
13.8
3-25-69
24.2
24.9
21.0
20.7
24.8
14.9
16.8
16.4
16.8
16.2
16.2
15.0
17.4
17.0
14.7
Digester 3-4-69 3-11-69 3-18-69 3-25-69 4-1-69
1A 13.9 13.5 22.0 24.2 16.5
IB 14.2 20.1 23.8 24.9 14.5
1C 10.1 18.3 19.4 21.0 18.4
2A 14.6 20.1 20.8 20.7 16.5
2B 14.4 20.0 24.6 24.8 23.0
2C 17.0 17.7 10.6 14.9 11.8
3A 16.1 16.9 18.6 16.8 10.5
3B 15.9 16.6 15.2 16.4 10.2
3C 12.7 16.2 8.9 16.8 14.9
4A 14.5 15.8 17.6 16.2 15.3
4B 5.0 16.9 15.1 16.2 13.4
4C 14.5 13.5 15.4 15.0 14.5
5A 14.0 14.8 17.2 17.4 13.4
5B 14.1 13.8 17.0 17.0 13.0
5C 12.3 14.4 15.4 14.7 13.3
52
-------
TABLE XIII
SOLUBI£ COD
Digester
1A -- 15.6 15.8 10.4 6.1 -- 18.9
IB 6.9 14.1 15.6 9.9 5.5 -- 21.8
1C 8.8 12.8 14.4 11.1 6.0 15.6 20.0
2A 7.4 15.0 17.7 9.7 6.6 20.3 23.1
2B 7.9 14.2 14.8 6.4 5.5 22.3 25.9
2C 8.1 13.7 14.9 12.4 8.3 16.2 19.6
3A 8.0 14.7 15.3 14.9 7.1 29.8 27.0
3B 8.8 12.9 15.0 11.7 10.2 19.3 29.0
3C 8.8 13.5 15.6 11.0 3.8 22.8 31.3
4A 7.3 12.4 11.5 12.2 6.9 — 24.4
4B 10.6 15.5 14.6 13.1 10.6 16.9 23.6
4C 7.9 14.5 19.7 14.8 13.2 16.5 21.9
5A 12.6 13.9 12.3 12.6 10.7 22.6 29.1
5B 8.5 14.0 14.2 13.8 12.3 32.7 31.1
5C 8.7 14.3 17.6 13.7 12.3 3.7 23.1
(grams /liter)
11-26-68
— —
6.9
8.8
7.4
7.9
8.1
8.0
8.8
8.8
7.3
10.6
7.9
12.6
8.5
8.7
12-3-68 12-10-68 1-7-69 1
15
14
12
15
14
13
14
12
13
12
15
14
13
14
14
.6
.1
.8
.0
.2
.7
.7
.9
.5
.4
.5
.5
.9
.0
.3
3-4-69
16.
18.
15.
15.
21.
15.
22.
24.
27.
16.
16.
21.
18.
21.
16.
1
1
5
0
1
5
1
2
4
0
1
0
6
8
0
15
15
14
17
14
14
15
15
15
11
14
19
12
14
17
.8
.6
.4
.7
.8
.9
.3
.0
.6
.5
.6
.7
.3
.2
.6
3-11-69
15
19
17
22
21
16
26
26
27
22
22
25
19
17
16
.9
.9
.1
.2
.2
.8
.0
.8
.6
.1
.6
.3
.6
.6
.2
10.4
9.9
11.1
9.7
6.4
12.4
14.9
11.7
11.0
12.2
13.1
14.8
12.6
13.8
13.7
-14-69 2-18-69
6.1
5.5
6.0
6.6
5.5
8.3
7.1
10.2
3.8
6.9
10.6
13.2
10.7
12.3
12.3
3-18-69
18.
25.
20.
23.
23.
21.
17.
21.
22.
18.
18.
17.
22.
22.
18.
9
5
7
5
4
3
6
2
2
5
0
5
0
0
4
..
--
15.6
20.3
22.3
16.2
29.8
19.3
22.8
--
16.9
16.5
22.6
32.7
3.7
3-25-69
26
21
18
19
17
19
17
18
21
17
17
18
21
21
16
.3
.2
.7
.7
.8
.6
.2
.5
.5
.5
.3
.6
.2
.0
.8
Digester 3-4-69 3-11-69 3-18-69 3-25-69 4-1-69
1A 16.1 15.9 18.9 26.3 31.0
IB 18.1 19.9 25.5 21.2 22.7
1C 15.5 17.1 20.7 18.7 23.7
2A 15.0 22.2 23.5 19.7 21.2
2B 21.1 21.2 23.4 17.8 22.6
2C 15.5 16.8 21.3 19.6 19.1
3A 22.1 26.0 17.6 17.2 15.2
3B 24.2 26.8 21.2 18.5 16.0
3C 27.4 27.6 22.2 21.5 20.6
4A 16.0 22.1 18.5 17.5 14.3
AB 16.1 22.6 18.0 17.3 17.0
AC 21.0 25.3 17.5 18.6 13.4
5A 18.6 19.6 22.0 21.2 17.5
5B 21.8 17.6 22.0 21.0 14.8
5C 16.0 16.2 18.4 16.8 16.3
53
-------
TABLE XIV
METHANE CONTENT OF DIGESTER GAS
(percent)
Digester 10-8-68 10-31-68 11-6-68 11-12-68 11-21-68 11-26-68 12-5-68 1-10-69 1-17-69
1A
IB
1C
2A
2B
2C
3A
3B
3C
4A
4B
4C
5A
5B
5C
46
45
57
50
50
62
41
48
44
55
55
44
52
50
51
45
49
46
47
49
49
45
47
47
46
48
47
44
47
48
57
55
55
54
45
57
53
56
53
54
55
55
53
55
55
56
57
53
58
35
57
52
57
56
56
55
55
53
57
58
50
49
49
47
46
48
49
48
49
43
52
52
45
33
50
55
56
53
54
55
55
55
55
54
54
55
55
50
52
53
54
61
56
56
54
54
57
53
55
55
54
54
55
55
54
57
57
57
58
58
41
34
42
49
43
44
39
41
37
38
64
63
62
63
65
52
48
50
52
52
46
38
41
42
34
-------
TABLE XIV (Concluded)
Digester
2-7-69
2-14-69
2-21-69
Ul
1A
IB
1C
2A
2B
2C
3A
3B
3C
4A
4B
4C
5A
5B
5C
50
42
51
50
50
50
41
58
43
53
52
50
50
44
46
53
54
54
50
55
53
53
47
55
55
56
51
51
51
f- f- i U7
49
47
47
50
42
52
51
50
19
50
54
54
51
47
53
i-^O-O^
56
57
56
62
44
58
58
57
34
51
61
9
56
26
59
j-i^-oy
56
50
53
52
54
50
51
51
52
52
52
49
52
52
52
3-21-69
31
50
50
52
54
52
52
52
43
53
53
54
51
51
52
3-28-69
38
52
52
53
51
58
53
53
52
55
54
55
53
53
53
4-4-69
50
48
46
47
46
48
45
48
46
32
49
52
40
41
50
5-16-69
54
52
54
54
53
53
55
53
53
54
54
52
51
52
52
-------
BIBLIOGRAPHY
1. PHS Publication No. 1065, "1962 Inventory Municipal Waste Facilities,"
U. S. Government Printing Office.
2. Loehr, R. C., and T. J. Kukar, "Removal of Lipids by Conventional
Waste Treatment Methods," International J. Air Water Poll., 9, 479
(1965).
3. Hunter, J. V., and H. Heukelekian, "The Composition of Domestic Sewage
Fractions," J. Water Poll. Control Fed., 37, 1142 (1965).
4. Davis, J. B., Petroleum Microbiology, Elsevier, New York, p. 260 (1967),
5. Raymond, R. L., and J. B. Davis, "n-Alkane Utilization and Lipid
Formation by a Nocardid," Appl. Microbiol. . 8, 329 (1960).
6. Beerstecher, E., Petroleum Microbiology, Elsevier Press, Houston,
p. 192 (1954).
7. Jeris, J. J., and P. L. McCarty, "The Biochemistry of Methane Fermen-
tation Using C14 Tracers," J. Water Poll. Control Fed.. 37, 178
(1965).
8. Chouteau, J., E. Azoulay, and J. C. Senez, "Anaerobic Formation of
n-Hept-1-ene from n-Haptane by Resting Cells of Pseudomonas
Aeriginoso." Nature. 194, 576 (1962).
9. Sarda, L., and P. Desnuelle, "Action of Pancreatic Lipase on Emulsified
Esters," Biochem. et Biophys. Acta. 30, 513 (1958).
10. McKinney, R. E., Microbiology for Sanitary Engineers, McGraw-Hill
Book Company, New York, 1962.
11. Dr. William Spangler, Midwest Research Institute, Personal Communica-
tion (1969).
12. McCarty, P. L., "Anaerobic Waste Treatment Fundamentals," Public
Works. November 1964, p. 94 and October 1964, p. 126.
13. Mixing Equipment Company, Inc., "Town of Tonawanda Eliminates Scum
Formation on Sludge Digestion Tank," News from Lightning. April 1966.
14. Basu, A. K., "Treatment of Effluents from the Manufacture of Soap and
Hydrogenated Vegetable Oil," J. Water Poll. Control Fed.. 39, 1653
(1967). ~~
56
-------
15. Technical Practice Committee, "Anaerobic Sludge Digestion - MOP16
Sec. 3 Raw Sludge Feed to Digesters," J. Water Poll. Control Fed.,
38, 1840 (1966).
16. Technical Practice Committee, "Operation of Wastewater Treatment
Plants," MOP11, p. 44.
17. Lyman, B., G. McDonnell, and M. Krup, "Startup and Operation of Two
New High-Rate Digestion Systems," J. Water Poll. Control Fed., 39,
518 (1967). ~~
18. Cookson, J. T., and N. C. Burbank, "Isolation and Identification of
Facultative Bacteria Present in the Digestion Process," J. Water
Poll. Control Fed.. 37, 822 (1965).
19. Filbert, J. W., "Procedures and Problems of Digester Startup,"
J. Water Poll. Control Fed.. 39, 367 (1967).
20. Grune, W. N., C. F. Chueh, and C. H. Kaplan, "Effects of C14 and
Sr90 on Anaerobic Digestion," J. Water Poll. Control Fed., 35,
493 (1963).
21. Alonzo, W. L., and P. L. McCarty, "The Role of Sulfide in Preventing
Heavy Metal Toxicity in Anaerobic Treatment," J. Water Poll. Con-
trol Fed., 37, 392 (1965).
22. Standard Methods for the Examination of Water and Wastewater. 12th
Edition, American Public Health Association, New York (1965).
23. Loehr, R. C., and G. A. Rohlich, "A Wet Method for Grease Analysis,"
17th Ann. Purdue Industrial Waste Gonf., p. 215 (1962).
57
4 U.S. GOVERNMENT PRINTING OFFICE : 1910 O - 408-301
------- |