600778174
Energy and Economic
ASSESSMENT
of ANAEROBIC DIGESTERS
for Rural Waste Management
Tom Abeles, David Freedman,
David Ellsworth, LUC DeBaere
JUNE 1978
EP 600/7
7S-174
c.l
OASIS 2000
Rice Lake, Wisconsin
-------
ENERGY AND ECONOMIC ASSESSMENT OF ANAEROBIC DIGESTERS AND
BIOFUELS FOR RURAL WASTE MANAGEMENT
by
T. P. Abeles
D. F. Freedman
L. A. DeBaere
D. A. Ellsworth
OASIS 2000
University of Wisconsin Center - Barren County
Rice Lake, Wisconsin 54868
Grant No. R-804-457
Project Officer
Charles J. Rogers
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
-------
ABSTRACT
A technological and socioeconomic assessment of anaerobic digester
feasibility for small to mid-size livestock operations was undertaken.
Three full scale digesters and one pilot scale facility were under various
degrees of monitoring and evaluation to assess design and operational
problems as they affect the adoption and establishment of farm scale
anaerobic digesters.
Materials handling presented the greatest obstacle to satisfactory
operation of the full scale systems. Conversion of the biogas to elec-
tricity via standard engine-generator sets is capital and maintenance
intensive. Electrical conversion requires engine-generator sets which can
add 30% to the cost of the system, and which have conversion efficiencies
of only 10-25%. The system becomes more economical if the biogas can be
used on site for direct thermal loads, suggesting that the economic
feasibility of anaerobic digesters is site specific and should be closely
integrated with the total farming operation. If excess engine heat can
be recovered from electrical conversion equipment, and if provisions can
be made on the farm to level electrical loads and conserve energy, then the
economics are enhanced. Analysis was performed for farms with 100 animal
units.
Laboratory studies using a 2:1 mixture of dairy manure to Municipal
Solid Waste (MSW) showed that the biogas production per pound of volatile
solids added is nearly the same as for straight manure. Addition of the
organic portion of MSW to small farm digesters could make marginal systems
economically attractive.
Preliminary refeed studies indicated that the digested manure may not
have the same nutritional value as raw manure, and that the cost of dehyd-
rating the effluent for refeed to the same animals could be cost prohib-
itive for the small to mid-size farming operation.
The feasibility of growing hydrogen producing algae on the effluent
to enhance biogas production was rejected for the northern United States
due to temperature extremes and to the difficulty in culturing selected
species.
Socioeconomic research revealed the more significant factors under-
lying the adoption of agricultural innovations that will play important
roles in determining the extent of digester establishment. The need for
adequate service and maintenance organizations, specific standards and codes
applicable to design and construction of these units, and the need for
iv
-------
-------
support from insurance and financing institutions cannot be stressed enough
if digesters are to be established on a wide scale. Combining wastes from
several farms and/or communities could provide economies of scale provided
management and social barriers were overcome. The sale of power back to
the utilities would prove reasonable provided quality and reliability
could be maintained.
This report was submitted in fulfillment of contract # R-804-457 by OASIS
2000 under the sponsorship of the United States Environmental Protection Agency.
This report covers the period September 1, 1976 - March 31, 1978, and work was
completed as of June 31, 1978.
-------
-------
TABLE OF CONTENTS
Title Page i
Disclaimer ii
Forward iii
Abstract jv
vi
Contents vii
Tables viii
Figures x
Acknowledgements xii
Section Page
1. Introduction 1
2. Conslusions and Recommendations 2
3. Overview 7
4. Systems Description 16
5. Technical Components 34
6. Biochemical Concerns and Advanced Research 47
7. Economics 55
8. Other Considerations 67
9. Market Feasibility 78
References 82
Appendices 87
A. Anaerobic Digestion of Dairy Cow Manure Plus the
Organic Fraction of Municipal Solid Waste: A Pilot
Feasibility Study 87
B. Feasibility of Dehydrating Anaerobic Digester Effluent
for Refeeding (Preliminary Study) 116
C. Comparison Between Influent and Effluent of Anaerobic
Digesters as Refeed 125
D. The Feasibility of Increasing Methane Gas Production
From Anaerobic Digesters Via Hydrogen Producing Algae ... 131
E. Assumptions Used In Economic Evaluations of Anaerobic
Digesters 149
F. Factors Underlying the Adoption of Agricultural In-
novating 154
G. Existing or Planned Digester Operations in USA 157
vii
-------
-------
TABLES
1. Brockman Stables—Assumptions & Systems Sizing 20
2. Brockman Stables—Summary of Results 21
3. Modes of Pump Operation—Green Bay System 24
4. Energy Balance - Green Bay Digester 25
5. Predicted Energy Balance (BTU/Day) 26
6. Loading Rates, Gas Production, and Chemical Parameters, Green Bay ... 27
7. Chemical Analysis* of Rice Hulls Used for Turkey Bedding 32
8. Analysis of Turkey Waste in Rice Lake System 33
9. Average Gas Sample Analysis Rice Lake System 43
10. Influent Analysis on Representative Samples* 54
11. Baseline Costs for a Covered Anaerobic Lagoon 56
12. Components and Installed Costs of Existing Rice Lake Digester 57
13. Annual Costs of Rice Lake Digester 58
14. Annual Biogas Production, Rice Lake 59
15. Cost/KWH of Rice Lake Digester Normalized over 20 Year Life 60
16. Cost/KWH of Optimized Digester Electrical System 62
17. Costs and Returns of Direct Biogas Combustion on Site 64
18. Comparison of Digester Systems 65
19. Numbers of Farms in the Following Wisconsin Counties (NW District) ... 79
20. Number of Digester Installations in Northwest Wisconsin 80
Al. Feed/Discharge Quantity 94
A2. Composition of MSW Shred-Combined Sample from Appleton,& Madison, WI . . 96
A3. Composition of Feedstock 2 96
A4. Summary of Background Information 97
A5. Biogas Production Data - Average for 13 Digesters 99
A6. Analysis of Variance of Biogas Production Data 101
A7. A Comparison of Means, By Feedstock, Over Time 101
A8. Average Biogas Composition for 13 Digesters 102
A9. Summary of Results - Chemical Analysis 104
A10. Comparison of Biogas Production Efficiency 106
All. Summary of Results 108
A12. Summary of Data on Anaerobic Digestion of Dairy Cow Manure 110
A13. Biogas Production From Combined Wastes 112
Bl. Companies Selling Solids Concentration Equipment 118
B2. Lear Filter Test Data (EIMCO corporation) ,119
B3. Analysis of Sample Cake and Filtrate (EIMCO Leaf Filter) 119
B4. Analysis of Vibroscreened Filter Sample (Kason Corporation) 121
Cl. Nutritive Analysis of Samples 126
C2. Total Amino Acids in Influent/Effluent Sample 127
C3. Concentration in (W/W x 100) of 8 runs 127
C4. Overview of-Results From Amino Acid Study 129
El. Assumptions Used In Economic Evaluations Anaerobic Digesters 149
viii
-------
-------
Tables continued
E2. Costs of Installed Components Used In Anaerobic Digester System . . . 150
E3. Annual Costs of Digester With Heat Recovery Using Deisel Powered
Generator 150
E4. Costs of Installed Components For Digester System With Direct
Combustion of Biogas on Farm 151
E5. Annual Costs of Direct Combustion Digester 151
E6. Heat Loss From Rice Lake Fermentation Tank 152
E7. Theoretical Daily Biogas Production Using Dairy - MSW Mix* 152
IX
-------
-------
FIGURES
1. 55 Gallon Drum Type Digester 8
2. Horizontal displacement digester (Fry) 8
3. Schematic of Ecotope's 100,000 gal. dairy cow manure Biogas
Demonstration Plant 9
4. Schematic of University of Minnesota Swine Digester 9
5. UW-Green Bay pilot scale digester 10
6. Two-phase anaerobic digestion, Institute of Gas Technology 11
7. Schematic of the experimental anaerobic digester installation .... 11
8. Earth supported flexible me-brane digester, Energy Harvest Co. ... 12
9. Redesigned fermentation tank-plywood and earth support, Energy
Harvest, Co 13
10. Patented solar heated digester, Biogas of Colorado 13
11. Anaerobic digesters under study. Wisconsin & Michigan 16
12. Brockman stables slurry and gas systems 18
13. Brockman stables heating system 18
14. Manure handling and heating systems, Green Bay 23
15. Gas Utilization System, UWGB 23
16. Schematic of fermentation tank feeding system, Ludington 28
17. Rice Lake control house 30
18. Temporary turkey manure collection, Rice Lake. 31
19. Rice Lake influent tank 34
20. Ludington Fermentation tank 36
21. Rice Lake fermentation tank 37
22. Schematic of original manure handling system, Rice Lake 39
23. Schematic of redesigned manure handling system, Rice Lake 40
24. Gas and effluent storage system 41
25. Gas scrubbing and compression sustem 42
26. Schematic of SIHI ring seal gas compressor, Rice Lake 42
27. Schematic of Rice Lake power system 44
28. Schematic of original heat exchange system 45
29. Schematic of rebuilt heat exchange system 45
30. Loading rates and chemical operating parameters of Green Bay system . 48
31. Cost of electricity from $40,000 digester vs. cost of electricity
from utility company 60
32. Range of costs for electricity from digester systems (100 cow dairy). 61
33. Btu/$ of different energy and digester types 63
34. Significant factors suderlying the adoption of agricultural
innovations 70
35. Chippewa River Basin, Upper Mississippe Region - Location map
(from Upper Mississippi River Basin Commission) 74
36. Barron County water quality problem areas 75
37. Northwest Wisconsin Crop Reporting District 81
-------
-------
Figures continued
Al. Reactor 89
A2. Top view of water bath 90
A3. Front view 90
A4. Biogas measurement system 91
A5. Average Biogas Production for 13 Digesters 100
A6. Total and Volatile Solids - Feedstock 2 105
A7. Ammonia and TKN - Feedstock . 106
A8. Percent of TKN in the form (NH4 - NHg)- N for Feedstock 2 107
A9. Chemical Oxygen Demand - Feedstock 107
Bl. Cost of drying waste at various moisture contents (heating oil
at $.50/gal) 117
B2. Possibilities for using effluent as refeed 120
Dl. Hydrogen and Rumen Methane Production 132
D2. Schematic of laboratory digester for combined algae/manure cultures .136
D3. Schematic of gas measuring system 137
D4. Average cell population in Experiment 1 138
D5. Results of Experiment 2a 139
D6. Results of Experiment 2b 139
D7. Results of Experiment 3a 140
D8. Results of Experiment 3b 140
D9. Results of Experiment 3c 141
D10. Total Algal-digester system design with algae introduced
directly into digester 143
Dll. Total system design using separate hydrogen producing chamber . . . .144
El. Daily electrical load profile, Rice Lake farm 153
-------
-------
ACKNOWLEDGEMENTS
Roland Brownlee, presently professor of Economics at the University of
Wisconsin Center - Barron County and director of OASIS 2000, was director
of the project for the duration of the study and was largely responsible for
its inception. Without his participation and encouragement during the
course of the work, this study would not have occurred.
Timothy Rowe and James Laundre prepared the material on hydrogen pro-
ducing algae for enhancement of methane production. James Berry, professor
of Economics at UW-Marinette, assisted in the socioeconomic assessment of
anaerobic digesters and was responsible for the initial groundwork: "Factors
Underlying the Adoption of Agricultural Innovations". Assistance from Elmer
Harris is also acknowledged in this respect. Susan Ernst did a great deal
of research on the impact of water quality regulations on the adoption of
digesters. Help from Eugene Hausner and Leonard Splett, of the Barron County
Soil Conservation Service, George Anderson of the DNR, and Kirk Johnson,
is acknowledged in this respect. Wayne Arntson was responsible for
performing the Rice Lake laboratory experiments and in the field monitoring
of the Rice Lake digester. Debra Patin and Janna King performed a great
deal of the market feasibility and socioeconomic data collection, were
responsible for compiling the list of digesters in the USA, assisted in
researching the institutional factors affecting feasibility, and provided
invaluable moral support. Gene Dale, Jerry Maelstrom, and Dick Vancura of
Energy Harvest provided valuable technical assistance throughout the project.
The cooperation of Phyllis, Leonard, and Len Schieffer, of Fertile Acres
Farm, Rice Lake, in the use of their farm and digester for field testing is
gratefully acknowledged, as is the assistance of Jerry Jerome and Bob Simons
of Jerome Foods.
Many thanks to all who assisted in providing information and assistance
throughout the project: Cal Sprain, Floyd Christiansen, and Ed Glass of
Northern States Power, Ken Smith of Ecotope, Phil Goodrich of the University
of Minnesota, Jan Freedie of Production Credit Association, the Waukesha
Engine division of Dresser Industries.
Particular thanks go to Lorrie Money and Leesa Roberts for final report
preparation assistance.
David Freedman built, operated, and collected two years of data on the
Green Bay pilot scale digester, performed the laboratory analyses throughout
the project, conducted and prepared the material on the combined MSW-Manure
experiments, and, in conjunction with Luc DeBaere prepared the material on
XII
-------
-------
dehydrating the digested effluent for refeed.
Luc DeBaere also prepared the comparative research on the refeed value
of the effluent versus that of the raw manure.
David Ellsworth coordinated and wrote the Economic, Social and Institu-
tional considerations, and Market Feasibility studies, and edited the final
report.
Graphics by: David Ellsworth, Karen Funk, JoAnn Swearingen, Oscar Mogen
Cartography by: Susan Ernst
xm
-------
-------
SECTION 1
INTRODUCTION
Recent concern over dwindling natural resources has enhanced the
development of technologies that are responsive to energy shortages,
resource conservation and environmental degradation. Growing interest in
the anaerobic digestion of farm animal wastes as a potential source of
fuel, animal refeed, and low cost fertilizer necessitated an assessment of
farm scale digester systems.
Three full scale digesters and one pilot facility were evaluated on
the basis of design and operational problems. A beginning assumption was
that small digester systems feasibility would transcend the the concerns
for engineering and chemical design alone. Methods of optimizing engineering
and economic feasibility were developed with particular emphasis on tech-
nology transfer to the many small to medium sized farms in the Upper Midwest,
where the study occurred.
Mechanisms which were considered to enhance the attractiveness of
these systems included the use of digester effluent for animal refeed and
aquaculture or algal farming. Because a combined Municipal Solid Waste
(MSW) and animal manure system could make small or marginal digester systems
economical, while at the same time help alleviate rural solid waste problems,
the addition of MSW to animal manures for enhancing biogas production was
experimented with on a laboratory scale. Socioeconomic research was also
conducted to determine the public acceptance of these systems. Institutional,
regulatory, and legal factors were examined to ascertain the impact of
attitudes and decisions made in these arenas on widespread establishment
of anaerobic digesters.
Characteristics of each site where the digesters were installed,
particularly farm energy utilization, were also examined to ascertain the
economic trade-offs between using the gas directly for thermal loads or
convertirg to electricity.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
TECHNICAL COMPONENT
1) Materials handling is the major barrier to widespread utilization
of anaerobic systems.
Four systems were under study during this contract. Two were built as
commercial ventures from new materials. One was contracted for but partly
built from used equipment, and one was a pilot plant built from new
materials. Additionally, over thirty existing systems were either visited
directly or were investigated via telephone conversations with owners,
designers, or operators. All systems had, or were having, significant
problems with routine conveyence of the animal wastes and other feed
materials into the fermentation tank. These problems were either related
to pumping or piping of the material. Few of the existing systems could
sustain unattended operation for any length of time. Whether the systems
were designed by professional engineers or lay people did not seem to matter.
Thus, systems became labor intensive.
Recommendation--
Research is needed on efficient feed design to allow for routine,
trouble free operations.
(2) Materials preparation for feeding was not carefully considered.
Too much energy in the past has focused on the chemical and bio-
chemical composition of the feedstocks for fermentation processes and not
enough attention given to the physical-mechanical characteristics. Some of
the related problems encountered in the systems under study included:
Heavy bedding of animals leads to excessive vegetable matter introduced into
the fermenter. Unprotected or unpaved animal facilities lead to the in-
clusion of rocks, dirt, and debris clogging or preventing adequate feeding
of the material. Lightweight materials create scumming problems in the
fermenters. The use of Municipal Solid Waste is limited because of its
variable nature and non-digestable components. Inert materials contained
therein may plug fermenters and create problems in effluent disposal.
Recommendation-
Work needs to be done on effective means of separating the inert
-------
materials from Municipal Wastes if they are to be used in small scale
fermentation systems.
3) Gas production on current designs for small scale systems does not
adequately address the daily and seasonal variations in energy use.
On farm energy consumption varies with the seasons. Daily farm
electric peaks also coincide with power company peaks. Gas storage is a
problem because biogas does not liquify readily. Carbon dioxide scrubbing
systems, even if low cost, efficient designs were available would not alle-
viate the storage problem. Few natural gas pipelines are sufficiently close
to allow for seasonal sales and buy backs. Conversion to electricity is
not, for the most part, cost effective. Sale of the electricity back to
the utility company would even be less economical due to the corresponding
peak loads. Therefore, more attention needs to be given to designing
systems which can follow seasonal loads.
Recommendation—
Waste storage-fermentation systems need to be designed to preserve the
"fermenter fuel" for seasonal peaking. Also, low cost storage and scrub-
bing systems need investigation.
4) Conversion of biogas to electricity is capijal and maintenance
intensive with an engine-generator set.
Standard power units, in field situations, produce electricity with
only a 10-15 percent conversion efficiency. Spark ignited or dual fuel
diesels offer a 25 percent conversion efficiency. Unless the waste heat
can be recovered for heating use, the cost of these units is prohibitive
for small scale use. Reformer-fuel cell combinations to produce electricity
from biogas, at 30-40 percent efficiency, offer a near term option which
appears more attractive than the current engine-generator sets. These
units can use the biogas directly, generate electricity at higher effi-
ciency, produce usable heat, and produce carbon dioxide which can be used
in greenhouses.
Recommendations—
More design work needs to be done on heat recovery systems from
onsite power production units. Fuel cells for biogas systems need careful
attention.
5) Neither this study or any other, up to this time, had addressed
alternative uses for the biogas.
Biogas contains the essential components for production of either
methanol or ammonia, as currently practiced in industry. The question of
scale and ease of production affecting the applicability to farm scale
systems is open to question. Exploratory talks with companies from which
-------
to obtain the necessary catalysts has been started under this contract.
Recommendation—
Research on uses of biogas for other than direct combustion or power
production needs to be started.
SOCIOECONOMIC COMPONENT
1) Electrical production with small scale systems, from an economic
perspective, is marginal and totally unoptimized.
Most farms are heavily electrified, thus, conventional wisdom dictates
that biogas be used to produce electricity. On a cost/Btu basis this is the
most ineffective means of using the gas. Thermal loads would be more cost
effective for biogas usage. No studies exist on daily farm energy profiles
or on the number of farms which could be converted from high electrical use
to gas and what the cost of doing so would be. The two systems studied
which utilized engine generators dumped 80 percent of the waste heat into
the environment. Optimization for utilizing this heat will be required to
enhance feasibility. Success of small scale digesters dictates that the
biogas be used efficiently.
Recommendation--
Energy-power studies need to be undertaken to successfully integrate
digester systems with normal farm operations with respect to energy utiliza-
tion.
(2) Maintenance and service of digester systems has not been adequately
addressed.
Contact with farmers during the course of this work indicated the
need for reliability and service of these systems. All systems under
study suffered from periodic breakdown and maintenance problems. Many of
these were the result of faulty or experimental design. Other problems
could be classified as routine maintenance which might occur under normal
farm operation. Neither of the commercial systems had a service component
in the contract agreement, nor had the company seriously considered such a
component. Acceptability of digester systems is dependant on service
organizations to maintain them. Without adequate service, safety and health
problems could be compounded and eventually lead to abandonment of the sys-
tems.
Recommendation-
Research needs to be carried out on technical, legal, and safety
considerations to assure reliable systems operation and maintenance. This
might include setting engineering and construction standards and lead to
licensing of the systems.
-------
(3) Preliminary mapping of farms in Barron County, Wisconsin, indicates
that there is a possibility of combining wastes from several farms and/or
municipalities.I
Combining wastes from several farms and/or municipalities could provide
some economies of scale, provided management schemes were developed for
payment, power, and fertilizer distribution. Also, transport, maintenance,
and legal constraints of commercial power production must be overcome.
Recommendation—
Guidelines need to be developed for community power producing, waste
handling systems to allow for small and intermediate scale systems.
(4) The sale of power to utilities may prove reasonable if quality and
reliability can be maintained and the technical concerns of three and four
can be met.
Small scale systems integrated with the power grid create management
problems for the utilities. Electrical feedback into the power grid from
many small generators creates potential dangers for linemen when a storm
knocks the power out. Also, the power generators examined produce power
below the quality maintained by utilities. This study began to investigate
microprocessor technology for systems monitoring and control. Micropro-
cessor technology can help to maintain levels of quality and provide the
necessary service tool to aide in the creation of systems reliability.
Recommendation—
Extensive work needs to be done on microprocessor controllers and
power conditioning systems to enhance alternative economic paths and, hence,
viability of small scale systems.
(5) The introduction of municipal solid waste to anaerobic digesters
enhances small systems feasibility and provides an alternative to rural
waste problems.
This research showed, in a preliminary study, that the addition of 2:1
manure and municipal solid waste could equal, on a volatile solids basis,
the gas production of straight manure. Other mixtures might provide a
synergistic effect and have hig.her gas production rates. Also, several
studies indicate that specialized microbial populations might affect gas
production; systems could be conditioned to enhance this. Work on the 2:1
mixture seemed to indicate this possibility.
No serious work has been done on wasteshed modeling from either a
transport, source separation, or quantity perspective in the case of diffuse
population areas. Some solid waste studies have shown that county level
systems in rural areas can often be more economical than regional systems.
No work has been done on even more diffuse systems which may collect from
a neighborhood with the farmer perhaps paying for the refuse.
-------
Recommendations--
a) Socioeconomic studies on microscale disposal of municipal solid
wastes via farm scale fermentation systems need to be carried out, including:
transportation studies, attitudinal studies, and wasteshed mapping.
b) Laboratory and field studies on mixtures of MSW with different
quantities of animal wastes, other organic matter such as field residues,
and different types of animal wastes should be initiated.
c) Studies on the microbial populations should be carried out to
ascertain if shifts in microbial populations do exist and can significantly
enhance organic decomposition and gas production.
-------
SECTION 3
OVERVIEW
Anaerobic fermentation processes have been used for many years in the
treatment of municipal sewage. Several years ago, as municipalities
replaced and enlarged their facilities, many of these anaerobic systems
were abandoned in favor of aerobic treatment which was lower in capital
cost. As the energy crisis developed, and lifetime costing clearly became
a better measure of the investment an industry or municipality would have
to make, anaerobic systems which produced energy as a byproduct became
attractive again.
Prior to the awakening of the urban environment to the potentials of
anaerobic fermentation, the rural sector became interested in the process
for obtaining energy while treating and disposing of animal manures and other
agricultural residues such as food processing wastes and cellulosic mater-
ials. These systems were used in rural Germany (1) during World War II
and more recently in India (2,3), and China (4,5). With cheap energy in the
U.S. and no severe pressure on the agricultural industry to meet pollution
control standards, the anaerobic process was virtually ignored until about
ten years ago.
Outside of a few farmers and some people seeking a self-sufficient
lifestyle (6, 7), little formal research was done to develop reproducible
systems. Most of the University research was in the departments of micro-
biology which were interested in the microprocess (8, 9) or in departments
of civil or sanitary engineering. (10) Some work on agricultural wastes
was undertaken by the Universities of Illinois, Wisconsin, and Cornell
(1, 11, 12) and some full scale systems were being developed on one shot
bases by entreprenuerial farmers.
Some of the earliest agricultural anaerobic systems to reach public
attention were those of Auerbach, Rutan, Merrill, and Fry (6, 7, 13), which
reached the public largely through alternative publications. Most of these
systems were designed for a small number of animals in which 55 gallon
drums, or the equivalent, were used for batch or continuous fermentation
(Figure 1). This experience was translated into some very simple design
manuals published by Merrill and Fry (13, 14), and Leckie (15) in the
early and mid 70's. Two exceptions to the small size are notable. The
first is the work of Fry in California, Figure 2, which resulted in a
rather extensive, but simple design manual.
-------
organic wastes
& water
methane gas
production
t>
affluent to
gardens
FIGURE 1 55 GALLON DRUM TYPE DIGESTER
(Auerbach, Merrill)
uum dug
coiTugatcd ito
msulilion
I scum
door —
t-.
»"•
i-^Sv
1 ^ ,•'-",, ,:.
^^--
tlllutnl pifi
••— — -^
^Z?(
^*S ,
(K— «
r ' ' want Wiling piii'
|«(t>
Figure 2. Horizontal displacement digester (Fry).
-------
The second is a unit which was built by Niel Huber in Connecticut in
the late forties. This system was developed for the purpose of selling the
nitrogen rich effluent and capitalizing on the government subsidies for
nitrogenous fertilizer. Since these subsidies were aimed at supporting the
ammonia plants built during World War II, the Huber facility did not qualify
and was abandoned as an unprofitable venture.
As the small scale research became better documented by a growing
movement of individuals labeled "appropriate or alternate technologists",
larger farm operations, major universities, companies and small entrep-
reneurs became interested in the feasibility of anaerobic digesters. Also,
as interest developed, definite philosophies of design emerged, each aimed
at improving the cost effectiveness of the process.
The majority of experimenters and researchers believed that operation
in the mesophillic region, 15 C-45 C, to be the most appropriate. It
seemed the easiest to control and design. Examples of these early systems
can be seen in units constructed by Ecotope, the University of Minnesota,
and the University of Wisconsin-Green Bay (Figures 3-5).
Figure 3. Schematic of
Ecotope's 100,000
gallon dairy cow
manure Biogas Dem-
onstration Plant.
CiMilety ul •COlOD* QroUO
C«rul Obirtim, btlphtc TiiKI
Figure 4. Schematic of University
of Minnesota swine
digester.
fltP W
J
-e—>-
[XSEOER MANURt BCIURM
TOR
Bt FE£D
-------
SAiMCTER
TO LA600N
Figure 5. UW-Green Bay pilot scale digester.
Another school advocated thermophilic digestion (temp.). One of the
major proponents of this theory was Hamilton Standard, Division of United
Technologies (16). Perhaps the first company to develop a package unit for
sale, the majority of their units are currently placed on sites being
funded by U.S. government studies. Several Universities have also opted to
focus primarily on the thermophilic process (17, 18).
The temperature question seems to have divided researchers into two
schools. There are other engineering factors which cut across these lines.
These include tank design, construction materials, and operational criteria
such as plug-flow, no mix, partial mix, completely mixed systems, and the
sophisticated two-phase operation developed and patented by the Institute
of Gas Technology (Figure 6). This two-phase system, originally intended
for municipal use, appears to offer real potential as a second generation
farm system.
Tank materials represent one of the largest costs for a farm system.
Thus, they must prove durable over a long period of time and/or be replace-
able at nominal costs. The option for extreme durability is exemplified by
the use of glass lined metal tanks, represented by the work of ECOTOPE in
Washington State (19, 20). These tanks are marketed by A.O. Smith's
Harvestore Division and were used by that company in an experimental sys-
tem on a Wisconsin farm (21). (Figure 7) The system is no longer opera-
tional and the company has not indicated what its plans are in this area.
One of the lowest capital cost fermentation tanks is a plug flow
design composed of a flexible membrane with a woven fabric reinforcement.
This is similar to the lined and covered lagoons designed by Environetics
(22).
10
-------
Figure 6. Two-phase anaerobic digestion, Institute of Gas Technology
SLURRTSTORE
DIGESTER
BOILER
MIXING
TANK
DILUTION WATER
MANURE
Courtesy of A.0.Smith Corporation
Figure 1. Schematic of the experimental anaerobic digester installation
11
-------
Figures 8 and 9 show two such configurations designed by Energy Harvest
Company. Figure 10 is a schematic of a similar system, with solar heating
proposed by Biogas of Colorado. A variation of this design is also under
investigation by Cornell University under Department of Energy funding.
These flexible membrane units all operate in the plug flow mode, meaning
the material goes in one end and out the other, as in a pipe. Fermentation
is controlled largely by temperature maintenance and flow rate (usually
15-30 days throughput). A number of these systems had gas jets installed
along the length to mix the effluent, because standard theory indicated
that mix-ing was imperative. Field experience with these units shows that
the same quality of fermentation is achievable without mixing; thus, most
of these mixers have never been used.
For the most part, anaerobic digesters in the U.S. and Europe have
been vertical tank configurations with intermittent mixing. The original
theory was that the microbial populations needed to be intimately con-
tacted with the substrate; mixing would aid this process. Experience with
packed bed reactors (1), expanded bed reactors (70), separated phase or
FLEXIBLE MEMBRANE:
BIOTAS STORAGE
Figure 8. Earth supported flexible membrane digester, Energy Harvest Co
12
-------
...•.•;.••..•. -. •.••.:•:.•>:•.•:•;••••«.' :•••••• .•-:••>!•
-.«- .»-.-;• ••;•:•*•;•!.:;.•...••.•>•.?.-.•••.:•- -
Figure 9. Redesigned fermentation tank-plywood and earth support,
Energy Harvest,
Figure 10. Patented solar heated digester, Biogas of Colorado
13
-------
stage operation (23), high solids reactors (24), and even land filled
systems (25) has shown that this assumption is not necessarily valid. This
has caused serious reevaluation of engineering design and operation of an-
aerobic fermenters, expecially where the influent is essentially uniform.
Cornell University is currently investigating the mixing question by sys-
tematically operating a digester in a variety of modes, from plug flow to
complete mix.
The concern for cheap energy or energy independence from conventional
utilities, rather than any inherent interest in pollution control, sparked
the current interest in farm scale systems. As interest grew, so did in-
vestigation of the hard economics. This is the economics based on conven-
tional cost/benefit analysis as opposed to crediting for externalities.
It was argued that the systems would have to pay for themselves or a farmer
could not afford to install a system just to cut off the local power and
fuel companies.
In attempts to show positive economics other benefits have been
factored in with economic values. These include enhanced fertilizer value
of the effluent, increased value of the effluent as animal feeds (26, 27),
and the use of these systems to control both water and air (odor) pollution.
Also, economies of scale have been factored in, since the cost is not a
linear function of the amount of waste available for fermentation.
A number of studies have attempted to ascertain the minimum size which
was economically feasible. From the examination testing of several full
scale systems, this study seems to indicate that the current state of the
art puts the 100 cow or equivalent unit as a lower limit, if a very
efficient system is designed for the production of gas but not electricity.
Almost 100 percent of the gas must be used efficiently, and then only a
five percent return on investment (ROI) is realized. This is a poor ROI
from a business perspective where industry would like to see a 13-15 percent
return. The case with electrical generation is even more precarious with
higher capital investments, high maintenance costs, and low conversion
efficiencies (at best 25 percent gas to electric). If the waste heat is not
utilized and almost 100 percent of the electricity is consumed, the cost/kw
for a biogas plant exceeds that supplied from the local utility even taking
all credits and escallations into account.
Our research clearly shows that at the present time and in most cases,
the interposition of a fermenter between an animal waste source and a well
designed storage system must be justified on the gas production alone.
Also, in most cases where the energy is to be used entirely on site, the
economics will dictate the use of the gas produced for thermal loads as
opposed to conversion for electrical energy. And, the economics require a
very high conversion efficiency with little rejection or loss of the ther-
mal energy produced. This analysis holds for the systems researched under
this present study; which implies beef and dairy operations using animal
wastes as a sole source of influent and which have only these same animals
available for refeeding the effluent.
14
-------
The above indicates that digester systems feasibility transcends the
concerns for engineering and chemical design alone. It must also take
into consideration the particular site in which it is to operate. Charac-
teristics of the site may override some purely engineering optimization
considerations.
There exist other mechanisms for enhancing the attractiveness of
these systems. As mentioned above, such areas as use of the effluent for
refeed is possible; in some areas, aquaculture and algal farming could be
significant in proving economic feasibility (27, 60).
Another area which was investigated in this study was the addition of
municipal solid waste (MSW) to animal manures. This could have a twofold
advantage. First, this system could help alleviate rural and in some cases
urban disposal problems. Also, combinations of MSW with animal manures
could make small or marginal digester systems economical.
Success of these units is also predicted on public acceptance of the
systems. Socio/economic parameters such as financing, safety, insurance,
service, and farm operational changes must be factored in.
Interviews with farmers indicates that the availability of operating
systems and established service organizations would be beneficial for
widespread adoption of anaerobic digesters. Also, federal or similar
support would help diffusion rates of these systems.
Standards and regulations dealing with construction, installations,
and inspection of digesters would enhance the ability to finance and insure
the projects. Certification would also be necessary before government
assistance could be provided.
15
-------
SECTION 4
SYSTEMS DESCRIPTION
During the course of this work, three full scale systems and one
pilot system were under various degrees of study. This included the proto-
type full scale system of Agricultural Energy Corporation in Ludington,
Michigan, the full scale system in Rice Lake, the pilot operation in Green
Bay, and the system in Neenah, Wisconsin (Figure 11). This latter unit,
designed for about 125 horses was constructed from old brewery tanks.
None of the full scale systems operated on a continuous, trouble free
basis. The pilot operation ran continuously for about two years with only
minor problems and occasional pump breakdown. The horse stable operation
is currently not operational and its fate is in doubt partly because of
faulty design but primarily due to the unanticipated expense which has to be
incurred to construct a barn cleaning system consistant with proper stable
management.
Figure 11. Anaerobic digesters under study.
Wisconsin - Michigan
™ Digester Locations
• Major Cities
•
100
16
-------
The Ludington system underwent a complete rebuild of the fermentation
section. While it is now functioning adequately, it too, suffers from a
materials handling problem on the front end. The feedlot is not winterized
and is subject to freeze-over which makes manures unavailable for fermenter
feedstocks.
The Rice Lake system underwent several major modifications in the
control, power producing, and materials handling design which prevented
extensive, continuous monitoring of operational parameters, though the
data obtained paralleled closely the laboratory models. Materials handling
problems here were due largely to the temporary nature of the feed system
used for the turkey waste study.
The overall conclusions as a result of this study indicate that
digester feasibility rests primarily in the development of reliable mater-
ials handling systems and that fermenter tank design and operation is of
secondary concern. Fermenter operation becomes a problem when there are
large quantities of bedding or other cellulosic material included with the
manures or essentially non-digestible bedding such as rice hulls are in-
cluded. Then one tends to see systems stoppage either through slugging or
scum buildup.
All physio/chemical data reported were obtained using procedures
described in Standard Methods (71) except as noted below. Gas analysis
was carried out in a Carle #8000 G C using two columns and standardized
against calibrated standards prepared by Scientific Gas Products. Phos-
phate analysis was carried out on a Technicon II auto-analyzer using
standard procedures described in the operations manual.
SYSTEMS DESCRIPTIONS
Brockman Stables, Neenah, Wisconsin
The system, schematically depicted in Figure 12, is constructed from
ten epoxy lined, used brewery tanks of approximately 37,850 litres
(10,000 gallons) or 40 M3 (1340 ft6}. The system is completely enclosed
in a standard metal farm building which is insulated with styrofoam sheet-
ing. The tanks, which are horizontal, are uninsulated and are heated by
coils made from black iron piping and right angle elbows. The coils, six
loops in each tank, are mounted so they follow the perimeter of the tank and
are held by clamps welded to the side of the tank. Figure 13 shows a
schematic of the tank heating system.
The tanks are configured in a series/parallel configuration. The con-
cept here was to parallel tanks to get proper volume (to create a pseudo
75,000 litre, '20,000 gallon ' tank). Tank Al is not depicted because it
was removed from service early in the system operation and converted to
gas storage. El serves as a settling tank and E2 as a mixing tank. The
original plan called for an aquaculture operation in the lagoon and it was
felt that some preliminary clarification would be needed. E2 would then be
17
-------
use
digestion tanks are epoxy lined,
used brewery tanks
effluent line
2 pipe
Figure 12. Brockman stables slurry and gas systems.
f
wm-
mm
89 gal.
-T
>
H20 heater
mm
mm
C2
boiler
275 gal.
di
flWWW-
\ h.p.
Figure 13. Brockman stables heating system.
18
-------
used to add nutrient supplement for the aquaculture operation.
The "mixing" tank received the input from the barn. As the stable was
about a quarter mile from the digester system, the original scheme had the
manure collected by a spreader. The spreader unloaded to a hay chopper
which conveyed the solids to the mixing tank. Early troubles with too much
bedding forced the stable to hand sort each stall to eliminate the bedding
and then hand load the mixing tank. This task became time consuming and
cost prohibative and eventually led to systems shut down.
The, system operated in a complete mix mode. Basically all tanks could
be isolated from each other with hand operated valves. Manure was added to
the mixing tank and water added to reach a total solids level of about ten
percent. This mixture was then pumped to tank A^ and valves opened to
allow excess volume to be displaced from the system to the lagoon. Each
tank pair could then be mixed in isolation by judicious choice of valves.
The system was supposedly designed for a 30 day retention time with
a full complement of 125 horse_s based on a volume of J)3 M3 (1 ft^) of
manure and bedding/day and a dilution with .04 - .06 MJ of H,,0. The
system was sized and designed by individuals not trained in either
science or engineering and standard design procedures were not followed.
Because of the underloading of the system, retention times were often 120
days or greater.
Table 1 shows the basis for the systems sizing including assumptions.
Table 2 shows a typical set of results throughout the system. It is clear
that the materials are quite stabilized and if filtered to remove parti-
culates and chlorinated to remove pathogens, the water would be potable.
Gas analysis on the system yielded CH./C02 ratio of approximately 45/55.
The gas system consisted of a pipe line which tapped all the tanks in
a paralled arrangement. The pipes led to a two stage Ingersol Rand com-
pressor (#LPG-5) with a 60 gal high compression tank, A, was later added as
gas storage. The gas fed the hot water heater which maintained fermenter
temperatures in the 95°F region and could also be piped to the stable for
heating; low feed rates kept the gas production at a level that this was
never routinely accomplished. No dewatering or hydrogen sulfide removal
was attempted.
The system cost in the neighborhood of $40,000 and operated essenti-
ally trouble free. All operations were manual except the gas compression
and distribution system which was controlled by pressure transducers. The
system was shut down largely because it became too labor intensive to
collect the manure with a minimal amount of bedding. Thus, the system did
not produce sufficient gas to warrant its continuance. Also, the original
intent of the system was to process the manure to reduce the mounds of
horse manure, which were self composting and hence both a fire and odor
problem. Since the digester system could not handle all the waste, the
composting was not eliminated and the dual operation became a source of
frustration and additional expense.
19
-------
TABLE 1. BROCKMAN STABLES-ASSUMPTIONS & SYSTEMS SIZING
Assumptions
1.) 30 day retention time
2.)
3.)
4.)
5.)
6.)
7.)
Raw manure characteristics:
80/o moisture (M)
20% total solids (TS)
80S volatile solids (VS) of TS
40% lbs/ft3 raw manure
1 ft3 Manure/horse/day
5 ft3 biogas/lb. of VS input
Biogas = 50% CH4 50% C02
100% recovery of waste
Tank volume = 10,000 gal = 1,334 ft3
Per Tank Calculations
ft manure added:
(gal)
(Ibs)
ft3 H20 added:
(gal)
ft3 total slurry:
(gal)
Ibs total solids
Ibs volatile solids
ft3biogas
ft3 CH4
btu value
# of horses supported
17.52
131.05
700.80
26.28
196.57
43.80
327.62
140.16
112.13
560.64
270.32
2.8xl04
18
Operationally, there were only minor design problems. First, inclusion
of too much bedding led to scum buildup initially. This scum blanket col-
lapsed a set of heating coils when the system was drained down. Rebracing
the heating system and eliminating much bedding solved the problem. Second-
ly, the larger pipes used butterfly valves which restricted flow and led to
plugging from bedding and particulates. Finally, the multiple tank system
led to time consuming mixing schedules and added complexity to the plumbing
system.
20
-------
TABLE 2. BROCKMAN STABLES—SUMMARY OF RESULTS
SAMPLE
PARAMETER LOCATION
% M of Total Sample
% TS of Total Sample
% VS of TS
% FS of TS
%VS of Total Sample £
% FS of Total Sample •?
cu
VS in lbs/ft3 >,
(v, VS in gm/1 "^
% VS destroyed
Total Alkalinity mg/1
Volatile Acids mg/1
pH: Filtered
pH: Nonfiltered
Total Nitrogen mg/1
Ammonia - N mg/1
Total Organic Carbon
Chemical Oxygen Demand mg/1
Gas Analysis: 49% CH4> 40% CO
MIXING
TANK
95.03
4.97
88.61
11.39
4.40
.57
.11
1.81
2040
1570
6.4
6.5
556
152
20,000
36,000
,, 11% 02 +
A- 2
98.79
1.21
73.90
26.10
.90
.32
.02
.37
79.55
1600
524
7.0
7.1
304
123
5700
N2
B Series
98,33
1.67
81.26
18.74
1.35
.31
.03
.56
1570
334
7.1
7.2
365
133
14,000
C Series
99.36
.64
64.94
35.06
.41
.22
.01
.17
90.68
1530
363
7.0
7.3
251
134
4100
D Series
99.59
.41
58.73
41.27
.24
.17
.01
.10
94.55
1490
343
6.7
7.1
210
133
3700
E Series
99.72
.28
47.20
52.80
.13
.15
.003
.055
97.05
1410
272
7.2
7.3
176
133
100
Key: M = Moisture
TS = Total Solids
VS = Volitile Solids
FS
CH4
\j\J f)
Fixed Solids
Methane
Carbon Dioxide
03 - Oxygen
N? = Nitrogen
mg/1 = Milligrams/liter
-------
Green Bay Pilot Plant
The Green Bay pilot plant was designed, constructed and operated by
undergraduate students at the University of Wisconsin-Green Bay, with a
National Science Foundation Student Originated Studies grant from the
National Science Foundation (NSF) and some University monies. Total cost
of the "3 cow" mesophillic (35°C) unit, including equipment, building, and
lagoon, was approximately $20,000. The cost of the system includes labor
and materials for the laboratory work to check design and labor for constr-
uction.
The system configurations for manure handling, gas utilization, and
heating are shown in Figures 14 and 15. The design uses a single vertical
concrete tank whose lid is flush with ground level. A small building over
the tank provides thermal insulation and access to the tank and other basic
components of the system. A single large pump periodically circulates the
slurry from the tank through various pipes above the tank, and back into the
tank, for the dual purpose of agitation and heating. Agitation is accom-
plished by circulation, while heating is achieved through an external heat
exchanger. It is heated from a thirty gallon water heater and circulated
to the heat exchanger by a small pump. Thermostats inside the digester are
used to provide feedback to the heating and pumping system in order to main-
tain a 350C (95°F) temperature within the tank.
Aside from agitation and heating, the main pump serves two other
functions: 1) It pumps fresh input manure from a mixing hopper, through
the heat exchanger into the tank; and 2) it pumps spent slurry out of the
tank into the lagoon. Thus, the main pump takes part in circulation/agita-
tion, heating, and input/output. Table 3 explains the different modes of
pump operation.
The hot water system serves a two-fold function: 1) to supply heat to
the heat exchangers, as previously described; and 2) to supply hot water
to the mixing hopper for the purposes of diluting and heating the manure.
Heating of the incoming manure is an especially critical operation, since
the manure will have been transported from the barn to the digester without
any protection from the severe climate conditions of winter.
The biogas passes through a hydrogen sulfide scrubber (iron filings in-
side of a five gallon drum) and a residential type gas meter, and into the
hot water heater for consumption. Any excess gas which is produced flows
outside of the building to a storage system, which consisted of a series
of tire inner tubes. A natural gas system is also tied into the hot water
heater to supply fuel for start-up and for back-up (primarily in winter, as
the energy balance in Table 4 indicates).
The system is maintained at an operating pressure of ten inches of
water by use of a pressure relief valve (a five gallon drum with ten
inches of water inside) which is vented to the outside of the building. The
maximum allowable pressure in the digester tank is 36 inches of water.
22
-------
modified
submersible
sewage pump
Input hopper
Figure 14. Manure handling and heating systems, Green Bay
to manometer -*-
-------
TABLE 3. MODES OF PUMP OPERATION—GREEN BAY SYSTEM
Valve Name
Lagoon Input
Digester Input
Hopper Input
Digester Output
Hopper Output
Abbreviation
LI
DI
HI
DO
HO
Location
Pump-Lagoon
Heat Exchange - Digester
Heat Exchange - Hopper
Digester - Pump
Hopper - Pump
Mode
Name
Warm-up
Input
Night
Inspect
Dump
Pumpout
Valve Positions
JLL2! JH52. HO
X X 0 X 0
X 0 X X 0
X 0 X 0 X
X X 0 0 X
0 X X X 0
0 X X 0 X
Effective
Circuit
Hopper-Hopper
Hopper-Digester
Digester-Digester
Digester-Hopper
Hopper-Lagoon
Digester-Lagoon
Purpose
Warm manure
Load digester
Warm digester
Unload digester
Dump slurry
Empty digester
How
Often
Daily
Daily
All night
Daily
Daily
X = valves closed
0 = valves open
The Digester Tank is a precast, reinforced, ASTM Class III sewer pipe,
with an inside diameter of six feet; and height of seven feet. The base is
a steel reinforced plate cast on. A precast lid with holes for observation
ports, slurry input, slurry output, gas outlet, and manhole, was gasked in
place after installation. The tank was sized based on monies available and
the ability to obtain materials within the time frame alloted for construc-
tion.
The system is actually too small to be self sustaining because of
heating requirements. Had the system been well insulated, breakeven in
energy might have been achieved. Tables 4 and 5 give power and energy re-
quirements and gas produced based on both field testing and design calcul-
ations. The system actually performed better than calculations indicated.
Table 6 summarizes loading rates and gas production during some typical
feeding periods and also gives basic chemical performance data.
24
-------
TABLE 4. ENERGY BALANCE - GREEN BAY DIGESTER
Total Energy Input
Month
3/76
4/76
5/76
6/76
7/76
8/76
9/76
10/76
11/76
12/76
1/77
2/77
3/77
4/77
5/77
6/77
7/77
Monthly
Sum ,
Btu x 10°
6.7
5.8
5.1
4.0
3.8
3.4
4.8
5.7
6.8
7.5
8.1
6.2
4.3
5.5
4.0
3.8
2.9
Daily
Mean fi
Btu x 10
.29
.20
.17
.13
.12
.11
.16
.18
.23
.24
.26
.22
.13
.18
.13
.13
.09
Total Energy Produced
Monthly
Sum c
Btu x 10
.120
.006
.130
.130
.073
.110
.170
.270
.420
.460
.620
.530
.410
.410
.850
1.400
1.600
Daily
Mean fi
Btu x 10°
.0050
.0002
.0042
.0043
.0023
.0034
.0006
.0086
.0140
.0150
.0200
.0190
.0130
.0140
.0270
.0480
.0510
Energy Balance
Monthly
Sum fi
Btu x 10
- 6.6
- 5.8
- 5.0
- 3.9
- 3.7
- 3.3
- 4.6
- 5.4
- 6.4
- 7.0
- 7.5
- 5.7
- 3.9
- 5.1
- 3.1
- 2.4
- 1.3
Daily
Mean fi
Btu x 10°
- .29
- .20
- .17
- .13
- .12
- .11
- .15
- .17
- .21
- .23
- .24
- .20
- .12
- .17
- .10
- .08
- .04
-------
TABLE 5. PREDICTED ENERGY BALANCE (BTU/DAY)
INPUT
Electricity
Heat
Total
OUTPUT
Gas Generated
Surplus/Deficit
YEARLY MEAN
1.52 x loj
7.87 x 10
9.39 x 104
1.19 x 10?
2.51 x 10
JANUARY
2.04 x 10?
1.16 x 10°
1.36 x 105
1.19 x 10?
-1.70 x 10
DESIGN LOW
2.04 x 10?-
1.31 x 10°
1.51 x 105
1.19 x 10?
-3.20 x 10
Rather than deal with the actual switching from biogas to propane
backup, the system was operated entirely on propane and gas production
noted. This was largely due to lack of time and funds to produce the dual
fuel system and the need for test data reliability.
Only two major problems were encountered. First, the pump which moved
the manures needed several modifications largely because budget constraint
did not allow for optimum pump selection. Secondly, abnormally high heat
losses were encountered because the tank was not insulated. Other problems
were minor and in more of a maintenance category.
Experience gained on this system and other full scale systems indicated
that this operation could have been run on a higher solids loading and thus
achieved the equivalent of a 6 cow operation. This would have been a net
energy producing system. The pump which was selected (Midland AFP submer-
sible pump) was unable to handle the high loadings. Funding was not
available to replace the pump, so the system was dismantled after exper-
iments under this contract were terminated.
26
-------
TABLF 6. LOADING RATES, GAS PRODUCTION, AMD CHEMICAL PARAMETERS
GREEN BAY PILOT PLANT
Load
Rate
Ibs/ft^5
.005
.010
.021
.031
.039
.043
.078
Gas
Produced i
ft3/day i
6.8
11.3
29.8
39.7
37.8
54.5
96.0
INFLUENT
% TS
4.99
4.76
7.31
8.48
9.65
7.11
7.42
% vs
of TS
80.20
84.00
87.60
87.10
88.30
82.50
75.60
Volatile
Acids
mg/1
2707
2466
2719
2925
3563
4380
5267
Alkal ,
mg/1
5233
4898
6008
7160
6663
6763
9337
pH
7.1
7.2
7.2
7.2
7.2
7.3
7.2
COD
48,333
67,400
96,667
97,500
103,333
84,500
100,667
EFFLUENT
% TS
3.55
2.22
4.20
4.82
5.01
4.04
5.17
% VS
of TS
64.86
68,27
80.57
81.87
83.09
78.42
78.60
Volatile
Acids
mg/1
1200
980
1587
1840
1390
1210
1890
Alkal
mg/1
5410
5964
6911
7085
6980
5778
8000
pH
7.45
7.88
7.75
8.30
8.05
7.60
7.80
COD
29,000
24,200
51,000
63,000
64,300
60,250
65,000
Abbreviations: TS - Total Solids
VS - Volatile Solids
mg/1 - milligrams/liter
COD - Chemical Oxygen Demand
Alkal. - Alkalinity
-------
Ludington (Cluster) MI System
This is the first system constructed by Agricultural Energy of Luding-
ton, Michigan (now Energy Harvest, Chicago, Illinois).
The Ludington unit was sized to utilize the waste from 350 head of
feeder cattle. The lot is open and collects rain water and snow. It
freezes in winter and dries hard in summer so that weather conditions
affect it adversely.
3 3
The manure is collected in a 136 M (4800 ft ) concrete tank. An
automatic system controlled with a timer clock fills and discharges a 300
litre (80 gallon) tank of 8 to 10 percent solids fed into the digester.
The system had a pneumatic tank which metered materials from the mixing pit
to the fermenter by first vacuuming up a volume and then using pressurized
biogas to force the material into the fermenter. The operation is schema-
tically depicted in Figure 16.
YACUU/yA FUtAP PRESSURIZED
-e
L D
VRRESU
MANURE
TO FERMENTATION
VALVES TANK
Figure 16. Schematic of fermentation tank feeding system, Ludington
28
-------
The basic innovation of this system was the use of a flexible membrane
for a fermentation tank and gas storage system. The original bag, manu-
factured from a fabric reinforced plastic and earth supported, suffered
from faulty seams. This unit was subsequently replaced by a wooden frame,
membrane lined unit which gave better structural control and insulation from
the ground heat loss. (Figures 9 and 20)
3
The digester is approximately 191 cubic meters (6750 ft ) liquid
volume. Level is controlled by a level probe and a diaphragm pump. The 3
bag expands when filled with gas and collects approximately 187 M-3 (6600 ft )
of biogas. The biogas from this bubble is compressed in a small vane pump
to 10 psi and is fed to an engine coupled with a 30 kw gas driven generator.
The excess heat from the generator is used to keep the fermenter at temper-
ature (35°C).
Aside from the structural problems in the original bag fabrication,
there were several major design problems. First, no provision was made to
keep the feedlot from freezing in winter. This led to digester shut down
in very cold months. Next, solenoid valves used in the automatic feed
system were unreliable leading to uneven feeding. Problems with the vacuum
pump also plagued the feeding operation.
The feedlot never had the full contingent of animals, thus the system
has never operated under full design load for extended periods of time to
test the efficiency of waste stabilization.
Rice Lake System
This unit was a spinoff from the Ludington design.
The Rice Lake unit is designed to use the manure from 110 head of
dairy cows in confinement with floor scrapers. From the scraper pit the
manure is pumped into the fermentation tank.
The original plan for the tank called for a half culvert to be placed
over the original bag digester so that the entire system could be earth
covered for insulation. Construction dictated that a complete culvert was
necessary. The final design led to a culvert with end plates and no liner.
Rather than use a one piece culvert, the 3.66 M (12 ft.) diameter by 14.6M
(48 ft) long tube was constructed from 24 rings each .6 metres (2 ft.) wide.
Each ring was made from three sections. The 72 pieces and the two end
plates together with flexible gaskets comprised the fermenter tank, along
with 8000 bolts which hold the system together. See Figure 21. The con-
cept behind the sectional nature was to test the idea of modularity for
flexibility in tank design, but the labor intensive construction of this
design makes it cost prohibitive.
The tank is buried under three feet of earth and has an additional
inch of styrofoam around it. Unfortunately, the water table in the area is
high and no provisions were made to tile away any moisture build up which
29
-------
could yield a high heat loss and possible floatation problems if the system
is drained at an inopportune time. The system is heated by coils running
longitudinally along the bottom. These pipes use hot water from the engine/
generator set. Currently there is no monitor on the flow rate and tempera-
ture of the water, so no accurate data as to tank heat loss can be obtained.
The control house (Figure 17) has undergone two evolutions as the
system was redesigned. The figure shows the current configuration. The
idea in compartmentalizing is to separate the electrical and engine controls
from the section where gas enters the structure to reduce the possibility
of explosion and fire. Also, the work area is isolated from the generator
system to reduce the effects of noise on operators. A fan installed in the
wall between the engine and gas section pulls fresh air through the work
area, through the engine area and to the gas area to keep a positive pres-
sure in the gas section. Gas detectors are located in the piping and en-
gine areas. They were designed to sound an alarm at 25 percent LEL and
shut the system down at 50 percent LEL. All motors are hydraulic to reduce
the possibility of explosion. Experience in installation and maintenance
seems to indicate that electrical systems are easier to -install, operate,
and maintain as well as being cleaner.
oumo room
k*pt at
poaitiv*
pr»aaurt
f«rm»nt»t!on tank
•outh and
~cs~
SI HI gaa compraaaor
affluent pump —
Idlaphragml
•cycle pump
L>'
O
gaa flltars
oo
hydraulic controls
and
tamo.
atoctrlcal control*
control room
id valvaa ^^»"L
p. gaugvs <^
^^C
hydraulic fluid
storaq*
•ngm«
garwrator
work
tnncn
watar
pump
angina room
radiator
Figure 17. Rice Lake control house.
30
-------
At the termination of this work, a small microprocessor for data acquis-
itions and control arrived and is being installed for monitoring and to as-
certain problems in using microprocessors over hard wired programmers and
timers.
The current engine generator is a 12 kw Waukesha engine coupled to a
Kato generator set. Power is supplied to a portion of the house.
A schematic of the piping for the dairy operation is shown in Figure 23.
The system was designed to operate on a 15 day hydraulic retention time (HRT)
and possibly as short as 10 day HRT. To prevent a wash out of the microbial
population on this short a retention time, a refeed pipe has been added.
While the effluent pipe removes material from the bottom of the fermentation
tank, another pipe, at the top of the liquid level, cycles spent materials
back into the tank with the new influent. This acts as a seed material,
preventing washout and enhancing the fermenter operation.
While the unit is set up to automatically feed on 24 hour basis, pro-
blems with materials handling has prevented the system from operating
in this mode and the fermenter has been batch fed approximately three times
a week with no adverse affects on systems operation.
pump
conveyor
* open ended turkey
manure collection
to fermenter
Figure 18. Temporary turkey manure collection, Rice Lake
Prior to operating as on dairy waste, the system was used as a pilot
facility for evaluation of turkey waste from a local feeder operation. The
schematic for the temporary operation is shown in Figure 18. Several pro-
blems appeared due to the temporary nature of the materials handling set up.
Most of these were due to improper weatherization since the original intention
of the experiment did not include winter operation.
Several problems arose in the materials handling systems operation which
were due to the nature of the turkey waste. First, turkey waste is self
composting like horse manure. This led to heating which had two effects,
bridging in the storage hopper and loss of nutritive value leading to low
gas production. Thus, turkey waste must be fresh for maximum benefit and ease
of handling. Secondly, the material was removed from unpaved pens. This
31
-------
led to the inclusion of dirt and other debris which clogged lines. Finally,
the turkeys were bedded with rice hulls which do not digest well and tend to
create plugging problems. Analysis of the rice hulls is shown in Table 7.
TABLE 7. CHEMICAL ANAYLSIS*
OF RICE HULLS USED FOR TURKEY BEDDING
Crude Protein%
Fat%
Fiber %
Ash %
NFE%
Water%
Si02%
Na20%
CaO%
MgO%
FeO%
P00C%
C 0
so3%
2.4
.5
41.0
18.3
21.9
10.5
1.60
1.58
1.01
1.96
.54
1.86
.92
Analysis supplied by J.B. Hunt Company,
Springdale, Arkansas
Table 8 is a typical analysis of the turkey waste in the system. The influ-
ent is equal portions of water, turkey waste solids, and recycled fluid.
Note from the recycle fluid and the effluent that there is stratification in
the fermentation tank. The analysis of the solid turkey waste is also shown
in Table 8.
32
-------
TABLE 8. ANALYSIS OF TURKEY WASTE IN RICE LAKE SYSTEM
CO
OJ
Diluted
Influent
Undiluted
Influent
Recycle Fluid
Effluent
Parameter
% M of Total Sample
% TS of Total Sample
% VS of TS
% FS of TS
% VS of Total Sample
% FS of Total Sample
Total Alkalinity
Volatile Acids
PH: Non-Filtered
PH: Filtered
Total Kjeldahl Nitrogen
Ammonia Nitrogen (as N)
Organic Nitrogen (as N)
Total Organic Carbon
Chemical Oxygen Demand
unit
0)
i
-Q
mg/1
mg/1
-
-
mg/1
mg/1
mg/1
mg/1
mg/1
analysis
96.30
3.70
69.65
30.35
2.58
1.12
8650
3530
7.3
7.4
3070
1690
1380
81,000
-4
0.16
0.16
0.80
0.80
0.14
0.02
41
60
-
-
1
6
5
1032
analysis
26.83
73.17
55.03
44.97
40.26
32.91
-
-
-
-
2.03°^
-
-
-
-
4
1.32
1.32
0.66
0.66
0.48
1.01
-
-
-
-
0.13
-
-
-
-
analysis
99.09
0.91
54.82
45.18
0.50
0.41
7620
1500
7.3
7.5
2020
1650
370
39,000
4
0.02
0.02
1.32
1.32
0.01
0.02
1
29
-
-
4
4
4
1220
analysis
88.16
11.84
57.77
42.33
6.84
5.00
8010
1820
7.5
7.5
4140
2060
2080
79,000
-A
0.49
0.49
1.60
1.60
0.43
0.17
45
0
-
-
4
1
3
1548
M = Moisture
TS = Total Solids
VS = Volatile Solids
FS = Fixed Solids
- Standard Deviation
-------
SECTION 5
TECHNICAL COMPONENTS
One can define three basic components of an anaerobic digester: the
influent tank which serves as a temporary storage of the material for fermen-
tation, the fermentation tank which stabilizes the residues and produces the
gas, and the lagoon for effluent storage. This particular study looked at
the problems in the design of these major components, as well as materials
handling, gas utilization, and power systems for full scale digesters.
INFLUENT TANK
The tank in the Ludington system was an underground concrete pit which
received the animal wastes from an open feedlot as they were introduced via
a scraper. (Figure 16) The Rice Lake system utilized a similar arrangement
which was fed by a barn scraper which cleaned a free stall barn where the
dairy herd was confined. (Figure 19) Several crucial design mistakes came
to light in this study.
free- stall barn
concrete
holding tank
I below grade!
Figure 19. Rice Lake influent tank,
34
-------
1) Both tanks lack insulation. Thus, the animal wastes soon reach
ground ambient temperature or lower. This reduces biological activity and
serves as a thermal schock to the fermenter when new material is introduced
from the influent tank. In fact, the major need for heating lies specifically
in bringing the animal wastes up to temperature.
The Rice Lake system was equipped with the capability for heating the
influent tank with surplus heat from the engine/generator, but the system
was never made operational. This is a large loss and causes a significant
dent in the engineering economics of the system.
2) Sizing of the influent tank is important both in initial cost and
in operation. The Rice Lake (RL) pit existed at the projects inception and
was designed to store animal waste for several weeks before spreading on
the fields. The large storage capacity was an advantage when the fermenter
was down for maintenance, repairs, or reconstruction and clearly pointed to
the need for reserve capacity and/or bypass from the influent tank to the
lagoon. On the other hand, the large size created pumping problems because
the manure tended to stratify. The solids settled to the bottom decreasing
the pumping capacity from the influent storage to the fermentation tank.
Neither tank had a built-in bypass for direct pumping to the lagoon
should operational conditions warrant. Based on the operational experience,
this feature is desirable, especially if only a short term capacity is
designed into the influent storage.
FERMENTATION TANK
Both Ludington and Rice Lake operations are horizontal, pseudo-plug
flow systems, in that the tank design did not prevent thermal mixing. Also,
there was no way to assure that the Hydraulic Retention Time (HRT) and the
Solids Retention Time (SRT) coincided. The Rice Lake system operated on a
very short (10-15 days) HRT and, thus, refed part of the effluent with the
influent to keep the bacterial population up (see systems operations).
The plug flow, or horizontal, mode of operation seems best from a
structural perspective and low installation costs. It appears that trouble
free operation with low mixing costs are possible with horizontal systems.
Neither of these systems, nor the ones under study at Cornell are mixed in
normal operation yet have comparable gas yields to completely mixed systems.
Thus, a costly installation and maintenance item, mixers, are eliminated.
The dimensions of the tank follow conventional wisdom for plug flow
units (length = 5x width). Experience at Cornell seems to indicate, though,
that tank shape is of little importance to efficient systems operation of
anaerobic fermenters.
35
-------
All tanks were insulated with sheet styrofoam to below the frost line;
the Rice Lake system was covered with styrofoam and three feet of earth. No
special provisions were made to provide drainage to keep moisture from the
tanks. Small changes in soil moisture can create large changes in heat loss,
so it appears that good drainage around the structure is imperative to mini-
mize heat losses. (28) One researcher has gone so far as to advocate not
burying tanks because of heat loss. (61) However, no data has been developed
to support the contention that a tank free of the earth is cheaper to build,
operate, or maintain.
Both the Rice Lake and Ludington systems had internal heat exchangers
running the length of the fermenter. They also had gas pipes placed so that
mixing, if desired, could be accomplished by introducing biogas under pressure.
(Figures 20 and 21)
SLURRY PUMPED INTO TANK. .
EFFLUENT DISCHARGED INTO LAGOON
BIOGAS SIPHONED OFF FOR USE.
WOOD FRAME
SUPPORTS
FLEXIBLE
HEATING PIPE FOR TEMPERATURE CONTROL
BIOGAS RE-CVCLED INTO SLURRY
PROVIDES MIXING ACTION (OPTION)
ELECTRICAL ENERGY
MCTMOO OF SEAL-
ING TOP 4 BOTTOM
PORTIONS OF BAG
Figure 20. Ludinqton fermentation tank.
The Ludington system appeared in two versions during the study (Shown
earlier in Figures 8 and 9). Both employed a fabric reinforced tank liner
manufactured by Staff Industries. The upper and lower segments of the tank
liner were field assembled. The first flexible membrane system was earth
supported while the second model also had a plywood frame for support. The
Ludington system floated styrofoam sheets on the effluent to protect against
heat loss. Gaps in the sheets permitted gas to escape the solution and to
fill the cover which served as storage.
36
-------
SLURRY PUMPED INTO TANK.
DISCHARGED INTO LAGOON
BlOOAS SIPHONCP OFF FO*. USB
FLUID RB-CYCL£D TO INTAKE
HEATING PIPE FOR
TEMPERATURE COMTOX.
BlOGAS WKYCtCO INTO
SLURRY PROVIDES MIYINC
ACTION (OPTION)
DETAILS OF TANK CONSTRUCTION
EACH 6EC~nOM->3PlECES
ALL JOINTS CASKETEP OF Yt CORRUCA1ED METAt.
Figure 21. Rice Lake fermentation tank.
The Rice Lake system was formed from segments of galvanized metal field
bolted together (Figure 21). The labor involved, however, precludes this
method of construction. The company, Agricultural Energy, feels confident
that a metal tank will survive the mechanical and chemical abuse from the
fermentation process. Their assumptions seem contrary to the published
literature (29) but are partially supported by the experiences of Biogas of
Colorado (30) who ran a portable unit for one summer. In the latter system
no corrosion was found in the anaerobic fermentation tank though there was
some deterioration in the metal influent tank which was exposed to the air.
Experience in India (3) seems to indicate that corrosion could be a problem
over the long run and that interior protective coatings should be applied at
regular intervals. The Rice Lake tank has -not been opened for visual inspec-
tion since start-up, thus, no data is yet available on actual field condition.
For low costs, the preferance is for the flexible membrane tankage
because it is easily installed, maintained, and also provides gas storage.
Another alternative is wooden tanks which have been used for fermentation
vessels in the past. They stand up well to anaerobic environments and are
used where harsh chemical environments would destroy steel. Both traditional
37
-------
stave structures or plywood could be used. (31) Metal tanks of mild steel
require more maintenance and are expensive, as is concrete. Galvanized tanks
such as the Rice Lake operation are worth watching. No data on longevity or
durability is yet available.
LAGOON SYSTEM
Both the Rice Lake and Ludington systems have standard farm lagoons for
effluent storage. Our studies, those at the University of Minnesota (36),
and data provided by Environetics (22, 23) leads us to conclude that anaero-
bic lagoons should be completely covered to prevent nitrogen loss from the
effluent. This cover would pay for itself in fertilizer savings. Lining the
entire lagoon with a membrane (creating a larger "fermenter") would allow
easier pumping of the liquid when the lagoon is emptied and also prevent
loss of solids and liquids which would be caught in a soil lined system.
MATERIALS HANDLING SYSTEMS
Both systems were designed with the idea that the daily influent would
be fed at a uniform rate throughout the day (24 hours). The intent was to
prevent shock loading of the fermenter with large slugs of cold, unfermented
material. Both systems suffer from intermittent clogging, failure of the
automatic or manual operation done on daily or twice daily operation. While
the principles behind the feeding systems are sound, they are unnecessary if
one judges by the operation of other systems. Shock loading would be pre-
vented if the influent were kept warm or preheated.
The Ludington system operates on a vacuum principle. (Figure 16)
A small vacuum pump evacuates the tank. A valve opens permitting manure to
fill the tank. Biogas then blows the manure into the fermenter. The system
is controlled by a timer adjustment to meet the needs of the daily loadings.
Because there is no pump in the system there is no opportunity
to chop the influent to break up large particles which may cause plugging.
A number of inline mascerators have been tried with little success. The
largest problem appears to be animal hairs which wrap around mascerator or
pump impellers bringing the operation to a halt.
The Rice Lake system is unique in that it is the first system to use
hydraulics to control valves and motors. Hydraulics were chosen: 1) to
test the concept of using hydraulics; 2) hydraulic motors are smaller and
more compact; 3) the hydraulic pump needed to run motors adapts readily
to engine generator set; 4) use of hydraulics reduces the potential
for an explosion. The hydraulics are run from a pump which is driven by
the engine/generator. Hydraulics then run off idle power similar to an air-
conditioning compressor on an automobile. The biggest problem here is there
are no manual over rides on the valves and no way to activate the pumps
should the generator system be down. Because of the late installation of
a satisfactory engine from which to operate the hydraulic pump, do data is
38
-------
available under this contract, but it has been operating satisfactorily.
Figure 18 schematically indicates how the Rice Lake unit was operated as
a test bed for turkey wastes. From a materials handling perspective, the
turkey waste proved to be the most burdensome because the fresh waste tended
to heat up, cake, and then bridge in the temporary hopper. Since the material
was cleared from unpaved houses, dirt, rocks, and similar debris were also
picked up. This caused problems in movement and pumping, because the debris
materials would not pass through pumps and pipes. Additional problems occur-
red because the addition of water caused tne rice hulls to settle out and
create plugs in low piping points. The system was not winterized because of
the temporary nature of the turkey test. Thus, freezing problems developed
during the winter of 1976 and 1977, which was unusually cold.
The basic idea was to work with an influent of 8 percent to 12 percent
total solids. This was accomplished by first loading a portion of raw effluent,
adding a measure of spent effluent as seed and then adding water. In the
turkey operation, these materials were mixed to assure uniformity and ease of
pumping. This was pumped into the fermenter after an equal volume of spent
materials had been removed.
Figures 22 and 23 show how the dairy handling was originally designed
and how it operates now. The original dairy system was to have operated
similar to the Ludington operation (a portion of spent effluent was to be
pumped into a tank along with a charge of influent and water. The mixture
was then to have been introduced to the fermenter via pressurized gas).
Prior to the startup of the dairy operation, however, the system was re-des-
igned eliminating the tank and operating as follows. After pumping effluent
from the bottom of the fermenter to the lagoon, a portion of effluent from
the top of the tank is diverted into the influent line joining manure pumped
from the influent pit. The operation is controlled by a timer which regulates
valves and controls quantity of input. Previously, quantity had been
controlled by level sensors in the tanks. These were high voltage, low
current sensors similar to those used in farm lines.
PRETSUR1Z1ED BlOGAS
BARN,
RJOOR
SCRAPER
LA60DKJ
MANURE
HOLDIKIQ-
PlT
Figure 22. Schematic of original manure handling system, Rice Lake
39
-------
WATER
t>^-——x
..M^J
BARN,
FLOOR
SCRAFER
'CHOPPER
PUMP
PUMP
LA500N
EFTUJEWT
PMKAP
Figure 23. Schematic of redesigned manure handling system, Rice Lake
About the time revisions to the Rice Lake system were made to the mater-
ials system, Cornell indicated that a fermenter could operate at high solids
concentration of manures (undiluted). This was the original intent of the
revised Rice Lake system but the installed pump could not handle the head
from the influent tank. This required that the operation be modified so
that the operator could add water to the influent tank prior to pumping. The
pump was then used in a mix or recirculating mode prior to pumping the mix-
ture to the fermenter. At the present time, this is forcing the operation to
be in semi-automatic mode.
Several factors became clear in observing both the Rice Lake and
Ludington operations. First, piping and pumps need to be oversized to take
into account the peculiar properties of the manures. Second, piping runs
with angular turns create obstruction problems, especially if bedding is
present. Thus, careful attention must be given to layout and inclusion of
sufficient cleanout ports to allow total access to all parts of the line.
Third, manual operation, at least for small operations, appears to be the
most efficient way to go. Finally, the pumping system should have a means of
being operated separately from an on sight generator set (more will be dis-
cussed about this in the gas utilization, technology and economics sections.)
GAS UTILIZATION SYSTEM
Both the Rice Lake and Ludington systems run an engine/generator set from
the biogas. Additionally, the gas can be burned directly on the Rice Lake
farm, though this is not being done currently.
Since the Rice Lake system has a rigid fermenter, gas storage was
40
-------
needed. This was accomplished by floating a flexible membrane on the lagoon
(Figure 24).
Several problems cropped up with this gas storage system. First, the
flexible membrane froze in the lagoon in winter and spring thaw caused the
ice to give unevenly allowing the bubble to tilt and release the gas. Also,
the high moisture of the gas may have caused icing problems in the gas line,
though this has not been verified. It should be noted that trace amount of
water can form hydrates with methane. These hydrates have a freezing point
well above the freezing point of water. This phenomena is well recognized
in the natural gas industry; pumping methane over long distances without
dehydrating it could cause freeze-up problems.
In both systems, concern was expressed for the quality of the gas
(approximately 600 Btu/cu feet) because of storage and efficiency issues.
Experience at the Ludington facility verified that lime water (limestone in
cold water) is a poor scrubbing medium, even under pressure. No signifi-
cant increase in quality was observed.
On the Rice Lake system the gas passes through a water trap and a hydro-
gen sulfide trap (copper wool in the form of pot scrubbers from the local
supermarket) before passing through a standard §as meter. The gas for the
engine manufacturers. When the system operated on turkey waste, the presence
of hydrogen sulfide was indicated by the formation of copper sulfide on the
piping; though it was not noticeable by odor.
• '.''.•.•. '. .•..."• • •!-j
••.'••••:•'. ••.<.-.*.::.1
flexible membrane
low pressure storage
Figure 24. Gas and effluent storage system.
41
-------
The Rice Lake system included a unique method for gas compression.
(Figures 25 and 26) The core of the system was a Sihi ring seal compressor.
The theory was that CCL could be dissolved in the water which served as a
seal compressor. After compression, both the water and gas were then sent
to a storage tank where the biogas separated and the water overflowed a
check valve into the lagoon. While the system worked well for low pressure
storage for the generator set and for other uses, the gas enrichment via
carbon dioxide removal proved to be insignificant. In fact, air trapped in
the well water was released into the gas mixture—a potential explosion pro-
blem. (Table 9)
The presence of a 10 percent concentration of carbon dioxide in the gas
is of concern only from the economic point because it is not usable in the
gas mixture and it requires storage volume. The quantity of C02 is too small
and the systems too simple to warrant a sophisticated amine scrubbing system
used in industrial applications. However, some chemical alternatives do
exist, and could significantly enhance the quality of the gas, as well as
economics of the system. This is discussed in the Biochemical section.
ou^S
aoz
HiO
VWUKESHA
SUfFLfcD
FILTERS
PQKO1RC
STORAGE.
TANK
Figure 25. Gas scrubbing and compression system.
CH,
ro ENP use
I
^ RW
-------
TABLE 9. AVERAGE GAS SAMPLE ANALYSES
RICE LAKE SYSTEM
Low Pressure System High Pressure System
/o Ln^i
% C02
% 02 + N2
58.23
36.96
4.82
58.27
29.51
12.22
ENGINE GENERATOR SYSTEM
The Rice Lake and Ludington systems utilize the entire gas production of
the fermenters to generate electricity via an engine generator unit. From
both an engineering and economic perspective, this is the most inefficient
means of utilizing the biogas. Direct combustion is the most efficient. The
economics are particularly bad when the engine generator unit runs 24 hours
a day, even under full load. This is almost a necessity in cold weather
since the cooling water is needed to provide heat to the fermenter. When
the systems are on automatic, the engine provides power to the material
handling system. Use of a large generator unit for providing waste heat and,
intermittently, pumping power is not an efficient use of resources.
Figure 27 shows the schematic of the Rice Lake power system. The
original design included a four-cylinder Ford industrial engine. Special
pistons were designed to increase the compression ratio to the level suggest-
ed by a Cornell study for optimum performance. These pistons were never
installed. Instead, a Ford spark ignited diesel was substituted. This
ran satisfactorily, however, no actual power curves were generated since the
digester never produced gas for an extended period of time, due to research
and redesign. This engine/generator set was replaced by a Waukesha engine/
generator unit specifically designed to run on biogas and/or propane.
Propane was used as a backup in case the fermenter was not working or producing
sufficient biogas.
The engine/generator unit did not supply enough electricity to meet
the farm/home requirements, thus, part of the house load, within the 12 kw
capacity, was transferred to the unit. The generator output was kept separate
from the utility lines so there was no need to provide a synching mechanism
to phase the two power systems. Clocks on the biogas generator circuit
indicated that good frequency control was hard to maintain with the Ford
engines. This was improved by a close tolerance governor on the Waukesha
unit, though still not completely satisfactory for delicate control mechan-
isms such as turntables for stereo units. Separate frequency conditioning
units may prove essential where good control is needed, or where motors (such
as compressors) could be subject to burnout from poor regulation in frequency
or power.
Both the Rice Lake and Ludington operations have heat exchange systems
43
-------
ELECTRIC
UTILTTV
IZKIIOWATTS
Z.4OV/OCT
SIN^LE PHASE
WITH
MECHANICAL.
QOVERWOR
KWER
SWITCH
I. UTILITY
FCOTOKOF
HOUSEHOLD
LCAO
Figure 27. Schematic of Rice Lake power system.
between the engine and the fermenter. (Figures 28 and 29) The design
includes two pumps and a thermostatic control. When the fermenter is warm
enough, the thermostatic control directs excess heat to a radiator/fan unit.
Design consideration must be given to p*ipe sizing so that the engine water
pump does not exceed its design capacity due to the additional heat created
in the heat exchanger network.
Several operational characteristics became clear during the course of
this study. First, continual running of the generator unit produced minimal
energy at maximum cost, since waste heat was used only for fermenter temper-
atures maintenance on both the Rice Lake and Ludington units. Greater
utilization of waste heat would improve energy costs somewhat. (See the
Economics section)
Next, most industrial engines have a finite time between overhauls. For
example, a standard automotive engine can get about 3,000 hours between
major overhauls; an industrial engine manufacturer may recommend about 60,000
hours between major overhauls. This implies that under continuous operation,
one major overhaul would be necessary every six years. Routine maintenanence
and normal breakdowns must also be anticipated. Routine engine maintenance
and overhaul costs on the Rice Lake system are estimated to be $.012/kwh
based on full load utilization. Thus, this study indicates that an anaerobic
digester should be integrated into the farm operation so that most of the
-------
gas is used for its thermal value directly, and a minimum amount of gas is
utilized for power generation.
CONTROL SYSTEMS
The control for both the Rice Lake and Ludington systems were designed
to be automatic. They consisted of electromechanical components such as
sensors, timers, and relays. The timers were set to start the systems cycling
"x" times/hour. During that cycle, level sensors were used to determine the
status of various tanks and regulate valves which controlled water, effluent
or influent. The timers and level sensors also activated the proper pumps.
The controls operated with minor problems, but there were no alarms to
indicate when malfunctions, such as plugging up or bridging in the hoppers,
occurred in the materials handling system. An alarm system would have been
too costly with electro/mechanical controls. During initial startup of the
Rice Lake system, one person was there full time for a period of one year to
monitor and work bugs out of the system.
A controller/monitor which works on the basis of a small microprocessor
has been under d sign and construction for the last six months under this
study. ———^——^»—__^__»^_—_
WVDIATCR
FERMENTS?
Figure 28. Schematic of original heat exchange system.
TANK
HEW"
0 PUMP
RAOIATDR 1NBUCWT
Figure 29. Schematic of rebuilt.heat exchange system.
45
-------
It is to be used to collect data and relay it on demand to a computer
located in Minneapolis. The unit, being designed by Automatic Hardware, also
will have some control functions via "on/off" relays. The cost of the pro-
cessor is low enough, that a number of desirable features can be incorporated.
They include safety, control and alarm monitoring functions, choice of manual,
semi-automatic, and automatic operation, and continual remote monitoring of
functions. This unit will be effective when the materials handling problems
of digesters have been reduced to a minimum. At the present time, the unit
serves as a data collection device for the digester operation.
46
-------
SECTION 6
BIOCHEMICAL CONCERNS AND ADVANCED RESEARCH
INFLUENT CHARACTERISTICS
Four systems were studied. Rice Lake, Ludington, Brockman Stables in
Neenah, Wisconsin and the pilot facility at Green Bay. They have been run on
dairy, beef, or horse manures, and the Rice Lake system has operated in an ex-
perimental mode on turkey wastes with a bedding of rice hulls. Table 10 lists
the systems, and influent analysis on representative samples.
In addition to strict biochemical analyses, several concepts were
developed to possibly enhance the economic feasibility of anaerobic systems.
Preliminary studies on algal growth on the effluent, uses of the effluent for
animal refeed, and the addition of municipal solid waste to the influent were
among the research topics considered to enhance feasibility.
The manure of various animals have different storage characteristics.
Collected turkey wastes will ensile themselves and heat up. In the turkey waste
system examined, the urine was lost because the houses were not cleaned daily.
This problem was exacerbated by long storage in the hopper prior to use. Rad-
ical changes in gas production could be observed depending on the age of the
waste. The longer the time between storage and use, the poorer the gas pro-
duction. No correlation was made because of the varying composition from the
different turkey houses. Ensiling would not be a problem for the unfermented
dairy or beef waste. One way to store these materials is in a covered lagoon,
which in the case of the digester can double as gas storage. Swine manure be-
haves in a manner similar to beef, while chicken and equine wastes react like
turkey wastes.
Rice hulls in the bedding of the turkey wastes proved to be a problem from
a materials handling perspective. The addition of water caused the rice hulls
to settle out, creating plugs in low piping points. As indicated by the analysis,
(Table 7), the rice hulls do not ferment well due to the high amount of inert
materials. Therefore, a shift in the type of bedding or elimination of bedding
material prior to fermentation is desirable. Should a digester system be in-
stalled in conjunction with the turkey operation, changes would be needed in
the choice of bedding materials and husbandry of wastes.
When the Rice Lake digester was shifted to operate on dairy input came from
a freestall dairy barn. The influent was essentially bedding free. The daily
production was metered to the fermenter uniformly over 24 hours under the
original design. The Ludington system's input was similar to that of the Rice
47
-------
Lake operation but came from an open feedlot and a mixing pit exposed to the
environment. Two unique feed problems occured in addition to winter freeze
up. First, there was a considerable amount of animal hair which plugged pipes
and tied up mascerators, which were installed to alleviate the problem. The
second problem was foaming in the mixing pit. This was eliminated by a
silicone surfactant (propriatary) which did not noticeably affect systems
operation.
The Green Bay pilot system was nand fed. Straw and bedding were omitted
intentionally to study animal waste characteristics free from beddings.
None of the influent tanks were heated; thus the temperature of the in-
fluent was reduced to ambient and/or soil temperature. No experiments were
made on changes in biological activity of the influent due to this cool down
period. Nor has there been a sufficient history of systems operation to
attribute any change in biological systems efficiency to this cooling effect.
However, additional energy is required to bring the material up to temperature
when it is introduced into the fermenter.
FERMENTER OPERATION
All three systems (Rice Lake, Ludington, and Green Bay) have essentially
been run as underloaded digesters. The Green Bay digester has the only
history of consistant operation during this work because of mechanical pro-
blems with the full scale systems. The loading rates and chemical parameters
of the pilot unit are shown in Figure 30. Table 6 under systems description
gives the data from which this graph is plotted.
input vs/ft3/day
Figure 30. Loading rates and chemical operating parameters of
Green Bay system.
48
-------
It is apparent from the graph that as loading rates were increased the
volatile acid content increased but so did the destruction of volatile acids.
As gas production and loading rates increase, so does the destruction of
volatile solids and volatile acids. The rate limiting loading rate, however,
was not determined. No data, is, as yet, available on long term operation of
a digester running on high load rates to ascertain the long term effect on
microbial populations and acidity inside the tank.
Experience with swine wastes by others (33,34,35) and our experience with
turkey, indicates that fermenters can operate at high pH or under other con-
ditions considered undesirable provided that the feed rate and materials are
kept within reasonable ranges on a daily basis. This experience is contrary
to recommendations and observations for sewage digesters. However, municipal
and industrial systems suffer the disadvantage of uneven loading rates coupled
with variable feed composition and lower solids composition. With low solids
concentration, slight changes in composition are liable to show up more quick-
ly.
With farm systems there is a possibility of drugs or other farm chemicals
entering the fermenter and upsetting the biochemical system. Cornell has
observed this with certain barn sprays. However, the systems observed in this
study have not reacted to normal farm practices. There were no attempts to
poison the systems during this study with common drugs or other materials;
therefore, no limits of tolerance have been discovered which could be used as
cautionary guidelines.
Of particular concern in digesters is the possibility of scum build-up
which can lead to crusting and a halt in gas production. This has not been a
problem with any of the systems which we have observed. One untested possib-
ility is that the insulation floats on the Ludington system keep the light
materials, which form scum mats, submerged in the liquor. Thus, the light
materials do not dry out and form a solid layer. Some credence to this theory
is the fact that in high solids digesters the materials are prevented from
separation and thus scum build-up. Another factor could be the short retention
time. This tends to wash partially digested materials out before build-up.
A third contributing factor could be that these systems are under loaded. The
dynamics of scum prevention are not well defined; pragmatic operational para-
meters seem to be the best guide at present.
MUNICIPAL SOLID WASTE/MANURE MIXTURES
These above factors become more important in light of the results of a
laboratory study which was conducted with a municipal solid waste: manure
mixture (2:1). The feasibility of using manures mixed with municipal solid
waste (MSW) appears to be attractive for small and large scale systems. As
mentioned earlier, small scale systems can be marginal economically. The
addition of the organic fraction of MSW to a fermenter can effectively increase
the size of the system to where it could be profitable for the farmer. The
effluent could still be used for fertilizer or other ends.
Appendix A points out the details of the laboratory studies on mixed
49
-------
manure and MSW. These systems were operated in the same fashion that a large
system would be. Sufficient numbers of identical systems were operated to
obtain statistical significance. The results indicate that the mix of 2:1,
MSW:Manure gives the same gas production as all manure over the time frame under
which the experiments were conducted. It remains to be seen what different
ratios will produce.
With MSW, one reintroduces the materials handling problem and the possib-
ility of introducing toxic materials. MSW also includes plastics which can
float to the surface and create scumming problems. Toward the end of the
experiment, gas production began to pick up. Time did not permit continuance
to verify the increased production but the question is raised as to whether the
population of microbial species was adapting to the new mixture. Other in-
vestigators have indicated a difference in gas production from the same
waste source depending on the seed materials used in starting the fermentation,
One farmer in Missouri, in cooperation with the University there, is running
a fermenter totally on hay. There is an attempt there to seek natural micro-
bial populations which enhance digestion of the organic matter. Thus, this
area needs to be explored more fully to ascertain whether there are selective
strains of microbial species which favor certain substrates. This might
eventually lead to the development of selected species for different
substrate materials.
The transition to MSW/manure mixed feedstocks was not tried on the full
scale systems due to lack of time. The Rice Lake operation did shift from
turkey to diary wastes by the gradual addition of dairy wastes. However, no
data was available on the shift from turkey to dairy because the system was
down for piping and engineering modifications. Turkey manure remained in
the digester at operational temperatures for two weeks before switchover
occured, thus, digestion may have been completed. Also, rice hulls remaining
from the turkey wastes caused severe plugging problems and necessitated a
partial pump out of the digester.
EFFLUENT CHARACTERISTICS
Effluent from the fermenter has a variety of uses: fertilizer, animal
refeed, algal growth and several other options. The use of the effluent for
algal growth has been studied by Golueke and others (37, 38, 60) and is
currently under investigation by Biogas of Colorado (27). Because of temp-
erature needs and large areas for solar absorption, this option is not
feasible for the Northern region of the United States where this study took
place. Thus, the two most feasible uses for digester effluent are fertilizer
and animal feedstuffs.
Use as fertilizer requires careful protection of the effluent to prevent
nitrogen loss. Spot samples from Green Bay, and studies by Moore at the
University of Minnesota (36) indicate rapid loss of this nitrogen via
volatilization from the effluent. This implies that a covered lagoon is a
cost-effective method to store the effluent and that knifing the material
into the ground is most efficient in preserving the fertilizer value.
50
-------
The possibility of developing an aquaculture system with digester
effluent was considered and would have been carried out on the horse farm had
operation been maintained by the owner-operators. Negotiations with a comm-
ercial fish farmer were underway when the system went down. The practice of
growing fish in sewage has been carried out for municipal wastes in countries
outside the United States and has been experimentally investigated in the
United States. Fish can grow in deeper ponds which allows for their use
with fermenter systems in colder climates.
Refeed
Using the fermenter effluent for animal refeed is a real possibility
and is under investigation in a number of places. USDA at Clay Center,
Nebraska (26) has a digester system designed by Hamilton Standard to check the
feasibility of using fermented effluent as a cattle feed. Biogas of Colorado
(27) has feeding trials using both fermented manures and algae grown on
effluent (the climate in the "Four Corners" area is suitable for this alter-
native).
Appendix B and C give the details of the preliminary "refeed" analysis
carried out under this contract. The first concern was the recoverability
of the material in a form suited for feedstocks.
A major problem in refeed is that digested effluent is dilute and
should probably be dewatered. This study looked specifically at the feasi-
bility of economically dew.atering the effluent to obtain a feedable cake.
Chemical flocculents were rejected because of costs and the problems of
obtaining FDA or local food code approvals. Obtaining all necessary approval
seemed to involve a prohibatively long and costly procedure.
Centrifugation was rejected for small operations (the focus of this
study) because of high-capital cost and maintenance problems. Hamilton
Standard has tested this process and it is under current investigation with
USDA and DOE funding (26, 16). The preliminary analysis indicate that the
procedures need more research.
Filtering via a filter press was also rejected because of high capital
costs. This left the option of a variety of low intermediate cost screens
and filters. It is possible to obtain a reasonable filter cake, but a large
fraction of very small matter is passed through the filters. It is these
particles which consist of high protein matter locked in cell walls.
Our studies indicate that there are no cost effective, FDA approved
precipitating materials for use with small scale farm systems. The only
viable alternative for reasonable, large scale capture is a centrifuge. This
conclusion was also reached by Biogas of Colorado and Hamilton Standard.
Capital and maintenance costs make these units cost prohibative for small
scale operations, though a lower feasibility limit was not ascertained.
Dewatering may be eliminated by utilizing high solids systems or mixing
the effluent in with haylage or other roughage and then feeding or ensiling
it for further fermentation. In arid regions or areas of high insolation,
51
-------
solar stills with water recapture is the most appropriate method. For the
region under study, aquaculture seems to be the method of choice. For
small operations the effluent can be fed directly to other animals, such as
swine, through the drinking water, as is done with undigested manures (41, 42).
For refeed to the same animals, there is a need for a moisture absorbing
nutrient supplement, due to the quality of the effluent. Addition of such a
supplement will create volumes too great for total refeed to the same
animals, thus, not all the effluent will be used for refeed.
Another question with regards to refeed is open. That is the nutrita-
tive value of the effluent. Three factors have to be taken into consideration,
the digestibility, palatibility and long term cumulative effects on the
animals and those who might consume the animal or their products such as milk
and eggs. The first two could be linked even if one does not anthro-
pomorphize. The digestibility of the effluent as opposed to the raw undi-
gestible manures also needs consideration in any economic evaluation.
At the present time, two extensive feeding trials are being carried
out. Biogas of Colorado is testing mesophilically digested effluent along
with algae grown on the waste water from the filtering process (27) The
USDA is testing the effluent from thermophillic systems. Research under this
contract was only preliminary because this was not a major emphasis of the
study. Results show that there appears to be a quantifiable increase in the
organic nitrogen portion of the effluent over that of the unfermented manures.
Biochemical breakdown of this material, simulating the ingestion by a
ruminant indicates that this material may be in a form somewhat less available
than that of the original manures. The use of this material by other animals,
however, is a genuine possibility. The results are only indicators of
possible differences and no firm conclusions can yet be drawn.
All indicators in this study point to the use of less capital and labor
intensive systems for processing the effluent. Thus, for small scale systems
it seems that the best alternative for uses of the effluent are as fertilizer
or as feedstocks for aquaculture and/or algal culture. The algae would be
used either as feedstock for animals or returned to the fermenter for enhanced
gas production.
GAS PRODUCTION
Both the quantity and quality of the gas produced is a function of the
digester operational parameters, but until special strains of microbial
populations are developed or innovative operational techniques are found, the
methane/carbon dioxide ratio will remain about 60 percent/40 percent.
There is some indication in the literature (43) that hydrogen is a rate
limiting factor. The direct addition of hydrogen to enhance the gas produc-
tion might be considered questionable unless a quantity of methane were pro-
duced which exceeded in energy value that of the hydrogen which was used.
Less directly hydrogen could be made available by cultivation of hydrogen-
producing algal microbial species. These organisms would also add cell mass
for refeed to the fermenter or animals.
52
-------
A preliminary study (Appendix D) was instituted to see whether such
species could be readily cultivated from existing algal species. Algal
growth on the effluent for use as either animal feed or biomass for the fer-
menter was initially rejected for this part of the country because of the
large acreage and shallow ponding required. This would have required too
much heat energy to maintain in the winter months.
The possibility of finding specific algae which could produce hydrogen
during biomass growth could possibly have made this work worthwhile. The
hydrogen is combined with carbon dioxide by microbial action to form methane
in the fermenter. This and the additional biomass could possibly enhance
gas production from digesters significantly. Results from this work, which
are preliminary, seem to bear out the experiences of Golueke and others (38,
39, and 60); the control of algal populations is difficult under all but
ideal conditions because of the inabliity to isolate pure strains and maintain
them. Additionally, the quantity of hydrogen produced per unit of biomass
or unit area is small.
But, the production of hydrogen from biological species has been under
investigation by a number of researchers (48) who are looking at modifying
existing species or finding different systems which can be induced to
provide high yields of hydrogen. Krampitz at Case-Western is currently
trying to create mutant species which can effectively produce hydrogen, (44)
and Mitusi, in Florida, is looking at tropical species which may be adaptable
to the United States for a variety of purposes. (45, 46, 47) In this
spirit, it may be possible to enhance the fermentation via some form of
pretreatment of the manures or other residues, such as plant biomass or MSW,
to enhance both solids digestion and/or gas production. The Environmental
Protection Agency and other groups have been sponsoring research on cellulosic
degredation via acid, basic and enzymatic hydrolysis. The Forest Products
Laboratory (USDA), Madison, Wisconsin,has been looking at various wood rot
fungi with the same interest (destruction and/or preservation of cellulosic
materials). (67, 68) Mitusi seems to be the only researcher taking a
systematic look at potential tropical species adaptable to the United States
for energy production via biological processes. (47)
The problem of gas quality and quantity is not severe if the biogas is
to be combusted directly and economics and efficiency are not a significant
consideration. Under most operations these constraints are severe. Thus,
to optimize the system it becomes necessary to decrease the carbon dioxide,
eliminate it entirely, spearate it for another use or transform it into a
useful product. As pointed out previously, it will not be economically
feasible in the foreseeable future to significantly decrease the production
or separate the CO^ for small systems. The possiblity of transformation in
an economical fashion is real and the technology is adaptable from industrial
knowledge; some which is readily available and some which is yet propriatary.
This is standard industrial practice for the production of either meth-
anal or ammonia. Both of, these are readily used on a farm and could be an al-
ternative use of the gas. The question of technology and economics of scale
are currently being investigated. This is an area for future research.
53
-------
TABLE 10. INFLUENT ANALYSIS ON REPRESENTATIVE SAMPLES"
Parameter
% M of Total Sample
% TS of Total Sample
% VS of TS
% FS of TS
% VS of Total Sample
% FS of Total Sample
Total Alkalinity
Volatile Acids
pH: Non-Filtered
pH: Filtered
Total Kjeldahl Nitrogen
Ammonia Nitrogen
Organic Nitrogen
Chemical Oxygen Demand
Unit
+->
.c
en
•i—
O)
3
>>
_a
>e
mg/1
mg/1
—
--
mg/1
mg/1
mg/1
mg/1
Rice
Lake1
97.09
2.91
69.52
30.48
2.02
.89
6360
2330
7.7
7.7
2260
1317
943
46,000
UWGB2
94.69
5.31
83.27
16.73
4.42
.89
5750
3880
7.0
7.0
1780
428
1352
84,600
Ludington
95.03
4.97
81.49
18.51
4.04
.93
2940
4340
5.5
5.7
1830
457
1373
67,000
Brockman
95.28
4.72
89.59
10.41
4.23
.49
1890
1690
6.6
6.7
670
130
540
28,000
en
M = Moisture
TS = Total Solids
VS = Volatile Solids
FS = Fixed Solids
mg/1 = Milligram/liter ^
(16,000 mg/1 = 1 Ib./ff3)
1) Sample was diluted, aged TURKEY manure
2) Sample was diluted DAIRY manure.
3) Sample was diluted CATTLE manure.
4) Sample was diluted HORSE manure.
* All of these systems were run as underloaded digesters because of the materials handling problems asso-
ciated with high load rates.
-------
SECTION 7
ECONOMICS
The economic feasibility of anaerobic digesters is predicated on a
number of significant parameters. These can be categorized into two areas of
concern: Design engineering and market economics. The amount and type of
waste available, type of farming operation the system is designed for, capitol
cost, end use of the biogas, operating expenses, insulation of influent and
fermentation tanks, engineering efficiency of the system, and the cost of
fuels the biogas replaces are just some of the factors.
If the gas can be burned directly on site, the recovery period and
annual returns are more favorable than if electrical conversion is necessary.
Direct combustion will lower initial investment costs and decrease operating
expenses. Coupling an engine generator of sufficient quality to provide
several years of reliable service can increase investment costs as much as
32 percent and annual operating costs up to 39 percent. Conversion efficiency
of biogas to electricity by a gas engine is quite poor (10-15 percent for
standard manufactured sets and up to 25 percent for mixed fuel-diesal/biogas
units), leading to low annual returns and a great deal of waste heat. Care-
ful siting of the engine generator would enhance the economics by using the
waste heat from the engine to heat outbuildings or, perhaps, a greenhouse.
The use of integrated systems with electrical generation similar to the
district heating concept, is an avenue that must be explored further. This
investigation indicates that the economics of producing electricity from
biogas on a small scale is unfavorable unless the waste heat is recovered.
The upper midwest, particularly Wisconsin and Minnesota, contain a
significant number of small dairy farms. Many of these operations are
economically marginal. As energy costs continue to rise and pollution regu-
lations become increasingly strict, a large percentage of these operations
may be forced out of business. It is for that reason that anaerobic digesters
are being evaluated for use on small dairy farms.
The Rice Lake digester is located on a dairy farm in Northwest Wisconsin.
The system was considered a prototype to determine the feasibility for
application on the many farms of this size in the region. The existing Rice
Lake digester is sized for 125 milking head, but the farm operated closer to
the 100 cow level. This has been the lower limit to which the economics of
dairy farm digester design have been field tested. Cornell has done theo-
retical studies showing that, in some cases, dairy farms with as few as 75
head can show anaerobic digestion to be an economical investment. (1) They
are just now beginning operation of a 65 cow digester.
55
-------
While dairy farm operations are basically similar, variance in acreage,
age of equipment, and current management practices does not allow for a sim-
ple economic analysis which can be immediately transferred from one operation
to the next, much less to other types of farms such as swine and poultry.
The issue is further complicated by potential uses of the biogas coupled with
uncertainty in energy prices.
An economic analysis was performed on a design similar to the existing
Rice Lake digester. To show returns based on energy production alone, the
digester was assumed to be an additional investment to the best possible
manure handling system in terms of nutrient conservation, the covered anae-
robic lagoon. Costs of the covered lagoon are found in Table 11. The costs
of the digester are based on installed system components found in Table 12.
To determine the relative significance of design parameters, the Rice Lake
system was used as a baseline operation. Specific parameters were then
altered to show the impact on economic feasibility. Engineering improvements
that were considered to enhance the feasibility include:
1) Increasing generation efficiency of the engine generator to
25 percent.
2) Insulate the influent tank and fermentation tank.
3) Utilize waste heat from the engine to heat outbuildings.
4) Combust the biogas directly on site.
TABLE 11. BASELINE COSTS FOR A COVERED ANAEROBIC
LAGOON
100 Cow Dairy
Barn Scraper, Earthen Manure $ 2,000
Storage Pit
Lagoon 2,000
Lagoon liner & cover 5,000
Pumps (2) 7,000
Piping 500
Large tank manure spreader 7,000
TOTAL $23,000
56
-------
TABLE 12. COMPONENTS AND INSTALLED COSTS OF
EXISTING RICE LAKE DIGESTER
Tank
Pumps
Controls
Hydraulics
Piping
Building
Engine-Generator
High Pressure Storage
Electrical Tie-In
Well
Miscellaneous (15%)
Engineering (15%)
Total
$ 6,250
1,500
1,500
2,500
1,000
3,600
12,000
2,000
1,500
600
3,250
5,350
$41,050
By calculating the present cost per kilowatt of digester produced
electricity and plotting these costs against the present cost and rate of
fuel price increases, the feasibility of anaerobic digesters can be deter-
mined. In most cases, the return on fuel savings would not begin until the
investment costs are paid off, the system fully depreciated, and operating
expenses are the only annual costs involved in the cash flow. Therefore,
the shortest possible loan and depreciation periods will prove to be the
most economical.
For comparative purposes the annual costs and returns from the Rice Lake
digester are shown in Tables 13 and 14.
A heat value of 580 Btu per cubic foot of biogas was assumed. If 85
percent of the manure produced is fed to the digester, approximately 720 x
10^ of biogas will be produced annually. Conversion efficiency of the gas
to electricity is low, 10 to 14 percent. Ecotope Group of Seattle has a
similar engine generator operating off biogas. Their average conversion
efficiency has been 11 percent. (19) Converting the gas produced by the
Rice Lake system to electricity, at 11 percent will yield 23,200 kilowatts
per year, or 22 percent of the operating capacity of the Rice Lake generator
and approximately 25 percent of the total electrical consumption on the farm.
An optimistic analysis was carried out by assuming that 100 percent of
the electricity and biogas produced would be consumed in useful work. The
graphs, then, show costs of the "best case" dairy farm system should the
engineering and marketing problems be worked out. In reality, a large por-
tion of the electricity generated goes unused, as witnessed by the farm load
profile shown in Figure El (Appendix). Between the hours of 10 p.m. and
5:30 a.m., there is virtually no electrical load, yet the generator is
57
-------
TABLE 13. ANNUAL COSTS OF RICE LAKE DIGESTER
5 yr. loan 10 yr. loan 20 yr. loan
Capital9 $41,050*00 $41,050.00 41,050.00
Annual Capital Costsb 10,691.00 6,538.00 4,658.00
Annual Operating Costsc 2,869.00 2,869.00 2,869.00
Equipment maintenance^
and repair $ 1,174
Labore 624
Taxes & Insurance^ 1,071
Total Annual Cost 13,560.00 9,407.00 7,527.00
Total Annual Costs Averaged
over 20 years 5,801.00 6,397.00 7,786.00
Annual Cost Per Cow
normalized over 20 yr. life 58.01 63.97 77.86
aActual installed cost of the equipment, site preparation, and building;
cost of land excluded.
^Annual capital costs are based on a yearly amortization rate of 9.5% for
5, 10, and 20 year periods. All capital is borrowed:'
cAnnual operating costs are assumed constant for the first 10 years, after
which they rise at the rate of 3% per year.
dEquipment maintenance and repairs includes the cost of maintaining the
engine/generator system, which comes to $200 per year in parts and labor,
based on maintenance procedures recommended by Waukesha. Also included is
the cost of an engine overhaul every 6 years at $2500. Maintenance and
repair of other system components was taken at 2% of the remaining $23,700
equipment costs.
eLabor. The present owner of the digester puts in approximately 20 hours
per week operating and maintaining the system. If the system is working as
designed the number of hours labor per week should be 10 or less. Four
hours per week at $3 per hour was used to obtain the labor cost figure.
fTaxes and insurance were estimated at 3% of the original equipment value
of $35,700.
58
-------
operating 24 hours a day. Load leveling would enhance the economics of
site generated electricity. Figure 31 shows the cost per kilowatt hour of
digester produced electricity from the Rice Lake digester versus electricity
purchased from the utility company for a range of current prices.
Because the farm is a business, tax deductions can be taken for
interest paid and depreciation of equipment. Costs per generated kilowatt
include tax deductions from interest and straight line depreciation for the
30 percent taxable income bracket. Annual operating costs for the first 10
years are assumed constant, after which they rise at a rate of 3 percent per
year.
At this present value of $.027/kwh to large northwest Wisconsin
electrical users such as the Rice Lake farm, (49) the returns from the
electricity generated will be $626 the first year. Over a 20 year period,
with electricity prices rising @11 percent per year, the returns would
total $40,191. Table 15 shows the 20 year lifetime costs and returns for the
different loan and write off periods. The cost per kilowatt hour, normal-
ized over the 20 year period, is compared to the same costs ef electricity
if purchased from the utility company.
TABLE 14. ANNUAL BIOGAS PRODUCTION, RICE LAKE*
Manure
Total Solids
Volatile Solids
Biogas Production
100% capacity
85% capacity
Btu content of gas produced 719.2
LBS. x 106
3.10
.39
.32
1.46 cf
1.09 cf
x 106 9 85% capacity
KG x 106
1.41
.18
.14
.041 m3
.031 m3
*Based on 100 animal units, avg. live weight 100 pounds. (1)
85 Ibs. of manure/cow/day
10.6 Ibs. total solids/cow (TS)
8.7 Ibs. volatile solids (VSAJ
4.56 cf biogas/lb. of
580 Btu/cf of biogas
59
-------
TABLE 15. COST/KWH OF RICE LAKE DIGESTER NORMALIZED
OVER 20 YEAR LIFE
Loan & Write
off period
5
10-
20
Cost over
20 years
$116,020
127,940
155,720
20 year returns
$40,191
40,191
40,191
Average 20 year
cost/kwh*
.219
.230
.275
(no digester,
electricity (40,191)
purchased 0 .027)
(@ .05) (74,475)
(.087)
(.160)
*Deductions accounted for.
From these figures and from the graph, it is apparent that biogas
conversion to electricity as currently designed on the Rice Lake farm, will
not become economical for another nine to 17 years, depending on the
current cost of electricity and on the period chosen for write off. In all
cases, the sooner the system is paid for, the sooner the returns will
begin. Should engineering design be optimized, the economics will be
affected as indicated in the following analysis.
_._._5 YEAK WRITE OFF
.-. 10 f*ff WHtTE OFF
20 rEA* WRITE OFF
8«»SE Or COSTS FOR
CUSCKASEO ELECTRICITY
I»C»SASIMG.1H i
0
(UTS)
15
20
Figure 31. Cost of electricity from $40,000 digester vs. cost of
electricity from utility csrop-any.
60
-------
Using Excess Heat to Heat Outbuildings
The amount of excess heat available from the engine cooling system is
significant. When operating at part load (50 percent) the heat output from
the cooling jacket is 69,000 Btu/hour. (50) The fermentation tank requires
23 percent of 50,700 Btu/hour, or $5.32 of propane, equivalent ($4.48
of fuel oil) heat per day is being vented to the outside air. If this
could be captured for use in outbuildings, the impact on economic recovery
would be noticable. Assuming 60 percent recovery of this waste heat for 200
days of heating, the annual savings, based on equivalent propane usage,
would be,$638 per year, and $538 of fuel oil savings. This is virtually
equivalent to the first year of electrical savings, cutting the recovery
period in half. Cost of heat recovery equipment was assumed to be $5,000
or $800 annually. If the heat is coupled with a productive unit, such as
a greenhouse, the returns, in the form of increased farm productivity, would
be even more significant.
Increased Generation Efficiency
Optimization of the system would include, not only heat recovery, but
increased generation efficiency as well. An engine generator that is 25 per-
cent efficient would be capable of producing 52,725 kwh/year, or two and one
quarter times the amount generated at current Rice Lake efficiency of 11 per-
cent. This could be accomplished, theoretically, by using a diesel engine
run on a 90:10 mixture of biogas and diesel. (51) Initial capital costs for
a diesel run generator would be approximately 25 percent above the costs of
a gas engine generator. The system investment would be increased $3,000,
while the operating costs would increase $376 annually, due to cost of the
diesel fuel required. (Table E3) Returns, on electricity alone, would be
$1,425 the first year. When coupled with heat recovery, returns total
$2,063. Figure 32 shows the cost per kilowatt of diesel-biogas generated
electricity and heat recovery over the 20-year life versus costs of electric-
ity purchased from the utility company. It becomes apparent that paybacks
could begin as early as six years with 25 percent generation efficiency and
a five year write-off period. Table 16 shows the 20-year costs and returns
for different write-off periods.
M.M.OT.M.
-------
TABLE 16. COST/KWH OF OPTIMIZED DIGESTER
ELECTRICAL SYSTEM
Loan Period
5
10
20
Equivalent
Electricity
20 Year
Costs
$134,637
148,927
183,487
20 Year
Returns
$110,041
110,041
110,041
Average cost/kwh*
for 20 yr. Period
.077
.081
.098
Cost/kwh with 80%
use factor
.096
.102
.123
(3 .027
@ .05
110,041
170,542
.072
.112
.072
.112
*Deductions accounted for
These figures show that for small scale dairy farms with electrical
operation, the economics of anaerobic digesters are marginal. Careful
optimization for use of waste heat, maximum generation efficiency, and use of
electricity, is required to achieve reasonable savings over a 20 year period.
If optimized, and electricity costs are above $.037/kw, then such a system
would pay for itself in a twenty year period if care can be taken to use all
the electricity produced. The alternative to electrical generation is to
burn biogas directly on site.
Combust Biogas Directly
Direct combustion of the biogas on site would replace the use of liquid
propane or fuel oil as current fuel sources. Significant savings could be
accrued if there was a use for this amount of gas. Total capital required
would be much lower because the engine generator, the electrical tie-in, and
engineering fees associated with the electricity would not be required. Costs
range from as low as $16,000 to $25,000. The capital costs used in this
analysis are shown in Table E4.
Operating costs would also be significantly lower. Twenty-two percent
of the biogas produced would be requred to heat the fermentation tank. Net
biogas production from a digester the size of the Rice Lake system would be
6,138 gallons of propane equivalent, or 4,493 gallons of fuel oil equivalent.
At present value of $.40/gallon of propane, the first year returns would be
$2,455. If fuel oil is displaced, the reutrns would total $2,067 the first
62
-------
year. Table 17 and Figure 33 show the costs and returns of direct biogas
combustion on site averaged over a 20-year life compared with equivalent
usage of propane or fuel oil.
With direct combustion , the system would pay for itself in less than
20 years. The ability of the farm to utilize all of the gas produced will
determine whether this analysis holds true. Also, the initial cost of
converting the farm operation to use biogas must be explored for those oper-
ations having high electrical and low fuel usage.
D'lrect Combustion Digester
Fuel 011 9 $.46/gal.
Electricity 9 J.04/kw
Optimized Digester. Electrical Conversion
Existing R1ce Lake Digester
Electrical Conversion
(1978)
Figure 33. Btu/$ of different energy and digester types,
63
-------
The amount of energy required to heat the influent to fermentation tank
temperature of 95 degrees is significant. This is compounded during cold
winter months if the influent tank is exposed to air or ground ambient temp-
eratures. There have been times during the past two winters when the Rice
Lake influent tank actually froze. If the influent can be kept at 55 degrees
during the winter months, versus 35° F, a savings of 170,000 Btu, or $.74
worth of propane equivalent can be saved per day. This would amount to
approximately $89 per year_in energy savings, which would be accrued in in-
creased biogas available for on-farm use. Close to seven percent of the
average daily biogas production could be saved by insulating the influent
tank. This added investment would pay for itself in less than two years.
TABLE 17. COSTS AND RETURNS OF DIRECT BIOGAS COMBUSTION
ON SITE
Loan
Period
5
10
20
No Digester
20 Year
Costs
$51,591
56,931
69,371
20 Year
Propane
73,105
73,105
73,105
Returns
Fuel Oil
64,013
64,013
64,013
20 Year Average
Cost/Gallon
Fuel Oil
.502
.529
.630
.712
20 Year
*
Average Cost
Gallon Propane
.368
.387
.461
.595
Deductions Accounted For.
CONCLUSIONS
Two factors should be noted. This analysis is predicated upon a small
number of animals. Increases in systems size rapidly change this economics.
Next, the economics are on energy alone since no proven credits can be taken
for animal refeed or fertilizer over that of undigested animal wastes. Cer-
tain externalities, such as odor control, or those which require a farmer to
install pollution abatement practices would nullify the economic analysis.
Future research in this area, and extenuating circumstances, change the
economics radically.
1) Direct combustion of biogas on the farm offers significant savings
and shorter payback periods over power production via an engine
generator system.
-------
TABLE 18. COMPARISON OF DIGESTER SYSTEMS
System *Cost Over **20 year ***Return on
20 years Returns Investment
Current Rice Lake System $116,020 $ 59,580 -6.9$
Optimized Rice Lake System
with diesel generator
and heat recovery 134,637 174,139 4.0%
Direct Combustion of
Biogas on site:
(replacing propane 51,568 72,759 5.8%
(replacing fuel oil) 51,568 61,545 2.7%
*Using 5 year payment plan and 5 year write-off.
**Based on electricity @.04/kwh, increasing 11%/year; Propane @.40/gallon,
increasing 4%/year; fuel oil @.46/gallon, increasing 4%/year.
***ROI = (20 year value of energy produced - 20 year cost of producing it)
20
2) The cost of producing electricity is high and conversion efficiencies
low for the existing engine generators. More efficient engines must
be used to improve economic recovery or alternatives such as load
following fuel cells need to be explored.
3) With electrical conversion careful siting of the digester and power
plant will allow waste heat recovery. Energy and actual dollar
savings could virtually double if the heat is recovered.
4) A diesel powered, duel fuel )oil and biogas mix) generator is twice
as efficient as present gas engines. This option, coupled with
waste recovery from the colling system, appears to be the most
economical form of converting biogas to electricity using current
commercially available systems.
5) It is essential that the fermentation and influent tanks be insu-
lated in colder climates. Insulation of both tanks is low cost
investment which can increase net biogas production up to 20 per-
cent.
6) The sooner the digester can be paid for, the more economical it
becomes.
65
-------
Economic models have a certain inflexibility which tends to lock us
into a narrow vision of lifestyles and development patterns. The develop-
ment patterns of farm energy use have dictated increasing electrical demand.
For this reason, most digesters being installed today are converting biogas
to electricity. The economics show that direct combustion, or careful
planning to recover waste heat and optimize conversion efficiencies have the
most promise for small farm use. The implications of this model can be
several depending on how it is viewed:
a) Digesters for small farm use may never be adopted on a widescale
because they are not economical.
b) Small to medium sized digesters will not solve energy use problems
by themselves. They must be effectively integrated into the total
farm operation through energy conservation, load leveling, and
matching low quality thermal loads for direct combustion and high
quality machinery loads for electrical use.
c) Diversification of small farm operations through refeed of animal
wastes, coupling a greenhouse with the digester to utilize the
waste heat and create year round farming opportunities, and pos-
ible aqua-culture or algalculture on the effluent will be necessary
to accomodate and enhance both digester feasibility and total
farm stability.
d) The anaerobic digester can become central to a farm integrated
utility system providing both power and heat to farming oper-
ations and also opening new avenues for farm production.
66
-------
SECTION 8
OTHER CONSIDERATIONS
QUANTITY OF WASTE
The amount of organic material available for anaerobic digestion will
have a significant impact on total biogas production and, hence, on the
economics of the system. Cost/benefit ratios of digesters as energy pro-
ducing systems have been shown, in 1975 dollars, to break even at around
250 animal units. (20, 57) More recent studies (30) show that below 400
animal units the system economics deteriorate rapidly. One company, Hamil-
ton Standard, has set 8,000 head as the minimum size feedlot for which
anaerobic digestion is economically feasible. (52)
While increasing biogas production is largely a function of input
material, a higher percentage of solids and increased retention times could
also enhance the amount of biogas produced per pound of manure added. By
increasing loading rates, Hamilton Standard is targetting system feasibility
at one to two thousand head. (52) The thermophilic digester temperatures
examined by Hamilton Standard could play an important role in determining
the necessity of large scale systems for economic feasibility.
The addition of other organic material, such as crop residues or munici-
pal solid waste (MSW), would increase the quantity of biogas produced.
Laboratory experiments during the course of this study showed that the
addition of MSW to dairy manure would increase biogas production, based on
the amount of digestible material, but would not increase the gas production
per pound of volatile solids added (VSA). For manures with higher nitrogen
contents, the gas production per VSA may be greater due to the better carbon
to nitrogen ratios. Further research must be accomplished in this area, but
it appears that the combination of MSW with animal manures could make small
systems economical if circumstances were right.
While MSW is a resource found in rather large quantities, the costs of
obtaining it may be prohibitive for small farms. These are costs associated
with separation, classification, hammer milling, and transportation of the
material. Currently MSW in rural areas is landfilled, and it is not avail-
able in a form suited for use in farm anaerobic digesters. If MSW processing
facilities develop in these rural areas, its use for farm scale digesters
must compete with the other uses of it as an alternative fuel. For this
reason, an economic analysis was not performed on the use of MSW in conjunc-
tion with animal wastes for 100 cow dairies. While the use of MSW to enhance
biogas production from farm animal wastes appears reasonable, the feasibility
67
-------
seems limited to large scale operations in conjunction with solid waste
processing facilities, or, to those farms located in close proximity to
such a facility.
TAX INCENTIVES
Incentives provided by governments in the form of tax credits or cost
sharing could give anaerobic digesters the necessary economic boost needed
to make the system competitive. Tax credits for solar energy conversion
devices, synthetic natural gas producing equipment, or water pollution con-
trol equipment would qualify anaerobic digesters, providing the system was
defined as such.
Present income tax laws would exclude farm scale digesters from pollu-
tion control deductions because 1) digesters are not classified as a non-
point pollution control measure, and 2) most farms have too few animals to
qualify as a point source feedlot operation. (53, 54)
The State of Wisconsin has proposed a bill that would give investment
tax credits to purchasers of alternative energy producing devices (55),
and in the 1978 National Energy Plan both the House and Senate versions
contain provisions for 10 percent tax credits on synthetic gas producing
equipment. The Senate version also contains provisions for bioconversion
equipment.
For the systems examined in this economic analysis, a 10 percent tax
credit would enhance the economics as follows: For the existing Rice Lake
digester, with no optimization, a 10 percent credit would make the system
competitive with electricity at $.26/kwh. For the optimized system, with
25 percent conversion and heat recovery, first year costs of the credit would
make the system competitive with electricity costs at $.09/kwh. For a
digester located on a farm using 100 percent of the first year gas directly,
a 10 percent tax credit would make the system competitive with propane at
$.45/gallon and fuel oil at $.61 gallon for the first year of operation.
These figures were generated for the five year annual cost models.
A one year tax credit, however, would not have any effect on the compet-
itiveness of the systems in the following years because of the high annual
costs involved.If the credit was available for a period of three to five
years, giving a total credit of 30 to 50 percent, then the systems could be
considered competitive with other forms of energy.
SOCIOECONOMIC
Social or cultural acceptability is often a major factor influencing the
success or failure of an innovation. Because this is so subjective and ill
defined, there is an obvious temptation to favor the more rational criteria
of engineering efficiency or economic viability. The feasibility for wide-
spread adoption of anaerobic digesters, however, transcends strict economic
analyses.
68
-------
This becomes apparent if one examines the relationships between innova-
tive technologies and the people to whom the innovation is addressed.
Successful innovations, be they industrial or agricultural, are character-
ized by the development of a new technology, or new form of organization, to
meet the need or demand for a new product or service. The innovation pre-
sumably corresponds to a pre-existing demand, which will allow it to diffuse
throughout the economic and social system. Success is not the general rule,
however, and a strategy for development of innovative technologies in energy
from agriculture must account for this.
Two approaches can be seen. The first is to help increase the total
number of innovations in the field, thereby increasing the liklihood that
one or several will succeed. The second approach is to try and reduce the
incidence of failure through systematic identification of the factors which
contribute to the success of innovations. (56)
Research concerned with the adoption of agricultural innovations revealed
some of the more significant factors underlying the success of innovations in
the agricultural sector. (63 through 66) (Appendix F) The more signifi-
cant of these, shown in Figure 34, include: (in addition to economics) The
social and environmental implications of the technology, institutional
factors such as the ability to insure and finance the project, assurance of
systems reliability and service of the technology, and regulatory and legal
decisions influencing farm practices. All of these factors are weighed
against the technology in light of the significance it may have for the over-
all farming operation. In essence, the perceptions of the people to whom
the technology is addressed and those of the innovator must be compatible.
The innovator is usually an outsider, highly educated, and familiar
with what is going on in the rest of the world. In this case, the innovator
is the person or company responsible for the design and construction of the
anaerobic digester. If an innovation is adopted and the promoter then
leaves town, it runs the risk of being neutralized or even rejected by the
community. A perfect example of this is the Rice Lake digester. It was
built by an outside company who made no provisions for local service people
to maintain the system in good operating order. The design is a prototype
and requires constant attention; time that is simply not available to
most farmers. Because the system has been inoperable most of the time, it is
not viewed by the local population as a successful venture.
Support of these factors affecting adoption can be found in a 1974
survey of Brown County, Wisconsin farmers concerning the attitudes of
farmers toward manure handling practices (57). Over ten percent of those
surveyed were actively interested in trying anaerobic digesters if they could
see an operational system and could be assured there would be adequate
service and trouble free operation. Interviews with farmers in Barren County,
Wisconsin substantiated this. When asked to weigh barriers preventing their
use of anaerobic digesters, virtually all of them gave as the most important
reasons (besides initial cost too high): "I would have to know of operational
systems on farms like mine", and "I would need to be assured of good reliable
service." (58) Thus, a primary factor contributing to the success of
69
-------
External
factors
H Regulatory
and Legal
H Systems 1
Design 1
-f
L
— {institutional 1—
Extent of economic
advantage over
alternatives
Monetary payoff
Replacement statu;
(condition) of
existing equipment
Tax Incentives
Pollution control
(Manure handling
rules and policy)
Zoning and Land
use regulations
! Divisibility for L
trial J
Reliability (Regu-l
larity of reward) r
Availability of
technical support
(Support of farm
finance sector
Support of InsuK-
ance sector
Infrastructure
for aupport and
dissemination of
information
(diffusion)
\ & \
Attitudes and
Values
Capital
, Value of land,
buildings, etc.
(Farm size and
| location
(Current income
Association wit
main enterprise
(dairying)
| 1
h
Overall signifiacance
of the practice -
for the farm operation
1 Farmer 1
Characteristics |
Ilnte
fact
|Farra 1
•Characteristics |
rnal
ora
Figure 34. Significant factors underlying the adoption of agricultural innovations.
-------
anaerobic digesters will be the availability of service for the systems.
REGULATORY AND LEGAL CONSIDERATIONS
The development of a strong service organization seems dependent on
sound engineering, in materials and operational guidelines, as well as
safety aspects of the system. Some day use of biogas may be commercialized
in much the same manner that natural gas or LP gas is now. But until the
equipment and operational practices are developed, special attention needs
to be given to the development of standards, codes, and regulations concern-
ing the safety of anaerobic digesters.
At present, there are a plethora of codes, regulations, and guidelines
which have applicability to the construction of anaerobic digesters, though
none are specific. The Occupational Safety and Health Administration (OSHA)
has adopted the National Electric Code, which makes reference to locations
in which flammable gases or vapors are present in sufficient quantities
to constitute an explosion. The National Gas Code, and guidelines establish-
ed by the American Society for Testing Materials are other regulations and
guidelines which must be considered by designers. Most of what is found
in these types of documents does not apply to biogas. Extracts from the
various agency guidelines should be combined to form a standard by which
farm scale digesters can be assured of providing safe, trouble free operation.
Two types of standards, or regulation seem necessary. The first would
deal with construction and installation standards. This would assure the
purchaser of a commercial system that the design was sound from an engineer-
ing standpoint. The second would be a licensing type of regulation which
would assure that the systems are maintained and could be deactivated in a
safe manner, protecting the public from a potentially hazardous situation.
INSTITUTIONAL CONSIDERATIONS
While the establishment of regulations and standards would assure the
purchaser(s) he was getting quality equipment and reliable design, they would
also increase the liklihood that insurance, and financing institutions
would support anaerobic digesters as a viable technology.
Most insurance companies base their premiums on past experience. With
new technologies, insurers do not know ahead of time what their "bsses and
loss adjustment expenses will be. Anaerobic digesters have no history or
demonstration of safety, thus insurance and products liability are difficult
to obtain. The American Mutual Insurance Alliance (59) suggested that to
break the initial barrier a demonstration of systems' safety to an innovative
member of the insurance community would be beneficial.
One hundred years ago, boiler insurance was difficult to get. Seventy-
five years ago, however, a group of engineers got totether to set up ASME
standards for boilers. Now boiler insurance is not difficult to obtain.
71
-------
The fact that anaerobic digesters are located in rural areas, away from
population concentrations, is an advantage in case of explosion. Basically,
the product must be designed, manufactured, and installed with quality
control to become readily insurable. Vulnerable areas of manufacturing include:
design errors, improper material selection, assembly errors, inadequate
testing, improper instruction on product use, inadequate warning of latent
dangers, insufficient or inadequate safety features, failure to anticipate
forseeable misuse of the product, and failure to take corrective action when
problems are discovered. (59) All of these could be accounted for with
proper standards for design and construction of anaerobic digesters.
Because anaerobic digestion is a relatively unproven technology,
lending institutions are also leary of supporting such a project. Discus-
sions with lenders revealed that long term money will be easier to obtain
until the system proves itself economically. As indicated by the economic
analysis in this report, long term money can put a severe dent in the
recovery of capital system invested in digesters. While initial cost is not
a significant factor, the system is viewed in terms of payback on the digester
and savings of time. An innovation that will save time will allow the
farmer to become a better manager of his farm operation.
Considerations which the lending institutions examine before committing
money for innovative projects fall under "farm and farmer characteristics"
of Figure 34. These include:
1) Will the digester make or save money?
2) Can the farmer afford the luxury of gambling on a somewhat
unproven technology? (In other words, what is the net
worth and capital income of the man?)
3) What is the experience and ability of the man who will
become the one to make the system work?
4) Is the innovation congruent with the existing operation,
and will it improve the efficiency of the farm operation?
5) What is the salvage value in case of failure?
ENVIRONMENTAL
In attempts to show positive economics for small scale anaerobic
systems, other benefits have been factored in with economic values. These
include enhanced fertilizer value of the effluent, increased value of the
effluent as animal feed, and the use of these systems to control water and
air (odor) pollution. Our studies have not shown the fertilizer and refeed
values of the effluent to be enhanced by anaerobic digestion. Thus, the
control of water and odor pollution must be evaluated.
72
-------
Odor Control
One argument for anaerobic digestion is that the process stabilizes
wastes and reduces odor. This practice has been particularly encouraged in
rural residential areas where odor is of concern. Presently, manure holding
ponds which go under the name of anaerobic lagoons, have such high BOD
loadings that anaerobic stabilization does not occur. When the systems
are agitated for removal of the materials odor becomes a nuisance.
If a completely covered anaerobic lagoon can be built, however, the
advantage of odor reduction via anaerobic digestion is eliminated. The
covered lagoon can be agitated, the effluent removed and knifed into the
soil in such a fashion that odor can be minimal. (22)
If a covered lagoon cannot be built, part of the digester can be
credited to odor pollution control. Also, some additional credit may be
obtained because the stabilized manure from the digestion tank would not
require so large a retention pond as unstabilized manure.
Water Quality
A study of the water quality in Northwest Wisconsin and elaboration
of nonpoint source planning problems was undertaken to determine the likli-
hood of environmental regulation stimulating the adoption of anaerobic
digesters.
In summary: Northwest Wisconsin does not have the serious pollution
problems prevalent in other portions of the state. Since there are few
large-scale industrial polluters, nonpoint agricultural pollution becomes
the major concern. High nutrient levels in the Red Cedar River and sediment
problems in the Chippewa, below Eau Claire, are considered problem areas.
(Figure 35) Though no studies have identified the exact cause of these
problems, non point agricultural sources are suspected to be the major
contributor.
Figure 36 identifies the problem areas the County Soil and Water
Conservation District (SWCD) has found attributable to waste management
practices in Barren County. The magnitude of these problems are not defined.
Polluted lakes and streams noted are combined sediment and fertility problems,
with agricultural runoff or cattle grazing in the stream considered the
sources. Barnyard and feedlot runoff indicated problems traced to their
source, many being dairy operations. Nutrient losses indicate high nutrient
levels in the waterways which are suspected to be agriculturally related.
The most significant affect land disposal of manure has on water
quality stems from proper timing and rate of application, the degree to
which manure is incorporated into the soil when spread, soil structure, and
the slope of the land on which it is spread. A well designed waste storage
facility and management plan can incorporate all of the above management
practices. Anaerobically fermented waste contains less solids than non-
fermented waste. This, perhaps, makes it more desireable for land application,
73
-------
UPPER MISSISSIPPI RIVER BASIN
BURNETT IWASHBURN "SAWYER
I | «
if Regional Center
Miles
Figure 35. Chippewa River Basin, Upper Mississippi Region - Location map.
(from Upper Mississippi River Basin Commission)
74
-------
SCALE
MILES
WISC. DEPARTMENT OF TRANSPORTATION
US. SOU CONSERVATION SERVICE
A Polluted Lakes and Streams
• Barnyard and Feed lot Runoff
I Nutrient Loss>e?
Figure 36. Barren County water quality
problem areas.
75
-------
if not disked in promptly, because it will soak into the soil more readily.
Whether or not digested slurry could significantly reduce water pollution
beyond liquid waste management systems remains to be determined. However,
it appears that the reduction would be minimal.
The interactions between manure management, soil, and water quality are
not completely understood nor quantified. Thus, it becomes difficult to
prove the necessity of waste management practices and the impact anaerobic
digesters would have on water quality, and, also, it becomes difficult to
enforce nonpoint water quality regulations. The 208 nonpoint source
planning program, being developed by the state DNR and Soil Conservation
Service (SCS), advocates the use of Best Management Practices (BMP's) of
manure to help alleviate deteriorating lake and stream water quality. These
include:
- A waste storage pond with a minimum of 120 day retention
time.
- Waste should not be spread on frozen soil, but spread and
and incorporated into the soil prior to the tillage for
planting.
- Waste may be spread on fields in the fall and incorporated
to prevent runoff.
- The amount of waste applied to the soil must not exceed
the requirements of the crop and the soil nutrient holding
capacity.
Actual regulation of these practices will be difficult if not impossible.
First, data assessing the ability of the BMP's to reduce water pollution
is not available. Second, the nature of the problem requires individualized
regulation for each farm. Manpower would be required that is not presently
available. Permit programs would need to be set up, with field representa-
tives and clerical staff required. Money put in to the regulatory beauro-
cracy would be better spent on direct alleviation of the problem, through
technical assistance and cost sharing. Third, the cost of annual waste
storage facilities may be prohibitive for the majority of small farmers.
Should forced regulation occur, many of these marginal enterprises could be
forced out of business.
Farmers, themselves, recognize the water quality problem and some of
them actually expect regulatory action. Those that have installed waste
management systems have done so with assistance from the county SCS, either
in the form of design assistance, cost sharing, or both. This funded,
volunteerism approach is a gradual means of improving water quality that
focuses on local initiative for problem solving. The strong selling points
of waste management systems, however, are the benefits such as reduced need
for fertilizer, lower annual fuel costs, and convenience. Improved water
quality is the desired goal, but to be widely accepted, waste management sys-
tems must be perceived as a significant contribution to the overall
farm operation.
76
-------
Cost sharing is available for liquid waste management systems but not
for anaerobic digesters, because digesters are not being viewed by nonpoint
planners as method of alleviating the water pollution problems. Regulations
and recommended practices affecting agricultural inputs to nonpoint pollution
may contribute to the adoption of anaerobic digesters by increasing the
number of liquid waste management systems. When digesters can be shown to be
an economical investment, based on energy production alone, as an addition
to the waste management systems, then those farmers who have installed the
waste management systems will be the likely ones to adopt anaerobic digesters.
This conclusion is based on the present course of planning for the 208 non-
point program and the attitudes of those involved in the program.
Utility Interest
Unlike other forms of renewable energy (solar, wind), methane gas con-
verted to electricity is not subject to variations in climate. The electri-
cal generation from a digester power plant is quite predictable and can
actually be controlled for seasonal peaking through manure storage systems.
Public utilities have expressed an interest in alternative energy
sources, if they can be done on a large scale. For smaller operations, the
electric utilities are accommodating alternate energy producing units by
purchasing power that is fed back into the line. Wind generators have been
the only ones in the North Central Region to date that have requested
this service. (72)
Safety is of prime concern to the power companies. When a storm knocks
the power out, there needs to be some assurance that the linemen will not
be zapped by numerous small generators feeding current into the line. If a
self-synchronous inverter with a DC generator is to be used, or if an AC
generator is involved, relaying must be incorporated which will recognize a
a situation where the generator is energizing a portion of the utility system
which is no longer connected to the main utility system and will interrupt
the back-feed.
Preliminary consultations with the power company in the region has
revealed that they are quite interested in alternate energy sources, and
that they fully intend to cooperate and provide technical support for such
connections. They have already expressed that the synchronous inverter
system would be designed, installed, owned, and maintained by the utility to
assure reliability and safety.
Also, if the farm could level their peak loads, or change the peaks so
they are not corresponding with the power company peaks, pricing mechanisms
could be worked out so it would be more beneficial for the farmer to send
current into the line at power company peaks. Should enough of these systems
be dispersed, there is a chance that the utilities would look to digesters
as load levelers to help meet their peaks.
77
-------
SECTION 9
MARKET FEASIBILITY
Anaerobic digesters, as indicated by the "Technical Components"
section, consist of a number of technical subsystems which require various
degrees of engineering, construction/installation services, and manufactured
components. Most of the equipment, such as pumps, energy conversion units,
boilers, piping, compressors, and hydraulics are readily available through
existing farm suppliers. Other components such as the fermentation tank can
be designed for modular construction, requiring minimum factory/warehouse
space. Most of the tank fabrication can occur in the field. Electrical
control panels can be easily fabricated in a small warehouse facility.
There are a number of small shops which specialize in subcontracting the
fabrication of these panels. Manure storage systems, the lagoon and influent
tank, are already found on many farms as part of liquid waste management
systems. If they are not already on site, the Soil Conservation Service
will provide design and, in some cases, cost sharing assistance.
The manure handling equipment has caused the most significant problem
of any digester component in the systems studied. The problem seems to stem
from lack of understanding and poor engineering rather than any inherent
lack of adequate equipment. There is manure handling equipment on the market
which, if properly sized, can move virtually any quantity or type of manure
solids concentration. Close work with manure equipment manufacturers by
the designer of the system would help to alleviate these problems.
There are a number of ways for a digester to be designed and installed.
Engineering is important. In an earlier section the need for standards
related to design and construction of these systems was outlined. Standards
and proper licensing of firms involved in digester design and construction
businesses would enhance marketability of the product. Currently, there are
approximately eight firms involved in the digester business. Most of these
provide both the engineering and construction services. Engineering, however,
could be specific and construction performed by local contractors.
Subsystems work could also be subcontracted to local manufacturers.
This could help alleviate many of the service problems that are currently
encountered by the firms in business now. It is crucial to both market
feasibility and to keep manufacturing costs to a minimum that the out of
town engineering firm work closely with the farmer in contracting a reput-
able service organization to keep the system in operating condition. This
could be done through a local contracting representative who is commissioned
to build the system.
78
-------
The potential farm market for anaerobic digesters includes all dairy
farms, poultry operations, beef cattle farms, large hog operations, and
meat packing plants with the equivalent of 125 animal units or greater. This
was seen as the lower limit to which digesters are, as yet, economical.
However, with improved conversion efficiencies, cost reductions, escalating
energy costs, and the availability of additional input material, such as
municipal solid waste or crop residues, this lower limit could become lower.
Coop systems involving several farmers would also increase the number of
systems that can be economically installed.
The Rice Lake digester is located in Barren County, part of the North-
west Wisconsin Crop Reporting District. (Figure 37) Of the 8,300 livestock
operations in the district, the greatest concentration of farms is found
in Barren, Chippewa, Polk, and Rusk Counties. Table 19 shows the numbers
and distribution of the different livestock operations. (69)
TABLE 19. NUMBERS OF FARMS IN THE FOLLOWING
WISCONSIN COUNTIES (NORTHWEST DISTRICT)
BARRON CHIPPEWA POLK RUSK
Total
Dairy
Beef
Hog
1,959
1,358
288
80
1,793
1,192
289
218
1,716
1,019
343
100
864
610
172
23
Barren County also contains a high concentration of turkeys, the majority of
which are in the ownership of one company.
The Statistical Reporting Service reports eight Barron County farms
with over 100 head of milking cows. Our data, gathered from Township
Assessors, generated closer to 50 farms with over 100 head. These include
all cows, however, not just milkers.
Initially, the use of these systems will also be restricted to the
more innovative farmers. Of the dairy farmers in Barron County there are
21 that can be considered innovative in terms of manure handling practices.
These 21 were the first to install liquid waste handling systems as pre-
scribed by the SCS. Interviews with the 21 farmers by outside sources, also
helped to characterize them as innovators. (62) As a group, these farmers
were better educated, had larger farm operations, earned more, and had more
community involvement than the average Barron County farmer.
Preliminary indications show anaerobic digesters being adopted for a
variety of reasons. Most evident is the rising cost of energy purchased for
79
-------
normal farm operation. Economics show the addition of a digester for
energy production alone, on this size farm, will require other changes in
farm operating procedures to enhance feasibility. Once an operating
digester can be seen by local farmers, and the implications of diversify-
ing operations becomes clear, more of these innovators will be considered
likely adopters.
Provided that a local service organization can be contracted to
maintain and perhaps build the digester the following two years could see
one digester per year installed in the county. Success of these installa-
tions could increase this rate to five per year after five years. At a
cost of approximately $40,000 per unit, the five year period would represent
a total Barren County sales volume of $560,000. With the adoption of stand-
ards for construction and installation, governmental tax incentives, and
assistance from farm agencies, such as extension, soil conservation services
and the USDA, digester adoption rates will be increased.
TABLE 2
YR. # OF DIGESTER SALES
INSTALLATIONS VOLUME
1
2
3
4
5
1
1
3
4
5
$ 40,000
40,000
120,000
160,000
200,000
TOTAL 14 $ 560,000
Extrapolations of this data to the remainder of the Northwest District
would show Chippewa County with about 18 digester installations after five
years and Polk County with close to ten. Total Northwest Wisconsin sales
volume for the five year period could total $1.5 million for dairy farms with
over 75 head. On a statewide basis, dairy farms could potentially represent
some $15 million in sales over the next five years.
One means of increasing feasibility and adoption of digesters for use
by greater numbers of farmers would be through the cooperative agreement of
several farmers or a township to invest in liquid waste management systems.
A centralized digester to provide power and gas could be sited so runs
by a large tank truck could be made to the farms on a regular basis. The
tank trucks would pick up one to two weeks of stored manure, truck it to the
digester, where gas and electricity would be produced, and return equivalent
amounts of slurry back to the holding ponds on the farmer's land. This type
of arrangement would be beneficial both for improving water quality and
alleviation of energy shortages. Electric utilities could help in construc-
tion and design of the power plant as it would be to their benefit to use
as a load leveling tool. If sited near an existing natural gas pipeline, the
gas company could purchase the gas from the coop for use in nearby towns
80
-------
where gas lines are prevalent. Arrangements could be made with participating
farmers to return the percentage of spent slurry and percentage of gas or
electricity provided by their animals' waste contribution. By working
closely with utility companies, the farms could arrange to adjust their
farming operations to coincide with off-peak electrical loads.
The engineering of such a system would be rather straight forward.
Digester designs are common civil engineering practice. Manure collection
would be coordinated similar to garbage collection routes and creamery
milk routing. The obstacles would be primarily management, legal problems,
and overcoming social barriers.
It seems that if a joint effort was made between the rural electric
coops, the soil conservation services, the participating farmers, with
cost sharing from other governmental agencies, such a plan would not only
prove feasible, but highly beneficial. Increased community self-relience
and uses of local resources to meet energy demands, along with soil and
water conservation would have long term benefits for the community and nation
as a whole.
Northwest Wisconsin
Figure 37. Northwest Wisconsin Crop Reporting District.
81
-------
REFERENCES
1. Jewell, W.J., et al., "Bioconversion of Agricultural Wastes for Pollu-
tion Control and Energy Conservation", Cornell University, Ithaca,
New York; NTIS # TID-27164; 1976.
2. Singh, R.B.; "The Bio-Gas Plant: Generating Methane from Organic Wastes";
Compost Science. Vol. 13:pp. 20-25. Jan-Feb 1972
3. Singh, R.B.; Biogas Plant Designs with Specifications. Rahstra Bhassha
Press, Raja Ki Mandi, Agra-2, India, 1973.
4. Po, C., Wang, H.H., Chen, S.K., Hung, C.M., Chang, C.I.; "Small Methane
Generator for Waste Disposal"; Managing Livestock Wastes. Proceedings
of the 3rd International Symposium, Urbanna, IL; pp 238-240, 1975.
5. Po, C.; "Production and Use of Methane from Animal Wastes in Taiwan";
Proceedings from the Biomass Energy Conference, Biomass Energy Insti-
tute, Winnipeg, Manitoba, 1973.
6. Merrill, R., Misser, G.C., Gage, T., Burkey, J.; Energy Primer. Portola
Institute, Menlo Park, CA, 1975.
7. Alternative Sources of Energy Magazine, Milaca, MN, 1974-1977.
8. McCarty, P.L., Jeris, J.S., McKinney, R.E., Reed, K.; Microbiology of
Anaerobic Digestion. MIT Press, Cambridge, MA, 19671
9. Toerien, D.E., Hattingh, W.H.; "Anaerobic Digestion: The Microbiology
of Anaerobic Digestion", Water Research. 3(6): 385-416, 1969.
10. McCarty, P.L.; "Anaerobic Waste Treatment Fundamentals - Part 4: Process
Design", Public Works. 95(12); 95-99, 1964.
11. Roll, J.L., Day, D.L., Pfeffer, J.T.; "Anaerobic Degradation of Swine
Manure with Municipal Digester Sludge"; Paper # 73-4525, American
Society of Agricultural Engineers, St. Joseph, MI, 1973.
12. Converse, J.C., Evans, G.W., Verhoevan, C.R.; "Performance of a Large
Size Anaerobic Digester for Poultry Manure", Paper # 77-0451, Ameri-
can Society of Agricultural Engineers, St. Joseph, MI, 1977
13. Fry, L.J. Merrill, R.; "Methane Digesters for Fuel, Gas, and Fertilizer";
New Alchemy Institute, Newsletter #3, Woods Hole, MA, 1973.
82
-------
14. Fry, L.J.; Practical Building of Methane Power Plants for Rural Energy
Independence, Standard Printing Co., Santa Barbara, CA, 1974.
15. Leckie, J., et al.; Other Homes and Garbage: Designs for Self-Sufficient
Living, Sierra Club Books, San Francisco, 1975^
16. Coe, W.B., Turk, M.; "Economic Evaluation of Asset Recovery from Animal
Waste by Anaerobic Fermentation"; Animal Waste Management Program,
Hamilton Standard Division, United Technologies, Windsor Locks, CN,
1973.
17. Converse, J.C., Graves, R.E., Evans, G.W.; "Anaerobic Degradation of
Dairy Manure under Mesophilic and Thermophilic Temperatures", Trans-
actions of the ASAE. 20(2); p. 336, ASAE, St. Joseph, MI, 1977.
18. Buhr, H.O., Andrews, J.F.; "The Thermophilic Anaerobic Digestion Process"
Water Research, 11(2); 129-143, 1977.
19. ECOTOPE GROUP; "Operation of a 50,000 Gallon Anaerobic Digester",
Quarterly reports submitted to the US Department of Energy, ECOTOPE
GROUP, Seattle, WA, 1978.
20. ECOTOPE GROUP, "Process Feasibility Study: The Anaerobic Digestion of
Dairy Cow Manure at the State Reformatory Honor Farm, Monroe,
Washington", ECOTOPE GROUP, Seattle, WA, 1975.
21. Pigg, D.L.; "Commercial Size Anaerobic Digester Performance with Dairy
Manure", Paper # 77-4055, ASAE, St. Joseph, MI, 1977.
22. Environetics Inc.; Bridgeview, IL, sales literature, 1977.
23. Ghosh, S., Klass, D.L.; "Two Phase Anaerobic Digestion", Symposium on
Clean Fuels from Biomass, Institute of Gas Technology, Chicago, IL,
~W7T.
24. Wong-Chong, G.M.; "Dry Anaerobic Digestion", Energy Agriculture and Waste
Management, W.J. Jewell, Ed.; Ann Arbor Science Publishers, Ann
Arbor, MI, 1975.
25. Augenstein, D.C., Wise, D.L., Wentworth, R.L., Cooney, C.L.; "Fuel Gas
Recovery from Controlled Landfill ing of Municipal Wastes"; Dynatech
Corp., Resource Recovery. No. 2, 1972.
26. Hashimoto, A.G., Chen, Y.R., Prior, R.L.; "Anaerobic Fermentation of
Animal and Crop Residues", US Meat Animal Research Center, Clay
Center, NE, 1977.
27. Biogas of Colorado, "Report to the Four Corners Regional Commission;
Energy Potential Through Bioconversion of Agricultural Wastes", Bio-
gas of Colorado, 5620 Kendall Ct., Arvada, CO, 1977.
28. Underground Space Center, Earth Sheltered Housing, University of
83
-------
Minnesota Press, Mpls, MM, 1978.
29. Straka, R.P., Nelson, G.H.; "Effect of Metal Containers on the Anaerobic
Fermentation on Cornstalk Flour by Thermophiles", Journal of Agri-
Cultural Research, 64(1): 19-30, 1942.
30. Biogas of Colorado, Methane on the Move, Biogas of Colorado, 5620
Kendall Ct., Arvada, CO, 1977. Also: Turnacliff, W., Personal
Communication.
31. National Tank and Pipe Co., "Wood Tanks for Industry", Engineering in
Wood, National Tank and Pipe Co., Clackamus, OR, 1977.
32. Jensen, A.M.; "Controliner Installation for Swine Wastes Lagoon",
Animal Science Dept., University of Illinois, Urbana, 1978.
33. Fischer, J.R., Sievers, D.M., Fulhege, C.O.; "Anaerobic Digestion in
Swine Wastes", Energy, Agriculture, and Waste Management, W.J.
Jewell, ed., Ann Arbor Science, Ann Arbor, MI, 1975.
34. Kroeker, E.J., Lapp, H.M., Schulte, D.D.; "Methane Production from
Animal Wastes 11: Process Stability", Paper # 76-208, Canadian
Society of Agricultural Engineers, Annual Meeting at Halifax, Nova
Scotia, 1976.
35. Lapp, H.M., Schulte, D.D., Kroeker, E.J., Sparling, A.B., Topnik, B.H.;
"Start-up of Pilot Scale Swine Manure Digesters for Methane Pro-
duction", Managing Livestock Wastes, Proceedings of the 3rd
International Symposium on Livestock Wastes, Urbana, IL, 1975.
36. Moore, J.A., Beehler, G.R., Horwath, N.J.; "Ammonia Volatilization from
Animal Manures", Biomass Utilization in Minnesota. Minnesota Poll-
ution Control Agency, Division of Solid Waste, 1977.
37. Benemann, J.R., Koopman, B.L., Baker, D.C., Weissman, J.C., Oswald, W.J.;
"A Systems Analysis of Bioconversion with Microalgae", Clean Fuels
from Biomass and Wastes, Symposium Papers, Institute of Gas Tech-
nology, Chicago, IL, 1977.
38. Boersma, L., Barlow, E.W., Miner, J.R., Phinney, H.K., Oldfield, J.E.;
"Protein Production Rates by Algae Using Swine Manures as a Sub-
strate", Energy and Agricultural Wastes, W.J. Jewell, ED., Ann
ArborScience Publishers, Ann Arbor, MI, 1975.
39. Golueke, C.G., Oswald, W.J.; "An Algal Regenerative System for Single
Family Farms and Villages", Compost Science, May-June, 1973.
40. Moore, J.D., Anthony, W.B.; "Enrichment of Cattle Manure for Feed by
Anaerobic Fermentation", Journal of Animal Science. Vol. 30,
p. 324, 1970.
41. Harmon, B.G., Day, D.L., Baker, D.H., Jensen, A.H.; "Nutritive Value of
84
-------
Aerobically or Anaerobically Processed Swine Waste", J. of Animal
Science, 37(2): 510-513, 1973.
42. Harmon, B.G., Day, D.L.; "Nutrient Availability from Oxidation Ditches",
Managing Livestock Wastes, Proceedings of the 3rd International
Symposium on Livestock Wastes, Urbana, IL, 1975.
43. Finney, D.C., Evans, R.S.; "Anaerobic Digestion: The Rate Limiting Pro-
cess and the Nature of Inhibition", Science. 190(12): 1088-1089,
Dec. 12, 1975.
44. Krampitz, L.O.; "Potentials of Hydrogen Production through Biophotolysis"
Clean Fuels from Biomass and Wastes, Symposium papers, Institute of
Gas Technology, Chicago, IL, 1977.
45. Mitsui, A.; "Long Range Concepts: Applications of Photosynthetic
Hydrogen Production and Nitrogen Fixation Research", Capturing the
Sun Through Bioconversion, Proceedings from Bioenergy Council,
Washington, D.C., 1976.
46. Mitusi, A.; "A Survey of Hydrogen Producing Photosynthetic Organisms in
Tropical and Subtropical Marine Environments", NSF, Semi-Annual
Report, 1975.
47. Mitusi, A. Ed., Biological Solar Energy Conversion, Academic Press, New
York, 1977.
48. International Conference on the Photochemical Conversion and Storage of
Solar Energy, University of Western Ontario, London, Ont.,
Canada, 1976.
49. Northern States Power Company, 1978.
50. Waukesha Engine Division, Dresser Industries, "Technical Data Sheet 5",
Waukesha, WI, 1975.
51. Vancura, R., Energy Harvest Inc., Sheaffer and Roland, Chicago, IL,
Personal Communication, 1978.
52. Coe, W.B., Hamilton Standard Division, United Technologies, Personal
Communication, 1978.
53. State of Wisconsin, Department of Natural Resources, "NR 200.12 Short
Form B., Agriculture", 1976.
54. Soil Conservation Service, Barron County, Wisconsin, 1978.
55. State of Wisconsin, Assembly Bill 1019, January, 1978.
56. Jequier, N.; "Low Cost Technology: An Inquiry into Outstanding Policy
Issues", Worldtech Report #2, Control Data, Corp., Mpls, MN, 1975.
85
-------
57. Abeles, T.P., Atkinson, P.; "Economic and Energy Considerations for
Anaerobic Digesters", Energy. Agriculture, and Waste Management,
W.J. Jewell, Ed., Ann Arbor Science, Ann Arbor, MI, 1975.
58. King, J.; "Survey of Barren County Farmers to Determine Attitudes toward
Manure Handling Practices", OASIS 2000, Rice Lake, WI, 1978.
59. American Mutual Insurance Alliance, "Products Liability", 20 W. Wacker
Dr., Chicago, IL, 1977.
60. Oswald, W.J.; "Productivity of Algae in Sewage Disposal", Solar Energy,
15:107, 1973.
61. Rutan, A.; The Do's and Dont's of Methane, Rutal Research, Stillwater,
MN,
62. Ernst, S.; "Role of Innovators in the Adoption of Liquid Waste Systems",
OASIS 2000, Rice Lake, WI 1978.
63. Cancian, F.; "Stratification and Risk Taking: A Theory Tested on Agri-
cultural Innovation", American Sociological Reviews, 1967.
64. Fliegel, F.C., Kirlan, J.; "Farm Practice Attributes and Adoption Rates",
Social Forces, 1962.
65. Kronus, C.L., van Es, J.C.; "The Practice of Environmental Quality
Behavior", Journal of Environmental Education, 1976.
66. Pampel, F. van Es, J.C.; "Environmental Quality and Issues of Adoption
Research", Rural Sociology, 1977.
67. Cowling, E.B., Kirk, T.K.; "Properties of Cellulose and Lignocellulose
Materials as Substrates for Enzymatic Conversion Process", Forest
Products Laboratory, USDA, Madison, WI, 1976.
68. Spano, L.A., Tassinari, T.H., Macy, C.F., Black, E.D.; "Pretreatments
and Substrate Evaluation for the Enzymatic Hydrolysis of Cellulosic
Wastes", Pollution Abatement Division, Food Sciences Laboratory,
U.S. Army Natick Research and Development Command, Natick, MA, 1977.
69. Wisconsin Statistical Reporting Service; "Assessor Farm Statistics",
State of Wisconsin, 1977.
70. Hayes, T.; Agricultural Engineering Department, Cornell University,
Personal Communication, 1977.
71. American Public Health Association; Standard Methods for the Examination
of Water and Wastewater. AMPH, 1015 Eighteenth St. N.W., Washington
D.C., 1976.
72. Northern States Power, Edward Glass, Director of Research, 1978.
86
-------
Appendix A
ANAEROBIC DIGESTION OF DAIRY COW MANURE PLUS THE ORGANIC FRACTION
OF MUNICIPAL SOLID WASTE: A PILOT FEASIBILITY STUDY
ABBREVIATIONS & TERMINOLOGY
ANOVA
C.O.D.
feedstock 1
feedstock 2
9
HRT
1
Ib
m
MSW
N
SRT
TS
TSA
TSW
VS
VS
retention time
analysis of variance
chemical oxygen demand
dairy cow manure only
dairy cow manure + MSW
gram
hydraulic
liter
pound
meter
municipal solid waste
elemental nitrogen
solids retention time
total solids
Total solids added
total sample weight
volatile solids
volatile solids added
volatile solids destroyed
SUMMARY OF FINDINGS
Completely mixed and continuously fed laboratory scale anaerobic diges-
ters (operating on a 16 day SRT) were used to determine the feasibility of
producing methane gas from a mixture of dairy cow manure (67% of the total
solids) plus the organic fraction of municipal solid waste (33% of the total
solids). Biogas production from this mixture was found to be 4.35 ft^/lb
VSA, or 17.7 ft3/lbVSD. Biogas production from the same set of digesters,
run on dairy cow manure alone, was found to be 4.56 ft-Vlb VSA, or 18.2
ft3/lb VSD. Although the difference in biogas production is slight an anal-
ysis of variance showed it to be statistically significant (p 0.0001).
Biogas composition from the dairy cow manure plus MSW mixture was found
to have a Cfy/CO^ ratio of 1.23 (*** 55% CH4); for the dairy cow manure only
phase the CH4/C02 ratio was found to be slightly higher at 1.34
87
-------
Chemical analyses of the influent and effluent during the dairy cow
manure plus MSW phase showed a reduction in total solids of 20.1%, a reduc-
tion in volatile solids of 24.6%, a reduction in chemical oxygen demand of
16.7%, and an increase in total Kjeldahl nitrogen (9.1%) and ammonia nitrogen
(19.3%).
Overall, the anaerobic digestion of a 2:1 mixture of dairy cow manure
plus MSW is nearly equivalent to the anaerobic digestion of dairy cow manure
alone. This finding is explainable by the fact that both dairy cow manure and
MSW have a low percentage of readily digestible organic matter; i.e., both
are composed of a high percentage of non-solubilized cellulose.
INTRODUCTION
Previous studies by Jewell (1,2) have shown that the use of anaerobic
digesters is not an economically feasible investment for dairy farms with
fewer than 125 cows. In Wisconsin, nearly 90% of the dairy farms have fewer
than 50 cows (3). Thus, the applicability of anaerobic digesters would appear
to be quite limited in the near future, at least until changes in a number of
factors (such as the price of natural gas, technological improvements in
digester design, and the number of larger size dairy farms) alter the feasib-
ility picture.
Another approach to the question of feasibility is on the basis of the
amount of organic material available to the farmer for digestion. On most
dairy farms the primary source will be dairy cow manure. The break even
point of 125 head of dairy cows represents 1325 pounds of manure total solids
production per day (1100 pounds of volatile solids per day) available for
digestion. Suppose that some fraction of this 1325 pounds of total solids
could be replaced with another type of organic material, which, when combined
with the manure, digests equally as well. Then the break even point, in
terms of number of cows, is lowered, depending on the quantity of that
"other" source of organic material in the digester feed.
One potential source of organic material is through the intentional
growing of field crops for digestion, perhaps a mini version of the Energy
Plantation scheme proposed by Intertechnology Corporation (4-9) amd others
(10, 11). Another potential source is the organic fraction of municipal
solid waste (MSW). The use of MSW in farm scale digesters would not only
improve the feasibility of digesters for dairy farms with fewer than 125
cows, but it would also attenuate the ever growing problems of MSW disposal.
Objective of this Study
The objective of this study was to compare the anaerobic biodegradability
of a 2:1 mixture containing dairy cow manure plus the organic fraction of
municipal solid waste with the anaerobic degradability of dairy cow manure
alone.
-------
MATERIALS AND METHODS
Apparatus
Thirteen laboratory scale anaerobic digesters were used to carry out
this experiment. Each digester was completely mixed, vertically placed, and
continuously fed (on a slug basis, i.e. one load/discharge per day). The
reactors consisted of one liter glass jars (8.0 cm in diameter, 17.0 cm in
height), sealed at the top with a #14 rubber stopper. Three glass tubes
passed through the stopper; one for extracting effluent and adding influent,
a second for biogas sampling, and a third for allowing the biogas to be
discharged (See Figure Al)
t
•quartern
b«ih
I
1
fut
0 5
Figure Al: Reactor
In order to maintain the digesters in the mesophilic temperature range,
they were kept inside a water bath (rectangular aquarium) with a heater/
thermostat, set at 35°C + 2°C. Air was continuously bubbled through the
water bath to ensure an even distribution of the heat.
Mixing of the digesters was accomplished with magnetic stirrers. A
5.2 cm long teflon stir bar was placed on the bottom of each reactor; the
entire aquarium rested on top of the 13 stirrers. Off-center drift of the
stirring bar is often a problem when a material as viscous as animal manure
is mixed. To help prevent this, the water bath was fitted with a "grid"
structure which held the reactors directly on-center in relation to the
stirrers (see Figure A2). Also, the small size of the reactors made stirring
that much easier.
An electric timer switch with a one-hour cycle was used to turn the
89
-------
stirring units on for one minute every hour.
12
13
10
Diogas sampling port
Figure A2: Top view of water bath
grid
structure
••aquarium
level I I
reactor
8
10
H
aquarium
leva! II
manometers
250 ml gas buffer bottles
pressure adjusting tubes
level III
water
Figure A3: Front view.
Measurement of the biogas produced was accomplished by displacement of
water. Figure A4 details the biogas measurement system. As the biogas
collected in the space above the digesting material, it flowed into a 250
ml glass bottle (5.0 cm in diameter, 12.8 cm in height; fitted with a three-
holed #10 rubber stopper) used to create a "buffer zone" in the gas system.
This buffer zone was needed so that, when gas samples were withdrawn from the
digesters, the pressure would not be lowered to a strong vacuum. One tube
leading from this container was attached to a water-filled manometer, having
a maximum/minimum pressure range of 5.0 cm water column. The other tube
was attached to the water reservoir container, from which water was displaced.
The 13 water reservoir containers consisted of one-gallon glass bottles
(each fitted with a two-holed #6 rubber stopper). Note that before the biogas
enters the atmosphere of the bottles, it must first bubble through a test
tube partially filled with water. This test tube separated the atmosphere
above the digesters from that inside the bottles. Thus, when the bottles
were opened for refilling with water (once every four days), the atmosphere
above the digesters was not exposed to oxygen.
As biogas collected in the one-gallon bottles, it displaced the water
through a siphon into one-liter collection bottles. Notice in Figure A4
that the siphon had a U-shape bend in it which prevented the siphon from
being broken in case air bubbles became entrained in the tube. The "pressure
adjusting tubes" were used in maintaining a relatively constant pressure in
90
-------
the atmosphere above the digesters; these tubes were adjusted daily according
to the water level in the one-gallon bottles. As the water level in the one-
gallon bottles dropped, the pressure adjusting tubes were moved downward.
The entire apparatus was contained on a moveable cart (see Figure A3)
with three levels: one for the water bath and magnetic stirrers; a second
for the manometers, buffer zone bottles, water reservoir bottles, and pressure
adjusting tubes; and a third for the water collection bottles.
Mo««« outtat
IIIM
tlpnon
Figure A4: Biogas measurement system.
Analytical Methods
Biogas Volume. To convert mis of water displaced into mis of biogas
produced at standard temperature and pressure, the following equation was
used:
X = V(pb - 20.1), where:
760
X = mis of biogas produced at STP
V = mis of water displaced
Pb = barametric pressure, mm Hg, measured once daily
20.1= vapor pressure, mm Hg, of water at 22.2° C.
(approximate room temperature)
760 = standard pressure, mm Hg
Each of the 13 daily readings for mis of water displaced was corrected.
91
-------
These values were also corrected according to the time of day that the read-
ings were taken. For example, if the time between two consecutive readings
was 24.5 hours, then each of the values was multiplied by 0.980 (i.e.,
24/24.5); this put the biogas production measurement on an hourly basis.
Biogas Quality. Analyses of biogas samples were performed on a Carle
Instruments Model 8000 Basis Gas Chromatograph, linked with a Heath Kit Servo
Recorder Model #EU-20B. The gas chromatograph was equipped with a two ml
sample loop and valve, as well as a column switching valve. Two columns were
used in each analysis: 1) Poropak Q (10' x 1/8", 80/100 mesh), used to
separate methane, carbon dioxide and air, Helium (used as the carrier gas)
flow rate through this column was 38.8 ml/min; and 2) Mol Sieve 5A (61 x
1/8", 80/100 mesh), used to separate the air peak into oxygen and nitrogen;
Helium flow rate through this column was 43.7 ml/min. The detector was kept
at a constant 83QC. Chart speed for output from the Porapak was 3"/min.;
for the Mol Sieve 5A it was 2"/min.
On each sampling day, four samples were taken per digester; two were
injected through the Porapak Q side, and two through the Mol Sieve 5A side.
The percentage of each gas present was calculated according to the area under
each peak, measured by the formula (height) x (base at one-half the height).
Using a set of six standards from Scientific Gas Products, calibration curves
were calculated for each gas.
pH. A Corning Model 7 pH meter coupled with a Corning #476054 glass
combination electrode was used to determine pH.
Solids, Nitrogen, C.O.D. The Determination of total and volatile
solids (TS and VS), total Kjeldahl nitrogen (TKN), ammonium plus free ammonia
( (NH4 +•-NHs) -N), and chemical oxygen demand (C.O.D.) were performed
according to Standard Methods for the Examination of water and wastewater
(12). Organic nitrogen concentrations were calculated by subtraction. Samp-
les analyzed for C.O.D. had to be diluted first (83.33X) in order to bring
them into the range specified for analysis (i.e., not greater than 2000
mg/1).
Computer Analyses. The analysis of variance of biogas production data
was performed by the NWAY1 program of STATJOB, on the Univac 1110 computer
of the University of Wisconsin-Madison. All other statistical analyses
and computer graphics were performed on the University of Wisconsin-Green Bay
Xerox Sigma 6.
EXPERIMENTAL DESIGN FOR STATISTICAL ANALYSIS OF BIOGAS PRODUCTION DATA
An experimental design utilizing analysis of variance was developed in
order to compare biogas production from dairy cow manure (labeled feedstock
1) versus production from a mixture of dairy cow manure and MSW (labeled
feedstock 2).
The basic design was to operate the 13 digesters first on dairy cow
manure only until some sort of consistent biogas production was reached.
92
-------
Then, all the digesters were to be switched to the mixed feedstock. After a
sufficient acclimation period (at least 1 full SRT), the digesters would be
operated until enough data was collected to make possible a comparison with
the feedstock 1 phase.
This design required the investigator to make several important choices
based on an examination of the data:
1) How could consistent biogas production in the feedstock 1 phase
be determined?
2) When should the switch in feedstocks be made?
3) For how long a period should the digesters be run on feedstock 2
before the project is terminated?
4) How many days from each feedstock phase should be chosen for
comparison?
The following criteria were developed as an aid in making these
decisions:
1) Consistent biogas production means:
a. at least 10 consecutive days of no substantial variation in
the overall mean; a substantial variation means a "wild fluctuation", char-
acteristic of some sort of "line out"; and
b. the absolute quantity of biogas production (ft^/VS^j must be
reasonably consistent with previously reported values for dairy cow manure.
2) The digesters must be operated with feedstock 2 for a period equal
to or greater than their SRT (16 days) before considering use of the biogas
production data for the ANOVA.
3) The design must be balanced, meaning that an equal number of days
in each phase of operation must b£ selected for comparison.
The specific ANOVA model used to make the comparison between feed-
stocks 1 and 2 was:
Yijk = H + Ai + Bj + e-jj + Ck + ACik + BC-jk + Sjjk
where: Yijk = biogas production of the i th digester, for the j th
feedstock, on the k th day
H = common mean
Ai= effect of the i th digester, i = 1, . . . , 13
Bj= effect on the j th feedstock, j = 1, 2
Ck= effect of the k th day, k = 1, . . . , 14
eij NID (0, oe) sijk MOD (0, o2)
Thus, biogas production is the dependent variable; the digesters,
feedstock and time are the independent variables.
93
-------
This design is based on the assumption that, except for a change in the
type of feedstock used, all other environmental conditions affecting per-
formance of the digesters would remain constant. These conditions include:
1) temperature of the water bath; 2) mixing of the digesters; and 3) con-
sistency within a feedstock phase of the quality of the feed, in terms of
such parameters as TS, VS and C.O.D.
PROCEDURE
On October 1, 1977, the 13 digesters were seeded with 800 m:ls of a very
dilute slurry containing dairy cow manure which had previously been anae-
robically digested. A gas space approximately 150 mis was left at the top of
the reactors. The digesters were allowed to acclimate for a period of 12
days. On October 12 (hereafter called day 1 as a reference point; see
Appendix A2), daily feeding/discharge began as show in table Al, the
digesters were fed increasingly greater amounts until a steady-state input/
output of 50 mis was reached on day 11; building up the feeding/discharging
quantity slowly was intended to prevent shocking of the digesters.
Table Al: Feed/Discharge Quantity
Day # mis of feed/discharge
1 10
2 12
3 14
4 16
5 18
6 20
7 25
8 30
9 36
10 42
11 50
12 50
82 50 .project terminated
Feed/Discharge Procedure
The feed/discharge quantity of 50 mis chosen for this study represents
a solids retention time (SRT) of 16 days (i.e., 50 mis in 800 mis of digester
slurry). Since the digesters were completely mixed, the hydraulic retention
time (HRT) was also 16 days.
Each day, at approximately the same time (between 12:00 p.m. and 2:00
94
-------
p.m.), recordings were made both of the pressure in the atmosphere above the
digesters and of the amount of water displaced by biogas production. Then,
50 mis of slurry were extracted from each digester and replaced with 50 mis
of new feed. A 25 ml wide mouth serological pipette was used in this process.
Note that the feed/discharge tube extends below the level of slurry in
the digester; thus, only a small area of the slurry was exposed to air
during the feed/discharge process.
Influent Preparation: Feedstock 1
Dairy cow manure (feces and urine) was collected every five days from
a local dairy farm, containing a herd of 22 Holsteins kept in a stanchion
barn. Straw bedding is used in their gutters, but the straw was not
collected with the manure. Approximately eight pounds of manure was taken
per visit. On the same day, the manure was mixed with water (added one to
one, by volume) to attain a slurry of 7% to 10% total solids. Because a
pipette was used in the feed/discharge process, the slurry was blended in
a high-speed commercial Waring Blender for about five minutes. The blending
action created substantial amounts of foam, some of which was removed prior
to use of the slurry as feed. When not in use, the blended slurry was placed
in a cooler maintained at 4°C. New feed was prepared every five days.
No samples of the dairy cow manure influent were taken during this
study; thus, it is not possible to state with complete certainty that the
total solids content was 7% to 10%. However, in a previous study made by
this investigator (13), manure was collected from a similar farm and prepared
in the same manner (i.e., diluted 1:1 by volume). For 13 of the influent
samples taken, the average total solids content was 7.6%, with a range of
6.1% to 9.3%.
Influent Preparation: Feedstock 2
Feedstock 2 contained two components: dairy cow manure and municipal
solid waste (MSW). The dairy cow manure was prepared in the same manner as
it had been for feedstock 1. On September 26, 1977, approximately 25 pounds
of MSW were collected from two locations: Appleton, Wisconsin, and Madison,
Wisconsin. Both cities use hammermills to shred their commercial and resi-
dential wastes; ferrous metals are then removed magnetically (with about
85% recovery). The MSW was combined in one container and, when not in use,
stored in a cooler maintained at 4°C.
Additional processing of the MSW shred was necessary before it could be
used in the laboratory digesters. Just prior to the start of its use, the
shred was manually inspected and all large pieces of metal, plastic, rags,
and cardboard were removed. The primary criterion used in the sorting
process was to exclude any large and/or plastic and aluminum pieces which
might damage the blender. Approximately ten of the original 50 Ibs. were
excluded from use.
95
-------
Actual preparation of a feedstock 2 mixture was as follows: 1) One
gallon of 7% to 10% dairy cow manure slurry was prepared, just as for feed-
stock 1. 2) This slurry was weighed and the quantity of dairy cow manure dry
solids (TS) was calculated, assuming 8.5% TS in the slurry. 3) An approxi-
mately equal wet weight of MSW shred (i.e., = to the dairy manure dry solids)
was added to the feedstock 1 slurry. 4) Water was added to this new mixture
until a "milk-shake" consistency was achieved. This additional water was
needed to account for the addition of MSW solids. 5) The dairy cow manure +
MSW mixture was then blended for approximately five minutes. New batches were
prepared every five days; a total of seven were prepared.
Thus, feedstock 2 contained a mixture of 50% dry weight dairy cow
manure solids and 50% wet weight MSW shred solids. To determine what this
ratio was on a dry weight to dry weight basis, an analysis of the MSW shred
was made, with the results shown in Table A2:
Table A2: Composition of MSW Shred - Combined Sample
from Appleton, WI., and Madison, WI.
% Moisture
%TS
% VS of TS
% VS of TSW
% FS of TSW
X
40.2
59.8
66.0
20.4
20.4
X
3.87
3.87
4.37
3.60
3.60
*standard deviation, from 3 replicate samples
If the %TS in MSW shred was 59.8%, then the ratio of dairy manure
solids to MSW solids was 1.0/0.598, i.e. 1.67/1.00. In other words, feed-
stock 2 contained 63% dairy manure solids, 37% MSW shred solids, on a dry
weight/dry weight basis.
Assuming that 83% of the dairy manure total solids was volatile (1),
then the ratio of dairy manure volatile solids to MSW shred volatile solids
was (0.83) (1.67) / (0.66) (1.00), i.e. 2.10/1.00. This translates to 68%
dairy manure volatile solids, 32% MSW shred solids. Table A3 summarizes
this information:
Table A3: Composition of Feedstock 2
Percentage
Dairy Cow
Dry Weight -
Dry Weight -
Solids
Volatile Solids
Manure
63%
68%
MSW
37%
32%
Ratio
(Dairy Cow Manure)
MSW
1.67/1.00
2.10/1.00
Biogas Sampling
Starting on day #2, and every seventh day thereafter, samples of biogas
96
-------
above each digester were withdrawn through the rubber septum (see Figure 2)
and immediately injected into the 2 ml sample loop of the gas chromatograph.
A 5 ml gas tight syringe was used to take the samples. The septums were
replaced at the midpoint of the project.
Influent and Effluent Sampling
On the same day that biogas analyses were carried out, the pH of the
effluent from each digester was measured.
In the period from day 57 through day 81, five influent and nine
effluent samples were taken and frozen. During January, 1978, they were
analyzed for total and volatile solids, total Kjeldahl nitrogen, ammonium
and free nitrogen, and chemical oxygen demand. The influent samples were
representative of a particular batch of feedstock 2, used over a five day
period, at five day intervals. The effluent were a composite of all 13
digesters for one particular day (i.e., the effluent from all 13 digesters
was mixed together and a sample from this was frozen).
Table A4 provides a summary of most of the background information
provided thus far:
Table A4: Summary of Background Information
Digester Specifications
# of digesters used 13
height of reactor 17.0 cm
diameter of reactor 8.0 cm
total useable volume 950 ml
slurry volume 800 ml
gas space 150 ml
mixing 1 min./hr.
temperature 35°C + 2°C
steady state feed/
discharge quantity 50 mis
SRT 16 days
Dates
duration of project 82 days (Oct. 12, 1977 to June 1, 1978)
# of days of feedstock 1 47 (day 1 thru day 47)
# of days of feedstock 2 35 (day 48 thru day 82)
# of days used in ANOVA
feedstock 1 14 (day 34 thru day 47)
feedstock 2 14 (day 69 thru day 82)
RESULTS
Biogas Production Data
All 13 digesters were operated on feedstock 1 from day 1 thru 47, and on
feedstock 2 from day 48 thru 82. Table A5 gives the mean, standard deviation,
97
-------
and range of biogas production for each day; Figure A5 displays the means
graphically.
Following the initial start-up period, the biogas production leveled
off starting on day 13, and remained consistent at approximately 452 mis/day
until day 27, when a "wild" fluctation began. Production leveled off again
starting on day 24, but at a definitely higher level. Thereafter, overall
production was quite consistent right through the day before the influent
was changed to feedstock 2.
The first loading of feedstock 2 caused a definite drop in the average
biogas production, a predictable response to a new feed, considering the fact
that the percentage of MSW shred was not increased gradually. However,
production increased the next day (49) and remained stable through day 59
when a sharp drop for 2 days was followed by the largest average production
for a single day for the entire project. The amplitude of this fluctuation
was probably aggravated by the fact that between day 60 and 61, the stirrers
were accidentally left off; when turned on again, production boomed the next
day. Beginning with day 64, one full SRT after the switch in feedstocks,
production leveled off and remained fairly consistent. The digesters were
operated on feedstock 2 for 18 more days (through day 82) before the project
was terminated.
An inspection of the graphs of biogas production for each individual
digester showed one phenomenon fairly common to each: spiking. That is, one
or two days of increasingly higher production, followed again by a drop.
This spiking cycle was common for production across both feedstocks. One
might expect (on the basis of a reduction in variance by a factor of 1/13)
that an averaging of the data for each day would result in a series of means
showing a much lesser degree of spiking. However, this is not what happened
(see Figure A5). The amplitude of the spiking appears to be every bit as
great for the means as for each digester.
ANOVA
Using the criteria outlined in section C, two sets of data (one each
from the different feedstock phases) covering 14 day periods were selected
for analysis of variance. For feedstock 1, days 34-47 were used; for feed-
stock 2, days 69-82 were used. Results from the ANOVA and table of two-
way means are shown in Tables A6 and A7, respectively.
The ANOVA shows that the effect of time on biogas production is statis-
tically significant (p 0.00.), meaning that the digesters behave differ-
ently from day to day. However, because the interaction of time x feedstock
is also statistically significant (p 0.001), the averages for each 14 day
period should not be combined. In other words, biogas production in the
feedstock 2 phase was lower than in the comparable phase for feedstock 1.
Table A7 shows the grand mean for feedstock 1 to be 583 mis/day; for
feedstock 2 it is 568 mis/day, a difference of 15 mis/day.
98
-------
TABLE A5: BIOGAS PRODUCTION DATA - AVERAGE FOR 13 DIGESTERS
FEEDSTOCK 1
Day No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
X
11
43
59
83
93
144
147
165
225
280
267
366
445
470
384
438
430
466
436
512
437
483
387
463
519
455
688
627
582
554
428
413
453
535
643
549
627
641
518
612
507
593
639
524
581
623
569
_**-
6.2
19.5
9.9
12.1
13.9
11.5
17.2
14.3
19.2
28.8
28.4
18.2
35.7
27.7
28.1
25.4
33.7
34.1
25.4
56.3
32.0
27.4
27.9
26.8
33.7
31.5
99.3
42.3
48.4
39.7
27.8
21.8
33.0
33.3
34.4
32.1
41.1
26.7
24.1
16.9
48.1
21.8
19.6
28.2
25.7
22.6
34.6
Range
20
63
39
49
55
38
76
54
85
122
95
68
122
93
83
82
98
96
97
252
112
84
90
104
119
98
311
121
183
114
86
70
110
123
143
96
147
80
82
64
178
77
57
105
99
71
105
Day No.
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
FEEDSTOCK 2
X
450
645
559
611
586
516
573
673
605
562
599
524
455
385
731
478
576
618
495
558
570
494
572
589
509
582
615
489
532
512
627
532
624
640
638
s—
42.0
88.5
46.4
42.2
32.8
44.5
22.3
30.8
31.9
40.3
49.4
26.7
20.8
21.4
89.3
38.1
32.4
24.3
37.2
21.4
38.4
40.7
33.4
32.3
34.1
24.3
32.9
32.4
23.3
32.6
24.7
27.7
36.2
23.7
34.8
Range
158
287
163
145
117
185
77
127
100
142
167
87
73
76
271
108
110
94
134
74
167
150
124
107
107
83
114
113
85
115
85
86
125
71
124
99
-------
o
o
Q
O
CC
03
O
1—4
CO
FEEDSTOCK 1:
DAIRY COW MANURE ONLY
T
27
FEEDSTOCK 2:
DAIRY COW MANURE + MSW MIX
Note: Enlarged points are
those used in ANOVA.
I
.L-.
45 54
_63
DAYS
800
700
600
500
400
300
200
100
FIGURE A5. AVERAGE BIOGAS PRODUCTION FOR 13 DIGESTERS
-------
TABLE A6: ANALYSIS OF VARIANCE OF BIOGAS PRODUCTION DATA
Source
d.f.
s.s.
m.s.
F.
Digesters
Feedstock
Error (a)*
Time
Time x Feedstock
Time x digesters
Error (b)*
Total
12
1
12
13
13
156
156
363
71,996.1
18,973.6
15,763.6
539,106.9
397,730.3
137,780.4
94,613.6
1,275,968.7
5,999.7
18,973.6
1,313.6
41,469.8
30,594.6
883.2
606.5
4.567
14.444
68.376
50.445
1.456
* Error (a) = Digester x Feedstock
Error (b) = Time x Digester x Feedstock
TABLE A7: A COMPARISON OF MEANS, BY FEEDSTOCK, OVER TIME
TIME*
Feedstock
1
2
x
T
535
494
515
2
643
572
608
3
549
589
569
4 5
627 641
509 582
568 612
6 7
518 612
615 489
567 550
8
507
533
520
9
593
512
553
10
639
627
633
n 12
524 581
534 625
528 603
13 14
623 569
640 638
632 603
x
583
568
576
* Time 1 in this table corresponds to day 34 for feedstock 1, day 69 for
feedstock 2; time 14 corresponds to day 47 and 82, respectively.
The statistical significance of the time x feedstock interaction is not
too surprising because two distinctly different 14 day periods were compared.
However, the time x digester interaction is not statistically significant
(p 0.001); meaning that the behavior of the digesters as a group varies
similarly over time; i.e., the digesters are not acting autonomously--each
changing in its own way; rather, the digesters change over time as a group.
Thus, the change in biogas production over time cannot be attributed to
random variations among the digesters. This suggests that the changes are
caused by the environment of the digesters, which includes not only the
feedstock but also several other independent variables not entered into the
analysis; namely the consistency of the influent over time (as measured by
parameters such as %TS, %VS, C.O.D.).
101
-------
Biogas Composition
Biogas composition was not significantly affected by the change in feed-
stocks, as shown in Table A8 and Figure A5. Once steady-state operation was
reached, the percent methane ranged from 50.5% to 55.4% for the duration of
the project, with an overall average of 53.3%. During the feedstock 1
operation, the average percent methane for all 13 digesters was 54.1%; for
the feedstock 2 phase, the average was 52.3%.
In all of the analyses, an air peak, split into nitrogen and oxygen,
was calculated. However, in a laboratory experiment headed by Klien (14),
tests demonstrated that virtually all the air present in their biogas
collectors resulted from contamination during feeding and was not a product
of the digestion process. All air was found to be expelled from the digesters
within ten minutes after feeding, so interference with the digestion process
was/ negligible.
In this study, the %No and %02 will be reported; however CH4/C02 have
also been calculated, which discounts the presence of air. For the feed-
stock 1 phase, the ratio of CH4/C02 was 1.34; for the feedstock 2 phase,
it was 1.23.
Table A8: AVERAGE BIOGAS COMPOSITION FOR 13 DIGESTERS
Day No:
2
9
16
23
30
37
44
51
58
65
72
79
x, 9-79
x, 9-44*
x, 51-79*
CH4/CO?
2.62
1.45
1.41
1.38
1.21
1.27
1.31
1.13
1.15
1.26
1.34
1.28
1.29
1.34
1.23
%CH
-------
Influent/Effluent Analyses
pH. All of the digesters, except #12 and #13, showed an acid pH at the
onset of the experiment (day 2). Once steady-state operation was reached,
pH in all of the digesters became slightly above neutral. Throughout the
remainder of the project (day 9-82), the pH of each digester fell in the
range of 7.0 to 7.4, with the overall mean closest to 7.2. The switch from
feedstock 1 to feedstock 2 appeared to have no effect on pH.
Total and volatile solids. Samples of influent and effluent were
analyzed for TS and VS only during the feedstock 2 phase, from day 57-82.
Results are summarized in Table A9 and Figure A6.
For the five influent samples analyzed, the TS averaged 4.93%, ranging
from 3.91% to 5.41%. This is a surprising result because the influents
appeared to be very viscous after they were homogenized in a blender. The
value of 4.93%TS is well below the desired range of 7% to 10% TS for an
influent.
Figure A6 shows that after an initial drop in influent %TS, there was
a very definite increase in the last three batches. This trend is signifi-
cant because it appears to correspond to a gradual increase in biogas prod-
uction for the same period (see Figure A6).
The %VS in the influent closely followed the %TS; for the five batches
sampled, 84.7% of the total solids were volatile (4.18% VS of TSW). This
value is higher than the corresponding one for the dairy cow manure, which
normally averages 83% VS of TS. Organic loading rate was calculated to be
an average of 0.11 Ib VS/ft3 day -1 (1.72 g VS/1 day -1).
Total solids in the effluent decreased steadily in the period when
samples were taken, as did the volatile solids. The overall mean for nine
samples was 4.17% TS, 3.45% VS (82.6% VS of TS). Four of the samples (days
70, 73, 78 and 81) fall into the same time frame as the biogas production
data used in the ANOVA (day 69-82). Because more than one full SRT had
passed from the time these four samples were taken, they are probably more
informative than the average for all nine samples. The mean %TS in these
four was 3.94%; the mean %VS was 3.23% (81.9% VS of TS).
Volatile solids reduction increased over time, indicating improving
efficiency in digestion as the reactors became acclimated to feedstock 2.
The overall %VS destruction was 17.0%, ranging from 9.5% to 31.7%. For the
12 day period, day 70-81, the average was 24.6% VS destroyed.
Using the average value discussed above, it was possible to calculate
several important values shown in Table A10.
Thus, when biogas production is measured on the basis of ft^/lb TS/\,
VS/\, and ft3/lb VSp the values are slightly higher for feedstock 1
than for feedstock 2.
103
-------
TABLE A9. SUMMARY OF RESULTS - CHEMICAL ANALYSES
INFLUENT
Date Range
from
12-7-77
12-12-
12-17-
12-22-
12-27-
to
12-11-77
12-16-
12-21-
12-26-
12-31- _
Day No.
from to
57 61
62 66
67 71
72 76
77 81
x
(NH4+-NH3)
mg/l-N
87.9
310
123
584
152
251
Organic
mg/l-N
608
810
907
956
948
846
N TKN
mg/l-N
696
1120
1030
1540
1100
1100
ZVS of
ZTKN as C.O.D. ZTotal ZVS Total
(NH/^-NHi) mg/1 Solids of TS Sample
13
28
12
38
14
21
51,000 4.93
52,000 4.60
53.000 4.73
52.000 4.99
60.000 5.41
54.000 4.93
87.
84.
84.
83.
83.
84.
19 4.29
89 3.91
40 4.00
78 4.18
27 4.51
71 4.18
Loading Rate
Ibs
(R
0.
0.
0.
0.
0.
0.
VS/ft^/day
VS/l/day)
11 (1.77)
10 (1.61)
10 (1.64)
11 (1.72)
12 (1.86)
11 (1.72)
EFFLUENT
Date
12-7-77
12-9-
12-13-
12-15-
12-17-
12-20-
12-23-
12-28-
Day No.
57
59
63
65
67
70
73
78
12-31- 81
Overall x
x for 70-81
(NllA+-NH3)
mg/l-N
193
178
149
188
226
228
284
375
356
242
311
Organic N
mg/l-N i
1020
1000
981
962
944
932
906
885
874
945
899
TKN
ng/l-N (
1210
1180
1130
1150
1170
1160
1190
1260
1230
1190
1210
ZTKN as
NH/+-NH-)
16
15
13
16
19
20
24
30
29
20
26
C.O.D.
mg/1
56,000
54,000
56,000
50,000
49,000
45,000
45,000
42,000
46,000
49.000
45,000
ZC.O.D. ZTotal
Reduction Solids
4-60
4.50
4.27
3.9 4.26
7.6 4.17
15 4.06
13 4.10
30 3.79
23 3.82
15 4.17
20 3.94
ZVS
ot TS
83.11
85.05
82.79
82.83
82.33
82.23
82.28
81.55
81.33
82.61
81.85
ZVS of
Total ZVS
Sample Destroyed
3.83 10.7
3.82
3.54
3.53
3.44
3.34
3.38
3.08
3.11
3.45
3.23
11.0
9.5
9.7
14.0
16.5
19.1
31.7
31.0
17.0
24.6
-------
5.5
5.0
4.5
4.0
3.5
3.0
ts= total solids
vs= volatile solids
5.5
5.0
4.5
4.0
3.5
3.0
time (no. of days)
FIGURE A6. TOTAL AND VOLATILE SOLIDS - FEEDSTOCK 2
Nitrogen. TKN values for the influent varied substantially, ranging
from 696 mg/1 to 1540 mg/1. Ammonia levels in the influent closely followed
the TKN. For the five samples analyzed, TKN averaged 1100 mg/1, (NH4 +
-NH3) -N averaged 251 mg/1 (see Figures A7 and A8.).
The effluent TKN and ammonia followed a much more consistent pattern
than the influent. Most significantly, Figure A8 shows a definite increase
over time of the percentage of TKN in the inorganic form, suggesting improved
efficiency of digestion farther into the feedstock 2 phase. This pattern
correlates well with the decrease over time of %TS and %VS (see Figure A7),
as well as C.O.D. (see below).
For the period day 70-81, effluent TKN averaged 1210 mg/1, (NH4 +
-NH3) -N averaged 311 mg/1.
Chemical Oxygen Demand. Figure A9 shows the change over time in C.O.D.
concentrations for the influent and effluent. The average influent value
was 54,000 mg/1. Values for dairy cow manure slurry have been reported in
the range of 70,000 mg/1 to 115,000 mg/1 (13). The discrepancy is explain-
able by the relatively low %VS in feedstock 2, since %VS and C.O.D. are
directly related.
Initially the effluent C.O.D. exceeded the influent C.O.D., probably
105
-------
Table A10: COMPARISON OF BIOGAS PRODUCTION EFFICIENCY
measured biogas production
mis of biogas/ml influent
%TS in influent
%VS of IS in influent
%VS destruction
ft3 biogas /lb TSA
(m3/Kg TSA)
ft3 biogas /lb VSA
(m3/Kg VSA)
ft3 biogas /lb VS,.
(m3/Kg VSD)
FEEDSTOCK 1
11.7 *
7.6 ?*
83.0 +
24.0 *
3.79
0.236
4.56
0.285
18.2
1.14
FEEDSTOCK 2
11.4 *
4.93
84.7
24.6
3.69
0.229
4.35
0.270
17.7
1.10
* based on the average biogas production results from the ANOVA
? no samples of influent or effluent were taken in the feedstock 1
phase; these values represent averages from a previous study(13),
in which the same type of influent was used; they also are con-
sistent with results from other, similar studies (1).
I
mg
I
CO
<
LU
O
o
ae
K-
i
§
TM FLUE NT
6O
T1ME (DAYS)
80
Figure A7. Ammonia and TKN - Feedstock.
106
-------
—I.
IQ
C
T
n>
c
-$
n>
gC.O.D.
mg/|
o
o
O)
n> —
O
n>
n>
o>
CL
o
o
3
00
o
CO
-a
n>
-i
o
n>
3
rl-
<-*-
3-
n>
Co
v—.
i
PERCENT
K
m
CO
co
tn
n>
d>
a.
1/1
r»-
O
O
ro
-------
because of a build-up of organic material following the switch to feed-
stock 2. However, the effluent C.O.D. dropped consistently after day 63,
before leveling off at approximately 45,000 mg/1. The % C.O.D. reduction
averaged 20% for the period day 70-81.
Odor. A homogenous b>end of dairy cow manure and the organic fraction
of municipal solid waste has a very offensive odor, much more so than dairy
cow manure alone. Following digestion of feedstock 2, the reduction in odor
was substantial, indicating stabilization of much of the volatile organic
compounds.
Table All provides a summary of the major findings of this study.
TABLE All: SUMMARY OF RESULTS
BIOGAS PRODUCTION & COMPOSITION
ft3/lb VSA
ft3/lb VSp
CH4/C02
FEEDSTOCK 1
4.56
18.2
1.34
FEEDSTOCK 2
4.35
17.7
1.23
FEEDSTOCK 2 - COMPOSITION
Dairy cow manure(TS)
MSW (TS)
Dairy cow manure (VS)
MSW (VS)
PERCENT
63%
37%
68%
32%
RATIO
1.67
2.10
FEEDSTOCK 2 - ORGANIC LOADING RATE
Ibs VS/ft3 day -1
g VS/1 day -1
0.11
1.72
FEEDSTOCK 2 - CHEMICAL ANALYSIS
% TS
% VS of TS
% VS of TSW
% VSn
TKN (mg. 1)
(NH4 + -NH3)-N
C.O.D. (mg/1)
INFLUENT
4.93
84.7
4.18
1100
251
54,000
EFFLUENT
3.94
81.9
3.23
24.6
1210
311
45,000
108
-------
DISCUSSION & CONCLUSIONS
A search of the recent literature on anaerobic digestion of farm wastes
found no previous work dealing specifically with the biodegradability of a
mixture containing dairy cow manure and the organic fraction of municipal
solid waste. However, there has been a significant amount of research on the
anaerobic digestion of dairy cow manure, which is summarized in Table A12.
Relative to most types of organic material, the anaerobic biodegrada-
bility of dairy cow manure is quite low, for an obvious reason: Readily
digestible material is broken down in the animal's rumen, and hence its
excreta contains only that fraction of the animal's feed intake which is
difficult to decompose, such as intricate poly-saccharide-lignin complexes.
In laboratory experiments, Morris, Jewell and Roehr (20) found the refractory
fraction (R) of dairy cow manure to be 0.575, i.e. the biodegradability of
dairy cow manure was 42.5% of the influent volatile concentration.
Likewise, about 30% to 60% of the organic portion of refuse is not
fermentable to methane and remains as a stable refractory residue (21).
Cellulose must be solubilized before it can be degraded. Because MSW is
not a good source of the enzymes which break down cellulose molecules, the
only point of attack for the anaerobic bacteria is at the partially dissolved
portion of the fibrils at the surface of the cellulose fibers (22).
Thus, the organic fraction of MSW and dairy cow manure appear to have
a similar level of anaerobic biodegradability. For this reason, the
addition of MSW to dairy cow manure should not substantially decrease the
efficiency of biogas production.
Table A13 provides a summary of biogas production from various com-
bined wastes. Of particular interest is the work done by Klien (23) with
chicken manure and components of municipal solid waste. Chicken manure
has a high concentration of nitrogen compared to other manures; its use
with highly carbonaceous refuse therefore improves the carbon/nitrogen
ratio of the feed much more so than other manures.
This study has demonstrated the feasibility of anaerobically digesting
a mixture of dairy cow manure and the organic fraction of municipal solid
waste, with a very small reduction in biogas production efficiency. Future
research must be addressed to the following areas:
1) What is the maximum of MSW which can be added to dairy cow
manure before biogas production is substantially reduced?
2) What is the affect of changes in the organic loading rate
and solids retention time on the anaerobic biodegradability
of a dairy cow manure and MSW mixture?
3) How does the presence of residual MSW in the digester
effluent affect the potential uses of this material as a
fertilizer or as a component in refeed?
109
-------
4) How much can pretreatment (by mechanical and/or biological
means) of the MSW improve the biodegradability of the
mixture as a whole?
5) What are the barriers to practical application of a waste
management/resource recovery system which combines animal
manures and MSW for a treatment in farm-scale anaerobic
digesters?
Table 12 Summary of Data on Anaerobic Digestion
Biosas Production
ft-Vlb VSA/day
(m3/kg VS4/day)
0.192 (0.012)
1.73 (0.108)
2.77 (0.172)
3.63 (0.230)
2.21 (0.137)
5.51 (0.344)
5-61 (0.350)
0.817 (0.051)
2.13 (0.133)
2.48 (0.155)
4.47 (0.279)
6.41 (0.400)
7.11 (0.444)
1.39 (0.087)
3.16 (0.197)
5.56 (0.347)
1.23*(0.077*)
2.49*(0.155*)
ft-Vlb VSD/day
(m3/k* VSP/day)
2.24 (0.140)
9.02 (0.563)
-
11.6 (0.721)
-
15.2 (0.950)
15.6 (0.972)
16.3 (1.02)
16.7 (1.04)
18.7 (1.17)
14.8 (0.927)
17.0 (1.06)
17.3 (1.03)
6.87 (0.429)
9.10 (0.568)
13.8 (0.859)
11.0 (0.69)
16.2 (1.01)
Loading Rate
Ib VS/ft3/day
(K VS/l/day)
1.74 (27.9)
0.874 (14.0)
1.09 (17.4)
0.436 (6.98)
0.544 (8.72)
0.218 (3.49)
0.145 (2.33)
0.231^(105^
0.461!'(2099)
0.692''(314(?)
0.236^(107^")
0.472!*(214@)
0.705!'(320
-------
Table 12 continued
Biogas Production
ft-Vlb VS./day
(m3/kg VSA/day)
1.67*(n.l04*)
2.33*(0.145*>
2.68*(0.167*)
2.97*(0.185*)
1.96*(0.122*)
2.65*(0.165*)
1.05*(0.065*)
1.36*(0.084*)
1.23*(0.076*)
1.57*(0.097*)
0.88 (0.055*)
0.79 (0.049*)
0.58 (0.036*)
1.05 (0.066*)
5.6 (0.349*)
4.82 (0.301)
~ • •
ft-Vlb VSD/day
(ra^/kg VSn/day)
16.1 (1.01)
14.3 (0.89)
6.39 (0.40)
6.15 (0.38)
4.11 (0.26)
4.97 (0.31)
4.80 (0.30)
5.08 (0.32)
6.33 (0.40)
8.88 (0.56)
1.94*(0.121*)
1.77*(0.110*)
1.53* (0.096*)
1.97*(0.123*)
12.4*(0.777*)
21.5 (1.34)
Loading Rate
Ib VS/ft3/day
(E VS/l/day)
0.202 (3.10)
0.215 (3.40)
0.155 (2.50)
0.176 (2.8)
0.185 (3.0)
0.216 (3.5)
0.120 (1.9)
0.12 (1.9)
0.24 (3.8)
0.24 (3.8)
0.10 (1.60*)
0.10 (1.60*)
0.18 (2.88*)
0.18 (2.88*)
0.50 (8.01*)
0.18 (2.88)
2CH-4
64
58
65
65
65
65
66
65
65
61
77
79
77
74
60*
55
Temp.
m.
23
35
35
35
35
35
32.5
32.5
32.5
43.6
35
35
35
35
35
35
SRTor
HRT
(days)
26.3
25.3
20
16.6
16.6
16.6
10
15
10
15
12
20
12
20
15
30
Scale
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
pilot
™«^"»"*ii»
Design
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
conven
I—"— W.M
Comment Ref No.
intermittent 15
mixing
15
natural 16
mixing
16
16
16
dairy bull 17
manure used
" " 17
17
17
18
18
18
18
high loading 19
rate
13
Notes on Table 12:
* Calculated from author's data
f Ib VS/reactor
G> gm VS/reactor
1. Scale was defined as one of the following categories:
a. laboratory
b. pilot
c. full-scale
2. Three types of design were considered:
a. Conventional (CONVEN), meaning vertically placed, completely
mixed, frequency of feeding often varies.
b. Batch, meaning the reactor is loaded once only.
c. Plug flow (PLUG), meaning horizontally placed.
Ill
-------
Table 13: BIOGAS PRODUCTION FROM COMBINED WASTES
Material
Chicken Manure
a
Chicken Manure
Paper Pulp
Chicken Manure
Newspaper
Q
Chick Manure
Grass Clippings
Steer Manure
Steer Manure
Grass Clippings
c
Dairy Cow Manure
Dairy Cow Manure
MSW
MSW d
MSW e
Sewage Sludge
MSW e
Sewage Sludge
Sewage Sludge
Proportion
100%
31%
69%
50%
50%
50%
50%
100%
50%
50%
100%
63%
37%
100%
90%
10%
75%
25%
100%
ft3 biogas/lb VSA
5.0
7.8
4.1
5.9
1.4
4.3
4.6
4.4
3.78
4.47
5.44
5.12
%CH, of bio gas
59.8
60.0
66.1
68.1
65.2
51.1
57.3
55.2
52.1
67.3
48.5
68
Notes:
a. Laboratory experiments by Klien, Ref. No, (23).
b. In New Alchemy Institute Newsletter 3, Ref. No. (25).
c. This study.
d. Laboratory experiments by Pfeffer and Khan, digestion at 60°C
with no pre-treatment, Ref. No. (24).
e. Laboratory experiments by Swartzbaugh, et al; average of
results, Ref. No. (22).
f. Reported by Meynell, Ref. No. (26).
112
-------
REFERENCES
1. Jewell, et al. 1976. Byconversion of Agricultural Hastes for Pollution
Control and Energy Conservation.Final Report, NSF-RANN & AERT
741222A01.NTIS # TID-27164.
2. Morris, G.R.; Jewell, W.J.; and Cosier, G.L. Alternative Animal Waste
Anaerobic Fermentation Designs and Their Costs. In: Energy,
Agriculture and Waste Management, ed. W.J. Jewell. Ann Arbor
Science, Ann Arbor, Michigan48106.
3. 1975 Wisconsin Assessor Farm Statistics. Wisconsin Statistical Reporting
Service, Publication #251-76, Madison, Wisconsin.
4. Fraser, M.D., 1977. The Economics of SNG Production By Anaerobic
Digestion of Specially Grown Plant Matter.Institute of Gas
Technology Symposium on Clean Fuels from Biomass, Orlando,Flordia,
January 24-28, 1977.
5. Fraser, M.D., et al. Design, Operation and Economics of the Energy
Plantation TMt Institute of Gas Technology Symposium on Clean
Fuels, Orlando, Flordia, January 27-30, 1976.
6. Inter Technology Corporation. Feasibility of Meeting the Energy Needs
of Army Bases With Self-Generated Fuels Derived from Solar Energy
Plantatio'nsT Report No. 260675, Defense Advanced Research Pro-
jects Agency, Contract No. DACA 23-74-0009.
7. Inter Technology Corporation. Solar SNG: The Estimated Availability
of Resources for Large-Scfale Production of SNG by Anaerobic
Digestion of Specially Grown Plant Matter.Report No. 011075,
American Gas Association, Project No. IU 114-1.
8. Kemp, C.C. and Szego, G.C. The Energy Plantation. 168th American
Chemical Society Meeting, Atlantic City, New Jersey, September
12, 1974.
9. Szego, G.C., et al. The Energy Plantation. Proceedings of the 7th
Intersociety Energy Conversion Engineering Conference, 1972.
10. Graham, R.W. 1975 (May). Fuels from Crops: Renewable and Clean.
Mechanical Engineering: 27-31.
113
-------
11. Greeley, R.S. and Spewak, P.C. Land & Fresh Water Energy Farming.
The MITRE Corporation, McLean, Virginia.
12. American Public Health Association, 1976. Standard Methods for the
Examination of Water and Wastewater, 14th edition. APHA,
Washington, D.C.
13. Freedman, D.L. 1978. Production of ^ethane Gas from Dairy Cow Manure:
Results of a Pilot Plant Study \ University of Wisconsin-Green
Bay, in preparation.
14. Klien, S.A. and Chan, D.B. 1969. Anaerobic Digestion. In:
Comprehensive Studies of Solid Waste Management, Second Annual
Report, ed. C.G. Golueke and P.Hr. McGaukey. SERL Report No.
69-1, U.C.-Berkeley. ,j
15. Hart, S.A. 1963. Digestion tests of livestock wastes. Journal
Water Pollution Control Federation 35: 748-757
16. Jeffrey, E.A.; Blackman, W.C. and Ricketts, R.L. 1964. Aerobic and
Anaerobic Digestion Characteristics of Livestock Waste,
University of Missouri Engineering Series Bulletin 57. "106 pp.
17. Gramms, L.C.; Polkowski, L.B. and Witzel, S.A. 1971. Anaerobic
digestion of farm ^imal wastes (dairy bull, swine, and poultry).
Transactions of the American Society of Agricultural Engineers 14:
7-11 and 13.
18. Dalrymple, W. and Proctor, D.E. 1967. Feasibility of dairy manure
stabilization by anaerobic digestion. Water Sewage Works: 361-364.
19. Halderson, J. 1972. Dynamic Response of an Anaerobic Digester with
Dairy Cow Manure Substrate"/ Ph.D. Dissertation,. Purdue University,
Lafayette, Indiana.
20. Morris, G.R.; Jewell, W,J,; and Loehr, R.C. 1977. Anaerobic Fermen-
tation of Animal Wastes: Design and Operational Criteria.
Unpublished paper, available from Dept. of Ag. Engineering,
Cornell University, Ithica, N.Y.
21. Gossett, J.M.; Healy, J.B.; Owen, W.F.; Stuckey, D.C.; and Young, L.Y.
1976. Heat Treatment of Refuse for Increasing Anaerobic Bio-
degradability. Final progress report, Stanford University, NSF-
RANN No. AER-74-17940-A01.
22. Swartzbaugh, J.T.; Miller, J.W.; and Wiles, C.C. Operating Experience
with Large Scale Digestion of Urban Refuse with Sewage Sludge.
Institute of Gas Technology Symposium on Clean Fuels, Orlando,
Flordia, Jan. 24-28, 1977.
23. Klein, S.A. 1972. Anaerobic digestion of solid wastes. Compost
Science 13(6).
114
-------
24. Pfeffer, J.T., Khan, K.A.; 1976, "Microbial production of methane from
municipal refuse", Biotechnoloqy and Bioenqineerinq XVIII: 1179-1191.
25. Merrill, R.; 1973, Methane Digesters for Fuel Gas and Fertilizer.
New Alchemy Institute Newsletter No. 3, Woods Hole, MA.
26. Meynell, Peter-John; 1976, Methane: Planning a Digester, Prism Press,
Dorchester, Dorset.
114
-------
-------
APPENDIX B
FEASIBILITY OF DEHYDRATING ANAEROBIC DIGESTER
EFFLUENT FOR REFEEDING
(Preliminary Study)
OBJECTIVES
Refeeding animal waste is economically attractive. The refeed value of
waste is higher than its soil amendment value; it does not change the quality
of the manure as fertilizer, and the ultimate disposal is via the soil anyway.
The low solids concentration in the effluent from anaerobic digesters
prohibits the immediate use of the sludge as a refeed or a feed-ingredient,
except as a possible source of food supplement in a watering system as prac-
ticed in swine operations. (1) Thus, an economical method of decreasing the
moisture content is of prime concern in refeeding anaerobically digested
sludge.
The preliminary study focused on finding a low-cost filter that could
be used on small and intermediate size farms in conjunction with anaerobic
digesters. The moisture content should be decreased to a level where the man-
ure can be used as a feed-ingredient or to a level where further dehydrating
with other methods, such as heat treatment, would become economically feasi-
ble. Figure Bl shows the cost of drying chicken manure at different moisture
contents down to 10 percent moisture. The graph indicates that removing
moisture by heat treatment is extremely expensive for mixtures of high
moisture content. The moisture content of anaerobically digested manure is
approximately 95 percent water. Zero-20 percent moisture removal is necessary
before refeeding.
PROCEDURE
Several companies were contacted for filters, but most of the products were
considered too costly or were not applicable to the desired process. (Table
Bl) Three products were selected for further testing. These were:
1) The trommel, a rotary vacuum drum filter produced by Technical
Fabricators, Incorporated, New Jersey.
115
-------
2) A series of different materials for vacuum filters,
produced by the Eimco Corporation from Palantine,
Illinois.
3) The Kason vibroscreen from Kason Corporation, Newark,
New Jersey.
Samples from the digester at Green Bay were sent to Technical Fabrica-
tors and Kason Corporation, while tests were run in Green Bay on the filters
from the Eimco Corporation.
The Eimco Corporation sent an "Eimco filter test leaf kit" on which our
samples were run. Two filters were feasible: Eimco filters #NY-415 and
POPR-898. The testing procedure consisted of a form time, which is the time
during which the filter is actually in the slurry, and a dry time, during
which drying air is pulled through the formed filter cake. The cycle time is
the form time plus the dry time. The form time was kept constant at one min-
ute, while the dry time was varied (1, 14, 29, and 59 minutes were used). A
dry time of more than twenty-nine minutes proved inefficient. The optimum dry
time lies between 14 and 29 minutes (Table B2). The samples had a solids
content or an increase of 6.6 percent. A further analysis was performed on
one sample cake and its filtrate (Table B3).
90 80 70 60 50 40 30 20 10
% Moisture Content
Figure B 1: Cost of drying waste at various moisture contents
(heating oil at $.50/gal)
116
-------
TABLE Bl. COMPANIES SELLING SOLIDS CONCENTRATION EQUIPMENT
Company
Type of Equipment
Approximate Cost
Hercules, Wilmington, OE
C. E. Bauer, Springfield.
Fait-Andritz, Austria
RE: Crane Corporation,
Appleton, VJI
Industrial Filter & Pump
Cicero, IL
Komline Sanderson,
Peopack, i!J
Pennwalt Corporation,
Sharpies Division
Oak Brook, IL
Rotex, Cincinatti, OH
Kason, Newark, NJ
Zimpro, Rothchild, WI
Parkson Corporation
Fort Lauderdale, FL
RE: Davco Incorporated,
St. Paul, MM
DeLaval Separator Company
Poughkeepsie, NY
Star Tank & Filter Company
Bronx, NY
Technical Fabricators,
Piscataway, NJ
Dart-Hoesch Filtration,
Paramus, NJ
Folcculant polymers
OH Hydrasive Screens
Pressure filters
Pressure filters
Rotary vacuum filter
Belt press filter
Decanter centrifuge
Vibrating screen
Vibrating screen
Cross-flow sieve
Filter press
Gravity thickener
Gravity thickener
Basket Centrifuges
Solid bowl centrifuges
Plate and Frame filter
presses
Rotary drum filter
(using newsprint)
Pressure Filter
700-12,000
75,§00-190,000
14,000-20,000
25,000
40,000
15,000-16,000
3,300-4,900
3,500
3,200
15,000
25,000
20,000
25,000
117
-------
ANIMAL WASTE
WATER
DIGESTER
EFFLUENT LIQUID MANURE
FILTER CAKE
FILTER(3)
HEAT DRYING
DEHYDRATED FEED
FILTER
J_
ENSILING
>. FILTRATE
FILTER(2)
PROTEIN'SUBSTRATE IRRIGATION
1. Extra cake that has to be disposed of can be mixed with the
filtrate and irrigated on the field.
2. This filter may be required if the filtrate proves to have
solids with a high nutritional value.
3. Further reduction of the -moisture content by means of a filter or press
would reduce the cost of heat drying considerably.
FIGURE 82: POSSIBILITIES FOR USING EFFLUENT AS REFEED
118
-------
TABLE 82: LEAR FILTER TEST DATA (EIMCO CORPORATION)
Cake
Total Formation Filtrate
Form Dry cycle (lbs/hr/ft2) Formation
#Eimco Thread time time time Vacuum Wet Dry rate »
Filter! count (min) (min) (min) (in hg) Solids Solids (lbs/hr/ft^)
1 NY-415
2 NY-415
3 POPR-
898
4 POPR-
898
5 NY-415
40x40
40x40
24x21
24x21
40x40
1
1
1
1
1
29
59
29
59
29
30
60
30
60
30
11
11
11
11
15
1.
0.
1.
0.
0.
1
55
3
62
95
0
0
0
0
0
.15
.075
.15
.085
.13
0
0
0
0
0
.19
.11
.20
.082
.21
Note 1: Doubling the total cycle time from 30 minutes to 60 decreases the
rate of formation by about half.
Note 2: Several other filters were used but these did not work, These were
-PO-808 with thread count, 40x23 -POPR-859 withhthread count, 68x30
-POPR-907 with thread count, 68x29 -PO-808Hf with thread count 48x30
TABLE B3: ANALYSIS OF SAMPLE CAKE AND FILTRATE
(EIMCO LEAF FILTER)
% Solids
% Volatile solids of total solids
% Volatile solids of total sample
% Fixed solids
% Total nitrogen
% Ammonia nitrogen
Cake
13.6%
80.0%
10.9%
2.72%
.36%
—
Filtrate
2.9*
64.3%
1.86%
1.04%
.17%
Note that the filtrate still contains 2.9% solids.
Samples were sent to the Kason Corporation in order to run tests with their
vibroscreen filter. The report is attached. The returned samples were
then analyzed in the Green Bay Labs. (See Table B4)
118
-------
-------
TABLE B4. ANALYSIS OF VIBROSCREENED FILTER SAMPLE
(KASON CORPORATION)
% Total Nitrogen
% Crude protein
Mg/1 Ammonia Nitrogen
% Solids
% Volatile Solids of
Unfiltered
Sample
.22%
1.37%
885.
6.66%
78.23%
Recovered
Solids
3.32%
2.08%
13.6%
84.1%
Filtrate
.20%
1.26%
920.
3.44%
65.4%
total solids
% Fixed Solids of 21.77% 15.9% 34.6%
total solids
% Volatile Solids of Total 5.21% 11.4 2.23%
Sample weight
% Fixed solids of Total 1.45% 2,16% 1.19%
Sample Wieght
A separate analysis of a sample showed a solids concentration of 16.4%.
119
-------
CONCLUSIONS
The results indicate that the use of filters could reduce the moisture
content from 95 percent to 85 percent. This is equivalent to the moisture
content of raw manure. Reviewing the literature on use of manure as a refeed
indicated that the nutritional value of the effluent is lower than the value
for whole manure. (3) /\ more concentrated effluent (less moisture) may be
desirable in order to obtain similar feeding value in the same volume. Figure
B2 shows various possibilities of using the effluent as a refeed.
The effluent from an anaerobic digester contains about 95 percent moisture.
The effluent can be used directly in the aqueous phase of a liquid manure
feeding system, as was done with swine manure from an oxidation ditch at the
University of Illinois by Harmon.(4) He noted that both gain and efficiency
values were significantly greater for pigs receiving the manure, but he also
stated that nutrient intake cannot be greatly reduced because the liquid
manure is 95 percent water. (Aeration probably would be required before re-
feeding anaerobic digester effluent.)
Running the effluent through a filter would give two products: a filter
cake, containing about 85 percent moisture, and a filtrate containing about
97 percent moisture. It may be important to investigate the refeeding
value of the 3 percent solids that are still in the filtrate. Harmon noted in
his report a linear increase in amino acid concentration as particle size
decreases. This is the reason a second filter was suggested before irriga-
tion in order to filter out the smaller size particles.
The filter cake is the basic product for refeeding. Filter cake that
cannot be recycled can be mixed with the filtrate and disposed of through
irrigation. The filter cake has approximately the same moisture content as
manure, depending on the efficiency of the filtering process. Therefore,
literature on recycling straight manure can be used to find an optimum sys-
tem.
On the other hand, previous studies indicated a lower nutritional value for
effluent compared to the influent (straight manure plus water) so that some
modifications will be necessary. The filter cake can be used in various ways
but three possibilities were selected here.
One possibility is to further dehydrate the filter cake in order to obtain
a dehydrated feed. Heat drying is an expensive method, unless a readily
available and cheap source of energy such as waste heat from the digester or
solar heat can be applied. Filtering or pressing before heat drying may also
prove to be economically sound. The cost of dehydration between 85 percent
and 70 percent is still very high. Heat treatment, however, has some major
advantages. The process assures the removal of pathogens and the product
has a higher chance of being accepted by the FDA as a feed ingredient. The
120
-------
dehydrated feed could be packaged for sale on the market.
Ensiling of whole manure has been extensively studied by Dr. W. B.
Anthony at the Auburn University in Alabama. According to Anthony, ensiling
raw manure is the most promising way of recycling. It may also prove to be
the most effective way of refeeding digester effluent. Anthony see six major
advantages in ensiling:(5)
1. All the nutritive value in the waste is retained for
feeding.
2. Ensiling is an effective process for eliminating pathogenic
organisms such as Salmonella and parasitic nematodes. There
is also a reduction in the number of Coliform bacteria.
3. It materially improves the esthetic aspect of using waste
as a livestock refeed.
4. Ensiling offers an important aid in pollution abatement.
5. It requires no major investment in facilities. Ensiling
is the lowest cost treatment process currently available.
6. Ensiling proves to be applicable to small and intermediate
size operations.
The filter cake could also be used as a substrate for protein produc-
tion by both unicellular organisms and invertebrates (algae, yeasts, fungi,
mixed cultures of microorganisms, housefly larvae and earthworms.)
The algae are the most efficient in converting the manure into usable
feed-grade protein. Drawbacks in algal recovery systems include: The amount
of space required to process large amounts of manure, the apparent high
capital outlay to establish the systems and the climatic and topographic
limitations on pond function and location. The technology is feasible however,
and if adequate markets for algal orotein could be established, this approach
may ultimately be practical in geographic locations that allow year-round
operation of ponds.(6)
The immediate future of anaerobic digesters is largely dependent on
benefits other than the methane gas. Our preliminary study indicates that the
effluent from anaerobic digesters can be effectively dehydrated with filters
to a level where its moisture content is about equal to the moisture content
of raw manure. This opens up a multitude of possibilities for the use of
the filter cake as a refeed or feed-ingredient.
Further research is needed in order to determine the effect of the
digestion process on the effluent, and to determine the nutritional value in
the filter cake as well as in the filtrate. The possibility that a signi-
ficant amount of nutrient is lost in the filtrate needs to be explored and
related to the cost effectiveness of centrifugation or other recovery options.
121
-------
REFERENCES
1. D. L. Day and B. G. Harmon. A recycled feed source from anerobically
processed swine wastes. Transactions of the American Society of
Agricultural Engineers, 17:82-84, 87, 1974.
2. R. Blair. Utilizing Wastes in Animal Feeds - A European Overview.
Feedstuffs, June 30, 1975. 33 pp.
3. Previous study under this grant on amino acid availability in the
effluent compared to the availability in the influent.
4. See 1
5. W. B. Anthony. Recycle manure to improve feed efficiency pf cattje/
(Paper read at Moscow, USSR, and Verona, Italy, 1975.)
6. C. C. Calvert. Animal Wastes as substrate for protein production.
Federation of American Society for Experimental Biology, Symposium at
the 57th annual meeting. April 18, 1973, New Jersey.
Federation proceedings, Vol. 33 n. 8, August 1974, pp. 1938.
122
-------
APPEMDIX C
COMPARISON BETWEEN INFLUENT AND EFFLUENT OF
ANAEROBIC DIGESTERS AS REFEED
Bacterial growth and methane formation alter the composition of manure
during anaerobic digestion and have an effect on the refeeding value of the
manure.
A preliminary study was conducted at UW-Green Bay in order to determine
the extent of this change in nutritive value. The study consists of two parts.
The first part is a rough determination of the composition and the second is
a more in-depth study on the amino acid content.
PROCEDURE
Four samples were collected from the UWGB digester that was running on
dairy cow manure. These four samples were used for every analysis described
in this study. Samples A and B are the influent and effluent collected on
July 11, 1977. Samples C and D are the influent and effluent collected on
July 18, 1977.
The influents A and C are compared to the effluents B and D.
Nutritive Composition—
A rough nutritive analysis consisted of the determination of the ash, crude
protein, fat and carbohydrate contents on a dry weight basis. Ash or mineral
content was determined by igniting the dried samples for two hours at 600 C.
Crude protein was obtained from the Kjeldahl Nitrogen method. Fat content
was determined through an ether extraction in an Soxhlet apparatus for 20
hours. The carbohydrate content was calculated by adding up the ash, crude pro-
tein, and fat content and subtracting this total from 100 percent.
123
-------
TABLE Cl. NUTRITIVE ANALYSIS OF SAMPLES
Ash
Crude Protein
Fat
Carbohydrate
INFLUENT
A
25.0
10.3
2.2
62.5
EFFLUENT
3
21.8
11.4
1.5
65.3
INFLUENT
C
22.0
11.2
2.5
64.3
EFFLUENT
D
23.1
12.0
1.1
63.8
Table Cl suggests that a decrease in fat content and an increase in crude pro-
tein may be expected. The impact that this may have on the refeeding value
should be explored further. The protein percentage seems quite low. Opti-
mum percent of ration of protein in feed for growing chickens and turkeys is
16 and 20 percent, and even higher for starting poultry. (1)
Amino Acids—
Proteins are polymerized units of amino acids which vary in relative amounts
and sometimes in kind from protein to protein. These amino acids are obtained
as the hydrolytic end products when proteins are boiled for many hours with
strong acids, or, when they are acted upon by certain enzymes. They are also
the end products of protein upon digestion, and the building stones from
which body protein is made, as well as intermediary products in protein cata-
bolism. (2)
A total amino acid analysis and two enzyme analyses were run on our samples.
(3) The total amino acid analysis (T.A.A.) indicate the availability of the
specific amino acids. The latter is thus, more important in determining the
value as a feed.
Total Amino Acid—One T.A.A. analysis was run on each sample. The samples
were nitrogen-purged before analysis. The results are shown,in Table C2.
The results do not indicate a difinite trend. Columns 5 and 6 are the aver-
ages of the influent and effluent samples respectively. These results can
be compared with data for either feeds. In our study, we included the ana-
lysis for essential amino acids from alfalfa and cornmeal. (4) The results
of the manure samples, both for the influent and the effluent, are in a
similar range.
124
-------
Enzyme Ami no Acid — The chemical changes within the digester may cause the
amino acids to be less available in intestinal digestion. A pepsin pancrea-
tin digest index of protein quality evaluation (5) was used to determine the
extent of the change in availability.
The Table C3 contains the results for these analyses.
twice (A]_, A2).
Each sample was analyzed
Run I may contain an inaccuracy of about 10 percent because the sample was not
brought up to volume before final analysis.
Sample 82 had been fat extracted and this may have caused the amino acids to
be available than the nontreated sample
TABLE C2. TOTAL AMINO ACIDS IN INFLUENT/EFFLUENT SAMPLE
Aspartic Acid (ASP)
Threonine (THR)
Serine (SER)
Glutamic Acid (GLU)
Proline (PRO)
Glycine (GLY)
Alanine (ALA)
Valine (VAL)
Methionine (MET)
Isoleucine (ILE)
Leucine (LEU)
Tyros ine (TYR)
Phenylaline (PHE)
Lysine (LYS)
Histidine (HIS)
Arginine (ARC)
Total
NH3
+NH3
A
.93
.46
.40
1.16
.57
.55
.66
.53
.24
.46
.71
.35
.51
.57
.17
.31
8.58
.31
8.89
~
.96
.48
.42
1.15
.46
.59
.67
.46
.24
.48
.79
.23
.46
.55
.19
.35
8.48
.32
8.80
C
.86
.48
.43
1.16
.42
.53
.68
.45
.20
.47
.77
.17
.40
.59
.18
.38
8.17
.30
8.47
~D
1.08
.56
.48
1.31
.53
.65
.78
.60
.27
.55
.85
.43
.77
.60
.18
.36
10.0
.36
10.36
Influent
A+C/2
.90
.56
.42
1.16
.50
.54
.67
.49
.22
.47
.74
.26
.46
.58
.18
.35
8.41
Effluent
B+D/2
1.02
.52
.45
1.23
.50
.62
.73
.53
.26
.52
.82
.33
.62
.58
.19
.36
9.28
Alfalfa
.62
.72
.13
.96
1.10
.65
1.08
.25
.76
Cornmeal
.34
.41
.12
.41
1.03
.38
.32
.23
.40
Nitrogen
1.72
1.87
2.49
Data are on a weight over weight basis and expressed as %. (w/w x 100)
Tryptophan was not analyzed for.
TABLE C3: CONCENTRATION IN (W/W x 100) OF 8 runs.
1.53
Asp
Thr
Ser
Glu
Pro
Gly
Ala
Val
Met
He
Lev
Tyr
Phe
Lys
His
Arg
iota:
.56
.23
.26
.74
.05
.16
.53
.51
.20
.41
1.07
.47
.49
.65
.05
.43
6.81
.35
.11
.13
.46
.06
.09
.33
.26
.13
.22
.59
.43
.37
.22
/
.36
4.11
.28
.09
.18
.21
.17
.21
.41
.19
.06
.12
.40
.29
.27
.27
1
.14
3.29
.11
.09
.08
.19
/
.11
.26
.18
.21
.18
.48
.50
.39
.36
1
.29
3.42
.39
.13
.15
.43
.05
.11
.29
.24
.11
.17
.63
.25
.34
.35
1
.25
3.89
125
.34
.23
.27
.84
.44
.26
.74
.55
.42
.54
1.22
.88
.77
1.12
1
.36
8.98
.23
.04
.07
.30
.09
.08
.19
.16
.04
.09
.33
.13
.21
.19
1
.05
2.2
.10
.02
.11
.47
.63
.96
.46
.18
.25
.48
.63
.54
.05
/
4.88
-------
The standard for run 2 gave erroneous results that had to be discarded
although the samples gave acceptable results.
Table C4 takes an overview of the results. Column 1 and 2 contain the
sum of all the influent and effluent samples. Columns 3 and 4 divide this total
by four in order to obtain an average concentration for the influent and
effluent.
The total of these columns indicate that the amino acids are less
available in the effluent than in the influent. (5.97 percent vs. 3.46 percent)
Column 7 compares the availability of the effluent and influent for each amino
acid. Only two amino acids, proline and glycine seem to be more available in
the effluent. Overall, the amino acids in the effluent have an availability
that is about 65 percent the availability in the influent.
Columns 5 and 6 compare E.A.A. vs. T.A.A. (the averages). The enzyme
analysis for the influent recovers about 80 percent of the total amino acids
present, while there is about a 42 percent recovery for the effluent. Note
that the enzyme analysis releases about twice as much tyrosine as the TAA.
This seems incorrect at first, but it is possible that tyrosine is formed
during the enzyme analyses.
126
-------
TABLE C4: OVERVIEW OF RESULTS FROM AMINO ACID STUDY
A
ASP
THR
SER
GLU
PRO
GLY
ALA
VAL
MET
ILE
LEV
TYR
PHE
LYS
HIS
ARG
INFL
1.64
.7
.81
2.47
.60
.62
1.89
1.56
.86
1.34
3.51
2.03
1.97
2.34
.05
1.40
EFFL
Bi n j.n _L
* i D « * U •» *
.72
.22
.35
.81
.73
1.03
1.82
.99
.49
.64
1.69
1.55
1.41
.87
--
.68
TOTAL
Infl
D-, Avq
.41
.18
.20
.62
.15
.16
.47
.39
.22
.34
.88
.51
.49
.59
.01
.35
5.97
Effl
Avq
.18
.06
.09
.20
.18
.26
.46
.25
.12
.11
.42
.39
.35
.22
--
.12
3.46
INFL
(Enz x 100
Total
45.6
38.3
47.6
53.4
30.0
29.6
70.1
79.6
100.0
72.3
119.0
196.1
106.5
101.7
100.0
E=1189.8
EFFL
Enz x 100 Effl
Total
17.6
11.5
20.0
16.3
36.0
41.9
63.0
47.2
46.2
30.8
51.2
118.2
56.5
37.9
33.3
E = 627.6
x 100 = Infl)
43.9
33.3
45.0
32.3
120.0
162.5
97.9
64.1
54.5
47.1
47.7
76.5
71.4
37.3
34.3
x = 79.3 x = 41.8 x = 64.5%
127
-------
An observation from Table C3 is that there are a few amino acids that are
higher in the influent for all cases. Comparing A, with EL, A with EL, C with
D, and C~ with D2, we find that the following amino acids are always higher in the
"2
influent!
Aspartic acid and Glutamic acid;
Threonine, valine, isoleucine and leucine.
It is interesting to see that both acids are present and that all the other
amino acids belong to the monoaminomonocarboxylic acids and that they are more
complex compounds of this group.
Iscoeucine, leucine, threonine and valine are considered essential amino
acids in animal feeds. This may be important in evaluating the difference in
availability between influent and effluent.
No amino acids are lower in the influent for all cases.
DISCUSSION
Our pilot study indicates that the amino acids may be less available
in the effluent and therefore the effluent may be considered a lower quality
refeed. It may suggest that the amino acids in the effluent may be part of the
bacterial cell more than in the influent, and that the cell wall resists the
enzymatic hydrolysis.
It would be interesting to treat the effluent in order to recapture the cell
walls and investigate whether the amino acids are more available after the
treatment.
In vitro analyses indicate a lower quality refeed but actual refeeding will
reveal more about the quality of the effluent and the influent.
BIBLIOGRAPHY
1. Duckworth, John, "Rearing and Table Birds: Scientific Principles of
Feeding Farm Livestock."
2. Maynard, L. A., Loosle, J. K., Animal Nutrition, McGraw-Hill Book
Company, 1969, P. 119
3. Laboratory analyses were performed on a Beckman Amino Acid analyzer at
Dr. Stahmann's Lab at UW-Madison, Wisconsin.
4. H. H. Williams, "The Essential Amino Acid Composition of Animal Feeds".
5. Akeson, W. R., Stahmann, M. A., "A Pepsin Pancreation Digest Index of Protein
Quality Evaluation", J. Nutrition. 83 (64): 257.
128
-------
Appendix D
THE FEASIBILITY OF INCREASING METHANE GAS
PRODUCTION FROM ANAEROBIC DIGESTERS VIA
HYDROGEN PRODUCING ALGAE
INTRODUCTION
The intent of this study is to investigate the potential for increasing
methane production from an anaerobic digester algal system. Our purpose
here is to increase the volume of methane gas produced for a given volume
of manure introduced into the digester. This could considerably increase
the economic feasibility of anaerobic digesters for small-midsize dairy
farms.
At the present time we can suggest investigations into two alternatives
which couple the anaerobic and hydrogen producing algal systems symbio-
tically. With one alternative hydrogen producing algae would be introduced
directly into the anaerobic digester, where any hydrogen evolved would be
used by methanogenic bacteria in the formation of methane. This relationship
has been observed by J.W. Czerkawski et. al. (1971). He found that intro-
duction of hydrogen into the gas phase of an artificial rumen, essentially
an anaerobic digester, increased methane production directly in proportion
to the amount of hydrogen added. This relationship is graphically displayed
in Figure Dl.
However, whether an anaerobic digester provides a suitable environment
for hydrogen production from algae is yet to be determined and will com-
prise the main focus of this study.
Another alternative system would consist of introducing hydrogen pro-
ducing algae into an anaerobic container separate from the digester, which
provides an optimum environment for hydrogen production. The hydrogen pro-
duced within this container could then be pumped into the anaerobic digester
to be used by the methanogenic bacteria, or simply mixed with the methane gas
to be burned directly.
HYDROGEN PRODUCTION EXPERIMENTS
Choosing the Algae
Choosing a specific type of algae for purposes of experimentation was
essentially a choice between greens and blue-greens, since all research we
129
-------
50r-
o
e
30
20
o
o
I
10
to)
O 10 2O
Conctntrotion of hydrogen in gas phasi
CH by volume)
FIG. 1. Relationships between the concentration of H in the gas phase and (a) the rato
of uptake of H and ('.») the rate of production of methane by strained rtsmen contents.
Kcsults were obtained from the experiment surnmnrized in Table 1. Initial volumes
of H (ml/vessel) were: (O). 35; (OK 86; (A) U>7; (A ). 256.
FIGURE Dl: HYDROGEN AND RUMEN METHANE PRODUCTION
could find was directed to one or the other. In the beginning of the study
we decided to work with both greens and blue-greens. However, due to time
constraints, we were forced to specialize. We chose green algae for a number
of reasons: (1) we were interested in adding algae directly into the anae-
robic digester for two reasons; a) if it worked, this seemed like the most
efficient arrangement; b) past studies have shown hydrogenase-containing
algae to produce H2 in nitrogen, argon, and helium atmospheres. We wanted
to see what would happen in a methane atmosphere. Blue-greens are known to
produce \\2 and 02 simultaneously, however, methanogenic bacteria within the
digester are inhibited by the slightest amount of 62. Green algae produce
only hydrogen and some C02. (2) blue-green algae produce the greatest
amount of hydrogen at high light intensities, therefore, they would be sev-
erely limited within a digester due to the opaqueness of the digesting me-
dium. Green algae are known to produce hydrogen in the dark. (3) when con-
sidering the production of hydrogen in a separate container, again we saw a
problem with blue-greens. Because blue-greens produce 02 and H2 simultane-
ously, separation of the 02 from the gas would be necessary before piping
130
-------
into an anaerobic digester. Gas produced from green algae could be piped
directly into the digester.
These reasons for choosing green algae in no way preclude future studies
with blue-greens. Blue-greens have been shown to produce hydrogen in quan-
tities eight times greater than that of green algae. We were not aware of
this information at the time of our choosing.
The green algal species used in the following experiments is Scenedesmus
obliquus. We chose this species because it was available and also the most
widely studied in the literature.
Hydrogen Experiments
A number of experiments have been designed to investigate the potential
of the digester-algosystems mentioned briefly in the introduction, and more
specifically in Section V. These experiments will take place during the
course of the summer.
Experiment 1: materials and methods
Scenedesmus obliquus algae will be introduced directly into lab anae-
robic digesters to investigate the change they might cause on methane pro-
duction. The lab digesters will be described in detail at the end of the
hydrogen production section. The procedure and conditions for the experi-
ment are listed.
1) There will be eight lab anaerobic digesters all operating at 35 C
and using equal volumes of cow manure as a substrate. By treating each
digester the same we hope to attain a low variance in gas production rates
between all the digesters. We will attempt stabilization of the digesters
at a 20-day retention time. The mixing rate is yet to be determined.
2) A mixture of cow manure and algae will be added to four of these
digesters, at a rate to be determined from experiment (3). The other four
digesters will act as a control group, and each digester will have a history
for use as a comparison. The algae to be used in this experiment will be
cultured in manure substrate medium.
3) The quantity and composition of gas will be measured regularly,
using gas chromatography techniques. Primary emphasis will be directed to
measuring methane increases, since the hydrogen is theoretically converted
to methane via anaerobic microbes. However, we do have hydrogen analyzing
equipment, and this will be used to measure hydrogen in the gas, first by
spot checks, and then, if necessary, regular measurement will ensue.
Discussion of Experiment 1
It is important to discuss how we hope to determine any change in gas
production with statistical significance, and what this would mean if a
change did occur. For purposes of statistical significance we hope to
stabilize the digesters to a point where we can predict, within a few mis,
the quantity and composition of gas which each digester will produce with
131
-------
respect to itself over time. The digesters will then be compared to each
other for gas composition and quantity. However, if we do find changes in
quantity and composition of the gas, this does not with any certainty, indi-
cate that hydrogen from the algae has caused this change. For example, it
may simply be the result of gas production from the decomposition of the
algae within the digester. Studies by Golueke C.G. et al (1957) found that
anaerobic digestion of a mixed culture of Scenedesmus and Chlorella algal
species produced two to four cubic feet per pound of volatile matter less
than anaerobically digested raw sludge. The percentage of methane was, how-
ever, similar. Therefore, with this information, it appears that a gas
production increase would most likely be a result of hydrogen conversion to
methane. However, more information would be necessary before any tight con-
clusions are attempted.
We could also look at the composition of the gas production. If the
relative concentration of methane increases, this might be due to increased
hydrogen in the gas phase. Smith P.N., et al (1976), found that hydrogen
and carbon dioxide were very rapidly converted to methane during anaerobic
digestion of organic molecules. This could effectively decrease the relative
composition of carbon dioxide in the gas phase. This would show a relative
methane increase. However, Smith also found that methane formation from
other substrates was decreased during the formation of methane via hydrogen.
This might decrease the relative methane concentration, Czerkawski I.W.
(1972), found that increased hydrogen in the gas phase of an anaerobic
digester directly increased methane production. This might suggest that,
even though methane formation from other substances is decreased, the amount
decreased is small relative to the amount produced from the hydrogen added
(see figure 1). Clearly, there is considerable difficulty in the analysis
of where the gas production changes are coming from. Experiments 3 and 4
will be designed to deal with this puzzle.
Experiment 2
This experiment will be set up in essentially the same manner as
experiment 1. The only difference will be the introduction of hydrogen
gas into the digester instead of algae. This experiment could give us some
indication of what happens with the direct introduction of hydrogen in the
digester. The technique for introduction, or the amount of hydrogen to be
introduced, have not yet been determined.
Experiment 3
This experiment will also be set up essentially the same as experiment 1.
The difference will be the control group. Known non-hydrogen producing algae
will be added to four digesters for use as a control group. Hydrogen pro-
ducing algae will be added to the other four digesters. Equal amounts of
algal biomass will be added to each digester. The main difficulty we per-
ceive with this experiment is getting a non-hydrogen producing algal species
similar enough in chemical makeup to the hydrogen producing algae, so both
species have similar decomposition characteristics.
From this experiment we hope to gain insight as to whether any changes
132
-------
in methane production are a result of hydrogen production, or just fermen-
tation. This experiment will probably not occur this summer simply because
experiments 1, 2, and 4, are of higher priority, and will most likely take
up all available time.
Experiment 4
This experiment will investigate the hydrogen producing potential of
Scenedesmus obliquus in varied conditions of lighting, temperature, and in
different mediums. The purpose for this experiment will be to (1) show us
whether the algal species we have actually produce hydrogen, (2) what con-
centrations of algae we should add to the digesters in experiment 1 to obtain
results, and (3) how these algae produce hydrogen in varied conditions of
light and temperature, and culture medium.
In this experiment, various concentrations of algae will be introduced
into containers with various mediums. These mediums will be described later.
The atmosphere within the container will then be filled with nitrogen gas or
argon to provide an anaerobic environment. Various light intensities and
mixing rates will be applied.
As discussed in section 2, light is an important variable in determining
hydrogen production rates. Mixing will be used with light experiments to
maximize the exposure of algae to light in the dark manure based medium.
Temperature is also an important variable in hydrogen production from algae.
A temperature of 35°C will be used as a basis for comparison to an anaerobic
digester system which is also at 35°C. Experiments will also be run at
25°C for comparison. Temperature control will be accomplished with a water
bath.
Pressure change will be measured with a manometer. Gas composition will
be measured with gas chromatography. The quantity of gas produced will be
determined by relating the manometer pressure to the volume of air space
within the container. Figure D2 is a sketch of this system.
Two different test mediums will be used: (1) Cow manure taken from an
active anaerobic digester; this will give us some indication of what will
happen when adding hydrogen producing algae to the digester, and (2) buffered
water, since this is the most widely used test medium in the literature.
From this we can compare the gas production we get to that of experiments
already undertaken with similar conditions. In this way we can test our
technique. We can then compare results from this test medium to those of
the digester effluent medium. Note: a blank will be set up every time a
test medium is used.
Description of Systems Operation
The digesters are of the continuous feed type. Digested manure is first
removed from the feed tube with a largemouth 10ml pipette. An equal amount
of raw manure is introduced. Presently, the digesters are fed 5ml per day.
We will increase feeding rate gradually until we reach a 20-day retention
time. The digesters are presently mixed for 12 minutes every four hours,
133
-------
with magnetic stirrers operated by an electric timing device.
As gas is produced, it flows in the direction of the orange lines as
shown, the pressure within the digester is checked by a manometer. Once the
gas flows through the test tube, it is cut off from the digester atmosphere.
The purpose of this is to allow measurement of the composition changes within
the digester. For example, if the atmosphere in the gallon bottle was not
cut off from the atmosphere of the digester, all the gas would mix together
and we would have no indication of the composition change. As pressure in
the gallon bottle builds up, water is forced up through the water outlet
tube. The pressure adjustment tube is moved up or down in the holder to
increase or decrease pressure within the system. The water is adjusted
daily to atmospheric pressure. The water then falls into the glass jar at
atmospheric pressure and is measured daily.
These lab anaerobic digesters have been operative for approximately
two months to date. During this time period we have manipulated feeding
rates and mixing rates in an attempt to stabilize the system. Ideally, for
purposes of experimental control, all eight digesters should be operating
at the same gas production rates. The system presently shows signs of
stabilizing, and we anticipate approximately one month before we can begin
experiments, if this trend continues.
rubber
stopper
gas sampling
port
pipette
manometer
erlenmeyer
flask
N2,CH4, argon,
or helium
rubber stopper
with holes
distilled water
with KOH
Figure D2 :
Schematic of laboratory digester for combined
algae/manure cultures.
134
-------
Lab Digesters
A series of eight anaerobic digesters were set up in a water bath
and maintained at 35°C. See sketch below.
sampling por
feed tub*
gas tub*
timer
, I -^ manom«t*r
,— with w»t*f
.water outlet
tub*
water lev*! \vr
t««t tub*
,op*n to atmoaphere
—9f»»»ur« adjustment
tube
holder
.glass j«r
//// water
Figure D?: Schematic of gas measuring system.
Algae Culturing Experiments
The primary focus of the algal culturing experiments was to provide cow
manure grown algae for the hydrogen production experiments. Having cow
manure grown algae was of high priority for purposes of adapting algae to
the cow manure prior to introduction into the lab anaerobic digesters.
Therefore, it was necessary to design experiments to examine the algae
producing potential of the manure. These experiments were accomplished
under controlled laboratory conditions and therefore are not reflective of
algae grown in the natural habitat. These experiments were designed to
examine the relative growth of Scenedesmus obliquus in various concentrations
of manure substrates. Scenedesmus obliquus grown on a known medium was
used as a control.
135
-------
The following experiments are to be considered preliminary to future
research in this area. In most cases these experiments are only capable
of predicting a relatively subjective indication as to the nature of
certain specified growth parameters for algae growth on manure substrates.
These experiments were generally not repetitive enough for conclusive
results.
Experiment 1 - Scenedesmus obliquus grown on TAP nutrient medium.
The average cell population/ml/day was recorded and graphed,
5.5 x
107
5io7
(/i
O)
ra
en
10'
T Range in
JL measurements
3
DAY
Experiment 2a - Scenedesmus obliquus growth on various concentrations of
lagoon supernatant
Bottle 1: 90 mis of distilled water and 10 mis of lagoon supernatant
Bottle 2: 80 mis of distilled water and 20 mis of lagoon supernatant
Bottle 3: 100 mis of undiluted lagoon supernatant
(see graph following page)
FIGURE D4: AVERAGE CELL POPULATION IN EXPERIMENT 1
136
-------
C3
-Starting Population
Final Population
\
#1
#2
DILUTION BOTTLE NUMBER
Figure 05. Results of Experiment 2a.
#3
Experiment 2b - Scenedesmus obi. growth on undiluted lagoon supernatant
with NH4C1 added.
Bottle 1:
Bottle 2:
50 mis of undiluted lagoon supernatant + .4 gms sodium
bicarbonate + .01 gms of NH4C1.
50 mis of undiluted lagoon supernatant + .4 gms sodium
bicarbonate.
10'
10'
<
=*t=
^-Starting Population
Final Population
Experiment 3a
Bottle 1:
Bottle 2:
Bottle 3:
Bottle 4:
+NH| No NHt
Figure D6. Results of Experiment 2b.
Scenedesmus obi. growth on low dilutions of strained
anaerobic digester effluent.
25 mis of a mixture of distilled water and strained anaerobic
digester effluent in proportions of 10/1, total volume of
solution to total volume of effluent.
25 mis of the same mixture in proportions of 20/1.
25 mis of the same at 40/1
25 mis of the same at 80/1
137
-------
1 107
io
•Starting Population
>Final Population
#1
#2 #3
FLASK NUMBER
Figure D7. Results of Experiment 3a.
Experiment 3b •
Bottle 1:
Bottle 2:
Bottle 3:
Bottle 4:
dio7
c_>
<
!,„«
Scenedesmus oJ3l. growth on high dilutions of strained
anaerobic digester effluent.
50 mis of a mixture of distilled water and strained anaerobic
digester effluent in proportions of 50/1, total solution to
total volume of effluent.
50 mis of the same mixture at 100/1
50 mis of the same at 200/1
50 mis of the same at 400/1
Starting Population
/•Final Population
#1
#2 #3
DILUTION BOTTLE NUMBER
Figure 08. Results of Experiment 3b.
#4
138
-------
Experiment 3c •
Bottle 1:
Bottle 2:
Scenedesmus obi, growth on strained digester effluent with
NH4C1 added.
50 mis of distilled water and anaerobic digester effluent
mixture, 200/1 total solution/effluent, .4 gms of sodium
bicarbonate, .01 gms of
the same without the
LU
10'
=*=
Starting Population
Final Population
No NH4+
Figure D9. Results of Experiment 3c
Discussion of Results
These preliminary results suggest that Scenedesmus obliquus algae grown
on manure substrates can attain population densities on the same order of
magnitude as when grown on TAP nutrient solution. For example, maximum
algae cells/ml of solution counted for TAP grown algae was 5.5 x 10' in
experiment 1. In experiment 3a algae grown on an 80/1 dilution of strained
digester effluent, and in experiment 2a algae grown on undiluted lagoon
supernatant attained populations of 1 x 10'.
When NH4 was added to the strained digester effluent in experiment 3c,
no significant population increases were recorded when compared to the same
medium without the NH4 added. However, in experiment 2b, with undiluted
lagoon supernatant, the greatest algal population density was attained when
NH4 was added. This might indicate that NH4 concentrations in the lagoon
supernatant may be limiting to algal growth. Chemical analysis of the
lagoon supernatant have found Nlfy concentrations to vary between 15 and 46
139
-------
ppm. The concentration of Ntfy in TAP nutrient medium is 200 ppm. Krauss
and Thomas (1954) used 139 ppm NFfy concentrations in the mass culture of
Scenedesmus obliquus. This information clearly adds support to the obser-
vation that NH4 added to lagoon supernatant in experiment 2b increased algae
population density over supernatant with no NH4 added. However, it is
interesting to note that the maximum population density in experiment 2a
with undiluted lagoon supernatant, no NH4 added, was approximately equal to
the population densities achieved in experiment 2b, with undiluted super-
natant and NH4 added. The reasons for this inconsistancy remains a puzzle.
However, this fact questions the validity of the two experiments.
Extrapolations
An attempt will be made in this section to give the reader some idea
of the volume of hydrogen to be expected from a one acre algae production
unit. It must be noted here that all calculations will be based on optimum
conditions for hydrogen production, algal culturing, and conversion
efficiency.
The maximum rate of hydrogen production from green algae recorded to
date has been shown to be approximately 4/d H2/mg of dry weight/hour, or
.064 cubic feet H2/lb dry weight per hour, Peak production lasts only for
a few hours, however, for the sake of simplicity we will assume algae always
operate at peak production. This would require introduction of new algae
into the hydrogen producing unit every few hours. For optimum conditions
we will assume system 11 of the following section is used.
The amount of algae that can be produced on one acre of land using
anaerobic digester effluent as a growth medium was predicted by Rowe (1976).
He predicted that 28,704 pounds of algae, dry weight, could be produced
during the summer season (92 days) on one acre of land. These predictions
were based on algae production rates on swine manure by Boersma et al.(1975).
If we assume algae are producing hydrogen for 12 hours/day in the summer,
then we might expect:
.064 ft3 H2/lb-hr x 28,704 Ibs/acre x 12 hrs/day = 22,040 ft3 H2/day.
or, 2.028 x 106 ft3 of H2 for the 92 summer days.
If we introduced this gas into a digester with 100% conversion effi-
ciency to methane (two H2 molecules to every one C02 molecule will form CH4)
then one could expect one half of 22,040 ft-* of H2, or 11,020 ft3 of CH4/day.
Depending on the size of the digester this could add a significant volume
of gas. However, this would only be significant if large amounts of energy
are not used to produce the hydrogen. For the same biomass production
rate from blue-green algae, several times this amount of hydrogen might
be expected.
140
-------
Total Systems Design
At the present time we have conceptualized two basic schemes for the
coupling of anaerobic digestion with algal-hydrogen producing systems.
System I is diagramed below.
algae return
pipe
hydrogen
production
unit
anaerobic adaptation chamber
low speed centrifuge
covered shallow
basin mass
culture algae unit
\
circulating pump
Figure D10;
Total algal-digester system design with algae introduced
directly into digester.
Description of System I
Cow manure is first introduced into the anaerobic digester. Digested
effluent is then pumped from the digester into the covered lagoon. The
lagoon is covered to prevent any NH4+ - NH3 losses. In the lagoon, manure
solids settle to the bottom and are either pumped back into the digester
for redigestion, or hauled out to the fields. The top few feet of the
lagoon is composed of a low solid algal growth sustaining liquid. This
liquid will be referred to as the supernatant liquid. Algae will grow
on the surface of the lagoon until sufficient population density, for mass
culture, is reached. Once adequate numbers are reached the supernatant
liquid is pumped into a shallow, covered, mass culture unit. During this
stage of operation algae population density will reach its maximum. Algae
will then be centrifuged to a liquid-paste consistency, with a low speed
centrifuge, and then pumped to an anaerobic chamber to incubate. Once
adapted, algae will be introduced directly into digester for hydrogen
production and subsequent digestion.
141
-------
gas pipe
• hydrogen production unit
with transparent cover
anaerobic adaptation chamber
—-low speed centrifuge
covered shallow
basin mass culture
algal unit
circulating pump
Figure Oil: Total system design using separate hydrogen producing chamber.
Description of System II
This system is essentially the same as system I until we get to the
hydrogen production unit. In this system algae are introduced into a
separate chamber outside the anaerobic digester. In this system algae are
subjected to an optimum hydrogen-producing environment. Hydrogen gas is
then piped into the digester and through the digesting medium. Once algae
have surpassed optimum hydrogen production levels, they are pumped into the
digester to undergo anaerobic digestion.
142
-------
REFERENCES AND BIBLIOGRAPHY
Ben-Amotz, A., D.L. Erbes, M.A. Riederer-Henderson, D.G. Peavy, and M.Gibbs.
1975. H2 Metabolism in Photo-synthetic Organisms: I. Dark H? Evolu-
tion and Uptake By Algae And Mosses. Plant Physiol. 56:72-77
Ben-Amotz, A., and M. Gibbs. 1975 H2 Metabolism In Photosynthetic Orga-
nisms: II Light-Dependent H2 Evolution By Preparations From
Chlamydomonas, Scenedesmus and Spinach. Biochem. Biophys Research
Comm. 64:355-359
Benemann, J.R., and N.M. Weare. 1974 Hydrogen Evolution By Nitrogen-
Fixing Anabae_na QyJ_[nd_Qca_ Cultures. Science 184:174-175
Benemann, J.R., and R.C, Valentine. 1972 The Pathways of Nitrogen
Fixation. Adv. in Mic. Physiol. 8:59-98
Boersma, L., E.W.R. Barlow, J.R. Miner, H.K. Phinney, and J.E. Oldfield.
1975 Protein Production Rates By Algae Using Swine Manure As A
Substrate. W.J. Jewell (Ed.) 1975 Energy, Agriculture and Waste
Management. Ann Arbor Science, Ann Arbor, Mich.
Brown, C.M., et al. Physiological Aspects of Microbial Inorganic Nitrogen
Metabolism. Adv. In Mic. Physiol. 11:1-45
Czerkawski, J.W., C.G. Harfoot, and G. Brechenridge. 1972 The Relation-
ship Between Methane Production And Concentrations Of Hydrogen In
Aqueous And Gaseous Phases During Rumen Fermentation In Vitro.
J. Applied Bact. 35:537-551
Dugan, G.L., C.G. Golueke, and W.J. Oswald. 1972 Recycling System For
Poultry Wastes. Jour. Water Poll. Control Fed. 44:432-440
Gaffron, H. 1942 Carbon Dioxide Reduction With Molecular Hydrogen In
Green Algae. Am. J. Bot. 27:273-283
Gaffron, H. 1942 Reduction Of Carbon Dioxide Coupled With The
Oxyhydration Reaction In Algae. J. Gen. Physiol. 26:241-268
Gaffron, H., and J. Rubin. 1942 Fermentative and Photo-
chemical Production of Hydrogen In Alaae. J. Gen.
Physiol. 26:219-240 "
143
-------
Gallon, J.R., T.A. LaRue, and W.G.W. Kurz. 1972 Characteristics Of
Nitrogenase Activity In Broken Cell Preparations Of The Blue-Green
Algae. Gloeocapsa sp. Canadian J. Microbio. 18:327-332
Gasper, E., L. Boersma, H.K. Phinney, and J.R. Miner. 1975 Utilization of
Waste Heat In A Biological System For The Conversion Of Swine Manure To
Single Cell Protein And Methane Gas. Presented at the 1975 Winter
Meeting of Amer. Soc. Agri. Engineers, St. Joseph, Mich.
Gest, H. 1972 Energy Conversion And Generation Of Reducing Power In
Bacterial Photosynthesis. Adv. in Mic. Physio!. 7:243-278
Golueke, C.G. Byconversion Of Energy Studies At The University Of
California (Berkeley). Univ. of Cal., Berkeley
Golueke, C.G., and W.J. Oswald. 1973 An Algal Regenerative System For
Single-Family Farms And Villages. Compost Science May-June:12-15
Golueke, C.G., and W.J. Oswald. 1959 Biological Conversion Of Light
Energy To The Chemical Energy Of Methane. Appl. Microbio. 7:219-227
Golueke, C.G., W.J. Oswald, and H.B. Gotaas. 1957 Anaerobic Digestion Of
Algae. Appl. Microbio. 5:47-55
Halderson, J.L. 1972 Dynamic Response Of An Anaerobic Digester With Dairy
Cow Manure Substrate. Ph.D. Thesis, Civil Engr. Dept., Purdue Univ.
Healey, F.P. 1970 The Mechanism Of Hydrogen Evolution By Chlamydomonas
moewusii. Plant Physiol. 45:153-159
Jackson, D.D., and J.W. Ellims 1896 On Odor And Tastes Of Surface waters.
(Referenced in Weissman and Benemann, 1977)
Jones, L.W., and N.I. Bishop. 1976 Simultaneous Measurement Of Oxygen And
Hydrogen Exchange From The Blue-Green Algae Anabaena. Plant Physiol.
57:659-665
Kaltwasser, H., T.S. Stuart, and H. Gaffron. 1969 Light-Dependent
Hydrogen Evolution By Scenedesmus. Planta. 89:309-322
Kessler, E. 1974 Hydrogenase, Photoreduction, And Anaerobic Growth.
pp 456-473 W.D.P. Stewart (Ed.), Algal Physiology And Biochemistry
Blackwell Scientific Pub!., London
Krampitz, L.O. 1977 Potentials Of Hydrogen Production Through Bio-
photolysis. Proceedings: Symposium On Clean Fuels From Biomass And
Wastes. Institute of Gas Technology, Chicago, 111.
Krauss, R.W., and W. H. Thomas. 1954 The Growth And Inorganic Nutrition
Of Scenedesmus obliquus In Mass Culture. Plant Physiol. 29:205
144
-------
Lapp, H.M., D.D. Schulte, E.J. Kroeker, A.B. Sparling, and B.H. Topnik. 1975
Start-up Of Pilot Scale Swine Manure Digesters For Methane Production.
Proceedings: The Third International Symposium on Livestock Wastes.
Champagne-Urbana, 111.
Morris, J.6. The Physiology Of Obligate Anaerobiosis. Adv. in Microb.
Physio 1. 12:169-233
Newton, J.W. 1976 Photoproduction Of Molecular Hydrogen By A Plant-Algal
Symbiotic System. Science 191:559-561
Oswald, W.J., and C.G. Golueke. 1964 Solar Power Via A Botanical Process.
Mechan. Engin. 86:40-43
Rabinowitch, and Govindjee. 1969 Photosynthesis John Wiley & Sons, Inc.,
N.Y.
Rao, K.K. L. Rosa, and D.O. Hall. 1976 Prolonged Production Of Hydrogen
Gas By A Chloroplast Biocatalytic System. Biochem. Biophys. Research
Comm. 68:21-28
Rowe, T. 1976 Photosynthetic Reclamation Of Animal Wastes Via Algae.
Research Report Univ. of Wis.-Green Bay, Wis.
Rychert, R.C. and J. Skujins. 1974 Nitrogen Fixation By Blue-Green Crusts
In The Great Basin Desert. Soil Science Sept.-Oct.
Singh, R.N. 1961 Role Of Blue-Green Algae In Nitrogen Economy Of Indian
Agriculture. Indian Council of Agricultural Research.Banaras Hindu
Univ., New Delhi
Smith, P.N., and P.J. Shuba. 1976 Terminal Anaerobic Dissimilation Of
Organic Molecules. Dept. of Microbio., Univ. of Florida, Gainesville,
_Tau
Stein, J.R. 1973 Handbook of Phycological Methods Cambridge University
Press, Cambridge
Stewart, W.D.P. 1974 Algal Physiology and Biochemistry Botanical Mono-
graphs Vol. 10. Univ. of California Press
Stuart, T.S. 1971 Hydrogen Production By Photosystem I Of Scenedesmus:
Effect Of Heat And Salicylaldoxime On Electron Transport And
Photophosphorylation. Planta 96:81-92
Stuart, T.S., and H. Gaffron. 1972(a) The Gas Exchange Of Hydrogen-
Adapted Algae As Followed By Mass Spectrometry. Plant. Physio!.
50:136-140
Stuart, T.S., and H. Gaffron. 1972(b) Mechanism Of Hydrogen Photopro-
duction By Several Algae. Planta 106:91-100
145
-------
Stuart, T.S., and H. Gaffron. 1972(c) The Mechanism Of Hydrogen Photo-
production By Several Algae, II: Contribution Of Photosystem II.
Planta 106:101-112
Stuart, T.S., and H. Gaffron. 1971 The Kinetics Of Hydrogen Photo-
production by Adapted Scenedesmus. Planta 100:228-243
Thomas, J., and K.A.U. David. 1971 Studies On The Physiology Of Heterocyst
Production In The Nitrogen-Fixing Blue-Green Algae Anabaena sp. In
Continuous Culture. J. Gen. Microbio. 66:127-131
Tilden, J.E. 1968 The Algae And Their Life Relations Hafner Pub. Co., N.Y.
Yates, M.G., and C.W. Jones. 1974 Respiration And Nitrogen Fixation In
Azotobacter. Adv. in Micro. Physio!. 11:97-130
Ward, M.A. 1969(a) Whole Cell And Cell-Free Hydrogenases of Algae.
Phytochem. 9:259-266
Ward, M.A. 1969(b) Adaptation of Hydrogenase In Cell-Free Preparations
From Chlamydomonas. Phytochem. 9:267-274
Weissman, J.C., and J.R. Benemann. 1976 Hydrogen Production By Nitrogen-
Starved Cultures Of Anaebena cylindrica. App. Env. Microbio.
33:123-131
Wolfe, R.S. 1971 Microbial Formation Of Methane Adv. in Microb. Physio!.
6:107-145
146
-------
APPENDIX E
TABLE El. ASSUMPTIONS USED IN ECONOMIC EVALUATIONS
ANAEROBIC DIGESTERS
To show returns on energy production alone, it was assumed that:
1. A covered anaerobic lagoon exists. This was chosen as
the best possible manure handling system in terms of organic
nitrogen conservation. The anaerobic digester is an
additional investment to the lagoon.
2. Annual capital costs are based on 9% percent interest
rates on loans to farmers.
3. Equipment maintenance and repair for equipment other than
engine generators is 2 percent of the capital invested. The
engine generator maintenance figures are based on manufacturers
recommendations for annual maintenance and overhauls every
six years.
4. Labor to operate and maintain the digester is an additional
four hours per week at $3 per hour.
5. Taxes and insurance are based on 3 percent of the capital
equipment costs.
6. Tax deductions from interest paid and straight line depreciation
are based on amount allowable for a 30 percent tax bracket.
7. One hundred percent of the biogas is used to produce electricity
or is combusted directly.
8. Annual operating costs are constant for the first 10 years, after
which they rise at 3 percent per year.
147
-------
TABLE E2 COSTS OF INSTALLED COMPONENTS USED IN ANAEROBIC DIGESTER SYSTEM
Fermentation Tank (25,000 gal.) $ 5,000 - 7,000
Pumps (2 diaphragm) 1 50o
Controls lf'500
1,500
Hydraulics 2 500
Building to house equipment 3,' 600
High pressure storage 2^000
Electrical tie in l)500
Well '500
Boiler and heat exchanger 1,300
Gas engine-generator (12 kw) 12^000
Diesel -generator (12 kw) 15JOOO
Heat recovery equipment 5^000
Miscellaneous
Engineering
TABLE E3: ANNUAL COSTS OF DIGESTER WITH HEAT RECOVERY USING DEISEL POWERED
GENERATOR
Initial Investment ~
aHnf ?LatcC°StS $12,774.00 $ 7,716.00 $ 5,556.00
Operating Costs 3,245.00 3,245.00 3 245.00
Equipment Maintenance $1,174
and repair
Labor 624
Taxes & Insurance 1,161
Diesel Fuel 286
TOTAL ANNUAL COSTS 16,019.00 11,061.00 8,811.00
Annual cost per cow 67.32 74. 46 91>74
normalized over 20 years
148
-------
TABLE E4: COSTS OF INSTALLED COMPONENTS FOR DIGESTER SYSTEM
WITH DIRECT COMBUSTION OF BIOGAS ON FARM
Digester Tank $ 5,000
Pumps 1,500
Controls 1,500
Piping 1,000
Boiler and heat exchange 1,300
Building 3,600
Miscellaneous (1556) 2,085
Engineering (15%) 2,400
TOTAL INVESTMENT IS,365
TAB1E E5: ANNUAL COSTS OF DIRECT COMBUSTION DIGESTER
5 year 10 year
Annual Capital Costs $ 4,788.00$ 2,929.00
Operating Costs 1,268.00 1,269.00
Equipment $ 320.00
Labor 468.00
Tax & Insurance 480.00
TOTAL ANNUAL COSTS 6,056.00 4,196.00
Annual Cost per cow
averaged over 10 years 36.62 41.96
149
-------
TABLE E6: HEAT LOSS FROM RICE LAKE
FERMENTATION TANK
ASSUMPTIONS: 1) Tank temperature kept at 95 F 24 hours
per day, 365 days per year.
2) Ground temperature around tank is con-
stant 50 F year round.
3) Area of the fermentation tank is 2,036
square feet.
4) Effective U value of tank and insulation
u «. -, , , is '2 Btu/ft2/F.
Heat loss calculation:
45 T x .2 x 24 x 365 x 2036 = 160.52 x 106 Btu/yr.
TABLE E7: THEORETICAL DAILY BIOGAS PRODUCTION
USING DAIRY - MSW MIX*
LBS.
Manure - MSW 9.56 x 103
Total Solids 1.69 x 103
Volatile Solids 1.29 x 103
Biogas production
100% capacity 5.61 x 103 cf
85% capacity 4.77 x 103 cf
Btu output per day 2.767 x 106 Btu/day
100 Animal Units 68% volatile solids -Manure
85 Ibs. manure/day 32% volatile solids -MSW
63% total solids -Manure 4.35 ch biogas/lb VS/\
37% total solids -MSW
*Based on 63% dairy and 37% MSW
150
-------
2 p.m
4 6
FIGURE El. DAILY ELECTRICAL LOAD PROFILE, RICE LAKE FARM
-------
APPENDIX F
FACTORS UNDERLYING THE ADOPTION OF
AGRICULTURAL INNOVATIONS
(Factor)
Cost
1.
2.
Initial cost
Continuing (operating)
costs
Returns
3. Rate of cost recovery
4. Monetary payoff
Profitability
5. Extent of economic ad-
vantage over alternatives
6. Replacement status (con-
dition) of existing equipment
"Efficiency"
7. Saving of time
8. Saving of discomfort
Risk and Uncertainty
9. Regularity of reward
(Shows results every time)
10. Divisibility for trial
(How easy to try; first,
on a small scale)
Congruence
11. Association with main
enterprise (dairying)
12. Advantage (overall sig-
nificance of the practice
for the entire farm program)
(Relative importance/comment)
Not an important deterrent
Minor significance
Minor significance
Significant
Significant
Significant
Minor significance
Minor significance
Significant
Significant
Significant
Significant
152
-------
13. Pervasiveness (of conse-
quences of adoption; leads
to other changes or practices)
Communicibility
14. Complexity (of understanding
and use)
15. Clarity of results (How
clearly do results show?)
Farm Characteristics
16. Farm size (total acreage,
dairy herd size)
17. Location of farmstead
(proximity to others; coop
unit possibilities)
Farm Financial Position
18. Capital (Curvilinear effect;
Proxies-estimated value of
land, buildings, equipment,
and livestock)
19. Current income (proxies-
gross farm sales, milk
receipts)
Farmer Characteristics
20. Age
21. Education
22. Tenure (owner/renter)
23. Years farmed
24. Attitudes and values
(proxies-affiliation with
farm organizations, activity
with social groups, contact
with "science" via reading
magazines and extension bulletins)
Regualtory and Legal
25. Pollution control (manure
handling rules and policy)
26. Safety requirements
27. Zoning and land use regulations
Institutional and Infrastructure
Not an important deterrent
Minor Significance
Minor significance
Significant
Significant
Significant
Significant
Significant
Significant
Significant
Significant predicator of
adopting environmental measures.
Significant
Significant
Precondition
Precondition
153
-------
28. Availability of technical Condition for widespread
support (engineering firms, adoption
university extension)
29. Convenience of repair and Condition for widespread
maintenance services adoption
30. Supportive (and familiarity Condition for widespread
with the innovation) farm adoption
finance and insurance sector
31. Information sources (mass
media-radio, TV, newspapers Significant for diffusion
and magazines; friends and
neighbors-mostly other farmers;
agricultural agancies-univ. ex.,
SCS, farm credit agencies and
the like; vendors and potential
suppliers)
The factors listed above were adopted, with modifications and expansion
from:
Cancian, Frank, "Stratification and Risk-Taking: A Theory Tested on
Agricultural Innovation", American Sociological Review, 1967.
Fliegel, Frederick C. and Joseph Kivlin, "Farm Practice Attributes and
Adoption Rates", Social Forces, 1962.
, "Attributes of Innovations as Factors of Diffusion", American
Journal of Sociology, 1966.
Kronus, Carol L. and J.C. van Es. "The Practice of Environmental Quality
Behavior", Journal of Environmental Education, 1976.
Mansfield, Edwin; The Economics of Technological Change, W.W. Norton, 1968.
Pampel, Fred Jr. and J.C. van Es;"Environmental Quality and Issues of
Adoption Research", Rural Sociology, 1977.
154
JEB/July 77
-------
APPENDIX G
EXISTING OR PLANNED DIGESTER OPERATIONS IN USA
LOCATION TYPE OF INFL
Guymon, OK
(no.l)
Guymon, OK
(no. 2)
Ripon, WI
Mt. Pleasant, IA
McCabe Farm
Ames, IA
State Univ.
Edwall, WA
Aurora, OR
Hartford, NY
Cattle
Cattle
Poultry
Swine
Beef
Cattle
All types
Cattle
TYPE
Full Scale
50,000 gal.
Full Scale
50,000 gal.
Full Scale
50,000 gal.
Full Scale
50,000 gal.
Pilot Scale
1800 gal.
Pilot Scale
Pilot scale
TYPE OF DIG. OPERATION
Continuous
feed
Complete mix,
Mesophilic
Continuous
feed, Complete
mix, Mesophilic
Continuous
feed, complete
mix, Mesophilic
Gas mixing
Complete mix
Mesophilic
Continuous
feed
Modular
Plug-flow
yes
yes
yes
yes
No
No
No
yes
IS GAS BEING USED WHO BUILT THE SYSTEM
In-house purposes
(sold to pipeline)
In-house purposes
(sold to pipeline)
Heat House
No, boiler
No
Experimentally
No
yes
Thermonetics
ific Recovery
Thermonetics
ific Recovery
Wayne Gibbons
Harold McCabe
University of
Earth Cyclers
David House
Bill Jewell,
Calor-
Calor-
LO
i-H
Iowa
Hartford, NY Cattle Designed Conventional Mo, in
for 65 Cattle Mis building
stage
No-space heat
Cornell
Bill Jewell,
Cornell
-------
APPENDIX G
PART II
LOCATION TYPE OF INFL
Hartford, NY
Dewey, IL
Lancaster, PA
Wyoming, MN
Verio Larson
Jefferson, WI
Habeck Farm
NorthGlenn, Co.
Karl 's Dairy Farm
(owned by city of
NorthGlenn)
Monfort, CO
Monfort feed lot
Bartow, FL
Kaplan Feedlot
Cattle
Poultry
Poultry
Swine
Dairy
Cattle
Dairy
Cattle
Dirt lot
Waste
Cattle
(feed lot
waste)
TYPE " TYPE OF DIG. OPERATION IS GAS BEING USED WHO BUILT THE SYSTEM
Designed Plug-flow no
for 65 "Bage type" (in build-
Cattle ing stage)
Full Scale Plug-flow no (in
25,000 chickens building
(5xl06 Btu/day) stage)
Full Scale Plug-flow no (in
75,000 chickens building
(15xlOb Btu/day) stage)
Full Scale Continuous yes
10,000 gal feed complet-
ely mixed
Full Scale Batch load no, not
No (space heating)
No (eventually)
run engine
Electric generator
and crop dryer
Heating digester
Heating digester
Bill Jewell
Cornell
Bill Jewell
Cornell
Energy Recovery
Phil Goodrich g
Univ. of MN ^
A.O Smith Corp.
32,000 gal. completely since 1976
mixed verti-
cal tank
Full Scale Plug flow no, (in
400 cows design
stages)
Pilot plant continuous yes
110 Ibs/day feed, completely
mixed
Full Scale Continuous or no (in
570,000 gal. semi-continuous building)
(in building feed-
completely stage)
mixed, thermophilic
Will run genera-
tor 52 KW
yes
no
Energy Harvest
Schaeffer &
Roland, Chicago
Hamilton Stand-
ard
Hamilton Stand-
ard
-------
APPENDIX G
PART III
LOCATION
Arvada.CO
Lamar, CO
West Union, IA
Sunny Times
Farm
Greeley, CO
Miller Feed lot
Henderson, MN
No. 1
Drury, MO
New Life farm
TYPE OF INFL.
Cattle
(feed lot)
Cattle
(feed lot
manure)
Chicken
Beef
Hog
Hay/Sludge
4,000 gal.
Custer, MI Cattle
Jim Allison farm
Rice Lake, WI
Schieffer Farm
Monroe, WA
Monroe State
Dairy Cow
Dairy cow
TYPE
Pilot plant
200 gal.
Full scale
6,000 gal.
Full Scale
160,000
layers
Full Scale
(2k x 106
Btu/day)
Full Scale
(5 x 106
Btu/day)
Full Scale
Full Scale
350 cattle
Farm Scale
110 cows
Full Scale
designed for
TYPE OF DIG.
Plug-flow
Mesophilic
Plug-flow
Mobile unit
Mesophilic
Gas mixed
Mesophilic
Plug-flow
Plug flow
Plug-flow
Plug-flow
bag type
Plug-flow
Complete mix
system, gas
OPERATION
yes
yes
yes
yes
yes (mi
April,1
yes
yes
yes
yes
IS GAS BEING USED
no
yes-to heat diges-
ter
yes-to heat diges-
ter and engine gen-
erator
heating building
and hot water
d Electric Generator
78)
no
yes-running 30 KW
generator
yes-running 12 KW
generator
yes-running
generator
WHO BUILT THE SYSTEM
Biogas of Colorado
Biogas of Colorado
Haeying Enterprises
Energy Recovery
Inc., Aravada ,-
u
1-
Energy Recovery
Inc., Aravada
Ted Landers
Energy Harvest
Shaeffer & Roland
Energy Harvest
Shaeffer & Roland
Ecotope Group
Dairy Farms
350 but uses recirculation
more
-------
APPENDIX G
END
.OCATION
TYPE OF INFL. TYPE
TYPE OF DIG. OPERATION IS GAS BEING USED WHO BUILT THE SYSTEM
Clay Center, NE
USDA Animal Meet
Research Center
Hastings, ME
Univers"ty ^ MO
Colimt 'a. M,:
University )f II
Champaign-Urbane
University Pa<"k
PA
Stillwater, MN
Cardiff, CA
Cattle
Cattle
Swine
Beef
Dairy
Poultry
Human
Vegetable
Sewage Pilot
Sludge and 200 gal
water hyacinth
Pilot Scale Semi-continuous
Lab Scale
40 sow herd
Pilot Scale
140 gals.
100 m3
Homestead
size
Semi-continuous
concrete stave
silo
completely mixed
stainless steel
tanks, thermophilic
Mixed 2 chamber
mesophilic
Plug-flow
mesophilic
Plug-flow
yes
yes
yes
yes
yes
yes
no (mainly inter- Hamilton Standard
ested in refeed
no
yes, some being
used for heating
Dr. Andy Hashimoto
Ag. Engineering Dept
of MO
no
yes Heating digester
Civil Engineering
Penn. State
University
00
Lf)
cooking and heating Rutan Research,
Al, Rutan
no
County owned water
treatment facility
------- |