PB83-183020
TO BIOGAS AND PROTEIN RECO?!!Y
Research Institute
s Folaad
Sstenaflta
-------
EPA-600/2-83-023
March 1983
OPTIMIZATION OF WASTES TREATMENT WITH
REFERENCE TO BIOGAS AND PROTEIN RECOVERY
by
Jan A. Oleszkiewicz
Szymon Koziarski
Research Institute on Environmental Development
Rosenberg6w 28, 51-616 Wroclaw, Poland
JB-5-534-7
Project Officer
Lynn R. Shuyler
Animal Production Section
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-83-023
2.
3. RECIPIENT'S ACCESSION>NO.
PB83-183020
4. TITLE AND SUBTITLE
Optimization of Wastes Treatment with Reference to
Biogas and Protein Recovery
5. REPORT DATE
March 1983
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Jan A. Oleszkiewicz and Szymon Koziarski
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Institute on Environmental Development
Rosenbergow 28, 51-616 Wroclaw, Poland
10. PROGRAM ELEMENT NO.
APBC
11. CONTRACT/GRANT NO.
JB-5-534-7
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Robert S. Kerr Environmental Research Laboratory
P.O. Box 1198
Ada, OK 74820
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
ProjectEconSuc£eaSwithin frame of bilaterial financial arrangement, the Maria
Sklodowska-Curie Fund for cooperative programs between Poland and the United States.
is. ABSTRACT Detailed tecnnoioglcal ana economic evaluation of the presently used
ment processes for the dilute wastewaters from hog farms, with capacity exceeding 10
thousand heads, is presented. The research part of the project was aimed at optimiz-
ation of the unit process and whole treatment trains selection, rather than unit
process operational parameters. The results indicate the need for diametrical shift
in research emphasis in animal wastes, towards high-rate, short detention time
anaerobic unit process combined with high-rate aerobic secondary treatment and
anaerobic-aerobic polishing treatment. Several full technological treatment trains
were evaluated and compared, from the standpoint of treatment efficiency, level of
recovery, ease of maintenance and economic efficiency indices. The economic analysis
has proved that the application of these new treatment trains can make industrial
scale farming more profitable with the increase of the size of the farm. The technol-
ogy proposed in the project will show increase of the economic efficiency, when
compared to conventional systems, with the increase of power costs, due to biogas
recovery and incorporation of sludge treatment subsystem in the overall treatment-
recovery train. Although the report is confined to swine wastes, the results are
applicable to other concentrated effluents from agricultural industry.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Swine; Agricultural Wastes
Treatment Processes;
Sedimentation; Coagula-
tion; Activated Sludge;
Anaerobic Digestion;
Anaerobic Biofilters
02/A, C, E
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
249
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
DISCLAIMER
Although the research described in this article has been funded wholly or
in part by the United States Environmental Protection Agency through contract
or grant JB-5-534-7 to Research Institute on Environmental Development - Poland,
it has not been subjected to the Agency's peer and policy review and
therefore does not necessarily reflect the views of the Agency, and no official
endorsement should be inferred. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
11
-------
FOREWORD
EPA is charged by Congress to protect the Nation's land, air and water
systems. Under a mandate of national environmental laws focused on air and
water quality, solid waste management and the control of toxic substances,
pesticides, noise, and radiation, the Agency strives to formulate and imple-
ment actions which lead to a compatible balance between human activities and
the ability of natural systems to support and nurture life. In partial
response to these mandates, the Robert S. Kerr Environmental Research Lab-
oratory, Ada, Oklahoma, is charged with the mission to manage research
programs to investigate the nature, transport, fate, and management of
pollutants in ground water and to develop and demonstrate technologies for
treating wastewaters with soils and other natural systems; for controlling
pollution from irrigated crop and animal production agricultural activities;
for controlling pollution from petroleum refining and petrochemical indus-
tries; and for managing pollution resulting from combinations of industrial/
industrial and industrial/municipal wastewaters.
This project was initiated to evaluate the presently used treatment
systems for wastes from large swine farms in Poland and to optimize the
system with the addition of a subsystem to produce biogas and to recover
protein. The project provided an opportunity to study large, full-scale
plants that do not now exist in this country. The results indicate that
biogas can be produced successfully from a much more dilute waste than had
been previously reported. It also optimized the systems both from a techno-
logical and economic standpoint. The information will be very useful in
this country as urban pressure requires more complex treatment of animal
wastes in the future.
Clinton W. Hall, Director
Robert S. Kerr Environmental
Research Laboratory
iii
-------
ABSTRACT
Detailed technological and economic evaluation of the presently used
treatment processes for the dilute wastewaters from hog farms, with capacity
exceeding 10 thousand heads, is presented. The present systems of treatment
for stream disposal encompass sophisticated multi-stage chemical-biological
treatment with high unit costs due to consumption of power, oil and chemicals.
The research part of the project was aimed at optimization of the unit
process and whole treatment trains selection, rather than unit process opera-
tional parameters. The unit processes investigated in laboratory and pilot
scale included: sedimentation, coagulation, activated sludge as a roughing
and as a polishing unit, algal-bacterial (oxidation) polishing ponds, anaero-
bic digestion in flow-through and contact reactors with suspended microorga-
nisms and in anaerobic biofilters, anaerobic ponds, aerated lagoons, and yeast
generation, as a method of treatment and protein recovery.
The results indicate the need for diametrical shift in research emphasis
in animal wastes, towards high-rate, short detention time anaerobic unit
process combined with high-rate aerobic secondary treatment and anaerobic-
aerobic polishing treatment. Several full technological treatment trains
were evaluated and compared, from the standpoint of treatment efficiency,
level of recovery, ease of maintenance and economic efficiency indices. The
systems recommended comprised of anaerobic biofiltration or contact digestion
followed by anaerobic biofiltration, anaerobic biofiltration and reaeration,
with anaerobic sludge digestion as a separate sludge train or incorporated
in the wastewater treatment train. The economic analysis has proved that
the application of these new treatment trains can make industrial scale
fanning more profitable with the increase of the size of the farm. This is
contrary to the presently observed trend toward limiting the construction of
IV
-------
large farms, due to the environmental constraints. The discouraging experi-
ences stem from the application of either conventional wastewater treatment
technology to these concentrated effluents, or application of agricultural
utilization practices as used for concentrated manures from smaller farms.
The technology proposed in the project will show increase of the economic
efficiency, when compared to conventional systems, with the increase of power
costs, due to biogas recovery and incorporation of sludge treatment subsystem
in the overall treatment-recovery train. Although the report is confined to
swine wastes, the results are applicable to other concentrated effluents from
agricultural industry.
This project has been conducted within the frame of a bilaterial financial
arrangement, the Maria Sklodowska-Curie Fund for cooperative programs between
Poland and the United States. The work has been accomplished between October
1, 1976 and November 30, 1980, by the Research Institute on Environmental
Development - Wroclaw Division and U. S. Environmental Protection Agency -
Robert S. Kerr Environmental Research Laboratory in Ada, Oklahoma.
-------
CONTENTS
Foreword ..............................
Abstract .............................. iv
Figures .................. ............ vlii
Tables ............................... xiii
Abbreviations and Symbols ..................... xv
Acknowledgment ........................... xviii
1. Conclusions ......................... 1
2. Recommendations ....................... 4
3. Introduction ........................ 6
Project aim ....................... 6
Project scope ...................... 8
4. Present Piggery Waste Treatment Practice .......... 10
Hog production and effluent treatment trends ....... 10
Wastewater quality and quantity ............. 12
Treatment Systems used; problems encountered ....... 18
Efficiency of treatment plants presently in use ..... 22
Economic efficiency of treatment ............. 28
Discussion and Conclusions ................ 31
5. Sampling and Analytical Methods ............... 37
Sampling ......................... 37
Analytical methods .................... 37
6. Primary and Secondary Aerobic Treatment ........... 40
Sedimentation ...................... 40
Coagulation of piggery wastewaters ............ 44
Activated sludge-secondary treatment ........... 50
7. Polishing Treatment ..................... 58
Activated sludge polishing treatment ........... 58
Algal treatment systems ................. 62
Polishing Anaerobic biofiltration ............ 73
VI
-------
8. Anaerobic Treatment 77
Introduction 77
Batch digestion 81
»
Continuous studies in ANFLOW reactors 84
Anaerobic digestion in ANCONT reactors 110
Anaerobic biofiltration in ANBIOF reactors 124
Discussion and conclusions 137
9. Production of Yeasts 142
Introduction ..... 142
Batch studies 143
Dynamic studies 150
Discussions and Conclusions 161
10. Wastewater Ponds Full Treatment Systems 164
Methods 164
Results and discussion 166
11. Protein Recovery 170
Introduction 170
Direct recovery 171
Conversion into bacterial SCP 173
Conversion into algal protein 175
Conversion into yeast protein 176
Nutritional value of recovered protein 177
Discussion and conclusions 181
12. Economics of Proposed Treatment Technologies 184
Outline of the economic analysis 184
Comparison of various systems 190
Combined treatment with other effluents 207
Conclusions 213
References 216
vii
-------
FIGURES
Number Page
1 (A) Overall hog production in two countries; (B) Planned
production increase of the industrial farms in Poland .... 7
2 Layout and cross-section of an Agrokomplex Farm A with 10.5
thousand head 11
3 (A) Water use at Farm D and Farm A before modification in
1979-curve 1, and after modification - curve 2; (B) Varia-
bility of wastewater concentration at Farm D 14
4 Raw wastes characteristics: (A) Concentration probability
based on pooled data from 14 farms; (B) Regression of COD ,
on total solids 16
5 Layout of Vidus type Plant A modified by authors 19
6 Presently used treatment systems for dilute piggery waste ... 20
7 (A) Efficiency of screening at Plant A and sedimentation at
Plant C; (B) BODg removal efficiency at Plant A as affected
by sludge loading - F/M 23
8 (A) Activated sludge performance at three Vidus plants;
(B) Temperature correction calculation for Plant A data ... 26
9 (A) Distribution of raw and effluent wastes COD after modifi-
cation; (B) Influence of COD ,. sludge load on effluent
quality; - Plant A data 27
10 Efficiency of unit treatment processes at two full scale
plants 28
11 (A) Capital costs of piggery farms and waste treatment plants -
WTP; (B) Economic efficiency of treatment as affected by WTP
size 30
12 Economic efficiency versus load removed on the basis of BODC ,.
5,nf
and COD _ 33
n±
Vlll
-------
Number
13 (A) Comparison of 10 minutes and 2h COD digestion; (B) Cali-
bration of COD against total solids - IS 38
14 Effects of preaeration on removal of: (A) Suspended solids;
(B) Organics; 1 - raw, fresh waste and 2 - after 0.5h aeration
on 24.IV.78; 3 - raw, fresh waste and 4 - after 19h preaera-
tion on 25.V.79 42
15 (A) Settling curves for Plant A raw wastewaters: line 1 - S =
12,500 mg TSS/dm3, 2 - aerated S = 12,700 mg TSS/dm3,
3 3
line 3 - S = 6,770 mg TSS/dm , 4 - aerated S = 6,230 mg/dm ;
(B) COD removal efficiency during settling of three plants
wastewaters 43
16 Effects of coagulation: (A) Waste stored for two days - 1978,
1979 data; (B) Fresh wastes - 1980 data 47
17 Comparison of coagulation effects with settling: (A) Effi-
ciency of COD , removal by various processes; (B) Sludge
volume at various alum doses 48
18 Layout of experimental activated sludge set-up 51
19 Effects of coagulation on COD removal rate in (batch) activated
sludge 52
20 Continuous activated sludge treatment of system (I) C-AS and
(I) S-AS-C: (A) CODf removal across the aeration tank; (B)
Overall sludge produced in the whole treatment train .... 54
21 Microscopic picture (40 X) of activated sludge: at t = Oh
tanks 1, 2 and 3 are respectively A, C, and E; at t = 24h
tanks 1, 2 and 3 are respectively B, D, and F: 1) agglom-
erated floes, 2) Zooglea uva Kolkw., 3) Opercularia sp. -
poor physical condition colony, 4) Opercularia sp. - two
individuals colony of good condition 56
22 Continuous extended aeration of activated sludge effluent from
Farm A: (A) Efficiency as affected by aeration time; (B)
Effluent quality versus sludge loading in the test units . . 60
23 Averaged effects of treatment in a series of four ponds. An
example for 18 and 25 days HRT runs 66
-------
Number PaSe
24 Efficiency of BOD- - removal and permanganate COD in the
3»i
effluent from four ponds system 69
25 Kinetics of BOD,. - removal in series of four ponds: (A) First
•*»t
order; (B) Authors' model - substrate kinetics data pool . . 70
26 Kinetics of BOD- f removal in algal ponds; four influent
3»t
concentration ranges - authors' substrate model 71
27 Efficiency of algal biomass production and optimum retention
for organics removal in algal ponds 72
28 Removal kinetics in full scale anaerobic polishing
biofilters 75
•
29 Layout of various anaerobic treatment process modifications
studied 78
30 Pressurized glass reactor used as batch and ANFLOW reactor . . 82
31 The course of batch anaerobic digestion in the third series of
experiments 85
32 Carbon and BOD content, versus COD 90
33 Anaerobic biodegradability of screened piggery wastes: (A)
Based on total volatile solids; (B) Based on COD _ 91
nt
34 VSS content and removal of COD in ANFLOW reactor 92
35 Gas production in ANFLOW reactor: (A) Overall average;
(B) Specific gas production based on COD input 93
36 Kinetics of the ANFLOW reactor: (A) Removal of biodegradable
BS; (B) Gas production; (C) Gas production from TVS intro-
duced (SGPa) 95
o
37 Gas production from removed BODC - i COD - ANFLOW (Based on
j,nr
daily loadings) 96
38 Gas production from biodegradable VS (BVS): (A) Based on
daily loads; (B) Volumetric 97
39 COD, BOD and SOC removal efficiency versus COD load in ANFLOW;
unit CH, production from COD 98
40 (A) COD dependence on TVS; (B) Morris, Jewell, Loehr kinetics
of BVS removal 100
-------
Number Page
41 Zero and first order correlations of COD removal kinetics:
(A) With active biomass (o) and zero order plot (•); (B)
Without biomass 101
42 Growth kinetics correlation by two methods of interpretation -
BOD- removal - ANFLOW reactor: (A) Eckenfelder and Ford;
(B) This study 103
43 Removal kinetics in ANFLOW reactor: (A) Michaelis (MBH) second
order; (B) Phase breakdown first order 104
44 Pseudo-first order correlation of COD load versus remaining
k
fraction of biodegradable COD 106
45 Layout of the ANCONT anaerobic digester Ill
46 Effects of hydraulic retention and organic loading on SRT,
ANCONT pH and effluent quality 115
47 Effects of hydraulic retention on: (A) Gas production (B) Ef-
fluent quality; (C) pH in the ANCONT reactor 116
48 (A) Kinetics of organics removal; (B) Effects of sludge age on
effluent BOD and pH in the ANCONT reactor 117
49 Gas production from removed COD - and efficiency of COD -
removal in ANCONT reactor 119
50 Kinetics of biogas production from the COD f introduced .... 121
51 Kinetics of BODC , and COD , removal in ANCONT reactor .... 123
j ,t nr
52 Layout of an anaerobic biofilter - ANBIOF - arrangement .... 126
53 Effects of organic loading on: (A) COD . removal; (B) ANBIOF
pH, methane content and alkalinity 129
54 Effects of CODnf loading on: (A) Effluent quality; (B) Solids
removal 130
55 Effects of COD - loading on composition of biogas from ANBIOF
reactor 131
56 Effects of COD loading on CH. production from: (A) COD ;
4 rem
(B) TVS ; (C) TKN and N-org removals 132
rem
57 Kinetics of CODf removal in ANBIOF: (A) Pseudo-first order
reaction; (B) Against sludge age SRT 135
58 Operational range of ANBIOF reactors 138
XI
-------
Number Page
59 Batch yeast production studies: (A) Experimental set-up;
(B) COD and TKN removals attained 146
60 Semi-dynamic yeast fermentation equipment 152
61 Semi-dynamic (batch-fed) fermentation: (A) Specific growth rate;
(B) Biomass increase 158
62 Layout of the wastewater ponds system 165
63 Effects of overall retention on organics removal in the ponds
system 167
64 Nitrogen compounds and pH in the effluent from polishing pond II
versus overall retention in the system 168
65 Layout of System I and II 191
66 Layout of System III 196
67 Layout of System IV 198
68 Layout of System V 199
69 Layout of System VI 200
70 Layout of System VII and VIII 203
71 Layout of combined wastes management - recycle system for pig farm
and yeast plant 208
72 Concept of combined municipal - piggery wastes treatment system . . 211
73 Concept of combined treatment with chemical industry wastes .... 214
XI1
-------
TABLES
Number Page
1 Theoretical Estimation of Pollutant Concentrations in Effluents
from Large Piggeries and Comparison with Actual Averages for
Plant D 15
2 Average Concentrations of Pollutants in Wastewaters from 14
Large Farms 17
3 Economic Efficiency of Five Studied Vidus-Type Treatment
Plants 32
4 Comparison of Three Pretreatment Methods - Batch Activated
Sludge 55
5 Activated Sludge Polishing Treatment of Effluent from the
Vidus-Type Plant at Farm A 59
6 Comparison of Intermittent Aeration in Polishing Activated
Sludge Treatment of Biological Effluent 61
7 Characteristic Overall Removal Efficiencies Attained in the
Series of Four Algal Ponds 67
8 Biogas Production in Series 3 83
9 Specific Gas Production from Introduced Organic Load - Batch
Study 84
10 Results of ANFLOW Reactor Performance 88
11 Comparison of Specific Gas Production SGP Rates as Quoted by
Different Authors 107
12 Summary of the ANCONT Performance Data 113
13 Results of ANBIOF Performance 127
14 Comparison of Optimum Design Parameters and Yields in Anaerobic
Reactors 140
15 Results of Batch Fermentation of Raw Centrate of Piggery
Effluent, by Yeast of Candida Type 145
Xlll
-------
Number Page
16 Biomass Yield in Batch Yeast Fermentation 148
17 Yeasts Production on Filtered Piggery Wastes Without Carbon
Enrichment - Removal Efficiencies 154
18 Tests Without Carbon Enrichment - Yeasts Yields and Nutrients
Use 155
19 Yeast Production on Filtered Piggery Wastewaters with Carbon
Enrichment - Removal Efficiencies 157
20 Tests with Carbon Enrichment - Yeasts Yields and Kinetic
Data 160
21 Wastewater Pond System's Parameters 164
22 Overall Efficiency of Wastewater Treatment in the Ponds
System 166
23 Total and Digestible Protein in Piggery Effluent Waste
Materials 173
24 The Composition of Silage Prepared with Excess Activated
Sludge 174
25 Amino Acids Composition of Protein from Raw Wastes Screenings,
Excess Activated Sludge and Duckweed 175
26 Amino Acids Content in Protein Recovered from Symbiotic Algal -
Bacterial Biomass 179
27 Amino Acids Content in Protein Recovered from Yeasts Grown on
Filtered Raw Piggery Wastes Enriched with Sucrose 180
28 Nutritional Value of Proteins Based on Methionine
Deficiency 181
29 Unit Energy Costs 189
30 Capital Costs for System II 194
31 Capital Costs for System VI 201
32 Economic Efficiency Indices for the Studied Piggery Wastewater
Treatment Systems 204
33 Cost Optimization for the Three Combined Treatment Plants . . . 210
xiv
-------
LIST OF ABBREVIATIONS AND SYMBOLS
AA
AB
AD
ANBIOF
ANCONT
ANFLOW
AL
AS
BOD
BVS
CMR
CNP
COD
CTP
D
DC
DI
DM
DO
DP
DPW
FAO
F/M
GC
GP
GP
avg
HRT
K
K
amino acid
anaerobic biofilter
anaerobic digestion
anaerobic biofilter
anaerobic contact digester with sludge recycle
anaerobic flow-through digester without sludge recycle
2
area loading (g 0_/m d)
activated sludge
biochemical oxygen demand
biodegradable volatile solids
— chemical oxygen demand
— combined treatment plant
3
— coagulant dose (mg/dm )
— direct costs
— digestion index
— dry matter (synonym of TS)
— dissolved oxygen
— digestible protein
— dry poultry waste
— food and agriculture
— food to microorganism ratio (kg 02/kg MLVSS d)
— general costs
3 3
— gas production from reactor volume (m /m )
3
— total gas production expressed per day (m /d)
— hydraulic retention time (d)
— removal rates
— Michaelis constant; i.e. substrate concentration
at which; y = 0.5 um
3
— organic loading of a reactor (kg/m d)
xv
-------
LW
MISS
MLSS
MLVSS
N
OM
R
RWL
S
o
S
e
Snb
S
SCP
SGP
'avg
SGP
SGP
SGP
a
"o,r
SOC
SRT
SS
STP
SVI
TF
TKN
TOO
TOTEM
TP
TS
TSS
TTC
TVS
— load introduced (kg/d)
— load removed (kg/d)
— liveweight
— mixed liquor suspended solids
— Xv (mg/dm3)
— removal efficiency (S - S )/S ; (%)
o e o
— operation and maintenance costs
— nonbiodegradable fraction
— raw waste load
3
— influent substrate concentration (mg/dm )
— effluent substrate concentration (mg/dm )
3
— nonbiodegradable substrate (mg/dm )
— Sr = SQ-Se (mg/dm3)
— single cell protein
— specific gas production; total volume of gas produced
3
divided by the total load throughput (m /kg)
— specific gas production; GP divided by the average
daily load input (m /kg)
— specific gas production based on load removal (m /kg)
3
— averaged SGP from overall cu-ulative data (m /kg)
— soluble organic carbon
— solids retention time (d)
— suspended solids
— standard temperature and pressure
— sludge volume index
— trickling filter
— Kjeldahl nitrogen
— total organic carbon
— total engery module
— total protein
— total solids
— total suspended solids
— dehydrogenase
— total volatile solids
xvi
-------
VSS — volatile suspended solids
2
X — assumed active (viable) cells concentration (mg/dm )
a 3
X — VSS at the end of treatment or in the effluent (mg/dm )
e 3
X — mixed liquor volatile suspended solids (mg/dm )
Y — yield coefficient = mass of synthesized cells divided
by mass of removed substrate (mg BSS/mg S )
b — endogenous respiration coefficient = mass of VSS used
per unit VSS in the system (mg VSS/mg VSS d) (1/d)
dm3 — liter
f — subscript denoting filtered sample
nf — subscript denoting non-filtered sample
q — substrate removal rate; q = S /X HRT ; (1/d)
r — X /X smaller than unity by def. (-)
U — growth coefficient; u = QY = \i S/(K +S); (1/d)
in s
y — maximum growth coefficient u = qY; (1/d)
m m TQ
zL — zloties, unit of Polish currency. $1 = 20 zloties
xvi i
-------
ACKNOWLEDGEMENTS
The report is authored by Jan A. Oleszkiewicz, Ph.D. - Principal Investi-
gator and Szymon Koziarski, M. Sc. - Project Co-Investigator. At the time of
the study, J. A. Oleszkiewicz was Head of the Research and Development Depart-
ment (R&D Dept.) - Research Institute on Environmental Development (RIED),
Wroclaw Division. He is presently with Duncan, Lagnese and Assoc., 3185
Babcock Boul., Pittsburgh, PA, 15237, USA. Mr. S. Koziarski is Chief of Animal
Waste Section at R&D Dept., RIED, Wroclaw.
The work has been performed by the staff of Animal Wastes Section -
Research and Development Dept., RIED Wroclaw Division, (in alphabetical order):
R. Domaradzki, M. Sc.-chemist; K. Jankowski, M. Sc.-sanit. engr.; J. Janson,
tech.; P. Ksiazek, tech.; K. KosiAska, M. Sc.-biologist; Z. Kwiatkowski, M.
Sc.-sanit. engr.; A. Lukasinska-Janiczek, M. Sc.-sanit. engr.; B. Majcher, M.
Sc.-chem. engr.; M. Pietralik, tech.; G. Ryznar, tech. The cooperation of
other persons from RIED, School of Management and Economy - Institute of
Chemical and Food Industry Technology (SME) and Wroclaw Technical University
(WTU) is acknowledged in alphabetical order: A. G6lcz, M. Sc.-RIED; A.
Grzesiak, M. Sc.-RIED; M. Karwecki, tech.-RIED; H. Kieloch, M. Sc.-WTU;
P. Ladog6rski, Ph.D.-Gas Works; T. Miskiewicz, Ph. D.-SME; M. Nawrocka, M. Sc.-
RIED; M. M. Sozanski, Ph.D.-WTU; K. Szczesny, M. Sc.-RIED; A. Wojda, tech.-
RIED; J. ZieliAski, tech.-Farm A; J. Ziobrowski, Prof.-SME; and personnel of
the Department of Wastewater Treatment and Reuse - RIED.
The help of Mr. W. Zalewski, M. Sc. in instrumental TOG, SOC analyses, is
gratefully acknowledged.
Particular thanks are due to Dr. Pawel Blaszczyk, Director - RIED, Warsaw,
and Ms. S. Ramotowska, M. Sc.-RIED, Warsaw for thier continuous liason efforts.
XVlll
-------
The consultations with Dr. F. J. Humenik of North Carolina State University
and Dr. R. C. Loehr of Cornell University have been very valuable in shaping
the course of this work.
The authors are grateful to Dr. H. Manczak, Professor and Director of
Wroclaw Division of RIED and to Mr. Lynn R. Shuyler for efficient supervision
and help during the project.
xix
-------
SECTION 1
CONCLUSIONS
The project is aimed at optimizing the presently used treatment systems
for dilute effluents from large piggeries, at optimization of loadings and
sequence of unit processes and operations for piggery waste treatment and at
placing the new waste treatment - recovery systems in the proper economic and
technological perspective. The project topics could be grouped as:
a) detailed technological and economic analysis of presently used treatment
systems (WTS) for dilute wastewaters for several large industrial pig farms;
b) an in-depth analysis of results of laboratory and pilot scale studies of
fourteen individual unit processes for treatment of pit wastes; and c) appli-
cation of obtained results to the design and economic assessment of 12 new
complete WTS for two types of large pig farms with partial effluent recycle,
3 3
i.e. water use 20 dm /hog/day and without recycle, i.e. 28 dm /hog/day.
It has been shown that the present highly sophisticated chemical-biologi-
cal WTS can be properly operated only in case of very careful cooperation
between the farm and the WTS personnel and in case of keeping the hydraulic
load within the design limits. It has also been shown that these systems are
vulnerable to influent variability and are highly inefficient and will cost
more with the increasing costs of power, imported chemicals and oil. The
recommended agricultural utilization of wastewaters is not always the desirable
alternative due to even higher costs, lack of area and relatively low recovery
value of nitrogen and phosphorus.
In the research part it has been shown that piggery WTS should include
a sequence of high-rate processes followed by low-rate low loading processes.
Chemical treatment should be replaced by plain sedimentation and/or anaerobic
pretreatment. The use of activated sludge should be limited in the high-rate
processes class and excluded from the low-rate processes class. The oxidation
-------
ponds as a polishing WTS should be used in combination with fish cultivation
as a method of biomass harvesting.
The comparison of the various modes of anaerobic digestion proves that
dilute piggery waste should be treated in high-solids retention time (SRT),
low-hydraulic retnetion time (HRT) systems, and that gas production and sludge
build-up decrease with increasing SRT, while the removal ratio and process
stability increase.' The recovery of gas is economically efficient already at
the present power costs, because the anaerobic processes proposed in this project
yield at the same time large removals of organics.
The recovery of single cell protein (SCP) through aerobic fermentation
is found feasible, however, the present costs of protein and N, P nutrients
make it an uneconomical venture from the standpoint of both the SCP production
and waste treatment. Large carbon supplementation is required in order to
fully utilize the nutrients contained in piggery wastes.
In all cases the alternative of combined treatment with other, nutrient
lacking effluents should be investigated because the benefits are usually much
higher than the cost of long-distance pressure transport systems.
It follows that the system of long detention time lagoons: anaerobic -
aerated - oxidation, is a viable WTS, easy to implement and operate in rural
conditions. The most efficient WTS, however, includes mesophilic anaerobic
biofiltration (ANBIOF) followed by aerobic biofiltration and polishing
anaerobic biofiltration. The system utilizes most of the methanogenic
potential of wastes by means of a separate anaerobic sludge digestion in an
ANCONT or ANFLOW type reactors, i.e. suspended growth reactors with and
without sludge recycle, respectively.
The ANBIOF type reactor should always be incorporated in any treatment
system as a first or second stage anaerobic digestion since it significantly
improves process stability and allows (as a second stage after an ANCONT
reactor) for an increase in organic loading without impairing the gas produc-
tion or organics removal efficiency.
-------
The major error in disposal of dilute piggery wastes in Europe, up to
recent times, was the application of either conventional wastewater treatment
technology or use of manure utilization systems applicable to concentrated
wastes. Based on numerous examples of difficulties in disposal of piggery
effluents, large industrial farm complexes are thought now to be economically
and technologically inefficient. This project has shown that this is not
necessarily so, at least from the standpoint of WTS. The accumulation of
large organic loadings and relatively low volumes of wastes may be regarded
as an advantage in novel, energy efficient, highly reliable gas and nutrient
recovery WTS, such as demonstrated in this report.
From the standpoint of the economics of wastes treatment, piggeries with
capacity smaller than 5000 head should rely on land disposal. If stream dis-
posal is the final goal, the size of the farm should not have to be limited,
since the proposed technology of anaerobic wastewater treatment and sludge
disposal improves its efficiency with the increase of the farm size. Another
factor improving the efficiency of the proposed systems is the decrease of
fresh water use through recycle of treated wastewaters.
-------
SECTION 2
RECOMMENDATIONS
Existing piggeries should begin the program of changes in the water dis-
tribution and sewerage systems to cut down the water use, increase the
temperature and concentration of raw wastes and apply recycle with purified
wastewaters for flushing purposes.
Full scale implementation of the proposed anaerobic treatment systems V and
VI is needed; the design work is presently being completed. Full scale
polishing oxidation ponds are being constructed. A three year period of
studies should encompass various techniques of biomass growth enhancement in
cold conditions such as greenhouses, mixed populations, recycle, etc. Fish
cultivation should be researched as a method of biomass harvesting and
biological sludge disposal.
New anaerobic treatment processes should be studied such as phase separa-
tion and selective organics removal systems to further cut down on the volumes
of anaerobic fermenters. The major trend in animal waste treatment technology
should be the further optimization of gas recovery and utilization of waste
heat, as the rising power costs will rapidly increase the applicability of
anaerobic digestion to concentrated organic effluents.
The work on yeast production should be continued with other types of
yeasts that require less of the readily available carbon. Studies on continuous
cultures, mixed yeasts populations need to be continued since further rises in
protein prices should yield the process more economical.
Methods of direct refeeding should not be pursued as much as the methods
of conversion into high protein feedstuffs. Further studies should be con-
ducted in open rather than in close cycles, i.e. feeding other kinds of
animals.
-------
Wide technology transfer and agricultural extension programs are needed
in order to show the growers, animal husbandry specialists and the agricul-
tural industry as a whole that animal wastewaters can be treated efficiently
at any level of dilution, at any volume and in any location. However, new,
more sophisticated technology is required, with significant recoveries
immediately available if proper liason between the producer and the sanitary
engineer is established.
-------
SECTION 3
INTRODUCTION
PROJECT AIM
The shortage of litter and the growing demand for animal protein have
caused several countries to turn to industrial scale animal production hus-
bandry. Large hog production plants in central Europe house usually from 10
to 40 thousand animals. In several instances larger farms were built, notably
in Romania and USSR where the size may approach 250 thousand hogs. Taking
into account the fact that a 36.5 thousand hog farm will require annually
Q f.
over 1.9 x 10 watthours of power, up to 18 x 10 kg of fodder and close to
5 3
3.2 x 10 m of water, one realizes the magnitude of operational problems ex-
perienced with the fulfillment of environmental requirements regarding air,
soil and water.
Figure 1-A illustrates the hog production trends in two countries:
Holland and Poland (1, 2, 6, 7). Figure 1-B shows the increasing participa-
tion of industrial sector in hog and cattle production in Poland (3). It
should be noted that industrial scale farms are responsible for much larger
segment of overall pig production in such countries as East Germany-29
percent, Hungary-47 percent, Romania-60 percent, and Bulgaria-65 percent (4).
Since, due to technical constraints, new farms are frequently sited in an area
unfit for land disposal, inevitable stream discharge of effluents requires the
highest practicable treatment technology to be applied.
At present, the industrialized hog production in Poland is limited to a
little over five percent, and the country is still in the process of testing
the various production technologies and wastewater/manure/disposal systems.
The present project will be confined to large piggeries as they create
a much larger environmental impact than the cattle farms due to more offensive
6
-------
21
20
«r 19
-18
o
o 16
15
14
13
12
22
20
oo 18
a
2516
1 14
Q12
=J10
8
6
4
2
• - POLAND
O- HOLLAND
1960
1970 1977
YEARS
CD
O
T—
X
CO
12
10
8
6
4
2
u.
o
1970 1975 1980
1985 1990 1995 2000
YEARS
Figure 1. (A) Overall hog production in two countries; (B) Planned production
increase of the industrial farms in Poland.
-------
odors, larger waste volumes and difficulties encountered during conventional
treatment and pretreatment before land disposal. The large piggeries were
purchased by Poland for trial purposes as the whole package with adjoining
waste treatment equipment. The practice has disproved many of the systems,
however, no formal assessment of the treatability of effluents from large
piggeries has been made. The project should answer the demand for: adequate
characterization of the raw effluents, knowledge of treatment efficiencies
attained by various unit processes and operations, alternative more economical
and more efficient treatment and pretreatment systems.
The overview of foreign practice in piggery wastewater treatment obtained
during visits to plants, in USA, Italy, Holland and Scotland as well as the
thorough literature perusal have documented the lack of data on treatment of
dilute wastewaters. The trends in these countries, although a rapid increase
in the overall pork production and the average farm size is evident, is to
keep the farms small and manure as concentrated as possible through recycle
and decreased water consumption. Thus, available information concerns con-
centrated effluents for which the process economics are different. The
present project will describe and charaterize the full scale practice of
treating the dilute pig wastes and evaluate the technological, economical
and environmental applicability of the novel alternative wastewater treatment
processes featuring biogas production and single cell protein recovery.
PROJECT SCORE
Based on mail surveys, literature data, field trips and on-site long-
term round-the-clock surveys, a summary of the present hog production and
wastewater treatment trends will be presented. Treatment effects will be
given and economic efficiency of various practiced unit processes will be
critically evaluated. Both practiced and promising future polishing treatment
process will be discussed. In-depth feasibility studies will be presented,
leading to the optimization of the operation of presently used systems treating
effluents for stream disposal and the novel systems proposed for the future
use. Finally, a set of proposals will be given, as to the required treatment
for stream disposal and for agricultural utilization or combined treatment
8
-------
with other effluents. The proposals will be based on comparison of costs,
treatment effects and non-economic factors which are beginning to play an
important role in the agricultural industry, e.g. the shortage of qualified
manpower, sight and odor nuisance, lack of adequate land for agricultural
disposal and other. The desirable development trends will conclude the report.
-------
SECTION 4
PRESENT PIGGERY WASTE TREATMENT PRACTICE
HOG PRODUCTION AND EFFLUENT TREATMENT TRENDS
There are close to twenty industrial hog production technologies used.
The most popular include: the Agrokomplex which is Hungarian, Emona which is
Yugoslavian, Gi-Gi which is Italian, Bisprol which is Polish, Achmidt-Ankum
which is West German, and Poznan which is Polish. The basics of production
are similar; the major differences are confined to: the methods of insemina-
tion, size and number of reproducing sows, the cycle of young pigs and fattened
hogs production, the age of wheaning, final product weight, degree of mechani-
zation and various technologies of wastes removal from the buildings.
The typical Agrokomplex technology consists of the following animals (5):
sows in various physiological stage - 870; young pigs up to 28 days old - 1,450;
young pigs 28 to 65 days old - 1,900; breeding sows - 147; breeding boars - 29;
fattening pigs - 6,200; which totals 10,596 animals. Since the usual loss
during the first 65 days is ten percent and only four percent during fattening,
animal production of the farm is (870 x 9 x 2.3) x (1 - 0.14) = 15,487 pigs/
year, at 105 kg/pig, assume 2.3 litters/year/sow and 9 pigs/litter, at 14
percent death rate.
The farm depicted in Figure 2 consists of seven multifunctional buildings
for sows with young and fattening pigs with 960 head each without natural
lighting, one building with weak lighting for pregnant sows and one well
lit building for loose sows, replacement sows and boars. Several other designs
are in use, notably radial arrangement of the major feeding buildings or two
story buildings fit for battery production, i.e. two- or three- story cages
housing young and fattening pigs.
10
-------
- SOWS. YOUNG & FATT. HOGS
- LOOSE SOWS. BOARS
- PREGNANT SOWS.
- ADMIN. BLDG.
- ENTRANCE
- WTP
SEWER- 100x600
Figure 2. Layout and cross-section of an Agrokomplex Farm A with 10.5 thousand head.
-------
There are two definite wastewater disposal trends which depend on the
concentration of effluents. For large farms with dilute effluents, the trend
is to treat the wastes for stream disposal as it is frequently impossible to
utiliize agriculturally the quantity of produced effluents on year-round basis.
For smaller farms various mechanical methods of manure removal are introduced
to limit the water content and different methods of odor removal are practiced
here are the distance for tank transport and quantity and availability of
diluting water that at times has to be applied.
WASTEWATER QUALITY AND QUANTITY
The evaluation of water and wastewater management systems is based on the
in-depth survey of three plants (7) and on a mail survey of other treatment
plants and regional water authorities.
Wastewater Quantity
Plants A and D have been studied in depth, serving respectively 10.5
and 18 thousand head farms - both are of the Agrokomplex technology. The
designed daily water use at these plants is (5): production zone - 145.0
33 3
m /d, sanitary zone 2.2 m /d, administration - 9.1 m /d, waste treatment
3 3
plant - 15.0 m /d which totals 171.3 m /d for 10.5 thousand head. The
2
quantity of wastewaters expected by designers is (171.3 - 50.5) = 120.8 m /d,
3
since close to 50 m /d is incorporated in the product (5). The waste treat-
ment facilities are designed based on the overall expected water demand, i.e.
3
on the average 17 dm /hog/day.
The present day experience with large hog farms indicates that very few
wastewater treatment facilities are able to meet desired effluent quality
because of the significantly increased water use at the farm and operational
problems at the treatment plant. The water use greatly exceeds both the basic
3
hog requirements and the hygienic needs, which are estimated at 9 to 13 dm /d/
3
hog over 90 kg liveweight (LW) (10), and amount to 20 to 40 dm /d/hog. The
reason for such an abundant water use lies in the design of the houses,
in improper washing practices, and leakage increasing with the size of farm,
and also with the design of treatment facilities, e.g. the use of tap water
for filter backwash.
12
-------
Typical animal houses are depicted in Figure 2-B. The collecting channels
are flushed by manure impounded with sluice gates, water is fed by nipple
feeders and is used for hoseflushing the fattening pens. The amount of water
saved when careful management practices are strictly observed is illustrated
in Figure 3 which shows the distribution of effluent quantity after the
completion of the modification program launched by the authors at Plant A,
a modified 10,000 head Agrokomplex farm, with significantly increased breeding
department (over 900 sows). The changes included both the farm and treatment
plant modifications (9). The installation of pressure tanks and pressure
nozzles on hoses for manual cleaning was one major cut in the water used.
Other cuts included elimination of sewers infiltration and introduction of
additional piping for filter backwashing with clarified effluent. Although
water management systems differ among farms, the quantity of inflowing raw
wastes has the most profound effect on the efficiency of the whole treatment
plant in any hog farm. Inherent hourly, daily and seasonal variability in
wastewater flow and concentration, usually assumed as (8): daily variability
Nd = 9««fd/Q«ve,d = 1'7 md hourly variability ^ = 24 Q^^/Q^^ - 1.8
were found much higher in this study and equalled N, - 2, N, = 3 to 5. When
this is superimposed on the excess base water use, it significantly adds to
an increase of instability factors in all unit treatment processes and large
fluctuations in the resulting final effluent quality.
In actual practice the unit and overall water use is much larger than
2
expected by designers. Plant D has had 29 dm /d/hog water use after significant
cuts were introduced (11), other farms were found to have even higher unit
3
water requirements, at extremeness, the values of up to 45 to 53 dm /d/hog were
quoted (10).
Wastewater Quality
Somewhat irregular cleaning and feeding procedures result in daily, weekly
and even in seasonal variations in wastewater quality and quantity. Figure 3
illustrates the BOD_ and COD monthly variability in raw wastewaters in the
half-year of studies at plant A. Daily and hourly variability is illustrated
elsewhere (12).
13
-------
-JS
•o
•
o
TD
to
13
UJ
tr
I
u_
400 -
300 -
200 -
'5 10 20 30 50 70 90 98
PROBABILITY OF OCCUR ^. I0/
15000
to
cr
10000
LU
t—
I/)
I
I
o:
** 5000
T3
LU
Q
LU
LJJ
o:
500
0
- 8000
- 7000
- 6000
- 5000
- AOOO
- 3000
2000
1000
o>iQO_o__ son
200
Q
O
CD
0
jo\
ii
MONTH
(1978)
B
IV V VI VII
Figure 3. (A) Water use at Farm D and Farm A before modification in 1979-curve 1, and after
modification - curve 2; (B) Variability of wastewater concentration at Farm D.
-------
Averaging the data collected by numerous authors and compiled by Loehf
(13): 0.12 kg BOD5/d/hog; 0.35 kg COD/d/hog; 0.121 kg SS/d/hog; 0.287 kg TS/d
and 0.254 kg VTS/d; and compiling results of Dragun's work (4) and authors'
own research, one obtains the basic unit raw waste load (RWL) from one hog
assumed to have a Iw of 50 kg. Dividing these values by average (50 percent)
wastewater volume output, which for plant D is 28 1/hog/d, one obtains the con-
centrations that compare fairly well with the data actually measured (Table 1).
TABLE 1. THEORETICAL ESTIMATION OF POLLUTANT CONCENTRATIONS IN
EFFLUENTS FROM LARGE PIGGERIES AND COMPARISON WITH
ACTUAL AVERAGES FOR PLANT D
Pollutant
BOD-
COD
Total Solids
Volatile Solids
Total SS
RWL g/d/hog
Loehr
124
352
287
254
121
Dragun
121
363
-
-
-
Authors
136
400
-
-
-
Loehr
4,430
12,570
10,250
9,070
4,320
3
Concentration mg/dm
Dragun
4,320
12,960
-
-
-
Authors
4,860
14,300
-
-
-
Actual
5,000
15,000
12,300
-
7,700
Note: Concentration based on q(50 percent) = 28 dm /d/hog.
In order to give the designer a set of numbers for making preliminary
design estimates fourteen large farms were surveyed. The analysis was per-
formed by fitting the normal probability curve as in Figure 4-A and finding
the average (50 percent), the standard deviation and the risk design value of
95 percent as the maximum. The characteristic values are given in Table 2.
Attempts were made to differentiate between the water management practices
of the three sizes of plants that made up the 14 plants studied, i.e.: 10, 18,
and 24 thousand head. It was confirmed that there are no statistically
meaningful differences between the various sizes.
The raw wastewater quality results were analyzed as shown in Figure 4-B.
Such correlations allow for establishing a set of design equations which are
valid in a very wide range of concentrations. This mechanism of verification
of animal wastes concentrations data was found of utmost importance due to
analytical difficulties experienced by various reporting sources.
15
-------
10 20 30 40 50 60 70 80 90 95 98 99
PROBABILITY <(%)
6 8 10 12 14 16
TOTAL SOLIDS (x103mg/dm3)
Figure 4. Raw wastes characteristics: (A) Concentration probability based on pooled
data from 14 farms; (B) Reggression of COD _ on total solids.
-------
TABLE 2. AVERAGE CONCENTRATIONS OF POLLUTANTS 11}
WASTEWATERS FROM 14 LARGE FARMS
Pollutant
BOD. - non-filtered
COD - (nf)
Permanganate value (nf)
N-NH3
Phosphorus - P«0.
Potassium - K_0
Total suspended solids
+3
Aluminum Al
+2
Zinc Zn
+2
Manganese Mn
_ +2
Copper Cu
•n ++•
Iron Fe
Chromium Cr (total)
Lead Pb+2
Mean (50%)
6,000
13,500
3,100
600
750
550
6,400
8.0
4.0
0.9
0.15
1.0
0.05
0.5
Max (95%)
11,800
28,300
-
1,700
1,700
920
-
30.0
9.0
3.0
2.2
-
-
_
Std. Deviation
3,000
8,500
-
550
600
240
-
12.0
3.2
1.3
1.20
-
-
The need for adequate screening of literature data has also been empha-
sized by other writers (14). The relationships for raw pig wastes attained
here are:
COD = 2.4 (BODC .) + 2000
nf 5,nf
valid for BOD- f >^ 1000 mg 02/dm , and
COD - = 3.09 TOC + 100
nt
COD , = 1.33 TS - 1200
nf
TS = 1.38 VS + 500
BOD. ,. = 1.66 BOD,. -
->, nt 5, f
(Farm A)
(Farm A)
(Farm A)
(Farm A)
(Farm A)
17
-------
Data for various farms yielded:
COD . = 2.4 BOD,. , + 1400
nf 5,nf
TREATMENT SYSTEMS USED; PROBLEMS ENCOUNTERED
There are several unit processes and waste treatment systems used before
stream disposal; while the technology of treatment prior to land disposal is
confined to lagooning with/or without aeration. The treatment systems
depicted in Figures 5 and 6 are the most common used.
The first system (Figure 5), Vidus (imported from Hungary), the most common
in this country, consists of fine dynamic screening (1 mm mesh), followed by
24 hr primary aeration and chemical coagulation and activated sludge (12 + 14
hr) aerated with surface aerators. This is followed by a submerged filtration
and chlorination. The excess sludge is fed into primary aeration tanks and
evacuated after alum coagulation with primary sludge to a separate gravity
thickener. Due to a large AL2 (SO,), coagulant dose (1000 mg/dm ), poorly
dewatering sludge is produced, the disposal still remains the major problem
although positive results are obtained with land disposal. The cost of the
treatment plant is high and amounts to 10 to 20 percent of the capital costs of
the farm, depending on size (15). The annual running costs amount are high
and run around 10 percent of the capital costs; however, the plants still
2
seldom achieve the expected effluent quality (BODC - = 50 mg 00/dm , COD .. =
- j, ni / nr
500 mg 02/dm ) due to operational difficulties encountered with this sophisti-
cated multistage treatment process.
The Vidus type plant has been modified within the realm of the project
3
in Farm A (9): the coagulant dose was optimized down to below 600 mg/dm ;
2
activated sludge content was increased to 5000 mg MLVSS/dm ; the anaerobic
3
sludge content was increased to 5000 mg MLVSS/dm ; the anaerobic denitrifying
biofilter was rebuilt and the sponge media was replaced with coke; and sludge
disposal problems were solved. The major disadvantage of the system is its
vulnerability to irregularities of flow and large quantities of chemical
sludge produced.
18
-------
I-
Ul
cr
en
LU
o
X
LU
FINAL
SETTLING
SLUDGE
DENITRIFICA-
TION FILTER
DISINFECTION
ACTIVATED
SLUDGE
/\
POPLAR
IRRIGATION
M
SLUDGE
HEAT
SETTLING
THICKENING
SLUDGE
pQNDS
RAPID MIX
EQULIZATION
(MIXED)
<3
SCREENS
L/H
WET WELL,
PUMPS
REFEEDING
Figure 5. Layout of Vidus type Plant A modified by authors.
19
-------
TYPE,,EMONA"
..WOSTOCHNYI"
HOLDING
LAGOON
0
AERATION
TANK 1°
CLARIFIER
O
AERATION
TANK 11°
CLARIFIER
Cl" CONTACT
TANK
RETENTION
TANK
VIBR. SCREENS
AERATION
TANK
CLARIFIER
RETENTION
TANK
O
VIBR. SCREENS
CLARIFIER
AERATION
TANK 1°
CLARIFIER
AERATION
TANK 11°
CLARIFIER
«
LAGOON
Figure 6. Presently used treatment systems for dilute piggery wastes.
20
-------
The overall sludge volume may amount up to 50 percent of wastewater
volume in heavily overloaded plants. The sludge is difficult to dewater;
the initial work on agricultural applicability has proved promising.
The Emona system, imported from Yugoslavia (Figure 6), features 5 months
lagooning followed by 18 hrs activated sludge treatment as first stage and 8
hrs aeration as second stage. The initial costs are low, approximately 3
percent of the farm cost, however, the system exhibits problems due to high
influent organics concentration after primary treatment due to sludge anaerobic
release and low temperatures in the activated sludge following the lagoon.
The Gi-Gi system (Figure 6), licensed from Italy, features primary
dynamic screening followed by aeration for 48 hours, at approximately 0.5
kg BOD,./kg MLSS,.d and final clarification. The system's capital costs are
some 8.5 percent of the farms cost, and the operation costs are 13 percent of
the plants capital costs (15). The major problems are hydraulic overloading
and the nature of excess sludge which results in very rapid clogging of drying
beds. The system has proved to yield inadequate effluent quality.
This brief comparison of technologies and costs (based on 35 thousand
annual pig production and the overall costs of the farm equal to 250 million
zloties, where 1 U.S. dollar = 20 zloties, zl) proves that designers have made
several errors, such as: placing the low rate processes before high rate
processes; or the general selection of wastewater treatment processes without
taking into account the resulting sludge problems. The performance failures
of several systems are related to the problems most common in this field, as
also noted recently by Evans (16), i.e. unsatisfactory operation and mainten-
ance and the type of design requires highly skilled personnel seldom
available for small flow-rate plants, in rural areas.
These rather complicated aerobic technologies are used in other countries
as well, often with acceptable results, however, at high operational costs.
The Wostochnyi plant (Figure 6) used in USSR, (18) features two-stage
activated sludge followed by lagooning (facultative) and reportedly yields
good removals in spite of heavy load of disinfecting and bactericidal agents.
21
-------
Activated sludge is also used in East Germany, as in the 12 thousand
head farm in Halle-Nord where 8 + 10 d aeration followed by 8 d lagooning
yields influent/effluent concentrations ratio of BOD5 - 13,800/180 and COD -
33,000/1500 at a cost of 0.9 mln DM and 0-M costs of 15.51 DM/ton LWK, where
energy use makes 48 percent of the running costs (17). One DM = 0.45 U.S.
dollars.
In the subsequent chapters the discussion will be confined to the results
of unit operations on wastes from Agrokomplex farms and the Vidus systems
since they have been used most often in this country.
EFFICIENCY OF TREATMENT PLANTS PRESENTLY IN USE
The efficiency of the various treatment system steps, at the plants
studied, varied depending on the irregularities of flow-rate biological process
loading and sludge content, coagulant dose, temperature, and to the major
extent on the level of skilled operation.
The efficiency of mechanical screening at Plants A and C is presented in
Figure 7. It follows from this plot that average COD removal is only 10 to 18
percent, while suspended solids (SS) removal amount to 22 percent. Comparing
these results with data for other plants, it follows that COD ,. removals vary
from 10 to 15 percent.
Full scale coagulation performed routinely in all plants of this type
yields removals of total COD ranging from 40 to 85 percent, with bulk of data
around 65 percent. Plant A data analysis revealed no correlation between the
dose and effect, due perhaps to the changing pattern of solids discharge from
the farm.
Biological Treatment
At Plant A two completely mixed (surface) aeration tanks are employed
o
treating primary effluent of average incoming BOD strength of 580 mg 0 /dm .
At present the concentrations have increased several times, however, the
22
-------
N>
A
^—.
60
o 40
o:
cc
o
Q
o
o
20
10 30 50 70 90 98
PROBABILITY OF OCCUR ^-1
200
CO
150
100
Q 50
o
CD
0
0.10
0,20
0.30 0,40 0,50
or nf -F/Mlg/g-d)
100
90
cc
o
Q
O
CD
80
70.
0,10 0,20 Q30 0,40 0.50 0,60 0.70 Q80
BOD5,f or nf -F/M(g/g-d )
Figure 7. (A) Efficiency of screening at Plant A and sedimentation at Plant C; (B) BOD5 removal
efficiency at Plant A as affected by sludge loading - F/M.
-------
retention time increased correspondingly and the mixed liquor suspended solids
(MLSS) were also increased. The design activated sludge parameters versus
the actual working regime during the time of this study (1978) were presented
in reference (11) - data before modification. The present parameters are X =
3 3
5000 mg/dm (values up to 7000 mg/dm MLSS were applied successfully), t =
3
24 hrs, effluent from chemical clarifier S = 2 + 4 g 02/dm - COD, food to
mircroorganisms ratio F/M = 0.5 to 0.7 kg 0-/kg/d.
The removal efficiency in the activated sludge system depends on many
factors, the most important being F/M or sludge loading, temperature and
sludge age, i.e. the sludge recycle practices. There are also indirect
factors that influence the biological system performance, such as total dis-
solved solids (TSS), sludge volume index (SVI), zone settling velocity, etc.
A rather well defined relationship between the sludge loading and BOD-
removal ratio is presented in Figure 7-B. The F/M ratio had a pronounced
effect on the effluent suspended solids carryover from the settling tank
overflow, before modification. The correlation of the effluent BOD_ versus
BOD-F/M for both soluble and the total values illustrates this best. The
significance of adequate final clarification is evident from this graph.
Soluble effluent BOD,, stays relatively unchanged over a large range of F/M
variations, while total BOD_ values increase rapidly with the increasing
loading. This should be well substantiated by sludge volume index changes.
The correlations of SVI versus F/M for both BOD,, and COD, were, however,
very weak, as described elsewhere (11). Since most regulations on effluent
concentrations are based on the total BOD,., maintaining the appropriate
SVI is of paramount importance, thus the need for polishing treatment.
It should be noted that data interpreted here comes from plants in normal
operating regime, although the authors have prepared instruction manuals for
the maintenance crew at Plant A. Thus, the operating parameters varied. This
is fully described by authors in other papers (11).
The equation used for calculation of the kinetic K constant is the sub-
strate kinetic model of Grau and Eckenfelder:
24
-------
s - s s
o e „ e
X . t " * S (1)
a o
The kinetic rate constants for Plant A and other plants expressed in COD units
are presented in Figure 8, where curve 1 is for soluble COD, data.
The operation of the process at varying temperatures allows for estimation
of the temperature correction factor for the rate constant, according to the
Arrhenius equation:
KT - Kzo ' 0 T " 2° (2)
The plot of log K- vs A T for BOD- data from Plant A prior to changes
-1
yields the value of 0 = 1.053 and the average K-. = 6.6 d (see Figure 8).
Judging by the spread of data, the value of 9 needs further verification, it
is, however, valid for estimation of dilute wastewaters from Plant A.
The performance of activated sludge after modification is presented in
Figure 9. Individual unit process efficiencies are presented in Figure 10.
Polishing Treatment
The unstable at times, operating conditions result in significant deteri-
oration of biological effluent quality due to solids carryover which may be
correlated as BODC - versus SS:
5,nf
BOD,. = BOD. -I- a . SS (3)
-j,nt o,r
where "nf" and "f" represent non-filtered and soluble values of BOD,., "a" is
the slope of the curve, and SS the total suspended solids. In the case of
Plant A:
BOD. = 30 + 0.20 TSS (4)
j,nr
In order to alleviate this problem in Plant A an existing anaerobic
flooded biofilter has been adopted as a filtration and denitrification filter.
The filter, filled with coke and an underlayer of gravel, yields good removals
25
-------
3000
S3
,- 1U
I g
5 8
7
§ 6
* 5
£ A
o
Z
a: 3
in
Q
o
m 2
1
(
^-«,
i
^"^
(
©
•
(
•»-*.
i
4
t (
•*»-.
i
> (
<
i
<
•
> ,
0=1
<
»
(
<
*^>
053
>
i
}
>
"*»»«.
i
^
4
4
>
i*»^
<
r*^<
•^^
<
) 1 2 3 4 5 6 7 8 9 10 11 12
500
1000 1500 2000 2500
EFFLUENT COD. Se ( mg02/dm3)
AT(°C)
Figure 8. (A) Activated sludge performance at three Vidus plants; (B) Temperature correction
calculation for Plant A data.
-------
1 IUU'
°E 1000'
•a
^900-
0>
E orin.
u. 80°
o 700-
S 600-
5 50°-
3 400-
u.
£ 300-
200-
100-
n-
(B)
^
. /-
g^
/
«/•
•' 0 0 JD'
J"1-*
^-°
owy O o
^0
X BEFORE
MODIFICATION
/.
/ AFTER
i MUDIHCATIUN
, \
o\<
1r"'
O^i^
& Q
., ,
\
I
o
CO
ro
o
a>
a
o
o
i-
co
18000
16000
14000
12000
10000
8000
2,0
CODNF-F/M(g02/gMLSS-d:
500
450 ~
400 2
350 %
300 \
250 §
200 z
Ul
150 3
LL
100 £
50
0
10 20 30 40 50 60 70 80 90 95 98
PROBABILITY OF OCCURRENCE ^ (%)
Figure 9 (A) Distribution of raw and effluent wastes COD after
modification; (B) Influence of CODnf sludge load on
effluent quality; - Plant A data.
27
-------
NJ
00
1500
°E 1000
"O
~7sj
o
o>
E
~jl 500
in
Q
o
CO
Q
~ PLANT A
I (1978)
— }
— 9 S^
• JQ
JO
- °/ A 4
I A*"*
— A minf
I 1- |A"T| 1 H
/•' 0/2 J
\o^ :
/°
j
^^A"^ -
^A
^
^WSA "4 ~
iii ii i~
ro
•o
IN
O
o>
_E
Q
O
O
Q
o:
LU
ti
u-
~z.
§
1 2 5 10 50 90959899
PROBABILITY OF OCCURRENCE ^ %
1- RAW BOD 5
2- SCREENED
3- AFTER CHEMICAL
A- AFTER ACTIVATED
PRECIPITATION
SLUDGE
5- RAW WAST WATERS
6- CHMICAL PRECIPITATION EFFLUENT
7 -EFFLUENT FROM
8 -FINAL EFFLUENT
ACTIVATED SLUDGE TANKS
FROM DRAINAGE FIELDS
ILJxIU
5x10A
2x104
10x103
8
6
4
2x103
10x102
8
6
Ax102
3
2
1x10
: :
r PLANT D :
7 (1978)
**-<*
^r*
^•"•^
^» 6
•^ ^
^^ .• jO^ "~~
- ^"^' ^ffs*l -
' O^A^ •
J3^ %'
o^^^ ^X^
o^ A 8
XA^
A^***** -
C ^^-**zf*
L A^ -
"
• -
1 1 I 1 1 I 1 i II
10 20 30 40 50 60 70 80 90 95 9i
PROBABILITY OF OCCURRENCE ^
Figure 10. Efficiency of unit treatment processes at two full scale plants.
-------
at varying influent loads and concentrations. The kinetic interpretation of
filter removal data according to the authors' model (20):
Se/SQ = exp (-K/L) (5)
3 3
where K (kg/m /d) is the gross removal rate coefficient, L (kg/m /d) is the
volumetric organic loading, yielded K_ = 2.9 and K...... = 1.1 kg/m /d.
The anaerobic biofilter at Plant A was responsive to influent substrate
concentration and load variations. Thus, maintaining good equalization is of
paramount importance at the influent to the biological part of the treatment
train. More detailed analysis of the anaerobic biofiltration polishing is
presented in Section 7.
ECONOMIC EFFICIENCY OF TREATMENT
The research on five of the large imported wastewater treatment plants
has shown high O&M costs which are due to large chemical use and use of oil for
heating wastewaters and high power consumption for pumping to subsequent
stages, as well as for aeration. The capital costs of the imported treatment
plants are also high and may run up to 20 percent of the overall farm costs,
as shown in Figure 11 (22, 23).
The analysis of efficiency of the existing treatment plants was performed
according to procedures binding designers of the so-called non-productive
structures in Poland (24) . The basic formula for the economic efficiency
index E:
J (r + > + K
where W - magnitude of the utility effect;
J - capital investment (zloties); equal to J = 1(1 +0.5 b.r) for b > 1
year, the value of b = 2 was used;
b - time needed for construction (years);
s - average amortization rate;
r - discount rate; r = 0.08 was used;
29
-------
CO
E
CO
o
N
O.
LU
400
300
200
N
CO
S?
, y\j
£ 80
70
S 60
£L
<
40
30
20
10
FARM*WTP) COST
I I I |
% OF TOTAL (COST
.13,919,622,014,1 13,5 13,6 16,1
30
Q
O
co 20
o»
Q
10
100
80
-60
40
20
15
10
Q
O
Q
o
200 400 600 800 1000
FLOW RATE p(m3/d)
Figure 11. (A) Capital costs of piggery farms and waste treatment plants
efficiency of treatment as affected by WTP size.
- WTP; (B) Economic
-------
K - annual running (O&M) costs decreased by amortization
I - raw investment costs (zloties).
The economic efficiency index is expressed against either volume of
wastewater treated in a year (E_-zl/m ) or in terms of weight of pollutant
removed in a year (E^-.-zl/kg BODC removed) , by dividing the cost by respective
D(JU j
utility effect, i.e. m /year of kg BOD,, removed/year. The evaluation of data
collected at five farms is presented in Table 3. Other data, reported by
3
Heidrich et al. (22) revealed the values of EQ varying from 13.7 to 67.5 zl/m
and E = 2.6 to 10.6 zl/kg BOD for the Vidus plants while the values for the
"Gdansk" treatment system were EO = 56 zl/m and ERO_ =9.4 zl/kg BOD^ - removed.
Graphical analysis of the economic efficiency data in Figures 11 and 12
shows that there is a definite decrease of the economic efficiency indices
with increasing throughput and increasing load removed. Data is significantly
scattered because of price changes and various periods in which plants were
erected. The indices are much higher, particularly E_, than the ones for
municipal sewage due to higher concentrations of organics, but also due to
overdesign of the equipment.
Taking into account the high costs of construction and O&M as well as
non-economical factors such as the difficulty in operation due to employment
of highly sophisticated technology and the fact that significant quantity of
nutrients are wasted, these waste treatment plants are considered to be a
temporary solution for the industry. The present trend is to find alternative
cost-effective and energy-wise solutions if the wastewaters are to be disposed
to stream. In all other cases smaller farms are to be built and sited so as
to enable agricultural utilization of wastewaters.
DISCUSSION AND CONCLUSIONS
Most of the farms seem to have expected smaller water usage, and thus, the
wastewater treatment plants are usually hydraulically overloaded. Standards
3
binding design engineers give the water use figures of 15 to 25 dm /hog/d
31
-------
TABLE 3. ECONOMIC EFFICIENCY OF FIVE STUDIED
V3DUS-TYPE TREATMENT PLANTS
No. of head
Flow Rate -Q
S -BODC ,.
o 5,nf
Sg -BOD5
S -COD '
o nf
S -COD .
e nf
Removal BOD^
Removal COD
I
K
EQ
E (BOD-removed)
E (COD-removed)
Unit
103
m3/d
g 02/dm3
g 02/dm3
g 02/dm3
g 02/dm3
%
%
106 zl
106 zl
zl/m
zl/kg BOD5
zl/kg COD
Farm A
10.5
250
4.0
0.08
13.2
1.72
98.0
87.0
20.23
2.03
46.7
11.9
4.1
Farm D
21.0
650
3.8
0.24
12.7 ,
0.77
93.7
94.0
54.70
4.96
46.2
13.0
3.8
Farm F
25.5
550
6.0
0.20
12.6
0.66
96.6
94.7
46.38
1.83
34.5 *
5.9
2.9
Farm G
28.0
900
6.4
0.26
12.1
1.82
96.0
85.0
45.82
3.84
27.05
4.4
2.6
Farm H
11.3
180
5.4
0.18
12.6
0.42
96.6
96.6
25.16
2.45
79.5
15.2
6.5
NOTE: 1 U.S. dollar is equivalent to 20 zloties (zl).
32
-------
40T
30+
- 20
>•
o
e
-------
3
while real-life practice yields numbers as high as 40 dm /hog/d; even higher
averages are quoted. Efforts will now have to be made to decrease this water
use by modifying the hydraulic transport system towards high pressure cleaning
and wastewater recycle for flushing purposes, as demonstrated by authors in
this study.
The mechanical (dynamic) screening apparently yielding low solid yields
(averaging 11 to 15 percent COD removal and some 20 percent TS) produces sludge
of very good dewatering characteristics, and in all cases screens should be
included in the wastewater treatment train. This is in accordance with other
research findings on solids effects on the treatment efficiency (19).
Coagulation seems to be the subsequent process that could be left in
existing plants since it dampens the load variability, however, the dose should
be optimized down, as will be demonstrated later.
Activated sludge is a large volume 25 to 48 hr retention process that
requires very skillful maintenance due to the nature of the biota (bacterial
sludge) with a tendency for bulking, solids carry-over, etc. Our studies
have shown that maintaining good settleability of the sludge is crucial to
the process overall efficiency. Contrary to other writers (25) who considered
it difficult to get down to very low BOD concentrations in the effluent, these
studies show good biodegradability of piggery effluents and rather low concen-
tration of refractory organics. The BOD removal rate coefficient at Plant A
was found to be approximately equal to 6.6 d . The process itself is quite
temperature sensitive. The value of temperature factor, equal to 1.053 is
higher than that usually assumed for municipal wastes.
The Plant A overall removal efficiency is presented in Figure 9 based on
1980 data. The individal unit process efficiencies (in Figure 10), 1978 data,
reveal fairly predictable values, noting that these are averages of the real
operating conditions without exclusion of upsets.
It follows that activated sludge is yielding poorer removals in Plant D
than in Plant A. Analysis of Plant D biological treatment performance
34
-------
revealed significant upsets of the biota, sludge index values above 600 to
900 (the "normal" value of SVI for pig wastewaters is 150 to 350) and resulting
very poor solids separation. Regardless of the reasons for such temporary
situations, it is apparent that in such cases the presence of chemical precipi-
tation, although costly, and additional polishing treatment steps buffer the
upset of one unit process in the whole treatment train. The soluble BOD-
3 3
values at Plant D have varied from 60 to 100 mg/dm (BOD^ = 75 to 300 mg/dm )
o j >nr
in activated sludge effluent and 25 to 40 mg/dm in the irrigation field
effluent (the final treatment step in Plant D).
The conclusions from this overview of the present treatment technology
are:
1. The variability of unit wastewater volume output is very significant
and apparently random in nature. It is difficult to maintain the unit water
use below 28 dm /hog/d, at farms above 10 thousand head.
2. The recommended variability coefficients underestimate the actual
conditions. The values estimated during this work were N, = 2, N. = 3 to 5.
3. The concentrations of manure should always be checked against the
unit pollutant load from one hog.
4. The combined high-rate chemical and biological treatment system, such
as studied here, is capable of producing high quality effluent in cases where
quantity is carefully kept at a constant level, within the hydraulic limits
of the treatment units. Inadequate equalization; activated sludge vulnerabil-
ity and bulking tendencies; solids carryover; temperature effects; hydraulic
overloading and first of all inadequate maintenance are the causes of poor
plant performance.
5. The analysis of the unit processes efficiencies revealed that the
high-rate treatment is uneconomical because of power use for pumping, aeration
and agitation and chemical costs.
6. Solids removal as primary treatment is essential.
7. Coagulation will have to be removed in favor of plain sedimentation
and longer aeration times, due to problems with sludge disposal and high O&M
costs.
8. Activated sludge is of bacterial type and quite sensitive to tem-
perature variations.
35
-------
9. Activated sludge solids carryover, i.e. poor solids capture, are
responsible for at least 20 percent of the non-filtered BOD,..
10. Polishing treatment is essential prior to the stream discharge of
piggery effluent. Anaerobic biofiltration yields over 50 percent of BOD,. .
removal.
11. The research needs promulgated by this work are:
a - methods of decreasing unit water demand;
b - introduction of low energy - low rate treatment units;
c - solution of chemical and biological sludge disposal problem
(present studies with irrigation of poplars need to be carried
for 5 more years);
d - economical recovery of by-products, SCP, gas, and treated
wastewater recycle for flushing; and
e - increasing concentration of effluents and methods of economic
land application as the final disposal of effluents.
36
-------
SECTION 5
SAMPLING AND ANALYTICAL METHODS
SAMPLING
Samples for analysis of efficiency of the Vidus type treatment plant
performance were collected immediately before and after the unit evaluated,
in one-hour intervals, through 72 consecutive hours. The samples were then
collected into containers; each contained the interval of one 8 hr shift; the
samples were added to containers in quantity proportional to the volume of
flow-according to the method described by Oleszkiewicz (174).
The wastes for studies at the departmental laboratory were collected from
Plant A, Agrokomplex type farm with 10.5 thousand head capacity: a) raw wastes
from the wet well and b) treated wastes for polishing studies, from the final
clarifier overflow.
ANALYTICAL METHODS
Organic content was measured by means of 5 day - Biochemical oxygen demand
both in filtered (BOD. ) and nonfiltered samples (BOD. f); by chemical oxygen
demand (CODf and COD f) and total and soluble organic carbon (TOG and SOC).
Due to lab space and time difficulties, a shorter, 10 min COD digestion
method was adopted (32, 98); method yielded comparable results constantly 5
percent lower than the two-hours-digestion-COD. The calibration is presented
in Figure 13. The same figure (B) gives a correlation between both COD 10
min and 2 hr against total solids (TS) as:
COD = 1 + 1.545 TS . . . (103 mg 02/dm3) (7)
used for qualifying outliers in the data pool. All other determinations were
37
-------
24 r—r—i—r
Co
oo
20
16
CO
T3
Q
O
O
12
r T f
O NON FILTERED
• FILTERED
COD (10') = 0,95 COD(2h)
O 10'CODNF
A 2h CODNF
• 10' COD F
A 2h COD c
COD = H1.545 TS
i I I I i
20
16
12
"
Q
O
12 16
2hCOD(g02/dm3)
2 4 6 8 10 12 14 16
TS(g/dm3)
Figure 13. (A) Comparison of 10 minutes and 2h COD digestion; (B) Calibration of COD against
total solids - TS.
-------
done according to EPA Methods (31). Additionally, permanganate value of
oxygen demand was analyzed according to Polish standards (32) to facilitate
an estimate of BOD dilutions.
The total and dissolved solids content were analyzed according to USA
ASTM Designation (33), while the determinations of suspended solids for both
total and volatile (TSS, VSS) were according to EPA Methods (31).
Nitrogen determinations N-NH,, N-org., TKN were made according to EPA
Methods (31). For the Kjeldahl nitrogen (TKN), a selenium catalyst was used.
Modifications according to Polish Standards (34, 35) had to be made for deter-
minations of N-NO_ and N-NO_ due to lack of catalists. The samples for
nitrates were prepared by filtration with activated carbon through a medium
filter paper and coagulation with zinc sulphate in alkaline conditions and
then filtered again. Then the sample was neutralized, urea was added (in
proportion to N-NO_ content), acidified with H-SO, and heated for 16 hours.
Somewhat different procedures was used for samples where the values of N0~
3
were much higher than NO-, i.e. NO- less than 0.1 mg/dm .
Other analyses were made in accordance with US EPA Methods or if the manual
did not elaborate on a particular analysis, Standard Methods were applied (36).
This applied to determination of phosphorus, chlorides, sulfides, sulphates and
heavy metals by atomic absorption spectrophotometry.
39
-------
SECTION 6
PRIMARY AND SECONDARY AEROBIC TREATMENT
SEDIMENTATION
The solids that are entrained in piggery wastes consist of unused fodder,
some bedding, urine, feces, large incidental objects, etc. These are usually
removed by screening. The openings of 1.0 to 1.5 mm were found to be the best
for the vibrating screens. The screening efficiencies were presented in Figure
7 and 10 based on references (26, 28). This section is aimed at evaluating the
sedimentation feasibility as a sole pretreatment to further biotreatment.
Optimum settling time will be selected and effects of primary aeration
(practiced now in some full scale plants) will be evaluated.
Methods
Settling tests were performed on site at Farm A. The wastewater was sub-
jected to the normal wet well mixing and pumping, and dynamic screening through
1.5 x 1.5 mm open screens. The tests were run in a series of 1 dm cylinders -
45 cm high. The samples were collected by pipette from a depth of 25 cm below
surface at selected time intervals: 1 min, 3 min, 5 min, 10 min, 0.5 hr, 1 hr,
2 hr, 3 hr, 24 hr, and 38 hr. One cylinder represented one sampling interval.
The efficiency was measured by COD, permangate values COD and BOD_ f analyses.
Results and Discussions
Most of the settling tests had evidenced very rapid initial solids removal
in the first few minutes followed by gradual settling of finer solids, which
was practically completed within 90 to 120 min. The difficulties in comparing
the relative rate of settling led to the development of the log-log plot of
effluent suspended solids, BOD and COD versus time, according to the equation:
St = Sj/t* (8)
40
-------
where S , S1 are respectively, effluent concentration after time - t, and
concentration after t = 1 min; t = settling time and a = the slope of the
settling curve representative of rate at which pollutant removal occurs. The
value of S.. is the intersect of the curve with the ordinate at t = 1 min, a
convenient time to record the initial or immediate removal.
Typical removals of TSS and organics are presented in Figure 14 which
shows effects of preaeration on the final TSS and BOD and COD concentrations.
The TSS removals, calculated as (1 - S /S ), indicate that the removal
is the most efficient in the first 30 minutes of the process and averages 75
percent. A one hour settling results usually in an average 78 percent solids
removal, while an additional hour yields an increase of only 3 percent up to
N = 81 percent. Further sedimentation, beyond the two-hour limit yields only
1 percent increase and at 3 hr, N = 82 percent curve 1 shown in Figure 15.
The removal of COD and BOD,., Figure 15, confirms the above conclusion.
The average organics removals curve forms platteau after two hours of settling.
Primary aeration was tried as a method of enhancing solids removal.
Figures 14 and 15 indicate the effects of 0.5 hr aeration and a prolonged 19
hr preaeration. It was found that in both cases, the primary aeration has
increased the solids removal by some 4 percent. A more pronounced effect was
evidenced in organics removals which were increased by over 10 percent. The
comparison indicates that primary aeration, of very short duration, should be
employed, e.g. as a method for mixing the wet-well contents.
Summary
Most of the imported piggery wastewater treatment plants apply screening
and chemical coagulation as primary treatment. This research has shown that
plain sedimentation, after fine screening 0.5 to 1 mm, which results in the
directly reusable by-product can be effectively substituted for expensive
alum coagulation. Figure 15-B illustrates statistical analysis of two-hour
settling tests conducted at other plants with annual production ranging from
14 to 35 thousand hogs. The tests yielded an average decrease of total COD
41
-------
10
° 8
2 ?
6
o^ A
co 3
CO
«" -
IS)
10
9
8
8 I
o 6
CO
E
^
o"
1 1 1 1 I I l I I
T T
Sx
i i
i i l l l l I l i
COD -
BOD
J L
2 3 A 5
10
.20 30 40 50 70 100 200 500
SETTLING TIME (mini
Figure 14. Effects of preaeration on removal of: (A) Suspended solids; (B) Organics; 1 - raw,
fresh waste and 2 - after 0.5h aeration on 24.IV.78; 3 - raw, fresh waste and 4 - after
19h preaeration on 25.V.79.
-------
2,0 2J5 30
SETTLING TIME(hours)
RAW SETTLED
PLANT A A A
PLANT B o
PLANT C Q
10
20 30 40 50 60 70 80 90 95 98
PROBABILITY OF OCCURRENCE
Figure 15. (A) Settling curves for Plant-A raw wastewaters: line 1 - S _= 12,500 mg TSS/dm , 2 -
aerated S = 12,700 mg TSS/dm , line 3 - S = 6,770 mg TSS/dm , 4 - aerated SQ - 6,230
mg/dm ; (B) COD f removal efficiency during settling of three plants wastewaters.
-------
3
down to about 4400 mg/dm , i.e. a removal of some 52 to 59 percent, which
compares well with the results of this study.
Full-scale studies with alum coagulation of very diluted wastes yielded
the following removals: COD - 62 percent; BOD5 nf - 39 percent; BOD5 f -
26 percent; and TSS - 82 percent. One has to take into account the fact that
diluted wastewaters yield lower removals (26), however, it is evident that
coagulation does not add that much in terms of additional removals. A study
at Farm D, 28 thousand head, yielded 65 percent COD removal in coagulation
of screened, more concentrated wastes.
The summary of findings is as follows:
1. In spite of attaining removals lower than with alum coagulation,
plain sedimentation appears to be the most economical method of
primary treatment, where sludge disposal is concerned, leading to
average removals expected: COD - - 55 percent; BOD_ . - 35 percent;
and TSS - 80 percent.
2. Primary advantage of sedimentation is the small volume of sludge
and its good dewaterability when compared with chemical sludge.
3. Preaeration should precede settling and could be effectively combined
with mixing of wet wells, presently turbine mixers are used that
result in particle shearing and decrease the settling efficiency,
and can yield up to a 5 percent improvement of solids removal.
4. Short, half-hour preaeration is sufficient to yield a 10 percent
improvement in organics removal.
5. The optimum settling time for piggery wastes, in quiescent, labora-
tory conditions, is two hours. Longer times yield only negligible
improvements in removal efficiencies.
6. A novel method of presenting settling data was employed which allows
for comparison of settling rates for various effluents based on rate
coefficient, "a," in the equation (S = S.,/ta).
COAGULATION OF PIGGERY WASTEWATERS
Screened piggery wastewaters contain unusually high amounts of non-
settling suspended solids and colloidal matter. Majority of piggery wastes
44
-------
treatment plants in Poland use alum coagulation as the primary treatment step,
prior to the subsequent biological processes. Due to the license recommenda-
tions and the lack of in-depth coagulant type and dose selection studies, it
3
is universally practiced to apply the alum dose of at least 1 kg/m . Operators
use larger doses when stronger raw wastes are received at the plant. Under
these circumstances, the cost of coagulant is a large component of the operating
costs (over 15 percent), while the volume of sludge, reported at some plants
over 30 percent of the daily wastewater flow results in the most formidable
ultimate disposal problem at the plant.
The practice of chemical sludge volume reduction and ultimate disposal
is in the early stages of development (30), particularly for the relatively
low volumes encountered at hog farms. Nevertheless, attempts have been made
to dewater the sludge by mechanical centrifuging. Favorable results were
obtained only at very high sludge conditioning polymer doses (29).
The present studies were undertaken to find the optimum alum coagulant
dose at varying wastewater quality and to evaluate the feasibility of using
other coagulants that are less expensive and that will produce lower volume
of sludge more amenable to air dyring or other simple dewatering means.
The second part of this work, dealing with comparison of two parallel
systems: coagulation followed by activated sludge and activated sludge
followed by coagulation was conducted in order to optimize coagulant cost
and the total volume of produced sludge.
Materials and Methods
Routine 1 dm jar test procedure was used. Rapid mixing lasted 3 min
at 80 rpm, and slow mixing lasted 20 min at 20 rpm. A two-hour settling
preceded supernatant sample collection and determination of sludge volume.
The efficiency of coagulation was determined as a function of dose D,
in terms of optimal dose D ; effects of various factors were also analyzed
such as the age of pig wastes and the use of Pollena JK disinfectant. The con-
clusions were drawn based on a relationship:
45
-------
(So - Se) = D.f(So) (9)
where S , S are initial and effluent wastes concentration; (COD and TSS were
used); f(S ) as a function of initial concentration.
o
Results
Full scale results at Plant A obtained at the varying dose of 1,000 to
1,300 mg/dm of H_(SO,)_ showed COD . removals to be 55 to 85 percent.
The laboratory jar tests were conducted at random since 1978 in order to get
enough data for statistical analysis. The raw wastes COD , concentrations varied
3
from 3,800 to 18,000 mg 0-/dm . Several correlations were analyzed. The basic
graph of S versus D was interpreted for an optimum dose D as in Figure 16-B.
The value of S at D , denoted S was compared against the values of S after
2 hr settling, with filtered sample concentration, and with the S at D = 1,000
3 e
mg/dm (Figure 17-A).
The analysis has been performed graphically and in tabular form. Graphical
analysis showed that there are two distinct segments of the dose-effects curve.
The first part, at doses below D ^ is linear and may be expressed as S /S =
_4 opt 3 e °
1 - a.D where a = (7 to 9) 10 and D is dose in mg/dm . The overall curve can
be expressed as an exponential curve S /S = D where a = 0.25 to 0.35. The
correlations were seldom satisfactory, and it was difficult to evaluate the
effects of increased disinfectant use or prolonged storage of manure in the
farm sewer system. The results indicate that coagulation is the most efficient
on fresh wastes, preaerated, and with the minimum amount of Pollena JK disin-
fectant used.
Figure 17 illustrates the effects of an averaged optimized dose of alum
coagulant (DQ =390 mg/dm of A12 (SO,)- .18 H20) compared with average (of 6
runs) effects of 2 hours and 4 hours sedimentation and the effect of coagulant
dose up to the assymptote (usually D=l,300 to 1,400 mg/dm ) and filtration.
3
Comparing the effects of practiced design alum dose of 1,000 mg/dm with effects
of D and the costs of sludge disposal which are over three times higher for
the design dose, it is evident that increased removal in the overall plant
efficiency analysis is offset by coagulant and sludge disposal costs.
46
-------
10
§ °.9
u
o
CO
co
0.8
07
0.6
0.5
04
03
0.0 0.2 0.4 0.6
0.8 10 1.2 1.4
ALUM DOSE (g/dm3)
0.9
c
Q
g OS
01
CO
to o.7
0.6
0.5
04
-i 1 ' '
A12(S04)3
J 1 L
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
DOSE-D(mg/cm3)
Figure 16. Effects of coagulation: (A) Waste stored for two days - 1978,
1979 data; (B) Fresh wastes - 1980 data.
47
-------
-p-
CD
AV2=83cm3/dm3
100 -
V SETTLING =110 cm3/dm3
02 0,4 0,6 0,8 1.0 1,2 1,4
ALUM DOSE D(g/dm3)
Figure 17. Comparison of coagulation effects with settling: (A) Efficiency of CODnf removal by
various processes; (B) Sludge volume at various alum doses.
-------
The analysis of the data gathered from 1978 to 1979 showed that:
3
The ratio of effluent COD concentration at alum dose 1,000 mg/dm to
concentration at an optimum dose - S varied:
opt
S (D=1000)/S ,. a 0.75 - 0.90 (10)
e opt
the ratio of S to the filtered COD (soluble)/:
opt
S /S- = 0.60 - 0.75; (11)
opt 1
the range of values of the removal by coagulation with an optimum dose over the
effects of two hours sedimentation:
*s
S (sedim. ) - S
(COD)
with a mean of 10 percent.
Conclusions
3
The presently used dose of alum, 1000 mg/dm , is too high from the stand-
point of technological and economic efficiency. The optimum dose of alum is
3
390 mg/dm in ideal laboratory conditions. The sludge volume at D is 0.03
m /m (3 percent) of wastewaters. Increase of the alum dose beyond D results
in only 15 percent increase of COD removal efficiency and over 300 percent
increase in the resulting sludge volume. The alternative; sedimentation, yields
removals of COD some 10 percent smaller than the optimum dose coagulation effects;
the sludge, however, has good dewatering characteristics.
The dose-response curves were analyzed by means of tangents to the respec-
tive ends of the curve in order to find D . Little correlation was found
opt
between the initial concentration S and the optimum dose and the coagulation
efficiency. The incidental use of disinfectants had detrimental effects on
organics removal.
In existing systems that use coagulation, alternative coagulants should
be researched. Lime proved equally effective in jar tests in approximately
50 percent of cases. At higher influent COD concentrations, the alum was by
far superior.
49
-------
ACTIVATED SLUDGE - SECONDARY TREATMENT
The activated sludge process has been used with various modifications
to treat raw pig wastes for over two decades. The most popular systems for
concentrated wastewaters are under-the—slatted—floor oxidation ditches (37)
or separated oxidation ditches (38), however, completely mixed reactors have
also been used (39). The data available on dilute effluent biotreatment is
ambiguous and lacks the full understanding of kinetics and reasons for the
bulking tendencies (25, 40). The present studies are aimed at testing the
reasons for better unit efficiencies and sludge yields experienced in activated
sludge treatment of primarily coagulated dilute wastes from Farm A. The tests
were conducted in laboratory units as in Figure 18.
Pretreatment by Coagulation
Two runs of batch tests were made. The run consisted of setting six
parallel activated sludge tanks to which enough activated sludge, acclimated
to chemically pretreated wastes, was added to attain X = 2.5 to 3.5 g/dm .
3 V
Equal 1 dm volumes of wastes were treated in a jar test apparatus with alum
3
dose of 0, 100, 200, 400, 800, and 1,200 mg/dm . The effluent was fed into the
activated sludge tanks and aerated for 24 hours. Samples were collected at
2, 4, 6, 8, 10, 12, and 24 hr in the first run and at 3, 6, 9, 12, and 24 hr
in the second run. The removal rates (K) for each individual unit were
calculated from:
Se/SQ = exp (-Kt) (13)
and expressed in (1/d). The results were plotted in K versus dose D to arrive
at an optimized coagulant dose. Figure 19 illustrates the selection of an
3
optimum dose as 400 mg/dm , the value coincidental with the results of the
coagulation optimization tests described in the previous chapter (the D
3
was equal to 390 mg/dm ). The second batch yielded different results,
indicating that beneficial effects of coagulation as primary treatment may
not be pronounced at all, at times. Due to the fact that it was not possible
to monitor the addition of disinfectants at the farm, it is still unclear why
such fluctuation of removals was attained.
50
-------
SCREEN
EQUALIZATION TANK
AIR
SETTLING POCKET
Figure 18. Layout of experimental activated sludge set-up.
-------
0.25
Q.
O
Q
S = 1 -*-2 g COD/dm3
y
3.5 g/dm3
0.12
"~ 0.08
0.04
0.113 after anaerobic digestion
S0 =2.9 -i-3.8 gCOD/dm3
— P
XV=4.3 +4.7 g/dm3
V(0
200 400 600 800 1000 1200
ALUM DOSE (mg/dm3)
Figure 19. Effects of coagulation on COD removal rate in (batch) activated
sludge.
52
-------
Continuous Studies - Pretreatment by Settling
Continuous studies (Figure 18) comprised of two systems: I) alum coagula-
tion - activated sludge, and II) settling - activated sludge - alum coagulation.
3
In the first system a massive dose of alum was used: 1200 to 1400 mg/dm . In
3
the second system the coagulant dose was 200 to 250 mg/dm as higher doses
yielded nothing in terms of additional removal. The tests were run on Farm A
wastewaters in 1978, prior to modification, thus, the influent concentrations
3
were very low. The average raw wastes CODf was 3870 mg 07/dm (SYSTEM I) and
3
3130 mg 02/dm (SYSTEM II). The aeration time was varied from 18.5 to 28.2 hr
and activated sludge loading F/M from 0.4 to 1.4 kg COD/kg MLSS/day. In order
to remove the difficulties in maintaining small flow-rates through aeration
tanks, varying volumes were used. Each test consisted of three parallel units
with volumes 11.4, 17.9, and 34.4 dm . Activated sludge concentration was
3
kept at 2.2 to 3.5 g/dm . The results are summarized in Figure 20. Part A
depicts the soluble COD removal across the aeration tank only indicating only
five percent higher removal efficiency, at the preferred F/M=0.5 kg COD/kg
MLSS/d for the first system, i.e. coagulation - activated sludge.
Sludge growth kinetics study has shown that SYSTEM I had sludge synthesis
_i
coefficient a=0.53 and sludge endogeneous respiration coefficient b=0.03 d
The coefficients for SYSTEM II were a=0.40 and b=0.008 d~ .
The analysis of performance data for the whole treatment trains reveals
that at 24 hr aeration time, SYSTEM I efficiency is approximately 5 to 6 percent
higher than the COD removal efficiency of SYSTEM II, settling-activated slude-
coagulation. Figure 20-B shows that the overall sludge production in SYSTEM I
rises by over 120 percent above the SYSTEM II. The data is verified by the
results of settling tests (Figure 17).
Biological Comparison of Various Pretreatment Methods
Three parallel batch activated sludge units were observed during 24 hr
aeration. Tank 1 treated effluent from 2 hr sedimentation, tank 2 from
anaerobic digestion in continuous flow contact digestors with sludge recycle,
3
and tank 3 treated effluent from coagulation with 1 g/dm -alum.
53
-------
100
1
Q
O
80
60
0.2
0.4 0.6 0.8 1.0
CODf - F/M (mg02 /mgMLVSS> d )
30
UI
20
§ 15
U)
10
SYSTEM I
•o
of
£
CM
SYSTEM II
18
20
22
24 26 28
AERATION TIME [ h J
30
Figure 20. Continuous activated sludge treatment of system (I) C-AS and (I)
S-AS-C: (A) CODf removal across the aeration tank; (B) Overall
sludge produced in the whole treatment train.
54
-------
Methods—
Activated sludge before being fed to the batch reactors was acclimated to
the wastes under study, e.g. tank 3 activated sludge was grown on substrate
subject to alum coagulation. Hydrobiological analysis encompassed microscope
studies at t=0, 12 and 24 hr. The method used (99) was confined to description
of floes, characterization of species and evaluation of the empty spaces
between floes. Photographs made are reproduced at 40 x magnification. The
activated sludge activity test TTC (100) supplemented biological analysis.
TABLE 4. COMPARISON OF THREE PRETREATMENT
METHODS - BATCH ACTIVATED SLUDGE
Settling
Tank 1
3
Influent COD - S mg O-Xdm
Sludge content - X mg VSS/dm
F/M mg COD/mg VSS
Sludge growth g.VSS/g BOD
Removal rate - K (h )
TTC (y mole TF/g VS) t = Oh
t = 24h
Cilliata sp. (Opercularia) fraction
of the total no. of organisms (%)
TIME: Oh
12h
24h
Visual analysis of floe type and
open spaces from photographs.
3,760
3,720
1.01
0.55
0.070
0.085
0.019
76.8
84.6
50.8
( — )
Anaerobic
Tank 2
2,700
6,424
0.42
0.78
0.09
0.045
0.010
67.7
70.1
91.9
X^^i \
\^r^r I
Chemical
Tank 3
3,180
4,301
0.74
0.42
0.062
0.032
0.023
35.1
49.3
72.1
(+-)
The physical-chemical parameters, particularly the shape of the removal
curve and the value of the removal rate coefficient indicate that anaerobic
pretreatment is better than the two remaining methods. The anaerobic-aerobic
system exhibits also the best sludge growth and the highest proportion of the
protozoans, Cilliata. The presence of these species is regarded traditionally
as the bioindication of the healthy state of the biota (101, 102).
Visual observation has also been performed. The photographs in Figure 21
illustrate on the left-hand column the sludge C40x) at t=0h, i.e. previously
55
-------
Reproduced from
best available copy.
Figure 21. Microscopic picture (40 X) of activated sludge: at t = Oh tanks
1, 2 and 3 are respectively A, C, and E; at t = 24h tanks 1, 2
and 3 are respectively B, D, and F: 1) agglomerated floes,
2) Zooglea uva Kolkw., 3) Opercularia sp. - poor physical
condition colony, 4) Opercularia sp. - two individuals colony of
good condition.
56
-------
acclimated to the wastes; at right-hand column the same sludge after 24 hr
batch aeration.
The tank 1 sludge at t=0 had compact floes and loose bacteria clustering
the free spaces. After 24 hr, the quality of floes deteriorated significantly,
the floes were still compacted, but they appeared as if coagulated with the
increased number of zoogleal agglomerations that occupied close to 50 percent
of the empty spaces were cleared of single bacteria.
The tank 2 sludge, after anaerobic pretreatment, exhibited little negative
changes during 24 hours. The floes were much better formed than in tank 1 or 3,
looked very fluffy and well multi-layered. The agglomerated or compacted floes,
i.e. very dark points evident in the photographs (a negative symptom) have
intensified in 24 hr, however, the empty spaces cleared and the overall picture
is superior to the parallel systems.
In tank 3 the flocks were formed spaciously, however, the appearance was
similar to activated sludge after prolonged aerobic stabilization, both at 0
and 24 hr. The flocks were not multi-layered as in tank 2 and were quite
dispersed with very weakly defined empty spaces. The initial dehydrogenase
activity (TTC) was the lowest of all three for tank 3.
Conclusions—
Pretreatment by settling only yielded sludge (in tank 1) which lacked
full development, was compacted into agglomerations, thus not exposed to sub-
strate. Problems may be expected in full scale operation.
The activated sludge in tank 2 had the most satisfactory appearance, was
well formed in multilayered floes, and had good bacteriological and protozoo-
logical characteristics. Anaerobic pretreatment has produced the most
appropriate effluent for further activated sludge treatment.
Pretreatment by coagulation results in the development of loosely formed
dispersed sludge, which appears the most vulnerable of all three to minor
changes in the technological regime.
57
-------
SECTION 7
POLISHING TREATMENT
The most popular secondary processes used so far in full scale encompassed
aerated lagoons and activated sludge. The GOD. concentrations of activated
3
sludge effluent from Plant A were equal to 300 to 500 mg/dm and further treatment
was needed. Solids overflow has prompted the designers to install the flooded
filtration units, filled with sponge. The units clogged very easily and it
was extremely difficult to backwash them. Within this project, the filter in
Plant A, Vidus type treatment plant, has been reconstructed and refilled with
coke media to be operated as a full scale polishing anaerobic biofilter.
Other methods of polishing treatment were tried in laboratory scale:
treatment in a series of four algal ponds and by extended aeration activated
sludge. Results of these three pretreatment methods will be reported to
evaluate the feasibility of biological polishing treatment in both low- and
high-level energy input processes.
For all polishing treatment tests, the wastewater was collected from the
final clarifier. The location of the clarifier is as in Figure 5.
ACTIVATED SLUDGE POLISHING TREATMENT
Four parallel aeration tanks were used as in Figure 18 loaded with BOD--
F/M 0.03 to 0.47 g 02/g MLSS/day. The results are summarized in Table 5.
The aeration volume was varied in order to better control the peristaltic
pumps delivery. Extended aeration periods were used, and it was found quite
difficult to build up the activated sludge biota. The removals attained at
the low values of activated sludge organic loading F/M were increasing linearily
with the aeration time shown in Figure 22A while the effluent concentration
has shown a parabollic increase with the F/M increase as shown in Figure 22B.
58
-------
TABLE 5. ACTIVATED SLUDGE POLISHING TREATMENT OF
EFFLUENT FROM THE VIDUS TYPE PLANT AT FARM A
Parameter Units
Aeration -
tank vol. dm
Aeration
time - t hours
F/M BOD5 g 02/g/d
Influent .,
S £ mg 02/dm
COD
BOD5
Effluent 3
S , mg 0,/dm
e, r t.
COD
BOD5
MLVSS - X mg/dm3
Removal rate ,
k d
COD
BOD5
Removal ef-
ficiency %
COD
BOD
Series I
1
5.7
28
0.47
142
86
84
44
170
0.494
40
49
2
11.4
41
0.29
138
76
72
29
170
0.432
48
62
3
17.9
55
0.19
142
64
60
20
190
0.445
58
73
4
34.4
79
0.03
138
78
51
13
780
0.091
63
83
1
5.7
33
0.15
110
44
66
22
210
0.253
40
50
Series II
2
11.4
37
0.09
110
44
62
20
220
0.251
44
55
3
17.9
57
0.05
110
44
57
17
370
0.116
48
61
4
34.4
62
0.02
110
44
55
15
850
0.05
50
66
59
-------
50
20
40 60 80
AERATION TIME(h)
CO
E
Q
O
CD
LU
D
30
20
10
= 110-140mg02/dm3
0,1 0,2 0,3 0,4 0,5
F/M(g02/gMLVSS-d)
Figure 22. Continuous extended aeration of activated sludge effluent from Farm A: (A) Efficiency as
affected by aeration time; (B) Effluent quality versus sludge loading in the test units.
-------
Relatively low removals at 24 hx detention time, equal approximately
to 45 percent BOD,. ,. and 40 percent CODf, when analyzed from the standpoint
j, £ I
of costs of aeration prove the process to be uneconomical.
It was then assumed that intermittent aeration, cutting the aeration costs
by over 50 percent, with the simultaneous increase of the aeration volume
should yield similar or better results. Intermittent aeration was practiced
in two parallel runs, 11 hr/d while the remaining 13 hr, the MLSS were allowed
to settle. The data is presented in Table 6.
The kinetic constants were calculated based on the standard kinetic model,
described before in the text, k.S /S ?=> S /2 .t.
TABLE 6. COMPARISON OF INTERMITTENT AERATION IN POLISHING
ACTIVATED SLUDGE TREATMENT OF BIOLOGICAL EFFLUENT
Tank No. Aeration Retention F/M
J
Volume Time hour/day hours kg BODe/kg MLSS/d Removal Percent
Tank 1
11.4 dm3
Tank 2
11.9 dm3
24 hr/d
11 hr/d
24 hr/d
11 hr/d
37
37
57
57
0.09
0.09
0.05
0.05
55
50
61
57
3 3
Note: Influent COD = 110 mg 0,/dm , BODS <= 44 mg 00/dm
MLSS = 140 - 230 mg/din *
Conclusion
Relatively low values of effluent BOD- and COD may be attained in polishing
treatment with activated sludge, however, prolonged aeration times make the
economic efficiency questionable. One feasible solution would be to use deep
facultative lagoons, aerated at a low power level or a completely mixed lagoon,
with a settling compartment, aerated intermittently. In rural circumstances
the costs of earthen tanks would not be high and the use of interrupted
aeration would result in low power consumption and extensive denitrification.
61
-------
A significant portion of the BOP and COD in these polishing teats was of
nitrogeneous origin, as most of the carbon was. removed easily in the first
biological stage. Based on these studies, low-energy-input aerobic-anoxic
(i.e. not strictly anaerobic) or a facultative system is recommended as a
polishing system for the biological effluents from the high rate full scale
plants.
ALGAL TREATMENT SYSTEMS
Introduction
Full scale algal systems have so far been applied only in regions where
there was adequate land area and sunlight, reasonable uniform warm water tem-
peratures, and when there was a demand to utilize the grown biomass while the
harvesting methods were economically acceptable.
Large scale demonstration plant on algal (Chlorella vulg.) treatment of
nitrogeneous factory wastes in Poland has demonstrated that problems in this
region are: low temperatures and high costs of extracted protein about equal
to the price of meat protein (41, 42, 103).
The major regular traits of all so far conducted experiments were well
described by Goldman (43) as follows:
1. light conversion is less than 5 percent;
2. short-term effects, in idealized lab conditions, yield the (limit)
2
maximum growth rates of 30 to 40 g.DM/m /day;
2
3. long-term effects are much lower and seldom exceed 15 to 24 g DM/m /
day (periods of 1 to 3 months);
4. there were no reported cases of pure monoculture growth for an
extended period of time;
5. the usually stray cultures, dominating after some, time, were
Scenedesmus, Chlorella or Micractinium; and
6. temperature is of smaller significance than light intensity.
The studies reported so far on piggery effluents treatment in algal systems
yield little more insight. The batch work of RogiAski (44) on filtered waste-
62
-------
waters has suggested the need for immediate dilution of piggery effluents,
and was quite optimistic. These laboratory shaker experiments yielded 0.3 to
2 g dry matter (DM)/dm /day.
The experiments in Israel (45) have shown that ponds, enriched with pig
wastes evidenced an increased fish production from 10 up to 30 kg/ha/day, due
to the increased plankton content by 100 to 1000 times, as compared to ponds
without fertilization. The plankton had 45 to 55 percent of crude protein.
The field studies in Ghent by de Pauw (46) on significantly diluted raw
2
pig wastes have shown yields 0 to 10 g DM/m /day depending on temperature and
solar radiation. The dry matter, i.e. total solids content was maintained at
23 3
30 g/m , i.e. 200 mg DM/dm varying as 30 to 220 mg DM/dm . In the greenhouse-
2
winter conditions these studies have proved yields of 1 to 3 g DM/m /d while
2 2
average summer yields were 4.8 g DM/m /d for Chlorella and 9.2 g DM/m /d for
Scenedesmus. The optimum retention time was 2 to 8 days depending on light
2
intensity which varied as 43 to 144 cal/cm /d. Nutrients conversion rates or
biomass production was Y(N)=10 g DM/g N and Y(P)=60 g DM/g P.
The laboratory studies in high temperatures (37 C) were reported by
Boersma, et al. (47) to give very high yields depending on dilution of pig
3
wastes and the resulting nitrogen content. At 63 mg N/dm , the yield was 8.9
2 3 . 2
g/m /d while at 125 to 250 mg N/dm , the yields increased from 22 to 45 g/m /d,
2
at light intensity of 381 uE/m second. The total amino acids content was
41 to 38 percent and crude protein was 55 to 50 percent in the Chlorella
vulg. cell mass. The product yield and cell density decreased with retention
time which varied from 2.5 to 6.7 day.
The aim of these studies was to find out the feasibility of algal treat-
ment of biological effluent from Pla,nt A,, the cell yields and protein content
of recovered biomass and to estimate des.ign parameters for a full scale
system.
Experimental Methods
The studies were conducted in a system of four ponds in series, with
facilities for collecting intermediate samples. The ponds were placed in a
63
-------
photostat and lighted continuously (24 hr/day) by fluorescent bulbs.
Temperature was approximately 23 + 3 C which was equal to room tempera-
ture.
Two parallel series of four ponds each, were used, with volumes of the
3
ponds: series A - 20.0, 21.45, 23.15 and 40.0 dm ; series B - 19.6, 21.0,
3 2
22.95, and 40.0 dm ; the overall surface areas were: F(A) = 4,669 cm and
2
F(B) = 4,707 cm . The corresponding retention times were: A = 3, 9, 18, and
21 days and B = 6, 12, 24, and 25 days in all four ponds.
The influent wastes were metered by peristalitic pumps. The wastes were
collected from the final clarifiers at the Vidus type treatment Plant A. The
wastewaters fed to the ponds had extremely variable quality which had a pro-
found effect on the performance and made interpretation quite difficult. The
3 3
values of BOD5 f varied 8 to 220 mg 02/dm , COD- = 9 to 1,038 mg 02/dm ,
PVf = 28 to 470 mg 02/dm3, N-NH3 - 147 to 900 mg/dm3, N-N03 = 1 to 153 mg/dm3,
N-N02 = 1 to 994 mg/dm3.
The treatment efficiency of the ponds was determined on the basis of
physical-chemical parameters: pH, BOD, COD, DO, SS, TKN, N-NH,, N-N02, N-NO-,
PO,, Total P, potassium; and on the basis of biological analyses and biomass
harvesting yields, protein content and amino acids composition. Biological
analysis included species definition, total cells counts and counting of
dividing cells.
The system was seeded with a mixture of Chlorella vulg. and Scenedesmus
pure algal cultures, the samples were collected after the passing of five
retention times. Only one seeding was made, thus, the system was subject to
cyclic (Harmonic in nature) variations in the biomass content.
Results
Biological data could not be correlated with physical-chemical data. It
is assumed that this was due to the contamination of the system with algae in
final clarifier effluent as compared to original seeding culture, the tempera-
ture variations, algae are quick to react to minor environmental changes,
extremely large influent concentration variability and the fact that 24 hr/day
64
-------
lighting was used. As noted by Matusiak fct al. (103), the lighting plays
a very significant role in the studies and intermittent lighting as used in
their experiments had profound beneficial effects on the algal performance.
The biological studies were also clouded because of the fact that experiments
were run in flow through, continuous systems maintained for several months
with one original seeding. As found by several other researchers (104) in
such systems, the biomass quality and quantity, particularly in laboratory
conditions exhibits oscillations with decreasing magnitude.
Since the detailed results contain close to 1,000 individual data points,
only major findings will be summarized here. Typical performance of the ponds
is depicted in Figure 23 for retention times of 18 and 25 days both during
October and November, 1979- In all experiments phosphates content increased
significantly, indicating decomposition of total phosphorus in the biological
effluent and coming from the decay of algal cells. In three instances there
was an increase of the content of N-NO., indicating completion of nitrification.
The removals of ammonia nitrogen were from 20 to 65 percent. The values of
BOD,., COD and permanganate COD decreased steadily in all cases. In long
hydraulic retention times (HRT) runs, the secondary decay of algae has intro-
duced error in these considerations. The data is condensed in Table 7. Each
efficiency is calculated as an average of the whole run, taking into account
the HRT.
Data Interpretation
The removal efficiency and effluent quality from algal ponds is dependent
inter alia on the HRT and surface area loading (AL) . So far in the literature,
there are very few instances where successful correlations or removal against
AL and HRT were attained and data is usually reported in terms of average bio-
mass yields and gross removal ranges without kinetic interpretation. The
first attempts in this study have revealed the reason for such approach. At
low influent waste concentration to the. ponds, the removal rate coefficient
was affected by the influent wastes concentration, which, made; the. interpreta-
tion impossible, without dividing the data into S concentration groupings.
65
-------
POND
X
3.6
36
3,6
IV
7,2
OVERALL
REMOV.
POND
X
IV
OVERALL
REMOV.
10
= 25d
BOD5
41
70
414%
57.1 %
BO
83
_7J
68
60
25%
80
75
66
28.7 %
COD
29.4 %
53.5 %
P04
47
19,5 %
increase
31
29
25
27
52 %
increase
TKN
36',
323
296
Y/A
202
42.5%
77)
202
50%
N-NH3
188
169
172
169
J4!
23 %
55%
N-N03
73,1 %
78
69
62
64
39,7 %
N-N02
56 %
2.6
94,2 %
Figure 23. Averaged effects of treatment In a series of four ponds. An
example for 18 and 25 days HRT runs.
66
-------
TABLE 7. CHARACTERISTIC OVERALL REMOVAL EFFICIENCY ATTAINED IS THE SERIES OF FOUR ALGAL PONDS
Retention
(daysi)
Dates
Removal (U)
EOD5,f
C01)f
PV-cod
N-SH.J
N-N03
N-NO,
ng O^/dm
SeB01)5.f,
mg 0.,/dm
Load
g EOD5/mZ/d
S PV
ng 0,/dm
S PV
mg 0,/da
3
Mar. 21
to
April 6
1979
1
11
4
0
- '
-
:
18
16
1.35
24
24
6
Mar. 21
to
April 6
1979
2
5.5
29.4
8.3
-
-
-
18
17
0.65
24
22
May 3
to
June 6
1979
3
62.8
-
42.6
22
3
«
113
42
2.88
258
148
9 ''
June 19
to
Aug. 10
1978
4
51.5
-
32.1
4
36
48
66
32
1.60
109
74
12
Feb. 5 May 3 April 19
to to to
Feb. 24 June 8 Aug. 10
1979 1978 1978
567
34 74.8 65
4
(+) 34.6 41
- 19 0
<+) 25
(+) 53
97 115 66
64 29 23
2.41 2.11 1.22
52 240 112
63 157 66
1
Feb. 5 Sept. 7
to to
April 24 Oct. 12
1979 1978
8 9
50.5 65.5
18.7
(+) 48.7
32
(+)
65
97 58
48 20
1.77 0.74
52 115
56 59
18
Oct. 20
to
Nov. 17
1978
10
41.4
29.4
25
23
73
55
70
41
0.88
80
60
21
Dec. 8
to
Jan. 16
1979
11
27
27.6
22.2
42.6
40
75
63
46
0.67
63
49
24
Nov. 29
to
Jan. 16
1979
12
19
32
3.2
41
(+)
96.6
63
51
0.56
63
'61
Sept. 7
to
Oct. 5
1973
13
53.8
-
62
56
(+)
96
58
26
0.52
115
44
25
Oct. 12
to
Dec. 17
1978
14
57.1
53.5
28.7
5.5
66
31
70
30
0.71
80
57
(+) Increase of concentration in the effluent.
-------
The effluent quality, expressed in permanganate COD (PVcod) showed a
'2
predictable increase with the increase of AL (g/m d) (see Figure 24-B). The
correlation for BOD_ was not acceptable, while the same correlation for COD^
(g 02/dm3) yielded a straight line (55):
S (COD) = 0.0278.AL; (14)
e
2
up to 20 g/m /d; beyond that point it begun leveling off toward an asymptote
parallel to AL axis.
Figure 24-A shows the correlation of removal efficiency against AL of
BOD,. ... It is characteristic that the efficiency increases with the influent
^ > ^
substrate concentration.
The scatter of data points, symptomatic for majority of algal studies in
flow-through systems, does not preclude a definite delineation of different
curves. This is further evident in the kinetic interpretation in Figures 25
and 26. The first order removal plot usually assumed for pond systems:
(SQ - Se)/HRT = k Se (15)
in Figure 25 shows poor agreement even when data is pooled into four raw
concentration groupings S £ 60, S = 61 to 105 units, S = 106 to 150 units
and S > 150 mg/dm - BOD5 -. A much better removal is shown in Figure 26
where the authors have applied a novel kinetic model:
(SQ - Se)/HRT = k (Se/SQ) (16)
which incorporates the effects of influent substrate concentration. The model
constant is named substrate kinetic removal constant and correlates the algal
ponds removal data with satisfying accuracy. It is interesting to note that
Hebrowska (104) has observed similar effects of concentration on the increase
of removal efficiency.
In order to find an optimum retention time, plots of efficiency and bio-
mass yield against HRT were made as shown in Figure 27. Plot A demonstrates
that biomass yields are not well correlated against time and that the expected
rt O
production based on laboratory data at 23 C, should be below 10 g DM/m /d at
retention times around 15 days. It follows from part B of the same figure that
68
-------
o
o
CO
u.
u.
UJ
o <60
A61-M05
A106-M50
456
AL -BOD^(g02/m2-d )
AL-PVCQD (g02/m-dl
Figure 24. Efficiency of BOD. . removal and permanganate COD in the
effluent from four'ponds system.
69
-------
0,1 0,2 0.3 04 0,5 0.6 0,7 0,8 0,9
Figure 25. Kinetics of BOD,. - removal in series of four ponds: (A) First order; (B) Authors' model
substrate kinetics data pool.
-------
CO
Q
8
cn
E
cc
I
O)
to
10
8
6
4
2
I T
S0<60
= 23
10
8
6
~ 4
T3
CO
k3=11,3
106 0 0,1 0,2 0,3 0,4 0,5 0,6
10
8
6
4
2
61 151
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9
0 0,1 0,2 0,3 0,4 0,5 0,6
Se/So
Figure 26. Kinetics of BOD_ , removal in algal ponds; four influent concentration ranges - authors'
substrate model.'
-------
•a
-------
the optimum retention time is 9 to 12 days, thus the expected algal biomass
2
yield is 3 to 4 g DM/m /day.
In the studied case of Plant A the full scale ponds are presently con-
structed. The design criteria and expected yields are: HRT = 15 d, AL = 10 g
o
0_/m /d - COD., removal efficiency - 50 percent for COD, BOD-, biomass yield 2
2
g/m /day. Fish will be planted in the final pond, thus the biomass will not be
harvested mechanically.
Conclusions
The system of four ponds was found very difficult to interpret due to
several overlapping interactions and disturbances. The biological information,
such as algal cells and bacterial cells counts could not be correlated with
physical-chemical parameters. Decay of algae and resulting secondary contain-
ation of the system has resulted in significant distortions of the data.
It has been found that the removal efficiency is increased with the
increasing substrate concentration. A new model (S -S /HRT = k(S /S ) intro-
o e e o
duced by the author was found adequate for data interpretation, provided data
was pooled into narrow raw wastes concentration ranges.
The expected efficiency of the algal ponds in summer conditions is over
50 percent removal of organics, at HRT = 10 days and with simultaneous biomass
2
production of over 3 to 5 g DM/m /day. It should be borne in mind that the
diurnal cycle of natural light will increase significantly the efficiency of
the algal ponds beyond yields obtained with this study. Based on those assump-
tions, a full scale fish pond system is presently installed.
POLISHING ANAEROBIC BIOFILTRATION
Full scale waste treatment train at Plant A contained an anaerobic filter
filled with sponge as the main polishing, mechanical filtration unit. Due to
clogging, overflowing and solids carryover, the filters were put out of opera-
tion. The authors have reconstructed the filter, its inlet and the washing
system and replaced the sponge media with coke on a gravel underlayer. The
studies reported have been conducted between January 1 and June 1, 1979.
73
-------
Detailed discussion of results obtained and the modifications on the two paral-
lel biofilters at Plant A is presented elsewhere (21, 105, 106).
The wastewaters were fed to the biofilters from the final clarifier
(Figure 5), and on the average had the following characteristics:
o 3
BODC . = 30 mg/dm (range 10 to 140), BOD,. . = 70 mg/dm (18 to 180), CODf =
j, £ ~ j,ni n r
205 mg/dm (range 44 to 480), CODnf = 410 mg/dm (100 to 1320), N-NH3 = 110 mg
NH4/dm3 (range 68 to 250); VSS = 78 (2 to 380), TDS = 1040 mg/dm3 (540 to 2400).
The characteristic removals attained were on the average 40 percent BODC ,.
j,nr
(30 percent BOD. -) and 38 percent COD . (30 percent CODf). Major problems
encountered were due to significant influent concentration variability as noted
in the range of numbers.
The kinetic of removal in the biofilters, interpreted by means of the
author's model, details in Section 4, are presented in Figure 28. Two fairly
distinct operations regimes are evident, the change of removal rate occuring
3
at L = 1.0 to 1.5 kg/m /d. Otherwise, it should be stressed that the biofil-
ters were exhibiting fairly stable removal characteristics, proportional to
the load applied.
Discussions and Conclusions
The full scale anaerobic biofilters in the modified version with coke
media offer short retention time and 30 to 40 percent removal of organics.
The denitrification occurs to only a small degree, due to rather high organic
loadings, high influent nitrogen concentrations. The removal of nitrites and
nitrates varies from 20 to 44 percent. Due to incomplete nitrification, the
total nitrogen content in the effluent is still excessively high. Ammonia
3
nitrogen content in the effluent is 40 to 160 mg N/dm . The biofilters
remove almost no ammonia nitrogen, at times an increase occurs which is indica-
tive of incomplete aerobic biooxidation in the aeration tank.
The biofilters exhibit up to 20 percent increase in the effluent suspended
3
solids, the concentrations being TSS = 60 to 270 mg/dm and VSS = 20 to 80
mg/dm , which indicates high level of mineralization in the sloughed-off
sludge.
74
-------
Ol
1000
80S
0.2 0.4 0.6
1.0 1.2 1.4 1.6
(m3d/kgCOD)
1.8
Figure 28. Removal kinetics in full scale anaerobic polishing biofilters.
-------
It is concluded that the biofliter working in the operational regime,
2
average values in parenthesis: L=0.3 to 2.8 (1.10) kg BOD/m /d; 0.2 to 1.6
(0.7) kg BOD/m3/d, 1 to 13 (3.7) kg COD/m /d at hydraulic loading of 6.5 to
3 3
11.5 (9.4) m /m /d provides insufficient nitrogen removal. Insufficient
breakdown of the organic matter in the preceding aerobic biotreatment yields
further production of N-NR, in the anaerobic biofilter. Due to high organics
loadings, the denitrification did not proceed to satisfactory levels.
76
-------
SECTION 8
ANAEROBIC TREATMENT
INTRODUCTION
The anaerobic treatment has been traditionally applied for almost a
century to sewage sludges, more as a method of sludge stabilization than
energy recovery. Recent increases in animal wastes generation rate and the
mounting energy crisis directed designers towards anaerobic digestion of
wastewaters. The conventional flow-through anaerobic digestors have been
regarded as applicable to wastewater with influent COD concentrations above
3
4000 mg 02/dm (13, 57). The other popular contention among the animal waste
anaerobic treatment researches was that the lower limit of total solids
qualifying wastewaters for anaerobic digestion was two percent. This stemmed
from the fact that the small farm wastes had much higher concentrations of
organics, that flow-through systems were used and no reliable data to the
contrary was reported. Recent re-introduction of the contact or sludge
recycle process to treatment of the more dilute effluents has shown that there
are no lower concentration limits to efficient anaerobic digestions. The
increasing interest was stimulated also by the fact that at higher concentra-
3
tions, e.g. above 20 g/dm COD, the aerobic treatment has been found to be
four times as expensive as equivalent anaerobic treatment (89).
This work will demonstrate that the flow-through anaerobic digestion
without sludge recycle is feasible on wastes with TS below one percent and
3
COD f around 10 g 02/dm . Several process modifications will be studied as
presented in Figure 29. The batch studies will be conducted to obtain the
adequate volume of seeding sludge. Following the batch tests, flow-through
studies will be conducted at varying detention times to test the lower limit
of the HRT and efficiency of methane generation from the unit of organics
removed. In order to further cut down on the HRT, contact digestion studies
77
-------
CH, * C02
: co2
ANFLOW REACTOR
ANBIOF
REACTOR
INFLOW
ANCONT REACTOR
INFLOW
EXCESS
SLUDGE
Figure 29. Layout of various anaerobic treatment process modifications
studied.
78
-------
will be run in a continuously fed, gas mixed reactor with sludge
recycle.
The main problems of the contact digestion system are the generation of
dispersed anaerobic sludge that settled poorly, and thus makes the control
of the sludge concentration MLSS in the reactor difficult. In order to
alleviate this problem anaerobic biofilters will be tested as the modification
of the contact process where sludge is attached to the filter media rather than
being retained in the system through recycle (Figure 29).
Anaerobic treatment of piggery wastewaters in concentrated form has so
far been confined to the batch or flow-through studies in laboratory scale
(67, 69, 70) or pilot studies (71, 72). So far there have been no anaerobic
studies on the treatment of dilute piggery wastes from large farms, and no con-
tact digestion studies on any hog wastes reported so far. The experiments
reported supply, then the parameters of the fermentation and methane generation
kinetics of dilute piggery wastes and will provide basic process design data.
Practical Aspects of Anaerobic Biodegradation
Anaerobic wastewater treatment processes have not found wider use for
several reasons:
1. Aerobic processes were considered more rapid.
2; Power costs expended for aeration were much less noticed before
the onset of current energy shortage.
3. Anaerobic processes were thought unreliable and fit mostly for
organic sludges.
4. Anaerobic pathways are still not agreed upon; the multitude of
kinetic equations used in modelling the process is not easy to
apply by the designers.
5. The optimum conditions and process operating parameters are not
well known; the designers in all cases have to rely on feasibility
studies.
6. There is generally established contention that anaerobic processes
are much more easily susceptible to toxicity and other environmental
stresses than aerobic processes, and that the recovery time is
significantly longer.
79
-------
Basically, during anaerobic fermentation, at least three intermediate
steps can be distinguished (90): 1) liquefaction or hydrolyzation of complex
polysaccharides (cellulose), proteins and lipids by fermentative bacteria to
organic acids, alcohols, H^ and CO^J 2) acetogenesis by the group of HL-pro-
ducing, acetogenic bacteria, which obtain energy for growth by producing
acetate C0_, H~ from acids and alcohols produced by liquefaction; and 3) methan-
ogenesis, i.e. utilization of H«, (XL, and acetate in the production of final
products, CH, and CO-. Hashimoto et al. (96) have recently presented a
mechanism with four groups: first, hydrolytic bacteria; second, hydrogen pro-
ducing acetogenic bacteria; third, homoacetogenic bacteria forming acetate
from H2, CO-, and formate; and fourth, the methanogenic bacteria. Detailed
knowledge of the process is still lacking as noted by Balch et al. (97).
Thus, for all practical purposes, a basically diphasic process is considered
here, with acidic phase preceding the rate determining (i.e. slower) methano-
genic phase.
In an actual fermenter the two processes are occurring simultaneously.
The methanogenic bacteria being strict, fastidious anaerobics are reproducing
slowly, and thus, the minimum design retention time is limited by the sludge age
or SRT (biomass). There are three distinct optimum operational temperature
ranges: 1) psychrophilic 5 to 15°C; b) mesophilic 30 to 35°C; and
c) thermophilic 50 to 55 C. The increase of temperature is accompanied by
the decrease of the minimum doubling time.
In view of the above, the decrease of the volume of the anaerobic reactor
(cost minimization) may be obtained by means of increasing the temperature or
sludge age, while keeping the HRT down.
With highly concentrated animal wastes, TS above 6 to 10 percent, the
increase of reactor temperature should be considered. Since sludge age cannot
be effectively increased due to solids separation problems, various researchers
have optimized the volume of the flow-through-anaerobic-reactors without sludge
recycle down to five days (HRT=SRT=5 d) at 55°C (91, 92, 93), i.e., operation
within the thermophilic range.
80
-------
With dilute effluents such as the piggery wastewaters described in Chapters
1 and 2, the best way of bringing the HRT down is to increase the SRT by solids
recycle or retention as in the heterogeneous (fixed film) reactor.
BATCH DIGESTION
Batch digestion tests initiated all anaerobic experiments and were con-
tinued throughout the studies in order to provide seed and at the same time
test possible inhibitory charges to the reactors.
Experimental Method
3
The studies were conducted in 3 dm glass vessels, kept in thermostats
at 35°C + 0.5°C. The method has been adopted after Chmielowski (52). The
setup is presented in Figure 30. The vessels were sealed with specially
designed valves containing thick rubber membrane that allows multiple per-
forations by hypodermic needle of a syringe feeding the tank and evacuating
accumulated biogas to the manometer. The biogas was vented once per day, the
pressure was compensated by the Boyle-Mariotte equation, so data is reported
in standard temperature and pressure conditions (STP). The error due to gas
solubility in liquid phase is + 1.5 percent. The changes of volume of the
digesting liquor due to the introduction of substrate were taken into account
in calculating the volume of biogas produced. After measuring the gas pres-
sure, calculating volume, and feeding the substrate, the pressure was equalized
with ambient pressure. Excess pressure during substrate feeding warranted
air-tight operation.
Three series of experiments were conducted, each was initiated differently.
In all experiments raw piggery wastes from Plant A were used after screening
and sieving. The first series consisted of introducing actively digesting
municipal sludge fermented in controlled laboratory conditions. Seven vessels
3
were fed 1 dm of sludge each, the air was evacuated by purging with propane-
butane mixture. The vessels were placed in 35 C and gas production (GP) was
monitored daily. When GP definitely slowed, to each of the six vessels, 0.1
3
dm aliquots of piggery wastes were added, the seventh was left unfed as a
reference for calculating GP from the seed sludge alone.
81
-------
HYPODERMIC
AIRTIGHTCONNECTOR
GAS VALVE
T = 35°C
V = 2dm3
GAS COMP.
DIGESTION COMP.
V= 1dm3
.MERCURY
MANOMETER
RUBBER
SmLLuiv\ MEMBRANE
1
t
*X BRASS
FEED/OUTPUT
PORT
D
-i
SYRINGE
Figure 30. Pressurized glass reactor used as batch and ANFLOW reactor.
82
-------
The second series was initiated with pig wastes and sludge from the first
3
as. Here 0.2 dm of wastes were added to 1 dm"
the first feeding was at the start of experiments.
3 3
series. Here 0.2 dm of wastes were added to 1 dm of seed mass in the vessel,
The third series was initiated with digested effluent from the second
series, and conducted the same as the second one, 0.2 dm of wastes were fed
at the decrease of GP.
Results—
The full results of all three experiments are summarized elsewhere by
Oleszkiewicz et al. (94) and Janiczek et al. C95). The quality of manure fed
into the laboratory vessels varied as follows: pH = 7.1 to 7.7; BOD- ,. = 1,500
3 3 '
to 14,000 mg 0,,/dm ; COD ,. = 4,000 to 2,600 mg 0,/dm ; total nitrogen = 600 to
,£ nr 3
1,700 mg N/dm ; chlorides = 100 to 190 mg/dm ; total alkalinity = 30 to 70
milivals/dm3; TS = 2,200 to 15,000 mg/dm3; TVS = 1,700 to 9,200 mg/dm ; TDS =
300 to 4,500 mg/dm3; VSS = 150 to 3,300 mg/dm3; TSS = 1,600 to 11,000 mg/dm ;
VSS = 850 to 6,700 mg/dm3.
The overall gas production from the introduced organic matter to six
vessels in Series III is presented in Table 8, while Table 9 presents the
arithmetic averaged specific gas production data from all three series, all
data is based on nonfiltered samples. The GP denotes volume of biogas
generated from the load introduced.
TABLE 8. BIOGAS PRODUCTION IN SERIES III
Tank
1
2
3
4
5
6
SERIES I
ZGP „
\CIH 7
SERIES II
ZGP _
I cm /
SERIES III
ZGP
o 3,
TOTAL
TVS =
1428
ZL :
o
2383
ZL :
0
6369
LOAD INTRODUCED ZL DURING
3.29 g, BOD
670
TVS = 2.848
2002
TVS = 6.016
7618
o
= 3.6 g, COD =
2196
g, BOD = 2.35 g
2060
g, BOD = 8.46 g
5713
EXPERIMENT
8.19 g
956
, COD = 6.
2283
, COD = 18
7274
1673
12 g
2484
.88 g
6685
685
2020
No Data
83
-------
TABLE 9. SPECIFIC GAS PRODUCTION FROM INTRODUCED
ORGANIC LOAD - BATCH STUDY
Series
I
II
III
SGP -
o .
TVS
0.383
0.775
1.120
Biogas Production Per Unit
BOD
0.450
0.940
0.796
of Introduced Load (m /kg) :
COD
0.154
0.361
0.356
The results indicate that only periodic acclimation of digesting biomass
to introduced wastewaters will produce satisfactory results in a short period
of time. In these experiments the initial seed to manure proportions were
8:1 to 10:1, and the minimum start-up time was 80 to 90 days. The gas produc-
tion data (SGP ) indicates that daily variations of wastewater concentration
do not influence directly the daily gas production (GP - cm /d) whose fluctua-
tions as seen in Figure 31 cannot be explained at this stage of experiments.
The increasing SGPQ from Series I to Series III indicates both the positive
influence of degree of acclimation and effects of increased concentration of
organics in the input.
Conclusions
The batch digestion tests have proved the feasibility of anaerobic fer-
mentation of dilute screened piggery wastes. Over 80 percent degradation of
organics was achieved. The SGP attained from the introduced organic load
varied with the increasing degree of acclimation of the fermenting biomass and
reached 0.36 m3/kg COD, 0.9 m3/kg BOD and 0.8 to 1.1 m3/kg TVS.
CONTINUOUS STUDIES IN ANFLOW REACTORS
A majority of the screening and treatability tests on anaerobic digestion
are run in batch or semi-continuous reactors, mixed once or twice per day.
Various researchers have found that such intermittent random shaking provides
completely mixed conditions from the standpoint of the process kinetics (52).
After testing the effects of intermittent mixing and the effect of the
length of acclimation to new environmental conditions on process efficiency
84
-------
350
oo
Ul
35
Figure 31. The course of batch anaerobic digestion in the third series of experiments.
-------
as discussed above, the continuous tests were initiated in completely mixed,
flow-through reactors without recycle called the ANFLOW reactors.
Experimental
The raw piggery wastewaters were collected in separate containers
throughout one day and then mixed to arrive at the desired concentration.
It was assumed that a relatively constant input COD concentration of approxi-
3 33
mately 2.5 x 10 to 14.5 x 10 mg 0«/dm should be maintained through-out
the experiment. The raw wastes were then sieved in order to facilitate
unobstructed addition by a hypodermic syringe through the rubber membrane
valve sho\
one week.
valve shown in Figure 30. The wastes were stored at 4 C for a maximum of
3
The active volume digesting liquid was 1 dm . Raw piggery wastes were
added once a day after the withdrawal of appropriate volume and determination
of gas pressure and quantity. The fermenters were manually shaken at least
two times a day to provide mixing. The volume of the output and input was
determined on the basis of desired retention times. Seventeen retention times
were tested: 1.5, 1.75, 2, 3, 4, 5, 6, 6.7, 8, 9, 10, 12.5, 15, 20, 25,
and 30 days.
Six parallel fermentors were incubated at 33°± 0.5°C for each experimental
period. The series at one retention time consisted of five individual runs,
each run was sampled at steady state conditions, a total of 17 x 5 = 85 runs
were made.
The analyses performed on samples from each run included: COD ,., BODC -,
nt j,nr
TOG, SOC, TSS, VSS, SS, TS, pH, alkalinity, acidity, nitrogen compounds and
random chlorides in raw and effluent concentrations were measured. A total
of 190 analyses series were performed. Additionally, 17 gas analyses were
made including C02, CH,, N« and water vapor. One complete gas analysis was
made for one retention time. Thus, the following data analysis is based on a
total of over 2000 individual data points.
Definition of terms—
Since the available literature on anaerobic digestion is full of ambiguous
notations of gas production, the following delineates unit GP parameters as
86
-------
used here. Two values for GP are given. The average gas production (GP )
•j a"S
denotes the ratio of total gas volume produced (m ) during the whole experimental
period in one reactor at one HRT.
n 33
GP = (S GP.) /n, . . . . (cm /d; m /d) .... (17)
avg i=1 i
Specific averaged gas production (SGP ) denotes the total volume of gas pro-
duced throughout the experiment at one retention time, divided by the total
cumulative load introduced into the reactor throughout the experiment:
n n ,
SGPa = (Z GP.)/(E L .),.... (in /kg) .... (18)
0 i=l * i=l °'X
Similarly SGPadenotes the production per cumulative load removed from the
reactor.
The GP expressed per average load introduced (SGP ) is determined as:
SGPQ = GP /LQ .... (m3/kg) .... (19)
where GP. is gas produced in day ^; n^ is the total number of days in the ex-
periment; L . is the load introduced in day jL. The daily load L . = Q.S .
varied with the inherent variations of daily S . as Q was constant. The
3. * 3,
average load during the experiment L = Q.S , where S is equal to an average
influent concentration, i.e.
So - (^So,i)/5 <20>
The specific production calculated on the basis of an average load
removed (L) is:
SGP = GP /L (21)
r avg r
where Lr = Q (Sa - Sa); and Sa = (E Se ±)/5
Theoretical GP is calculated on the basis of a stoichiometric equation of
GP from a kg or removed COD. The value used as a maximum should refer to TOD
2
rather than COD, according to Buswell, is SGP = 0.334 N m /kg COD reduced at
STP (61). The values quoted in literature differ to a great extent (see Table
10). In this study as a rule, the values of SGP were significantly larger
than the values of SGPQ and correspondingly, the values SGP > SGPa.
87
-------
TABLE 10. RESULTS OF ANFLOW REACTOR PERFORMANCE
CO
CO
Retention
Time In Days
HRT
1.5
1.75
2
3
4
5
6
6.7
8
9
10
12.5
15
17.5
20
25
30
Concentration
pH
6.3
6.4
6.5
6.6
6.3
6.9
7.2
7.5
7.6
7.7
7.7
7.3
7.9
7:9
8.0
8.0
8.2
CH4 (%).__
25
26.1
28
30.6
34.6
38
47
50.1
52.5
52.7
56.5
58.2
56.8
60.6
62.5
68.6
68.8
COD , ng/dm
So
14,370
14,370
12,870
12,870
12,370
12,870
12,730
12,570
12,870
12,730
12,570
12,730
12,670
12,730
12,570
12,730
12,570
S£
11,530
11,220
10,520
10,090
8,410
8,180
7,430
7,620
6,810
6,290
6,560
6,170
5,300
5,320
4,230
3,480
2,640
BOD5 nf rag/dm
So
5750
5750
5160
5160
5160
5160
5090
5010
5160
5090
5010
5090
5010
5090
5010
5090
5010
Se
3990
3940
3550
3370
2800
2720
2640
2380
2110
1970
2060
1870
1500
1690
1320 .
860
610
TOC rag/dm3
S
o
4500
4500
3900
3900
3900
3900
3650
3450
3900
3650
3450
3650
3450
3650
3450
3650
3450
Se
4180
4028
3530
3150
2800
2910
2320
1550
1570
1570
1470
1680
179C
1730
2050
2080
2250
SOC mg/ilra3
So
1500
1500
1200
1200
1200
1200
1150
1075
1200
1150
1075
1150
1075
1150
1080
1150
1075
S
e
1220
1170
1150
900
840
1030
470
300
250
260
250
240
260
200
220
200
210
TVS mn/dn>3
So
5200
5199
4631
4630
4630
4630
4310
4500
4630
4310
4500
4310
4500
4310
4503
4310
4500
Se
4760
4664
4045
3620
3100
3015
2730
2155
1845
2150
2330
2500 .
2590
2720
2620
2770
2600
VSS
-tank X
2510
2566
2274
2400
2500
2520
2410
' 2410
2740
2210
2030
1970
1940
2120
1960
1930
1780
TSS
-tank X
3760
3870
3470
3495
3700
3820
3490
3610
4030
3450
3D30
2910
2940
3340
3040
3830
2710
VSSend
-tank Xve
1630
1590
1025
1480
1080
1355
1410
910
1670
1450
1400
1340
. 1170
1250
1330
1310
910
*£ach number is an aritliraetic average of the whole run where several determinations were tjiade.
-------
An attempt was made to evaluate the viable or active cell content in the
digester. Substrate addition was halted and digestion was continued until
there was no more gas produced. Then it was assumed that the VSS left were
equivalent to the active biomass X . The ratio of the initial VSS - X to X
a v a
varied fairly narrowly and as the average equalled 0.55, the value used was
found as:
X = X .0.55 (22)
a v
Results
Table 10 lists major analytically obtained parameters, without showing
the calculated values of specific gas production and removals.
The analysis of the correlations between COD, BOD, TOC, SOC, and TVS
showed that, based on Figures 32 and 33, there is a nonbiodegradable COD ,. of
3
approximately 1200 mg 0_/dm , and that organic carbon data is not reliable as
a straight line should have been obtained (Figure 32). Analysis of TVS bio-
degradability in Figure 33 shows that only TVS (1-R) = BVS =0.49 TVS are
biodegradable, and that there is a limit to COD . removal of some 80 percent.
n, f
The course of anaerobic degradation in the ANFLOW without recycle is
depicted in Figure 34. It follows that the rate of effluent nonfiltered COD
concentration decrease changes or slows down at t = 6 to 8 days at which point
the content of the VSS is at a maximum mode at 8 days. The content of methane,
which is an indicator of the degree of conversion of carbonaceous compounds,
increases much faster than in data reported by Andrews (59).
3
The GP has a maximum at 440 cm /day. This is equivalent to a produc-
avi?3 3 33
tion of 0.44 m /m /d at STP since the digesting volume was 1 dm = 0.001 m
(see Figure 35). The maximum occurred at t = 8 d and then GP started
avg
to decrease rapidly, while its calorific value, expressed as methane content
increased steadily from 25 percent at 1.5 days through 53 percent at 8 days
and to a maximum of 68 percent at a detention of 30 days. Other workers have
noted higher CH, contents and in Figure 35, it is shown that the conversion
of COD into methane continues beyond HRT = 30 days.
89
-------
**>
(N
0
5
A -
3 -
•n" 2 h
1 -
O
V •
BOD5 nf =0.42 COD* 1.2
O.A2
^
2 4 6 8 10 12 U
S0,Se-CODnif (g02/dm3)
Figure 32. Carbon and BOD content, versus COD.
90
-------
00
1.0
0.8
0.6
0.4
0.2
t>12.5 d
a) TVS
bJCODnf
30 60 90 120 0 10
1/(S0-HRT l-
20 30 40
Figure 33. Anaerobic biodegradability of screened piggery wastes: (A) Based on total volatile
solids; (B) Based on COD ..
nr
-------
70
60
50 f*
•40
LU
t-
o
LD
X
-30
5 10 15 20 25 30
HYDRAULIC RESIDENCE TIME - HRT = t h ( d )
L20
Figure 34. VSS content and removal of COD in ANFLOW reactor.
92
-------
16 20 24 28
HRT(d )
Figure 35. Gas production in ANFLOW reactor: (A) Overall average;
(B) Specific gas production based on COD input.
93
-------
3 3
Figure 36 illustrates the average methane production in m /m /day (B) at
STP, from the ANFLOW reactors; and, the specific biogas production based on
the introduced load of BVS.
The decrease in SGPa - BVS at HRT = 8 days is interpreted as the comple-
tion of the liquefaction of organics. This is further documented in the A
portion of the same figure which shows the decrease of the BVS concentration
with time and proves that at 8 days, there is an abrupt change in the rate of
BVS decrease, coincidental with change in the rate of pH changes (Figure 35)
and indicative of start of a more balanced system of acidifiers and methanogens.
The data on biogas production from the removed load of organics should
yield an aymptote close to theoretical carbon - methane conversion. Figure 37
shows that after HRT = 8 days, the rate of biogas production, Part A, is
considerably slower, however, only in the SGP based on COD evidences a
plateau like stretch. In the B portion of the same figure, at and beyond 8
days, the conventionally accepted maximum of SGP (CH.) is reached, i.e.,
3 r
0.334 m CH,/kg COD removed at STP, however, at 15 days the curve has a ten-
dency to increase again. Literature perusal shows that other researchers have
3
reported data for SGP (CH,) up to 0.64 m /kg. The error cannot be excluded
from pur studies, however, since the four data points for HRT, 17.5 to 30 days,
are from very small samples collected over a period of time.
Comparing gas production from the unit load of BVS, it appears that the
SGP values are higher than obtained elsewhere, and there is a tendency to
increase beyond 30 days (see Figure 38), which is not observed at data inter-
preted from the overall gas produced per overall load input (Figure 38-B) -
The technical design parameters may be obtained from Figure 39 where the
3
decreasing COD load below 6 kg 02/m /d yields continuous increase in the
treatment efficiency. From the detention time, considerations (Figure 34) and
from the analysis of gas production rates (Figures 35 and 36) at HRT = 8 to
10 days, there is a definite change in process efficiency. At these HRT values,
3
the COD load is equivalent to 1.6 to 1.3 kg COD -/m /d and the expected
94
-------
•o
2.0
-------
o>
12 16 20 24
HRT (d)
28
Figure 37. Gas production from removed BOD i COD-- ANFLOW (Based on
daily loadings). '
96
-------
HRT (d)
12 16 20 24 28
0.5 1.0 1.5
BVS LOAD -I (kg/m3-d)
ZO
Figure 38. Gas production from biodegradable VS (BVS):
loads; (B) Volumetric.
97
(A) Based on daily
-------
24 6 8 10
LQ-CODnf LOAD (kg02 /m3-d )
Figure 39. COD, BOD and SOC removal efficiency versus COD load in ANFLOW;
unit CH^ production from COD .
98
-------
removals will be approximately: 5Q percent COD f, 50 percent TVS, 60 percent
BODC ,., 58 percent TOG and 78 percent SOC; at the influent concentrations
5,nr - 2
(mean from 85 data points): COD * 12,917 mg 0,/dm , TVS = 4,566 mg/dm ,
3 ° 3 3
BOD5 nf = 5'164 m8 °2/dm» TOC ~ 3>764 m8 C/dm and SOC = 1,184 mg C/dm .
The removals cited above for HRT ^ 8 to 10 days refer to the BOD_ loading
of 0.65 kg 02/m3/d and TVS load of 0.6 kg/m3/d or BVS of 0.3 kg/m3/day.
Since the wastes in this study are screened and sieved, thus considerable
differences are noted when comparing this with other data. It is seen in
Figure 40 where COD versus TVS concentration yields a relationship:
COD = M . TVS = 2.66 TVS (23)
while e.g., Morris et al. (80) have found for cow manure M = 1.43 g COD/g
TVS.
Kinetics of growth and kinetics of removal—
Various interpretation methods are used, all empirical in nature. A
simple and effective one has been suggested by Morris, Jewell and Loehr (80)
as illustrated in Figure 40. The model is a hyperbolic relationship between
time and fraction remaining of TVS. After transformation, it is R as the
nonbiodegradable fraction = 0.51 g TVS/g TVS:
(TVS) Se/SQ = R + II k . HRT (24)
and the plot yields k = 0.89 d for this study, which compares well with
0.85 d obtained by Morris et al. although the correlation is not satisfying.
Subjecting the data to the reaction order test, as used by Levenspiel
(63), it is evident that the anaerobic degradation cannot be described by the
zero or first order reactions with the active biomass (Figure 41-A). The
value used for the active biomass concentration X = 0.55 X where 0.55 is the
a v
average ratio of VSS at the end to the initial VSS solids at the start of
e o
batch digestion without substrate addition which completed each series at a
given HRT. Average values of the dimensionless ratio is used rather than
individual values which ranged from 0.45 to 0.65.
Figure 41-B shows first order kinetic, simplified from Michaelis kinetics
for K > > S where the fit yields a reasonably straight line with slope
99
-------
AVERAGED TVS (10mg/dm3)
2345
T3
IS)
>
co
CO
t>6d|
ft."". •
I
2 3
TVS-S0/HRT(gTVS/dm3-d;
Figure 40. (A) COD dependence on TVS; (B) Morris, Jewell, Loehr kinetics
of BVS removal.
100
-------
HRT
o
x
(/I
I
o
to
CO
o
•o
oo
ot
-------
K. = 0.11 d"1. Gaddy et al. (B6)_ have shown, for cattle waste., K^ - 0.125 d~
from a similar plot.
Figure 42-A presents the growth correlations based on the BOD,. - data
from screened wastes according to a standard model, a transformation of the
mass balance equation:
YS
xa • TTTT <25)
or
S /Xa = 1/Y + /b/Y/t (26)
17 3.
where t = HRT. This method of interpretation proposed by Eckenfelder and Ford
(58) yield a two-phase curve, the change of acidic into methanogenic phase
occurring at HRT = 8 days and the coefficients equal to Y(a) = 1 mg VSS pro-
duced per mg BOD5 removed; Y(n) = 0.59 (mg VSS/mg BOD5); b(a) = 0.16 mg VSS
used for endogeneous respiration per mg VSS in the reactor per day; b(m) =
0.11 d"1.
Plotting the model of growth in the form:
1/t = y - b = qY - b = (S /X t)Y - b (27)
L Si
which is identical to the one above, one obtains a straight line as in Figure
42-B from which the yield coefficient Y = 1 and b = 0.11 d~ as for the
methanogen phase in the A portion of the graph. This excessively large value
of Y and b, the values found in literature do not exceed Y = 0.35, b =
0.05 d for anaerobic treatment is most likely the result of errors intro-
duced by the TVS determination, noted by several researchers (87) and
difficulties in determining the exact amount of biodegradable-active fraction
of the biomass.
The applicability of the second order mechanism of growth (Monod empirical
model) and removal (the Michaelis - Briggs - Haldane - MBH model) has been
tested. Both proved inadequate, yielding, as in Figure 43-A, the MBH model,
a curve rather than a straight line. Finally, the first order correlation was
tried (Figure 43-B) :
(Se - Snb) ' (Se ~ Snb) = exp(-kt) (28)
102
-------
o
x
Jp
0,6
0,5
04
2 0.3
0,2
0,1
0,0
bs-on
I
I
I
8 12 16 20
(S0-se)xat-(d-i)
24 28 32
t = HRT(d)
Figure 42. Growth kinetics correlation by two methods of interpretation -
BOD removal - ANFLOW reactor: (A) Eckenf elder and Ford; (B) This
study.
103
-------
REMOVAL RATE MBH FUNCTION
§25
o
en
E
« 20
E
•a
•6
"*" is
o
-i 10
t/>
J?
~ 5
i i i
-
o o
o
0 ,. ds v S
o rit "vmax « +c
0 o QT m
a
00°
rt o t _ Km 1 4 1
sr vmaxSe vmax
"
i i i
°0 1 234
COD 1/S 10"^dm3/mq
0.9
0,8
0,7
0,6
"f 04
^ °'3
^
1/1
01
W ' ' ' '
^ \,D/ -
•|°\NS ;
.yy\. e
LO ^v
^•L Ne
FULLY \
METHANOGENIC ° '
i i i
' 0 10 20 30 40
HRT(d)
Figure 43. Removal kinetics in ANFLOW reactor: (A) Michaelis (MBH) second
order; (B) Phase breakdown first order.
104
-------
which is a plug flow approximation of a series of CMR. The rate coefficient
k = 0.065 d is adequately representing the data for HRT > 10 d. The value
m — 3
of non-biodegradable COD was earlier estimated at 1200 mg 07/dm . It is
L _1
interesting to note that for poultry wastes Yang (82) has obtained k = 0.09 d .
Two distinct process kinetic constants are seen from Figure 43, i.e., for
the acidic phase with k = 0.092 d and the one suggesting a fully developed
a -1
balanced methanogenesis k = 0.065 d .
Finally, plotting the data in the pseudo-first order plot of biodegradable
fraction remaining of COD versus total COD load (L ), function one obtains two
distinct data pools. For retention time HRT > 8 days, the load removal rate
3
coefficient k = 0.158 kg/m /d as shown in Figure 44.
m
Discussion and Conclusions
Anaerobic digestion of piggery effluent in conventional ANFLOW reactors
without recycle has received much attention. The data published is, however,
full of ambiguous figures on gas yields and efficiencies; there also is a
lack of concrete evidence as to the minimum critical TS concentration that can
be effectively digested. Fischer et al. (67) have worked with a 35 C diges-
tor with separate removal of sludge and liquid, thus, it was a semi-contact
3
system in reality. They have found stable operation at 2.3 to 2.8 kg TVS/m
3
with gas production of 0.45 m /kg TVS removed (time unit not specified by
authors). Extremely scattered data for those concentrated wastes contrasts
with data from this project as well as the gas production, which at 15 days,
3
amounted to 1.4 m /kg TVS removed (see Figure 35-A). Similarly Morris et
3
al. (68) cites Loehr's data for swine wastes as SGP = 0.434 m /kg TVS intro-
a
duced. Our data shows gas production to reach a steady state level at SGP =
0.85 m /kg TVS introduced (Figure 36-C) Kroeker et al. (69) in his studies
3
of ANFLOW with swine wastes loadings of 2 to 1 kg TVS/m /d obtained SGP =
3 3 °
0.68 to 0.82 m /kg TVS intr. and SGPe = 1.98 to 1.62 m /kg TVS destroyed at
SRT =15 and 30 days.
The range of gas production reported in some references is depicted in
3
Table 11. Larger H^ production values than 0.334 m /kg CODremoved reported as
105
-------
.Q
ooc
co
0,8
0,6
Q5
0,4
0.3
CO
CO
,* 02
0,1
Km = 0,158
Q5
= 1200mg0/dm3
1,0 1,5 2,0 2,5
LOAD'1 1/L(m3/d/kgCOD)
Figure 44. Pseudo-first order correlation of COD load versus remaining
fraction of biodegradable COD.
106
-------
TABLE 11. COMPARISON OF SPECIFIC GAS PRODUCTION SGP RATES
AS QUOTED BY DIFFERENT SOURCES
Type of Wastes
(Author) :
"Theoretical"
maximum
Poultry manure
Poultry manure
Poultry manure
MPS various
sources
Fruit slops
Molasses wastes
Dairy cow
Dairy cow
Industr. sanit.
sludge
Various fruit-
veget. wastes
SWINE WASTE
Fischer, et al.
Fischer, et al.
Solly
Longsdom
Overcash, et al.
Sparling
Pipyn, et al.
Kroeker, et al.
Gas
M
M
M
B
B
B
B
M
M
M
M
B
M
B
B
M
B
M
B
SGP
0.334
0.66
0.7 (m)
0.93
0.5-1.8
0.47
0.50
0.77
0.68-1
0.3-0.64
0.2-0.42
0.6-1.8
0.77
0.4-0.6
0.6
0.31
0.5-1
0.3-0.65
0.68-0.82
^nits
m /kg of
COD
TVS
TVS
o
TVS
o
TVS
CODr
CODr
TVSr
TVS
COD
TVS
o
TVS
r
TVS
COD
o
COD
o
COD
0
TVSr
CODr
TVS
o
Reference
Notes Number
61
3 - 7% TS 75
74
74
85
72
72
60% 84
3 - 7% TS 75
73
HRT=32d,-0.7-1.6 155
kg TVS/m /d
L=2.4-3 TVS 67
Calcul.: 65% M 67
Calcul. 77 •
Calcul. 79
Calcul. 76
57 - 60% M 78
2-5% TS, t=10-30d 73
SRT = 15.30d 69
(Continued)
107
-------
Table 11 Cont.
Type of Wastes
(Author) :
Kroeker, et al.
Summers, et al.
Morris, et al.
Ifeadi, et al.
MPS guide
Scharer, et al.
Thomas, et al.
Thomas, et al.
Knol, et al.
Ngian, et al.
THIS STUDY
ANFLOW
HRT = 9 d
L-1,4 kg COD/
in /d
Gas
B
B
B
B
B
M
M
M
M
M
M
B
B
B
B
M
B
SGP
1.62
0.30
0.434
0.5-1.3
0.75
(0.62-0.94)
0.54
0.25
0.33
0.26
0.334
0.88
0.60
1.2
0.3
0.76
0.45
gnits
m /kg of
TVSr
TS
o
TVS
o
COD
o
TVSr
TVS
o
TVS
o
TVS
o
TVS
o
TVS
o
COD
TVS
o
COD
BOD,
5,r
COD
o
TVS
_
Reference
Notes Number
L=1.5 kg TVS/m3/d 69
6% TS raw 70
68
Calcul. 83
20°C, 12.5 d 85
Value in m3/m3/d 153
HRT 8.5 d; 79% M 154
L=1.25 kg TVS/m /d
HRT 2.6 d; 66%.M 154
L=5.6 kg TVS/m /d
HRT 32 d; 74% M 155
L=l kg TVS/m /d
Batch, 38°C 156
L=l kg COD/m3/d
L=0.5 kg TVS/m3/d
L=0.6 kg BOD/m3/d
SGP* =0.35 m3/kg
SGP*
3 3
value in m /m /d
m - minimum value is quoted
B - Biogas
M - Methane
Subscripts "r" - removed or destroyed; "o" - introduced "Calcul." denotes the
use of 400 g COD/hog/d; at 100 kg live weight. Temperature is 35°C in all
cases, except when specified.
108
-------
maximum by McCarty et al. have b,een reported by Pypln and Verstraete (73)
who attained 0.3 to Q.65 m CH4/kg caDremaVed for plg ^rastes>
It follows from the analysis of this data that gas production, although a
reliable control parameter, should be carefully evaluated and procedures
leading to the values quoted should be outlined.
The conclusions from this work may be itemized:
1. Flow-through digestion in ANFLOW without sludge recycle is feasible
2
at very low concentration of TVS (0.5 percent) and COD (1.2 g/dm );
thus, the lower limit for optimized digestion set by various
researches as 2 percent TS is proven to be invalid.
2. Minimum required time for methanogenic digestion to develop is
8 days HRT.
3. Liquefaction is virtually completed at HRT = 8 days.
4. Maximum methane production is not coincidental with optimum reten-
tion for removal (i.e., 8 days), thus, sludge recycle is mandatory
if one is to increase the reseeding and decrease the washout of
methanogens and get lower values of HRT.
5. The method of seeding the digesters with acclimated digesting
municipal sludge was applied successfully but long acclimation
periods were necessary, over four months.
6 . Lack of kinetic data evidenced in literature is the result of inade-
quate data base and perhaps also the inapplicability of the second
order (hyperbolic) Monod or MBH kinetic models.
7. Piggery wastes are found to be easily biodegradable, with optimized
conditions for biogas production and removal occuring at HRT _>_
8 days.
8. The volatile suspended solids (MLVSS) content increases to a certain
3
level and assumes steady state at t ^ 10 days at VSS = 1750 mg/dm .
9. The parameter of TVS is inadequate due to analytical errors inherent
in the standard test, the 600 C combustion includes some non-
biodegradables .
109
-------
10. The results indicate the need for reliable re-evaluation of the
theoretical and practical approach to gas production and for the
establishment of criteria for evaluation of anaerobic digestion
process performance.
11. Errors in evaluating SGP based on daily readings may be due to
erratic daily GP; periods of low or inhibited gas generation lasted
for more than 24 to 36 hours and were followed by intensive produc-
tion; an average value then represented a better estimate, this
was despite of mixing the ANFLOW contents.
ANAEROBIC DIGESTION IN ANCONT REACTORS
The anaerobic digestion studies in the ANFLOW reactor have shown feasibil-
ity at the hydraulic retention time, HRT _> 8 to 10 days. The decrease in the
volume of the digestor can be made only through retention of sludge in the
sludge recycle system similar to an activated sludge system. The ANCONT
reactor designed in the project serves the purpose of complete mixing and
sludge recycle, and was used to study the possibility of decreasing the
HRT below 8 days.
Methods
Four parallel ANCONT reactors were used in the study. Each reactor, shown
in Figure 45, was fed screened (15) raw pig wastes by a separate peristaltic
pump (11) from tank (14) mixed through recycle (13) through conduit (12) and
the feed pipe (6). The raw substrate enters the recycled sludge conduit and
through injector (8) and recycle pipe (9) enters the central cylindrical
completely mixed part of the reactor (2). After the contact with fresh sludge,
the wastes enter concentric clarifier (3) and leaves the strictly anaerobic
reactor through siphon (7) and (17). The siphon serves the purpose of level
control and water seal. The treated wastes were collected in tank (16) and
analyzed daily. The whole ANCONT (1) is made of plexiglass and placed in water
bath (4), the temperature was controlled by thermoregulator (22), through
electric heaters (21) with contact thermometers in the bath (2) and in the
reactor (19). Gas was collected in a gas tank (5) with water seal (28) (NaCl
brine in 15 percent ^SO^). The gas tank is equipped with counterweight (27),
manometer (26), thermometer (25), and gas valves (23) for collecting gas
110
-------
19 20 23 22 21 5 25 23 26 28 27
Figure 45. Layout of the ANCONT anaerobic digester.
Ill
-------
samples and venting. Mixed liquor samples were collected through valve (24).
Reactor mixing is induced by gas compressor (10) through pipe (18) and jet
mixer (8) as shown in Figure 45.
3
The overall volume of the reactor is 5.10 dm of which approximately 35
percent is the volume of the gas mixed reactor. Since sludge wasting was not
practiced, the total volume was used in the calculation, since removal and
gasification occurred in the settling compartment as well. The only sludge
evacuation mechanism was the effluent suspended solids.
Twelve runs were made. Each run encompassed, on the average, 14 indivi-
dual analytical series performed every second day. Each series usually
included analyses of both influent and effluent BOD5 (nf, f), COD (nf, f), TS,
TVS, NH*, TKN, N-NO~, N0~, and pH. Additionally, MLSS and MLVSS and ANCONT
pH were determined every fourth day, the sample was collected from valve (24)
(Figure 45). The foregoing analysis is based on approximately 3000 individual
data points. Arithmetic averages are reported in Table 12.
The experiment was started with HRT equal to 7.4 days, and the HRT was
gradually lowered down. This report is written with data available for HRT
as low as 0.5 days. First results of the tests with HRT = 0.25 days and 0.125
days indicate feasibility of methanogenesis, however, they could not be
included in the report due to time constraints. In the continuation of this
work (Project JB-5-534-7, reference 169) phase separation and multi-stage
anaerobic processes will be further evaluated.
Results and Discussion
The ANCONT reactors were operated without intended sludge wasting;
sludge carryover was responsible for partial evacuation of excess sludge.
The concentration of MLSS (X) and MLVSS (X ) has gradually increased with the
increased organic loading and decreasing HRT. The value of X has doubled
3 v
from 15.4 to 34.6 g TVS/m , with HRT decreasing from 7 to 0.5 d, while the
3
values of volumetric loading increased from 1.51 to 22.09 kg TVS/m /d and 1.5
o
to 28.65 kg COD /mj/d. This indicated excellent liquifying potential of the
ANCONT reactor even at very low retention times.
112
-------
TABLE 12. SUMMARY OF THE ANCONT PERFORMANCE DATA*
HRT days
Parameter Unit 7.4 6.1 5.2 ^ 4.0 3.0 2.0 1.76 1.54 1.26 1.0 0.75 0.50
1 2 34~ 56 7 8 910 11 12
SRT d 36.8 30.6 25.1 20.2 16.1 13.2 12.4 11.0 9.9 8.5 7.2 5.0
pH - 7.3 7.2 7.3 7.2 7.2 7.2 7.1 7.1 7.0 6.9 6.9 6.8
D ,-S
nf o
COD -S g/dm3 14.27 14.64 14.27 .14.64 14.27 14.64 14.48 14.48 14.48 14.23 14.36 14.33
COD ,-S g/dm3 1.87 1.74 1.99 1.89 2.24 2.62 3.06 3.07 3.33 3.68 4.31 5.74
nf e
TS - SQ g/dm3 10.98 12.05 10.98 12.05 13.71 12.05 11.37 11.37 11.37 11.05 11.10 11.05
TS - Se g/dm3 5.79 6.30 6.65 6.78 6.47 6.74 5.73 7.09 6.74 6.56 6.24 6.17
Xv g/dm3 15.40 15.45 18.94 19.06 18.00 21.30 24,90 29.14 30.73 33.85 32.90 . 34.60
L-CODnf kg/m3d 1.96 2.40 2.76 3.61 4.71 7.21 8.23 9.41 11.59 14.22 19.17 28.65
GP ' dm3/d 1.90 2.28 2.48 '2.90 3.35 4.07 5.11 5.62 6.22 7.06 8.02 9.21
CH4 % 72.0 72.0 72.0 71.4 75.2 75.8 73.0 75.0 75.0 73.6 75.8 76.7
*Each number is the arithmetic average of 15 separate determinations which made
up one run at a predetermined hydraulic retention time.
-------
3
The laboratory scale ANCONT system with 5.1 dm active volume required
large volumes of wastes for feed at low HRT. The system was stable even at
low retention times since sludge recycle allowed for much faster recovery
from upsets and failures than in the ANFLOW reactor. As found by other
authors, e.g. Lin Chou, Speece and Siddiqui (170), the increase in SRT over
HRT has created an inherently stable system as compared to the ANFLOW type
reactor.
The increase of SRT did not effect directly the organics removal effi-
ciency. The value of SRT was directly related to HRT and exerted a much
different effect on removal and gasification efficiency than the organics
loading, the primary and direct factor of influence (Figure 46-A). Similar
conclusions are drawn by Carr and O'Donnell (172), who have also found that the
viability or bacterial activity of the digester solids is not directly linked
to the TS concentration. In fact in their work, COD was better removed at
lower TS, although the solids served as an excellent buffer against shocks
and sudden load changes.
The initial concentrations of carbon and nitrogen in this study were
3 3
4600 mg/dm and 830 mg/dm , respectively, a ratio of C:N = 5.54. Several
authors indicate the need for maintaining the optimum ratio of at least 16:1
(173). On the other hand, recent studies of Sroczynski and Kokuszko (171)
show that the differences in fermentation efficiency at optimum C:N (their
optimum was found experimentally as 11.5:1) compared to the ratio they used
to operate their system C:N = 6.5:1 was only 3 to 4 percent, expressed in
terms of COD ,. removal.
n±
The first graphical correlation from the data in Table 12 is presented
in Figure 47 and shows effects of increased HRT on gas production, effluent
quality and pH in the ANCONT reactor. It is characteristic that gas produc-
tion from the unit ANCONT volume as well as the pH curves show abrupt change
at HRT = 2 days, being formed of two straight lines of the y = ax + b type.
This finding may be compared with data from the ANFLOW studies where the
linear increase in pH slowed abrubtly at approximately HRT = SRT = 8 to 10
days. In the ANCONT study HRT = 2 days corresponds to SRT = 13 days. The
correlation of pH versus SRT is presented in Figure 48-B.
114
-------
I
*•—
c
Q
O
LLJ
QL
ro
E
LU
ID
90
80
70
60
50
.£" 4
o
a
<-> o
30 gj
LU
20 a
CO
10
6 7
HRT(d)
8 12 16 20 24 28
LOAD CODnf (kg/m3 d)
73
71
69
67
65
Figure 46. Effects of hydraulic retention and organic loading on SRT, ANCONT
pH and effluent quality.
115
-------
12
T3
CO
_£
CO
|,s
o
CO
£
3,0
2
s 2p
o
o
^
u.
10
0
7,5
x
Q.
O
0
7,0
6.5
\
S =532(CODf)
o r
INFLUENT pH= 6,4 -7,1
0 1
5 6
HRT(d)
Figure 47. Effects of hydraulic retention on: (A) Gas production; (B) Efflu-
ent; (C) pH in the ANCONT reactor.
116
-------
o:
co
100-1
75-
50-
co 25'
>
0J
0,2
0,16
0.12
0,08
0.04
0
b
15
CO
.£
r 1125
-------
Gas production changes, linearly with HRT until HRT = 2 days; then
after an abrupt change in rate (see Figure 47-A), it decreases linearly
following equation:
GP = 0.62 - 0.051 HRT; for HRT >_ 2 days (29)
The decrease of gas production with HRT is the result of decrease of
organic loading introduced, which has a more profound influence than the
increase of unit utilization of organic matter in the system, i.e., the more
complete fertmentation at higher HRT.
Effluent concentration of organics evidences a hyperbolic decrease with
HRT (Figure 47-C), similar to removal efficiency as shown in Figure 46-A
which stabilizes at 87 percent COD , at HRT ^.3.5 days. The HRT in ANCONT
should, however, be regarded as a supplementary design factor for volume
determinations, since the major effects on effluent quality and gas produc-
tion is exerted by organic loading. All subsequent correlations are against
volumetric loading and/or SRT.
Figure 46-B illustrates a linear increase in effluent quality expressed
in COD ,. with volumetric COD ,. load:
nf nf
S (COD ,) = 1.5 + 0.151 L (30)
e nf
Adequate accuracy is attained in this graph, however, one should bear in mind
the fact that the concentration of biological solids in the reactor X (MLVSS)
increases with the increase of L and that the removal efficiency is in fact
affected by the food to microorganisms ratio (F/M) as in case of activated
sludge. Similar accuracy is attained for COD f removal correlation in Figure
49-C:
CODnf(%) = 89 - 1.01 L (31)
The curve for effluent CODf versus L (COD .) yields two straight sections;
for L £ 4 kg/m3/d:
Se(CODf) = 1.1 + 0.0618 L (32)
Beyond the inflexion point (L < 4 kg COD _/m /d), the line tends to go to
zero which could indicate that almost all of soluble COD is biodegradable.
118
-------
E
QI
L.
C
Q
O
O
J?
CO
E
CL
O
CO
U/3
0,20
i
oo 015
0
O
03 0,10
0,20
0,15
~r
5 0,10
009
0,08
nr»7
^j • i • i • i • i i i i i i
°***°**0 x-v
^^*o-***».^ Q\)
9= -1,77 ] -^Q^
O ^^^^"^i
^^^^o
*0.
®"*rt
-o
m = -275 T**^*1*^*,^
' ' ^^^^^^^^ Q
0 "^"O^^Q
^*""**°**^^^ "
^^*^o
1 1 1 . 1 . 1 . 1 1 1 1 1 1
c
Q
O
O
i
90
80
70
60
4 5 6 7 8 9 10 20
L(kg CODnf/m3 d
30
50
ro
E
2 —
**_
Q
O
O
00
12 16 20 24 28 30
LtkgCOD
Figure 49. Gas production from removed COD and efficiency of COD _ removal
in ANCONT reactor.
119
-------
From Figure 46-B, it can be incurred that the nonhiodegradable COD is in the
solid, non-filterable form and is equal to 1500, mg O./dm .
Figure 50 illustrates biogas production from the introduced COD f versus
load, in arithmetic and log x log scales, indicating good approximation of
the hyperbolic equation:
(COD ,) SGP = L"1'44 (33)
nr o
Similarly in Figure 49-A and -B, very good fit is obtained for biogas and
methane production from removed COD _:
(COD ,) SGP = L""1*77; biogas (34)
nr r
(COD -) SGP - L~2'75; methane (35)
nr r
Comparing the efficiency of fermentation versus SRT with correlations
versus COD load or HRT, it is evidenced in Figure 48-B that SRT exerts,
similar to HRT, curvilinear effect on effluent quality expressed here as
BOD,. ... It follows from this graph that the fermentation of TVS, expressed as
Jȣ
removal efficiency, stays steady at approximately 50 percent at SRT 5 to 40 days.
Kinetics—
The kinetics of biological growth from unit of removed substrate is
presented in Figure 48-A. Equation 27 was applied, where the value of
q = (S - S )/(X .HRT) (36)
O 6 3
The plot for COD f and BOD f data, with assumed active (viable) biomass
concentration X = 0.55 X /(equation 22) yielded the values of biomass yield
coefficients Y(COD) = 0.213, Y(BOD) = 0.625 and the decay coefficients,
b(COD) = 0.016 d~ , b(BOD) =» 0.026 d~ . These are lower values than Y = 1,
obtained from the ANFLOW study, however, errors are present here too, due to
the already mentioned problems with X , X determinations.
Correlation by means of the Eckenfelder-Ford method (Equation 26) yielded
values of Y » 1. This could indicate inapplicability of standard biomass
growth models to studies where the active (viable) biomass is not determined
accurately and where there is little control on sludge recycle (sludge age).
120
-------
g J
o
^ "£
§ o
cr p>
Q. <->
00 QL°
< O
O CO
o
CD
018
0.14
010
0-06
04
0,3
0.2
0,1
009
' 0,08
0.07
0,06
0,05
0,04
0,03
002
48 12 16 20 24 28
L(kgCODnf/m3d)
g=-1(44
2 3 4 5 6 8 10 20 30 40
L(kgCODnf /m3 d )
Figure 50. Kinetics of biogas production from the COD - introduced
nf
121
-------
Finally, substrate removal kinetics was evaluated graphically, according
to equation 1 modified for constant S :
a
(37)
Figure 51-A and -B shows this correlation for COD - and BOD, ,, respectively.
_i n -}>^
The plot for COD - yielded K = 0.09 d~ . The values quoted by Oleszkiewicz,
Koziarski, et al. (56) for various full scale activated sludge plants were
K = 1.70 to 1.75 d~ . The value for BOD,. , in ANCONT reactor was K = 0.22 d~ ,
J»i
while the aerobic activated sludge system removed BOD- , with a rate coef-
-1
ficient of K = 3.35 d (56). Based on the theory of biological treatment,
it is known that aerobic processes are many times faster. Here the removal
rate for activated sludge is 15 to 30 times higher than anaerobic contact
digestion. However, due to technical constraints and other parameters imposed
on aerobic systems such as permissible organic loading L and F/M, the resulting
ANCONT volume should exceed the volume of an activated sludge tank (with the
clarifier) only by a factor of three to four.
Discussion and Conclusions
The ANCONT reactor studies have shown that it is possible to decrease the
anaerobic digester volume down to 3 to 4 days HRT. The reactor COD - loading
3
at 4 days HRT can be as high as 4 kg COD f/m /d, yielding constantly, removals
equal to or better than 85 percent COD f> and the following effluent qualities
nr 2
(averages of 15 steady state series) - values in mg/dm : COD , = 1900; COD, =
1200; BOD. , = 700; BOD, , = 500; and TVS = 4100.
j,nt j,r
The gas has a very steady methane concentration of 72 to 76 percent. The
average gas production from the unit volume of the reactor decreases with the
increasing retention time, and decreasing organic loading. At the design load
3 33
of 4 kg COD ,/m /d, the gas production is GP = 3.15 m /m /d (75 percent CH, ) .
The unit or specific gas production from the introduced SGP at L = 4 kg COD ,/m /d
equals: SGP - COD , = 0.18 m3/kg; SGP - BOD, , = 0.35 m3/kg; SGP - TVS =
, o nf o 5,nf o
0.25 m /kg.
The respective specific biogas production from the unit of SGP , at L =
4 kg COD /m3/d equal to 0.17 m3/kg CODnf, 0.51 m3/kg CODf, 0.41 m /kg
122
-------
0,5
04
03
Q2
0,1
020
015
00
I
o
00
005
K = 0,09 d
r1
1 2 3
EFFLUENT CODnf
4 5
Se (g/dm3)
K=022 d
' L
02 04
06
08 10
12
EFFLUENT BOD5f Se(g/dm3)
Figure 51. Kinetics of BODg - and COD f removal in ANCONT reactor.
123
-------
3 33
0.74 m /kg BOD,. f, 0.34 m /kg TS, and 0.5 m /kg TVS., These, yalues correspond
33
to an approximate production of 2.20 m b.iogas/m piggery wastes (75 percent)
at HRT = 3.5 d and L « 4 kg CODnf/m3/d or 3.30 kg TVS/m3/d or 3.8 kg TS/m3/day.
These values are much lower than the values attained in the flow-through
DW reactors (HRT = SRT) where SGP-COD = (
nr
at an optimum retention of HRT - 8 to 10 days.
ANFLOW reactors (HRT = SRT) where SGP-CODnf = 0.4 to 0.5 m3/kg (SGPQ-0.3-036),
l£ is concluded that the saving in construction costs of the volume of the
ANCONT anaerobic digester is balanced by the lower gas production. The removal
efficiencies attained in ANCONT reactors are much higher than those attained in
3
the ANFLOW units. For the corresponding COD - load of 4 kg/m /d, the ANFLOW
yields some 26 percent while ANCONT 85 percent COD removal.
In both studies nonbiodegradable COD appeared to be equal to 1500 to
3
mg/dm and \
organic matter).
3
2000 mg/dm and was in the non-filtered form (suspended solids and dispersed
As in other anaerobic reactors, in this study the system with wastewaters
exhibited good buffering capacity, the ANCONT pH was practically stabilized
beyond 2 d HRT (SRT = 13d) at 7.2 to 7.3.
ANAEROBIC BIOFILTRATION IN ANBIOF REACTORS
Anaerobic digestion with suspended cultures of dilute wastewaters, i.e.
with water content exceeding 98 percent, is usually difficult to achieve due
to the washout of biota and difficulties in separating the secondary sludge.
Frostell (1979) notes that the doubling time for methanogenic population tends
to increase with the wastes becoming more diluted. Anaerobic biofliters offer
an alternative to the sludge return (contact) system by providing supportive
media for anaerobic organisms and thus, effectively increasing the SRT over the
usually short HRT.
The interest in anaerobic biofiltration has been recently activated by
Frostell (48), Anderson et al. (49), Mosey (50), Genung et al. (51), Mueller
and Mancini (88) and others, however, these authors used synthetic soluble wastes
124
-------
or effluents otherwise totally different from animal wastes, the latter were
never before treated in heterogeneous-fixed film reactors.
The aim of this study was to define the feasibility of attaining high
removal efficiency at low retention time and at low temperatures and high
organic loadings in anaerobic biofilters (ANBIOF) treating raw-screened
piggery wastes. It is expected to develop a treatment system characterized
by lower initial costs and significantly reduced power requirements.
Methods
The experiments were conducted in laboratory conditions, at 23 to 26 C,
in a setup as in Figure 52. Three filters were used in parallel. The ANBIOF
1 and 2 had ID 0.12 m bed height H = 1.90 m (liquid height 2.10 m) and overall
volume V = 0.023 m3, while ANBIOF 3 had ID = 0.010 m, H = 1.35 m (1.57 m
3
liquid), and V = 0.011 m . The specific surface area of the media, which
2 3
consisted of 20 mm expanded polyethylene spheres, was 100 m /m . Gas production
was measured in brine filtered gas tanks since the methods of measuring through
meters were discredited earlier, as unreliable for small gas volumes. The
large gas tanks have also served as an equalization tank.
The studies were initiated after stable effluent quality was attained and
after a time equal to five HRT had elapsed. The anaerobic processes were seeded
several times with inocculum from batch digestion and contact digestion units.
Analytical determinations included TS, TSS, VSS, DS, pH, alkalinity, nitrogen:
N-NH., N , COD-, COD .., BODC .., and BODC -. All results in Table 13 are the
"t org I nr 3,1 D,nt
average of ten determinations. Since each run encompassed some 200 individual
analyses (10 analyses of 20 parameters), the foregoing analysis is based on
1800 data points. The samples were 24 hr average composites.
Gas analysis (Orsat method) included CH,, C0_, CO, C H , H0, H0S, N0, and
4 / n m 2. 2. 2.
02 as a control for airtightness of the system. The reading of the accumulated
gas volume was brought to STP conditions knowing the temperature, gas and
barometric pressure.
Results
The studies have proved the expected reliability and efficiency of the
pilei
125
anaerobic biofilters; the data is compiled in Table 13. The COD . removals
nr
-------
raw
wastes
CM
o
o
IT)
microscreen
1.5mm
\
vent
^
thermometer
manometer
A
f-
junterweight
sampling
f
brine filled
gas tank
water seal
anbiof media
effluent collection
feed tanks and pump
Figure 52. Layout of an anaerobic biofilter - ANBIOF - arrangement.
126
-------
TABLE 13. RESULTS OF ANEIOF PERFORMANCE
Loading with
Parameter
HYDRAULIC LOAD
HRT
.SRT
S
o
S
o
S
o
S
S
e
S
e
S
e
S
e
S
e
- (hydraulic)
- (solids)
= COD ,
nf
= CODf
- BOD, ,
5,nf
= VS
= COD -
nf
= CODf
- B°D5,nf
- B°D5,f
= VS
CH^, TOTAL
production
CH,
CONTENT
Unit
3, 3
m /m
hours
days
mg 02
mg 02
mg 02
rag 02
mg 0,,
mg 02
mg 02
mg 02
mg /dm
3, 3
m /m
/d
/dm3
/dm3
/dm3
/dm3
/dm3
/dm3
/dm3
/dm3
3
/d
Percent
0.40
0.04
343
698
9,400
4,130
4,030
6,040
660
300
210
150
1,020
0.009
14.2
0.77
0.08
120
497
9,400
4,130
4,030
6,040
570
320
300
205
800
0.067
62.6
1.75
0.17
59.4
211
10,070
4,120
3,290
4,630
1,080
570
705
330
680
0.103
n.d.*
(70)
Non-Filtered COD
2.40
0.25
38.6
66
9,400
4,130
4,030
6,040
2,130
1,050
710
380
1,210
0.242
76.6
3.50
0.36
25.7
57
9,560
4,070
3,220
4,540
3,070
1,420
1,220
680
1..310
0.366
77.8
kg/m3/d
3.60
0.34
29.2
52
10,590
4,520
3,870
5,670
2,340
1,090
930
540
1,360
0.358
80.4
5.60
0.61
20.7
46
9,610
4,030
3,175
4,560
3,020
1,610
1,120
860
1,040
0.642
n.d.**
(80)
6.55
0.74
13.6
34.6 :
10,590
4,520 ,
3,870
5,670
3,170
1,680
1,405
690
1,630
0.499
80.0
* n.d. (70) means
** n.d. (80) means
"not determined," assumed equal 70 percent.
"not determined," assumed equal 80 percent CH,.
-------
were 70 to 93 percent for HRT - JL3.6 to 343.4 hr; has. met;han.e. content increased
with the decrease of HRT and the increase of the ANB.I.OF organic loading (L).
3 3
Gas production was 0.024 to 0.172 m CH,/kg COD removed and 0.03 to 0.222 m
3
CH,/kg TVS removed, at loadings from 0.4 to 6.5 kg COD/m /d.
The efficiency of COD ,. removal and effects of anaerobic biofilter vol-
umetric loading on the basic process parameters, pH, alkalinity, and methane
content are depicted in Figure 53. Figure 54 shows effluent total and soluble
BOD- (A) and effects of loading on solids removed in Figure 54-B. The BOD- f
removals were 64 to 95 percent yielding effluent quality S - = 1400 to 210
3 e,ni
mg 02/dm .
The biogas composition has changed gradually versus the increasing COD
loading, while the unit gas production has evidenced an increase with the
COD load (Figures 55 and 56). The average daily methane production was 0.024
2
to 0.171 m /kg COD removed at the content of 14 to 80 percent (by volume) in
the biogas. The long term gas production (based on storage in the gas tank)
for the various ANBIOF units (three in parallel were used) showed differences.
Chmielowski's (53) conclusions that the duration of the acid phase of
digestion can be minimized when soluble substrate able to enter cells for
internal gasification to CH, and CO- is introduced, seem to be verified by
this work. The low solids content of wastes at high loadings yielded effi-
cient gasification at HRT. The methane generation rate was the highest at
the lowest retention times equal to 12 to 24 hr, corresponding to the lowest
organics removals.
The maximum methane generation rate observed was equal to only half of
3
the theoretical maximum stoichiometric value of 0.35 m CH,/kg COD removed,
which was easily attained in the anaerobic completely mixed reactors without
sludge recycle. The authors see this as a result of lower temperatures of
the biofiltration and the fact that other organisms are present in the biomass
of the fixed film reactor when compared to classic conditions of the meso-
phillic methane generation at 35 C, in a completely mixed reactor.
128
-------
01
34567
CODnf LOADING - L ( kg02/m3-d )
Figure 53. Effects of organic loading on: (A) COD removal; (B) ANBIOF pH,
methane content and alkalinity.
129
-------
CO
o
L-COD (kg02/m3-d)
L-COD (kg02/m3-d)
Figure 54. Effects of COD loading on: (A) Effluent quality; (B) Solids removal.
nf
-------
— 70 -,
c
-------
200
20
50
t) 1 2 3 4 5 6
COD LOAD -L(kg02/mJ-d)
Figure 56. Effects of COD loading on CH, production from: (A) COD
(B) TVS ; (C) TKN and N-org removals.
rem
rem'
132
-------
It is interesting to note in Figure 55 the presence of nitrogen and
carbon monoxide at very long retention times and low loadings. The air con-
tamination was excluded as oxygen was not found in the gas, and the gas
composition is based on the average of three samples collected independently
and denitrification processes occurring at the highest retention, i.e. at HRT
above Id or SRT above 50d.
The removals of TKN and organic nitrogen were respectively 30 to 50 percent
and 52 to 72 percent (Figure 56-C). The increase of COD loading from 0.4 to
3
3.5 kg/m /d resulted in ammonia nitrogen removal 10 to 25 percent, coupled with
nitrates removal (30 to 50 percent) and significant nitrite concentration
increase (100 to 300 percent).
The studies on the breakdown of the stable digestion process showed that
instability was reached at temperatures below 17 C and at reaction variations
beyond the easily tolerated range of pH = 6 to 8. The optimum process para-
3
meters were L = 2.5 to 4 kg COD/m /d, HRT = 24 hr, SRT above 50 days. The
3
effects to be expected are methane generation rate of 0.160 m /kg COD-removed,
3
some 70 percent COD removal which means S = 1500 mg 02/dm (COD -) for Farm
A wastewaters.
Comparison of results of the direct aerobic treatment in one-stage acti-
vated sludge with the ANBIOF units performance, performed in another phase of
the project, indicates that anaerobiosis breaks down complex and hardly
biodegradable aerobically substances. The lowest concentrations of COD
3
attained in the course of anaerobic biofiltration were 300 to 570 mg 0_/dm ,
3
while activated sludge yielded effluent 460 to 730 mg 0_/dm (based on COD).
It seems that anaerobic treatment is the best method of breaking down complex
substances facilitating at the same time rapid and more complete aerobic
polishing treatment.
Treatment Kinetics
Two first order equations were used to determine the rate of organics
biodegradation:
Se/SQ = exp C-K HRT) (38)
and
133
-------
Se/So = exp (-K SRI) (39}
where S , S are respectively the Influent, effluent concentrations, K is the
rate coefficient; HRT and SRT. The model derived by Oleszkiewicz (27) has also
been tried:
Se/SQ = exp (-k/L) (40)
where L (kg/m /d) is the volumetric organic loading. The latter model yielded
the most satisfactory fit (see Figure 57-A). Two phases of the process are
3
distinguished from this plot. For COD - load above 2.0 kg 0-/m /d, the removal
rate is 3.66 kg/m /d. Below that loading, the rate drops down to 0.42 kg
3
O^/m /d. The two rate constants were identical for the non-filtered and fil-
tered COD data. The respective removal rates (k) expressed per unit area of
2
the media surface (A) are equal to 20.0 and 4.2 g 02/m /d COD (K=kA).
Similarly, the plots of log S /S versus HRT yielded two rates K, = 3.7d~
i e o JL
and K- = 0.21d , the change of rate occurred at HRT = 1.5 days. The log
—1 —1
S /S versus SRT correlation has yielded TO. = 0.17d and K_ = 0.002d , the
rate changed at SRT equal approximately to 65 days (Figure 57-B). Interpre-
tation of the log S /S against HRT correlation of data provided by Young and
McCarty (54) has yielded similar results, however, the rate change occurred
at HRT = 3.5d, probably because wastewaters were more diluted in their study.
Discussion
The results indicate that the ANBIOF system of contact digestion with
attached biological slime is capable of treating piggery effluents with low
solids content. The development of adequate microorganisms composition in
the slime at temperatures of 20 C to 25 C, i.e. lower than the optimum meso-
phyllic range (35 C) can be achieved only after prolonged adaptation. With
large multiple seedings done with actively digesting mesophyllic cultures, the
break-in was 100 to 120 days. Without seeding, the spontaneous optimization
of the bacterial composition took several months.
After attaining stability, the system was resistant to short-term temper-
ature changes within 20 to 25°C, to pH variation within 6 to 8, and organics
concentration variability. It is interesting to note that Mueller and Mancini
(88) in their study of highly loaded anaerobic biofilters have found small
134
-------
0
o
o
V)
1.0
0.8
0.6
0.4
0.2
0.1
0.08
0.06
0.04
i
1.0
0.8
0.6
0.4
K,= 3.7 kg/m3-d
kg/m -d
0.4 0.8 1.2 1.6 2.0 2.4
[m3-d/kgCODnf ]
o
I
(SI
0.2
0.10
0.08
0.06
-1
K, = 0.17 d
-1
K2 = 0.002 d
100 200 300 400 500 600 700
-SRT- (d)
Figure 57. Kinetics of CODf removal in ANBIOF: (A) Pseudo-first order
reaction; (B) Against sludge age SRT.
135
-------
differences in removal rates at temperatures 25 C and 35 C for acldogenic
phase and no difference for methanogenic phase.
3
Due to the alkalinity of piggery wastes equal to 25 to 70 mval/dm , the
ANBIOF units exhibited significant buffering capacity at all loadings, in our
studies.
The studies in this proejct have also indicated that change of substrate
structure occurring during anaerobic degradation makes the process an ideal
alternative to physicochemical pretreatment before aerobic polishing treat-
ment. Lower gas production attained with the anaerobic filtration is offset
by the lower temperature requirements, lower maintenance requirements, and
high removal efficiency.
The data by Frostell (48) and van den Berg and Lenz (139) indicate that
the type of media and mode of application have a small effect on removal ef-
ficiency. These conclusions supply the author's contention (94) that
efficiency of biological removal is dependent primarily on sludge loading.
However, the results of these studies show that in upflow ANBIOF unit most
of the removal takes place in the first 30 percent of the ANBIOF height.
The control sampling revealed that COD . at H = 0.25 m measured from the inlet
3
at the bottom of ANBIOF was close to 200 g/dm due to the fact that there was
a large biomass of suspended microorganisms. The nonfiltered COD gradually
33 3
decreased every 0.5 m to 45 g/dm , 6 g/dm and 2 g/dm (ANBIOF 3) which indi-
cated that the conditions in an upflow ANBIOF resembled more closely the
fluidized bed principle than the classic concept of fixed film reactor. Thus,
the type and the specific surface of the media have less importance than in
aerobic trickling filters, where the hydraulic regime and media porosity decide
whether aerobiosis or anaerobiosis are the dominant organic removal pathways.
Anaerobic conditions may dominate in case of ponding and overgrowth of biolo-
gical slime in aerobic trickling filters.
The kinetics of biochemical changes in the ANBIOF units are best expressed
with a pseudo first order model S^s „ exp C-K/L) or by S /S = exp (-K.SRT).
The efficiency of treatment is dependent primarily on the volumetric loading
136
-------
and influent concentration. The relationship in .Figure 58 shows the response
of anaerobic biofilters to COD loading as reported by different authors for
different effluents, against the data from this study.
Conclusions
The design criteria for anaerobic biofilters treating piggery wastes are
as follows: for SQ (CODnf) = 12,000 mg 02/dm , Se (CODnf) = 1,500 mg 02/dm
at L = 2.5 to 4.0 kg COD /m /d, SRT = 56 d, HRT = Id, and methane generation
3 nl
approximately 0.16 m CH,/kg COD removed. For the studied piggery effluent
some 70 to 90 percent COD f removal can be attained during HRT = 15 to 60 hours,
at 20 to 25 C. The COD . load removal rate expressed per unit area of the
nr 2
media (in the recommended operating range of the ANBIOF) is 20 g 0«/m /d and
3
per volume of ANBIOF 3.66 kg 02/m /d.
Anaerobic biofiltration should be more widely accepted as an alternative
treatment characterized by high efficiency of removal, good organic load shock
tolerance, ease of maintenance, low land surface area requirements, energy
generation potential, ability to operate at temperatures lower than the standard
mesophyllic range, the 20 to 26 C temperatures can always be provided by low
parameter waste heat at any industrial plant. The ANBIOF reactors is particularly
well suited for soluble organic effluents with TS content below 0.5 percent.
In this study the anaerobic biofiltration has been proved applicable to
both roughing and polishing treatment indicating very large flexibility of the
basic operational parameters, i.e. organic and hydraulic loading.
DISCUSSION AND CONCLUSIONS
The various processes studied have shown that the piggery wastewaters are
easily biodegraded anaerobically, with no lower TS concentration limiting the
process performance. The anaerobic process in general, offers the best possible
pretreatment before aerobic polishing treatment and before agricultural utili-
zation. In the latter case there is only 10 percent loss of TKN during anaerobic
digestion and almost complete destruction of pathogenic organisms. The effluent
is almost non-odors. All process modifications have exhibited resistance to
shocks (introduction of air or load variation), the most rapid recovery was
evidenced by the ANCONT reactor and ANBIOF, i.e. was increasing with the increase
of SRT.
137
-------
100
90
80
70
60
50
0= 40
o 30
20
10
SYMBOL
A.
•
O
D
a
RANGE (mg02/dm3)
1500 — 3000
8400
6400 - 12500
10000 - 18000
6300 — 13400
AUTHORS
Young Me Cart y
Plummer et al
Lettinga et al
Mosey
This study
8 10 12 14 16
CODnf LOAD - L ( kg 02/m3-d )
Figure 58. Operational range of ANBIOF reactors.
138
-------
The studies have shown that the volume of the digester cap, be. significantly
reduced by increasing SRT through recycle or retention on the biofilter media.
When one adds to this the value of recovered biogas, increasing with the ongoing
energy crisis, the anaerobic digestion of dilute piggery wastewaters may become
competitive to other methods of treatment.
Further studies are presently being conducted on possible further decrease
of digester volume through phase separation and the increase of the overall
organics removal rate, including nitrogen compounds, by combined anaerobic-
aerobic-anaerobic systems.
Table 14 illustrates the performance parameters obtained in this study
from the three basic anaerobic process modifications. It is interesting to
note that anaerobic biofilter, ANBIOF, is much more efficient than an ANFLOW
reactor and at the same time has a much smaller gas production rate, the latter
increased with the increase of HRT and decrease of SRT. The increase of SRT
has resulted in an increase of methane content in the biogas. Taking into
account the design loads of the three reactors, the required volumes, the
resulting effluent quality and the necessary level of maintenance, it is
apparent that the anaerobic biofilter is the best system for dilute piggery
wastes.
The dilution of wastewaters directly affects the selection of the process
modification and the resulting economic implications. With the most concen-
trated animal wastes (TS above 2 percent) thermophilic digestion (55 to 60 C)
in an ANFLOW type reactor may be feasible since the heat losses decrease with
the decreased dilution, 90 percent of heat is required to heat up the incoming
fresh wastes. In case of effluents with TS ^ 2 mesophyllic digestion (30 to
35°C) in ANFLOW or ANCONT reactors is feasible. Lower concentrations of TS
(below 0.5 to 1 percent) call for ANBIOF, i.e. heterogenous, biofilm reactors
with well developed biomass, operating in the psychrophilic regime (15 to 25 C).
Since gas production in the homogeneous reactors, ANCONT and ANFLOW, is not
completed, they should be followed by an ANBIOF reactor, which will benefit
from the heat contained in the incoming wastes and yield much better gas pro-
duction than those reported in Table 14. Studies, presently run by the authors
(169), on two stage digestion, verify this hypothesis.
139
-------
TABLE 14. COMPARISON OF OPTIMUM DESIGN PARAMETERS AND
YIELDS IN ANAEROBIC REACTORS
Reactor
Parameter
COD - LOAD
nf
HRT
SRT
BIOGAS PRODUCTION*
SGP - COD ,
o nr
SGP - COD -
r nf
SGP - TVS
o
CH4 CONTENT
REMOVAL EFFICIENCY
COD ,
nf
EFFLUENT COD .
nr
BOD. .
5,nf
BOD, .
Units
kg/m3/d
day
day
3
m /kg
o
m /kg
m /kg
Percent
Percent
g/m3
0
g/m
g/m
BATCH
n.d.**
n.d.
n.d.
0.36
n.d.
1.0
n.d.
n.d.
n.d.
n.d.
n.d.
ANFLOW
1.40
9.0
9.0
0.29
0.60
0.76
55.0
63.0
6.5
2.0
ANCONT
4.0
4.0
20.0
0.15
0.17
0.25
75.0
85.0+
1.9
0.9
0.44
ANBIOF
4.0
1.0
50.0
0.13
0.17
0.23
80.0
73.0
2.5
1.1
0.67
RESISTANCE TO SHOCKS
IN OPTIMUM CONDITIONS
EASE OF OPERATION
*The SGP values for ANFLOW reactor are SGP , I.e. are based on the ratio of
total volume of gas produced to the total COD , or TVS load removed or
introduced during the whole experimental run.
**n.d. - not determined.
140
-------
The analysis of the overall design of a piggery wastes treatment plant
revealed that the optimum system should feature high efficiency of wastewater
treatment, sludge stabilization, high specific gas production and constant
generation of anaerobic seed, a measure against sporadic disinfection practices
and occasional antibiotic shocks. The system consists of a thickener which
separates supernatant for anaerobic biofiltration and sludge for anaerobic
digestion in a modified ANFLOW-ANCONT type reactor with additional clarifying
sludge recycling facility and a HRT of 10 to 15 days. The digester receives
sludge from the polishing treatment steps and discharges supernatant to the
ANBIOF reactor system.
141
-------
SECTION 9
PRODUCTION OF YEASTS
INTRODUCTION
Mounting protein deficit and difficulties encountered with conventional
treatment of industrial piggery wastewaters or problems with the year-around
agricultural utilization have directed the attention towards the possibilities
of yeast production on these effluents for use as feed additive in feeding
swine or other animals. Most of work published so far on yeast production from
animal wastes has been confined to cattle wastewaters, Anthony (109), Calvert
(112), and Singh and Anthony (120). The papers on yeast production from swine
wastes show that researchers are looking for methods of enriching this sub-
strate with easily available carbon and for more economical methods of
substrate sterilization (113).
Yeast fermentation for SCP recovery has been extensively used by industry
on such waste products as whey, molasses and sulphite liquors. Wastewaters
from numerous food and organics processing industry plants have been success-
fully fermented by Candida utilis, Saccharomyces cervisiae, Trichoderma viride,
Rhodotorula glutinis and other species in both laboratory and pilot scale by
Tomlinson (122).
Concentrated piggery wastes have been fermented in laboratory, sometimes
with the addition of cracked corn or with addition of other carbohydrate
substrates such as sucrose in fungi fermentation by Aspergillus niger (123).
Up to date there have been no data presented on fermentation of dilute piggery
wastes.
The ultimate goal of the presented research is evaluation of economic
feasibility of producing pig feed supplement consisting of yeast derived from
142
-------
the farm wastewaters through fe.rmentati.on. The. system, could perhaps close
the nutrient cycle in the farm, and produce, high quality protein and low con-
centration liquid effluent for stream disposal, after adequate polishing
treatment.
The present work has been aimed at evaluating:
- feasibility of fermentation of dilute manure;
- eventual need for carbohydrate supplementation;
- selection of the group of the most promising organisms;
- delineation of further pilot scale research priorities;
- testing selected species in semi-dynamic and dynamic studies; and
- evaluating economic efficiency of yeast fermentation as a method of
SCP recovery and wastewater treatment.
The studies were conducted on Farm A raw (screened) wastewaters in batch mode
and in batch-fed semi-dynamic mode as reported exhaustively by the authors
elsewhere (117 and 125).
BATCH STUDIES
Methods and Equipment
Initially, screening tests of a large number of yeast species obtained
from the Institute of Fermentation Industry in Warsaw were performed. The
tests consisted of growing yeasts on agar, solidified piggery wastes in micro-
biological vials. The four species selected from these tests were Candida
tropicalis, strain 11 strain 8, Candida robusta and Candida utilis, strain 3.
These species have proved their growth potential on the manure substrate while
other species failed to reproduce. They are characterized by intensive growth
rate and high protein content of good quality when grown in optimum conditions.
It should be noted that other strains, such as Torula casei, Torulopsis Candida
or Rhodotorula glutinis have not survived the pig manure substrate. It is of
interest also that the four selected species exhibited only 30 percent survival,
as determined by LBffler blue staining vitality test.
Further screening of the initial preparation methods and species was
performed in batch tests. These consisted of various methods of hydrolytic
breakdown of cellulolytic materials to bring up the energetic potential of
143
-------
the piggery wastewater prior to fermentation,. Since, various, authors recommend
different variations of preparatory treatment (124), the following methods of
hydrolysis were tested in the batch mode:
- Sulphuric acid hydrolysis (pH 1.0) at 1.5 atmosphere in an autoclave
for 2 hours;
- Sulphuric acid hydrolysis at atmospheric pressure;
- Heating on boiling water bath for 2 hours; and
- Decanting the manure, evaporation, acid hydrolysis and dilution to
previous concentration.
The preparation of manure through hydrolysis was aimed at increasing the
carbon to nitrogen C/N ratio in respect to easily biodegradable carbohydrate
carbon. In the batch tests that followed, hydrolysis was not used. The reason
is evident when one compares the total sugar after preparatory treatment (Table
15) with raw untreated manure. The raw wastewater concentration of total sugars
3
varied between 200 to 530 mg/dm while hydrolysis by the various methods,
3
outlined above, yielded total sugar content of 230 to 740 mg/dm . The addition
3
of molasses of up to 2 percent boosted total sugars to over 19,000 mg/dm .
Because of these results, all subsequent work was done on untreated raw
piggery wastewaters. As a routine, the wastewaters were brought to pH 5 to
3
5.5 with sodium hydroxide, put into 250 cm beakers, sterilized, seeded with
some 50 cm of monoculture seed grown on brewery mash and incubated for 3 to
5 days in 30°C.
The beakers were aerated with warm air, through sterile cotton air filters,
in thermostatic incubators. If foaming became excessive, a few drops of sterile
soya oil were added (see Figure 59).
After incubation, the samples were centrifuged at 3000 rpm, cake was
weighed, dried, ground, weighed again and analyzed. The weight of the centri-
fuged manure sludge from a sterile reference sample and of the seed was
substracted from the total obtained in order to get the true net microbial
mass yield; supernatant was also analyzed.
144
-------
TABLE 15. RESULTS OF BATCH FERMENTATION OF RAW CENTRATE OF PIGGERY
EFFLUENT BY YEAST OF CANDIDA TYPE
Unit3
Dig/dm
Parameter as
COD 02
TKN N
N-NH3 NH+
N-N02 N02
N-N03 N03
N-Org. N2
K>J P04
Tot. P P
Sugars C6H12°6
Untreated Molasses
Wastes Enriched
A B
4,136 19,666
650 1,348
288 500.
t t (trace)
1.1 16.4
421 848
453 464
966 1,037
500 19,600
C.robusta
A
1,382 5
261
219
t
t
68
303
575
41-7
1
B
,821
496
76
t
t
420
17
78
58
C.tropicalis
A
1,270 7
261
193
t
t
68
232
812
420 1
11
B
,738
373
45
t
t
328
30
37
,760
C.tropicalis
A
1,382 8,
250
183
t
t
67
284
265
460
8
B
656
550
78
t
30
472
26
65
48
C.utilis
A B
1,428 13,
276
186
t
t
91
236
838
468 1,
3
318
739
103
t
35
736
175
284
336.
Removals of:
COD %
TKN %
N-NH3 %
N-Org. %
Phosphates %
67
60
24
84
34
70
64
85
50
96
69
60
33
84
49
61
72
90
61
93
67
62
36
83
37
56
59
73
44
94
65
57
84
78
48
32
45
36
13
62
-------
COMPRESSOR
MOLASSES
34123412341234
Figure 59. Batch yeast production studies: (A) Experimental set-up;
(B) COD and TKN removals attained.
146
-------
All analyses were performed according to the. V. S. EPA guidelines or
Standard Methods except for the colorime.tric anthron.e determination of the
energetic material present, performed in accordance with Polish Standards
PN-76/C-04628/02 and except for the vitality L8ffler blue staining test.
Results
The following results are confined to the analysis of two substrates:
raw untreated pig manure and raw untreated manure enriched with molasses both
adjusted to pH 5.5. No hydrolysis was used for pretreatment because the process
was considered not feasible in full scale.
The tests were run on two parallel batches of wastewaters with and without
molasses each in four units to accommodate all four Candida type yeasts. The
results of these eight combinations are presented in Table 15, which provides
raw wastes centrate data, centrate (effluent) quality data and percent removals
of individual parameters. Thus, Table 15 serves as an estimation of the
potential of yeast fermentation in treatment of soluble fraction of piggery
wastes. This is better illustrated in Figure 59 where influent and effluent
COD and TKN concentrations are compared.
Table 15 shows raw and yeast treated centrate data because the only way
to remove yeasts, commercially, would be by centrifuge. The tests were run
on untreated total wastewaters. The cake from the centrifuge after fermenta-
tion consists of manure solids, seed solids and biomass culture. The actual
values, compiled in Table 16 were calculated by subtracting from the total
cake mass the masses of seed and the manure solids. The table also lists
unit cells yield calculated from the difference between the influent centrate
concentrations of nitrogen according to Kjeldahl method and carbon calculated
as 25 percent of COD and their respective effluent concentrations.
Discussion
Results in Table 15 indicate good COD removals by all species of the
Candida yeasts. The raw wastes. COD removals oscillate closely in the range of
65 to 69 percent. The nitrogen removals expressed in TKN show a range of 57 to
62 percent, similarly N removal is stable. Phosphorus removals seem to
org
favor C.tropicalis 8 (73 percent) and C.robusta 1 C40 percent) against C.tropi-
calis 11 (16 percent) and C.utilis 3 (13 percent). C.utilis 3 shows good
147
-------
removal of N-NH, (84 percent). The. data on fermentation of untreated raw
wastes for soluble pollutant removal Indicates the suitability of the first
two species mentioned.
TABLE 16. BIOMASS YIELD IN YEAST FERMENTATION
Yeast
C.robusta 1
C. tropi-
calis 11
C.tropi-
calis 8
C.utilis 3
Yeast yield
kg d.w./m
Raw
2.13
2.22
2.74
2.05
Enrich.
5.86
7.58
6.11
2.85
Unit cells yield kg cells /kg
nutrient removed
Nitrogen
Raw
5.5
5.7
6.8
5.5
Enrich.
4.6
5.8
4.8
2.3
Carbon
Raw
3.1
3.1
4.0
3.0
Enrich.
1.7
2.5
2.2
1.8
Phosphorus
Raw
16.6
44.4
12.0
49.0
Enrich.
18.7
80.0
19.0
8.4
NOTE: "Raw" refers to filtered raw wastewaters. "Enrich" refers to filtered
raw wastewater enriched with molasses.
The data for enriched wastes clearly discriminates between the four species
in favor of C.robusta 1 with the highest COD removal C70 percent) and constantly
high removals of other nutrients, 64 to 96 percent. Definitely second best
seems to be C.tropicalis 11, followed by C.tropicalis 8, while C.utilis yields
the poorest removals, 13 to 73 percent.
Analyzing the batch fermentation from the biomass production angle, in
Table 16, it becomes apparent that the two C.tropicalis give the best cell
yields in both raw (2.22 and 2.74 kg dry weight/m ) and in enriched wastes.
The interpretation of unit cells yield reveals much better utilization of
nitrogen, carbon and phosphorus in case of raw wastes. This means that enrich-
ment with molasses does not improve the unit yields, i.e. kg cells/kg N, P, C
removed, and that it leaves a considerable energy potential in the soluble
centrate in all cases exceeding the raw centrate COD value.
An important point should be stressed here, namely the initial ratios
C/N = 1.7 and C/P = 11 and for enriched C/N = 3.64 and C/P = 32.7, are much
below the average optimum ratios practiced in the fermentation industry. The
reason lies in the unusually high nitrogen and phosphorus content of piggery
148
-------
wastewaters, and leads to the content;ton. that without carbon, enrichment, yeast
fermentation will always retain a lot of nutrients in. the soluble effluent.
The analysis of data in Table 16 indicates a somewhat better performance
of Candida tropicalis strains 8 and 11 over C.rqbusta 1 and C.utilis 3, both
on enriched and raw wastes. The unit cells yield more and favor the two C.
tropicalis strains, however, the margin is not large.
Preliminary estimates of protein content reveals dry weight protein con-
tent from approximately 46 percent for Candida utilis 3 to over 60 percent
for C.tropicalis 11. The protein content was calculated by multiplying the
TKN content by 6.25.
Conclusions
1. The preliminary hydrolysis to increase the carbohydrate content of
piggery wastes proved ineffective and will not be feasible in full
scale.
2. The solidified substrate tests have screened out four Candida type
strains: C.robusta 1, C.tropicalis 11 and 8, C.utilis 3.
3. The batch liquid fermentation for 120 hours at an optimum pH 5.5
yields good removals of COD from the soluble fraction of wastewaters,
i.e. 65 to 69 percent for all species and TKN removals of 57 to 62
percent, for raw wastes without enrichment.
4. The enrichment of piggery wastewaters results in:
a. production of biomass apparently more lively than raw unadjusted
wastes;
3
b. production of larger biomass yield (kg/m ), however, at the
overall lower efficiency expressed in (kg cells) kg nutrient
utilized; and
c. increased removal of nitrogen and phosphorus from soluble
effluent.
5. The results indicate also that enrichment with molasses may be
economically difficult to justify, because:
a. enrichment does not increase the unit cells yield per nutrient
utilized;
b. it retains considerable concentration to residual COD and certain
amount of nutrients; and
149
-------
c. technical cost of supplying waste molasses may be.come, forbidding,
since it may be used directly as feed.
6. Considering the influent and effluent C/P, C/N ratios and their
respective changes, it is concluded that raw, filtered piggery
wastewaters are an unbalanced medium for aerobic aseptic fermentation.
7. The biomass of all yeast species, after drying and grinding, reveals
no trace of specific pig manure odor.
8. Out of the four screened species, the Candida tropicalis 8 was the
most suited for piggery wastes treatment and biomass production,
while Candida utilis 3 showed the poorest performance based on batch
tests.
9. Preliminary contents estimation in the dry biomass reveals high
protein content ranging from 46 to 61 percent, dry weight.
DYNAMIC STUDIES
The batch tests have shown inadequacy of the presently available methods
of pretreatment to increase the available carbon content. It has been found
in this work that acidogenic anaerobic digestion yields an increase in the
3
volatile acids content up to 10,000 mg/dm . Evison (114) expects to attain
3
the values as high as 20,000 mg/dm . Due to the lack of space in this project,
the system: anaerobic acidogenesis, yeast fermentation will be studied sepa-
rately in the future. The subsequent work is on untreated pig wastes from
Farm A, with and without addition of sucrose or sugar beet molasses (the solids
were removed by screening).
Materials and Methods
The studies were conducted on fresh piggery wastewaters, stored for 12
hours in 10 to 12°C and then centrifuged for 10 min at 1500 rpm.
Four yeast species selected on the basis of batch tests were used: Candida
tropicalis 8, C.tropicalis 11, C.robusta and C.utilis 3. When tests with carbon
enrichment were run, then 10 percent water solution of sucrose or beet molasses
diluted to 10 percent, by weight, of sucrose were used. Correction of pH was
1 Molar NaOH and 0.5 Molar H2SO,.
150
-------
3
Yeast production was conducted in a 3 dm active volume fermenter at a
constant air flow of 4Q dm /hr, mixed at 500 rpm, at pH - 5.0 and at 30°C
as shown in Figure 60.
The growth of yeast was limited by the quantity of easily available carbon.
The tests were concluded as follows: the fermenter was filled with filtered
3
raw piggery wastes (1.5 to 2.0 dm of wastes) and yeasts were added. Partial
dissolved oxygen (DO) pressure was measured. When the oxygen demand by yeasts
dropped, DO went up to 70 percent saturation, an aliquot of fresh substrate was
added automatically. The process DO controlled as described by Miskiewicz et
al. (115) and Robinson (119). In the test on wastes without enrichment, at the
3
DO meter signal 0.1 dm of wastes were added, until the liquid volume of V =
33
2 dm was reached (V = 1 dm ).
In the tests with sucrose the initial volume was 1.5 dm , the DO sensor
activated the sucrose feed tank until there was an inhibition of the biomass
growth.
In the tests with sugar beet molasses V = 1.5 dm , the tests were ter-
minated after 7 hours in order to keep the length of the run comparable with
3
the previous tests, thus, the final volume of approximately 2 dm was depen-
dent upon the amount of 10 percent molasses solution added.
The following analyses were performed on both raw and treated effluents:
COD, TKN, N-NH, by colorimetric nesslerization; N-NO~ colorimetrically with
sulphonyl acid and naphtylamine, N-NO, colorimetrically with phenylbisulphonyl
-3
acid, P-PO, colorimetrically and SOC.
Dissolved oxygen or 0« was measured potentiometrically by means of a mem-
brane DO sensor, as described by BorkoAski et al. (110), while yeast content
was evaluated turbidimetrically, as described by Miskiewicz (115). Total
protein content was determined in yeast cells as described in Methods in
Microbiology (107); the amino acid AA composition was determined, after 20 hours
hydrolysis Q5 N KC1 at 105°C) by a Czech AA analyzer.
151
-------
Ol
NJ
U)
2
o
5
K)
U)
U)
5
o
r~ •
^— v.
— •
VI
i %
•^_^ — ^.
— n
LJ
<— i
I_J
• °'a ' ••
v •..•: °J
0^0
DO - CONTROL
AIR
Figure 60. Semi-dynamic yeast fermentation equipment.
-------
The observed growth yield was calculated as:
Y = ^ (41)
where: Y - observed growth yield (g/mole)
S - assimilated carbon (mole glucose)
AX - net biomass solids increase (g)
Results and Discussion
Tests with raw wastes without enrichment—
3
The time needed for assimilation of available carbon in 2 dm of piggery
wastes, without carbon addition, varied from 2 hr for C.tropicalis 8 and
C.robusta to 3 hr for C.tropicalis 11.
3
The biomass increase was small and amounted to 0.385 g/dm for C.tropi-
3
calls 8 and 1.395 g/dm for C.utilis 3 and corresponded to the generally low
3 3
productivity of 0.192 g/dm hr for C.tropicalis 8 to 0.51 g/dm hr for C.utilis
3 (see Tables 17 and 18).
Table 18 describes also the utilization of carbon, phosphorus and nitro-
geneous compounds which were best utilized by C.robusta. This finding is
coincidental with the results of batch tests reported above. The TKN removals
in this study were 64 percent (60 percent in batch tests), where N-NH, de-
creased by 37 percent and N-NO- by 72 percent, with simultaneous increase of
3 -
the nitrates from 11 to 15 mg/dm N-NO-. Phosphate removals were low, some
18.6 percent (and also low in batch tests, 34 percent) while COD and SOC
removals were only 24 percent and 35 percent, respectively.
The C.utilis 3 species has removed 64 percent SOC while C.tropicalis 8,
58 percent, however, the TKN removals by these species were smaller than
C.robusta, and amounted to 11.5 and 14.3 percent, respectively. Higher COD
removals were obtained by the two species: 48 percent for C.utilis 3 and 59
percent for C.tropicalis 11.
These results show that yeast production on piggery wastes without carbon
enrichment is at present not economical from the standpoint of biomass produc-
tion. It is interesting to note that removal rates calculated from the formula:
153
-------
TABLE 17. YEASTS PRODUCTION ON FILTERED PIGGERY WASTEWATERS
WITHOUT CARBON ENRICHMENT - REMOVAL EFFICIENCIES
en
C.tropicalis 8
COD
TKN
N-NH+
N-NO~
N-NO~
P-PO'3
soc
Raw „
mg/dm
3,860.
875.
400.
0.
5.
736.
800.
Removal
0
0
0
01
0
0
0
C.tropicalis 11
Raw 3
mg/dm
3,040.
1,057.
370.
0.
13.
680.
1,125.
0
0
0
034
0
0
0
Removal
59.0
11.5
23.0
68.0
increase
to 2834
mg/dm
8.0
58.0
C . rubusta
Raw _
mg/dm
2,430
1,442
342
0
11
860
850
.0
.0
.0
.018
.0
.0
.0
Removal
24.0
64.0
37.0
72.0
increase
to 3338
mg/dm
19.0
35.0
C.utilis
Raw
mg/dm
3,860.
875.
400.
0
0
0
0.01
5..
736.
800.
0
0
0
Removal
48.0
14.0
25.0
28.0'
increase
to 2730
mg/dm
3.4
64.0
-------
TABLE 18. RESULTS OF YEAST PRODUCTION WITHOUT CARBON ADDITION -
YEASTS YIELDS AND NUTRIENTS USE*
Parameter
Initial yeasts
Net biomass
increase
Duration of
experiment
Maximum specific
growth rate .
Productivity
Cell yield per
nutrient used
Carbon
Phosphorus
COD
Average removal
rate coefficient
Average removal
rate coefficient
Symbol
Xo
A X
t
p
P
Y i
k
(SOC)
k
(COD)
Units
g DM
g DM
hr
hr-1
g DM/dm3/hr
R DM
g nutrient
g/g C/hr
g/g P/hr
g/g 02/hr
10""3dm3/mg C/d
10~3dm3/mg 02/d
C.tropicalis 8 C.tropicalis 11
10.02 8.74
0.77 1.96
2.0 3.0
0.05 0.10
0.19 0.33
0.52
10.2
0.18
1.0
1.1
C.robusta
5.22
1.06
2.0
0.10
0.26
0.88
1.65
0.45
1.0
0.6
C.utilis
fresh
filtered
11.71
2.79
2.75
0.12
0.51
0.45
31.7
0.19
1.0
0.6
*Wastewater volume 2 dm .
-------
' " Se (42>
where: S is the removed mass of pollutant including added carbon in test
with enrichment (g),
3
S is the effluent concentration (mg/dm ),
e
(X + AX) are the biological volatile solids (g DM); are comparable
o
with removals attained in the activated sludge systems. As reported by
Oleszkiewicz, Koziarski, et al. (11), this value for full scale systems was
3 3
200 to 400 mg/dm /hr, which is similar to the 180 to 390 mg/dm /hr attained
in the yeasts tests.
Yeast production on wastewaters enriched with sucrose—
The introduction of sucrose has resulted in elongated biomass production
time of 6.5 to 7 hr as shown in Table 19. Sucrose was not a suitable carbon
source for C.robusta and C.tropicalis 8 as far as the biomass yield was con-
cerned. The specific growth rates for the two species have reached a level of
0.083 h~ for C.robusta and 0.052 h~ for C.tropicalis 8, and then decreased
gradually to zero as shown in Figure 61. The biomass increase was small and
3 3
amounted to 1.56 g/dm for C.robusta and 1.49 g/dm for C.tropicalis 8.
The yeasts C.tropicalis 11 and C.utilis had multiplied according to a
different pattern. The specific growth rate has reached, respectively, 0.189
and 0.206 h , and then decreased to zero (Figure 61); while the productivity
3 3
was 0.934 g/dm /hr and 1.26 g/dm /hr, respectively. The biomass increase was
3 3
7 g/dm for G.tropicalis and 9.13 g/dm C.utilis 3.
It was characteristic for all cultures that the biomass increase has
gradually deteriorated to zero (Figure 61). Although the contents of nitrogen,
phosphorus, and DO were sufficient, the yeasts had halted the uptake of suc-
rose. It seems that this phenomenon is the result of accumulation of inhibiting
metabolites, a thing to be expected from periodic or semibatch cultures where
the biomass and substrate are not exchanged.
Candida robusta has again proved to utilize the most of nitrogen (78
percent removal of N-NO^t whlle removing other nutrients to the degree similar
156
-------
TABLE 19. YEASTS PRODUCTION ON FILTERED PIGGERY WASTEWATERS
WITH CARBON ENRICHMENT - REMOVAL EFFICIENCIES
C.tropicalis 8
+ sucrose
COD
TKN
N - NH?
N~- N0~
N - N0~
P - PO"3
soc
Raw
mg/dm
3,860
875
400
0.01
5.0
736
900
Removal
%
35
35
25
20
Increase
6
15
28
C.tropicalis 11
+ sucrose
Raw Removal
mg/dm X
3,040
1,060
370
0.03
13
680
1,125
C.robusta
+ sucrose
Raw
mg/dm
24
14
542
0.02
11
860
850
Removal
%
41
60
27
78
increase
15
5
41
C.utilis
+ sucrose
Raw
mg/dm
3,860
875
400
0.01
5
736
800
Removal
%
54
62
74
30
increase
8
76
37
+ molasses
Raw
mg/dm
5,568
1,158
643
0.206
0.30
468
2,000
Removal
%
60
76
84
13
0
84
69
-------
4567
INCUBATION TIME-t(h)
Figure 61. Semi-dynamic (batch-fed) fermentation:
(B) Biomass increase.
(A) Specific growth rate;
158
-------
to tests without enrichment (Table 19). The level of nutrient utilization
for C.utilis 3 has increased and the removals were 62.3 percent TKN, 74 percent
N-NH,, 30 percent N-N07, and 75.5 percent P-PO^. In spite of adding carbon
the COD removal increased to 54.3 percent. Contrary to C.tropicalis 8 and C.
robusta, this species showed good cell yield, similar to C.tropicalis 11.
The protein content in C.utilis 3, C.tropicalis 8, C.tropicalis 11, and
C.robusta was, respectively, 429, 437, 476, and 434 mg/g. The aminoacids compo-
sition was similar to C.utilis from spent sulphite liquors, or C.utilis grown
on molasses by Peppier (118). A somewhat lower content of leucine, isoleucine,
treonine, methionine and cystine has been found for the C.utilis 3 and C.tropi-
calis 11. The digestible protein varied from 60 to 64 percent of the total
protein.
It is interesting to compare the specific cell yields Y per g of nutrient
s
removed for the enriched wastes (Table 20) with those for wastes without
enrichment (Table 18), as well as respective SOC and COD removal rates. The Y
values show excellent utilization of carbon (and nitrogen) by C.robusta and
C.tropicalis as compared to poorer use of the nutrients by C.utilis. This is
coincidental with the findings of the batch study preceding this work (117).
The comparison of removal rates, calculated as in wastewater treatment opera-
tions, shows that enrichment provides for higher rates. It should be pointed
out, however, that the increase is not large enough to offset the high cost of
added sucrose, and that C.robusta and C.tropicalis are still excellent yeast
species for piggery wastes.
Yeast growth on wastewaters enriched with molasses—
Only C.utilis 3 species were tested on beet molasses enriched wastes, which
was found to be a better carbohydrate substrate than sucrose. The productivity
3
was equal to 1.58 g/dm /hr and was higher than that obtained with sucrose
3
(1.26 g/dm /hr - Table 20). Similarly, higher conversion of substrate components
into cell biomass was found. The cell yield was equal to 115 g/mole as compared
to the yield with sucrose (95 g/mole) and was higher throughout the test. A
milder inhibiting rate was found during the test with molasses, the biomass
increase was still quite intense after 7 hours, i.e. time of almost complete
159
-------
TABLE 20. TESTS WITH CARBON ENRICHMENT - YEASTS YIELDS AND KINETIC DATA
CM
O
Parameter
Initial yeast
quantity
Net biomass
increase .
Introduced carbon
calculated as
{•lucose
Duration of
experiment
Maximum specific
growth rate
e
Productivity
Cell yield per
nutrient used
(removed)
Carbon
Nitrogen
Phosphorus
COD
Average removal
rate coefficient
Symbol Units
Xo g DM
Ax g DM
C m mole
t h
v h-1
m
P g DM/dm3/h
Y g DM/g nutrient
g DM/g C/h
g DM/g N/h
g DM/g P/h
g DM/g 02/h
k(SOC) 10~3dm3/mg C/d
k(COD) 10~3dm3/mg 02/d
C.tropicalis 8
+ sucrose
10.79
2.24
5.83
7
0.05
0.21
0.3
0.7
2
0.09
0.04
0.4
C.tropicalis 11 C.robusta
+ sucrose + sucrose
7.54 4.71
10.51 2.34
116.67 17.5
7.5 6
0.19 0.08
0.93 0.26
0.22
0.30
6.5
0.13
2
1.2
Candida
sucrose
8.40
13.7
137.67
7.25
0.21
1.26
0.19
2.28
2.28
0.04
3
3
utilis 3
molasses
10.8
11.05
110.83
7
0.17
1.58
0.16
1.19
2.67
0.04
2.6
2.6
-------
halt of activity in the sucrose system (Figure 61). The phenomenon of better
yields with molasses may be explained by the presence of additional micro-
nutrients in molasses that have a positive effect on yeasts.
The use of beet molasses instead of sucrose yielded the following increase
+ -3
of removals: TKN from 62 to 76 percent, N-NH, from 74 to 85 percent and P-PO,
from 76 to 84 percent (Table 20), in spite of the fact that molasses brought in
an additional load of pollutions.
DISCUSSION AND CONCLUSIONS
The research has shown that various physicochemical methods of primary pig
wastes treatment to increase the content of easily available carbon are not
efficient technically and economically. It may also be shown that enrichment
with molasses is economically unfeasible as the molasses is becoming an expen-
sive and scarce component of animal feeds.
With the selected species of Candida yeasts, it is feasible to produce
SCP for animal feeding purposes. It is, however, important to keep in mind
the fact the full scale yeast production process may require continuous cul-
ture, sterile conditions which may not be feasible with piggery effluents.
At the same time, the efficiency of wastewater treatment will not offset the
high cost of yeast production, and the carbon that will be incorporated in
yeast cells will originate primarily from added molasses.
In view of above the future work will have to go along the lines of
finding mixed yeasts cultures with low carbon requirements as well as finding
alternative low-cost waste carbon source in agriculture such as silage efflu-
ents or potato processing plant wastes.
Preliminary studies with four species of Hansenula yeasts: Hansenula
canadensis (W), henricii (W), nonfermentans (W), Wickerhamii (W) obtained
from Dr. A. Kockova-Kratochvilova catalog (Btatislava) have proved better
adaptability to the pig wastes conditions. The Hansenula yeasts are capable
of utilizing for metabolic purposes other forms of carbon than the easily
biodegradable carbohydrates, e.g. fatty acids, alcohols, etc.
161
-------
The conclusions of the work done on Candida yeasts may be summarized as
follows:
1. The use of raw piggery wastes without carbon enrichment has yielded
unsatisfactory effects both in case of biomass yield and wastewater
treatment, although the unit utilization of CNP nutrients was quite
high.
2. Sugar beet molasses was found the most appropriate source of easily
available carbon.
3. The introduction of carbon results in significantly larger decrease
of nitrogen and phosphorus compounds from piggery wastes.
4. When sucrose was used, the best yields and pollutant removals were
found for C.utilis 3 and C.tropicalis 11 in dynamic tests.
5. The enrichment with molasses (C.utilis 3) yielded the following in-
creases in removals, compared with the sucrose system: TKN from
-3
31 to 76 percent; COD from 48 to 58 percent; PO, from 42 to 84 per-
cent; and SOC from 64 to 69 percent.
6. The productivity of C.utilis 3 on molasses-enriched piggery wastes
was increased, in comparison to the raw wastes system: 0.507 to
3 -1
1.578 kg/m /hr while the maximum specific growth rate from 0.125 h
to 0.172 h"1.
7. The productivity of C.tropicalis 11 increased, after enriching with
o
sucrose, from 0.327 to 0.934 kg/m /hr and the maximum specific growth
rate from 0.105 to 0.189 h~ .
8. The metabolic by-products of yeast production result in the biomass
growth inhibition. The effluent quality is still not satisfactory
from the standpoint of the follow-up treatment.
9. The yeasts grown on piggery wastes have aminoacids composition simi-
lar to those grown on spent sulphite liquor and on molasses alone and
evidence a lack of swine wastes odors.
10. The unit carbon utilization rates (g DM/g C removed/hr) are generally
higher for wastes without enrichment; C.robusta has shown the best
conversion rate in both the raw and enriched wastewaters.
11. The removal rate coefficients for systems without enrichment are
3
equal to 0.001 (SOC) and 0.0006 to 0.0011 (COD) dm /mg/d and are
comparable to activated sludge systems working on presettled wastes
-------
(K = 0.0004 dm /mg/d). The coefficients for enriched systems are
higher due to the selective removal of more available carbon.
12. The protein content in the Candida species is 43 to 47 percent, the
digestible protein content was 263 to 305 g protein/kg.TS. The
amino acids content indicates excellent applicability of these yeasts
as feed components certain methionine deficiency calls for supple-
mentation.
163
-------
SECTION 10
WASTEWATER PONDS FULL TREATMENT SYSTEMS
METHODS
A series of four ponds was studied in laboratory conditions in a setup as
in Figure 62. The system consisted of an anaerobic pond followed by an aerated
pond and two oxidation-stabilization ponds. The relative capacities are pre-
sented in Table 21 together with percent HRT in each pond.
TABLE 21. WASTEWATER POND SYSTEM'S PARAMETERS
Pond
Anaerobic
Aerated
Oxidation - 1
Oxidation - 2
TOTAL
Arga
m
0.180
0.090
0.090
0.090
0.45
Volume
dm
46.00
21.75
19.75
19.00
106.50
% HRT
43.20
20.42
18.54
17.84
100.00 %
The anaerobic pond was covered but was not air-tight, the aerated lagoon
was both mixed and aerated. The oxidation ponds were continuously mixed to
assure full utilization of their active volume.
The polishing (oxidation) ponds were lighted 12 hr/day. The seeding with
algae, collected from natural pools contaminated with piggery wastes biological
effluent, was done once throughout the experiment. Anaerobic pond was seeded
several times with actively digesting anaerobic sludge.
The whole experiment lasted 16 months. Two parallel systems were inves-
tigated at one time. The steady state conditions were assumed after passing
of time equal to 4 to 5 HRT. The whole system was closely investigated as a
164
-------
J^SOrpm I >50rpm
|500rpm \ V
• V / S / f\/ / / /////S///\/f /f /////// /S S//
^. , ^ , . _ _ r • — •*••••• IT"1
T /IN ' /i\ /l\- /IN p^
I
a-FEED TANK
b-ANAEROBIC POND
c-AERATED POND
d-ALGAL POND I
e-ALGAL POND II
f-COLLECTION TANK
Figure 62. Layout of the wastewater ponds system.
-------
series of individual unit processes. The samples were collected in five loca-
tions as in Figure 62. Additionally, samples for TVS and TSS were collected
from the anaerobic and aerated lagoon to evaluate the concentration of active
biomass. The analyses performed routinely on the samples one thru five
included: pH, DO, BOD5 f, CODnf, CODf, TS, TVS, TSS, VSS, DS, TKN, N-NH3>
N0_, NO-, and N-org. Other parameters were determined sporadically. The
following presentation is based on at least 6 systems, 10 runs, 5 samples
and 16 analyses = 4800 individual data points.
RESULTS AND DISCUSSION
Due to space limitations, only the overall removal data for the whole
treatment train will be presented in Table 22. Other results are presented
graphically in Figures 63 and 64. It follows from the table and Figure 63
that the removal of soluble organics gains little after HRT = 60 days and that
the stability of effluent quality improves at retention times HRT >_ 30 days.
Comparing this with data in Figure 64, which presents concentration of various
forms of nitrogen versus HRT, one arrives at conclusion that basic carbon and
TKN biodegradation is completed at HRT >^ 30 days and that nitrification begins
to take place beyond that retention.
The kinetics of biological processes in individual ponds and analysis of
nitrogen pathways indicate several areas where full scale research should be
directed. The problems of wastewater recycle to improve nitrogen removal,
biomass harvesting and heat conservation are all considered as factors that
could make the process more efficient.
TABLE 22. OVERALL EFFICIENCY OF WASTEWATER
TREATMENT IN THE PONDS SYSTEM
Hydraulic
Parameter (g
Influent (S
o
Effluent (S
02/dm3)
) CODnf
BODnf
) CODnf
BOD
nf
Average temperature
of the run (°C)
7
14.07
5.51
1.46
0.75
21.5
14
5
0
0
21
retention in all ponds days
14
.07
.51
.91
.43
.5
30
13.82
5.44
0.45
0.22
22.5
45
14.00
5.51
0.35
0.17
22.4
60
13.82
5.44
0.29
0.14
22.5
90
14.00
5.51
0.20
0.11
22.4
166
-------
05
? 04
OJ
to
in
Q
O
m 0.3
LU
13
LU
0.1
0-
10
08
CO
-2^06
in
Q
o
^04
02
10
20
I
30
40
50
60
I
70 80
HRTId
97
96
95
94 1
LU
QL
93 •£
in
Q
O
CD
92
91
90
89
100
98
96 ^
O
94
Q
O
CJ
92 z
90
90
Figure 63.
Effects of overall retention on organics removal in the ponds
system.
167
-------
400
on
E
LU
CO
Q
•z.
z>
o
Q.
S
o
o
o
o
a:
10
20
30
40
50 60 70
HRTtd
80 90
Figure 64. Nitrogen compounds and pH in the effluent from polishing pond II
versus overall retention in the system.
168
-------
To summarize these findings, it is assumed that in full scale, a system
with HRT = 60 days should yield 97 percent COD - and 97 percent BOD
removal and an effluent COD, and BOD5 -, respectively, 250 and 150 mg 0»/dm
in summer conditions.
In order to improve the efficiency of the overall treatment system an
additional polishing step is required, particularly to account for the winter
conditions. Soil filtration with wastes introduced three feet underground,
such as used at some rural treatment plants could be a natural solution, in
line with the simple to operate, main waste treatment system.
169
-------
SECTION 11
PROTEIN RECOVERY
INTRODUCTION
The protein deficit in the developed countries directly related to the
deficit of animal feed. In Poland close to 85 percent of protein is used as
animal feed and only 15 percent for direct consumption. The forecast for 1990
is 1.1 million tons for direct consumption and 7.7 million tons as animal
feedstuff. The conversion of plant protein to animal protein involves signifi-
cant costs and high losses which may amount to over 80 percent of the applied
feed.
The increasing feed deficit in animal husbandry can be partly alleviated
in conjunction with measures leading to improvement of the wastewater efficien-
cies. Two methods of utilization of the energetic potential of animal wastes
are used: direct recovery and recycling of manure into feed and indirect
method of conversion into microorganisms biomass, the SCP. The literature
perusal reflects the increasing interests particularly in the second method
(111, 122, 126, 127, and 128).
This section will discuss the feed values of screenings separated from
piggery wastes, and of excess activated sludge, yeasts of the Candida type
and of algae grown on treated effluents. Yields and effects will be presented
and amino-acids content will be defined and compared with other feed compon-
ents. Possible full scale application feasibility will be reviewed.
Methods
As noted by Martin et al. (137), animal manures can be divided into
three general categories: 1) energy feedstuffs; 2) protein feedstuffs; and
3) forages. Energy feedstuffs, such as corn, contain less than 20 percent
protein, less than 18 percent fiber and are high in easily available energy.
170
-------
High protein feeds contain more than 20 percent protein. Forages consist of
vegetative materials high in fiber, usually over 18 percent. Protein feeds
have the highest monetary value, followed by energy feeds and then forages.
The following analysis will be based on protein and amino-acids determin-
ations in recovered material. References to feeding trails reported by others
will be made based on literature.
The estimate of total protein (TP) content was made on the basis of total
Kjeldahl nitrogen analysis and calculation of TP = 6.25 TKN. The digestible
protein (DP) content was based on the same TKN analysis made on material
after 48 hours digestion with pepsin in presence of hydrochloric acid (129).
The ratio of DP/TP is equal to digestibility index (DI). The amino-acids
content was determined by means of an amino-acids analyzer.
DIRECT RECOVERY
Recycling of solids from piggery wastes has been practiced on experimental
scale in several countries and is still not sanctioned by the Polish health
authorities or the U.S. Food and Drug Administration. Several studies have
shown negative effects, notably the one by Harmon (131) who noted weight loss
in swine fed anaerobically predigested solids.
In one Polish study (132) 64 percent and 50 percent of dried piggery solids
were added to the following respective feed components in two mixes: 22 percent
and 34 percent of barley, 5 percent and 7 percent of beans, 4 percent of corn,
molasses and additional minerals. In the first case the feed gave a yield of
865 g LW/day while the second feed, the yield was 1019 g LW/day of cattle.
The studies of meat quality revealed no significant difference between the
cattle fed rations with wastes and the controls.
In another study by Flachowski et al. (134) the piggery wastes screen-
ings were preserved with urea and treated with sodium hydroxide and fed to
bulls with other feedstuffs, with positive effects.
171
-------
The work of Dzieniechowicz (150) on direct recycle of piggery waste
screenings has demonstrated that both sheep and cattle can take up to 50
percent of dried solids in their diet. The feed containing 50 percent of
screenings had 23 percent of protein, 4.23 percent fat, 18.71 percent of fiber
and 33.37 percent nonnitrogeneous substances yielded a daily gain of 1326 g
in beef cattle, without an apparent sign of health or meat quality deteriora-
tion.
It is noted that numerous studies on ensiling prove increasing applicabil-
ity of the process to protein recycling (135, 136). Ensiling improves the
digestibility of the waste material and increases the value of the by-product
when incorporated in feedstuffs.
Table 23 illustrates the protein content of various materials obtained
from Farm A effluent studied in this project.
The protein and amino-acids analyses of screenings have shown that they
can serve as a volumetric feed (forage feed) ingredient for ruminants. The
screenings contain 7 to 9 percent TP of 45 percent digestibility, large amounts
of amino acids and 200 g/kg DM of mineral components. The low digestibility
is the result of fiber content.
An interesting experiment is presently run by authors at Plant A. The
screenings are placed in 20 to 40 cm layers on a sheet of plastic foil. The
seeds of oats, rape and lupin are planted and the resulting forage, together
with the base substrate undergoes ensiling or is fed to the drier. The pro-
cedure should allow for 3 to 4 crops/year provided they are covered in the
winter time.
Direct recovery of solids, as screenings in the process of dynamic
sieving (sieve 1 to 1.5 mm mesh) is practiced at large farms as a method of
solids removal for disposal rather than a direct economic incentive. The
low protein content places screening in the group of low cost forage feedstuffs.
The economic comparison of costs of recovery with monetary value of solids
(approximately 5 to 6 tons/day wet weight are recovered at Farm A) makes
172
-------
solids removal by screening an unnecessary treatment step. This is particularly
true if settling tanks and/or anaerobic digesters are introduced.
TABLE 23. TOTAL AND DIGESTIBLE PROTEIN IN
PIGGERY EFFLUENT WASTE MATERIALS
1.
2.
3.
4.
5.
6.
Material
Solids from dynamic screens
Excess activated sludge
Mixed algal - bacterial
cultures - Chlorella and
Scenedesmus predominant
Yeasts monocultures:
C.robusta 1
C.tropicalis 11
C.utilis 3
Lemma minor from the algal
ponds (lab.)
Daphnia Magna** on pig manure
diluted 1:1
Protein (mg/g DM)*
Total Digestible
84.2
413.4
376.2
225.3
372.1
269.1
429.2
437.5
433.0
274.6
500.0
37.0
251.5
237.4
110.5
191.4
133.7
271.0
281.1
262.7
216.0
.
Digestibility
Index (%)
45.8
60.8
63.0
49.0
49.9
49.7
63.1
64.2
60.5
78.7
.
*DM - dry matter
** - based on data by Jarocka (130)
Other efforts in direct recovery include separation of feces by hanging
nets below slatted floors (133), but the economics of the process are still
questionable.
In conclusion sieving to recover solids does not seem to be economically
efficient. Recovery and processing through high temperature driers should
be practiced at plants where dynamic screens are already installed, however,
alternative methods such as composting and ensiling should be introduced.
CONVERSION INTO BACTERIAL SCP
Excess activated sludge may contain significant quantities of protein
from 380 to 530 g/kg DM and beneficial quantities of vitamins and minerals.
Several attempts on direct refeeding of activated sludge from under-floor-
oxidation ditch back to swine were shown successful by Harmon (131).
173
-------
In this country centrifuge concentration was tried in two cases to dewater
excess sludge for recovery. In experiments in Kolbacz 22 percent DM (138) was
attained with 24 percent raw protein in the dry mass. The excess sludge was
from the 48 hr aeration tank operating at F/M = 0.5 kg 02/kg MISS/d (BODj, in
the Gi-Gi system shown in Figure 6. The centrifuge cake was used for preparing
silage of the composition as in Table 24. The prepared silage had good taste
and odor and was only 9 percent less digestible than the normal silage. The
first silage contained protein 36 g/kg and 440 g DM/kg while the second, res-
pectively, 40 g/kg and 430 g DM/kg.
In the authors' experiments (29) excess sludge from Plant A of the Vidus
3
type was centrifuged in pilot scale in 5 m /hr installation of a decanting
Humboldt-Vedag centrifuge. Various water quality grade coagulant aids were
used to improve the dewaterability of excess activated sludge; some 13 percent
of total solids (DM) were attained in the cake with some 39 percent of crude
protein in the DM. The digestible protein was 237 to 251 g/kg with 61 to 63
percent DI. The material thus recovered was fed into an industrial dryer.
Strong odors have precluded longer experiments, however, the final pelleted
product had very good organoleptic properties. The amino-acids spectrum is
presented in Table 25, against the values for chicken eggs.
In conclusion excess activated sludge is an excellent high protein feed,
with high vitamins content and high digestibility. The present processes of
piggery wastes treatment are based on activated sludge and thus recovery seems
feasible. The major problem is the high cost of dewatering and drying.
Presently, these costs_make the process uneconomical.
TABLE "24. THE COMPOSITION OF SILLAGE PREPARED~ "
WITH EXCESS ACTIVATED SLUDGE
-% Weight
Components °
Corn grain
Barley grain
Hay cut
Straw cut
Grass silage
Activated sludge cake
TOTAL
Silage I
27
-
7
-
23
43
100%
Silage II
-
25
-
6
28
41
100%
174
-------
TABLE 25. AMINO ACIDS COMPOSITION OF PROTEIN FROM RAW WASTES
SCREENINGS, EXCESS ACTIVATED SLUDGE AND DUCKWEED
Amino Acids
mg/g DM protein
1
Lysine
Histidine
Arginine
Asparg. Acid
Treonine
Serine
Glutamic Acid
Glycine
Alanine
Valine
Methionine
Isolwucine
Leucine
Tyrosine
Phenyloalanine
TOTAL
mg/g DM
Excess
vated
Run 1
2
4.2
2.0
3.6
7.1
3.7
3.2
7.6
4.3
6.2
4.6
1.1
3.4
6.0
2.5
3.8
63.2
261.1
acti-
sludge
Run 2
3
4.9
2.0
3.4
6.6
4.3
3.1
8.2
5.0
7.7
5.6
1.2
3.9
6.7
2.9
4.0
70.5
265.1
Solid
Screenings
Run 1
4
3.5
1.6
3.2
5.2
2.8
2.6
7.9
3.7
4.1
3.7
0.4
2.7
4.5
2.1
3.0
51.1
43.0
Run 2
5
3.3
1.5
3.2
5.3
2.9
2.7
8.4
4.4
4.5
3.8
0.2
2.6
4.2
1.8
3.0
51.8
37.4
Duckweed
LGHHHcL
Minor
6
4.1
1.9
7.4
14.2
3.6
3.1
8.5
4.0
5.1
4.3
0.7
3.4
5.6
2.9
4.1
72.7
199.5
Chicken
Egg
7
6.3
2.1
6.4
-
5.0
-
-
-
-
7.1
3.1
6.8
9.0
4.4
6.0
-
_
CONVERSION INTO ALGAL PROTEIN
Numberous studies were conducted on the growth of algae on wastes, very
few so far on piggery wastes. Most of the studies are on pure monocultures,
but there are beginning to appear papers on mixed or symbiotic growth of algae
and bacteria on wastes (140). In this project symbiotic continuous algal-
bacterial cultures were grown in four laboratory ponds in series, 30 cm deep,
illuminated 24 hr/day and kept at 22+3 C. Biologically treated piggery
wastes were used, the effluent from Plant A final clarifiers.
Various species were tried, however, only Chlorella vulgaris and Scene-
desmus survived the piggery wastes environment. The biomass was harvested by
175
-------
sieving, centrifuging and drying at 55 C. The average yields were presented
in Section 7. The harvested biomass contained 22.5 to 37.2 percent of protein
with DI = 49 percent and containing also 10 to 17 percent of raw cellulose and
3 to 8 percent fat.
Biomass harvesting and the problem of breaking the hardly digestible cell
walls are presently the two major obstacles in fuller application of algal
ponds (45, 141, 142). Comparison of sedimentation, filtration and centri-
fuging has proved the latter to be the preferred harvesting method although
it is still far from satisfactory.
The cellulose and chemicellulose which makeup some 50 percent of the cell
wall mass are only partly digested by ruminants. Thus, the final applicability
of algae as feed can be evaluated only after a thorough disintegration of the
cell wall and on samples that are free from bacteria.
The studies conducted in field scale by de Pauw (46) have yielded similar
2
results. De Pauw attained yields of 1 to 10 g DM/m /d at various temperatures
in different seasons and has managed to grow on raw and pretreated piggery
wastes, Scenedesmus, Chlorella v. and Coelastrum probascid. Other researchers,
MUntz (143) and Goldman (43), have also found cell was digestibility to be
the major obstacle. The future of algae application lies in solving the two
problems: separation and digestibility.
Based on present difficulties, the solution is to use the algal ponds in
natural ecosystems of aquaculture and using the developed biomass for fish
production. Other methods to be evaluated include the use of algae as supple-
ment to biogas generation such as pilot work in Italy by Micheli (72) and the
development of mutants with low-cellulose-walls or larger volume to cell wall-
surface-area ratio.
CONVERSION INTO YEAST PROTEIN
The production of protein by yeast is over two hundred times faster than
in the conventional agriculture supplying the farms, and the studies show that
higher protein contents are attained in yeasts than in field crops.
176
-------
The fermentation yields in this project are described in Section 9. In
the raw wastes fermentation the specific growth rate, productivity and the
level of C, N, and P nutrients utilization were unsatisfactory. Much higher
yeast production was attained on enriched wastes, however, then the cost was
comparable to yeast production from molasses alone. The yeasts contain 43 to
47 percent crude protein of 60 to 64 percent DI. The results indicate the
technological feasibility of protein recovery through yeast fermentation.
The high costs of yeasts fermentation due to the use of expensive and not
readily available molasses suggest the use of integrated treatment-recovery
systems. The combined system of feed yeast production, animal husbandry,
and a joint water, wastewater management and treatment systems will allow the
decrease of the overall costs.
It should be pointed out that a lot needs to be done to increase the
overall efficiency of the process. In particular mixed yeast cultures should
be considered, the mastering of the principle of continuous fermentation, and
the decrease of costs of additional carbon.
Preliminary trials have verified the possibility of applying the continu-
ous process. The maintenance of the steady state for prolonged periods of time
without the need for inocculum and culture replacement is of major importance
to process cost decrease.
It seems that before the solutions to these problems are found, yeast
protein from piggery wastes will be economically unfeasible until further pro-
tein price increases occur.
NUTRITIONAL VALUE OF RECOVERED PROTEIN
Although the feeding trials are the best method of evaluating the nutri-
tional potential, the chemical methods give rapid answers and usually are
highly correlated with the biological tests.
Evaluation of the protein quality was made on the basis of analysis of 15
AA and comparing them to the whole egg protein, looking for the most deficient
177
-------
AA as proposed by Ruszczyc 051). It is assumed that the deficient AA decides
about the value of protein. Calculating in this manner the nutritional value
of the recovered protein we have obtained: activated sludge 38 percent, yeasts
31 percent, algae 29 percent, Lemma minor 24 percent, and screenings 10 percent.
In all cases methionine is the limiting acid, however, it is an inexpensive
commercially available component. Although thyrosine, valine and lisine are
present at lower concentrations, the recovered SCP does contain the full
spectrum of exogennic AA and as such is considered adequate feedstuff.
It should be noted that several sources suggest the use of Food and
Agriculture Organization (FAO-UNO) reference protein standard which is given
in Table 26. When comparing the AA content from Tables 25 and 27 with FAO
standard, the nutritional values of recovered proteins are much higher as
shown in Table 28. This and the fact that species grown on pure cultures,
without inhibitors also reveal significant deficiencies, while at the same
time they are considered fully acceptable, lead to the conclusion that the
materials recovered from piggery wastes processing are adequate feedstuff
supplements or substitutes.
All recovered protein may be used only as feed supplement, up to a
maximum of 50 percent, due to low digestibility. Activated sludge, yeasts
and Lemma Minor (duckweed) have proved to be the most easily digestible source
of SCP; cellulollytic material was responsible for low digestibility of other
SCP sources, particularly in case of algae.
The AA of algae is 162 to 145 mg/g DM, activated sludge 265 mg/g DM and
yeasts 246 to 291 mg/g DM. When calculating these amounts in terms of mass
per 100 g of protein, the AA content is 64 to 74 g for algae, yeast and
activated sludge and some 50 g for solid screenings.
The possibility of using yeasts and algae has been demonstrated experi-
mentally (147): yeasts can be used in feed up to 20 percent by fattening
swine, up to 15 percent by small piglets and 10 percent by sows and boars.
In feeding poultry, layers, and cocks one can use up to 20 percent of SCP
yeasts.
178
-------
TABLE 26. AMINO-ACIDS CONTENT IN PROTEIN RECOVERED FROM
SYMBIOTIC ALGAL - BACTERIAL BIOMASS
Values in
mg/g DM Protein
Lisine
Histidine
Arginine
Asparg. Acid
Threonine
Serine
Glutamic Acid
Glycine
Alanine
Valine
Methionine
Isoieucine
Leucine
Tyrosine
Phenylalanine
SUM
Total mg/g DM
Algae
Reference
Mixed cultures - Plant A Pure Cultures
I
3.9
1.8
3.9
7.5
4.1
3.5
8.2
4.9
6.7
4.9
0.8
3.8
6.5
3.1
3.9
67.6
152.3
II
3.5
1.7
4.1
3.2
4.0
3.3
7.9
7.9
6.8
4.8
0.9
3.6
6.7
3.0
3.9
65.9
245.0
Chlorel-
III la
4.9 8.0
1.6
3.4
7.2
3.7 5.3
3.4
8.1
4.1
5.5
3.9 6.2
0.9 1.8
3.0 4.8
5.8 9.3
2.4
3.7 6.6
61.4
165.2
Scene-
desmus FAO
5.7 4.2
-
-
-
5.2 2.4
-
-
-
-
7.2 4.2
1.4 2.2
4.4 4.2
9.3 4.8
4.8
4.6 2.8
-
_ _
NOTE: Data from continuous culture four bacterial-algal ponds in series,
treating biological effluent from Plant A.
Reference data on pure culture are from Boersma et al. (47).
179
-------
TABLE 27. AMINO-ACIDS CONTENT IN PROTEIN RECOVERED FROM YEASTS GROWN
ON FILTERED RAW PIGGERY WASTES ENRICHED WITH SUCROSE
Amino-acids
mg/g DM
Protein
Lisine
Histidine
Arginine
Asparg. Acid
Treonine
Serine
Glutamic Acid
Glycine
Alanine
Valine
Methionine
Isoleucine
Leucine
Tyrosine
Phenylalanine
TOTAL
mg/g DM
Yeasts monocultures
C.robusta
4.9
3.2
3.0
7.1
3.8
3.3
6.8
2.8
4.7
4.1
1.0
3.9
4.7
2.6
2.9
62.9
269.7
C.tropi-
calis 11
5.5
2.2
3.4
6.8
4.3
3.9
9.8
3.7
4.9
3.9
0.9
4.0
5.3
2.5
4.2
65.3
285.4
of Candida
C.tropi-
calis 8
5.5
1.8
2.8
8.1
4.0
3.4
9.2
3.1
4.5
3.8
0.9
3.5
5.0
2.6
3.0
61.3
291.2
kind
C.utilis 3
5.1
1.8
2.7
10.8
3.8
3.1
5.3
2.4
4.2
2.3
1.0
2.9
4.2
2.1
2.4
58.9
246.7
Reference
Sulfite
Liquor Molasses
6.7
1.9
5.4
-
5.5
-
-
-
-
6.3
1.2
5.3
7.0
3.3
4.3
-
_
10.7
2.8
4.7
-
4.8
-
-
-
-
5.7
1.4
7.3
8.1
1.4
4.1
-
_
NOTE: Reference data is for pure Candida utilis grown on spent sulfite liquor
and molasses (Peppier, 144).
180
-------
The experiments have documented the beneficial role of the vitamins,
minerals and carbohydrates in algae, which when fed to swine (Chlorella) up
to 25 percent in the diet resulted in better weight gains (149).
TABLE 28. NUTRITIONAL VALUE OF PROTEINS
BASED ON METHIONINE DEFICIENCY
1.
2.
3.
4.
5.
6.
7.
8.
Protein From
Screenings
Activated sludge
Algal-bacterial biomass
Pure algae: Chlorella
Scenedesmus
Yeasts (this study)
Yeasts: sulfite liquor
molasses
Lemma minor
Against Whole
Egg
(%)
7-13
35-39
26-29
58
45
29-32
39
45
23
Against FAO
Standard
(%)
9-18
50-55
36-41
82
64
41-45
54
64
32
DISCUSSION AND CONCLUSIONS
In ideal conditions 1000 kg LW of beef cattle produces 1 kg/day of pro-
tein, 1000 kg soybean yield 100 kg/day, 1000 kg yeasts (DM) yield 100,000
kg/day and 1000 kg of bacteria may yield daily 100,000,000,000 kg of protein/
day (148). This short comparison although biased, since the wastewater based
crop is collected on year-around basis, as a whole plant with "roots," etc.
as opposed to land based yields, demonstrates the nutritional potential of
SCP as the substitute of other relatively inefficient agricultural products.
The SCP is synthesized in numerous countries from petroleum derived
hydrocarbons. Contrary to such sources, where the danger of introducing
carcinogens is quite apparent, the use of agricultural by-products for SCP
synthesis has to receive a much wider attention. The energetic potential
of animal wastes is 30 percent of the input feedstuff, sometimes in an un-
changed form.
181
-------
Numerous problems need to be. solved before, wider application of recycle
and SCP conversion will be practiced. It has already been demonstrated, for
instance that anaerobically digested dried pig wastes sludge yields definite-
ly negative effects when fed to swine.
As a rule, higher benefits are attained when recovered protein is fed to
another group of animals, as for example with dried poultry waste (DPW) fed to
cattle, or swine solids fed to cows and sheep.
The conclusions may be itemized as follows:
1. There are potential possibilities of direct recovery of protein in
waste solids from pig wastes. Complex animal hygienic studies are,
however, needed before full application, since the present experi-
ence is short-termed and narrow in scope although very encouraging.
The sterilization and/or drying is not the best method of screenings
preparation due to cost and odor problems. The desired alternative
is ensiling with other feedstuffs. The direct recycle should be in
open cycles, i.e. fed to animals of other kinds.
2. Based on protein content and digestibility analyses, the materials
recovered from piggery wastes can be classified as: proteinaceous
feedstuffs; excess activated sludge yeasts and Lemma minor; forage
feedstuffs; raw waste solids and algae.
3. The following determinations on the recovered material were made:
protein content (g protein/kg DM); screenings 34 to 37, activated
sludge 240 to 250, algal-bacterial biomass 110 to 192, yeasts 263
to 281, L.minor 216; with respective digestibility indices 45
percent, 62 percent, 50 percent, 63 percent, and 78.8 percent;
nutritional value compared to full egg AA content; screenings 10
percent, L.minor 24 percent, algae and bacteria 29 percent, yeasts
29 to 32 percent, and a.sludge 38 percent.
4. Based on analysis of AA content in the FAO reference protein and in
the SCP recovered from other substrates, it is concluded that al-
though the material recovered in this study have generally lower AA
content, they can be used as adequate feedstuff components.
182
-------
5. Algae SCP grown on piggery wastes are. an excellent food source with
high protein, vitamin and minerals content- The research should in
the future solve the problem of low digestibility of cell walls and
the difficulties in harvesting. At present the immediate application
would be in closing the cycle through aquaculture, fish cultivation,
as other methods are too expensive and not simple enough for the
rural environment.
6. Yeasts grown on piggery wastes have quality similar to the commercial
product. There are no additional cost benefits when the pig wastes
are used as substrate since large quantities of costly beet molasses
need to be added, while the danger of contamination of the culture
is much more eminent.
The possible more economical alternatives are: integration of waste
treatment and recovery systems of regular yeast manufacturing plant
and pig farm; and the use of mixed yeast cultures and the use of
yeast with low carbohydrate requirements; and the use of other inex-
pensive waste carbohydrate sources.
7. The most promising techniques of piggery wastes utilization involve
combined waste treatment and direct as well as SCP protein recovery
and industrial complexes that attempt at closing the nutrient and
food cycles: nutrients, SCP, aquatic animals, feed, and farm
animals.
8. Excess activated sludge is found to be of the highest value as
feedstuff, however, as the SCP source, it is of minor importance
since the activated sludge is becoming uneconomical for piggery
wastewaters and will have to be replaced in the future by low-
energy-consumption treatment systems.
183
-------
SECTION 12
EVALUATION OF VARIOUS WASTE TREATMENT AND RECOVERY SYSTEMS
OUTLINE OF THE ECONOMIC ANALYSIS
Economic optimization of waste treatment, recovery systems is a complex
process, which if to be done correctly, requires sophisticated mathematical
modelling and computer simulation of unit processes performance within the
treatment train. It is observed, however, that the effects of the sophisti-
cated approaches which include presumably all of the foreseeable variables
may not be representative, frequently due to an inadequate data base assump-
tion. In case of animal wastes the researchers are not consistent as to the
actual price of biogas and recovered protein, the costs of comparable unit
processes, etc.
Due to variations in the technological and constructional know-how of
treatment facilities, it is not possible to apply literature data to processes
derived in this work. For example, Genung et al. (51) calculates power use for
3
three plants (design flow 3780 m /d): activated sludge (AS), trickling filter
(TF) and anaerobic biofilter (AB), respectively as 963, 623, and 105 kWh/day
with sludge handling included. It occurs to the author of this report that:
a) removal efficiencies of each treatment train in this comparison are different,
and b) the power input for the AS is 1.5 times the power input for TF and over
9 times that of an AB. In another paper Mills and Tchobanoglous (162) quote,
for two plants (Q = 4250 m /d) with AS and TF both featuring anaerobic digestion
of sludge, the respective power use of 2671 kWh/day for an activated sludge
plant and 723 kWh/day for trickling filter treatment train, where as the AS
system requires 3.7 times more power than the TF system. The latter does not
compare well with the 1.5 times factor quoted by Genung et al.
Similar large discrepancies are evident in the literature concerning
strictly animal wastes treatment economics. For example, there is little
184
-------
agreement between various sources, on such items, in the economic analysis, as:
the power level for mixing the anaerobic digester, quantity of recoverable and
recovered energy, unit costs of gas production, and removal efficiencies (69,
91, 160, 162).
Another conclusion from the perusal of literature on recovery from animal
wastes is that economic conditions are so much different in various countries
that the comparison yields contrasting results. Mills (161) comparing an
aerobic and an anaerobic system for swine wastes has concluded that for Scotish
farms above 100 head capacity the anaerobic treatment for biogas recovery is
becoming economical. On the other hand, Hashimoto and Chen (91, 93) show that
anaerobic digestion (AD) for cattle wastes become feasible economically above
several thousand head capacity, U.S. conditions. Other authors show: the
feasibility of anaerobic digestion for piggery farms of 24,000 + head capacity
9
at natural gas costs of $2/10 J; and lack of economic justification for gas
9
recovery at gas costs of $0.8/10 J, for farms as large as 240,000 head capacity
(83).
The selection of the type of anaerobic digestion system is also not agreed
upon by writers: some favor the thermophillic process, others indicate that
there is more energy recovered from the long-term mesophyllic digestion.
Some writers use 10 to 14 days retention in a mesophyllic digestor, while
others base their calculations on 20-day retention. This is an additional
source of differences since the increased detention time results in more
efficient unit gas recovery, however, the overall economy may deteriorate
particularly in the flow-through digesters due to an increase in volume.
Summing this brief literature perusal: it is impossible to gain any
knowledge of the true economic efficiency of the anaerobic digestion as a
waste treatment method, with an additional gas recovery bonus, based on
literature data. Similar situations exist in literature on economics of SCP
recovery as noted in the chapter on protein recovery.
185
-------
It was then decided that a special economic analysis, will have, to be made
in this case of dilute piggery wastes from a typical industrial piggery farm.
Basic Assumptions
Almost all economic data reported on animal wastes deals with very con-
centrated or even semi-solid manures. In this project diluted piggery
wastewaters will be evaluated from the standpoint of economic efficiency of
their treatment for stream disposal. As opposed to the literature data
where efficiencies of individual anaerobic digesters and their gas production
are compared, full treatment trains will be evaluated with associated sludge
handling facilities. Uniform standards for effluent quality and final sludge
characteristics will be assumed for all compared systems. In this project
the profit from recovery has a secondary meaning; the primary objective being:
the decrease of the overall operation and maintenance costs, cutting down on
imported materials (fuel oil and chemicals) and obtaining low quantity of well
stabilized sludge, and acceptable year-around quality of treated effluent.
It is usually assumed that agricultural utilization of effluents, raw
or pretreated, is always the preferred disposal method, since it results in
the least harm to the environment and yields direct benefits. Jewell and
Loehr (159) show that the nitrogen value of manures is $0.013/hog/d while the
energy potential of piggery manure is only $0.0068/hog/d, which is two times
less. The net benefit is still smaller when the difficulties in recovering the
energy contained in significantly diluted wastes are compared with the relative
ease of nitrogen application to crops. The conditions at numerous large farms
make agricultural utilization impossible, and hence, this analysis deals with
full treatment before stream disposal of wastewaters from two types of farms.
One is a conventional farm of typical Agrokomplex technology, called Farm A,
2
with the basic wastewater generation of 28 dm /hog/da^nd 10,500 head capacity.
The other is a similar farm with a modified wastewater system, flushing of
3
sewers with treated wastes and wastewater generation of 20 dm /hog/d with 15,OOC
head capacity called Farm B.
It has been shown so far that the presently used methods of piggery
wastes treatment are inadequate because of low efficiencies attained, high
186
-------
process instability, resulting effluent variability, and mounting operational
costs. It has. also been documented that AD should be the basic biological
preparatory process that alleviates several of these problems, yielding an
effluent very well fit for further aerobic treatment. It is also fit for
land disposal as shown in our studies, the effluent retains 90 percent N and
P. Thus, the processes selected for comparison feature AD as a method of
either sludge stabilization or wastewater treatment or both. Eight basic
processes shown in Figures 65 through 70 were selected.
The diagrams depict only the unit operations that are included in the
economic analysis. As a reference, a basic Vidus treatment system VIII
(Figure 5 and 70) costs are used, calculated similarly as other systems based
on current 1980 prices.
Wastewater loads and removal efficiencies—
3
Hypothetical Farms A and B are characterized by flows Q. = 300 m /d =
QB; LA(COD) = 4200 kg 02/d, LB = 6000 kg 02/d; LA (TS) = 3013 kg/d, LB = 4305
kg/d: LA(BOD5) = 1430 kg 02/d, and LB(BOD5) = 2040 kg 02/day.
The removal efficiencies are based on our experimental data and are
expressed only by means of COD and BOD,.. Influent to effluent concentration
3
(mg 02/dm nonfiltered) ratios are: CODA - 14,000/200 and CODg - 20,000/200;
BOD - 4,770/45 and BOD,. - 6,800/45. The effluent qualities assumed in
j ,A J ,B
this analysis were attained in the course of this study.
Conversion efficiencies and unit costs—
Efficiency of gas/electric power conversion is combined with the efficiency
of the biogas fueled internal combustion engine and with the synchronous gener-
ator. Neyeloff and Gunkel (164) show the respective efficiencies as 16 percent
and 90.0 percent which yield a product of 14.5 percent. Hashimoto and Chen
(91) suggest an engine-generator efficiency of 38 percent as a maximum at 100
percent methane content, which for 60 percent CH, mixture yields overall ef-
ficiency of 23 percent. Recent work of Michelli (72) contains the most
reliable data from full scale operation of the TOTEM (Total Energy Module)
units at S.Agreste Farm at Todi, Italy, where 8600 pigs and 45,000 chickens
are raised and effluent undergoes combined anaerobic treatment. The TOTEM is
187
-------
based on Fiat 127 engine and has the efficiency of 25 percent gas/power con-
version and additional recovery of heat from the engine cooling water some
64 percent of the input energy. The writers have visited, the operation at
Todi and found that it yields enough energy to power and heat the treatment
plant while the excess is utilized in the farms production sector. The ana-
erobic digester effluent undergoes anaerobic lagooning and aeration in an
aerobic lagoon before being recycled to the farm.
In this analysis the 25 percent gas/power and 64 percent gas/heat conver-
3
sion ratios are assumed. The gas calorific value is assumed 23.45 MJ/m (M =
6 3
10 ), i.e. 5600 Kcal/m . The unit energy costs are given in Table 29. The
conversion for electric energy is assumed based on TOTEM experience as 3384
Kcal/kWh.
It should be noted that 0.50 zl/kWh is an average for 1978-1979, the
1980 lowest present price is 0.66 zl/kWh for city users and 1.50 zl/kWh for
individual use for heating during peak hours. However, two values will be
used, 0.50 and 1.50 zl/kWh. The costs in dollars and energy densities in
Table 29 are after Martin and Loehr (163). The analysis will be appropriate
to the Polish conditions, as the prices of certain commodities, such as fer-
tilizers, electricity, gas or coal on domestic market are set rigidly and
are usually low when compared to other goods or to the fluctuating prices in
the world market.
Based on these figures, the following recovery calculations can be made.
Energy recovery may be effected through: 1) direct sale of gas, then the
price of 1.20 zl/m could be assumed; 2) conversion to power, then one obtains
(5600 Kcal/m3 0.25 percent efficiency)/(860 Kcal/kWh = 1.63 kWh/m3), i.e.
0.81 to 2.44 zl/m and an additional 60 percent recovery of heat yields 3360
3 3
Kcal/m , which at 232 zl/Gcal is equivalent to an additional 0.78 zl/m ; and
3) direct use as heat in the plant, which at 70 percent efficiency of the
3 3
gas fired boiler yields 3920 Kcal/m or 0.90 zl/m , at city heat prices. In
a remote location of the piggery farm the above underestimates the actual
revenues from gas recovery.
188
-------
TABLE 29. UNIT ENERGY COSTS
Energy Source
Gasoline
(regular)
Fuel oil
Liquif.
petroleum gas
Coal
(anthracite)
Natural gas
Energy Density
35.6 MJ/dm3
40.1 MJ/dm3
25.5 MJ/dm3
30.0 MJ/kg
36.3 MJ/m3
$
USA
0.283 /dm3
0.268/dm3
0.171/dm3
0.083/kg
0.133/m3
Unit Costs
Zloties
Poland
16 zl/dm3
14 zl/dm3
1.18 zl/m3
0.55 zl/kg
1.14 zl/m3
Notes
78 octane
Mixture of
natural and
coal gas
Industrial
Electricity
City heat (coal
fired plants)
Nitrogen
(fertilizer)
Phosphorus
(fertilizer)
3.60 MJ/kWh 0.0434/kWh
0.5 to 1.5
zl/kWh
232 zl/Gcal
0.013/hog/d 10.4 zl/kgN
n.a.
8.5 zl/kg P
consumer may
pay up to,
2.40 zl/m
Valid for a
large city
NOTE: The prices quoted are sale prices based on 1980 averages.
equivalent to 20 zl.
1 U.S. $ is
189
-------
Methods, of analysis —
Equation 6 from Section 4 will be used here. The. method ass.umes approxi-
mately 10 years period of amortization of the capital investment:. The costs
and economic efficiencies of COD removal an.d of volume of flow removal
and EQ) are based on current 1980 data. The data in Table 3 are calculated
based on surveys of full scale plants erected between 1973 and 1976, thus,
the numbers cannot be directly compared with the results of this economic
analysis .
The recovery of biogas will be taken into account by subtracting from the
maintenance and operation costs. When applicable, nitrogen and phosphorus
costs will be substrated. The value of recovered protein is not taken into
account, which is consistent with the findings of this study.
COMPARISON OF VARIOUS SYSTEMS
The resulting economic efficiencies expressed in zl/kg COD , removed and
3 nr
in zl/m of wastewater treated are compiled in Table 32. The two variants
A-10,500 and B-15,000 head capacity, of one treatment system are calculated
at two power costs levels: 1 to 0.50 zl/kWh and 232 zl/Gcal; 2 to 1.50 zl/kWh
and 500 zl/Gcal. In order to save space only two examples of calculations are
given below and shown in Tables 30 and 31.
The costs of land is excluded from these considerations:
System I
The system (Figure 65) is based on studies presented in Section 10. The
lagoon system shall yield, in favorable conditions, effluent COD . = 380 mg/dm
* nr
and BOD,. ,. = 235 mg/dm . It is designed based on the following loadings:
3 3
anaerobic lagoon 0.5 kg COD /m /day, aerated lagoon 0.4 kg COD f/m /day,
oxidation ponds 0.08 kg COD /m /day. For variant A (10.5 thousand head),
these loads yield respectively 30 days, 10 days, and 20 days HRT. To arrive
3
at the design effluent COD and BOD^ - of respectively 200 and 45 mg 02/dm ,
1 cal = energy to heat 1 g water by IxC; 1 cal = 4.1868 J.
10 Btu - 10 J = 10 MJ = 1 GJ. 860 Kcal = 1 kWh (2850 J5
190
-------
sludge
AERATED
LAGOON
AEROBIC
STABILIZA-
TION POND
sludge (periodically)
SYSTEM I
H
SOIL
FILTER
gas
VO
GAS
TANK
2500
ACTIVATED
SLUDGE
1°
H
CLARIFIER
sludge
ACTIVATED
SLUDGE
SYSTEM
STORAGE
M
CLARIFIER
Figure 65. Layout of System I and II.
-------
soil filters are designed as a polishing treatment step. These should have
2
an area of 10,000 m (1 ha) and will yield, based on full scale data from
Plant D, Section 4, an effluent quality better than the design values on a
year-around basis. Sludge treatment will include storage, in the anaerobic
2
lagoon and dewatering in the drying beds of 5000 m area. The layout is
consistent with the now prevailing concept of rural treatment plants con-
sisting of earthen basins and simple structures that can be built by local
agricultural enterprises and that are the easiest to operate. It should be
noted, however, that the costs here are already high (Table 32) without even
accounting for the costs of land required by this low-efficiency system.
System II
This system (Figure 65) features an ANFLOW reactor with 10 d HRT followed
by a thickener (HRT = 1 d) and two-stage activated sludge tanks. The thickener
reduces the volume of sludge to be dewaterd by centrifuge and allows for sludge
recycle to ANFLOW in case of upsets and during the initiation of the anaerobic
process. The activated sludge process is designed at HRT = 3 d, Farm A, and
HRT = 6 d, Farm B, overall retention, with intermediate settling tanks. Staging
of the process and constant temperature T = 20 to 25°C allow to attain steady
effluent quality exceeding the design standards, temperature correction factor
1.053, K2Q = 6.6 d"1.
Capital costs —
In order to illustrate the calculation procedures, capital costs for
System II are presented in Table 30.
Further exemplary calculations for System II are performed for variant
A-l; the results for System II: A-2, B-l, and B-2 are given in Table 32.
With the excess activated sludge that undergoes gasification with raw wastes,
the value of SGP = 0.43 m /kg COD . (in) and the calorific value of gas =
~ o nf °
5600 Kcal/m are conservative assumptions.
Energy balance —
Electrical energy recovered is:
EEr = 0.43 X 4200 X 5600 X -- = 2980 kWh/d
m
192
-------
Heat energy recovered from the. TOTEM engines:
3 v i
HE = 1800 2- X 5600 *&r X 0.64 = 6.5 Gcal/d =* 2372.5 Gcal/yr
a J
m
The gross profit from recovered energy:
PE = 2980 X 0.5 X 365 - + 2372.5 — X 232 - = 1,095,000 zl/yr
Average heat losses are calculated as follows, based on temperature of
raw wastes 14 C:
3
HL = 0.023 £—•', T300 j- X 365 ^ X 232 -^- = 590,000 zl/yr
m
Electrical energy expenses for 100 kW installed power are calculated as:
EE = 110 kW X 24 ^ X 0.5 ^r- X 365 — = 482,000 zl/yr
exp d kwh yr
Operation and maintenance costs —
The total operation and maintenance costs (OM), were calculated as:
OM = DC + GC - (PE - HL)
where direct costs (DC) were calculated as:
DC = (material costs) + (labor costs) + EE
and general costs (GC) were calculated as:
GC = 0.4 DC = 766,400 zl/yr
Thus, the total OM costs are:
OM = 1.916 + 0.766 - (1.095 - 0.590) = 2.2 106 zl/yr
Economic efficiency indexes —
For System II - A-l, the EO index expressed per volume of wastewaters
treated in a year is:
., 41,500,000 (0.08 + 0.022) + 2,200,000 ,D , - , 3
EQ -- 300 . 365 58'7 Zl/m
The index expressed per COD ,. load removed (COD ) :
A TIX IT
(14 - 0.2) ^| X 300 j- X 365 — = 1,511,100 kg CODr/yr
m
F - 41,500,000 (O.Q8 + 0.022) + 2,200,000 _,..,, /lr& rnn
ECOD -- - 1,511,100 - - - ~ 4'24 Zl/kg C°Dr
193
-------
TABLE 3Q. CAPITAL COSTS FOR SYSTEM II
Unit
Volume Cost _ Equipment
Unit m 10 zl/m KT zl
Pump house
ANFLOW
Gas tank
Aeration
tanks
Clarifiers
Thickener
Recycle pump
Building
Centrifuge
Sludge
storage
Landscaping
Piping (m)
110 3.0 0.2
3000 4.0 0.5; 0.65
2500 2.2
900;
1800 2.0 0.2; 0.4
100 2.2
300 2.8
50 3.0 0.03
600 4.0
0.2
1300; 2
1700 m 0.4
_
1000 1.2
SUBTOTAL
Design cost, geological survey: 15 percent
SUBTOTAL
Unexpected purchases and labor: 20 percent
Including freezing
investment I1
GRAND TOTAL I (106)
of capital fi
= 1.08' I (10 )
Cost -
A
0.53
12.50
5.50
2.00
0.22
0.84
0.18
2.40
0.20
0.52
1.70
1.20
27.80
4.17
31.97
6.39
38.40
41.50
no6 zi
B
0.53
12.65
5.50
4.00
0.22
0.84
0.18
2.40
0.20
0.68
1.70
1.20
30.10
4.51
34.61
6.92
41.50
44.80
System III
The system is similar to System II in the primary, anaerobic part of the
treatment train. Following the ANFLOW reactor, the supernatant from the
thickener undergoes anaerobic biofiltration in the first stage ANBIOF 1°
where further biodegradation of liquified organics occurs. Thus, the gas
production is more complete than in System II; it is expected that it will
add at least 0.1 m CH,/kg COD removed at L = 2 kg COD/m /d; i.e., some 110 m3
biogas/d, removing at least 80 percent of the incoming COD - load. The effluent
from ANBIOF I > containing at the most 800 mg 00/dm3 COD ,, will be fed to an
*• nf
aerobic biofilter with high void ratio to maintain good aeration in the first
194
-------
layer of the biofilter media. As found experimentally by Oleszkiewicz and
23
Eckenfelder (175), the oriented plastic media at 100 m An specific surface
area increases the DO level to near saturation conditions in the first 0.5 m
3
of the filter height. The aerobic biofilter at L < 0.5 kg COD ,,/m /d should
— ni
yield at least 60 percent COD . removal and a well nutrified effluent with
ni A
COD - of approximately 300 mg/dm . Anaerobic biofiltration in ANBIOF II ,
judging by the results of Section 7 should yield an effluent COD below 150
3
mg/dm . Clarification and 1 hr aeration will be applied before discharge in
order to remove gaseous nitrogen and introduce oxygen to the effluent stream,
yielding further removal of COD.
The primary advantages of the system are the use of easily operated bio-
filters, two-stage, without phase separation, though anaerobic treatment for
more complete gasification, much better resistance to shocks due to high SRT
in the biofilters.
System IV
The system features short detention time anaerobic digestion in the ANCONT
reactors with 3.5 days HRT. Compared to System III, lower gas production is
expected in the ANCONT reactor than in the ANFLOW reactor and higher in ANBIOF
I than in System III. Based on our experimental evidence, it is expected that
the overall biogas production will not be much smaller in System IV. On
3
purpose, however, low overall gas recovery value is assumed 0.25 m /kg COD f
introduced, 50 percent smaller than in System III in order to account, in an
indirect way, for the higher level of operational difficulties with the ANCONT
reactor in System IV.
3
The treatment efficiency of the ANCONT reactor at L = 4 kg COD/m /d is
85 percent which compares favorably with 65 percent COD removal in ANFLOW
A
System III at L = 1.4 kg COD/m /d. The subsequent removal train: ANBIOF 1°,
aerobic biofilter, ANBIOF II in System IV is designed at 10 percent lower HRT
than in the System III.
It should be noted that in both treatment trains the actual performance
of the ANBIOF reactor will be higher than the 73 percent COD removal attained
195
-------
SYSTEM
ON
Figure 66. Layout of System III.
-------
in this study at 20°C, due to higher temperature, of the effluent from the
thickener, designed to conserve heat and recover gas.
System V and VI
Both systems start with a settling or thickening tank with HRT = 1 day
which feeds clarified wastes, through a heat exchanger to anaerobic biofilter,
ANBIOF 1° where 80 percent COD - removal is assumed at L = 3.0 kg/m /d in both
variants A and B. The actual removals that can be expected from ANBIOF I will
be higher, as the process will be ran at 35 C, i.e. higher than the experi-
mental data collected in approximately 20 C.
The secondary treatment in System V will be in a two-stage activated
sludge system designed for approximately 25°C for 1° and 18°C for II, with the
overall HRT in both stages equal to 2 days (A) and 3 days (B). In System VI
the aerobic biofilter will be used with a design loading of 1.0 kg/m /d for COD,
HRT equal to 1.4 and 2.4 days followed by an ANBIOF 11°, a polishing unit with
HRT = 0.67 and 1.0 days.
Both systems are treating sludge in a separate ANFLOW reactor (HRT = 15d)
3
designed to stabilize sludge and produce gas at a maximum rate of 0.5 m /kg
COD . introduced. This arrangement separates the digestion process into two
systems, the ANBIOF reactor treating low TS, low COD concentration wastewater
and the ANFLOW reactor treating sludge of at least 3 to 4 percent TS and thus,
results in optimum conditions for both systems. As expected, this shows in the
economic efficiency indices which are the lowest for System VI.
The capital cost calculation for System VI are presented in Table 31.
Systems VII, VH-a and Vll-b
These three systems compare efficiency of land disposal and agricultural
utilization of effluents. Systems VII and VH-a (Figure 69) feature anaerobic
pretreatment of clarified wastewaters for pathogens destruction and odor stabil-
ization and partial carbon removal followed by three months storage, a minimum
retention, if year-round, operation is to be practiced in the Polish conditions.
The sludge undergoes separate anaerobic digestion, and thus, maximum gas produc-
tion is effected by the system. System VII is calculated without accounting for
the value of recovered N and P while Vll-b includes these values.
197
-------
SYSTEM IV
vo
oo
ANBIOF
AEROBIC
BIOFILTER
H
ANBIOF
1°
gas
t»
STORAGE
1
GAS
TANK
Figure 67. Layout of System TV.
-------
SYSTEM V
Figure 68. Layout of System V.
-------
NJ
O
O
ANBIOF -
CLARIFIER
GAS
TANK
ANFLOW
15d
*
CENTRIFUGE
M
STORAGE
SYSTEM VI
STORAGE
LAGOON
90 days
LAND
DISPOSAL
GAS
TANK
ANFLOW
15 d
*
CENTRIFUGE
H
STORAGE
SYSTEM VII
Figure 69. Layout of System VI.
-------
TABLE 31. CAPITAL COSTS FOR SYSTEM VI
Unit
1
Pump house
Thickener
ANBIOF 1°
Clarifiers (2)
Aerobic biofilter
ANBIOF 11°
ANFLOW
Gas tank
Centrifuge
Sludge storage
Pump-recycle
Building
Landscaping
Piping
3
Volume (m \
or area (in )
2
220
300
840; 1200
115
420; 720
200; 300
800; 1000
800; 1000
-
1300; 1700
50
600
-
-
Unit Co§£
zl 10 /m
3
3.0
2.8
2.5
3.0
2.0
2.5
4.0
2.2
-
400
3.0
4.0
-
-
SUBTOTAL
GRAND TOTAL
Equipment
(10 zl)
4
0.40
-
-
-
-
-
-
-
0.20
-
0.03
-
-
-
(+15% + 20%)
Coat -
.A
5
1.06
0.84
2.10
0.35
0.84
0.50
3.20
1.80
0.20
0.52
0.18
2.40
1.70
1.20
16.88
23.50
I (106zl)
B
6
1.06
0.84
3.0
0.35
1.44
0.75
4.00
2.20
0.20
0.68
0.18
2.40
1.70
1.20
19.99
27.60
To arrive at an adequate irrigation dose, we have compared the data
quoted by Loehr (13) 26 g TKN/hod/d; by Overcash and Humenik (168) 0.230 kg
N/450 kg LW/d and 0.068 kg P/450 kg LW/d; with our own data reported in Section
4, and have assumed: 30 g N/head/d and 9.3 g P/head/d for this analysis.
The agricultural utilization was calculated based on Majdowski's elaborate
lysometric studies on piggery wastes (166) conducted from the standpoint of
effluent quality. From his long-term studies, conducted at various concen-
3
trations and doses, a 25 mm dose (250 m /ha/yr) was selected since at S =
3 ° 3
4000 mg/dm of BOD5> it yields an effluent (drainage) of 1.4 to 45 mg O./dm of
BOD-. Thus, an area of 430 ha for both Plants A and B is assumed, which at
53,000 zl/ha for a spray irrigation system (167) yields capital costs of 22.8
X 10 zl for the land disposal alone and an overall cost of 71.8 X 106 zl.
In practice this area could be doubled increasing the costs, since the recent
201
-------
USDA manual (165) recommends that only half of the nitrogen dose, be applied
as manure and the other half as artificial fertilizer in order to control
phosphorus runoff because, there is an excess of phosphorus, in piggery wastes.
System VII-b features land disposal only after 90 days retention in a
holding lagoon and no other pretreatment. The costs are significantly reduced
when compared with VII and Vll-a, however, are still much higher than the costs
of other treatment systems for disposal. This illustrates the reasons why
designers are becoming discouraged with large piggery farms: the economics
of land disposal were against this form of wastes disposal; the artificial
treatment trains as used so far were offering no better alternatives.
System VIII
This is the reference Vidus-type treatment system which has been made
comparable with Systems I through VI, and is calculated to bring comparable
effluent quality. The activated sludge is designed as a one-stage process at
HRT (A) = 3 d, and HRT (B) = 4.5 d and ANBIOF 11° HRT (A) = 1 d, and and HRT (B)
= 1.5 d. Full sludge stabilization is applied to bring sludge to quality
comparable with other systems.
Discussion
Table 32 lists the economic efficiency indices for all systems studied.
Systems V and VI are found to be the least expensive. It is characteristic
that the so-called "natural" treatment systems I and VII, Vll-a and VII-b
are more expensive than some of the new stream disposal systems. This means
that further wastewater volume reduction is necessary in order for the land
disposal systems to become economically efficient as the costs of fertilizers
are still low (Vll-a and VII-b). Anaerobic digestion is the preferred method
of pretreatment in all systems, including VII due to: energy recovery, 90 +
percent nitrogen conservation and pathogen stabilization.
It is noted that increased power costs beyond 1.50 zl/kWh, in the near
future will make systems with gas recovery even more attractive provided that
the ANCONT- or ANBIOF-type reactors are used.
202
-------
ISJ
O
Co
SYSTEM VIII
PRIMARY
AERATION
H
COAGULATION
CENTRIFUGE
SLUDGE
STABILIZATION
STORAGE
Figure 70. Layout of System VII and VIII.
-------
TABLE 32. ECONOMIC EFFICIENCY INDICES FOK THE STUDIED PIGGERY WASTEWATER TREATMENT SYSTEMS
System
1
I
II
III
IV
V
VI
VII
Vll-a
Variant
2
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
Power
Cost
Level
3
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Recovered
Gas
loV
4
1.8
2.4
2.2
3.0
1.1
1.5
2.2
3.C
2.2
3.0
1.8
2.4
1.8
2.4
Capital
Costs-1
106zl
5
30
36
38.4 \
41.5
40.0
42.0
24.5
27.0
22.7
27. C
23.5
27.6
71.8
75.1
71.8
. 75.1
Recovery
Power +
Heat PE
106
6
n.a.
n.a.
n.a.
n.a.
1.10
2.80
1.45
3.70
1.35
3.45
1.80
4.65
0.70
1.75
0.90
2.35
1.30
3.25
1.80
3.30
1.35
3.45
1.80
4.65
1.10
2.80
1.45
3.75
1.10
2.80
1.45
3.75
Power
Consumption Efficiency
EE
exp
zl/yr
7
0.45
1.30
0.60
1.70
0.50
1.45
0.50
1.45
0.20
0.50
0.20
0.50
0.20
0.50
0.20
0.50
0.50
1.45
0.50
1.45
0.20
0.50
0.20
0.50
0.20
0.50
0.20
0.50
0.20
0.50
0.20
0.50
EQ
zl/m3
8
43.9
54.8
52.6
67.3
58.7
61.5
59.3
56.5
53.9
45.7
52.2
36.7
40.6
41.1
40.7
33.0
39.1
39.2
38.9
31.2
35.5
35.5
36.0
20.4
98.7
93.2
100.2
90.1
85.0
79.5
80.6
70.1
ECOD
zl/kg
9
3.2
4.0
2.7
3.4
4.2
4.4
3.0
2.9
3.9
3.3
2.6
1.8
3.0
3.0
2.0
1.9
2.8
2.8
2.0
1.6
2.6
2.5
1.8
1.0
7.1
6.7
5.1
4.5
6.1
5.8
4.1
3.5
Notes
10
activated
sludge
ANBIOF and
activated
sludge
Without
N and P
bonus
N and P
bonus
-------
Table 32 Continued
o
tsi
System
Variant
Power
Cost Recovered Capital
Level Gas Costs-1
103m3 106zl
1
Vll-b
VIII
2
A
B
A
B
345
1
2
1
2
1
2
1
2
58.5
59.6
19.2
22.2
Recovery
Power +
Heat PE
106
6
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n. a.
n.a.
Power
Consumption Efficiency
EE
exp
zl/vr
7
0.20
0.50
0.20
0.50
0.55
1 . 70
0.55
1.70
EQ
zlM3
8
67.1
70.3
64.6
68.0
47.6
62.2
50.7
65.3
ECOD
Zl/kR
9
4.8
5.1
3.3
3.4
• 3.4
4.5
2.6
3.3
Notes
10
N and P
bonus
Modified
Vidus
system
;NOTE: Heat loss is 1) 0.59 10 ; and 2) 1.26 10 zl/r, both variants.
i land (430 ha) and power for irrigation by spraying.
1 U.S. $ is equivalent to 20 zl.
Costs for Vll-VII-b do not include
-------
The secondary treatment systems utilizing activated sludge are very costly
because of high, power consumption, difficult to operate and should not be
included in the piggery wastes treatment plants.
System's I principal advantages are low level of sophistication in con-
struction and in operation, resistance to shocks and power failures (gravity
flow). All these factors are very important in rural conditions. The disad-
vantages are lack of recovery incentive and power costs.
System II features perhaps the most complicated operation and treatment
process vulnerable to shocks, however, offers an advantage of combining waste-
water and sludge treatment in one train.
System III offers similar advantages to System II, and a much simpler
operation of the secondary system, larger specific biogas production and high-
rate low-volume secondary treatment in an aerobic biofilter and an anaerobic
polishing biofilter.
System IV offers much small volume of the anaerobic digester and higher
removal efficiency than System III, thus, a more stable operation of the sub-
sequent treatment units.
Systems V and VI offer separate sludge digestion and the easiest to operate
waste treatment trains. The arrangement (patent applied for by authors) now
implemented at Farm A allows for rapid removal of soluble pollutants in a
series of biofilters with biogas bonus and significant denitrif ication in
ANBIOF 11° and through recycle of aerobic biofilter effluent to the ANBIOF 1°.
The costs of the two Systems V and VI are the lowest of all systems compared.
The beneficial effects of increased concentration of wastes and biogas recovery
are apparent at the high electricity cost level which for System VI B-2: 20.4
3
zl/m and 1.0 zl/kg COD as compared to A-2 (high-cost low-concentration)
' '
n
35.3 zl/m and 2.5 zl/kg CODrem (Table 32). These indices are in great con-
trast with the Vidus-type plant (VIII) which for E-2 are 65.3 zl/m , and 3.3
zl/kg COD , and for A-2 are 62.2 zl/m3 and 4.5 zl/kg COD
rem ° rem
206
-------
Finally, it should be noted that the presently us.ed aerobic, systems, like
System VIII and the land disposal Systems Vll-a and YH-b, here calculated
with underestimated power costs, will show an increase in, the index values due
to increasing power costs while gas recovery Systems II through VI will improve
the economic indices since the rising power costs usually decrease their value.
The use of the System VI makes the plant independent of the fluctuation of
volume of wastewaters since biofilters are less vulnerable to hydraulic over-
loading than suspended sludge reactors.
COMBINED TREATMENT WITH OTHER EFFLUENTS
Large piggeries are frequently sited close to large municipalities and
within industrial complexes or areas otherwise unfit for land disposal. Three
possible solutions to combined treatment of animal wastes result from these
studies presented here so far:
a. feed yeasts production with the use of piggery wastes and combined
wastes treatment;
b. combined anaerobic digestion with municipal sludge for biogas
recovery; and
c. combined treatment with nutrient deficient industrial wastes.
Yeast Production
The studies of Candida yeasts production on piggery wastes with and with-
t
out the addition of easily available carbon have proved much higher nutrients
(C, N, and P) utilization in the runs with carbon supplementation. Technical
feasibility of yeast production was verified in semidynamic tests, however, it
should be noted that at present the high costs of molasses and unusually low
costs of yeasts make the concept economical only when the two plants are within
one industrial complex, and when there is the benefit of joint management of
recovered gas and waste heat. In the conceptual layout in Figure 71 piggery
wastes are clarified, and the sludge is thickened or centrifuge is an alterna-
tive and fed to the combined mesophyllic anaerobic digestor operating as an
ANFLOW1 reactor. The raw supernatant or centrate is directed to the yeast
plant as N and P rich substrate, carbon is fed as 6600 kg/d of molasses for
yeast production. Effluent (centrate) from yeast production is fed to the
ANFLOW for gas generation and stabilization. The ANFLOW effluent is subject
207
-------
DISTILLERY
AND/OR
YEAST PLANT
DEWATERING
AEROBIC
POST-TREATMENT
METHANE
DIGESTION
SUPERNATANT
cc
UJ
UJ
o
LU
THICKENING ;
DEWATERING
COMPOSTING
LAND
DISPOSAL
Figure 71. Layout of combined wastes management
farm and yeast plant.
- recycle system for pig
208
-------
to separation into liquid phase whi.ch undergoes further treatment, while the
solid phase after dewatering and composting is used in agriculture.
Recycle of treated wastewaters for pig farm use and recycle of waste
heat from the yeast plant, as well as full utilization of all carbonaceous
matter generated at both plants into methane generation increase the economic
feasibility of this joint venture.
3
The efficiency indices are calculated based on 300 m /d piggery wastes
flow for variant A-l assuming System V, Figure 68 and Table 33. The fodder
yeast plant is calculated as using and discharging 300 m /d of water. In a
3
combined plant where raw wastes are fed, an effluent will be 400 m /d. The
values of Ert = 34.7 zl/m and 2.5 zl/kg COD ,, a considerable difference
Q . » e removed
from the separate plants when the fact of decreased volume of wastes is taken
into account. Thus, although the use of yeast production as a method of
piggery wastes treatment is still uneconomical, in a combined production,
waste treatment, recovery system the economics are in favor of combined pig
wastes fermentation. The benefits are:
a. decreased use of water by 50 percent and decreased volume of
wastewaters;
b. treatment of very concentrated and difficult to treat aerobically
effluents from yeast plants;
c. higher treatment efficiency due to better C/P/N ratio in yeast
wastes;
d. rational use of waste heat from yeast plant;
e. better use of nutrients in pig wastes and higher carbon removal than
in case of separate treatment systems; and
f. increased energy recovery and decreased investment, operation and
maintenance costs.
As found in this work, anaerobic processes are the best for both full
treatment and pretreatment before agricultural wastes disposal. This is
utilized here as shown in Figure 72. Municipal sewage is transported by
gravity or pumped to the combined treatment plant (CTP) located 2 to 5 km
from the town. At the CTP sewage undergoes screening, grit removal, primary
209
-------
TABLE 33. COST OPTIMIZATION FOR THREE COMBINED TREATMENT PLANTS
Capital
W.T. Plant Costs
Type
Piggery WTP
(System V)
Yeast fact.
W.T. P.
Combined
W.T. P.
Piggery WTP
Municipal
WTP
Combined
WTP
Piggery WTP
(System VIII)
Chemical
factory WTP
Combined
WTP
Flow
in /d
300 23
300 23
400 29
300 23
2,500 50
2,800 67
300 20
25,000 210
25,300 230
Economic
0 - M Recovery Power Net Efficiency
Coats EE, HE Use Profit Indices
EQ
106 zl/yr zl/m3
2.5 1.3 0.5 0.7 39.3
2.5 1.3 0.5 0.7 39.3
3.1* 2.0 0.57 1.22 34.7
2.5 1.3 0.5 0.7 39.3
4.5 0.5 0.87 0.03 11.0
5.0 0.43 0.88 0.45 11.8
3.1 - 0.55 - 47.6
10.0 - 5.7 - 3.4
10.0 0.3** 5.7 - 3.6
ECOD
zl/kg
COD
2.8
2.8
2.5
2.8
16.3
2.2
3.4
4.8
4.0
*Profit from N and P recovered in yeast biomass (1.5 million zl) not included.
**Phosphorus recovery included only 90 kg P/d 0.3 mln zl/yr.
1 Dollar is equivalent to 20 zl
210
-------
MUNICIPAL
SEWAGE
COMPOSTING
PRIMARY
AERATION
LU
O
Q
13
CO
V
X
x
PRIMARY
CLARIFIERS
miiiiir
ANAEROBIC
DIGESTOR
o.
Z)
CO
THICKE-
NING
I
ce
LU
CO
DEWA-
TERING
^
r
r^
RETENTION
TANK
LAND
DISPOSAL
Figure 72. Concept of combined municipal - piggery wastes treatment system.
211
-------
aeration, settling, activated sludge treatment and final clarification. The
primary and excess activated sludges are mixed with piggery wastes and fed
to the anaerobic flow-through digester (ANFLOW) operated conventionally for
maximum biogas recovery (HRT - 10 days). Effluent from ANFLOW is thickened
and dewatered for composting while supernatant is directed to an equalization
tank for agricultural disposal or to the aerobic treatment train. In the
latter situations the aeration tank has to be preceded by an anoxic compart-
ment to account for excessive concentrations of nitrogen in wastes stream.
Anoxic zone is equipped with mixers for contacting incoming wastes with
carbon-deficient recycled sludge.
3
The economic comparison for CTP serving 15,000 inhabitants (2,500 m /d)
3
and a farm with 10,500 hogs (300 m /d) is presented in Table 33.
The capital costs may be further diminished by substituting the ANCONT
in place of the ANFLOW digester although maintenance costs may go up and
biogas recovery will be less efficient.
The combined municipal pig wastes systems are recommended for wider use in
Czechoslovakia (152, 157). The city of Trebon (30,000 population equivalents)
and a 25,000 hog farm have erected a CTP which, in an initial phase of operation,
3 3
has attained 3 m CH,/kg TS/d producing some 4000 m /d of biogas at 67 percent
methane. With the foreseen increase of pig farm to 60,000 head capacity, the
plant operators expect to attain 0.150 m /hog/d of biogas and the aeration
power input 2W/hog.
It should be noted that high content of nitrogen in the supernatant
poses serious problems. A system as proposed here in Figure 72 with an anoxic
sludge compartment offers a solution, however, adequately large capacity and
skillful maintenance is required. A conventional activated sludge system,
as used in Trebon (152) will not remove nitrogen to the degree allowing for
3 3
stream disposal, since the supernatant may contain 10 to 15 g/dm TS, 3 g/dm
3
TKN and up to 5 g/dm BOD_ -.
o,ni
212
-------
Combined Treatment with Industrial
A combined treatment plant in Czechoslovakia treats 120, m /d of pig waste
with 60,000 m /d of kraft pulp effluents (152). In another case, analyzed in
this project, an organic chemicals plant treats it wastes with activated sludge.
The performance of aeration basin, without addition of nutrients, was some 30
percent, thus, the chemical plant has to spend close to 0.9 10 zl/year on
^ fi
artificial fertilizer in the first stage (25,000 m /d) and some 1.7 10 zl/year
3
in the second stage (75,000 m /d).
Similar systems are designed by authors at a 24,000 head capacity
Agrokomplex farm which produces 600 to 800 m /d of wastes, twice the existing
Vidus plant hydraulic capacity. In order to accomodate these wastes a pressure
3
(150 mm ID) pipe was designed to pump the excess 300 m /d of screened, pre-
aerated piggery wastes to a chemical plant located 11 km away (see Figure 73).
The chemical plant nutrient demand: 320 kg N/d and 90 kg P/d (I stage)
will be adequately covered by raw pig wastes. The analysis in Table 33 does
not include savings due to lower costs of equipment feeding nutrients ready
to the aeration tank and includes only the cost of phosphorus.
The benefits are apparent and show an excellent route for the use of
animal wastes as natural nutrients in waste treatment of unbalanced industrial
wastes.
CONCLUSIONS
The so-called "natural" treatment systems featuring lagoons, oxidation
ponds earthen structures and agricultural utilization (land disposal), Systems
I and Vll-b, have been proved less economical for the modern large scale
3
industrial pig farm, which use water in excess of 20 dm /hog/d, than the new
systems proposed in this project utilizing full biogas recovery and anaerobic
treatment, Systems V and VI.
Systems I and Vll-b are, however, still more economical than the presently
used chemical-biological systems (System VIII) and some of the anaerobic treat-
ment systems now proposed by various sources.
213
-------
pH CONTROL
INDUS! WASTE
EQUALIZATION
AERA
["ION
o oOj. o
PIG. WASTE
ACTIVATED
SLUDGE TANKS
Figure 73. Concept of combined treatment with chemical industry wastes.
214
-------
The high concentration of nutrients in. piggery wastes inak.es. these effluents
ideal substrate, for combined treatment with high-volume, low-concentration,
nutrient deficient industrial wastes.
Location of pig farms close to municipalities creates an opportunity of
combining the separated solids from municipal sewage with pig wastes for a
more efficient biogas recovery operation.
The location of pig farms within the complexes of other agricultural
industry segments creates an opportunity to combine other concentrated effluents
for gas recovery, waste heat utilization operations in ANFLOW or ANCONT type
digesters.
Analyzing the trends in neighboring countries with similar animal wastes
problems (152) and the recent recommendation in Poland (158), as well as the
results of this economic analysis, it is conluded that the pig farms should have
a capacity much lower than 10,000 head if agricultural utilization is to be
practiced. If larger farms are erected or in cases where land disposal is not
feasible, the piggery wastes should be treated in combination with other indus-
trial effluents or municipal sewage. Due to increased size of the combined
waste treatment facilities and significant generation of marketable products,
biogas and digested sludge or compost, better quality of operation can be
maintained than in local plants. The decrease in capital costs is usually
significant (e.g. 20 to 25 percent) in spite of frequent need for transporting
the wastes to the CTP site. The decrease in running costs, even without
accounting for the recovered biogas, for the joint treatment facilities may be
as large as 20 to 50 percent of the sum of these costs at individual plants.
In all cases advanced biogas recovery high-rate treatment systems should
be applied, such as the AflBIOF 1° AEROBIC BIOFILTER - ANBIOF 11° System VI,
which are more efficient technologically and economically than the presently
used treatment systems and will become the only alternative with the increasing
energy shortage.
215
-------
REFERENCES
1. Anon. Statistical Annual. Main Statistic Office, Warsaw, 1979.
2. Anon. Pig Production in the Netherlands. Public. E 102, Min.
Agriculture and Fisheries, Hague, AVB-EB-1975.
3. Kutera, J. and M. Kutera. Utilization of Liquid- jtnure from Large
Animal Farms in Poland. (Great Britain) 1981. (in print) Agricultural
Wastes.
4. Dragun, W. Quantitative and Qualitative Characteristics of Manure from
Closed Cycle Pig Farms (in Polish). Proceed. Rural Design Off. Assoc.,
Szczecin, May, 1978.
5. Wieckowicz, E. and S. Tereszczuk. Industrial Methods of Pig Production
(in Polish), PWRL, Warsaw, 1978. 283 pp.
6. Zebrowski, Z., et al. Industrial Scale Hog Production in Poland (in
Polish). Nowe Rolnictwo, 13, 24-29, 1978.
7. Poelma, H. R. De opslag van dunne mest. IMAG, Wageningen, Publ. 97,
1978.
8. Sobieracki, Z., et.al. Water Use in Animal Husbandry (in Polish); Gaz,
Woda i Techn. Sanit., _52, 8, 245-247, 1978.
9. Koziarski, S., et al. Upgrading of the Vidus-Type Piggery Wastewater
Treatment Plant (in Polish) National Agriculture Wastes Symposium,
Slupsk, 24-26 September, 1979.
10. Melaniuk, J. and Z. Suligowski. Water Requirements of Large Animal
Farms.(in Polish). Gaz, Woda i Techn. Sanit., 52, 3, 78, 1978.
11. Oleszkiewicz, J. A., S. Koziarski, and Z. Kwiatkowski. Chemical-Biological
Treatment of Piggery Wastewaters. Environment Protection Engineering,
J5, No. 2, 109-125, 1979.
12. Dragun, W., et al. Wastewater Quality and Quantity at Farm B. (Manu-
script). Institute of Irrigation, Wroclaw, 1977.
13. Loehr, R. C. Agricultural Waste Management. Academic Press, Inc., New
York, 1974.
216
-------
14. Taiganides, P. and R. K. White. Typical Variations Encountered in the
Measurement of Oxygen Demand of Animal Wastes. Cornell University
Conference on Agricultural Waste Management, Jan. 13-15, 1969.
15. Matysiak, A. Evaluation of Imported Piggery Wastewater Treatment Plants
in Poland (in Polish). Przeglad Hodowlany, 44., 5, 12-15, 1978.
16. Evans, F. L. Summary of National Operational and Maintenance Cause and
Effects Survey, U.S. EPA Technological Transfer, July, 1979.
17. Fuss. E. and D. SUssenbach. Mechanisch-Biologische Aufvereitung von
SchweinegUlle im VEG Halle - Nord. Reprint No. 60, DDR, 1975.
pp. 468-471.
18. Adam, F. Dewatering of Fine Suspensions by Means of Settling Centrifuge.
(in German) Environment Protection Engineering, 4, 1, 5-26, 1978.
19. Harper, J. P., et al. Enhanced Treatment of Livestock Wastewater (by
solids removal). J. Agric. Eng. Res., Vol. 19, 1974. pp. 353-363.
20. Oleszkiewicz, J. A. Rational Design of High-Rate Trickling Filters,
Based on Experimental Data. Environment Protection Engineering, 2, 2,
86-105, 1976.
21. Oleszkiewicz, J. A. Anaerobic and Aerobic Biofiltration of Agricultural
Wastes (in print). Agricultural Wastes, Vol. 2, 1980.
22. Heidrich, Z. Initial Analysis of Economic Efficiency of Piggery Waste
Treatment Plants (in Polish). XIX Konfer. Nauk. Techn. NOT, Katowice,
1976. pp. 240-250.
23. Roman, M. and Z. Heidrich. Economics of Large Piggeries Wastewater
Treatment (in Polish). Nat'l. Agric. Waste Mgmt. Conf. PZITS-NOT,
Slupsk, 1976. pp. 117-133.
24. Anon. Methods of Evaluating Economic Efficiency of Nonproductive
Enterprise. Planning Commission, Board of Ministers of Poland (in
Polish). 26 VII 1974.
25. Kalisz, L., et al. Activated Sludge Treatment of Piggery Wastewaters.
RIED Papers (in Polish). Warszawa, 1978. pp. 23-37.
26. Oleszkiewicz, J. A. Sedimentation of Suspended Solids from Piggery
Wastes. Environment Protection Engineering. _5, 4, 313-321, 1979.
27. Oleszkiewicz, J. A. Theory and Design of High-Rate Plastic Media
Trickling Filters. Progress in Water Technology £, Pergamon Press,
1977. pp. 777-785.
28. Szymanski, S. Screening of Solids from Piggery Wastes (in Polish).
Nat'l. Agric. Waste Mgmt. Conf., PZTIS-NOT, Slupsk, September, 1978. pp. 155-
168.
217
-------
29. Humboldt-Vedag. Dewatering Piggery Waste Sludges with a Solid Bowl
Centrifuge. Tests Conducted at Plant A. RIED, Wroclaw (manuscript and
data), 1977.
30. Oleszkiewicz, J. A. and M. SozaAski. Factors Affecting the Dewatering
Efficiency of Chemical Sludges. Environment Protection Engineering _3_»
1-2, 55-78, 1977.
31. Anon. Methods for Chemical Analysis of Water and Wastes. U.S. EPA,
16020-07/71, Cincinnati, 1971.
32. Anon. Determination of Chemical Oxygen Demand by Permanganate
Method. Polish Standard: PN-74/C-04578.03.
33. Anon. Standard Method of Test for Particulate and Dissolved Matter
in Industrial Wastes. USA ASTM Designation D1069-66.
34. Anon. Determination of Nitrite Nitrogen Colorimetrically with
Sulphanyl Acid and 1-Naphtylamine. Polish Standard PN-73/C-04576.06.
35. Anon. Determination of Nitrate Nitrogen Colorimetrically with
Phenyldisulphonyl Acid. Polish Standard PN-73/C-04576.08.
36. Anon. Standard Methods for the Examination of Water and Wastewater.
13th Ed., APHA, AWWA, WPCF, Washington, 1971.
37. Murray, I., J. W. Parsons, and K. Robinson. Interrelationships Between
Nitrogen Balance and DO in an Oxidation Ditch Treating Farm Animal Waste.
Water Research £, 1975. pp. 25-30.
38. Taiganides, E. P. and K. R. White. Automated Handling and Treatment of
Swine Wastes. Research Report (reprint) U.S. EPA Grant, No. 13040 EOL,
OSU Research Foundation Project 2980, 1972.
39. Owens, J. D., M. R. Evans, M. R. Thacker, R. Hisset, and S. Baines.
Aerobic Treatment of Piggery Waste. Water Research ^, 1973. pp. 1745-
1766.
40. Bartoszewski, K. and M. Soroko. Oczyszczanie sciekow z ferm tuczu
trzody metoda osadu czynnego. Proceed. XVIII Scient. Techn. Confer.,
NOT-PZITS-BPBK, Katowice, 1975.
41. Oleszkiewicz, J. A., S. Koziarski, A. Janiczek, and K. Kosinska. Nowe
Technologie w oczyszczaniu sciekiw z ferm hodowlanych (in Polish).
Proceedings X Techn. Conference "Dni Techniki," Jelenia Gira, 1979.
42. Oleszkiewicz, J. A. A Note on Protein Deficit and Environmental Engineer-
ing. Environment Protection Engineering, 4_, 4, 1978. pp. 359-361.
43. Goldman, J. C. Outdoor Algal Mass Cultures-II. Photosynthetic Yield
Limitations. Water Research, 13, 1979. pp. 119-136.
218
-------
44. RogiAski, J. and G. Fafara. Oczyszczanie gnojowicy przez sedymentacje,
napowietrzanie oraz hodowle glon6w (manuscript in Polish). Institute
IBMER, Warsaw, 1975.
45. Schroeder, G. L. Agricultural Wastes in Fish Farming. A Commercial
Application of the Single-Celled Organisms for Protein Production.
Water Research, U, 1977. pp. 419-420.
46. De Pauw and L. Leerhear. Mass Culturing of Marine and Freshwater Algae
on Aerated Swine Manure. Conference Proceedings. "Cultivation of Fish
Fry and its Live Food." Szymbark, Poland, September 23-28, 1977.
47. Boersma, L., et al. Protein Production Rates by Algae Using Swine
Manure as a Substrate (in) Energy, Agriculture and Waste Management.
Ann Arbor Scien. Publ., 1975. pp. 475-493.
48. Frostell, B. Anaerobic Filter Treatment of a Dilute Synthetic Waste-
water. Vatten, 35/3, 1979. pp. 169-175.
49. Anderson, G. K. and A. B. Ibrahim. Treatment of High Nitrate Waste-
waters by Plastic Media Anaerobic Filters with Reference to Latex
Processing. Prog. Wat. Techn., 10, 5/6, 1978. pp. 237-253.
50. Mosey, F. Anaerobic Filtration: A Biological Process for Warm Trade
Wastewaters. WRC Technical Report 48, July 1977.
51. Genung, R. K., et al. Pilot-Plant Demonstration of an Anaerobic Fixed-
Film Bioreactor for Wastewater Treatment. Biotechnology and Bioengineer-
ing Symposium. No. 8, (J. Wiley), 1978. pp. 329-344.
52. Chmielowski, J. Kinetics and Mechanism of Methane Digestion (in Polish).
Zesz. Naukowe Polit. Slaskiej, No. 165, Gliwice, 1966.
53. Chmielowski, J. Principles of Biochemical Changes in Anaerobic Digestion
(in Polish). Prace Nauk. Inst. Inz. Ochr. Srodow. Wroclaw Technical
University, No. 20, 1973.
54. Young, J. C. and P. L. McCarty. The Anaerobic Filter for Water Treatment.
Journal Water Pollution Control Fed., 41, 5, P-140-R, 1969. 173 pp.
55. Oleszkiewicz, J. A., S. Koziarski et al. Piggery Wastewater Treatment
in Poland: Full Scale Practices and Promising Future Techniques.
Proceedings of the IV Int'l Symposium on Livestock Wastes/1980. ASAE.
56. Oleszkiewicz, J. A. and S. Koziarski. Management and Treatment of Wastes
from Large Piggeries (in print). Agricultural Wastes, 3, 1981. pp.
123-144.
57. Cillie, G. G., M. R. Henzin, G. J. Stander, and R. D. Bailtie. Anaerobic
Digestion. IV Application of the Process in Waste Purification. Water
Research, 3., 1969. pp. 623-644.
219
-------
58. Eckenfelder, W. W. and D. L. Ford. Water Pollution Control. Jenkins
Publ. Co., Austin, 1970.
59. Andrews, J. F. Dynamic Model of the Anaerobic Digestion Process. Sanit.
Eng. ASCE, 95_, SAI, 95, 1969.
60. McCarty, P. L. Anaerobic Waste Treatment Fundamentals. Part IV:
Process Design; Public Works, No. 12, 1964. pp. 95-99.
61. Buswell, A. M. and H. F. Mueller. Mechanisms of Methane Fermentation.
Industr. and Engin. Chem., 44, 1952. pp. 550-552.
62. McCarty, P. L. Anaerobic Waste Treatment Fundamentals. Part I: Chem-
istry and Microbiology: Public Works, No. 9, 1964. pp. 107-112.
63. Levenspiel, 0. Chemical Reaction Engineering. J. Wiley and Sons, Inc.,
New York, 1972.
64. Lawrence, A. W. and P. L. McCarty. A Unified Basis for Biological Treat-
ment Design and Operation. J. Sanit. Eng. Div., ASCE, 96, SA3, 1970.
pp. 757-778.
65. Speece, R. E. and P. L. McCarty. Nutrient Requirements and Biological
Solids Accumulation in Anaerobic Digestion. Proceedings First Int'l.
Conf. Water Pollut. Research, Pergamon Press, 1962 and 1964.
66. Metcalf and Eddy, Inc. Wastewater Engineering. McGraw-Hill Book Co.,
New York, 1974.
67. Fischer, J. R., D. M. Sievers, and C. D. Fulhage. Anaerobic Digestion
of Swine Wastes. In "Energy, Agriculture, and Waste Management." Ann
Arbor Scien. Publ., 1975. pp. 307-316.
68. Morris, G. R., et al. Alternative Animal Waste Anaerobic Fermentation
Design and Their Costs. In "Energy, Agriculture and Waste Management."
Ann Arbor Scien. Publ., 1975. pp. 317-335.
69. Kroeker, E. J., et al. Cold Weather Energy Recovery from Anaerobic
Digestion of Swine Manure. In "Energy, Agriculture and Waste Management."
Ann Arbor Scien. Publ., 1975. pp. 337-352.
70. Summers, R. S. and S. Bousfield. A Detailed Study of Piggery Waste
Anaerobic Digestion. Agricultural Wastes. ^, 1980. pp. 61-78.
71. Hobson, P. N. and B. G. Shaw. The Anaerobic Digestion of Waste from an
Intensive Piggery Unit. Water Research. 1_, 1973. pp. 437-449.
72. Michelli, A. Possibilities for Energy Recovery in the Treatment of
Effluents from Distillery and Feedlot Operations. Environment Protec-
tion Engineering, £, 3, 1980.
220
-------
73. Pipyn, P. and W. Verstraete. Waste Classification for Digestibility in
Anaerobic Systems. First Symposium on Anaerobic Digestion and Methane
Gener. Cardiff, Wales, 1979.
74. Hawkes, D., et al. The Application of Anaerobic Digestion to Producing
Methane Gas and Fertilizer from Farm Wastes. Process Biochem., No. 3,
1976. pp. 32-36.
75. Hart, S. A. Digestion Tests of Livestock Wastes. Journal Water Pollut.
Control Fed., ^5, 6, 1963. pp. 748-757.
76. Humenik, F. J., et al. Cleaning the Animal Farm Environment. Technology,
JJ, 984, 1974.
77. Solly, R. K. A Study Methane Digesters in the South Pacific. Appropriate
Technology, _3, 4, 1977. pp. 23-25.
78. Sparling, A. B. University of Manitoba, Interdisciplinary Project -
Energy Recovery from Agricultural Wastes. Proceedings Int'l Biomass Energy
Conference, Winnipeg, Manitoba, Canada, 1973.
79. Longsdom, G. How Taiwan Homestead Farms Make Methane Work for Them.
Compost. Science, 16, 5, 30, 1975.
80. Morris, R. G., W. J. Jewell, and R. C. Loehr. Anaerobic Fermentation of
Animal Wastes. Design and Operational Criteria. In: Food, Fertilizer
and Agricultural Residues. Ann Arbor Science. Publ., Inc., 1977.
pp. 395-412.
81. Berg van den, L. and C. P. Lentz. Methane Production During Treatment of
Food Plant Wastes by Anaerobic Digestion. In: Food, Fertilizer and
Agricultural Residues. Ann Arbor Science Publ., Inc., 1977. pp. 381-393.
82. Yang, P. Y. and K. K. Chan. Anaerobic Digestion of Poultry Waste With
and Without Acid Hydrolysis. In: Food, Fertilizer and Agricultural
Residues. Ann Arbor Science Publ., Inc., 1977. pp. 423-435.
83. Ifeadi, C. N. and J. B. Brown. Technologies Suitable for Recovery of
Energy from Livestock Manure. In: Energy, Agriculture and Waste Manage-
ment. Ann Arbor Science Publ., Inc., 1975. pp. 373-396.
84. Wong-Chong, G. M. Dry-Anaerobic Digestion. In Energy, Agriculture and
Waste Management. Ann Arbor Science Publ., Inc., 1975. pp. 361-372.
85. Midwest Plan Service: Livestock Waste Management with Pollution Control.
North-Cent. Reg. Res. Publ., 222, 1975.
86. Gaddy, J. L., E. L. Park, and E. B. Rapp. Kinetics and Economics of
Anaerobic Digestion in Animal Waste. Water, Air and Soil Pollut., _3,
1974. pp. 161-169.
221
-------
87. Chen, J. R. and A. G. Hashimoto. Kinetics of Methane Fermentation.
Biotechn. and Bioeng. Symposium, No. 8, J. Wiley, Sons, 1978. pp. 269-
282.
88. Mueller, J. A. and J. L. Mancini. Anaerobic Filter-Kinetics and Applica-
tion. Proceedings 31st Industrial Waste Conference, Purdue University,
1976. pp. 423-447.
89. Anderson, G. K. and T. Donnelley. Anaerobic Digestion of High Strength
Industrial Wastewaters. Public Health Engineer, _5, No. 3, 1977. 64 pp.
90. Bryant, M. P. Animal Wastes Utilization Symposium. J. Animal Science,
.48, 1, 1979. pp. 193-200.
91. Hashimoto, A. G. and J. R. Chen. Economic Optimization of Anaerobic
Fermenter Designs. Proceedings of the IVth International Symposium on
Livestock Wastes/1980. ASAE.
92. Hashimoto, A. G. and J. R. Chen. Anaerobic Fermentation of Beef Cattle
and Crop Residues. 3rd Annual Biomass Energy Systems Conference, Golden,
CO, 5-7, 1979.
93. Hashimoto, A. G., J. R. Chen, and R. L. Prior. Methane and Protein
Production from Animal Feedlot Wastes. J. Soil and Water Conservation,
34, No. 1, 16-19, 1979.
94. Oleszkiewicz, J. A. Convergence of Activated sludge and Trickling
Folter Design Formulas. Environment Protection Engineering, 5^, 1, 27-32,
1979.
95. Janiczek, A., K. Jankowski, S. Koziarski, and J. A. Oleszkiewicz.
Beztlenowa fermentacja gnojowicy swinskiej (in Polish). Seminar Pro-
ceedings, National Seminar on Agricultural Wastes, NOT-PZITS, Slupsk,
September 1979. pp. 148-161.
96. Hashimoto, A. G., J. R. Chen, and V. H. Varel. Theoretical Aspects of
Methane Production: State-of-the-Art. Proceedings of the IVth Inter-
national Symposium on Livestock Wastes/I980. ASAE.
97. Balch, W. E., G. E. Fox, L. J. Magrum, C. R. Woese, and R. S. Wolfe.
Methanogens: Reevaluation of a Unique Biological Group. Microbio.
Reviews, ^3, 2, 1979. pp. 260-296.
98. Anon. Standard Method of Test for Chemical Oxygen Demand of Wastewater.
ASTM Part 23 D 1252-67.
99. G6lcz, A. On the Evaluation of Microscopic Methods for Activated Sludge
Analysis. Environment Protection Engineering, 6^, 1, 1980. pp. 103-111.
100. Malina, J. F., D. L. Ford, W. W. Eckenfelder, and E. M. Davis. Analytical
Methods for Wastewater Studies. Project Poland 0026, WHO, Vol. 1, Warsaw,
1970.
222
-------
101. Baines, S., H. A. Hawkes, C. H. Hewitt, and S. H. Jenkins. Protozoa as
Indicators in Activated Sludge Treatment. Sewage Industrial Wastes, 25,
1953. pp. 1023-1033.
102. Cramer, R. The Role of Protozoa in Activated Sludge. Industrial
Engineering Chemical, 23, 1931. pp. 309-313.
103. Matusiak, K., R. Mycielski, M. Blaszczyk, and A. Bisz. Studies on Purti-
fication of Wastes from Nitrogen Fertilizer Industry by Intensive Algal
Cultures. IV Growth of Chlorella Vulg. in Wastes with High Nitrogen
Content in Continuous and Intermittent Light. Acta Microbiologica
Polonica, 26, 1, 1977. pp. 79-93.
104. Hebrowska, A. Removal of Nutrients in Algal-Bacterial Ponds-Laboratory
Experiments. Environment Protection Engineering. 6_, 2, 1980. pp. 89-107.
105. Mazurek, A. Oczyszczanie sciek6w zwierzecych na zlozach beztlenowych.
Masters Thesis, Wroclaw Technical University, 1979. 82 pp.
106. Koziarski, S., Z. Kwiatkowski, J. A. Oleszkiewicz, and J. Zielinski.
Usprawnienia oczyszczalni typu Vidus. Budownictwo Rolnicze, 2_, 6, 1980.
107. Anon. Methods in Microbiology. 5B, Norris, J. R. and D. N. Ribbons,
(Editors). Academic Press, London-New York, 1971.
108. Miskiewicz, T., W. Lesniak, and J. Ziobrowski. Control of Nutrient Supply
in Yeast Propagation. Biotechnological Bioenergy 17, 1829, 1975.
109. Anthony, W. B. Cattle Manure. Reuse Through Wastelage Feeding. Proc.
Conf. Animal Waste Management., Cornell University, Ithaca, NY, 1969.
105 p.
110. Borkonski, J. and M. J. Johnson. Long-Lived Steam-Sterilizable Membrane
Probes for Dissolved Oxygen Measurement. Biotechnology Bioengineering,
9., 1967. 4 pp.
111. Brown, D. E. and S. Reddington. The Production of Microbial Pig Feed
Protein Supplement from Piggery Waste. Manuscript: University of Man-
chester, Inst. Scien. Technology, 1976.
112. Calvert, C. S. Use of Animal Excreta for Microbial and Insect Protein
Synthesis. J. Animal Science, 48, 178, 1979.
113. Cena M. Utilization of Manure from Large Industrial Farms (in Polish).
Przeglad Hodowlany, 12, 1976. pp. 20-21.
114. Evison, L. M. Production of Protein from Digester Liquids and Sludges.
Proceedings Paper, No. 7, "Protein and Livestock Feed from Biological
Wastewater Treatment Plants." University Manchester, 1ST, 14 XII,
1976.
223
-------
115. Miskiewicz, T. Application of the Turbidimetric Method for Determining
the Yeasts Concentration (in Polish). Przemysl Fermentacyjny i Rolny.
3, 17, 1978.
116. Olbricht, H. Malasse als Rohstoffproblem der Backhefeindustrie,
Branntweinwirtschaft, 113, No. 4, 1973. 53 pp.
117. Oleszkiewicz, J. A., K. Kosinska, and S. Koziarski. Piggery Wastewater
Fermentation by Yeast. Environment Protection Engineering, _5_, No. 3,
1979. pp. 155-162.
118. Peppier, H. J. Food Yeasts in: The Yeasts. 3^, Rose A. H. and J. S.
Harrison, Eds, Academic New York, NY, 1971. 421 pp.
119. Robinson, K. and D. Fenton. A Comparison of pH-Controlled and Dissolved
Oxygen-Controlled Nutrient Addition for the Maintenance of Steady-State in
Mixed Continuous Culture. J. Appl. Bacteriol., 42, 1977. 393 pp.
120. Singh, Y. K. and W. B. Anthony. Yeast Production in Manure Solubles. J.
Animal Science, 27, 1968. 1136 pp.
121. Thomas, K. Acid Digestion of Pig Waste to Produce a Suitable Substrate
for Protein Recovery. M. Sc. Dissert. Dept. Civil Eng. Newcastle upon
Tyne, 1975.
122. Tomlinson, E. J. Utilization of Strong Organic Effluents. Water Research
Centre. TR 15, 1975. 69 pp.
123. Nagy, G., M. Vaillant et al. Fermentation of Agricultural Wastes by
Yeast. Biotechn. Bioeng., 17, 1975. pp. 1823-1826.
124. Weiner, B. A. A Fermentation Process for the Utilization of Swine
Waste, in: "Food, Fertilizer and Agricultural Residues," Ann Arbor,
Science Publ., 1975. pp. 621-635.
125. Miskiewicz, T., J. A. Oleszkiewicz, K. Kosinska et al. Dynamic Tests
on Yeast Production from Piggery Effluents (in Print). Agricultural
Wastes, _3, 1981.
126. Girzynska, J. and T. Krzywacka. Protein from Algae (in Polish). Przem.
Farm, i Rolny., XX, 10, 1976. pp. 24-25.
127. Han, J. W., C. E. Dunlap, and C. D. Gallihan. Single Cell Protein from
Cellulosic Wastes. Food Technol., 25, 1971. pp. 130-135.
128. Hauri, K. S., et al. Production of Protein by Fungi from Agricultural
Wastes. Zbl. Bakt. II, Abt., 133, Parts I, II, III, and IV, 1978.
pp. 588-618.
129. Grzesiak, K. Analysis of Fodder Value of Protein Recovered from Pig
Wastes (Manuscript in Polish). Res. Devel. Dept., RIED, Wroclaw, 1979.
224
-------
130. Jarocka, D. Increase of Zooplankton Production by Addition of Pig
Wastes. Inform. Sygn. (in Polish) No. 3/79, (8) ITORG - Szczesin, 1979.
131. Harmon, B. G., et al. Nutritive Values of Aerobically or Anaerobically
Processed Swine Waste. Journal Animal Science, 37, 1973. 510 pp.
132. Mazurkiewicz, W. Use of Animal Waste in Feeding Ruminants, (in Polish)
Przegl. Hodowk., XLV, 21, 1977. pp. 16-17.
133. Kroodsma, W. Separation of Pig Feces from Urine Using Synthetic Netting
Under a Slatted Floor. Proceedings of the IVth Int'l Symposium on Live-
stock Wastes/1980. ASAE.
134. Flachowsky, G., A. Henning et al. Field Scale Trials of Feeding
Fattening Bulls Screened Solids from Pig Manure. Proceedings of the IVth
Int'l Symposium on Livestock Wastes/1980. ASAE.
135. Wawrzynczak, S. The Effectivenss of Composite Silages Containing Recycled
Bovine Wastes in Beef Cattle Fattening. Proceedings of the IVth Int'l
Symposium on Livestock Wastes/1980. ASAE.
136. Naugle, B. I., et al. The Dynamics of the Ensiling of Swine, Wastes
with Ground Corn. Proceedings of the IVth Int'l Symposium on Livestock
Wastes/1980. ASAE.
137. Martin, J. H., T. E. Pilbeam, and R. C. Loehr. Characterization of Animal
Manures as Feedstuffs. Proceedings of the IVth Int'l Symposium on Live-
stock Wastes/1980. ASAE.
138. Legiec, A., et al. Full Silage with Activated Sludge from Piggery Waste
(in Polish). Przegl. Hodowl., XLVI. 22, 20, 1978.
139. Berg van den, L., and C. P. Lenz. Comparison Between Up End Downflow
Anaerobic Fixed Film Reactors of Varying Surface-to-Volume Ratios for the
Treatment of Bean Blanching Waste. 34th Purdue University Industrial
Waste Conference. May 8-10, (NRCC 17452), 1979. 17 pp.
140. Nye, J. C. and R. N. Felix. Symbiotic Growth of Algae and Bacteria on
Swine Waste. Proceedings of the IVth Int'l Symposium on Livestock Waste/
1980. ASAE.
141. Anon. Algae as a Food Source (in Polish). Przemysl Spozyw. XXV,
1971.
142. Burczyk, J. Problems in Utilizing Algae as Animal Fodder, (in Polish)
Przem. Ferment, i Rolny. XXI, 10, 15-17, 1968.
143. Mtlntz, K. Algae as Potential Food Source (in Polish). Przem. Spozyw.,
XXV, 12, 1971.
144. Peppier, H. J. Food Yeasts, in "The Yeasts," edit. Rose, A. H. and J. S.
Harrison, Academic Press, NY, _3_, 1971. 421 pp.
225
-------
145. Pronczuk, A. Production Possibilities and Nutritional Value of Protein
from Bacteria, Yeasts and Alage (in Polish). Przemysl Spozywczy, XXI,
1970. pp. 7-8.
146. FAO-UN. Protein Requirements. FAO Nutritional Studies. 16, 1957. 52 pp.
147. Bednarski, W., et al. Use of Bacterial - Fungal Biomass in Enrichment
of Carbohydrate Fodder. Przem. Ferment, i Rolny. XIV, 10, 1970. pp.
21-25.
148. Bhattacharje, J. K. Microorganisms as Potential Sources of Food. Grant
No. GB-12330., National Science Foundation, Fac. Ros. Miami University,
Oxford, OH, 1969.
149. Anon. Microbiological Protein Synthesin in USRR (in Polish).
Przem. Ferment, i Rolny. XIV, 3, 1970.
150. Dzieniechowicz, M. Recycle of Animal Waste Solids as Fodder for Cattle
and Sheep (in Polish). Nowe Rolnictwo, 29, 4, 1980. pp. 32-33.
151. Ruszczyc, Z. Animal Feeding and Feedstuffs (in Polish). PaAstw. Wyd.
Roln. i Lesne. Warsaw, 1973.
152. Jonas, J. Aerobic Stabilization of Pig Feedlot Wastewaters in
Czechoslavakia in: "Animal Wastes." Applied Science Publ. London, 1977.
pp. 233-240.
153. Scharer, J. M. and M. Moo-Young. Methane Generation by Anaerobic Digestion
of Cellulose Containing Wastes. Adv. Biochem. Eng. 11, 1979. 85 pp.
154. Thomas, K. and L. Evison. Process Control of Anaerobic Digestion for SCP
Recovery. Symposium Papers: New Processes for Waste Treatment and
Recovery. Horwood Publ., Chichester, G. B., 1978. 349 pp.
155. Knol, W., et al. Biogas Production by Anaerobic Digestion of Fruit
and Vegetable Waste. A Preliminary Study. Science Foundation Agricul-
ture., 29, 1978. 822 pp.
156. Ngian, M. F., et al. Anaerobic Digestion of NaOH Pretreated Pig Feces
Using Sheep Rumen Liquor or Anaerobic Digester Liquor as Inoculum. Agri-
culture and Environment 4, 1978. 139 pp.
157. Hojovec, J. Feedlot Waste Management in Czechoslovakia in: "Animal
Wastes." Applied Science Publ., London, 1977. pp. 409-412.
158. Anon. Waste Treatment in Agriculture and Environmental Protection
(in Polish). IV Krajowa Kursokonfer. Nauk. Techn., PZITS-NOT, Slupsk, 1979.
159. Jewell, W. J. and R. C. Loehr. Energy Recovery from Animal Wastes in:
Animal Wastes Applied Science Publ., London, 1977. pp. 273-294.
226
-------
160. Smith, J. J., et al. The Role of an Anaerobic Digester on a Typical
Central Iowa Farm, in: Food, Fertilizer and Agricultural Residues.
Ann Arbor Science Publ., 1977. pp. 341-371.
161. Mills, P. J. Comparison of Anaerobic Digester and Aeration System Treating
Piggery Waste from the Same Source, in: Food, Fertilizer and Agricultural
Residues. Ann Arbor Science Publ., 1977. pp. 415-422.
162. Mills, R. A. and G. Tchobanoglous. Energy Consumption in Wastewater
Treatment, in: Energy, Agriculture and Waste Management. Ann Arbor
Science Publ., 1975. pp. 151-185.
163. Martin, J. H. and R. C. Loehr. Economic Analysis of Biogas Production and
Utilization. Proceedings of the IVth Int'l Symposium on Livestock Wastes/
1980. ASAE.
164. Neyeloff, S. and W. W. Gunkel. Methane-Carbon Dioxide Mixtures in an
Internal Combustion Engine, in: Energy, Agriculture and Waste Management.
Ann Arbor Science Publ., 1975. pp. 297-408.
165. Anon. Animal Waste Utilization on Cropland and Pastureland. USDA
Util. Research Dept. No. 6, EPA-600/2-79-059.
166. Majdowski, F. Lysometric Investigations on Hog Manure Land Disposal.
Environment Protection Engineering, ^_, 1, 1978. pp. 49-61.
167. Zawilski, J. Economics of Various Methods of Agricultural Utilization of
Manures (in Polish). National Conference on Agricultural Management of
Wastes. PZITS-NOT, Slupsk, Proceedings, 1976. pp. 60-79.
168. Overcash, M. R. and F. J. Humenik. State-of-the-Art: Swine Production
and Pretreatment Processes. Report EPA-600/2-76-290, 1976. 170 pp.
169. Oleszkiewicz, J. A. (Principal Investigator): "Optimization of Anaerobic
Digestion Systems for Industrial Wastes." Agreement J5-534-7, Marie
Sklodowska-Curie Fund, Research Institute Environmental Development -
Wroclaw and U. S. Environmental Protection Agency realized 1980-1983.
170. Lin Chou, W., R. W. Speece, and R. H. Siddiqi. Acclimiation and Degrada-
tion of Petrochemical Wastewater Component: by Methane Fermentation.
Biotechnology and Bioengineering Symposium, No. 8, 1978. pp. 391-414.
171. SroczyAski, A. and Z. Koluszko. Treatment of Starch Wastes (in Polish).
Zesz. Nauk. Politechn. Lidzkiej., No. 169, Chemia Spozyw., z.23, 1975.
pp. 149-161.
172. Carr, Ad. and R. C. O'Donnell. The Dynamic Behavior of an Anaerobic
Digester. Prog. Wat. Tech., £, PergamonP., 1977. pp. 727-738.
173. Imhoff, K. Taschenbuch der Stadtentwasserung. Verlag Oldenborg,
MUnchen, 1969. 300 pp.
227
-------
174. Oleszkiewicz, J. A. An Approach to Establishing Effluent Guidelines and
Evaluating Plant Performance. Environment Protection Engineering, 7_,
No. 1, 1981.
175. Oleszkiewicz, J. A. and W. W. Eckenfelder. The Mechanism of Substrate
Removal in High Rate Plastic Media Trickling Filter. Vanderbilt University
Press, 1974. 277 pp.
228
------- |