MECHANISM, KINETICS AND MICROBIOLOGY OF SELECTION

PRESSURE DRIVEN BIOMASS RETENTION IN SOLIDS

CONTAINING AGRICULTURAL WASTE TREATMENT

By

JINGWEI MA

A dissertation submitted in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

WASHINGTON STATE UNIVERSITY Department of Biological Systems Engineering

December 2012

ii

To the Faculty of Washington State University:

The members of the Committee appointed to examine the dissertation of JINGWEI MA find

it satisfactory and recommend that it be accepted.

___________________________________ Shulin Chen, Ph.D., Chair

___________________________________ Birgitte Ahring, Ph.D.

___________________________________ Pius Ndegwa, Ph.D.

___________________________________ Haluk Beyenal, Ph.D.

iii

ACKNOWLEDGMENTS

I would like to express my deep appreciation and sincerity to Dr. Shulin Chen, for his direction,

advice and funding support on my dissertation research. He invested me plenty of time and effort

to make me what a Ph.D. should be. He critiqued my study and challenged my thought to

achieve higher goals, and he also encouraged me when I faced the frustration. I sincerely thank

the other committee members, Dr. Birgitte K. Ahring, Dr. Pius M. Ndegwa, and Dr. Haluk

Beyenal, for their professional guidance and support in completion of this study.

I would like to thank Dr. Craig Frear who has generously provided the invaluable mentorship

and funding support. Many thanks go to my fellow colleagues: Zhiwu Wang, Baisuo Zhao, Yu

Liang, Quanbao Zhao, Zhanyou Chi, Yubin Zheng, Tingting Li, Jie Liu, Chenlin Li, Anping

Jiang, Nick Kennedy, Timothy Ewing, Paul Gamble and all group members. Thanks are also

extended to Jonathan Lomber, Kathy Dorgan, John Anderson, Joan Hagerdorn, Cindy Alwine,

Pat Huggins, and Pat King for their laboratory and administrative support.

I would like to thank China Scholarship Council (CSC) for providing me with this precious

opportunity to pursue Ph.D. degree aboard and thank for their funding support throughout my

dissertation research. I also would like to thank Washington State University Agriculture

Research Center, WA DOE, WSDA, and USDA for providing the financial support.

My deepest gratitude goes to my parents and family as well. Their unconditional love and

unending support gave me the strength to complete my PhD program.

I would like to express my sincere gratitude to all the people that contribute to the completion of

this thesis. With you all, living and studying in Washington State University was a great

experience!

iv

MECHANISM, KINETICS AND MICROBIOLOGY OF SELECTION

PRESSURE DRIVEN BIOMASS RETENTION IN SOLIDS

CONTAINING AGRICULTURAL WASTE TREATMENT

Abstract

by Jingwei Ma, Ph.D.

Washington State University

December 2012

Chair: Shulin Chen

The aims of this dissertation were to explore the mechanism of biomass retention in

solids containing waste under selection pressure, and to develop a high-rate anaerobic digester

treating agricultural waste containing solids. Active biomass retention is a cost-effective strategy

for uncompromised anaerobic digestion rate at lower temperature without requiring heating

energy. This is especially crucial for the anaerobic digestion of flushing diary manure in large-

scale dairies where flushing manure management system is employed.

A methodology for determining rate-limiting step in anaerobic digestion of complex

substrates was developed by supplementation of metabolic intermediates from each step of the

digestion process. The concept of microbial community ratio (r) in the anaerobic degradation

system was introduced and investigated in this study. The results revealed that the rate-limiting

step changed according to the variation of r.

v

The dissertation study was also designed to test both Gravity settling (GS) and selection

pressure (SP) theories applied for biomass retention, and to investigate their effects on active

biomass retention in Anaerobic Sequencing Batch Reactor (ASBR) treating flushing dairy

manure. The mechanism behind the opposing roles of GS and SP in settling time was also

studied. Results revealed that both short and long settling time were able to retain high

concentration of active microbes, though in disparate living forms.

A new strategy, improving biomass retention with fiber material presented within dairy

manure as biofilm carriers, was then developed and evaluated for treating flushing dairy manure

in a psychrophilic ASBR. A kinetic model for the process was also derived. It was proved that

ASBR using manure fiber as support media not only improved methane production but also

reduced the required HRT and temperature to achieve a similar treating efficiency compared

with current technologies.

The methanogenic community from ASBR was evaluated by Terminal Restriction

Fragment Length Polymorphism (T-RFLP) and clone libraries for both 16S rRNA gene and

mcrA gene. Results revealed that a Methanosarcina dominated methanogenic community was

successfully established in the ASBR digesters at short HRT. Diversity of methanogenic

community changed with variation of HRT. The performance of the digester was also related to

the diversity of microbial community.

vi

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ·················································································· iii

ABSTRACT··································································································· iv

LIST OF TABLES ··························································································· xi

LIST OF FIGURES ·························································································· xii

CHAPTER ONE

INTRODUCTION ···························································································· 1

1.1. Introduction ························································································· 1

1.2. Dairy Waste management in US ································································· 4

1.3. Anaerobic digestion ················································································ 6

1.4. Flushing manure treatment technology ························································· 9

1.5. Biomass retention practice in dairy manure AD ·············································· 11

1.6. Biomass retention mechanisms·································································· 13

1.7. Selection pressure driven biomass retention ·················································· 14

1.8. Conclusion ························································································· 15

1.9. References ·························································································· 17

CHAPTER TWO

DEVELOPMENT OF A SIMPLE METHODOLOGY FOR RATE-LIMITING STEP

DETERMINATION FOR ANAEROBIC DIGESTION OF SOLIDS CONTAINING

SUBSTRATE AND EFFECT OF MICROBIAL COMMUNITY RATIO ·························· 22

2.1. Abstract ····························································································· 22

2.2. Introduction ························································································ 23

vii

2.3. Methods ····························································································· 24

2.3.1. Substrate and inoculum ····································································· 25

2.3.2. Rate-limiting step evaluation ······························································· 25

2.3.3. Effect of microbial community ratio ······················································ 27

2.3.4. Analytical methods ·········································································· 27

2.3.5. Kinetics model ················································································ 29

2.3.5.1. Biogas production simulation ························································ 29

2.3.5.2. Kinetic model for microbial community ratio (r) affecting rate-limiting step 29

2.4. Results and discussion ············································································ 31

2.4.1. Rate-limiting step evaluation during the anaerobic digestion ························· 31

2.4.2. Effect of microbial community ratio on kinetics of anaerobic digesting dairy

manure ························································································· 35

2.4.3. Implications for anaerobic co-digestion ·················································· 38

2.5. Conclusion ························································································· 39

2.6. References ·························································································· 40

2.7. Nomenclature ······················································································ 44

CHAPTER THREE

BIPOLAR EFFECTS OF SETTLING TIME ON ACTIVE BIOMASS RETENTION IN

ANAEROBIC SEQUENCING BATCH REACTORS DIGESTING FLUSHING DAIRY

MANURE ····································································································· 46

3.1. Abstract ····························································································· 46

3.2. Introduction ························································································ 47

3.3. Materials and methods ············································································ 49

viii

3.3.1. Experimental setup and operation ························································· 49

3.3.2. Microscopic visualization of microbial distribution ···································· 50

3.3.3. ATP determination ··········································································· 51

3.3.4. Methanogens activity measurement ······················································· 51

3.3.5. Other analytical methods···································································· 52

3.4. Results and discussion ············································································ 52

3.4.1. Process of active biomass retention in ASBRs at various settling time ·············· 52

3.4.2. Effect of settling time on active biomass retention ····································· 53

3.4.3. Active microbial response to settling time ··············································· 55

3.4.4. Living form of retained microorganisms. ················································ 57

3.4.5. Mechanism behind bipolar effects of settling time on active biomass retention ···· 59

3.4.5.1. Model development ··································································· 60

3.4.5.2. Model verification ····································································· 61

3.4.5.3. Model limitation ······································································· 62

3.5. Conclusions ························································································ 65

3.6. References ·························································································· 67

CHAPTER FOUR

KINETICS OF PSYCHROPHILIC ANAEROBIC DIGESTER WITH BIOFILM SUPPORTED

BY SOLIDS FROM FLUSHING DAIRY MANURE ················································· 70

4.1. Abstract ····························································································· 70

4.2. Introduction ························································································ 71

4.3. Materials and methods ············································································ 74

4.3.1. Feedstock and seed ·········································································· 74

ix

4.3.2. Experimental setup and operation ························································· 74

4.3.3. Chemical analytical methods ······························································· 75

4.3.4. Development of kinetic model ····························································· 77

4.4. Results and discussion ············································································ 82

4.4.1. Effect of HRT on biogas production ······················································ 82

4.4.2. Effect of OLR on biogas production ······················································ 87

4.4.3. Kinetic modeling ············································································· 87

4.4.3.1. Evaluation of Kinetic models ························································ 87

4.4.3.2. Model simulation ······································································ 88

4.4.3.3. Model prediction ······································································· 91

4.4.4. Process performance comparison ·························································· 93

4.4.5. Implications for dairy AD process design ················································ 93

4.5. Conclusion ························································································· 94

4.6. References ·························································································· 96

4.7. Nomenclature ···················································································· 100

CHAPTER FIVE

METHANOSARCINA DOMINATION IN ASBR DIGESTER AT SHORT HRT ··············· 102

5.2. Introduction ······················································································ 103

5.3 Materials and methods ·········································································· 107

5.3.1 Feedstock and seed ········································································ 107

5.3.2 Experimental setup and operation ······················································· 107

5.3.3 DNA extraction············································································· 108

5.3.4 PCR amplification of the 16S rRNA gene for T-RFLP ······························ 108

x

5.3.5 PCR Amplification of Functional Gene Marker for T-RFLP························ 109

5.3.6 T-RFLP analysis ··········································································· 109

5.3.7 Clone Library of the 16S rRNA Gene ·················································· 110

5.3.8 Clone Library of Functional Gene Marker ············································· 110

5.3.9 Phylogenetic analysis ······································································ 111

5.3.10 Nucleotide sequence accession numbers ··············································· 111

5.4 Results and discussions ········································································· 112

5.4.1 Digesters performance ···································································· 112

5.4.2 Methanogenic community diversity ····················································· 113

5.4.3 Phylogenetic analysis ······································································ 116

5.4.3.1 16S rRNA clone library ····························································· 116

5.4.3.2 mcrA gene clone library ···························································· 117

5.4.3.3 Statistical analysis of clone libraries ·············································· 121

5.4.3.4 Comparison of clone library for 16S rRNA gene and mcrA gene ············ 122

5.4.4 Methanogenic pathway in psychrophilic ASBR ······································· 124

5.5 Conclusion ······················································································· 127

5.6 Reference ························································································· 128

CHAPTER SIX

CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK ······················· 133

xi

LIST OF TABLES

Table 1.1 Dairy Manure Collection Systems .................................................................................. 5

Table 2.1 Experimental design for rate-limiting step evaluation .................................................. 26

Table 2.2 Microbial community ratio experimental design. ......................................................... 28

Table 2.3 Kinetics parameters for methane production with metabolic intermediates

supplementation ............................................................................................................ 34

Table 4.1 Operating condition for ASBR ..................................................................................... 76

Table 4.2 Kinetics models used in this study ................................................................................ 81

Table 4.3 Summary of model comparison with kinetic coefficients and goodness of fit ............. 90

Table 4.4 Performance data for different anaerobic biofilm reactors treating dairy manure ....... 92

Table 5.1 Typical competitive microbes in anaerobic digester .................................................. 105

Table 5.2 Diversity and richness indices for 16S rRNA and mcrA clone libraries .................... 120

Table 5.3 Cluster of Archaea clone sequences for 16S rRNA clone libraries and its affiliation

regarding BLAST search ............................................................................................ 121

Table 5.4 Cluster of Archaea clone sequences obtained for mcrA gene clone libraries and its

affiliation regarding BLAST search ........................................................................... 123

Table 5.5 Gibbs free energy of reactions associated with methane generation .......................... 125

xii

LIST OF FIGURES

Figure 1.1 Anaerobic Digestion Process ......................................................................................... 7

Figure 2.1 Cumulative methane productions with metabolic intermediates supplementation.

Profiles are from triplicated experiments; values represent average and error bars

mean standard deviation ............................................................................................. 32

Figure 2.2 Maximum methane production rate changes against NS to HS ratio R, (●)

experimental results, and (—) simulation profile ....................................................... 36

Figure 2.3 Maximum methane production rate changes against H to M ratio r, (●) experimental

results, and (—) simulation profile ............................................................................. 38

Figure 3.1 Schematic of ASBR operation in a serum bottle ......................................................... 50

Figure 3.2 volumetric biogas production profiles in ASBRs operated at various settling time ... 53

Figure 3.3 Effect of settling time on steady state volumetric biogas production rate (a), mixed

liquor ATP concentration (b) and methanogenic activity (c) ..................................... 54

Figure 3.4 Effect of settling time on washout coefficients for abiotic and biotic particles,

respectively ................................................................................................................. 57

Figure 3.5 Confocal visualization of rod and coccoid microbial cells (green), a) attaching on

manure fiber surface in steady state ASBR running at settling time of 0.5 min and b)

remaining suspended in the bulk solution of steady state ASBR running at settling

time of 60 min, respectively. c) SEM image showing the detail of rod and coccoid

microbial cells attachment on fiber surface in steady state ASBR running at settling

time of 0.5 min ............................................................................................................ 58

xiii

Figure 3.6 Eq. (3.6) simulated active biomass retention (a), inactive biomass retention (b),

volumetric biogas production rate (c), ATP concentration (d) and methanogentic

activity (e), (●) experimental results, (—) simulation profile .................................... 65

Figure 4.1 Volumetric biogas production rate at various HRT .................................................... 83

Figure 4.2 Changes of volumetric methane production rate and specific methane productivity

against HRT and OLR, (●) experimental results, and (—) simulation profile ........... 86

Figure 4.3 Comparison of simulation with different kinetics models .......................................... 89

Figure 4.4 Changes of effluent substrate concentration and treatment efficiency with derived

model, (●) effluent VS data, () VS removal efficiency data, and (—) predicted

profile .......................................................................................................................... 91

Figure 5.1 Relationship between digester performance and microbial community diversity .... 114

Figure 5.2 Relative abundance of the Archaea 16S rRNA gene fragments (A) and mcrA gene

fragments (B) retrieved from the biomass in ASBR. The length of T-RFs in base pairs

(bp) is indicated in parenthesis ................................................................................. 115

Figure 5.3 Phylogenetic relationship among 18 OTUs with partial 16S rRNA gene sequences of

known methanogenic Archaea based on the neighbor-joining analysis ................... 118

Figure 5.4 Phylogenetic relationship among 21 OTUs with partial mcrA gene sequences of

known methanogenic Archaea based on the neighbor-joining analysis ................... 119

Figure 5.5 Rarefaction curves generated for 16S rRNA genes and mcrA genes clone libraries.

Dotted lines indicate the 95% confidence intervals .................................................. 120

Figure 5.6 Speculated Methanogenic Pathway in ASBR ........................................................... 126

xiv

Dedication

This dissertation is dedicated to my family, especially father, mother and my wife, who were

always willing to support me in whatever way necessary.

1

CHAPTER ONE

INTRODUCTION

1.1. Introduction

With the anticipated decline of fossil fuel, there is a great opportunity for renewable energy,

especially biomass derived energy, to play an important role in near future. Anaerobic digestion

(AD) technology is gaining more and more attentions and has been widely implemented for the

treatment of animal manure, energy crop, industrial wastewaters, organic fraction of municipal

solid waste, and other biomasses for both energy generation and biodegradable organic waste

mineralization at the meantime.

AD as a mean for stabilization of livestock manure has been carried out for many years. This

practice involves a mixed culture of symbiotic bacteria that mediate the degradation of organic

matter in manure solids ultimately to methane, carbon dioxide and mineralized nutrients. This

microbial mixture includes symbiotic groups of hydrolytic, acidogenic and methanogenic

bacteria. The performance of AD thus depends on the availability and activity of these symbiotic

bacteria. Hydrolysis and methanogenesis are two crucial steps in the course of anaerobic

digestion of solids wastes. Hydrolysis rate has been found correlated to the coverage of

hydrolytic bacteria grown on solids substrate surface (Sanders 2001). Immobilization of

hydrolytic bacteria on solids substrate thus directly promotes hydrolysis progress. Due to the

slow growing nature of methanogenic bacteria and their limited numbers presented in fresh

manure (Griffin, McMahon et al. 1998; McMahon, Stroot et al. 2001; Stroot, McMahon et al.

2001), the intermediates products in the course of anaerobic digestion which are volatile and

2

often a source of odors will become accumulated, leading to low methane gas production and

odor emission. Therefore, an effective anaerobic digester should have capacity to immobilize

high symbiotic biomass concentration on solids surface so that a faster and more complete solids

to biogas conversion efficiency can be achieved, and the odor potential of the manure is also

greatly reduced.

In order to achieve an enhanced solids wastes to biogas conversion efficiency, three major

approaches have been implemented so far for anaerobic digestion of livestock manure, i.e. (i)

prolong digestion time; (ii) maintain high temperature; and (iii) retain high biomass.

Corresponding digester types that have been developed include various plug-flow digesters,

complete-mix digesters, covered lagoons, biofilm reactors and some forms of their

hybridizations (Dennis and Burke 2001).

Hydraulic retention time (HRT) has been manipulated to prolong the digestion time in plug-flow

and complete mixed manure digesters. It is a very common practice to design a 20 to 40 days

HRT for complete anaerobic digestion of dairy manure in these two types of digesters. However,

one should be realized that the volume of the digester is directly proportional to the designed

HRT, i.e. a greater HRT leads to large digester volume. This requires that feedstock should have

as high TS content as possible to minimize the excess water volume that takes up digester

volume.

To minimize HRT, plug-flow and complete mixed digesters are usually operated at mesophilic

temperatures (usually 35 oC) to have a faster anaerobic digestion rate. This approach inevitably

3

involves additional energy input needed to heat the wastewater. In practice, a portion of biogas

recovered in the course of anaerobic digestion is recycled back to maintain this mesophilic

temperature, which results in less biogas production from the system. Likewise to the constrain

of HRT, this requires that the manure fed to these reactors should have as high a total solids (TS)

content as possible to minimize the water content that must be heated. In temperate climates,

often dilute manure with 2% TS or less will fail to provide enough biogas to heat the mixed

liquor to 35 oC.

In recent years, it becomes a popular trend to use large amount of water to flush livestock

manure out of barn for the sake of the convenience and labor saving (Dennis and Burke 2001). A

direct consequence from this is the extensive dilution of livestock manure from 10 to 15 % solids

concentration all the way to about 1 to 2 %, resulting in excessive amounts of dilute wastewater

in flushing manure (Dennis and Burke 2001). As aforementioned, the dilute nature of flushing

manure effectively precludes the approaches that depend on HRT and/or temperature as

practiced in conventional plug-flow and complete-mix manure digesters in view of the

uneconomical digester volume requirements and the excessive energy required to heat the dilute

water to mesophilic temperatures for effective digestion. Therefore, there is a need for

developing a new high rate anaerobic digester design enabling efficient flushing dairy manure

treatment.

Before proceeding directly to the research as described in Chapters Two, Three, Four and Five of

this dissertation, it is important to outline and summarize some practical knowledge regarding

4

dairy manure management, AD process, the dairy manure AD practice and biomass retention for

high rate digester for background purposes.

1.2. Dairy Waste management in US

Dairy farming has been cited as the largest agricultural source for greenhouse gas (CH4) mission

(Weiske, Vabitsch et al. 2006). It has been estimated that, if AD process is exploited to recover

methane energy from dairy manure, around 222,142 MMBtu thermal energy can be annually

recycled from each head of dairy cattle (Garrison and Richard 2005). Among the 9,000,000

heads milk cows raised in United States (USDA 2005), only 150,000 heads are currently served

with AD system (U.S. EPA 2007), implying a 98.3% shortage in AD application for dairy. Poor

cost-effectiveness of current AD design pertaining to the characteristics of dairy manure is

responsible for limited adoption of AD. Manure characteristics are closely associated with the

way it is collected. There are mainly two types of manure collection systems employed in dairy,

namely scraped and flushing systems (Dennis and Burke 2001). Dry manure collection systems,

resulted to 20-25 % of total solids, also been used under certain circumstance. Flushing system

has been considered to be the more economical and less labor-intensive means over scraped

system for dairy manure collection (Dennis and Burke 2001). Dairy management survey in

California revealed that flushing system is the major means for dairy manure collection, and its

adoption popularity has kept a increasing trend from 61.7 % in 1985 to 77.1% in 1997 (Fairbank

1986; Meyer, Garnett et al. 1997). Similar predominant usage of flushing system was also

reported for Florida (Wilkie, Castro et al. 2004).

5

Unsafe and improper disposal of animal and farm wastes results in land and surface and ground

water pollution, ammonia emission, methane causing greenhouse gas effect, and odors; while

treatment of these wastes by anaerobic degradation produces biogas, bio-fertilizer, reduced

pollution, reduced GHG emission, and reduced odor. Adopted AD process varies with the

feedstock generated by different manure collection systems. Solid digester was employed for

high solids containing dry manure. Scraped manure could be effective treated by conventional

plug flow digester or complete mix digester. Flushing manure challenges conventional digesters

due to the large amount of water usage, and covered lagoon is the most prevalent technology

treating flushing dairy manure.

Table 1.1 Dairy Manure Collection Systems

Options Dry system Scrape system Flush system

Total solids 20 – 25 % 8 – 10 % 1 – 2 %

Classification Solid Slurry Liquid

Treatment Solid digester Plug flow digester or complete mix digester

Covered lagoon

Cattle manure is a complex type of substrate, composed of carbohydrates, proteins and fats, and

anaerobic digestion process has been successfully be applied to treat cattle manure. Application

of AD process for dairy manure treatment resulted in both environmental and economic benefits.

First, AD process stabilizes volatile organic waste through BOD/COD reductions, which helps

6

enhancing water quality; second, volume of solids waste is reduced, contributing to water quality

protection and reduction of farm lagoon maintenance; third, manure associated odor problem is

controlled because of volatile solids and volatile fatty acids (VFA) reductions, which improved

community relations and air quality; fourth, pathogen reduction is another contribution of AD

process, caused by partial zoonotic destruction from extended anaerobic and temperature

conditions; moreover, application of AD process will bring extra CHP and carbon, energy and

tax credits, making AD process more compelling than other manure management technologies.

1.3. Anaerobic digestion

The anaerobic digestion of complex organic material has been described as a multi-step process

that consists of four stages in series: hydrolysis, acidogenesis, acetogenesis and methanogenesis

(Pavlostathis and Giraldogomez 1991; Batstone, Keller et al. 2002). This process involves a

number of microorganisms such as fermentative bacteria, hydrogen-producing acetogenic

bacteria, hydrogen-consuming acetogenic bacteria, carbon dioxide-reducing methanogens, and

aceticlastic methanogens (methane-forming bacteria) and enzyme secreted by some bacteria

(Chynoweth and Isaacson 1987; Appels, Baeyens et al. 2008).

Disintegration and hydrolysis

The first step of AD process can be further differentiated into disintegration and hydrolysis two

steps. Disintegration is the degradation of the whole biomass. Disintegration occurs primarily by

physicochemical processes and is dependent primarily on the physical characteristics of the

system including temperature and pH. This process breaks down the initial solid biomass into

separate components, namely carbohydrates, lipids, proteins, soluble inerts, and particulate inerts.

7

Hydrolysis is the breakdown of relatively large organic compounds, lipids, carbohydrates and

proteins to long chain fatty acids, monosaccharides, and amino acids, respectively (Bagi, Acs et

al. 2007). Hydrolysis is carried out by both physicochemical means as well as by break down

from extra-cellular enzymes.

Figure 1.1 Anaerobic Digestion Process

Acidogenesis/fermentation

Acidogenesis, also known as acid fermentation, involves acid-forming fermenters and hydrogen

producers. Monomer or oligomer from hydrolysis step are further degraded to volatile fatty acid,

8

hydrogen and carbon dioxide. Degradation of monosaccharides follows glycolytic pathway,

while Stickland oxidation-reduction paired fermentation governs the conversion of amino acids

(Batstone, Keller et al. 2002). Degradation of long chain fatty acids is through beta-oxidation

pathway.

Acetogenesis

Acetogenesis is syntrophic process and need collaboration with intermediate consumer,

methanogens for example. Volatile fatty acids other than acetate, such as propionate and butyrate,

are converted to acetate and produce hydrogen at the same time. To make these reactions

thermodynamically favorable, a low hydrogen partial pressure is required (Zinder 1993). The

presence of hydrogenotrophic methanogen and other hydrogen utilizers could maintain a low

hydrogen concentration for the oxidation reaction to be thermodynamically possible.

Methanogenesis

Methanogenesis is the last step of the anaerobic conversion of organic wastes to methane.

Formic acid, acetic acid, methanol, and hydrogen can all be used as energy sources by the

various methanogens. Methane formation is generally accomplished by two main pathways, i.e.

aceticlastic methanogenesis and hydrogenotrophic methanogenesis. Aceticlastic methanogenesis

is the primary route converting the major fermentation product of the acid forming phase, acetic

acid, to methane and carbon dioxide, which account for about 70% of methane productions. So

far, only the Methanosarcinales, specifically Methanosarcina and Methanosaeta, have been

shown to be capable of utilizing acetate with Methanasarcina being uniquely capable of utilizing

all three forms of substrate for methane production and as such is not an obligatory acetotroph

9

(Boone, Whitman et al. 1993). Methanogens from orders, Methanobacterials, Methanococcales,

and Methanomicrobiales reduce CO2 to CH4 using H2 or formate through hydrogenotrophic

methanogenesis pathway.

In such a multi-step complex process, the kinetics of the slowest step will account for the overall

kinetics. Anaerobic digestion systems are limited by two major steps depending on the nature of

the substrate. Hydrolysis is often limited if the substrate is complex organic solids while in the

digestion of soluble organic matter, the rate-limiting step has been identified as methanogenesis

(Vavilin, Rytov et al. 1996; Tomei, Braguglia et al. 2009). Determination of rate-limiting step is

critical for a new anaerobic digestion process design treating specific feedstock, establishment of

a stable process performance, and management for full scale of application of anaerobic

digestion processes. Hence, there is an important need for an experimental method capable of

determining precisely which of the different steps is limiting for a complex substrate.

1.4. Flushing manure treatment technology

The efficiency of anaerobic digestion is controlled by several factors. For a given type of

feedstock, the rate of anaerobic digestion increases with the amount of active biomass retained in

the digester as well as temperature (Lettinga, Rebac et al. 2001; Batstone, Keller et al. 2002).

Active biomass retention provides a cost-effective method for uncompromised anaerobic

digestion rate at lower temperature with less heating energy consumption (Lettinga, Rebac et al.

2001; Connaughton, Collins et al. 2006). This is especially critical for the anaerobic digestion of

flushing diary manure in large-scale dairies where flushing system is employed.

10

The huge dilute water flow from flushing system makes conventional dairy AD design

inapplicable owing to the requirements of massive digester size and heating energy consumption.

For this account, solids/liquid separation prior to AD strategy has been approached for AD of

flushing dairy manure. Following conventional AD design for scraped manure, two modified

plug-flow digesters have been practiced in north America to concentrate flushing dairy manure

form 1-2 % back to a desired solids concentration of 8-10% within the capacity of conventional

plug-flow digester through discarding flushing water prior to AD. Although this approach

considerably saved digester space and heating energy, evidence showed that a great proportion of

readily digestible organic matter is actually dissolved in the liquid phase of flushing dairy

manure (Mackie, Stroot et al. 1998), and thus discarding flushing water will inevitably reduce

methane recovery efficiency and emit greenhouse gas (Wilkie, Castro et al. 2004). Accordingly,

an alternative approach based on biofilm driven high rate AD process has been taken towards the

methanation of flushing dairy manure (Wilkie, Castro et al. 2004). This approach allows rapid

digestion of large flow of flushing dairy manure at psychrophilic temperature with comparable

digester size (Vartak, Engler et al. 1997). However, costly biofilm carrying media have to be

applied in this approach to retain high biomass concentration in digester (Vartak, Engler et al.

1997). Clogging of these media by manure solids seems unavoidable in the course of AD

operation, which reduces digester robustness, displacing reactor volume and short-circuiting

wastewater around the biofilm media (Wilkie, Castro et al. 2004). Consequently, a solids/liquid

separator has to be installed to remove coarse manure solids for a steady biofilm reactor

performance. It was held that these cattle rumen indigestible solids should own a low

biodegradability, and thus their disposal should not cause much reduction in methane production

(Wilkie, Castro et al. 2004). However, accumulated evidences show that considerable amount of

11

readily digestible COD is actually mingled with those manure solids and removed in separator,

e.g. approximately 54 % and 80 % of methane production potentials were found to be carried in

discarded manure solids (Hills and Kayhanian 1985; Chastain, Vanotti et al. 2001). Thus,

current high rate AD process is also subjected to recoverable energy lost, and the high cost for

expensive biofilm media impede the popularity of the approach in dairy as well. Latest EPA

census shows that there is merely one full-scale biofilm based AD reactor in service for flushing

dairy manure throughout United States (U.S. EPA 2007). Apparently, a process with no need of

exotic media addition would be preferred for flushing dairy manure anaerobic digestion.

1.5. Biomass retention practice in dairy manure AD

High biomass retention would be the best means for anaerobic digestion of dilute livestock

manure in an efficient and economical way. Biofilm formation on support media with large

surface area for bacterial attachment offers a greater concentration of bacteria for accelerated

digestion of organic matter. Moreover, retaining these support media in digester through gravity

settling and/or mechanical screening methods is able to uncouple bacterial retention time from

mixed liquor HRT. Therefore, biofilm based high biomass retention reactor offers a possibility to

operate anaerobic digester at much shorter HRTs and lower temperatures while achieving similar

treatment efficiencies as those in conventional plug-flow and complete-mix systems.

Currently, there are designs available for high-rate anaerobic digestion using biofilm based

biomass retention strategy. With different morphology and physical characteristics of supporting

media, these biofilm based anaerobic digester are known as fixed-, fluidized- or expanded-bed

reactors, in which microorganisms are retained by attachment to support media in fixed-,

12

fluidized- or expended- state within the processing system (Borja, Sanchez et al. 1994; Wilkie,

Castro et al. 2004; Demirer and Chen 2005; Umana, Nikolaeva et al. 2008; Zaher, Frear et al.

2008). However, although a wide variety of supporting media including sand, anthracite,

activated carbon, PVC materials, or diatomaceous earth have been practiced, none of those

support materials can work with wastewater having significant levels of suspended solids such as

those found in livestock manure.

Livestock manure often includes rumen indigestible forage fibers as well as materials used for

bedding, such as hay, sawdust or sand. These large solids particles contribute to clogging of all

aforementioned packing material. They also hinder the attachment of bacteria to the media,

ultimately, lead to short-circuiting of the anaerobic system, which reduces the effectiveness of

the biological treatment system and results in the failure of anaerobic digester. For these reasons,

a gravity settling and/or mechanical screening based separator has to be placed in front of all

existing biofilm manure digesters to exclude large solids from entering anaerobic digestion. It is

obvious that this separator increases the complexity of digester apparatus. Moreover, it has been

estimated that about 54 to 80 % of readily methane recoverable organics mingling with separated

solids will be lost after separator (Hills and Kayhanian 1985; Chastain, Vanotti et al. 2001). It

should be pointed out that all existing biofilm reactors rely on exogenous support media and thus

cannot work in harmony with high solids containing wastewater. In addition, expensive capital

investment in support media also possesses an impediment to the implementation of existing

biofilm digester, in that support media used in biofilm digesters accounts for a significant portion

of the capital investment.

13

1.6. Biomass retention mechanisms

Gravity settling is a conventional method for biomass retention with no requirement for addition

of supporting materials. It takes advantage of a prolonged settling time to separate bacterial cells

from the supernatant. Based on Eq. (1.1), a gravity settling theory (GS) was established

𝑉𝑐 =𝐿𝑑𝑡𝑠

(1.1)

in which Vc represents a critical settling velocity created by the depth of discharge zone (Ld) and

the settling time (ts) (Vesilind 2003). It means that all particles with settling velocity (Vp) greater

than Vc will settle beneath the discharge zone together with their surface attached cells and be

retained (Vesilind 2003). Due to the minor settling velocity of planktonic bacterial mass, a rather

long ts of 1 to 3 hours is typically needed to retain bacterial cells at a given Ld (Lee 2000;

Wilderer, Irvine et al. 2001). It should be noted that influent solids, e.g. undigested dairy manure

fibers, may have settling velocities greater than those of planktonic bacterial cells, and thus

application of GS in an anaerobic digester may retain not only active but also inactive biomass

(Lott, Loch et al. 1994). Obviously, inactive biomass retention is undesirable, as it takes up

reactor volume. Selection pressure (SP) driven cells immobilization is another theory developed

in recent years for active biomass retention with no need of external media (Liu, Wang et al.

2005). SP theory is based on an equation similar to Eq. (1.1), but it requires an extremely short ts,

typically less than 5 min, to create a large Vc to drive cellular self-immobilization for retention

(Qin, Liu et al. 2004). According to SP theory, microorganisms are able to actively respond to a

short ts to avoid being washed out of a reactor (Qin, Tay et al. 2004). In support of this theory,

successful bacterial retention has been achieved in solids containing dairy wastewater

(Schwarzenbeck, Borges et al. 2005). SP theory takes advantage of bacterial initiative behavior,

14

and thus has the potential to selectively favor the retention of active biomass over inactive

biomass. Such a possibility has yet been studied in solids wastes anaerobic digestion.

1.7. Selection pressure driven biomass retention

Ideally, a successful anaerobic digestion reactor should not only be able to operate at lower

temperatures and shorter HRT, but also capable of tolerating high level influent solids for long

term operation. As a matter of fact, the influent solids themselves have the potential to work as

in-situ support media for biomass immobilization. This is the exact mechanism that is working in

animal rumen digestion, i.e. retention of rumen bacteria is realized through establishing biofilm

on the surface of fed forage for accelerated rumen digestion (Cheng, Fay et al. 1981). In order to

drive bacterial immobilization on solids surface, a selection pressure that suppress suspended

bacteria growth has to be applied in reactor to favor immobilized biofilm growth (Hulshoff Pol,

Heijnekamp et al. 1988; Liu, Wang et al. 2005). Recent development in selection pressure

driven cells immobilization technology offers practical means to realize such a biomass

immobilization process in bioreactors such as anaerobic sequencing batch reactor (ASBR) and

upflow anaerobic sludge blanket reactor (UASB) (Hulshoff Pol, Heijnekamp et al. 1988; Liu,

Wang et al. 2005). Both types of operations have capacities of uncoupling SRT from HRT so

that a minimized reactor volume can be achieved with maximized digestion performance

(Hulshoff Pol, Heijnekamp et al. 1988; Liu, Wang et al. 2005). Further, by virtue of the high

biomass retention in ASBR and UASB, they have been proved applicable even at psychrophilic

condition for high rate anaerobic digestion. Therefore, implementing selection pressure in ASBR

for biomass immobilization on influent solids allows anaerobic digestion of diluted solids wastes

15

at low temperature and short HRT with no need of pretreatment as well as exogenous support

media.

1.8. Conclusion

To date, there remain technological deficiencies in the economical anaerobic digestion of the

entire flow of flushing dairy manure on farms within cold climates. The huge dilute water flow

from flushing dairy manure system makes conventional dairy AD design inapplicable owing to

the requirements of massive digester size and heating energy consumption. Flushing dairy

manure possesses challenges to conventional high temperature anaerobic digesters because of the

large volume of flushing water that must be heated. To address this issue, fixed-film digesters

with submerged high specific surface area media were introduced to retain active biomass

through bacterial attachment (Vartak, Engler et al. 1996; Wilkie, Castro et al. 2004; Umana,

Nikolaeva et al. 2008; Nikolaeva, Sanchez et al. 2009). Nevertheless, fixed-film was classified as

an inappropriate process for dairy manure anaerobic digestion due to its vulnerability to media

clogging (Burke 2001). Apparently, a new process for flushing dairy manure anaerobic digestion

should be developed and the following criteria need to be satisfied:

• With no need of expensive exotic biofilm support media addition to save capital

cost and reduce digester volume;

• Capable of treating both solids part and liquid part of dairy manure in one single

digester without prior separation;

• Without concerns about clogging hazards caused by fibrous solids presented

within dairy manure.

16

The high affinity of microbes to dairy manure fibrous solids seems to point to an alternative for

biomass retention in an anaerobic digester providing a natural biofilm support medium for high

biomass retention as opposed to using external media that might clog and add cost. As a matter

of fact, forage like straw fibers have been evaluated as excellent support medium for

methanogenic biofilm establishment. Dairy manure fibrous solids have the potential to act as

effective natural biomass carriers for high rate digester.

Therefore, a new strategy, improving biomass retention with fiber material present within the

dairy manure as biofilm carriers, was proposed and the biofilm formation mechanism under

selection pressure was investigated in this research. The performance of this technology was then

evaluated for treating flushing dairy manure in a psychrophilic anaerobic sequencing batch

reactor (ASBR) with a derived kinetic model. The unique Methanoarchaea community structure

and methanogenic pathway were also elucidated in this study.

17

1.9. References

Appels, L., J. Baeyens, et al. (2008). "Principles and potential of the anaerobic digestion of

waste-activated sludge." Progress in Energy and Combustion Science 34(6): 755-781.

Bagi, Z., N. Acs, et al. (2007). "Biotechnological intensification of biogas production." Applied

Microbiology and Biotechnology 76(2): 473-482.

Batstone, D., J. Keller, et al. (2002). Anaerobic digestion model no. 1 (ADM1), IWA Task

Group for mathematical modelling of anaerobic digestion processes. London, UK, IWA

Publishing.

Boone, D. R., W. B. Whitman, et al. (1993). Diversity and Taxonomy of Methanogens.

Methanogenesis: Ecology, Physiology, Biochemistry and Genetics. J. G. Ferry. New

York, Chapman & Hall: 35–80.

Borja, R., E. Sanchez, et al. (1994). "Kinetics of anaerobic degestion of cow manure with

biomass immobilized on zeolite." The Chemical Engineering Journal and the

Biochemical Engineering Journal 54(1): B9-B14.

Burke, D. A. (2001). Dairy waste anaerobic digestion handbook. 6007 Hill Street, Olympia, WA

98516, Environmental Energy Company.

Chastain, J. P., M. B. Vanotti, et al. (2001). "Effectiveness of liquid-solid separation for

treatment of flushing dairy manure: A case study." Applied Engineering in Agriculture

17(3): 343-354.

Cheng, K. J., J. P. Fay, et al. (1981). "Formation of Bacterial Microcolonies on Feed Particles in

the Rumen." Appl Environ Microbiol 41(1): 298-305.

Chynoweth, D. P. and R. Isaacson (1987). Anaerobic Digestion of Biomass. London and New

York, Elsevier Applied Science.

18

Connaughton, S., G. Collins, et al. (2006). "Psychrophilic and mesophilic anaerobic digestion of

brewery effluent: A comparative study." Water Research 40(13): 2503-2510.

Demirer, G. N. and S. L. Chen (2005). "Anaerobic digestion of dairy manure in a hybrid reactor

with biogas recirculation." World Journal of Microbiology & Biotechnology 21(8-9):

1509-1514.

Dennis, A. and P. E. Burke (2001). Dairy waste anaerobic digestion handbook. 6007 Hill Street,

Olympia, WA 98516, Environmental Energy Company.

Fairbank, W. C. (1986). Dairy design and waste management practices in southern California,

San Luis Obispo, CA, USA, ASAE, St. Joseph, MI, USA.

Garrison, A. V. and T. L. Richard (2005). "Methane and manure: Feasibility analysis of price

and policy alternatives." Transactions of the Asae 48(3): 1287-1294.

Griffin, M. E., K. D. McMahon, et al. (1998). "Methanogenic population dynamics during start-

up of anaerobic digesters treating municipal solid waste and biosolids." Biotechnology

and Bioengineering 57(3): 342-355.

Hills, D. J. and M. Kayhanian (1985). "Methane from settled and filtered flushing dairy wastes."

Transactions of the ASAE 28(3): 865-869.

Hulshoff Pol, L. W., K. Heijnekamp, et al. (1988). The selection pressure as a driving force

behind the granulation of anaerobic sludge. Granular anaerobic sludge: microbiology and

technology. G. Lettinga, A. J. B. Zehnder, J. T. C. Grotenhuis and L. W. Hulshoff Pol.

Netherlands, Wageningen: 153-161.

Lee, C. C. (2000). Handbook of environmental engineering calculations. New York, McGraw-

Hill.

19

Lettinga, G., S. Rebac, et al. (2001). "Challenge of psychrophilic anaerobic wastewater

treatment." Trends in Biotechnology 19(9): 363-370.

Liu, Y., Z. W. Wang, et al. (2005). "Selection pressure-driven aerobic granulation in a

sequencing batch reactor." Appl Microbiol Biotechnol 67(1): 26-32.

Lott, S. C., R. J. Loch, et al. (1994). "Settling characteristics of feedlot cattle feces and manure."

Transactions of the American Society of Agricultural Engineers 37(1): 281-285.

Mackie, R. I., P. G. Stroot, et al. (1998). "Biochemical identification and biological origin of key

odor components in livestock waste." Journal of Animal Science 76(5): 1331-1342.

McMahon, K. D., P. G. Stroot, et al. (2001). "Anaerobic codigestion of municipal solid waste

and biosolids under various mixing conditions - II: Microbial population dynamics."

Water Research 35(7): 1817-1827.

Meyer, D. M., I. Garnett, et al. (1997). "A survey of dairy manure management practices in

California." Journal of Dairy Science 80(8): 1841-1845.

Nikolaeva, S., E. Sanchez, et al. (2009). "Kinetics of anaerobic degradation of screened dairy

manure by upflow fixed bed digesters: Effect of natural zeolite addition." Journal of

Environmental Science and Health Part a-Toxic/Hazardous Substances & Environmental

Engineering 44(2): 146-154.

Pavlostathis, S. G. and E. Giraldogomez (1991). "KINETICS OF ANAEROBIC TREATMENT

- A CRITICAL-REVIEW." Critical Reviews in Environmental Control 21(5-6): 411-490.

Qin, L., Y. Liu, et al. (2004). "Effect of settling time on aerobic granulation in sequencing batch

reactor." Biochemical Engineering Journal 21(1): 47-52.

Qin, L., J. H. Tay, et al. (2004). "Selection pressure is a driving force of aerobic granulation in

sequencing batch reactors." Process Biochemistry 39(5): 579-584.

20

Sanders, W. (2001). Anaerobic Hydrolysis During Digestion of Complex Substrates. Ph.D.,

Wageningen University, The Netherlands.

Schwarzenbeck, N., J. M. Borges, et al. (2005). "Treatment of dairy effluents in an aerobic

granular sludge sequencing batch reactor." Applied Microbiology and Biotechnology

66(6): 711-718.

Stroot, P. G., K. D. McMahon, et al. (2001). "Anaerobic codigestion of municipal solid waste

and biosolids under various mixing conditions - I. Digester performance." Water

Research 35(7): 1804-1816.

Tomei, M., C. Braguglia, et al. (2009). "Modeling of Anaerobic Digestion of Sludge." Critical

Reviews in Environmental Science and Technology 39(12): 1003-1051.

U.S. EPA (2007) " Guide to Anaerobic Digesters." Anaerobic Digester Database DOI:

http://www.epa.gov/agstar/operational.html.

Umana, O., S. Nikolaeva, et al. (2008). "Treatment of screened dairy manure by upflow

anaerobic fixed bed reactors packed with waste tyre rubber and a combination of waste

tyre rubber and zeolite: Effect of the hydraulic retention time." Bioresource Technology

99(15): 7412-7417.

USDA (2005). Agricultural Statistics, National Agricultural Statistics Service, States

Government Printing Office, Washington, DC.

Vartak, D. R., C. R. Engler, et al. (1996). Attached-film media performance in psychrophilic

anaerobic treatment of dairy cattle wastewater. 7th National Bioenergy Conference,

Nashville, Tennessee.

Vartak, D. R., C. R. Engler, et al. (1997). "Attached-film media performance in psychrophilic

anaerobic treatment of dairy cattle wastewater." Bioresource Technology 62(3): 79-84.

21

Vavilin, V. A., S. V. Rytov, et al. (1996). "A description of hydrolysis kinetics in anaerobic

degradation of particulate organic matter." Bioresource Technology 56(2–3): 229-237.

Vesilind, P. A. (2003). Wastewater treatment plant design. London, IWA Pub.

Weiske, A., A. Vabitsch, et al. (2006). "Mitigation of greenhouse gas emissions in European

conventional and organic dairy farming." Agriculture Ecosystems & Environment 112(2-

3): 221-232.

Wilderer, P. A., R. L. Irvine, et al. (2001). Sequencing Batch Reactor Technology. London, UK,

IWA.

Wilkie, A. C., H. F. Castro, et al. (2004). "Fixed-film anaerobic digestion of flushing dairy

manure after primary treatment: Wastewater production and characterisation."

Biosystems Engineering 89(4): 457-471.

Zaher, U., C. Frear, et al. (2008). "Evaluation of a new fixed-bed digester design utilizing large

media for flush dairy manure treatment." Bioresour Technol 99(18): 8619-8625.

Zinder, S. H. (1993). Physiological ecology of methanogens. Methanogenesis: Ecology,

Physiology, Biochemistry and Genetics. J. G. Ferry. New York, Chapman & Hall: 128–

206.

22

CHAPTER TWO

DEVELOPMENT OF A SIMPLE METHODOLOGY FOR RATE-LIMITING STEP

DETERMINATION FOR ANAEROBIC DIGESTION OF SOLIDS CONTAINING

SUBSTRATE AND EFFECT OF MICROBIAL COMMUNITY RATIO

2.1. Abstract

Anaerobic digestion of complex substrates is a multi-step process, which is kinetically controlled

by an individual rate-limiting step. Balanced anaerobic microbial community structure plays a

pivotal role in an efficient anaerobic digestion process. A methodology for rate-limiting step

determination for anaerobic digestion of complex substrates was developed by supplementation

of metabolic intermediates from each step of the digestion process. This method elucidated that

hydrolysis of dairy manure was the rate-limiting step when normal anaerobic sludge was used as

inoculum. Furthermore, the concept of microbial community ratio in the anaerobic degradation

system was first introduced in this study and investigated by manipulating two different inocula,

i.e. normal anaerobic sludge (NS) and heated anaerobic sludge (HS). The results revealed that

the rate-limiting step changed according to the variation of microbial community ratio (r).

Results indicated a critical ratio r*=24 between hydrolytic bacteria (H) to methanogens (M) at

which as r decreased or exceeded from this value, hydrolysis or methanogenesis limited the

anaerobic digestion process, respectively.

Key words: Rate-limiting step, kinetics, microbial community ratio, anaerobic digestion, dairy

manure

23

2.2. Introduction

The anaerobic digestion of complex organic material is a multi-step process that consists of four

stages in series: hydrolysis, acidogenesis, acetogenesis and methanogenesis (Batstone et al.,

2002;Pavlostathis and Giraldogomez, 1991). In such a multi-step complex process, the kinetics

of the slowest step will account for the overall kinetics. Anaerobic digestion systems are limited

by two major steps depending on the nature of the substrate. Hydrolysis is often limited if the

substrate is complex organic solids while the rate-limiting step in the digestion of soluble organic

matter is methanogenesis (Tomei et al., 2009;Vavilin et al., 1996). Determination of rate-limiting

step is critical for anaerobic digestion process design treating a specific feedstock, establishment

of a stable process performance, and management anaerobic digestion processes. Hence, there is

a need for an experimental method capable of determining precisely which of the different steps

is limiting for a complex substrate.

Anaerobic degradation process is dependent not only on the physic-chemical characteristics of

the substrates, but also on the concentration and quality of anaerobic microbial community

composed of symbiotic microbes responsible for each disparate metabolic step. The stability of

the process is dependent on the inoculum mass and critical balance of different trophic groups

(Angelidaki et al., 2009). The impact of inoculum concentration on biochemical methane

potential (BMP) for various organic wastes has been widely studied (Chen and Hashimoto,

1996;Eskicioglu and Ghorbani, 2011;Gonzalez-Fernandez and Garcia-Encina, 2009;Hashimoto,

1989;Koksoy and Sanin, 2010;Liu et al., 2009;Lopes et al., 2004;Neves et al., 2004;Raposo et al.,

2006;Raposo et al., 2009;Zhou et al., 2011). Relatively high hydrolysis rates were reached in

anaerobic biodegradability tests with a high inoculum to substrate ratio, showing some degree of

24

dependence of hydrolysis to biomass concentration or activity (Fernandez et al., 2001). Similarly,

it is believed that inoculum concentration has influence on methane production rate and methane

yield. Equally important is the population or composition of the microbial community. For

examples, the microbial community ratio, defining the ratio of microbial concentration of each

symbiotic growth group, is a key parameter governing the anaerobic digestion process. Balanced

microbial population can be translated into proper microbial community ratio. However, research

on microbial community ratio and its effect on kinetics of anaerobic degradation process is

limited and is the focus of this research paper.

Therefore, the aim of the present work was to develop a simple method for rate-limiting step

evaluation through kinetic characteristics of a series batch tests with metabolic intermediates

generated from the conversion of macro-molecule to methane and carbon dioxide. Moreover, a

new concept, microbial community ratio, i.e. hydrolytic bacteria to methanogens ratio, was first

proposed in this paper, and its effect on the kinetics of anaerobic digestion process was

investigated with two different sources of inocula.

2.3. Methods

The methodology developed for rate-limiting step identification is based on a modified

biochemical methane potential (BMP) assay. Except for substrate and inoculum, metabolic

intermediates produced from each step were also supplemented into closed vessels which were

incubated at 35ºC and daily methane production was monitored. Experiments were conducted for

as long as required to reach a plateau in methane production. The rate-limiting step was then

evaluated based on BMP data and some assumptions (see descriptions in Section 2.3.2).

25

2.3.1. Substrate and inoculum

Flushing dairy manure, a representative of complex substrate with undetermined

hydrolysis/methanogenesis rate-limiting step, was collected from the Washington State

University Dairy Center in Pullman, WA, USA and stored at 4ºC prior to use. Before adding into

reactors, manure was screened through a 2 mm sieve to remove coarse solids, which resulted in

mixed liquor containing 9.1 g/L total solids (TS) and 7.6 g/L total volatile solids (VS). Anaerobic

sludge was sampled from an anaerobic digester at the Pullman Wastewater Treatment Facility

with TS of 17.1 g/L and VS of 11.7 g/L. Two different inocula, namely normal anaerobic

digester sludge (NS) and heated anaerobic digester sludge (HS), were used in this study. The HS

was anaerobic digester sludge heated in an oven at 105 ºC for 2 h in order to kill non-spore-

forming methanogens (Logan et al., 2002) then cooled to room temperature in a desiccator. As a

result, NS included all three groups of microorganisms, namely hydrolytic bacteria, acidogens,

and methanogens while HS consisted of hydrolytic bacteria and acidogens only.

2.3.2. Rate-limiting step evaluation

The aforementioned dairy manure, the basic substrate, was added to serum bottles with working

volume of 200 mL. Glucose was considered as the main intermediate of hydrolysis of

carbohydrate complex organics while sodium acetate was used as model intermediate of

acidification. Dairy manure, glucose, sodium acetate and NS were added into bottles according

to the experimental design summarized in Table 2.1. The concentration of glucose, sodium

acetate, and NS were 3 gCOD/L, 3 gCOD/L, and 1 gVS/L, respectively. A bottle filled with

dairy manure only served as a control while a bottle with NS only served as a blank. Each bottle

was magnetically stirred at a speed of 300 rpm and maintained at 35 ºC. Methane production was

26

monitored using a respirometer (Challenge technology AER-200 respirometer, Springdale, AR,

USA). Sodium hydroxide was used as scrubbing media to purify methane by adsorbing carbon

dioxide and hydrogen sulfide, so that only methane production was recorded. Triplicate analysis

was carried out to ensure reproducibility of results. To identify the rate-limiting step in anaerobic

digestion of dairy manure, the following assumptions were made: (1) if glucose increased

methane production rate while acetate did not, then hydrolysis would be the rate-limiting step; (2)

if acetate increased methane production rate while glucose did not, then acidification would be

the limiting step; and (3) if both glucose and acetate did not affect methane production rate, then

methanogenesis limits the overall process. Usually, the anaerobic digestion process was

considered either limited by hydrolysis or methanogenesis, while acidification step was not

expected as rate-limiting step since acidogenesis is usually the fastest reaction in the anaerobic

conversion of complex substrates during digestion (Mosey and Fernandes, 1989).

Table 2.1 Experimental design for rate-limiting step evaluation*

No 1 2 3 4 5 6 7

Substrate DM DM DM DM DM DM -

Inoculum - - - NS NS NS NS

Intermediates - GL SA GL SA -

* All experiments were conducted in triplicate.

27

2.3.3. Effect of microbial community ratio

Experiments for the influence of microbial community ratio on anaerobic digestion of dairy

manure were conducted in a series of serum bottles with working volume of 200 mL. Methane

production was monitored using a respirometer and maintained at the same operating conditions

as the rate-limiting step evaluation experiment. Dairy manure was autoclaved at 121 ºC and 15

psi for 30 min (PB-series Steam Sterilizers, Consolidated Stills & Sterilizers, Boston, MA, USA)

in order to sterilize all of the microorganisms. Different volumes of NS and HS were added into

bottles 1 to 22 giving rise to a wide spectrum of NS to HS ratios according to the experimental

design presented in Table 2.2. Bottle 23 and 24 were used as blanks. The total sludge

concentration in each bottle was 2 gVS/L. Triplicate analysis was also carried out to ensure

reproducibility of results.

2.3.4. Analytical methods

Total solids, VS analyses were done according to Standard Methods (APHA, 1998). Biomass

concentration was indirectly determined by measuring adenosine 5’-triphosphate (ATP)

concentration using a method detailed in (Wang et al., 2011). Briefly, samples were mixed with

Tris/EDTA buffer in 1:10 volumetric ratio and boiled for 90 seconds to extract ATP. After

equilibration to room temperature, the ATP extract was reacted with an ATP determination kit

(A-22066, Molecular Probes, Eugene, OR) and the luminescence intensity was measured with a

SpectraMax L microplate luminometer (MDS Analytical Technologies, USA).

28

Table 2.2 Microbial community ratio experimental designa.

No NS

(gVS/L)

HS

(gVS/L)

R r

1 2.00 0.00 - 2.65

2 1.60 0.40 4.00 3.32

3 1.33 0.67 2.00 3.98

4 1.00 1.00 1.00 5.30

5 0.67 1.33 0.50 7.96

6 0.50 1.50 0.33 10.61

7 0.40 1.60 0.25 13.26

8 0.33 1.67 0.20 15.91

9 0.29 1.71 0.17 18.57

10 0.25 1.75 0.14 21.22

11 0.22 1.78 0.13 23.87

12 0.18 1.82 0.10 29.17

13 0.15 1.85 0.08 34.48

14 0.13 1.87 0.07 39.78

15 0.12 1.88 0.06 45.09

16 0.11 1.89 0.06 50.39

17 0.10 1.90 0.05 55.70

18 0.08 1.92 0.04 68.96

19 0.07 1.93 0.03 79.57

20 0.04 1.96 0.02 135.26

21 0.02 1.98 0.01 267.87

22 0.00 2.00 0.00 -

23 0.00 0.00 - -

24b 2.00 0.00 - - a All experiments were conducted in triplicate.

b manure was added into each bottle except for bottle 24.

29

2.3.5. Kinetics model

The modified Gompertz model was adopted to characterize methane production and microbial

community ratio model was developed to investigate microbial community ratio on rate-limiting

step. The values of parameters for each experiment were estimated using the solver function in

Microsoft Office Excel (2007, Microsoft) with a Newtonian algorithm.

2.3.5.1. Biogas production simulation

The modified Gompertz equation 2.1 was developed to describe the cumulative methane

production curve in a batch culture (Lay et al., 1997;Zwietering et al., 1990).

𝐺 = 𝑃 exp �−exp �𝑅𝑚𝑒𝑃

(𝜆 − 𝑡) + 1�� (2.1)

When glucose was supplemented, a biphasic methane production curve was presented. To

characterize each phase, eq. (2.1) was used twice to separately best-fit the two methane

production phases which resulted in two sets of kinetic parameters. Cumulative methane

production exceeding 95% of methane potential (P1) for the first phase was taken to be the end

of the first phase and as well as the beginning of the second phase.

2.3.5.2. Kinetic model for microbial community ratio (r) affecting rate-limiting step

The ratio of two inocula, which contain different proportion of hydrolytic bacteria and

methanogens, may affect the anaerobic digestion process. There should be a critical R* of the NS

to HS ratio; when R < R*, methanogenesis could be rate-limiting step; while when R > R*,

hydrolysis could be rate-limiting step, which can be represented as eq. (2.2)

30

𝑣 = � 𝑞𝑀 ∙ 𝑋𝑀, 𝑅 < 𝑅∗ 𝑞𝐻 ∙ 𝑋𝐻, 𝑅 > 𝑅∗ (2.2)

in which

𝑋𝑀 =𝑅

1 + 𝑅𝛼𝑀𝑋 (2.3)

𝑋𝐻 = 𝛼𝐻𝑋 (2.4)

𝑞𝑀 =𝜇𝑀𝑌𝑀

(2.5)

𝑞𝐻 =𝜇𝐻𝑌𝐻

(2.6)

Substitute eq. (3) – (6) into (2) yields:

𝑣 = �

𝑅1 + 𝑅

𝛼𝑀𝑋𝜇𝑀𝑌𝑀

, 𝑅 < 𝑅∗

𝛼𝐻𝑋𝜇𝐻𝑌𝐻

, 𝑅 > 𝑅∗ (2.7)

By definition, r represents the true ratio of hydrolytic bacteria to methanogens in the digester and

significantly affects the anaerobic digestion process. The critical r* for the H to M ratio was

hypothesized here. When r < r*, hydrolysis should be rate-limiting step while when r > r*,

methanogenesis should be rate-limiting step, which can be represented as eq. (8)

𝑣 = � 𝑞𝐻 ∙ 𝑋𝐻, 𝑟 < 𝑟∗ 𝑞𝑀 ∙ 𝑋𝑀, 𝑟 > 𝑟∗ (2.8)

31

in which

𝑋𝑀 =1𝑟𝛼𝐻𝑋 (2.9)

Substitute eq. (4) – (6) and (9) into (8) yielding:

𝑣 =

⎩⎨

⎧𝛼𝐻𝑋𝜇𝐻𝑌𝐻

, 𝑟 < 𝑟∗

1𝑟𝛼𝐻𝑋

𝜇𝑀𝑌𝑀

, 𝑟 > 𝑟∗ (2.10)

2.4.Results and discussion

2.4.1. Rate-limiting step evaluation during the anaerobic digestion

The methane production profile in each bottle was closely related to substrate and metabolic

intermediates supplementation. The two tests with glucose supplementation experienced a

biphasic methane production process. As Figure 2.1 shows, the steep slope of methane

production rate at the beginning indicated significant increase of methane production with

degradation of glucose. Gas production started immediately without lag phase after incubation.

Theoretically, total glucose or acetate supplemented in bottles can be converted to 217 mL

methane at 35 ºC. The experimental data shows a good agreement with theoretical calculation.

When glucose was depleted, methane production stopped before manure degradation. The

methane production rate elevated again when the microbes were adapted to the manure, with the

highest total methane production compared with other tests. The methane production process

with acetate supplementation exhibited a similar trend as that of manure control but with higher

maximum methane production rate and higher total methane production as expected with the

theoretical calculation. Sludge supplementation shortened the lag phase prominently. Moreover,

32

the total methane production was enhanced, which indicated that high microbial concentration

can help to overcome process inhibition, or anaerobic sludge microbe is capable of consuming

certain substrate that manure microbes are unable to consume. Chen and Hashimoto reported a

similar result that increase of inoculum improved both methane production rate and ultimate

methane yield (Chen and Hashimoto, 1996;Hashimoto, 1989).

Figure 2.1 Cumulative methane productions with metabolic intermediates supplementation.

Profiles are from triplicated experiments; values represent average and error bars mean standard

deviation

Time (d)

0 10 20 30 40

Acc

umul

ated

met

han

prod

uctio

n (m

l)

0

200

400

600

800

1000

1200

DMDM+NSDM+GL DM+SA DM+NS+GLDM+NS+SANS

33

The kinetics parameters varied dramatically due to metabolic intermediates supplementation.

Table 2.3 compared the kinetics parameters calculated with eq. (2.1) for each condition. As

regards P1, it seems nearly all of the glucose was converted into methane and carbon dioxide on

the basis of this calculation (in Table 2.3, 185 mL increase from DM+GL vial compared with

DM vial, and 198 mL increase from DM+NS+GL vial compared with DM+NS vial). Since

glucose was consumed at the very beginning of the test, Rm1 was affected by glucose

supplementation. Rm1 increased dramatically (almost 3 times increase compared with that of

control) when glucose was supplemented. However, Rm only rose about one fourth by sodium

acetate and sludge supplementation. In accordance with the given assumptions, it can be

concluded that the hydrolysis of carbohydrate could be limiting the overall anaerobic digestion

rate. Noike et al. (Noike et al., 1985) studied the characteristics of carbohydrate degradation and

found the rate of cellulose hydrolysis was so low that it was shown to be the rate-limiting step in

overall anaerobic digestion.

This study demonstrated a simple and fast approach to identify the rate-limiting step in anaerobic

digestion of a complex substrate. First, metabolic intermediates produced from each step of the

anaerobic digestion process are supplemented into BMP test vials containing the testing substrate.

Then based on the methane production profile, a maximum methane production rate (Rm) is

calculated. The rate-limiting step can be identified according to the notably increased methane

production rate. The step that produced the particular intermediate which elevate Rm, limits the

overall digestion process. The prerequisite of this method is to understand the major composition

of the substrate, such as carbohydrate, protein and lipid, so that possible intermediates can be

chosen as candidates for each step of degradation. The results also showed that the lag phase (𝜆)

34

was considerably affected by microbe concentration. In each test when sludge was supplemented,

𝜆 was shortened to half of that from control, which indicates anaerobic digestion of dairy manure

can be accelerated by dosing anaerobic sludge seed and potentially can reduce hydraulic

retention time (HRT) of digester. Additionally, the ultimate biogas yield was also augmented by

sludge supplementation, showing the possibility of more energy yield in a biogas plant when

operated with higher biomass concentration.

Table 2.3 Kinetics parameters for methane production with metabolic intermediates

supplementation

Substrate λ Rm P

(d) (mL/d) (mL)

DM 11.6 44.7 716

DM+NS 6.4 55.3 751a

DM+GL 0.0 (λ1)b

11.3 (λ2) b

173.6 ( Rm1) b

56.7 ( Rm2) b

212 (P1) b

689 (P2) b

DM+SA 11.5 56.7 852

DM+NS+GL 0.0 (λ1) b

6.9 (λ2) b

191.2 ( Rm1) b

56.4 ( Rm2) b

242 (P1) a, b

707 (P2) b

DM+NS+SA 6.5 54.3 853 a a The methane production potential has been deducted with that from NS;

b λ1, Rm1, and P1 represent parameters for first phase; λ2, Rm2, and P2 represent parameters for

second phase.

35

2.4.2. Effect of microbial community ratio on kinetics of anaerobic digesting dairy manure

The maximum methane production rate was obviously affected by the ratio of two inocula as

shown in Figure 2.2. The concentration of methanogens elevated with growth of R while

concentration of hydrolytic bacteria remained unchanged. As a result of that, the maximum

methane production rate increased with the NS to HS ratio (R) before a plateau was reached. It

seems that a concentration enhancement of one species may not always ensure a rate elevation of

anaerobic digestion process, or a balanced microbial structure is desired. The relationship

between the maximum methane production rate and R was simulated with eq. (2.7). As Figure

2.2 shows, the simulated curve exhibits a good agreement with experimental data. The

simulation profile revealed that there existed a critical R* at 0.13. When R < R*, methane

production rate was hyperbolically related to R, where the whole process was governed by

methanogenesis; when R > R*, methane production rate remained unchanged with the raise of R,

where hydrolysis limits the anaerobic digestion process.

Effects of H to M ratio r on rate-limiting step can be derived from above results. ATP analysis

indicated that 𝛼𝐻 and 𝛼𝑀 in NS was 72.6% and 27.4%, respectively, giving rise to an H to M

ratio of 2.65 in NS. Thus it is possible to obtain a wide range of r from 2.65 to infinity via

blending certain proportions of NS and HS as inocula. Concentration of methanogens decreased

with a raise in r but concentration of hydrolytic bacteria stayed the same. The maximum methane

production rate kept constant as r was raised to a certain value then decreased until approaching

zero. The simulation of eq. (2.10) with experimental data is shown in Figure 2.3. A critical r* of

24 was found from the curve. When r < r*, methane production rate held the line with the

increase of r, where hydrolysis limits the anaerobic digestion process; when r > r*, methane

36

production rate was inversely proportional to r, where the whole process was governed by

methanogenesis. Due to the low r value of 2.65 in normal anaerobic sludge compared with r*of

24, the hydrolysis of complex substrates is the rate-limiting step of the whole process with

normal anaerobic sludge as inoculum degrading dairy manure. The proportion of hydrolytic

bacteria in the inoculum should be elevated to the same order of r* to overcome this kinetic

limitation.

Figure 2.2 Maximum methane production rate changes against NS to HS ratio R, (●)

experimental results, and (—) simulation profile

R

0.0 0.2 0.4 0.6 0.8 1.0

Max

imum

met

hane

pro

duct

ion

rate

(ml/l

/d)

0

20

40

60

80

100

Hydrolysis rate-limitingMethanogenesisrate-limiting

v = k1R/(1+R)k1 = 469.24R2 = 0.88

v = c1

c1 = 51.98R2 = 0.82

37

This study showed that the microbial community ratio has a strong effect on the anaerobic

digestion rate. The kinetics of methane production is determined by both the quantity and the

characteristics of microorganisms. Hydrolysis and methanogenesis are usually considered as the

major rate-limiting steps during anaerobic digestion of complex substrate. The volatile fatty

acids (VFA) produced by hydrolysis/acidogenesis can depress the acetogenesis and

methanogenesis (Vavilin and Angelidaki, 2005). To balance the VFA production and

consumption, a balanced population of these two groups of microorganisms (hydrolytic bacteria

and methanogens) is indispensable. This balance may be affected by type of feedstock, organic

loading rate, temperature, pH, inhibitors, etc. If the rates of hydrolysis and methanogenesis are

unbalanced, the system either accumulates VFA or suffers a low methane production.

A balanced microbial community could be reached by adding other sources of biomass to adjust

r back to a proper range. Addition of cellulolytic bacteria enriched culture, cow rumen

microorganisms for instance, could improve r when hydrolysis of cellulosic substrate is limiting

the process (Weimer et al., 2009). If methanogenesis is the rate-limiting step or amount of

methanogens is not enough to consume the VFA produced during acetogenesis, accumulation of

VFA and pH drop will inhibit hydrolysis step, and subsequently limit the overall anaerobic

biodegradation (Angelidaki et al., 1993;Salminen and Rintala, 2002;Vavilin, 2010). Then acetate

feeding sludge abundant with methanogens could be added to the digester so that r would be

regulated to the appropriate point, resulting in an accelerated VFA consumption and mitigated

inhibition. Hence, the modification of microbial community structure in terms of H to M ratio

could promote both hydrolysis and methanogenesis and speed up the anaerobic digestion process.

38

Figure 2.3 Maximum methane production rate changes against H to M ratio r, (●) experimental

results, and (—) simulation profile

2.4.3. Implications for anaerobic co-digestion

Anaerobic co-digestion studies have gained momentum in recent years (Li et al., 2009;Romano

and Zhang, 2008;Yen and Brune, 2007). It is believed that the dilution of potential toxic

compounds, improved balance of nutrients, and synergistic effects of microorganisms by co-

digestion enhance biogas yield (Labatut et al., 2011). In addition, microbial community ratio

would be another factor important to co-digestion and biogas yield. Mono-digestion of dairy

manure was limited by hydrolysis step so that an inoculum with higher hydrolytic bacteria

concentration (i.e., higher r value) is preferred to balance the digestion process. Food waste is

r

0 20 40 60 80

Max

imum

met

hane

pro

duct

ion

rate

(ml/l

/d)

0

20

40

60

80

100

Methanogenesis rate-limitingHydrolysis rate-limiting

v = k2/rk2 = 1248.71R2 = 0.88

v = c2

c2 = 51.98R2 = 0.61

39

relatively easily degradable and therefore high methanogens concentration (i.e., lower r value) is

required to convert VFA as soon as possible and prevent digester from souring. When co-

digestion of dairy manure and food waste was employed, characteristics of feedstock can be

adjusted to fit for a certain inoculum, or a certain microbial community ratio. Therefore, an

enhanced digester performance could be accomplished by the co-digestion of two substrates with

suitable proportion, which could be derived based on microbial community ratio of inoculum

used.

2.5. Conclusion

A methodology described in this research showed that a simple procedure, namely revised BMP

test with metabolic intermediates supplementation, could be used to determine the rate-limiting

steps in the degradation of complex substrates. With this method, it was illustrated that

hydrolysis of dairy manure was the rate-limiting step when anaerobic sludge was used as

inoculum. The results also showed that an increase of microbial concentration could shorten the

digester start-up time and elevate the efficiency of the anaerobic digestion process.

The effects of microbial community ratio on the anaerobic degradation of dairy manure were

investigated by manipulating two different inocula. Results revealed that the rate-limiting step

changed with the variation of microbial community ratio. The relationship between methane

production rate and ratio of H to M was further derived with ATP data. It showed that there

exists a critical ratio r* at 24 regarding dairy manure; a value below which yields hydrolysis as a

potential rate-limiting step; conversely, when r exceeded this value, methanogenesis limits the

anaerobic digestion process.

40

2.6. References

Angelidaki, I., Ellegaard, L. and Ahring, B.K. (1993) A mathematical model for dynamic

simulation of anaerobic digestion of complex substrates: Focusing on ammonia inhibition.

Biotechnology and Bioengineering 42(2), 159-166.

Angelidaki, I., Alves, M., Bolzonella, D., Borzacconi, L., Campos, J.L., Guwy, A.J., Kalyuzhnyi,

S., Jenicek, P. and van Lier, J.B. (2009) Defining the biomethane potential (bmp) of solid

organic wastes and energy crops: A proposed protocol for batch assays. Water Sci

Technol 59(5), 927-934.

APHA (1998) Methods for examination of water and wasterwater, American Public Health

Association/American Water Works Association/Water Environment Federation,

Washington, DC, USA.

Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., S.G., P., A., R., WTM., S., H., S. and

V.A., V. (2002) The iwa anaerobic digestion model no 1 (adm1). Water Science and

Technology 45(10), 9

Chen, T.-H. and Hashimoto, A.G. (1996) Effects of ph and substrate:Inoculum ratio on batch

methane fermentation. Bioresource Technology 56(2-3), 179-186.

Eskicioglu, C. and Ghorbani, M. (2011) Effect of inoculum/substrate ratio on mesophilic

anaerobic digestion of bioethanol plant whole stillage in batch mode. Process

Biochemistry 46(8), 1682-1687.

Fernandez, B., Porrier, P. and Chamy, R. (2001) Effect of inoculum-substrate ratio on the start-

up of solid waste anaerobic digesters. Water Science and Technology 44(4), 103-108.

Gonzalez-Fernandez, C. and Garcia-Encina, P.A. (2009) Impact of substrate to inoculum ratio in

anaerobic digestion of swine slurry. Biomass & Bioenergy 33(8), 1065-1069.

41

Hashimoto, A.G. (1989) Effect of inoculum/substrate ratio on methane yield and production rate

from straw. Biological Wastes 28(4), 247-255.

Koksoy, G.T. and Sanin, F.D. (2010) Effect of digester f/m ratio on gas production and sludge

minimization of ultrasonically treated sludge. Water Science and Technology 62(7),

1510-1517.

Labatut, R.A., Angenent, L.T. and Scott, N.R. (2011) Biochemical methane potential and

biodegradability of complex organic substrates. Bioresource Technology 102(3), 2255-

2264.

Lay, J.-J., Li, Y.-Y. and Noike, T. (1997) Influences of ph and moisture content on the methane

production in high-solids sludge digestion. Water Research 31(6), 1518-1524.

Li, R., Chen, S., Li, X., Saifullah Lar, J., He, Y. and Zhu, B. (2009) Anaerobic codigestion of

kitchen waste with cattle manure for biogas production. Energy & Fuels 23(4), 2225-

2228.

Liu, G., Zhang, R., El-Mashad, H.M. and Dong, R. (2009) Effect of feed to inoculum ratios on

biogas yields of food and green wastes. Bioresource Technology 100(21), 5103-5108.

Logan, B.E., Oh, S.-E., Kim, I.S. and Van Ginkel, S. (2002) Biological hydrogen production

measured in batch anaerobic respirometers. Environmental Science & Technology 36(11),

2530-2535.

Lopes, W.S., Leite, V.D. and Prasad, S. (2004) Influence of inoculum on performance of

anaerobic reactors for treating municipal solid waste. Bioresource Technology 94(3),

261-266.

Mosey, F.E. and Fernandes, X.A. (1989) Patterns of hydrogen in biogas from the anaerobic

digestion of milk sugars. Water Science and Technology 21, 10.

42

Neves, L., Oliveira, R. and Alves, M.M. (2004) Influence of inoculum activity on the bio-

methanization of a kitchen waste under different waste/inoculum ratios. Process

Biochemistry 39(12), 2019-2024.

Noike, T., Endo, G., Chang, J.-E., Yaguchi, J.-I. and Matsumoto, J.-I. (1985) Characteristics of

carbohydrate degradation and the rate-limiting step in anaerobic digestion. Biotechnology

and Bioengineering 27(10), 1482-1489.

Pavlostathis, S.G. and Giraldogomez, E. (1991) Kinetics of anaerobic treatment - a critical

review. Critical Reviews in Environmental Control 21(5-6), 411-490.

Raposo, F., Banks, C.J., Siegert, I., Heaven, S. and Borja, R. (2006) Influence of inoculum to

substrate ratio on the biochemical methane potential of maize in batch tests. Process

Biochemistry 41(6), 1444-1450.

Raposo, F., Borja, R., Martin, M.A., Martin, A., de la Rubia, M.A. and Rincon, B. (2009)

Influence of inoculum-substrate ratio on the anaerobic digestion of sunflower oil cake in

batch mode: Process stability and kinetic evaluation. Chemical Engineering Journal

149(1-3), 70-77.

Romano, R.T. and Zhang, R. (2008) Co-digestion of onion juice and wastewater sludge using an

anaerobic mixed biofilm reactor. Bioresource Technology 99(3), 631-637.

Salminen, E.A. and Rintala, J.A. (2002) Semi-continuous anaerobic digestion of solid poultry

slaughterhouse waste: Effect of hydraulic retention time and loading. Water Research

36(13), 3175-3182.

Tomei, M., Braguglia, C., Cento, G. and Mininni, G. (2009) Modeling of anaerobic digestion of

sludge. Critical Reviews in Environmental Science and Technology 39(12), 1003-1051.

43

Vavilin, V. (2010) Anaerobic degradation of organic waste: An experience in mathematical

modeling. Microbiology 79(3), 334-341.

Vavilin, V.A., Rytov, S.V. and Lokshina, L.Y. (1996) A description of hydrolysis kinetics in

anaerobic degradation of particulate organic matter. Bioresource Technology 56(2–3),

229-237.

Vavilin, V.A. and Angelidaki, I. (2005) Anaerobic degradation of solid material: Importance of

initiation centers for methanogenesis, mixing intensity, and 2d distributed model.

Biotechnology and Bioengineering 89(1), 113-122.

Wang, Z.-W., Ma, J. and Chen, S. (2011) Bipolar effects of settling time on active biomass

retention in anaerobic sequencing batch reactors digesting flushing dairy manure.

Bioresource Technology 102(2), 697-702.

Weimer, P.J., Russell, J.B. and Muck, R.E. (2009) Lessons from the cow: What the ruminant

animal can teach us about consolidated bioprocessing of cellulosic biomass. Bioresource

Technology 100(21), 5323-5331.

Yen, H.-W. and Brune, D.E. (2007) Anaerobic co-digestion of algal sludge and waste paper to

produce methane. Bioresource Technology 98(1), 130-134.

Zhou, Y., Zhang, Z., Nakamoto, T., Li, Y., Yang, Y., Utsumi, M. and Sugiura, N. (2011)

Influence of substrate-to-inoculum ratio on the batch anaerobic digestion of bean curd

refuse-okara under mesophilic conditions. Biomass and Bioenergy 35(7), 3251-3256.

Zwietering, M.H., Jongenburger, I., Rombouts, F.M. and Van 'T Riet, K. (1990) Modeling of the

bacterial growth curve. Appl. Environ. Microbiol. 56(6), 1875-1881.

44

2.7. Nomenclature

DM – dairy manure

GL – glucose

SA – sodium acetate

NS – normal anaerobic sludge

HS – heated anaerobic sludge

H – hydrolytic bacteria

M − methanogens

G – cumulative methane production (mL)

λ – lag time (d)

P – methane production potential (mL)

Rm – maximum methane production rate (mL/d)

e – 2.718281828

R – NS to HS ratio

r – H to M ratio

𝑣 – methane production rate (mL/L/d)

𝑞𝐻 – specific substrate utilization rate when hydrolysis is rate-limiting step (gCOD/gVS)

𝑞𝑀 − specific substrate utilization rate when methanogenesis is rate-limiting step (gCOD/gVS)

𝑋 − total biomass concentration (gVS)

𝑋𝐻 − concentration of hydrolytic bacteria (gVS)

𝑋𝑀 − concentration of methanogens (gVS)

𝛼𝐻 − percentage of hydrolytic bacteria in biomass

𝛼𝑀 − percentage of methanogens in biomass

45

𝜇𝐻 – specific hydrolytic bacteria growth rate (/d)

𝜇𝑀 – specific methanogens growth rate (/d)

𝑌𝐻 – yield coefficient of hydrolytic bacteria (gVS/gCOD)

𝑌𝑀 – yield coefficient of methanogens (gVS/gCOD)

46

CHAPTER THREE

BIPOLAR EFFECTS OF SETTLING TIME ON ACTIVE BIOMASS RETENTION IN

ANAEROBIC SEQUENCING BATCH REACTORS DIGESTING FLUSHING DAIRY

MANURE

3.1. Abstract

Active biomass retention is a technical challenge in anaerobic digester treating dilute animal

manure. A strategy was tested using fibers in the dairy manure as biomass carriers by controlling

settling time. Settling time ranging from 0.5 to 60 min were applied to eight anaerobic

sequencing batch reactors to investigate their effects on active biomass retention in anaerobic

digestion of flushing dairy manure. Results revealed a critical settling time of 2 min at which

only minimum amount of active biomass was retained, while biomass retention improved on

either lower or higher settling time. Gravity settling and selection pressure theories were

suggested to account for the results. A model integrating these two effects was developed and

verified with the experimental data. Knowledge derived from this study may lead to innovative

bacterial retention technology for cost-effective anaerobic digestion of dairy wastes.

Key words: biomass retention, dairy manure, gravity settling, selection pressure, biofilm

47

3.2. Introduction

Anaerobic digestion converts organic matter in dairy manure to biogas that can be used as

renewable energy to reduce greenhouse and odor gas emission (US EPA, 2007). For a given type

of feedstock, the rate of anaerobic digestion increases with the amount of active biomass retained

in the digester as well as the temperature (Batstone et al., 2001; Lettinga et al., 2001). Active

biomass retention provides a cost-effective method for uncompromised anaerobic digestion rate

at lower temperature with less heating energy consumption (Connaughton et al., 2006; Lettinga

et al., 2001). This is especially critical for the anaerobic digestion of flushing diary manure in

large-scale dairies where flushing system is employed as a major means of manure collection for

the sake of its efficiency and less labor requirements (Burke, 2001). As pointed by Wilkie et al.

(2004), flushing dairy manure poses challenges to conventional high temperature anaerobic

digesters because of the large volume of flushing water that must be heated. To address this issue,

fixed-film digesters with submerged high specific surface area media were introduced to retain

active biomass through bacterial attachment (Nikolaeva et al., 2009; Umana et al., 2008; Vartak

et al., 1996; Wilkie et al., 2004). Nevertheless, fixed-film was classified as an inappropriate

process for dairy manure anaerobic digestion due to its vulnerability to media clogging (Burke,

2001). Apparently, a process with no need of exotic media addition would be preferred for

anaerobic digestion of flushing dairy manure.

Gravity settling is a conventional method for biomass retention with no requirement for addition

of supporting materials. It takes advantage of a prolonged settling time to separate bacterial cells

from the supernatant. Based on Eq. (3.1), a gravity settling theory (GS) was established

48

𝑉𝑐 =𝐿𝑑𝑡𝑠

(3.3)

in which Vc represents a critical settling velocity created by the depth of discharge zone (Ld) and

the settling time (ts) (Vesilind, 2003). According to this equation, all particles with settling

velocity (Vp) greater than Vc will settle beneath the discharge zone together with their surface

attached cells and be retained (Vesilind, 2003). Due to the minor settling velocity of planktonic

bacterial mass, a rather long ts of 1 to 3 hours is typically needed to retain bacterial cells at a

given Ld (Lee, 2000; Wilderer et al., 2001). Influent solids, e.g. undigested dairy manure fibers,

may have settling velocities greater than those of planktonic bacterial cells, and thus application

of GS in an anaerobic digester may retain not only active but also inactive biomass (Lott et al.,

1994). Inactive biomass retention is undesirable, as it takes up reactor volume. Selection pressure

(SP) driven cells immobilization is another theory developed in recent years for active biomass

retention with no need of external media (Liu et al., 2005). Selection pressure theory is based on

an equation similar to Eq.(3.1), but it requires an extremely short ts, typically less than 5 min, to

create a large Vc to drive cellular self-immobilization for retention (Qin et al., 2004). According

to SP theory, microorganisms are able to actively respond to a short ts to avoid being washed out

of a reactor (Qin et al., 2004). In support of this theory, successful bacterial retention has been

achieved in solids containing dairy wastewater (Schwarzenbeck et al., 2005). SP theory takes

advantage of bacterial initiative behavior, and thus has the potential to selectively favor the

retention of active biomass over inactive biomass. Such a possibility has yet been studied in

solids wastes anaerobic digestion. This study is designed to test and compare GS and SP theories

in terms of active biomass retention in anaerobic digestion of flushing dairy manure by using

anaerobic sequencing batch reactor (ASBR) as a platform. For this purpose, a wide spectrum of ts

49

ranging from 0.5 to 60 min was employed to test their effect on active biomass retention. The

mechanism behind the opposing roles of GS and SP in settling time was also investigated.

3.3. Materials and methods

3.3.1. Experimental setup and operation

Flushing dairy manure and anaerobic digester sludge were inoculated at a 1:1 volume ratio in

eight identical serum bottles each with 1 L working volume. The source of flushing dairy manure

was from the Washington State University Dairy Center in Pullman, WA, USA. Manure was

flushing hydraulically from the concrete feeding lots to alley boxes, giving rise to 9.58 g L-1 total

suspended solids (TS) mixed liquor. Anaerobic sludge was sampled from a 35oC anaerobic

digester operating in the City of Pullman wastewater treatment plant.

These serum bottles were sparged with nitrogen gas for 5 min before incubation at 37ºC on a 150

rpm shaker. The operation was in a sequential batch mode consisting of 5 steps, i.e., (i) filling, (ii)

reaction, (iii) discharging mixed liquor, (iv) settling and (v) discharging supernatant. The cycle

time was set at 3 days, at the end of which 3 cm of the total 10 cm working height of mixed

liquor was removed in a homogenous state (Figure 3.1), followed by 0.5, 1, 2, 5, 10, 20, 30 or 60

min of settling in respective bottles, and then a 2 cm height of supernatant was dispensed out of

each bottle before refilling with flushing dairy manure (Figure 3.1) and starting the next cycle.

This operation resulted in a 50% volume exchange ratio, 6 days hydraulic retention time and 10

days designed solids retention time (SRT).

50

3.3.2. Microscopic visualization of microbial distribution

Confocal laser scanning microscopy (CLSM; Zeiss LSM 510, Germany) was employed to

visualize the distribution of microorganisms in samples collected from aforementioned ASBRs.

SYTO 9 (Invitrogen, USA), which is a cell-permeative nucleic acid dye, was dripped onto

samples, allowed to stand for 15 min to label the microorganisms and then rinsed twice with

deionized water. SYTO 9 fluorescence was detected with CLSM at excitation/emission of

480/500 nm. Scanning electron microscopy (SEM) (Hitachi S-570, Japan) was employed to

study the morphology of fiber surface attached cells. Sample was prepared with lyophilizer and

coated with gold sputter coater. Fixation was with 2% glutaraldehyde and 2% paraformaldehyde

in 0.1 M phosphate buffer followed by 1% osmium tetroxide in 25 mM phosphate buffer for

overnight.

Figure 3.1 Schematic of ASBR operation in a serum bottle

10 c

m

3 cm

2

cm

Mixed liquor

Supernatant

51

3.3.3. ATP determination

Since the quantity of adenosine 5’-triphosphate (ATP) per cell remains fairly constant, and ATP

is stable only in living cells with functioning catabolic and anabolic processes, ATP

concentration has been used as a standard method for quantifying living microorganisms (ASTM,

2002). In this study, the method of microbial ATP extraction was adopted from that of Lundin

and Thore (1975). Briefly, 1 volume of sample was mixed with 9 volumes of boiling Tris/EDTA

buffer for 90 seconds. Tris/EDTA buffer was made of 0.1 M Tris and 3 mM EDTA titrated to pH

7.8 with acetic acid. After cooling, the extract was assayed using an ATP determination kit (A-

22066, Molecular Probes, Eugene, OR). This assay is based on luciferase’s reaction with the

available ATP, producing light with maximum 560 nm emission at pH 7.8 (ASTM, 2002). The

SpectraMax L microplate luminometer (MDS Analytical Technologies, USA) was used to

quantify luminescence. Triplicate analysis was carried out to ensure results reproducibility.

3.3.4. Methanogens activity measurement

A respirometer (AER-200, CES Inc., USA) was employed to determine the methanogenic

activity in the manure samples. 3000 mg COD L-1 equivalent glucose was added as substrate to

200 mL of fresh sample and incubated at 35 ºC and 300 rpm in serum bottle submerged in a

thermal water bath. Potassium hydroxide was used as a scrubbing medium to purify produced

methane gas by adsorbing carbon dioxide and hydrogen sulfide, and its effectiveness was

verified using gas chromatography (GC, CP-3800, Varian, Inc., USA). Homogenous mixing was

provided with magnetic stirring bars. Methane gas production was automatically measured and

logged in 5 min intervals using a monitoring computer. Control tests with no glucose addition

were carried out in parallel to offset any possible background noise. All measurements were

52

done in triplicates. Methanogenic activity was defined as methane production per sample volume

per day.

3.3.5. Other analytical methods

Daily headspace biogas productions were measured using a glass syringe. TS and chemical

oxygen demand (COD) analyses were done according to standard methods (APHA, 1998). All

samples were measured at the end of each cycle.

3.4. Results and discussion

3.4.1. Process of active biomass retention in ASBRs at various settling time

The rate of volumetric biogas production is associated with the amount of active biomass

retained in a bioreactor. As shown in Figure 3.2, the entire microbial retention process can be

divided into four phases based on the trends in biogas production rates during 90 days (about 30

cycles) operation. The first phase is characterized by a decreasing biogas production rate,

indicating a continuous washout of inoculums. This phase took about 30 days to reach a minimal

value which remained stable for another 10 days in the transitional phase, in which microbial

washout was offset by growth. By about the 40th day, the remaining microorganisms seemed to

have adapted to ASBR operation, entering a rapid growth phase with a steep increase in the

volumetric biogas production rate. In comparison with the washout phase, this growth phase

indicates successful active biomass retention in ASBR serum bottles. After 65 days, stationary

phase was reached in all ASBRs with stabilized biogas production rates; the difference in these

stabilized rates indicate that settling time affects the amount of active biomass retention (Figure

3.2).

53

Figure 3.2 volumetric biogas production profiles in ASBRs operated at various settling time

3.4.2. Effect of settling time on active biomass retention

So far, there is no standard method available for measuring active biomass concentration in

solids wastes, especially when dealing with complicated samples like dairy manure (Grady et al.,

1999). In order to have a better estimation, three parameters were employed in parallel in this

study to assess the actual level of active biomass retention in stationary phase after 90 days

cultivation, i.e., (i) volumetric biogas production rate; (ii) ATP concentration and (iii)

methanogenic activity. Volumetric biogas production rate strongly relies on the amount of

fermentative and methanogenic microorganisms retained in the bioreactor. Its relationship with

settling time in Figure 3.3a shows a “tick” curve with a minimum at 2 min and peaks at both the

Washout Transitional Growth Stationary

Time (days)

0 20 40 60 80

Vol

umet

ric

biog

as p

rodu

ctio

n ra

te (m

L L

-1 d

-1)

50

100

150

200

250

0.5 min 1 min 2 min 5 min10 min20 min30 min60 min

54

Figure 3.3 Effect of settling time on steady state volumetric biogas production rate (a), mixed

liquor ATP concentration (b) and methanogenic activity (c)

Settling time (min)

0.5 1.0 2.0 5.0 10.0 20.0 30.0 60.0

Vol

umet

ric b

ioga

s pro

duct

ion

rate

(mL

L-1 d

-1)

180

195

210

225

240

255a

Settling time (min)

0.5 1 2 5 10 20 30 60

ATP

( m

ol L

-1)

0.15

0.18

0.21

0.24

0.27

0.30

0.33b

Settling time (min)

0.5 1.0 2.0 5.0 10.0 20.0 30.0 60.0

Met

hano

geni

c act

ivity

(L C

H4 L

-1 m

ixed

liqu

or d

-1)

8.0

8.5

9.0

9.5

10.0

10.5 c

55

shortest and longest settling time. This “tick” curve implies the existence of a critical settling

time (ts)c at which ASBR retained the minimum amount of active microorganisms. It is

interesting to see that retention of active biomass increased in both directions away from this (ts)c

to the two extreme ends (Figure 3.3a). Since ATP is mainly found in living cells (Lundin and

Thore, 1975), ATP concentration was used here as an indicator of active biomass. It can yet be

regarded as a better tool for dairy manure full of dead biomass debris. ATP profile in Figure 3.3b

appears to be in line with that of the Figure 3.3a, i.e., a “tick” curve with (ts)c at 2 min settling

time and maximum retention at both poles. Comparing to the ATP content in original flushing

dairy manure which is as low as 0.07 µmol L-1, it is obvious to see the ASBR operation and

settling effects on active biomass retention in Figure 3.3b. Methanogens are the slowest growing

organisms in anaerobic digester. Their retention can be regarded as a good sign of successful

active biomass retention, and thus mixed liquor methanogenic activity was determined in Figure

3.3c. Once again, the profile resembles those in Figure 3.3a and b, exhibiting a “tick” curve with

(ts)c at 2 min settling time (Figure 3.3c). This consistency across three different measures of

active biomass retention indicates that settling time exerts bipolar effects on active biomass

retention, and that either long or short (but not intermediate) settling time may be favorable for

anaerobic digestion in ASBRs.

3.4.3. Active microbial response to settling time

According to the GS theory, a long settling time is favorable for biomass retention because it

allows physical retention of whatever settles beneath the effluent zone during the time. However,

the substantial active biomass retention at very short settling time in Figure 3.3 implies a possible

microbial initiative response to settling time in favor of their own retention. In order to examine

56

differences between biotic and abiotic particle response to settling time, a washout coefficient,

namely wε , is defined as

i

ew C

C=ε (3.1)

in which Ci and Ce represent particle concentrations in ASBR mixed liquor and effluent

supernatant, respectively. Technically, wε represents washout strength imposed by settling time

on settling particles.

Since the concentration of biotic solids can be considered minor compared to abiotic solids in

dairy manure, stationary phase TS is used here to stand for abiotic particles concentration. As for

the biotic washout coefficient, the stationary phase methanogenic activity measured in mixed

liquor and effluent supernatant is employed. The results in Figure 3.4 demonstrate that abiotic wε

monotonously decreases with settling time. In other words, there are always more abiotic

particles washed out of the ASBRs at shorter settling time, which is in line with GS theory

prediction. In contrast, biotic wε demonstrates reversed trends on each side of the 2 min (ts)c.

Moreover, both biotic and abiotic wε values overlap at 2 min (ts)c (Figure 3.4). Considering the

“washout strength” meaning of wε , the (ts)c of 2 min appears to be a threshold value at which

both biotic and abiotic particles were washed out at an equivalent percentage, which explains the

observed minimum active biomass retention at (ts)c in Figure 3.3. Biotic wε on the left-hand side

of (ts)c is the only increasing trend in Figure 3.4. This increase actually implies a lesser washout

of active biomass at shorter ts, which is in opposition to the decrease in abiotic wε in the same

region (Figure 3.4). This suggests that biotic particles, unlike their abiotic counterparts, are able

57

to actively respond to settling time, i.e., more active biomass is actually going to be retained at

stronger abiotic washout strength (Figure 3.4).

Settling time (min)

0.5 1.0 2.0 5.0 10.0 20.0 30.0 60.0

ε w

0.70

0.75

0.80

0.85

0.90

0.95

1.00

1.05

1.10

AbioticBiotic

Figure 3.4 Effect of settling time on washout coefficients for abiotic and biotic particles,

respectively

3.4.4. Living form of retained microorganisms.

The reversed retention trends on both sides of the (ts)c of 2 min, and the maximal retention of

active biomass at 0.5 and 60 min settling time, imply that there might be different living form of

microorganisms in each case (Figure 3.2 to Figure 3.4). CLSM and SEM images indicate that

biofilm attaching on dairy manure fiber is the major living form of retain microorganisms at ts =

0.5 min, while ASBR running at 60 min settling is dominated with planktonic bioflocs (Figure

3.5).

58

Figure 3.5 Confocal visualization of rod and coccoid microbial cells (green), a) attaching on

manure fiber surface in steady state ASBR running at settling time of 0.5 min and b) remaining

suspended in the bulk solution of steady state ASBR running at settling time of 60 min,

respectively. c) SEM image showing the detail of rod and coccoid microbial cells attachment on

fiber surface in steady state ASBR running at settling time of 0.5 min

a b

c

59

3.4.5. Mechanism behind bipolar effects of settling time on active biomass retention

Figure 3.3 to Figure 3.5 clearly demonstrated that settling time is indeed playing bipolar effects

on microorganisms retention in ASBRs, i.e., both extremely long and short settling time were

able to help retain active microbial biomass. The major difference between the two extremes lies

in the living form of those retained microorganisms: long settling time retains bioflocs, and the

short settling time promotes biofilm formation. For decades of conventional ASBR application in

biological wastewater treatment, settling time is usually set at greater than 45 min (Irvine et al.,

1977; Lo et al., 1985; Woolard and Irvine, 1995). Like other gravity settling devices, the design

of ASBR settling time is guided by Eq. (3.1), according to which ASBR discharge height (Ld)

and settling time (ts) create a designed critical settling velocity (Vc) (Figure 3.1). Particles with

settling velocity Vp > Vc are completely retained in ASBR (Vesilind, 2003). Accordingly, a minor

Vc, or in other words a longer ts, should be favorable for particle retention. This holds true for

both biotic and abiotic particles. At least two important points are conveyed from Eq. (3.1): i) ts

actually plays its particle retention role via Vc; and ii) ts should be positively related to particle

retention, so that a maximum retention would be achieved at ts → +∞. This description fits with

data on ts > (ts)c side in Figure 3.3; however, it does not explain the retention downturn at ts <

(ts)c (Figure 3.3).

Extremely short settling time has been be applied in ASBR since 1999 (Beun et al., 1999). It was

found that microorganisms were able to retain themselves in such a strong washout strength, but

they were only retained in the form of biofilm (Beun et al., 1999). Later studies indicated that

short settling time was also playing its retention role via Vc as defined in Eq. (3.1) (Wang et al.,

2006), but in this case, it showed a totally reversed relationship with reference to GS theory, i.e.,

60

short ts turns out to be a positive factor for microbial retention. Thereupon, the SP theory was

established to interpret this microbial active response to washout selection phenomenon, i.e., a

selection pressure will be created by Vc as defined in Eq. (3.1) to wash out slow settling

bioparticles; only those capable of immobilizing themselves in the form of biofilm can become

retained and dominant at short-settling time (Liu et al., 2005). In fact, this mechanism also

explains retention of fermentative microorganisms in animal rumen, where there is a fairly short

hydraulic retention time (McAllister et al., 1994). Thus, the SP theory provides an appropriate

explanation for our observations on ts < (ts)c side in Figure 3.3. It should be emphasized that the

SP theory only applies to active bioparticles capable of self-retention in response to selection

pressure, and not to inactive abiotic particles. Instead, abiotic particles should presumably always

follow GS theory regardless of the range of ts. This integrated interpretation of GS and SP sheds

light on the observations reported in Figure 3.4.

3.4.5.1. Model development

As discussed above, the extent of active biomass retention is governed by two individual theories

on the left- and right- hand sides of (ts)c, respectively. It should be a reasonable assumption that

these two mechanisms always exist and jointly function together in ASBR, and the only

difference is that one may gradually overtake the other as Vc in Eq. (3.) strays from a critical

settling velocity (Vc)c at (ts)c. At Vc = (Vc)c, the two effects just trade off each other giving rise to

a minimum active biomass retention (Figure 3.3). This collective effect can be described as

( ) ( )ccr VSPVGSX += (3.2)

where Xr represents the extent of active biomass retention, and GS(Vc) and SP(Vc) represent the

Vc on Xr effect predicted by GS and SP theories, respectively. According to GS theory (Metcalf

61

and Eddy, 2003), the extent of biomass retention is always inversely related to Vc, and a

maximum retention would be thus achieved when Vc approaches zero, namely infinite ts. This

relationship can be expressed by,

( )

bVaVGS

cc += (3.3)

where a and b are constants. At Vc = 0, a maximum retention, namely a/b, would be

accomplished, and when Vc >> b, the magnitude of retention would approach a/Vc. The unit of “a”

should be the product of the units of the biomass retention and Vc, and “b” should be in Vc unit.

In spite of the novelty of SP theory, all results so far derived from it point to a first-order

dependence of biofilm formation on Vc (Li et al., 2008; Liu et al., 2005; Wang et al., 2006), i.e.,

( ) cc cVVSP = (3.4)

where c is a constant. Its unit should be the quotient of the units of biomass retention over Vc.

Substitution of Eqs. (3.3) and (3.4) into (3.2) gives

c

cr cV

bVaX ++

= (3.5)

Eq. (3.5) describes the integrated effect of Vc on Xr by GS and SP theories, respectively. It has a

minimum value at ( ) bcaV cc −= , which gives ( ) bcacX r −= 2min .

3.4.5.2. Model verification

In order to verify the fitness of Eq. (3.5) for description of the joint effects of GS and SP theories,

the extent of active methanogen retention in Figure 3.4 is expressed by wε−1 , and its

62

relationship with Vc is simulated in Figure 3.6a using Matlab 7.0. The simulated profile exhibits

a “tick” curve in good agreement with experimental data. Since a maximum wε−1 , namely 1,

will be obtained at Vc = 0 m h-1, the “a” value become equal to “b” in this case (Figure 3.6a). Eq.

(3.5) is also applied to the TS retention data, which represents abiotic particles. This simulation

in Figure 3.6b shows a best fit at c = 0, i.e., no SP effect, in agreement with the conjecture that

SP theory does not apply to abiotic particles at all. To further confirm the fitness of Eq. (3.5), it

is also applied to simulate the active biomass retention in terms of other parameters such as

volumetric biogas production rate, ATP concentration and methanogenic activity. The good

agreement between simulation profiles and experimental results in Figure 3.6c, d and e further

confirms the fitness of the Eq. (3.5) model developed above in description of the Vc dependent

active biomass retention. And moreover, it verifies for the first time the speculation that bipolar

effects of settling time on active biomass retention observed in this study are in fact the

integrated results of GS and SP theories in one system.

3.4.5.3. Model limitation

Eq. (3.5) has its limitations, especially at high Vc values. Eq. (3.5) predicts infinite active

biomass retention at Vc → +∞, which is certainly untrue. This false prediction can be ascribed to

the immature development of SP theory at the moment. Little is currently known about what will

happen to active biomass retention at ts = 0, or Vc → +∞ in ASBR. It is expected that the

proposed active biomass retention model will be further improved with the development of the

SP theory in the future.

63

Vc (m/h)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

1- w

-0.1

0.0

0.1

0.2

0.3a = 0.01b = 0.01c = 0.04R2 = 0.92

a

Vc (m/h)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

1- w

-0.02

0.00

0.02

0.04

0.06

0.08

0.10a = 0.002b = 0.002c = 0.000R2 = 0.96

b

64

Vc (m/h)

0.0 0.5 1.0 1.5 2.0 2.5

Vol

umet

ric b

ioga

s pro

duct

ion

rate

(mL/

L/d)

160

180

200

220

240

260a = 273.31b = 1.13c = 63.39R2 = 0.94

c

Vc (m/h)

0.0 0.5 1.0 1.5 2.0 2.5

ATP

( m

ol/L

)

0.15

0.20

0.25

0.30 a = 0.08 b = 0.28c = 0.10R2 = 0.88

d

65

Figure 3.6 Eq. (3.5) simulated active biomass retention (a), inactive biomass retention (b),

volumetric biogas production rate (c), ATP concentration (d) and methanogentic activity (e), (●)

experimental results, (—) simulation profile

3.5. Conclusions

This study revealed bipolar effects of settling time on active biomass retention in ASBR; both

short and long settling time were able to retain high concentrations of active microbes, though in

disparate living forms. Settling time affected active biomass retention through the combined and

opposing effects of gravity settling and selection pressure theories. This observation is supported

by a model integrating the effects of gravity settling and selection pressure which shows good

agreement with experimental results. The effects of gravity settling gradually overtook those of

Vc (m/h)

0.0 0.5 1.0 1.5 2.0 2.5

Met

hano

geni

c ac

itivi

ty (L

CH

4/L/

d)

8

9

10

11

a = 14.52 b = 1.43 c = 2.56R2 = 0.98

e

66

selection pressure as settling time exceeded a critical point; below that critical point, selection

pressure dominated.

67

3.6. References

APHA, 1998. Standard methods for the examination of water and wastewater. 20th ed. American

Public Health Association, Washington DC, USA.

ASTM, 2002. Standard test method for adenosine triphosphate (ATP) content of microorganisms

in water. in: American Society for Testing and Materials (Ed.), ASTM D4012 -81.

Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., Pavlostathis, S.G., Rozzi, A., Sanders,

W.T.M., Siegrist, H., Vavilin, V.A., 2001. The IWA Anaerobic Digestion Model No. 1

(ADM1) 9th World Congress on Anaerobic Digestion, Antwerp, Belgium, pp. 65-73.

Beun, J.J., Hendriks, A., Van Loosdrecht, M.C.M., Morgenroth, E., Wilderer, P.A., Heijnen, J.J.,

1999. Aerobic granulation in a sequencing batch reactor. Water Res. 33, 2283-2290.

Burke, D.A., 2001. Dairy waste anaerobic digestion handbook. Environmental Energy Company,

6007 Hill Street, Olympia, WA 98516.

Connaughton, S., Collins, G., O'Flaherty, V., 2006. Psychrophilic and mesophilic anaerobic

digestion of brewery effluent: A comparative study. Water Res. 40, 2503-2510.

Grady, C.P.L., Daigger, G.T., Lim, H.C., NetLibrary Inc., 1999. Biological wastewater treatment.

2nd edn ed. Marcel Dekker, New York.

Irvine, R.L., Fox, T.P., Richter, R.O., 1977. Investigation of fill and batch periods of sequencing

batch biological reactors. Water Res. 11, 713-717.

Lee, C.C., 2000. Handbook of environmental engineering calculations. McGraw-Hill, New York.

Lettinga, G., Rebac, S., Zeeman, G., 2001. Challenge of psychrophilic anaerobic wastewater

treatment. Trends Biotechnol. 19, 363-370.

Li, Y., Liu, Y., Xu, H.L., 2008. Is sludge retention time a decisive factor for aerobic granulation

in SBR? Bioresour. Technol. 99, 7672-7677.

68

Liu, Y., Wang, Z.W., Qin, L., Liu, Y.Q., Tay, J.H., 2005. Selection pressure-driven aerobic

granulation in a sequencing batch reactor. Appl. Microbiol. Biotechnol. 67, 26-32.

Lo, K.V., Bulley, N.R., Kwong, E., 1985. Sequencing aerobic batch reactor treatment of milking

parlor wastewater. Agr. Wastes. 13, 131-136.

Lott, S.C., Loch, R.J., Watts, P.J., 1994. Settling characteristics of feedlot cattle feces and

manure. Trans. ASAE. 37, 281-285.

Lundin, A., Thore, A., 1975. Comparison of Methods for Extraction of Bacterial Adenine

Nucleotides Determined by Firefly Assay. Appl. Environ. Microbiol. 30, 713-721.

McAllister, T.A., Bae, H.D., Jones, G.A., Cheng, K.J., 1994. Microbial attachment and feed

digestion in the rumen. J. Anim. Sci. 72, 3004-3018.

Metcalf and Eddy, 2003. Wastewater engineering: treatment and reuse. 4th edn ed. McGraw-Hill,

Boston.

Nikolaeva, S., Sanchez, E., Borja, R., Raposo, F., Colmenarejo, M.F., Montalvo, S., Jimenez-

Rodriguez, A.M., 2009. Kinetics of anaerobic degradation of screened dairy manure by

upflow fixed bed digesters: Effect of natural zeolite addition. J. Environ. Sci. Health, Part

A: Toxic/Hazard. Subst. Environ. Eng. 44, 146-154.

Qin, L., Liu, Y., Tay, J.H., 2004. Effect of settling time on aerobic granulation in sequencing

batch reactor. Biochem. Eng. J. 21, 47-52.

Qin, L., Tay, J.H., Liu, Y., 2004. Selection pressure is a driving force of aerobic granulation in

sequencing batch reactors. Process Biochem. 39, 579-584.

Schwarzenbeck, N., Borges, J.M., Wilderer, P.A., 2005. Treatment of dairy effluents in an

aerobic granular sludge sequencing batch reactor. Appl. Microbiol. Biotechnol. 66, 711-

718.

69

Umana, O., Nikolaeva, S., Sanchez, E., Borja, R., Raposo, F., 2008. Treatment of screened dairy

manure by upflow anaerobic fixed bed reactors packed with waste tyre rubber and a

combination of waste tyre rubber and zeolite: Effect of the hydraulic retention time.

Bioresour. Technol. 99, 7412-7417.

US EPA, 2007. Guide to Anaerobic Digesters. Anaerobic Digester Database.

Vartak, D.R., Engler, C.R., McFarland, M.J., Ricke, S.C., 1996. Attached-film media

performance in psychrophilic anaerobic treatment of dairy cattle wastewater 7th National

Bioenergy Conference, Nashville, Tennessee, pp. 79-84.

Vesilind, P.A., 2003. Wastewater treatment plant design. IWA Pub, London.

Wang, Z.W., Liu, Y., Tay, J.H., 2006. The role of SBR mixed liquor volume exchange ratio in

aerobic granulation. Chemosphere 62, 767-771.

Wilderer, P.A., Irvine, R.L., Goronszy, M.C., 2001. Sequencing Batch Reactor Technology.

IWA, London, UK.

Wilkie, A.C., Castro, H.F., Cubinski, K.R., Owens, J.M., Yan, S.C., 2004. Fixed-film anaerobic

digestion of flushing dairy manure after primary treatment: Wastewater production and

characterisation. Biosystems Eng. 89, 457-471.

Woolard, C.R., Irvine, R.L., 1995. Treatment of hypersaline waste-water in the sequencing batch

reactor. Water Res. 29, 1159-1168.

70

CHAPTER FOUR

KINETICS OF PSYCHROPHILIC ANAEROBIC DIGESTER WITH BIOFILM

SUPPORTED BY SOLIDS FROM FLUSHING DAIRY MANURE

4.1.Abstract

Microbial biomass retention is critical for ensuring the performance of an anaerobic digester. In

applications such as treatment of animal waste, however, the current biomass retention

technologies using artificial biofilm support media are vulnerable to clogging caused by manure

fiber. In this study, a new strategy, improved biomass retention with fiber material present within

the dairy manure as biofilm carriers, was evaluated for treating flushing dairy manure in a

psychrophilic anaerobic sequencing batch reactor (ASBR). A kinetic study was carried out for

process control and design by comparing four microbial growth kinetic models, i.e. first-order,

Grau, Monod and Chen & Hashimoto models. A volumetric methane production rate of 0.24

L/L/d of and specific methane productivity of 0.19 L/gVSloaded were achieved at 6 days HRT. It

was proved that ASBR using manure fiber as support media not only improved methane

production but also reduced the necessary HRT and temperature to achieve a similar treating

efficiency compared with current technologies. The kinetic model can be used for design and

optimization of the process.

Keywords: psychrophilic, kinetics, ASBR, anaerobic digestion, dairy manure

71

4.2. Introduction

Livestock farms in US produce a total of about 2 billion tons of manure each year (Gillespie &

Flanders, 2010), which accounts for 8% of the total US anthropogenic bio-methane emissions

(USEPA, 2010). Anaerobic digestion (AD) is an alternative to livestock waste management that

offers economic and environmental benefits. Besides alleviating manure-associated greenhouse

gas (GHG) emissions and farm-generated odors, AD of animal waste provides fertilizers rich in

nutrient, and biogas as renewable energy.

Wider adoption of AD for animal manure management has been limited primarily by economics.

This is especially true in some applications where the wastewater is relatively dilute such as in

flushing dairies. Flushing manure handling systems are widely employed within large-scale dairy

farms due to their reduced labor and mechanical failures (Powers et al., 1997). However,

flushing systems produce a waste stream with total solids of 1–2 %, negatively impacting

conventional AD treatment processes due to the fact that diluted manure increases digester size

and heating requirements. Anaerobic digestion at psychrophilic temperature can alleviate this

concern, if corresponding losses in biogas production rates due to the lower utilized temperature

can be overcome through high microbial accumulation (Kashyap et al., 2003). By inference,

assuming adequate psychrophilic operation, the main concerns with using an anaerobic digester

for dilute manure treatment is the challenges in achieving higher solids retention time (SRT)

required to retain microbial biomass and reducing required size. Typical designs such as

continuous stirred-tank reactor (CSTR) or plug flow (PF) digesters cannot accomplish such

decoupling of SRT and HRT (hydraulic retention time) (Zaher et al., 2008).

72

Many efforts have been made to increase microbial biomass retention with different digester

configurations, such as fixed-bed and hybrid reactors (Borja et al., 1994; Demirer & Chen, 2005;

Umana et al., 2008; Wilkie et al., 2004; Zaher et al., 2008). A variety of external artificial

biofilm support media, such as spherical plastic trickling filter media, floating support media,

automobile tires, and zeolite have been employed in anaerobic biofilm digesters to enhance

biomass retention. The addition of artificial support media occupies substantial digester volume,

which automatically lowers the digester efficiency. Moreover, the artificial biofilm support

media are vulnerable to clogging caused by manure fiber, which impedes commercialization.

A concept of biofilm retention with influent solids was presented in the authors’ previous studies

(Frear et al., 2010; Wang et al., 2011). It was reported that anaerobic microorganisms have a

strong affinity to manure fiber and they can serve as natural biomass support media in a high rate

digester. Biomass retention using manure fiber as natural support media seems a promising

approach for anaerobic treatment of flushing dairy manure. In virtue of no artificial biofilm

support media, the concern regarding mechanical failure caused by media clogging is removed.

Along with low maintenance, the required digester size and cost are reduced. Anaerobic

sequencing batch reactors (ASBR) are known to be capable of uncoupling HRT with SRT for

biomass retention with particular sequence of operation of ASBR exerting selection pressures to

microbes for immobilization (Liu et al., 2005; Wang et al., 2011). Wang et al. (2011) showed

that an ASBR digester, which retained high concentration of biomass in the form of fiber-

attached biofilm by selection pressure, exhibited comparatively high methane yield and

production rate. However, applications of this technology require technical information for

process design and optimization.

73

Although simple, ASBR operation involves complex processes whose design and optimization

can be facilitated using mathematical models. Kinetic modeling, being a useful tool in process

analysis, design, and system control can be established by precise determination of kinetic

coefficients. Process kinetics also details the effects of operational and environmental factors on

substrate utilization rates. A variety of kinetic models have been developed to describe microbial

growth kinetics. First order model is the simplest model for microbial growth with the

assumption of first-order degradation, which has been used often to describe hydrolysis limited

digestion with respect to particulate substrate (Gavala et al., 2003). Monod model is the most

widely used kinetic model which was developed as a result of empirical analysis (Monod, 1949).

Grau et al. (1975) and Chen and Hashimoto (1978) improved the Monod model by predicting

that effluent substrate concentration is proportional to influent substrate concentration. However,

it was assumed that microbial growth kinetics in anaerobic biofilm reactors followed Monod or

first-order models in most literature (Batstone et al., 2002; Buffière et al., 1998; Huang & Jih,

1997; Rittmann & McCarty, 2001). The lack of appropriate models in the literature though

shows that improvements can still be made. For example, it has been hypothesized that the Chen

& Hashimoto model is capable of characterizing biofilm growth kinetics with an improved

performance, compared to the Monod model, due to dependence on influent substrate

concentration.

The main objective of this study was to evaluate the performance and kinetics of the new

biomass retention strategy during psychrophilic ASBR digestion of flushing dairy manure. A

kinetic model with aim to find a more appropriate biofilm growth model was derived and

assessed for substrate utilization and methane production. Both HRT and OLR (organic loading

74

rate) are considered as the most important parameters for digester operation. Hence, the effects

of HRT and OLR on methane production were the primary output of the kinetic model.

4.3. Materials and methods

The aforementioned biomass retention technology for treating flushing dairy manure at

psychrophilic temperature were evaluated in five lab scale digesters operated in sequencing batch

mode. Kinetic properties of psychrophilic ASBR digesters were then analyzed and a kinetic

model was derived for system optimization.

4.3.1. Feedstock and seed

Fresh dairy manure was collected from the Washington State University Knott Dairy Center in

Pullman, WA, USA and stored at 4 °C prior to use. Before addition to digesters, manure was

diluted with tap water to mimic flushing dairy manure, which resulted in mixed liquor containing

9.1 g/L total solids (TS) and 7.6 g/L total volatile solids (VS). Anaerobic sludge was sampled

from an anaerobic digester in Pullman Wastewater Treatment Facility with TS of 17.1 g/L and

VS of 11.7 g/L.

4.3.2. Experimental setup and operation

Five digesters (64 cm in height and 10 cm in diameter) with working volume of 6 L were

operated as ASBR at respective cycle times of 2, 4, 6, 8, and 10 days while the other operation

conditions remained constant (50% exchange ratio, 5 min settling time and 20 days SRT). This

operation produced HRT of 4, 8, 12, 16, and 20 days in each digester (Table 4.1). Settling time

was determined based on the authors’ previous study to form biofilm which attached on the

75

surface of manure fiber (Wang et al., 2011). Each digester was mixed with a separate impeller

driven by a respective motor at 100 rpm. Intermittent mixing was carried out for 10 min every 2h.

Manure and anaerobic sludge were introduced to each digester at 1:1 volume ratio when

experiments started. Digesters were then placed in a low temperature chamber (22 °C) and

operated in sequencing batch mode which consists of 5 stages: filling, reaction, desludge, settling

and discharging in one cycle. At the end of the reaction stage, mixed liquor was discharged under

complete mixed state to control SRT then supernatant was discharged after settling stage to

control HRT in each digester. Evaluation of system performance for each HRT was carried out

during pseudo steady state conditions, when biogas production, methane content and effluent

COD variations were less than 10% (Karim et al., 2005).

4.3.3. Chemical analytical methods

TS, VS and COD analyses were done according to Standard Methods (APHA, 1998). The

volume of biogas from the digester was determined by water displacement method. Content of

CH4 and CO2 were determined using a Varian gas chromatograph (Palo Alto, CA, USA)

equipped with a thermal conductivity detector. A Restek (Bellefonte, PA, USA) shincarbon

column (2 × 1/16 inch) with silcosteel packing material (100/120 mesh) was used, and nitrogen

served as the carrier gas. Each type of gas was quantified based on a calibration curve. All

samples were measured at the end of each cycle.

76

Tabl

e 4.

1 O

pera

ting

cond

ition

for A

SBR

Rea

ctor

H

RT

(d)

OLR

(g V

S loa

ded/L

/d)

Exch

ange

ratio

(%)

Tota

l

cycl

e

time

(d)

C

ycle

tim

e

Feed

ing

(min

)

Rea

ctio

n

(d)

Des

ludg

e

(min

)

Settl

ing

(min

)

Dec

antin

g

(min

)

R1

4

2.0

50

2

5 2

5

5 5

R2

8

1.0

50

4

5 4

5

5 5

R3

12

0.7

50

6

5 6

5

5 5

R4

16

0.5

50

8

5 8

5

5 5

R5

20

0.4

50

10

5 10

5

5 5

77

4.3.4. Development of kinetic model

Kinetic models predicting methane production were derived assuming digesters operated at

steady state conditions. Four model types were considered in this study in order to choose a

model with appropriate fit for microbial growth kinetics during anaerobic digestion of flushing

dairy manure in ASBR digester.

First order: 𝜇 =𝑘𝑘

𝑘0 − 𝑘− 𝑏 (4.1)

Grau: 𝜇 =𝜇𝑚𝑘𝑘0

− 𝑏 (4.2)

Monod: 𝜇 =𝜇𝑚𝑘𝐾𝑠 + 𝑘

− 𝑏 (4.3)

Chen & Hashimoto: 𝜇 =𝜇𝑚𝑘

𝐾𝑘0 + (1 − 𝐾)𝑘− 𝑏 (4.4)

Microbial growth mass balance was considered as the basis for model derivation, and rate of

microbial growth can be written as:

𝑟𝑚 =𝑑𝑋𝑑𝑡

= 𝜇𝑋 (4.5)

Thus, mass balance equation for microbial growth can be given as follows:

𝑉𝑑𝑋𝑑𝑡

= 𝑄𝑋0 − 𝑄𝑋 + 𝑟𝑚𝑉 (4.6)

78

At steady state (𝑑𝑋𝑑𝑡

= 0), the above equation can be simplified as:

𝜇 =1𝑏

(4.7)

By use of each of the above models, steady state effluent substrate concentrations were derived

and are listed in Table 4.2. Taking Chen & Hashimoto model as an example, by substituting Eq.

(4.4) in Eq. (4.7), effluent substrate concentration can be expressed as

𝑘 =𝐾𝑘0(1 + 𝑏𝑏)

(𝐾 − 1)(1 + 𝑏𝑏) + 𝜇𝑚𝑏 (4.8)

If B denotes the volume of methane produced at standard condition per gram of substrate loaded

to the digester, and B0 is the volume of methane produced at standard condition per gram of

substrate loaded at infinite retention time, then the biodegradable substrate in the digester will be

directly proportional to B0 – B, and B0 will be directly proportional to the biodegradable substrate

loading (Chen & Hashimoto, 1978). Therefore, the following relationship can be derived:

𝑘𝑘0

=𝐵0 − 𝐵𝐵0

(4.9)

The above equation can be rearranged to give:

𝐵 = 𝐵0 �𝑘0 − 𝑘𝑘0

� (4.10)

The volumetric methane production rate equals B times OLR:

𝑀 =𝐵𝑘0𝑏

(4.11)

79

By use of Eq. (4.10), Eq. (4.11) can be expressed as:

𝑀 =𝐵0𝑏

(𝑘0 − 𝑘) (4.12)

The volumetric methane production rate with respect to HRT can be derived by substituting

steady state substrate concentration of effluent into Eq. (4.12). Using Chen & Hashimoto model

as an example, Eq. (4.12) can be written as follows:

𝑀 =𝐵0𝑏�𝑘0 −

𝐾𝑘0(1 + 𝑏𝑏)(𝐾 − 1)(1 + 𝑏𝑏) + 𝜇𝑚𝑏

� (4.13)

The volumetric methane production rate with respect to OLR can be expressed as:

𝑀 = 𝐵0 �𝐿 −𝐾𝐿(𝐿 + 𝑏𝑘0)

(𝐾 − 1)(𝐿 + 𝑏𝑘0) + 𝜇𝑚𝑘0� (4.14)

The specific methane productivity with respect to HRT can be expressed as:

𝑃 = 𝐵0 �1 −𝐾(1 + 𝑏𝑏)

(𝐾 − 1)(1 + 𝑏𝑏) + 𝜇𝑚𝑏� (4.15)

The specific methane productivity with respect to OLR can be expressed as:

𝑃 = 𝐵0 �1 −𝐾(𝐿 + 𝑏𝑘0)

(𝐾 − 1)(𝐿 + 𝑏𝑘0) + 𝜇𝑚𝑘0� (4.16)

The VS removal efficiency with respect to HRT can be written as:

𝐸 = 1 −𝐾(1 + 𝑏𝑏)

(𝐾 − 1)(1 + 𝑏𝑏) + 𝜇𝑚𝑏 (4.17)

80

The volumetric methane production rate, specific methane productivity and removal efficiency

with respect to the other three models were also derived for model comparison and listed in

Table 4.2.

In order to determine B0, mass balance equation of substrate consumption can be written as:

𝑉𝑑𝑘𝑑𝑡

= 𝑄𝑘0 − 𝑄𝑘 + 𝑟𝑠𝑉 (4.18)

Rate of substrate consumption

𝑟𝑠 =𝑑𝑋𝑑𝑡

= −𝑘′𝑘 (4.19)

At steady state (𝑑𝑆𝑑𝑡

= 0), effluent substrate concentration can be expressed as:

𝑘 =𝑘0

(𝑏𝑘′ + 1) (4.20)

Substituting Eq. (4.20) into Eq. (4.12) yields:

𝑘0𝑀

=𝑏𝐵0

+1

𝐵0𝑘′ (4.21)

81

Ta

ble

4.2

Kin

etic

s mod

els u

sed

in th

is st

udy

Kin

etic

coef

ficie

nts

Firs

t-ord

er

Gra

u M

onod

C

hen

& H

ashi

mot

o

Spec

ific

grow

th ra

te

𝜇=

𝑘𝑘𝑘 0−𝑘−𝑏

𝜇=𝜇 𝑚

𝑘𝑘 0

−𝑏

𝜇=

𝜇 𝑚𝑘

𝐾 𝑠+𝑘−𝑏

𝜇=

𝜇 𝑚𝑘

𝐾𝑘 0

+(1−𝐾

)𝑘−𝑏

Efflu

ent

subs

trate

conc

entra

tion

𝑘=

𝑘 0(1

+𝑏𝑏

)𝑏(𝑘

+𝑏)

+1

𝑘=𝑘 0

(1+𝑏𝑏

)𝜇 𝑚

𝑏

𝑘=

𝐾 𝑠(1

+𝑏𝑏

)𝑏(𝜇 𝑚

−𝑏)−

1 𝑘

=𝐾𝑘 0

(1+𝑏𝑏

)(𝐾

−1)

( 1+𝑏𝑏

) +𝜇 𝑚

𝑏

Vol

umet

ric

met

hane

prod

uctio

n ra

te

𝑀=𝐵 0 𝑏

(𝑘0−

𝑘 0(1

+𝑏𝑏

)𝑏(𝑘

+𝑏)

+1)

𝑀=𝐵 0 𝑏

(𝑘0−𝑘 0

( 1+𝑏𝑏

)𝜇 𝑚

𝑏)

𝑀=𝐵 0 𝑏

(𝑘0−

𝐾 𝑠( 1

+𝑏𝑏

)𝑏(𝜇 𝑚

−𝑏)−

1) 𝑀

=𝐵 0 𝑏

(𝑘0−

𝐾𝑘 0

( 1+𝑏𝑏

)( 𝐾

−1)

( 1+𝑏𝑏

) +𝜇 𝑚

𝑏)

Spec

ific

met

hane

prod

uctiv

ity

𝑃=𝐵 0

(1−

1+𝑏𝑏

𝑏(𝑘

+𝑏)

+1)

𝑃=𝐵 0

(1−

1+𝑏𝑏

𝜇 𝑚𝑏

) 𝑃

=𝐵 0

(1−

𝐾 𝑠( 1

+𝑏𝑏

)𝑘 0𝑏(𝜇 𝑚

−𝑏)−𝑘 0

) 𝑃

=𝐵 0�1

−𝐾

(1+𝑏𝑏

)( 𝐾

−1)

(1+𝑏𝑏

)+𝜇 𝑚

𝑏�

Rem

oval

effic

ienc

y 𝐸

=1−

1+𝑏𝑏

𝑏(𝑘

+𝑏)

+1

𝐸=

1−

1+𝑏𝑏

𝜇 𝑚𝑏

𝐸

=1−

𝐾 𝑠( 1

+𝑏𝑏

)𝑘 0𝑏(𝜇 𝑚

−𝑏)−𝑘 0

𝐸

=1−

𝐾(1

+𝑏𝑏

)(𝐾

−1)

( 1+𝑏𝑏

) +𝜇 𝑚

𝑏

82

B0 of manure depends on the species, ration, age of the manure, collection, and storage and

bedding material. Values of B0 for dairy manure range from 0.21 to 0.27 (Chen & Hashimoto,

1978; Husain, 1998). B0 of flushing dairy manure used in this present study was determined by

plotting HRT versus S0/M according to Eq. (4.21). The slope of the curve was used to calculate

B0 as 0.24 L CH4/gVSloaded.

An endogenous decay constant b of 0.03 /d was used for the model simulations (Husain, 1998).

Values of kinetic parameters for each model were estimated using the Curve Fitting function in

SigmaPlot 11 (Systat Software, Inc.) using the Marquardt-Levenberg algorithm with 200

iterations.

4.4. Results and discussion

4.4.1. Effect of HRT on biogas production

Biogas generation in the ASBR digesters with different HRT is represented in Figure 4.1. It can

be seen that the time needed to reach steady state condition is associated with HRT. The longer

the HRT, the more time required for start-up. After operating periods of 34, 62, 72, and 80 days,

digesters R1, R2, R3, R4 and R5 corresponding to HRT of 4, 8, 12, 16, 20 days reached steady

state condition, respectively.

83

Figure 4.1 Volumetric biogas production rate at various HRT

The volumetric biogas productivity during manure bio-methanization was also related to HRT in

each digester. The highest volumetric biogas production was observed at the shortest HRT of 4

days (digester R1) with rate of 0.37 L/L/d, while digester R5, running at the longest HRT of 20

days, showed the lowest rate of 0.14 L/L/d. Zaher et al. (2008) reported similar results

(volumetric methane production rate of 0.20 L/L/d at 5 day HRT and 0.10 L/L/d at 17 day HRT)

from a tire supported fixed-bed digester treating flushing dairy manure but under mesophilic

temperature. A much lower rate using a zeolite supported CSTR digester even at mesophilic

conditions was obtained by Borja et al. (1994). Powers et al. (1997) reported a fixed-bed digester

84

with a low methane production at 2.3 day HRT. The manure fiber supported psychrophilic

ASBR digester showed good performance as compared against these other technologies.

Biofilm was expected to be formed in all digesters due to selection pressure in terms of settling

time (Wang, Ma et al. 2011). However, washout of biomass was present and digester failure

occurred when the HRT was further shortened to 2 days, which is because too short of an HRT at

start-up period exceeded the microorganism growth limits (Rittmann & McCarty, 2001). A

practical limit with a minimum HRT of 4 days was required at the start-up period for biomass

retention within the digester.

Compared with our previous studies, the present study was conducted under psychrophilic

temperature (22 °C) instead of mesophilic condition (35 °C). However, the digester in this study

showed a comparable performance with that of a mesophilic digester (volumetric biogas

production rate of 0.25 L/L/d and specific methane productivity of 0.20 L/gVSloaded), indicating

that biomass retention provides a cost-effective method for uncompromised anaerobic digestion

rate at lower temperature with less energy consumption due to the reduction of heat required

(Connaughton et al., 2006; Lettinga et al., 2001).

85

HRT (d)

5 10 15 20

Vol

umet

ric

met

hane

pro

duct

ion

rate

(L/L

/d)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

A

OLR (gVS/L/d)

0.5 1.0 1.5 2.0 2.5

Vol

umet

ric

met

hane

pro

duct

ion

rate

(L/L

/d)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

B

86

Figure 4.2 Changes of volumetric methane production rate and specific methane productivity

against HRT and OLR, (●) experimental results, and (—) simulation profile

HRT (d)

5 10 15 20

Spec

ific

met

hane

pro

duct

ivity

(L/g

VS

load

ed)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

C

OLR (gVS/L/d)

0.5 1.0 1.5 2.0

Spec

ific

met

hane

pro

duct

ivity

(L/g

VS

load

ed)

0.10

0.12

0.14

0.16

0.18

0.20

0.22

0.24

D

87

4.4.2. Effect of OLR on biogas production

OLR in each digester varied according to different HRT since influent substrate concentration is

fixed. The effects of OLR on volumetric methane production rate and specific methane

productivity were plotted in Figure 4.2, C and D. Specific methane productivity showed an

inverse relationship with respect to OLR, reducing slightly until OLR reached 1 g VSloaded/L/d at

which point further OLR extension led to steep drops in specific methane productivity.

The volumetric methane production rate and specific methane productivity are often two

competing performance parameters. An increasing of OLR favors volumetric methane

production rate but leads to impaired specific methane productivity. It seems a compromise in

OLR should be employed with an OLR of around 1 g VSloaded/L/d yielding high values for both

parameters. The same situation is applied HRT with a HRT around 4–6 days being favored for

both parameters.

4.4.3. Kinetic modeling

4.4.3.1.Evaluation of Kinetic models

First-order, Grau, Monod and Chen & Hashimoto models were chosen to determine the most

appropriate model for the kinetics of methane production from flushing dairy manure in an

ASBR digester. Results are presented in Figure 4.3 and goodness of fitting is listed in Table 4.3.

On the account of poor correlation with data sets, Monod model was not recommended for

kinetic analysis. Furthermore, the Monod and first-order model failed to predict volumetric

methane production rate decreases at extremely short HRT, which led to reduced accuracy in the

simulation. The limitation of the Monod model resides in the effluent substrate concentration (S)

being independent of the influent substrate concentration (S0), with organic loading notably

88

having been found to affect digester performance. Saravanan and Sreekrishnan (2006) pointed

out that the assumption of substrate degradation described within the Monod model is

questionable in biofilm reactors. The Chen & Hashimoto model and Grau model explicitly

account for the influent substrate concentration, so they are able to overcome this constraint and

predict S as a function of S0, with effluent substrate concentration being directly proportional to

the influent substrate concentration. The Chen & Hashimoto model successfully fit the peak

volumetric methane production rate at 4 days HRT and the microorganisms’ washout at 2 days

HRT. The Chen & Hashimoto model included the influence of S0 in the kinetic expression in

order to express mass transfer limitations (Chen & Hashimoto, 1978). Mass transfer limitation

can lead 𝜇 to vary with initial substrate concentration. Therefore, the Chen & Hashimoto model

was selected for development of a derived model to conduct kinetic analysis for an ASBR

digester treating flushing dairy manure.

4.4.3.2.Model simulation

In the derived model, values of 𝜇𝑚 and K were the variables identified to characterize the

digester performance and for fitting. 𝜇𝑚 is the maximum specific growth rate of microorganisms

expressed as per day. K is a dimensionless kinetic parameter indicating digester performance. K

is equal to Yc, where Y is growth yield coefficient and c is the Contois coefficient. From the data

presented in Figure 4.2, it can be seen that the experimental data fit nicely with the derived

model (R2 > 95 %), showing the validity of the model. The 𝜇𝑚 and K values calculated from the

derived model are 0.36 d–1 and 0.23, respectively.

89

Figure 4.3 Comparison of simulation with different kinetics models

The value of 𝜇𝑚 is at the lower side of the wide range (0.041–0.912 d–1) reported for mixed and

pure cultures of methanogens at temperature between 35–37 °C (Pavlostathis & Giraldogomez,

1991). This may be due to the lower temperature and mixed culture used in the kinetic

coefficient determination. K is an indicator of the overall performance of the digester (Chen,

1983). An increasing K indicates inhibition of fermentation while low K indicates rapid substrate

degradation. K value for typical digesters range from 0.6 to 2.0 with a mean of 1.06 (Hashimoto,

1982). The K value obtained from this study is lower compared with other studies, implying only

short HRT is needed for anaerobic digestion of flushing dairy manure in an ASBR digester.

Furthermore, K and 𝜇𝑚 were reported to be independent of the influent substrate concentration

HRT (d)

5 10 15 20

Vol

umet

ric

met

hane

pro

duct

ion

rate

(L/L

/d)

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Experimental dataFirst-order modelGrau modelMonod modelChen & Hashimoto model

90

for diluted organic waste stream (Chen & Hashimoto, 1978). Hence, the lower K value could be

attributed to high amount of biomass retained within ASBR digester.

From simulated profiles, digestion performance, as indicated by volumetric methane production

rate, was highest at near 4 days or at an organic loading rate of 2 g VSloaded/L/d (Figure 4.2, A

and C). Digester performance can also be represented by specific methane productivity with this

indicator pointing to maximums at 5–20 days and OLR of between 0.5–1.3 g VSloaded/L/day

(Figure 4.2, B and D). Comparison of the two performance indicators shows that an ideal HRT

for ASBR feeding with flushing dairy manure would be around 4–6 days while OLR can either

be limited for enhanced specific methane productivity or slightly increased for improved

volumetric performance.

Table 4.3 Summary of model comparison with kinetic coefficients and goodness of fit

Models Kinetic coefficients R2 SRSE*

First-order 𝑘= 0.43 0.92 0.073

Grau 𝜇𝑚= 0.67 day-1 0.96 0.032

Monod 𝜇𝑚= 0.07 day-1

𝐾𝑠= 0.24 g VS 0.76 0.102

Chen & Hashimoto 𝜇𝑚= 0.36 day-1

𝐾= 0.23 0.99 0.011

*The sum of residual squared error

91

4.4.3.3.Model prediction

The derived model was then used to predict digester performance, namely effluent VS

concentration and VS removal efficiency. As Figure 4 shows, the model successfully predicted

effluent VS concentration with R2 of 0.91. However, the calculated results of VS removal

efficiency do not fit well with the measurements. This is also because influent substrate

concentration was not included in the VS removal efficiency calculation equation. Although

Chen and Hashimoto (1980) states that the treatment efficiency does not depend on influent

substrate concentration, it was not validated here by experimental data.

Figure 4.4 Changes of effluent substrate concentration and treatment efficiency with derived

model, (●) effluent VS data, () VS removal efficiency data, and (—) predicted profile

HRT (d)

5 10 15 20

Eff

luen

t VS

(gV

S/L

)

5.2

5.4

5.6

5.8

6.0

6.2

VS

rem

oval

eff

icie

ncy

(%)

0

10

20

30

40

50

Effluent VS VS removal efficiency

92

Tabl

e 4.

4 Pe

rfor

man

ce d

ata

for d

iffer

ent a

naer

obic

bio

film

reac

tors

trea

ting

dairy

man

ure

Dig

este

r typ

e In

fluen

t O

LR

Tem

pera

ture

H

RT

Met

hane

co

nten

t Sp

ecifi

c m

etha

ne

prod

uctiv

ity

Lite

ratu

re

(g V

S/L)

(g

VS/

L/d)

(°

C)

(d)

(%)

(L/g

VS l

oade

d)

ASB

R w

ith m

anur

e fib

er a

s su

ppor

t med

ia

7.6

0 1.

26

22

6.0

73

.4

0.19

Th

is st

udy

Fixe

d-fil

m re

acto

r with

sp

heric

al p

last

ic tr

ickl

ing

filte

r med

ia

1.3

0a 4.

07

23-2

4 2

.3

65.0

0.

10

(Pow

ers e

t al.,

199

7)

Ana

erob

ic h

ybrid

reac

tor

with

floa

ting

supp

ort m

edia

9

.87b

7.30

36

15

.0

63.5

0.

19

(Dem

irer &

Che

n,

2005

)

Fixe

d be

d re

acto

r with

au

tom

obile

tire

s 13

.83c

2.77

d 35

5

.0

NA

e

0.19

f (Z

aher

et a

l., 2

008)

CST

R w

ith z

eolit

e su

ppor

t 47

.10c

9.42

d 35

5

.0

NA

0.

12 f

(Bor

ja e

t al.,

199

4)

Fixe

d be

d re

acto

r with

tire

ru

bber

and

zeo

lite

75.0

0 4.

40

22-2

6 5

.5

NA

0.

18 f

(Um

ana

et a

l., 2

008)

a %

TS; b %

VS;

c g C

OD

/L; d g

CO

D/L

/d; e n

ot a

vaila

ble;

f L/g

CO

D lo

aded

93

4.4.4. Process performance comparison

Comparisons of the operation parameters and the methane productivities obtained in this study

with the performance data from other anaerobic biofilm reactors treating dairy manure are

presented in Table 4.4. As can be seen, except for this study, a variety of different types of

external artificial biofilm support media were employed in those anaerobic biofilm digesters to

enhance biomass retention. It is clear from Table 4.4 that ASBR using manure fiber as support

media not only improved methane production but also reduced the necessary HRT and

temperature to achieve a comparable treating efficiency. This digester expanded the capacity of

anaerobic digestion to dilute solid wastes treatment with no requirement of prior solids

separation or the risk of biofilm support media clogging. The specific methane productivity

obtained in this study was higher than most of others as shown in Table 4.4. High performance

was attained at relatively low HRT and low temperature so that this digester may be more cost-

effectiveness than others.

4.4.5. Implications for dairy AD process design

A high rate AD process driven by high biomass retention instead of mesophilic temperature (35

◦C) appears to be an economical approach for methane recovery from flushing dairy manure

(Frear et al., 2010). This study demonstrated a new biomass retention strategy with biofilm

supported by manure fiber. Fibrous solids content in flushing dairy manure act as a natural

biofilm support medium for high biomass retention as opposed to using external media that

might clog and add cost. As attributed to high concentration of biomass, the performance of this

technology at psychrophilic temperature is comparable to that of other technology under

mesophilic conditions. Taking advantage of successful in vitro biomass immobilization on dairy

94

manure fibrous solids, ASBR might be employed as an optimum means to achieve high rate AD

in flushing dairy manure.

This novel biomass immobilization process expands the capacity of anaerobic digestion to dilute

solids wastes with no need for prior solids separation and no hazard for media clogging. As a

result of no artificial biofilm support media, the digester structure will be very simple and of

lower cost. Sequencing batch mode operation procedures of ASBR are tailored to cater to non-

continuous manure production and collection practices. The multi-feeding procedures are well

adapted to infrequent barn flushing. Owing to simple digester configuration and low

maintenance requirement, this technology is suitable for application to both small farms and

large CAFOs (confined animal feeding operations). This technology is also able to handle a wide

range of TS from 1% to 5% caused by varying flushing intensity and water usage. The kinetic

model and kinetic parameters obtain from this study can be used for design and optimization of

the process. A six-day HRT and an OLR of 1.0–1.5 g VSloaded/L/day are recommended by the

kinetic model prediction. At optimized conditions, a volumetric methane production rate of 0.24

L/L/d of and specific methane productivity of 0.19 L/gVSloaded are expected.

4.5.Conclusion

A successful biomass retention technology for treating flushing dairy manure at psychrophilic

temperature was presented in this study. A Chen & Hashimoto based model gave the best

simulation with R2 of 0.99. The simulation of kinetic modeling indicated the best HRT and OLR

were 4–6 days and 0.5–1.3 g VSloaded/L/day, respectively. Extended SRT was important to retain

high concentration of biomass at low temperature, and to enhance the digester performance.

95

When compared with other research, this technology exhibited a better performance in terms of

specific methane productivity while at shorter HRT and lower temperature.

96

4.6.References

APHA. 1998. Methods for Examination of Water and Wasterwater. 20th ed. American Public

Health Association/American Water Works Association/Water Environment Federation,

Washington, DC, USA.

Batstone, D., Keller, J., Angelidaki, I., Kalyuzhnyi, S., Pavlostathis, S., Rozzi, A., Sanders, W.,

Siegrist, H., Vavilin, V. 2002. Anaerobic digestion model no. 1 (ADM1), IWA Task

Group for mathematical modelling of anaerobic digestion processes. IWA Publishing,

London, UK.

Borja, R., Sanchez, E., Weiland, P., Travieso, L., Martin, A. 1994. Kinetics of anaerobic

degestion of cow manure with biomass immobilized on zeolite. The Chemical

Engineering Journal and the Biochemical Engineering Journal, 54(1), B9-B14.

Buffière, P., Steyer, J.P., Fonade, C., Moletta, R. 1998. Modeling and experiments on the

influence of biofilm size and mass transfer in a fluidized bed reactor for anaerobic

digestion. Water Research, 32(3), 657-668.

Chen, Y.R. 1983. Kinetic analysis of anaerobic digestion of pig manure and its design

implications. Agricultural Wastes, 8(2), 65-81.

Chen, Y.R., Hashimoto, A.G. 1978. Kinetics of methane fermentation. in: Biotechnol. Bioeng.

Symp. 8, pp. Pages: 269-282.

Chen, Y.R., Hashimoto, A.G. 1980. Substrate utilization kinetic model for biological treatment

process. Biotechnology and Bioengineering, 22(10), 2081-2095.

Connaughton, S., Collins, G., O'Flaherty, V. 2006. Psychrophilic and mesophilic anaerobic

digestion of brewery effluent: A comparative study. Water Research, 40(13), 2503-2510.

97

Demirer, G.N., Chen, S.L. 2005. Anaerobic digestion of dairy manure in a hybrid reactor with

biogas recirculation. World Journal of Microbiology & Biotechnology, 21(8-9), 1509-

1514.

Frear, C., Wang, Z.-W., Li, C., Chen, S. 2010. Biogas potential and microbial population

distributions in flushing dairy manure and implications on anaerobic digestion technology.

Journal of Chemical Technology & Biotechnology, 86(1), 145-152.

Gavala, H., Angelidaki, I., Ahring, B. 2003. Kinetics and Modeling of Anaerobic Digestion

Process. in: Biomethanation I, (Eds.) B. Ahring, I. Angelidaki, E.C. Macario, H.N.

Gavala, J. Hofman-Bang, A.J.L. Macario, S.J.W.H.O. Elferink, L. Raskin, A.J.M. Stams,

P. Westermann, D. Zheng, Vol. 81, Springer Berlin Heidelberg, pp. 57-93.

Gillespie, J.R., Flanders, F.B. 2010. Modern Livestock and Poultry Production. 8th ed, Clifton

Park, NY.

Grau, P., Dohányos, M., Chudoba, J. 1975. Kinetics of multicomponent substrate removal by

activated sludge. Water Research, 9(7), 637-642.

Hashimoto, A.G. 1982. Methane from cattle waste: Effects of temperature, hydraulic retention

time, and influent substrate concentration on kinetic parameter (k). Biotechnology and

Bioengineering, 24(9), 2039-2052.

Huang, J.-S., Jih, C.-G. 1997. Deep-biofilm kinetics of substrate utilization in anaerobic filters.

Water Research, 31(9), 2309-2317.

Husain, A. 1998. Mathematical models of the kinetics of anaerobic digestion--a selected review.

Biomass and Bioenergy, 14(5-6), 561-571.

98

Karim, K., Thomas Klasson, K., Hoffmann, R., Drescher, S.R., DePaoli, D.W., Al-Dahhan, M.H.

2005. Anaerobic digestion of animal waste: Effect of mixing. Bioresource Technology,

96(14), 1607-1612.

Kashyap, D.R., Dadhich, K.S., Sharma, S.K. 2003. Biomethanation under psychrophilic

conditions: a review. Bioresource Technology, 87(2), 147-153.

Lettinga, G., Rebac, S., Zeeman, G. 2001. Challenge of psychrophilic anaerobic wastewater

treatment. Trends in Biotechnology, 19(9), 363-370.

Liu, Y., Wang, Z.-W., Qin, L., Liu, Y.-Q., Tay, J.-H. 2005. Selection pressure-driven aerobic

granulation in a sequencing batch reactor. Applied Microbiology and Biotechnology,

67(1), 26-32.

Monod, J. 1949. The Growth of Bacterial Cultures. Annual Review of Microbiology, 3(1), 371-

394.

Pavlostathis, S.G., Giraldogomez, E. 1991. KINETICS OF ANAEROBIC TREATMENT - A

CRITICAL-REVIEW. Critical Reviews in Environmental Control, 21(5-6), 411-490.

Powers, W.J., Wilkie, A.C., VanHorn, H.H., Nordstedt, R.A. 1997. Effects of hydraulic retention

time on performance and effluent odor of conventional and fixed-film anaerobic digesters

fed dairy manure wastewaters. Transactions of the Asae, 40(5), 1449-1455.

Rittmann, B.E., McCarty, P.L. 2001. Environmental biotechnology: principles and applications.

McGraw-Hill, Boston.

Saravanan, V., Sreekrishnan, T.R. 2006. Modelling anaerobic biofilm reactors--a review. J

Environ Manage, 81(1), 1-18.

Umana, O., Nikolaeva, S., Sanchez, E., Borja, R., Raposo, F. 2008. Treatment of screened dairy

manure by upflow anaerobic fixed bed reactors packed with waste tyre rubber and a

99

combination of waste tyre rubber and zeolite: Effect of the hydraulic retention time.

Bioresource Technology, 99(15), 7412-7417.

USEPA. 2010. Draft Inventory of US Greenhouse Gas Emissions and Sinks: 1990–2008, (Ed.)

h.w.e.g.a. 09.18.2012).

Wang, Z.-W., Ma, J., Chen, S. 2011. Bipolar effects of settling time on active biomass retention

in anaerobic sequencing batch reactors digesting flushing dairy manure. Bioresource

Technology, 102(2), 697-702.

Wilkie, A.C., Castro, H.F., Cubinski, K.R., Owens, J.M., Yan, S.C. 2004. Fixed-film anaerobic

digestion of flushing dairy manure after primary treatment: Wastewater production and

characterisation. Biosystems Engineering, 89(4), 457-471.

Zaher, U., Frear, C., Pandey, P., Chen, S. 2008. Evaluation of a new fixed-bed digester design

utilizing large media for flush dairy manure treatment. Bioresour Technol, 99(18), 8619-

25.

100

4.7.Nomenclature

𝜇 specific microbial growth rate (/d)

𝜇𝑚 maximum specific microbial growth rate (/d)

𝑟𝑚 microbial growth rate (g/L/d)

X0 influent biomass concentration (g/L)

X effluent biomass concentration (g/L)

V digester working volume (L)

Q flow rate (L/d)

S0 influent substrate concentration (g/L)

S effluent substrate concentration (g/L)

𝑏 hydraulic retention time (d)

b endogenous decay constant (/d)

k maximum specific substrate utilization rate (gVS/g/d)

k′ first order rate constant (/d)

Ks half-saturation constant (gVS/L)

K dimensionless kinetic parameter of Chen & Hashimoto model

Y growth yield coefficient

c Contois coefficient

B volume of methane produced under standard condition per gram of substrate loaded (L

CH4 STP/gVSloaded)

B0 volume of methane produced under standard condition per gram of substrate loaded at

infinite retention time (L CH4 STP/gVSloaded)

M volumetric methane production rate (L CH4/L/d)

101

P specific methane productivity (L CH4/gVSloaded)

L organic loading rate (gVSloaded/L/d)

102

CHAPTER FIVE

METHANOSARCINA DOMINATION IN ASBR DIGESTER AT SHORT HRT

5.1. Abstract

The microbial population of Anaerobic sequential batch reactor (ASBR) featuring cycle

operations operated under short HRT and low OLR was evaluated for treating a dilute waste

stream. T-RFLP and clone libraries for both 16S rRNA gene and mcrA gene were employed to

characterize the methanogenic community structure. Results revealed that a Methanosarcina

dominated methanogenic community was successfully established when using an ASBR digester

at short HRT. Also the Diversity of the methanogenic community increased as the ASBR was

operated at shorter HRT. It was revealed that both 16S rRNA and mcrA clone library could not

provide complete community structure, while combination of two different clone libraries could

capture more archaea diversity. Thermodynamic calculations confirmed a preference for the

observed population structure. The results both experimentally and theoretically confirmed that

Methanosarcina dominance emphasizing the important role ASBR’s in treating low strength

wastewater as Methanosarcina will be more adaptable at overcoming temperature and shock

loadings experienced with treating this type of wastewater.

Keywords: microbial community structure, ASBR, clone library, T-RFLP, methanogenic

pathway

103

5.2. Introduction

High activity of methanogens is important for maintaining efficient anaerobic digestion and

avoiding accumulation of volatile fatty acids (acetate) and hydrogen, two of the most important

methanation substrates. Methanoarchaea, or methanogens involved in anaerobic digestion, cover

four orders (Liu & Whitman, 2008). All species within three of these orders, Methanobacteriales,

Methanococcales, and Methanomicrobiales, are hydrogenotrophic methanogens reducing CO2 to

CH4 using H2 or formate. The other order, Methanosarcinales, comprising two families,

Methanosaetaceae and Methanosarcinaceae, is unique. The former family contains a single

genus, Methanosaeta, an exclusively acetotrophic species which utilizes acetate as its sole

energy source, while the latter one, including Methanosarcina and several other genuses, can

metabolize both hydrogen and acetate as energy source (Boone et al., 1993).

Methanogens, along with other hydrogenotrophs and acetotrophs compete for acetate and

hydrogen, forming complex Archaea community structures, which vary under different

anaerobic environments (Table 5.1). Affinity to substrate (Ks), growth rate (µmax), and substrate

utilization rate are the key factors governing the dominance of species within a microbial

community. As noted, Methanosarcina are the most versatile methanogens (Zinder, 1993),

offering methanation processes, which when compared to hydrogenotrophic as well as

aceticlastic methanogens (Methanosaeta), are higher in substrate utilization rate, growth rate and

cell yield while exposed to an environment with relatively high acetate and hydrogen

concentration (Daniels et al., 1984). Thus Methanosarcina is favored under conditions in which a

high input of organic matter leads to rapid accumulation of acetate and hydrogen (Zinder, 1993).

consequenctly, digesters dominated by Methanosarcina are more capable of handling increased

104

loads and therefore would be less prone to upset by feeding increases; promoting more stable

digestion (Conklin et al., 2006). However, previous research has repeatedly shown that

Methanosaeta dominance was found in most steady state anaerobic digesters, such as CSTR

(continuous stirred tank reactor) and UASB (upflow anaerobic sludge blanket) (McHugh et al.,

2003; Raskin et al., 1995; Schmidt & Ahring, 1999; Sekiguchi et al., 1998). Raskin et al. (1995)

investigated 21 conventional sewage anaerobic digesters with a wide variation in digester design

and operating conditions by means of molecular probes, and found that Methanosaeta sp.

dominated in all digesters. Their dominance was consistent with the low acetate concentrations

present in all of the digesters conditions, which provided competitive advantage for

Methanosaeta sp. due to their low Ks and threshold compared to Methanosarcina sp. (Table 5.1).

McHugh et al. (2003) examined six granular digesters of varying scale, design, feedstock, and

temperature and the dominance of Methanosaeta sp. was discovered across all digesters,

indicating these filamentous, acetate-utilizing methanogens had a crucial role in the formation

and maintenance of stable anaerobic granules. Correspondingly, it was generally assumed that

Methanosaeta sp. improved granulation and resulted in more stable and higher rate reactor

performance (Schmidt & Ahring, 1999).

Methanosarcina outcompeting Methanosaeta though has been reported under certain operating

conditions (short hydraulic retention time (HRT) or high acetate concentration) (Leclerc et al.,

2004; Mladenovska et al., 2003). Leclerc et al. (2004) reported the dominance of

Methanosarcina sp. in fluidized-bed, fixed-film, ASBR and CSTR, although detailed digester

information was not provided. Prevalence of Methanosarcina sp. in biofilm reactors has also

been reported by Schmidt and Ahring (1999). They concluded that Methanosarcina sp. formed

105

Tabl

e 5.

1 Ty

pica

l com

petit

ive

mic

robe

s in

anae

robi

c di

gest

er

Com

petit

ive

subs

trate

A

ceta

te

H2

Mic

robe

s M

etha

nosa

rcin

a

Met

hano

saet

a

Synt

roph

ic

acet

ate-

oxid

izin

g ba

cter

ia

Hom

oace

toge

ns

Hyd

roge

ntro

phic

m

etha

noge

ns

Sulfa

te re

duci

ng

bact

eria

Ksb

3.0–

4.5

0.49

–0.8

6 —

6

μM o

r 800

Pa

4–8

μM o

r 550

–110

0 Pa

2

μM o

r 250

Pa

Thre

shol

d (m

M)

0.62

–1.2

0.

005–

0.06

9 >

0.2

320–

710

nM o

r 43–

95

Pa

23–7

5 nM

or 3

–10

Pa

6.8

nM o

r 0.9

Pa

Yc 1.

1–3.

1 1.

1–1.

4 —

—

0.

6–6.

4 5.

8 kd

11

.4

9.4

—

—

—

—

μ max

e

0.55

–2.0

4 0.

08-0

.69

—

—

0.70

–3.4

6 1.

44

Prod

ucts

C

H4,

CO

2 C

H4,

CO

2 H

2, C

O2

Ace

tate

C

H4

H2S

ΔG

0’f

–31.

0 –3

1.0

+104

.6

–104

.6

–135

.6

–152

.0

Favo

red

cond

ition

s H

igh

acet

ate

conc

entra

tion

Low

ace

tate

co

ncen

tratio

n

Hig

h hy

drog

en

conc

entra

tion,

low

te

mpe

ratu

re.

Low

hyd

roge

n co

ncen

tratio

n, h

igh

tem

pera

ture

.

Not

e

Met

hano

sarc

ina

will

out

com

pete

Met

hano

saet

a w

ith a

vera

ge a

ceta

te c

once

ntra

tions

hig

her t

han

1.9

mM

.

Hom

oace

toge

ns a

re b

ette

r ada

pted

to g

row

at l

ower

tem

pera

ture

s. Th

e cr

itica

l tem

pera

ture

is 2

0–25

°C.

Met

hano

gens

will

be

inhi

bite

d by

sulfa

te re

duci

ng b

acte

ria w

hen

H2

conc

entra

tion

is b

elow

5μM

. a D

ata

from

Rob

inso

n an

d Ti

edje

(198

4), D

anie

ls e

t al.

(198

4), C

ord-

Ruw

isch

et a

l. (1

988)

, Zin

der (

1993

), C

onkl

in e

t al.

(200

6),

Dem

irel a

nd S

cher

er (2

008)

; b H

alf-

satu

ratio

n co

nsta

nt (m

M);

c Cel

l yie

ld (g

cel

l/rnx

); d M

axim

um sp

ecifi

c su

bstra

te u

tiliz

atio

n ra

te (g

CO

D/g

cel

l·d);

e Max

imum

spec

ific

mic

robi

al g

row

th ra

te (d

–1);

f G

ibbs

free

ene

rgy

unde

r sta

ndar

d co

nditi

ons (

kJ/rn

x).

106

biomass clumps consisting of large numbers of individual cells surrounded by a thick polymeric

wall, as opposed to the filamentous type biofilm consisting of long multicellular rod-shaped

Methanosaeta sp.; noting that the immobilization process of Methanosarcina sp. was even faster

than that of Methanosaeta sp. The stratification of methanogens along the height of an UASB

reactor was noticed with Methanosaeta sp. predominating in the top of the granular layer while

Methanosarcina sp. primarily presented in the bottom of the reactor, concepts which are

consistent with acetate concentration distribution in an UASB (Schmidt & Ahring, 1999). The

presence of Methanosaeta ensured a better performance due to their low acetate threshold,

however, in granules where Methanosarcina sp. were the only acetate-utilizing methanogen

present, a syntrophic acetate-oxidizing system was found (Schmidt & Ahring, 1993).

Operation of an ASBR, with its infrequent feeding and intermittent mixing protocol, creates a

high acetate concentration and dynamic condition within each cycle. Additionally by being able

to uncouple HRT and SRT (sludge retention time), ASBR often employ uniquely short HRT,

adapted to treat low strength waste streams while still treating high flow rates. Although above

advantages for Methanosarcina dominance have been studied in UASB and CSTR, limited data

exists showing methanogenic community dynamics in ASBR digester. The goal of this research

was to investigate that if the unique hydraulic regime of ASBR may select Methanosarcina as

the dominant species, which can lead to a more efficient and more stable anaerobic digestion

process. In this study, the diversities of methanogens in an ASBR digester operated at different

HRT were compared with T-RFLP. The composition of the methanogenic community of the

digester with the best methane production performance was then revealed by sequence analysis

of the partial 16S rRNA gene (~800 base pairs (bp)) and functional gene marker (mcrA gene,

107

about 400 bp) generated from two constructed clone libraries. The correlation between

methanogenic diversity/structure and digester performance was also investigated.

5.3 Materials and methods

Experiments of anaerobic digestion in ASBR digesters with varying HRT were conducted with

the aim to establish a Methanosarcina dominated methanogenic community. Methanogenic

community from steady state digester was characterized by T-RFLP and clone libraries. Flushing

dairy manure, representative of dilute low strength wastewater was selected as the feedstock in

this study.

5.3.1 Feedstock and seed

Fresh dairy manure was collected from the Washington State University Dairy Center in Pullman,

WA, USA and stored at 4 ºC prior to use. Before addition to digesters, manure was diluted with

tap water to mimic flushing dairy manure, which resulted in a waste stream containing 9.1 g/L

total solids (TS) and 7.6 g/L total volatile solids (VS). Anaerobic sludge, for later use as reactor

inoculum, was sampled from an anaerobic digester at the Pullman Wastewater Treatment

Facility. The TS and VS of the sludge were 17.1 g/L and 11.7 g/L, respectively.

5.3.2 Experimental setup and operation

Five digesters each with a working volume of 6 L were operated as ASBR at respective HRT of

4, 8, 12, 16, and 20 days while all other operating conditions remained constant. Each digester

was mixed with a separate impeller driven by a respective motor at 100 rpm. Intermittent mixing

108

was carried out for 10 min every 2h. Manure and anaerobic sludge were introduced to each

digester at a 1:1 volume ratio when experiments started. Digesters were then placed in a low

temperature chamber (22 ºC) and operated in sequencing batch mode, which consisted of 5

stages: filling, reaction, desludge, settling and discharging in one cycle. Steady state conditions

were defined as biogas production, methane content and effluent COD variations less than 10%.

The experiments were conducted for over 120 days.

5.3.3 DNA extraction

Samples of mixed liquor were collected from the five steady-state digesters. Genomic DNA was

extracted and purified using the PowerSoil DNA isolation kit (Mo Bio Laboratories, Inc.,

California) according to the manufacturer’s instructions.

5.3.4 PCR amplification of the 16S rRNA gene for T-RFLP

A fragment of 822 bp length of the DNA was amplified using primers Ar109f and 6-FAM

labeled Ar915r (Grosskopf et al., 1998). PCR amplification was carried out using standard buffer

conditions, DNA template and GoTaq DNA Polymerase (Promega, WI, USA) with a total

volume of 50 μl in a PTC-200 DNA Engine Peltier Thermal Cycler (Bio-Rad Laboratories, Inc.,

CA, USA) with initial denaturation at 95°C for 5 min, 25 cycles of denaturing at 95°C for 0.5

min, annealing at 55°C for 0.5 min and extension at 72°C for 1.5 min, plus a final extension at

72°C for 10 min. Amplified DNA was verified by electrophoresis of aliquots of PCR mixtures (5

µl) in 1.0% agarose in 1× TAE buffer and EtBr staining.

109

5.3.5 PCR Amplification of Functional Gene Marker for T-RFLP

Fragments of mcrA gene encoding methyl coenzyme M Reductase in methanogenic Archaea was

amplified with primers MCRf and 6-FAM labeled MCRr (Springer et al., 1995). PCR

amplification was performed with the same procedures for 16S rRNA gene.

5.3.6 T-RFLP analysis

After purification with QIAquick spin columns (Qiagen, Alameda, CA), approximately 100 ng

of the amplicons was digested separately with 15 U of the restriction endonucleases TaqI

(Thermo Fisher Scientific Inc., Rockford, IL) for 16S rRNA PCR products (Lueders & Friedrich,

2000) and Sau96I (New England Biolabs, Ipswich, MA) for mcrA gene PCR products (Lueders

et al., 2001). The digestions were carried out in a total volume of 30 µl incubating at 37°C

overnight. Subsequent electrophoresis was performed on an ABI 3730xl sequencer (Applied

Biosystems, Carlsbad, CA) by GENEWIZ Inc. (South Plainfield, NJ). Based on the internal LIZ-

500 size standard, the terminal restriction fragments (T-RFs) were sized using GeneMarker V 2.2

(SoftGenetics LLC., USA).

Fragments with a peak height below 0.5 % (regarded as background noise) as well as those with

a size less than 50 bp were excluded from the analysis. Peaks over a threshold of 50 relative

fluorescent units above background fluorescence were analyzed by manually aligning fragments

to the size standard. To avoid detection of primers and uncertainties of size determination,

terminal fragments smaller than 50 bp and larger than 500 bp were excluded from the analysis.

Peaks with less than 0.5 bases differences were merged. The T-RFs profiles were normalized by

110

calculating the relative abundance (percentage) of each fragment (each T-RF peak area divided

by the total peak area of each sample). The predicted T-RF lengths of known Archaea species

were obtained using the Phylogenetic Assignment Tool (PAT) (Kent et al., 2003), with a

modified database from the Ribosomal Database Project (Cole et al., 2009) and our database

from either the 16S rRNA gene or mcrA gene clone libraries generated by using the ISPaR tool

of MiCA, a web-based tool for the analysis of microbial communities (Shyu et al., 2007).

5.3.7 Clone Library of the 16S rRNA Gene

For clone library construction and phylogenetic analysis, PCR products were obtained in

triplicate as described for PCR of 16S rRNA gene except that the Ar915r primer was not labeled

with 6-FAM. The purified amplified products were pooled, cloned into pGEM-T Easy vector

using the TA cloning kit (Promega, USA), and transformed into competent Escherichia coli

JM109 for sequencing. Random selection of white colonies were performed after E. coli M109

cells grown at 37°C overnight on LB plates containing 100 µg/ml ampicillin, 0.5mM IPTG, and

40 µg/ml X-Gal. The colonies that contained an appropriately sized insert were sequenced using

primers SP6 by GENEWIZ Inc (South Plainfield, NJ).

5.3.8 Clone Library of Functional Gene Marker

For clone library construction and phylogenetic analysis, PCR products were obtained in

triplicate as described for PCR of mcrA gene except that the MCRr primer was not labeled with

6-FAM. The rest of the procedure followed the same as the clone library of the 16S rRNA gene.

111

5.3.9 Phylogenetic analysis

All sequences in each clone library were compared to entries of the NCBI nucleotide sequence

database using BLAST. The 16S rRNA amplicon sequences and deduced amino acid sequences

from mcrA sequences were respectively aligned with selected known sequences from reference

strains by means of software tool MEGA 5 (Tamura et al., 2011). Subsequently, sequences were

assigned to operational taxonomic units (OTUs) by comparing sequences on the basis of 97%

similarity between sequences. Diversity indices, such as the Shannon, the Evenness, the Chao1,

ACE and coverage were used to measure the diversity of each library. Diversity indices, OTU

clustering, rarefaction analysis of library structure and multiple sample statistics were computed

using the MOTHUR programs package (Schloss et al., 2009). The steepness (θ) of rarefaction

curve was calculated from the equation y = mx + c where ‘x’ and ‘y’ are the coordinates of the

points ‘m’ is the gradient and ‘c’ is the ‘y intercept’ of the straight line. Phylogenetic trees were

constructed by means of the MEGA 5 (Tamura et al., 2011) using the Neighbor-Joining method

and Bootstrap resampling analysis for 1000 repeats. The scale bar represents 5% substitutions

per nucleotide sequence position. Numbers at nodes represent bootstrap values. Only bootstrap

values above 50% are displayed.

5.3.10 Nucleotide sequence accession numbers

The nucleotide sequences reported in this paper have been deposited in the Genbank, EMBL, and

DDBJ nucleotide sequence database under the accession numbers JQ684533-JQ684646 (16S

rRNA gene) and JQ697374-JQ697491 (mcrA gene). Accession numbers of clones representing

OTUs are depicted in Figure 5.3 and 5.4.

112

5.4 Results and discussions

5.4.1 Digesters performance

The experiments of anaerobic digestion of flushing dairy manure in ASBR digesters were

conducted for over 120 days, and all digesters reached respective steady-state after 80 days. The

amount of methane production was affected by HRT as shown in Figure 5.1A. The longer the

HRT, the more methane production was achieved. However, the highest volumetric methane

production was observed at the shortest HRT of 4 days at a rate of 0.26 L/L/d, while the longest

HRT of 20 days resulted in the lowest rate of 0.08 L/L/d.

A relatively high acetate concentration (above 1 mM) and hydrogen partial pressure (about 60 pa)

was detected within the ASBR at an HRT of 4 days. With the extension of HRT, concentrations

for both acetate and hydrogen in the digesters approached to zero. Methanosaeta grows at low

acetate concentration, whereas Methanosarcina consumes acetate only at high acetate

concentrations (Boone et al., 1993; Zinder, 1993). Methanosarcina have a higher threshold and

maximum substrate utilization rate compared to Methanosaeta (Table 5.1), due to different

mechanisms of acetate metabolism, i.e., by acetate kinase/phosphotransacetylase versus acetyl-

CoA synthetase, respectively (Jetten et al., 1992). During each cycle of the digester running at 4-

day HRT, the overall high acetate concentration varied from 6 mM after feeding to 1 mM at the

end of cycle, favoring growth of Methanosarcina, leading to a Methanosarcina dominated

community. The digesters with longer HRT functioned like a regular plug flow digester; since

acetate concentration was not detected at most of cycle times in these digesters. Under these high

HRT conditions, Methanosaeta could outcompete over Methanosarcina by taking advantage of

high acetate affinity and dominated in Archaea community.

113

5.4.2 Methanogenic community diversity

Methanogenic community diversities at different HRT were revealed by T-RFLP. The number of

T-RFs represents the community diversity. As Figure 5.1A shows, the number of T-RFs

decreased with the extension of HRT, indicating low community diversity within ASBR

digesters at long HRT. The reason for this result could be limited carbon source. As the dilute

organic waste stream was used as a feedstock, the available substrate for microbes was reduced

while conducting anaerobic digestion at a longer HRT. Presumably, with limited substrate

availability, some species disappeared at the longer HRT, which led to decrease in diversity of

microbial community.

When plotting the number of T-RFs and the methane production rate together in Figure 5.1B, a

high correlation between the two was found, which revealed that digester performance was

related to community diversity. A digester with high microbial community diversity showed

somewhat high methane production rate, while a digester of low performance was associated

with low diversity of microbial community. The enhanced performance with high community

diversity might be due to the presence of multiple pathways conducted by a more complex

community. Krakat et al. (2010) reported that HRT decrease not only resulted in increased

species diversity, but also led to slightly higher specific methane production rate. Presumably, as

availability of organic substrate increased with HRT reduction, proliferation of underrepresented

or rare methanogens occurred, leading to enhanced Archaea diversity.

114

Figure 5.1 Relationship between digester performance and microbial community diversity

115

Figure 5.2 Relative abundance of the Archaea 16S rRNA gene fragments (A) and mcrA gene

fragments (B) retrieved from the biomass in ASBR. The length of T-RFs in base pairs (bp) is

indicated in parenthesis

A

B

116

Figure 5.2 clearly showed an Archaea community shift with change in HRT. At short HRTs,

substrate concentration and acetate concentration varied rapidly during each cycle, resulting in a

stressed condition for microbial community. Species affiliated to Methanosarcinaceae displaying

competitive advantage via high growth rate and high substrate utilization rate predominated in

the Archaea community. With long HRT, the performance of ASBR tended to be similar to a

steady state CSTR or plug-flow digester featuring low acetate concentration in the effluent;

leading to Methanosaeta outcompeting Methanosarcina. The growth of hydrogen-oxidizing

methanogens from Methanomicrobiales and Methanobacteriales were promoted by the

proliferation of exclusive acetate utilizing Methanosaeta, so that hydrogen methanogenesis was

significant at long HRT (Figure 5.2). Hydrogenotrophic methanogen dominated biogas plants

have been reported by several studies (Nettmann et al., 2008; Rastogi et al., 2008), which mostly

included Methanomicrobiales, while Zhu et al. (2011) reported that Methanobacteriales were

the dominant species in digester fed with swine manure.

5.4.3 Phylogenetic analysis

5.4.3.1 16S rRNA clone library

The results of the phylogenetic cluster of assigned OTUs indicated that Methanosarcina was the

dominant methane producing Archaea in the ASBR digester (Figure 5.3). A total of 114 16S

rRNA clone sequences were analyzed and a total of 18 OTUs generated were affiliated to

Methanobacteriales, Methanomicrobiales, and Methanosarcinales, while Methanococcales was

absent. Five OTUs (68.42%) were assigned to Methanosarcinales which included aceticlastic

strains of Methanosaeta sp. (23.68%) and the most metabolically diverse Methanosarcina sp.

(44.74%). Obligate hydrogenotrophic Methanobacteriales and Methanomicrobiales included a

117

total of 9 clones (7.89%) in 5 OTUs and 27 clones (23.68%) in 8 OTUs, respectively (Table 5.3).

Methanomicrobiales showed a high diversity with 8 OTUs covering Methanomicrobiaceae and

Methanospirillaceae. Only 9% of clone was assigned to Methanobacteriales, which mainly

consisted of Methanobacteriaceae.

5.4.3.2 mcrA gene clone library

A similar result was obtained from the mcrA clone library while more OTUs were detected

compared to the 16S rRNA clone library (Figure 5.4). A total of 118 mcrA clone sequences

were analyzed and a total of 21 OTUs were generated that were affiliated to Methanobacteriales,

Methanomicrobiales, and Methanosarcinales. 60.17% of clones (3 OTUs) were allocated to the

order of Methanosarcinales, while 13.56% (5 OTUs) and 26.26% (13 OTUs) of clones were

assigned to hydrogen-oxidizing Methanobacteriales, Methanomicrobiales, respectively (Table

5.4). No Methanosaeta sp. was detected in Methanosarcinales and all clones belong to the

family Methanosarcinaceae. Methanobacteriales, Methanomicrobiales showed a similar

diversity and proportion to results from to16S rRNA clone library.

118

Figure 5.3 Phylogenetic relationship among 18 OTUs with partial 16S rRNA gene sequences of

known methanogenic Archaea based on the neighbor-joining analysis

OTU-03 (JQ684535) (5 clones) uncultured Methanomicrobiales (AB236081.1)

OTU-04 (JQ684592) (11 clones) Methanolinea sp. (JN836394.1)

Methanolinea sp. (AB447467.1)

Methanosphaerula palustris (EU156000.1) OTU-11 (JQ684579)

Methanomicrobiales (JN836393.1)

OTU-08 (JQ684598) (4 clones) Methanospirillum hungatei (AY196683.1)

OTU-12 (JQ684583) OTU-13 (JQ684585)

OTU-09 (JQ684568) (4 clones) Methanospirillum lacunae (AB517986.1)

OTU-05 (JQ684537) (12 clones) Methanosaeta sp. (HQ133141.1) OTU-07 (JQ684558) (15 clones) uncultured Methanosarcinales (CU917405.1)

Methanosaeta thermophila (AB071701.1)

Methanosarcina lacustris (AF432127.1) Methanosarcina mazei (DSM 2053)

Methanosarcina siciliae (FR733698.1) uncultured Methanosarcina sp. (AB636574.1) OTU-01 (JQ684634) (49 clones)

OTU-15 (JQ684610) uncultured Methanosarcinaceae (AJ879017.1)

Methanotorris formicicus (AB100884.1) OTU-16 (JQ684623)

Methanobacterium alcaliphilum (AB496639.1) OTU-18 (JQ684644)

OTU-06 (JQ684539) uncultured Methanobacteriaceae (FJ579962.1)

OTU-14 (JQ684593) uncultured Methanobacteriaceae (FJ579348.1)

Methanobacterium beijingense (AY552778.3) OTU-17 (JQ684635) (3 clones)

OTU-02 (JQ684557) (2 clones) Methanobacterium formicicum (DSMZ1535) Methanobacterium petrolearium (AB542742.1) OTU-10 (JQ684560)

Methanobacterium kanagiense (AB368917.1)

100

100

100

100

100

99

100

100

92

99

90

82

88

61

100

100

71

93 97

100

100

58

99

100

100 99

71 96

50

100

100 97

98

99

99 99

71

0.05

Methanomicrobiales 23.68%

Methanosarcinales

68.42%

Methanobacteriales 7.89%

119

Figure 5.4 Phylogenetic relationship among 21 OTUs with partial mcrA gene sequences of

known methanogenic Archaea based on the neighbor-joining analysis

OTU-16 (JQ697474) OTU-20 (JQ697481)

OTU-11 (JQ697443) Methanospirillum hungatei (AAK16835) OTU-21 (JQ697486) (3 clones) OTU-12 (JQ697450)

uncultured archaeon (ACL80596)

OTU-15 (JQ697462) (2 clones)

OTU-9 (JQ697430) (11 clones)

Methanoculleus thermophilus (AAK16834) Methanoculleus bourgensis (AAL29285)

Methanomicrobiaceae strain (AAC43430)

OTU-18 (JQ697477) (2 clones) OTU-6 (JQ697405)

OTU-14 (JQ697485)

Methanomicrobium mobile BP (AAL29293) Methanocorpusculum bavaricum (AAL29298)

Methanocorpusculum labreanum (AAP20896) uncultured methanogenic archaeon (ABI18566)

Methanocorpusculum aggregans (AAL29283) Methanocorpusculum parvum (AAP20900)

OTU-3 (JQ697396)

OTU-2 (JQ697389) (5 clones) OTU-8 (JQ697419)

Methanosaeta concilii (AAK16832) Methanosaeta concilii (AAL29286) Methanosaeta concilii (AAK16833)

Methanolobus bombayensis (AAC43407) Methanosarcina vacuolata (AAC43428)

Methanosarcina acetivorans (AAC43405) Methanosarcina barkeri (CAA68357)

Methanosarcina mazei (AAC43413)

OTU-7 (JQ697418) Methanocaldococcus jannaschii DSM 2661 (AAB98063)

Methanococcus voltae PS (CAA30633) Methanothermobacter thermautotrophicus (AAA73439)

uncultured archaeon (ACL80573)

Methanopyrus kandleri (AAB02003) Methanothermus fervidus (AAA72197)

OTU-1 (JQ697374) Methanobacterium bryantii (AAK16836) Methanothermobacter thermautotrophicus (CAA30639)

Methanothermobacter wolfeii (BAF56659) OTU-5 (JQ697404) Methanobrevibacter arboriphilus (AAL29284)

Methanobacterium formicicum (AAL29299) OTU-17 (JQ697476) (10 clones)

OTU-4 (JQ697398) (3 clones)

95

76

100

96

76

74 92

87

96

54 56

84

75

95

96 99

93

61 81

61

87 96

99

67 53

78

86 73

97 98 60

65

0.05

OTU-13 (JQ697460)

OTU-19 (JQ697381) (3 clones)

OTU-10 (JQ697431) (67 clones)

Methanomicrobiales 26.26%

Methanosarcinales 60.17%

Methanobacteriales 13.56%

120

Figure 5.5 Rarefaction curves generated for 16S rRNA genes and mcrA genes clone libraries.

Dotted lines indicate the 95% confidence intervals

Table 5.2 Diversity and richness indices for 16S rRNA and mcrA clone libraries

Clone library Na nb Shannon (H) Evenness (E)c Coveraged Chao1 ACE

16S rRNA 114 18 2.01 0.69 0.93 36 33

mcrA gene 118 21 1.80 0.60 0.90 34 55

aN: number of clones in library;

bn: detected OTUs defined at 97% sequence identity;

cE: evenness calculated based on the formula of H/Hmax;

dCoverage: calculated by the formula of [1-(n/N)].

121

5.4.3.3 Statistical analysis of clone libraries

The rarefaction curves for both 16S rRNA and mcrA clone libraries were generated in Figure 5.5.

The curve steepness is a function of the community taxon evenness, while its height indicates its

taxon richness. The sequence population was more diverse for the mcrA gene than for the 16S

rRNA gene, as the rarefaction curve of the mcrA gene clone library showed a higher slope than

the curve for 16S rRNA gene. The steepness for 16S rRNA and mcrA clone libraries was 7.31°

and 8.85°, respectively. As revealed by rarefaction analysis, both curves reached plateau, an

indicator that each library size was sufficient.

Table 5.3 Cluster of Archaea clone sequences for 16S rRNA clone library and its affiliation

regarding BLAST search

OTUs clones % Methanobacteriales 5 9 7.89 Methanobacteriaceae 5 9 7.89 Uncultured Methanobacteriaceae 2 2 1.75 Methanobacterium sp. 2 6 5.26 Methanobrevibacter sp. 1 1 0.88 Methanomicrobiales 8 27 23.68 Methanomicrobiaceae 3 17 Methanolinea sp. 1 5 4.39 Uncultured Methanomicrobiaceae 2 12 10.53 Methanospirillaceae 5 10 8.77 Methanospirillum sp. 5 10 8.77 Methanosarcinales 5 78 68.42 Methanosaetaceae 2 27 23.68 Methanosaeta sp. 2 27 23.68 Methanosarcinaceae 3 51 44.74 Methanosarcina sp. 1 49 42.98 Uncultured Methanosarcinaceae 2 2 1.75

122

Richness of the two libraries was assessed by two nonparametric richness estimators, Chao1 and

ACE, to determine whether the library was of sufficient size to obtain meaningful stable richness

estimates. Chao1 and ACE showed higher estimation values of expected OTUs for the available

clone number. The respective coverage for 16S rRNA and mcrA clone libraries were 0.93 and

0.90, which also confirmed that the major part of diversity in each library was detected, leaving

some rare species from the digester unrevealed (Table 5.2). The Shannon index (H) and the

evenness (E) are commonly used to characterize species diversity in a community. The Shannon

indices for the 16S rRNA clone library (H = 2.01) was higher than that for the mcrA clone

library (1.80). H accounts for both diversity and uniformity of the species and is maximum only

when all species have perfectly even distribution of abundances thus indicating that detected

species are more uniformly distributed in the 16S rRNA clone library than in the mcrA clone

library. The evenness of 0.60 of the mcrA clone library was lower than the evenness of the 16S

rRNA clone library (E = 0.69), which also confirmed that the species in the 16S rRNA clone

were distributed more equitably among Archaea community compared to the mcrA clone library.

5.4.3.4 Comparison of clone library for 16S rRNA gene and mcrA gene

Comparative study of the two clone libraries constructed for 16S rRNA gene and mcrA gene

revealed that the gene of mcrA is capable of detecting more diversity of overall methanogenic

communities, especially for the orders Methanomicrobiales and Methanobacteriales. However, it

explored few OTUs in the order Methanosarcinales, and even failed to detect Methanosaeta sp.

This result was consistent with the observation of other researchers (Chin et al., 2004; Nettmann

et al., 2008). Nettmann et al. (2008) evaluated the archaea community diversity of a biogas plant

utilizing herbal biomass with 16S rRNA and mcrA clone library. The results showed that the

123

mcrA clone library explored 3 times the number of OTUs that of 16S rRNA clone library, while

the mcrA clone library could not detect Methanosaeta sp., which led to a low diversity in

Methanosarcinales. Chin et al. (2004) analyzed archaea community structure on rice roots and

found that only 1 OTU (1 clone) of mcrA clone was assigned to Methanosaetaceae while 16S

rRNA gene revealed 4 OTUs (7 clone) were affiliated to Methanosaetaceae, since the mcrA

primer set was not be able to detect Methanosaetaceae (Lueders et al., 2001). As this study as

well as previous research revealed that, neither 16S rRNA none mcrA clone library could not

provide complete community structure. It is possible to capture the whole spectrum of archaea

diversity of an environmental sample by combining two different clone libraries.

Table 5.4 Cluster of Archaea clone sequences obtained for mcrA gene clone library and its

affiliation regarding BLAST search

OTUs clones % Methanobacteriales 5 16 13.56 Methanobacteriaceae 5 16 13.56 Methanobacterium sp. 4 15 12.71 Uncultured Methanobacteriaceae 1 1 0.85 Methanomicrobiales 13 31 26.26 Methanomicrobiaceae 6 10 8.47 Methanolinea sp. 2 6 5.08 Uncultured Methanomicrobiaceae 4 4 3.39 Methanospirillaceae 7 21 17.79 Methanospirillum sp. 7 21 17.79 Methanosarcinales 3 71 60.17 Methanosarcinaceae 3 71 60.17 Methanosarcina sp. 3 71 60.17

124

5.4.4 Methanogenic pathway in psychrophilic ASBR

The formation of methane is subject to a functional anaerobic microbial community, where

methanogen interact with other anaerobic microbial consortia. The methanogenic pathways are

often affected by their interaction with homoacetogens and syntrophic acetate-oxidizing bacteria.

Homoacetogen reduce CO2 to acetate using H2 while syntrophic acetate-oxidizing bacteria

convert acetate to H2 plus CO2. Competitions among methanogens and syntrophic microbial

community would be thermodynamically affected by the changes of acetate and H2 concentration.

For a chemical reaction (Eq. (5.1)), the ∆𝐺′ can be estimated using the Nernst equation (Eq.

(5.2)).

aA + bB → cC + dD (5.1)

∆𝐺′ = ∆𝐺0′ + 𝑅𝑇 𝑙𝑛[𝐶]𝑐[𝐷]𝑑

[𝐴]𝑎[𝐵]𝑏 (5.2)

where ∆𝐺0′is Gibbs free energy under standard conditions, while the second item describes the

thermodynamics of the reaction under real conditions, in which R is ideal gas constant, and T is

absolute temperature in kelvin.

The Gibbs free energy of reactions related to methane formation was calculated and listed in

Table 5.5. Under common digester conditions, aceticlastic methanogenesis is the most

thermodynamically favorable reaction thus it will be the main pathway for methane formation

125

(Table 5.5). In an ASBR digester though, due to the unique operation conditions, the

thermodynamically favorable methane production reaction was shifted by the changing

concentration of metabolic intermediate, i.e. acetate and hydrogen (Table 5.5). The dynamic

changes of acetate concentration within each cycle as a result of infrequent feeding strategy was

actually a shock loading, functioned as selection pressure, which led to Methanosarcina

outcompeting over Methanosaeta and dominating in the ASBR methanogenic Archaea

community.

Table 5.5 Gibbs free energy of reactions associated with methane generation

Reactions

(kJ/rnx)

Aceticlastic methanogenesis

Homoacetogenesis Hydrogenotrophic methanogenesis

Homoacetogenesis + Aceticlastic methanogenesis

ΔG’a

–20.1 +7.5 –12.6 –12.6

ΔG’b

–20.5 –36.8 –57.3 –57.3 aΔG’ calculated for typical digester conditions: T 308K, pH 7, pH2 1Pa, pCH4 70kPa, HCO3

–

0.1M, and acetic acid 1mM (Batstone et al., 2002).

bΔG’ calculated for present study: T 295K, pH 7, pH2 60Pa, pCH4 70kPa, HCO3– 0.1M, and

acetic acid 1mM.

126

Figure 5.6 Speculated Methanogenic Pathway in ASBR

Homoacetogens and hydrogenotrophic methanogens are competitors for hydrogen consumption.

Homoacetogens growth is fast but requires a high hydrogen threshold, while hydrogenotrophic

methanogens can utilize hydrogen at much lower concentration and at a considerably lower rate

(Table 5.5). High hydrogen partial pressure was detected in the ASBR digester headspace at

short HRTs, which favored the growth of homoacetogens over hydrogenotrophic methanogens,

therefore homoacetogens became the main hydrogen-consuming microorganism. Low

temperature was another factor accounting for the prevalence of homoacetogens, due to

decreased affinity of methanogens for hydrogen at low temperature (Kotsyurbenko, 2005). As

127

Table 5.5 shows at the above mentioned ASBR conditions, homoacetogenesis plus aceticlastic

methanogenesis is the most thermodynamically favorable pathway. Homoacetogenic bacteria,

together with aceticlastic methanogens, especially Methanosarcina, predominated for hydrogen

and acetate consumption, so that played a key role in the ultimate degradation of organic

substrate and production of methane. It could be speculated that the main pathway for methane

formation in ASBR at short HRT was homoacetogenesis plus aceticlastic methanogenesis

(Figure 5.6).

5.5 Conclusion

As revealed by phylogenetic analysis of constructed clone libraries in this study, the ASBR at

short HRTs was capable of establishing a Methanosarcina predominated methanogenic Archaea

population, which led to a more stable and efficient anaerobic digestion process. This study also

proved that reactor performance was related to microbial community diversity; the community

showed higher diversity when reactor performance was superior. The comparison study showed

that the mcrA clone library detected more Archaea diversity, especially in Methanomicrobiales,

while failed to detect Methanosaeta sp., compared to 16S rRNA clone library. The

experimentally shown dominance of Methanosarcina was explained by thermodynamic

calculation by assessing different digester conditions. It was speculated that homoacetogenesis

plus aceticlastic methanogenesis are the main pathways for methane production in the ASBR

digester.

128

5.6 Reference

Batstone, D., Keller, J., Angelidaki, I., Kalyuzhnyi, S., Pavlostathis, S., Rozzi, A., Sanders, W.,

Siegrist, H., Vavilin, V. 2002. Anaerobic digestion model no. 1 (ADM1), IWA Task

Group for mathematical modelling of anaerobic digestion processes. IWA Publishing,

London, UK.

Boone, D.R., Whitman, W.B., Rouviere, P. 1993. Diversity and Taxonomy of Methanogens. in:

Methanogenesis: Ecology, Physiology, Biochemistry and Genetics, (Ed.) J.G. Ferry,

Chapman & Hall. New York, pp. 35–80.

Chin, K.J., Lueders, T., Friedrich, M.W., Klose, M., Conrad, R. 2004. Archaeal Community

Structure and Pathway of Methane Formation on Rice Roots. Microbial Ecology, 47(1),

59-67.

Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., Kulam-Syed-Mohideen, A.S.,

McGarrell, D.M., Marsh, T., Garrity, G.M., Tiedje, J.M. 2009. The Ribosomal Database

Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Research,

37(suppl 1), D141-D145.

Conklin, A., Stensel, H.D., Ferguson, J. 2006. Growth Kinetics and Competition Between

Methanosarcina and Methanosaeta in Mesophilic Anaerobic Digestion. Water

Environment Research, 78(5), 486-496.

Cord-Ruwisch, R., Seitz, H.-J., Conrad, R. 1988. The capacity of hydrogenotrophic anaerobic

bacteria to compete for traces of hydrogen depends on the redox potential of the terminal

electron acceptor. Archives of Microbiology, 149(4), 350-357.

Daniels, L., Sparling, R., Sprott, G.D. 1984. The bioenergetics of methanogenesis. Biochimica et

biophysica acta, 768(2), 113-163.

129

Demirel, B., Scherer, P. 2008. The roles of acetotrophic and hydrogenotrophic methanogens

during anaerobic conversion of biomass to methane: a review. Reviews in Environmental

Science and Biotechnology, 7(2), 173-190.

Grosskopf, R., Janssen, P.H., Liesack, W. 1998. Diversity and Structure of the Methanogenic

Community in Anoxic Rice Paddy Soil Microcosms as Examined by Cultivation and

Direct 16S rRNA Gene Sequence Retrieval. Appl. Environ. Microbiol., 64(3), 960-969.

Jetten, M.S.M., Stams, A.J.M., Zehnder, A.J.B. 1992. Methanogenesis from acetate: a

comparison of the acetate metabolism in Methanothrix soehngenii and Methanosarcina

spp. FEMS Microbiology Letters, 88(3–4), 181-197.

Kent, A.D., Smith, D.J., Benson, B.J., Triplett, E.W. 2003. Web-Based Phylogenetic Assignment

Tool for Analysis of Terminal Restriction Fragment Length Polymorphism Profiles of

Microbial Communities. Applied and Environmental Microbiology, 69(11), 6768-6776.

Kotsyurbenko, O.R. 2005. Trophic interactions in the methanogenic microbial community of

low-temperature terrestrial ecosystems. FEMS Microbiol Ecol, 53(1), 3-13.

Krakat, N., Schmidt, S., Scherer, P. 2010. Mesophilic Fermentation of Renewable Biomass:

Does Hydraulic Retention Time Regulate Methanogen Diversity? Applied and

Environmental Microbiology, 76(18), 6322-6326.

Leclerc, M., Delgènes, J.-P., Godon, J.-J. 2004. Diversity of the archaeal community in 44

anaerobic digesters as determined by single strand conformation polymorphism analysis

and 16S rDNA sequencing. Environmental Microbiology, 6(8), 809-819.

Liu, Y., Whitman, W.B. 2008. Metabolic, Phylogenetic, and Ecological Diversity of the

Methanogenic Archaea. Annals of the New York Academy of Sciences, 1125(1), 171-

189.

130

Lueders, T., Chin, K.-J., Conrad, R., Friedrich, M. 2001. Molecular analyses of methyl-

coenzyme M reductase α-subunit (mcrA) genes in rice field soil and enrichment cultures

reveal the methanogenic phenotype of a novel archaeal lineage. Environmental

Microbiology, 3(3), 194-204.

Lueders, T., Friedrich, M. 2000. Archaeal Population Dynamics during Sequential Reduction

Processes in Rice Field Soil. Appl. Environ. Microbiol., 66(7), 2732-2742.

McHugh, S., Carton, M., Mahony, T., O'Flaherty, V. 2003. Methanogenic population structure in

a variety of anaerobic bioreactors. FEMS Microbiology Letters, 219(2), 297-304.

Mladenovska, Z., Dabrowski, S., Ahring, B.K. 2003. Anaerobic digestion of manure and mixture

of manure with lipids: biogas reactor performance and microbial community analysis.

Water Science & Technology, 48(6), 271-278.

Nettmann, E., Bergmann, I., Mundt, K., Linke, B., Klocke, M. 2008. Archaea diversity within a

commercial biogas plant utilizing herbal biomass determined by 16S rDNA and mcrA

analysis. Journal of Applied Microbiology, 105(6), 1835-1850.

Raskin, L., Zheng, D., Griffin, M.E., Stroot, P.G., Misra, P. 1995. Characterization of microbial

communities in anaerobic bioreactors using molecular probes. Antonie van Leeuwenhoek,

68(4), 297-308.

Rastogi, G., Ranade, D.R., Yeole, T.Y., Patole, M.S., Shouche, Y.S. 2008. Investigation of

methanogen population structure in biogas reactor by molecular characterization of

methyl-coenzyme M reductase A (mcrA) genes. Bioresource Technology, 99(13), 5317-

5326.

131

Robinson, J.A., Tiedje, J.M. 1984. Competition between sulfate-reducing and methanogenic

bacteria for H&lt;sub&gt;2&lt;/sub&gt; under resting and growing conditions. Archives

of Microbiology, 137(1), 26-32.

Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski,

R.A., Oakley, B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G.,

Van Horn, D.J., Weber, C.F. 2009. Introducing mothur: Open-Source, Platform-

Independent, Community-Supported Software for Describing and Comparing Microbial

Communities. Applied and Environmental Microbiology, 75(23), 7537-7541.

Schmidt, J.E., Ahring, B.K. 1993. Effects of magnesium on thermophilic acetate-degrading

granules in upflow anaerobic sludge blanket (UASB) reactors. Enzyme and Microbial

Technology, 15(4), 304-310.

Schmidt, J.E., Ahring, B.K. 1999. Immobilization Patterns and Dynamics of Acetate-Utilizing

Methanogens Immobilized in Sterile Granular Sludge in Upflow Anaerobic Sludge

Blanket Reactors. Applied and Environmental Microbiology, 65(3), 1050-1054.

Sekiguchi, Y., Kamagata, Y., Syutsubo, K., Ohashi, A., Harada, H., Nakamura, K. 1998.

Phylogenetic diversity of mesophilic and thermophilic granular sludges determined by

16S rRNA gene analysis. Microbiology, 144(9), 2655-2665.

Shyu, C., Soule, T., Bent, S., Foster, J., Forney, L. 2007. MiCA: A Web-Based Tool for the

Analysis of Microbial Communities Based on Terminal-Restriction Fragment Length

Polymorphisms of 16S and 18S rRNA Genes. Microbial Ecology, 53(4), 562-570.

Springer, E., Sachs, M.S., Woese, C.R., Boone, D.R. 1995. Partial Gene Sequences for the A

Subunit of Methyl-Coenzyme M Reductase (mcrI) as a Phylogenetic Tool for the Family

Methanosarcinaceae. International Journal of Systematic Bacteriology, 45(3), 554-559.

132

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S. 2011. MEGA5:

Molecular Evolutionary Genetics Analysis Using Maximum Likelihood, Evolutionary

Distance, and Maximum Parsimony Methods. Molecular Biology and Evolution, 28(10),

2731-2739.

Zhu, C., Zhang, J., Tang, Y., Zhengkai, X., Song, R. 2011. Diversity of methanogenic archaea in

a biogas reactor fed with swine feces as the mono-substrate by mcrA analysis.

Microbiological Research, 166(1), 27-35.

Zinder, S.H. 1993. Physiological ecology of methanogens. in: Methanogenesis: Ecology,

Physiology, Biochemistry and Genetics, (Ed.) J.G. Ferry, Chapman & Hall. New York,

pp. 128–206.

133

CHAPTER SIX

CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK

In this study, a novel strategy, improved biomass retention with fiber material present within the

dairy manure as biofilm carriers under selection pressure, was developed and evaluated for

treating flushing dairy manure in a psychrophilic ASBR. The main conclusions from this

research are summarized below:

1. Methodology for rate-limiting step determination for anaerobic digestion of solids

containing substrate and effect of microbial community ratio

A simple and fast procedure, revised BMP test with metabolic intermediates supplementation

was developed in this study to determine the rate-limiting steps in the degradation of complex

substrates. With this method, it was illustrated that hydrolysis of dairy manure was the rate-

limiting step when anaerobic sludge was used as inoculum. The results also showed that an

increase of microbial concentration could shorten the lag phase at the beginning of digestion

process, which indicates anaerobic digestion of dairy manure can be accelerated by dosing

anaerobic sludge seed and potentially can reduce HRT of digester. Additionally, the ultimate

biogas yield was also augmented by sludge supplementation, showing the possibility of more

energy yield in a biogas plant when operated with higher biomass concentration.

This study revealed that the microbial community ratio has a strong effect on the anaerobic

digestion rate. Results showed that the rate-limiting step changed with the variation of microbial

community ratio. It showed that there exists a critical ratio r* at 24 regarding dairy manure; a

134

value below which yields hydrolysis as a potential rate-limiting step; conversely, when r

exceeded this value, methanogenesis limits the anaerobic digestion process. This study opened a

new way to enhance anaerobic digestion process by manipulating microbial community ratio to

fit for a particular feedstock.

2. Bipolar effects of settling time on active biomass retention in anaerobic sequencing

batch reactors digesting flushing dairy manure

Settling time exerts bipolar effects on active biomass retention in ASBR. Both short and long

settling time were able to retain high concentrations of active microbes, though in disparate

living forms. Biofilm attaching on dairy manure fiber is the major living form of retain

microorganisms at short settling time, while ASBR running at longer settling is dominated with

planktonic bioflocs.

Gravity settling and selection pressure theories were accounted for the results. These two

mechanisms always exist and jointly function together in ASBR, and the only difference is that

one may gradually overtake the other strays from a critical settling velocity, in which the two

effects just trade off each other giving rise to minimum active biomass retention. The effects of

gravity settling gradually overtook those of selection pressure as settling time exceeded a critical

point; below that critical point, selection pressure dominated.

135

3. Psychrophilic anaerobic digester with biofilm supported by solids from flushing dairy

manure

A successful biomass retention technology with fiber material present within the dairy manure as

biofilm carriers was developed in this study. ASBR using manure fiber as support media not only

improved methane production but also reduced the necessary HRT and temperature to achieve a

comparable treating efficiency. This technology expanded the capacity of anaerobic digestion to

dilute solid wastes treatment with no requirement of prior solids separation or the risk of biofilm

support media clogging.

A kinetic model was derived for process control and design by comparing four microbial growth

kinetic models, i.e. first-order, Grau, Monod and Chen & Hashimoto models. Chen & Hashimoto

model was proved capable of characterizing biofilm growth kinetics with an improved

performance, compared to the Monod model, due to dependence on influent substrate

concentration. At optimized conditions, a volumetric methane production rate of 0.24 L/L/d of

and specific methane productivity of 0.19 L/gVSloaded are expected.

4. Methanosarcina dominated in ASBR digester at short HRT

The ASBR at short HRT was capable of establishing a Methanosarcina predominated

methanogenic Archaea population. Reactor performance might be related to microbial

community diversity, the community shows higher diversity when reactor performance is

superior. The comparison study showed that the mcrA clone library detected more Archaea

diversity, especially in Methanomicrobiales, while failing to detect Methanosaeta sp., compared

to 16S rRNA clone library, indicating a combination of two different clone libraries could be

136

able to capture more archaea diversity. The experimentally shown dominance of

Methanosarcina was explained by thermodynamic calculation by assessing different digester

conditions. It was speculated that homoacetogenesis plus aceticlastic methanogenesis are the

main pathways for methane production in ASBR digester.

5. Future work

This technology needs to be applied to even lower temperature (10–15°C) to yield more net

energy in cold region during the winter. Psychrophilic microorganism culture sourced from low

temperature environment could be used as inoculum for psychrophilic digester. Biofilm

formation mechanism at such low temperature needs to be investigated. In addition, quantitative

PCR and high throughput pyrosequencing could be used to better understand the function of

microbial community. The speculated methanogenic pathway needs to be confirmed by methods

such as stable-isotope technique. Further research on the syntrophic relationships between

methanogenic archaea and bacteria are also needed.