Anaerobic Digestion of Biofuel Production Residues

MISSING IMAGE

Material Information

Title:
Anaerobic Digestion of Biofuel Production Residues
Physical Description:
1 online resource (189 p.)
Language:
english
Creator:
Tian,Troy
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Agricultural and Biological Engineering
Committee Chair:
Pullammanappallil, Pratap C
Committee Co-Chair:
Ingram, Lonnie O
Committee Members:
Koopman, Ben L
Correll, Melanie J
Chadik, Paul A

Subjects

Subjects / Keywords:
anaerobic -- biofuel -- methane -- microbial
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre:
Agricultural and Biological Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The production of renewable fuels is expanding globally as the supply of petroleum reserves diminish and concerns regarding environmental impacts rise. Fuels produced from biomass have the potential to reduce reliance on petroleum resources and reduce greenhouse gas emissions. Commercially produced liquid biofuels are biodiesel and ethanol. Obtaining from coproduced residues remains one of the major challenges in full scale operations. This study investigated the potential of using anaerobic digestion to convert three waste residues generated from biodiesel production and cellulosic and non-cellulosic ethanol production. The wastes were glycerol byproduct from biodiesel production, distillery wastewater (stillage) from sugarcane bagasse and hardwood-based cellulosic ethanol production and tailings from sugarbeet-based noncelluloisc ethanol production. Anaerobic digestion is an engineered biochemical process that converts organic matter to biogas (a mixture of methane and carbon dioxide) by the concerted action of syntrophic microbial populations under oxygen free conditions. Biogas can be used as a fuel at the biorefinery displacing fossil fuel use. Biochemical methane potential assays were performed on glycerol byproduct at mesophilic temperature (37 oC). The methane-producing potential was determined to be 450 ml CH4 (ml sample)-1. The produced methane is able to generate 1058 KJ (kg biodiesel )-1, which is sufficient to offset the fuel energy consumption of 495 KJ (kg biodiesel )-1 in a typical small scale biodiesel production, leaving in excess of energy that can be sold to make profit. A semi-continuous anaerobic digestion of glycerol byproduct was also carried out at laboratory scale. It was found that if glycerol byproduct was fed by itself, the degradation rate was only 0.67 ml (day)-1(L) -1. Ethanol production using fermentation process produces a stillage waste from distillation of fermentation broth. Stillage is generated from ethanol distillation that is usually conducted at 60-70oC. To take advantage of these warm temperatures, anaerobic digestion was carried out at thermophilc temperature (55 oC). Stillage obtained from cellulosic ethanol production process using sugarcane bagasse as well as hardwoods, as feedstock was used in these studies. Along with an aqueous fraction, the stillage contained of unconverted ligno-cellulosic material. The biochemical methane producing potential of these two different fractions of the stillage was first determined. It was observed that 70% of the methane-producing potential was in the aqueous filtrate. This meant that only the filtered fraction needs to be digested. The lignin-rich fibrous residue can be used for other applications like making bioproducts. Biochemical methane potential studies were followed by long term digestion of stillage in a continuously fed laboratory scale (13L) anaerobic digester. The digester was operated at HRTs of 21 and 14 days for 90 days. The methane yield averaged 10 ml at STP (ml stillage)-1. The organic matter removal efficiency was about 80%. An energy balance developed for integrating anaerobic digestion with the ethanol production process showed that up to 70% of the energy consumed in steam generation can be displaced using biogas from anaerobic digestion of stillage. Sugar beet is a widely used feedstock for sugar-based non-cellulosic ethanol production in Europe. Tailing waste is generated when raw sugar beets are washed and mainly consists of sugar beet pieces, weeds, sugar beet tops, and other debris and soils. Anaerobic digestion of sugar beet tailings was implemented in a single-stage, batch system. Digestion performance was investigated and compared under mixed and non-mixed conditions. A higher methane yield of 0.35 L CH4 (g VS)-1 was obtained under non-mixed conditions, whereas only 0.25 ml CH4 at STP (g VS)-1 was obtained at mixed condition. The rate of methane production was also higher in the unmixed digester. The microbial community structure was investigated using 16s rDNA analysis for both mixed and non-mixed digesters, revealing marked differences. For example, an abundance of hydrogen-producing bacteria was detected in the mixed digester.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Troy Tian.
Thesis:
Thesis (Ph.D.)--University of Florida, 2011.
Local:
Adviser: Pullammanappallil, Pratap C.
Local:
Co-adviser: Ingram, Lonnie O.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043261:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 ANAEROBIC DIGESTION OF BIOFUEL PRODUCTION RESIDUES By ZHUOLI TIAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Zhuoli Tian

PAGE 3

3 To my parents who have always been supportive of my education

PAGE 4

4 ACKNOWLEDGMENTS I would like to extend my deepest appreciation to my parents, Mr.Tian and Mrs. Zheng for their continuous support and encouragement towards my academic endeavors. I express my sincere gratitude to my adv isor and committee chairman, Dr. Pratap Pullammanappallil, for his c ontinual support and guidance during my time at the University of Florida. I am grateful to Dr. Lonnie O. Ingram for his support throughout this research study. I would like to thank Dr. Ben Koopman and Dr. Paul A. Chadik for serving on my committee and their inspirational lectures. I also own gratitude to Dr. Melanie Correll for her devotion and patience in helping improve my writing skills. I give thanks to Dr. Adriana Giongo and Ms. Diane C hauliac for their guidance and assistance in carrying out microbial anaysis for this re search. I also would like to thank Dr. Claudia C. Geddes and Dr. Mike Mullinnix for providing necessary materials and data for this research, as well as their insightful ideas and assistance s in solving analytical issues. I would like to thank the following indivi duals for their supports and assistances during my studies, Abhay Koppar, Cesar Mo reira Congrong Yu, Gayathri Ram Mohan, Ioannis M. Polematidis, Jaime Chavez Leon, Mandu Iny ang, Patrick Dube, Samriddhi Buxy, Yuan Tian and all the fri ends in the Department of Agricultural and Biolo gical Engineering. I gave my special thanks to Lushun Xiang and Ying Yao for their wholehearted supports and companies to relieve my frus tration and stress. This would never have been possible without them.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4 LIST OF TABLES............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 ABSTRACT................................................................................................................... 11 CHA PTER 1 INTRODUCTION.................................................................................................... 14 Project Foc us.......................................................................................................... 14 Glycerol Byproduct from Biodi esel Production Process .......................................... 15 Residues from Bi oethanol Pr oduct i on..................................................................... 16 Research Ob jectives ............................................................................................... 18 Anaerobic Digestion Technique .............................................................................. 19 The Microbiology of Anaerobic Di gestion ......................................................... 19 Methane Production and COD Degr adation ..................................................... 21 Configurations of Anaerobic Di gesters ............................................................. 22 Factors Affecting An aerobic Digestion ............................................................. 23 Methods for Investigation of Micro bial Communities........................................ 25 Current Technologies for 16S rD NA S equencing............................................. 25 16S rDNA Sequencin g Data An alysis .............................................................. 26 2 BIOCHEMICAL METHANE POTENTIAL OF GLYCEROL BYPRODUCT FROM BIODIESEL PRODUCTION US ING WASTE VEG ETABLE OIL ............................. 31 Summary................................................................................................................ 31 Background ............................................................................................................. 32 Material and Met hod ............................................................................................... 33 Feedsto ck......................................................................................................... 33 BMP A ssay ....................................................................................................... 34 Semi-continouous Dige stion of GBP ................................................................ 35 Analys is............................................................................................................ 36 Results.................................................................................................................... 38 Characterization of the GBP ............................................................................. 38 Methane Pot ential ............................................................................................ 38 Soluble Chemical Ox ygen Demand Profiles ..................................................... 42 Volatile Organic Acid Pr ofiles ........................................................................... 43 Methane Production from Semi-c ontinuous Digesti on of GBP ......................... 43 Discuss ion........................................................................................................ 45 Closing Re marks .................................................................................................... 48

PAGE 6

6 3 ANAEROBIC DIGESTION FOR TREATMENT OF STILLAGE FROM CELLULOSIC BIOETH ANOL PR ODUCTION ........................................................ 55 Summary................................................................................................................ 55 Background ............................................................................................................. 56 Materials and Methods............................................................................................ 58 Feedsto ck......................................................................................................... 58 Biochemcial Methane Po tential Assa y (BMP) .................................................. 59 Continuous Digestion of Bagasse Still age Filtrate ............................................ 60 Monitoring and Analysis ................................................................................... 62 Results.................................................................................................................... 64 Characterization of the Stillage ......................................................................... 64 Methane Pot ential ............................................................................................ 65 Soluble COD and Volatile Or ganic Acids in BMP Assays ................................ 66 Continuous Digestion of Baga sse Stillage Filtrate (BF) .................................... 67 Discuss ion........................................................................................................ 68 Closing Re marks .................................................................................................... 75 4 COMPARISON OF BATCH ANAEROBI C DIGESTION OF SUGARBEET TAILINGS AT MIXED AND NON-MIXED CO NDITIONS ........................................ 88 Summary................................................................................................................ 88 Background ............................................................................................................. 89 Material and Met hod ............................................................................................... 91 Feedsto ck......................................................................................................... 91 Anaerobic Digesters ......................................................................................... 91 Experiment Description .................................................................................... 92 Temperature Monitor ........................................................................................ 93 Chemical Analysis ............................................................................................ 94 Microbial Communi ty Anal ysis .......................................................................... 94 Results.................................................................................................................... 96 Characterization of Subs trates ......................................................................... 96 Temperature Profiles ........................................................................................ 96 CH4 Production................................................................................................. 97 Profiles of sCOD De gradation Co mparison ...................................................... 98 Profiles of Volatile Organi c Acid ....................................................................... 98 Microbial Communiti es Structure ...................................................................... 99 Discuss ion ............................................................................................................ 100 Determination of Mi x ing Intensity................................................................... 100 Digestion Perform ance Com parison............................................................... 107 Microbial Community ...................................................................................... 108 COD Bala nce ................................................................................................. 111 Closing Re marks .................................................................................................. 113 5 MICROBIAL POPULATION IDENTIFI CATION AND DY NAMICS DURING BATCH THERMOPHILIC ANAEROBIC DIGESTION AT MIXED AND NONMIXED CONDI TIONS........................................................................................... 126

PAGE 7

7 Summary.............................................................................................................. 126 Background ........................................................................................................... 127 Method and Materials ........................................................................................... 128 Feedsto ck....................................................................................................... 128 Anaerobic Reactors and Oper ation ................................................................ 128 Chemical Analysis .......................................................................................... 129 Molecular Biologi cal Anal ysis ......................................................................... 129 DNA Extraction and Purifica tion ..................................................................... 130 Polymerase Chain Reac tion Amplif ication ...................................................... 130 Results.................................................................................................................. 132 Characteristics of Feed Subs trates ................................................................ 132 Methane Pr oduction ....................................................................................... 132 Degradation of Organic Ma tte rs..................................................................... 133 Microbial Communi ty Anal ysis ........................................................................ 134 Discuss ion ............................................................................................................ 136 Digestion Perform ance Com parison............................................................... 136 Microbial Community ...................................................................................... 137 Closing Re marks .................................................................................................. 147 6 CONCLUSIONS AND FUTURE WORK ............................................................... 158 Conclusi ons .......................................................................................................... 158 Future Wor k.......................................................................................................... 162 APPENDIX A EMERGY ANALYSIS OF BIODI ESEL PRODCTION FROM WASTE VEGETABLE OIL .................................................................................................. 163 B DEGRADATION OF PROPION IC ACID IN M IXED ANAEROBIC DIGESTION... 164 LIST OF RE FERENCES ............................................................................................. 170 BIOGRAPHICAL SKETCH .......................................................................................... 189

PAGE 8

8 LIST OF TABLES Table page 1-1 Comparison of Anaerobic Digesters ................................................................... 27 2-1 Contents and consti tuents of t he assays ............................................................ 50 2-2 Paramerers of Gompertz Equatio n ..................................................................... 50 3-1 Contents and constit uents of the BMP a ssays ................................................... 77 3-2 Characteristics of bagasse whol e stillage ........................................................... 78 3-3 Monitored parameter s of anaerobic digestion.................................................... 79 4-1 Substrate characteristics and loading amount s for di gester 1 and 2................ 114 4-2 Power input per unit v olume............................................................................. 115 4-3 Bacterial 16S rRNA gene library ....................................................................... 116 4-4 Archaeal 16S rRNA gene library ....................................................................... 117 5-1 PCR amplifications of 16s rRNA genes using 12 barcoded pr imers ................. 149 5-2 Substrate characteristics and loaded quantities for di gesters 1 and 2.............. 150 5-3 Analysis of 16S-rDNA sequenc e at tax on rank of order................................... 151 A-1 Emergy evaluation of biodiese l pr oduc tion....................................................... 163

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Transesterificati on of trig lyceri de ........................................................................ 29 1-2 Ethanol production proc ess from s ugar cane b agasse........................................ 30 2-1 Cumulative methane yield from assay A-1, C-II, 1 and A-3 ................................ 51 2-2 sCOD profile in assays A-1, A-2, C-II, 1 and CII, 2 ............................................ 52 2-3 Propionic acid profile s from assay A-1 and A-2.................................................. 53 2-4 Daily methane yields from se mi-c ontinuous diges tion of GBP........................... 54 3-1 Custom-built tripod sta nd for anaerobic digesters.............................................. 80 3-2 Continuous anaerobi c reactor sch ematic........................................................... 81 3-3 Methane yields of Assay BW-1 B W-2, BF, BR, HW -1 and HW-2...................... 82 3-4 COD and VOA profiles of assay B W-1 and BW-2.............................................. 83 3-5 COD and VOA profiles of assay H W-1 and HW-2.............................................. 84 3-6 Daily methane yield from continuous digestion of st illage .................................. 85 3-7 Profiles of SCOD, VOA, phosphorus and am monia as N ................................... 86 3-8 Mass balance of t he in tegrated process............................................................. 87 4-1 Digester and gas mete r confi guration............................................................... 118 4-2 Inoculum used for ex perimental set 1 and 2..................................................... 119 4-3 Temperature profiles of inc ubator, digester 1 and digester 2 ........................... 120 4-4 Methane yield in digeste r 1 and 2 from all 6 runs ............................................. 121 4-5 SCOD concentration in digester 1 and 2 fro m 6 runs....................................... 122 4-6 VOA concentration in digester 1 and 2 from three runs .................................... 123 4-7 The rank abundances of bacte ria l and archaeal phylo types............................. 124 4-8 Average COD balance of digesters 1 and 2 ..................................................... 125 5-1 Methane yield from digesters 1 a nd 2............................................................... 153

PAGE 10

10 5-2 SOCD and VOA pr ofiles of digester 1 and diges ter 2 ....................................... 154 5-3 Estimate of Chao1 co verage ............................................................................ 155 5-4 Microbial community dy namics of diges ter 1 and 2.......................................... 157 B-1 pH prof ile .......................................................................................................... 168 B-2 VOA prof ile ....................................................................................................... 169

PAGE 11

11 Abstract of Thesis Pres ented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy ANAEROBIC DIGESTION OF BIOFUEL PRODUCTION RESIDUES By Zhuoli Tian August 2011 Chair: Pratap Pullammanappallil Major: Agricultural and Biological Engineering The production of renewable fuels is expanding globally as the supply of petroleum reserves diminish and concerns regarding environmental impacts rise. Fuels produced from biomass have the potential to reduce relianc e on petroleum resources and reduce greenhouse gas emissions. Commerc ially produced liquid biofuels are biodiesel and ethanol. Obtaining from copr oduced residues remains one of the major challenges in full scale operations. This st udy investigated the potential of using anaerobic digestion to convert three wa ste residues generated from biodiesel production and cellulosic and non-cellulosic ethanol production. The wastes were glycerol byproduct from biodiesel production distillery wastewater (stillage) from sugarcane bagasse and hardwood-based cellulosi c ethanol production and tailings from sugarbeet-based noncelluloisc ethanol production. Anaerobic digestion is an engineered biochemical proce ss that converts organic matter to biogas (a mixture of methane and carbon dioxide) by the conc erted action of syntrophic microbial populations under oxygen free conditions. Bi ogas can be used as a fuel at the biorefinery displacing fossil fuel use.

PAGE 12

12 Biochemical methane potential assays we re performed on glycerol byproduct at mesophilic temperature (37 oC). The methane-producing potential was determined to be 450 ml CH4 (ml sample)-1. The produced methane is able to generate 1058 KJ (kg biodiesel )-1, which is sufficient to offset the fuel energy consumption of 495 KJ (kg biodiesel )-1 in a typical small scale biodiesel production, leaving in excess of energy that can be sold to make profit. A se mi-continuous anaerobic digestion of glycerol byproduct was also carried out at laborator y scale. It was found that if glycerol byproduct was fed by itself, the degr adation rate was only 0.67 ml (day)-1(L) -1. Ethanol production using fermentation process produces a stillage waste from distillation of fermentation br oth. Stillage is gener ated from ethanol distillation that is usually conducted at 60-70oC. To take advantage of these warm temperatures, anaerobic digestion was carried out at thermophilc temperature (55 oC). Stillage obtained from cellulosic ethanol production process using sugarcane bagasse as well as hardwoods, as feedstock was used in t hese studies. Along with an aqueous fraction, the stillage contained of unconver ted ligno-cellulosic material. The biochemical methane producing potential of these two different fractions of the stillage was first determined. It was observed that 70% of t he methane-producing potential wa s in the aqueous filtrate. This meant that only the filt ered fraction needs to be digest ed. The lignin-rich fibrous residue can be used for other applications like making bioproducts. Biochemical methane potential studies were followed by long term digestion of stillage in a continuously fed laboratory scale (13L) anaerobic digester. The digester was operated at hydraulic retention times (HRT) of 21 and 14 days for 90 days. The methane yield averaged 10 ml at STP (ml stillage)-1. The organic matter remo val efficiency was about

PAGE 13

13 80%. An energy balance developed for integr ating anaerobic digest ion with the ethanol production process showed that up to 70% of the energy consumed in steam generation can be displaced using biogas from anaerobic digestion of stillage. Sugar beet is a widely used feedstock for sugar-based non-cellulosic ethanol production in Europe. Tailing waste is generated when raw sugar beets are washed and mainly consists of sugar beet pieces, weeds, sugar beet tops, and other debris and soils. Anaerobic digestion of sugar beet tailings was implemented in a single-stage, batch system. Digestion performance was in vestigated and compared under mixed and non-mixed conditions. A higher methane yield of 0.35 L CH4 (g VS)-1 was obtained under non-mixed conditions, whereas only 0.25 ml CH4 at STP (g VS)-1 was obtained at mixed condition. The rate of methane production was also higher in the unmixed digester. The microbial community structur e was investigated using 16s rDNA analysis for both mixed and non-mixed digesters, revealing marked differences. For example, an abundance of hydrogen-producing bacteria was detected in the mixed digester.

PAGE 14

14 CHAPTER 1 INTRODUCTION Project Focus The rapid increase in demand for petroleum, its finite reserves, and concerns regarding it s environmental impac t are driving the search for new energy sources and alternative ways to power the worlds economy An alternative fuel is expected to be technically feasible, economically competitive, environmentally acceptable, and readily available ( Meher, Sagar et al. 2006). Numerous alternative fuels have been proposed, including bioethanol, biodiesel, methanol, hydrogen, natural gas, liquefied petroleum gas (LPG), FischerTropsch fuel, electr icity, solar fuels and algae based fuels ( Balat 2011). Among these, biomass-based fuels or biofuels have several advantages over petroleum with a major one being easy avai lability from locally grown biomas s ( Demirbas 2008). The most commonly produced biofuels include ethanol an d biodiesel because these can be used as a substitute fuel for motor vehicles ( Demirbas 2011). Biofuel production has increased in recent ye ars. This, however, has also resulted in increases in the amount of associated wa stes. For example, biodiesel production generates crude glycerol byproduct and et hanol production generates distillery wastewater (stillage) and fermentation resid ues as wastes, which present significant disposal or treatment problems. The studies presented in this dissertation investigated anaerobic digestion as a viable and sustainable approach for simultaneous waste treatment and energy production. The biogas pr oduced from anaerobic digestion can be used as a fuel on site in the bioref inery to displace fossil fuels. Anaerobic digestions of three different residues from biofuel production process are considered in this dissertation: 1) glycerol byproduct (GBP) from biodiesel

PAGE 15

15 production, 2) stillage from sugarcane-bagas se and hardwood-based cellulosic ethanol production and 3) tailings from sugarbeet based noncelluloisc ethanol production. Glycerol Byproduct from Biodiesel Production Process Biodiesel is typically produced by reacting vegetable oil or animal fat (triglycerides) with methanol or ethanol in the presence of an alkali catalyst. Glycerol is generated during the transesterification proce ss, as shown in the stoichiometric reaction in Figure 1-1 ( Parawira 2009). Transesterification is the process of ex changing the alk -oxy group of an ester compound with another alcohol. The overall process is a seque nce of three consecutive and reversible reactions, in which digl ycerides and monoglycerides are formed as intermediate compounds ( Ma and Hanna 1999). In general, transesterification process produces about 1 kg glycerol per 10 kg bi odiesel. Glycerol byproduct stream may contain impurities like unreacted oil, exc ess alcohol, catalyst and soap. Since the transesterfication reaction is reversible, al cohol is usually added in excess to shift the equilibrium in favor of production of biodiesel ( Meher, Sagar et al. 2006). Excess methanol is distributed between the ester and gl ycerol phases. Methanol is typically recovered even by small-scale biodiesel proc essors from these phases for reuse as reactant. Purified glycerol, or glycerin, is a fairly high value chemical that is primarily used in the manufacture of various foods and beverages, pharmaceuticals, cosmetics, and other personal care products. Thus, a po ssible option to add value to GBP would be to refine it to pure glycerol or small scale biodiesel manufactures can directly sell it. However, as the biodiesel industry has ex panded at a global scale, greatly increased GBP production has impacted the market, c ausing a 10-fold slump in crude glycerol price in recent years and recession among glycerol refineries (Sabourin-Provost and

PAGE 16

16 Hallenbeck 2009). As a result, biodiesel refi ners are faced with limited options for managing GBP, which has essent ially gone from being a potentially valueable product to a waste stream in the biodiesel industry ( Johnson and Taconi 2007). Currently utilize d options for glycerol management incl ude selling it as boiler fuel or as a supplement for animal feed. However, when burned to produce thermal energy, the impurities contained in GBP ma y create a significant amount of ash with potential to cause environmental and health problems, and the presence of water decreases its heating value. When sold as animal feed, t he market value of GBP is too low to be economically attractive ( Hu and Wood 2010). Consequently, it is crucial to develop an alternative process to convert GB P to a value-added product to increase the competiveness of biodiesel industry. Residues from Bioethanol Production Bioethanol can be produced from different ki nds of feedstocks. The materials can be classified into three categories: sucr ose-containing feedstocks (e.g. sugar cane, sugar beet, sweet sorghum and fruits), starch -based materials (e.g corn, wheat, rice, potatoes, cassava and sweet potatoes) and lignoc ellulosic materials (e.g. wood, straw and grasses) ( Balat 2011). Sugar beet provides an abundant sour ce of sucrose which can be easily converted to ethanol by yeas t fermentation, wher eas producing ethanol from maize and wheat requires enzymes to convert starch to sugar first ( Antoni, Zverlov et al. 2007). On a per hectare basis, sugar beet is one of the most productive sources of ethanol and was the second most common feedstock used in European Union for ethanol production in 2009 ( Panella 2010; Takara, Nitayavardahana et al. 2010). Howev er, commercial ethanol production from sugar beet has not been implemented in United States to date ( Outlaw, Ribera et al. 2007).

PAGE 17

17 Sugarbeet processing generates signific ant quantities of both solid and liquid wastes. Raw sugar beets, when brought into the processing plant from storage in outdoor stockpiles, are first washed and separa ted from tailings. The washed beets proceed for further processing and juice ex traction generating ano ther solid waste stream, the spent beet pulp. The tailings main ly consist of 10% to 30% sugar beet chips and 70 to 90% weeds, sugar beet tops, debris, and soil ( Kumar, Rosen et al. 2002). Sugar beet tailings ha ve been anaerobi cally digested to produce biogas (( Weiland 2003; Klocke, Mhnert et al. 2007; Felde 2008). The alternate option to produce ethanol from nonfood biomass resources like forestry and agricultural residues and urban wast es is currently gaining ground. Biofuel produced this way is termed second generation biofuel. Lignocellulosic materials serve as a cheap and abundant feedstock and have the potential to produce up to 442 billion liters of ethanol that could help meet the future demand ( Kim and Dale 2004). Ninety percent of dry matter i n li gnocelluloses consists of cellulose, hemicellulose and lignin with the rest consisting of extractables and ash. Ethanol is produced from cellulose, hemicellulose and other sugar components in the feedstock. ( Balat 2011). The difficulties in bioconversion of lignocellulos ic materials to ethanol are (1) the resistant nature of biomass to breakdown; and (2) i nefficient fermentation of sugars released from degradation of cellulose and hemicellulose by naturally occurring organisms. At the Biofuels Pilot Plant in University of Florida, a recombinant E. Coli KO101 strain developed in the Microbiology and Cell Science Department is employed to ferment both the hexose and pentose sugars to ethanol ( Yomano, York et al. 1998; Yomano, York et al. 2008; Yomano, York et al. 2009). Figure 1-2 shows the flow diagram of

PAGE 18

18 bioethanol production from s ugarcane bagasse conducted in the Biofuel Pilot Plant. The flow quantities were calculated by assumi ng the plant being abl e to produce 1 million gallon of ethanol annually. As shown in Fi gure 1-2, the process generates 12 liters of stillage per 1 liter of ethanol produced. Stillage organic content can exceed 100 g COD/L and exhibits a consi derable pollution potential (Sheehan and Greenfield 1980; Lele, Rajadhyaksha et al. 1989; Yeoh 1997). Wilkie et al. ( Wilkie, Riedesel et al. 2000) conducted an extens ive literatur e review to investigate methods to process and utilize the stillage associated with ethanol production from conventional and cellulosic feedstocks and revealed a consensus toward anaerobic digestion as an economical technology. Research Objectives The overall goal of this research was to investigate the feas ibility and ascertain environmental benefits of depl oying anaerobic digestion for wastes, byproducts and residues generated in biofuel production process. Thre e different types of feedstocks were separately investigat ed. 1) mostly defined liquid feedstock (GBP), 2) fiber containing slurry (stillage), and 3) solid feedstock (tailings) Specific objectives were to 1) determine the biochemical me thane potential of the feedstocks 2) investigate the long term operability of the anaerobic digestion process to operate on these feedstocks by employing appropriate di gester designs at the laboratory scale. Operational parameters monitored included a. quality of the effluent b. biogas production rate c. biogas composition

PAGE 19

19 d. microbial popul ation and dynamics 3) estimate fuel savings when anaerobic digestion is integrated with the biofuel production process. Anaerobic Digestion Technique Anaerobic digestion is an engineered bioche mical process that converts organic matter to biogas (a gas mixture of methane and carbon dioxide) by the concerted action of microbial populations under oxygen-free condi tions. The bioconversion of organic matter into biogas is accomplished by a seri es of interdependent metabolic reactions in which different classes of microorganisms take part ( Alkaya and Demirer 2011). Hydrolysis and acidogenesis reac tions are conducted by bacterial populations that produce hydrogen and organic acids, and the methanogenesis reaction is performed by archaeal group that produce methane using t he products of acidogenesis reactions. Organic matter is converted into biogas as well as new bacterial biomass ( Romano and Zhang 2008). The Microbiology of Anaerobic Digestion The microbiological nature of anaerobi c digestion was discovered more than a century ago ( Koster 1988). While a diverse array of microorganis ms is implicated, anaerobic digestion is mostly driven by bacteria and methanogens. Although the microbiology and chemistry of anaerobic de composition are complicated, it can be conceptualized into three transformation st eps: 1) hydrolysis of complex organic molecules to soluble monomer molecules such as amino acids, fatty acids, glucose and glycerol; 2) fermentation of those solubl e substrates to pr oduce carbon dioxide, hydrogen and organic acids with acetic acid being the main product, with smaller

PAGE 20

20 amounts of propionic and butyric ac ids; 3) conversion of the ac etic acid, hydrogen and a portion of the carbon dioxide to methane. Four categories of microorganisms: 1) hydrolytic bacteria, 2) fermentative acidogenic bacteria, 3) acetogenic bacteria and 4) methanogens are involved in step 1, 2 and 3 respectively. A large number of obligate and facultative anaerobic bacteria carry out the hydrolysis and ferment ation steps. These include Bacteroides Bifidobacterium Clostridium Lactrobacillus and Streptococcus ( Bitton 2005). The Syntrobacter and Syntrophomonas genera are known to include acetoge nic species that convert propionic and butyric acid to acetate, hydrogen and carbon dioxide ( McInerney, Bryant et al. 1981). This group requires low H2 partial pressure for acetate production. There is a synergistic relationship between acet ogenic bacteria and methanogens, in which acetogenic bacteria provide substrates for methanogens and methanogens help to achieve low H2 tension by converting it to CH4. Distinct from the other 3 categories, methanogens belong to a separate domain, Archaea and are only able to use a limited number of substrates t hat include acetate, H2, CO2, formate and methanol. The bioconversion to CH4 by methanogens can be divided into 2 pathways: hydrogenotrophic and aceticlastic met hanogenesis. Hydrogenotrophic methanogens, which belongs to the orders of Methanobacteriales Methanomicrobiales Methanosarcinales and Methanococcales reduce CO2 to methane by using H2. Aceticlastic methanogens produce CH4 and CO2 from acetic acid. These are considered as the dominating methanogens in many anaerobic reactors treating waste water ( Leclerc, Delgenes et al. 2004; Karakashev, Batstone et al. 2005). An estimated 65-70 % of CH4 produced in an anaerobic digester is be lieved to originat e from acetate

PAGE 21

21 ( Mackie and Bryant 1981; Liu.Y. and Whitman 2008). Unlike the diverse hydrogenot rophic methanogens, acetotrophic me thanogens were only identified among the order of Methanosarcinales ( Liu.Y. and Whitman 2008). Methane Production and COD Degradation Anaerobic digestion produces CH4 which is a sparingly soluble gas, and most is degased from solution and can be recovered eas ily for subsequent use. Stoichiometry of methane oxidation shows the chemical oxygen demand (COD) equivalent of methane is 4 g COD (g CH4)-1. At standard temperature and pre ssure (0 C and 1 atm), it corresponds to 0. 35 L of CH4 produced per gram of COD converted to CH4. This directly relates CH4 production and COD removal and provides a way to estimate CH4 production by knowing how much COD has been removed in anaerobic digester. The production of CO2 does not contribute to COD reduction because the carbon in CO2 is in the maximum oxid ation state. The CH4 and CO2 content of biogas varies depending on the nature of the subs trate with CH4 content ranging between 50 to 70% ( Parkin and Owen 1986). Anaerob ic digesters are typically utilized to stabilize organic matter in wastes with biodegradable COD concentrati on greater than 1 g/L. Compared to other waste treatment methods, anaerobic processes offer advantages including less solid production, lower nutrient requirement, lo wer energy consumption and production of a potentially useful product of CH4 ( Sahm 1984; Lettinga, Field et al. 1997). The first step in developing an anaerobic di gestion project involves determining the methane potential of the feedstock using the biochemical methane potential (BMP) assay. BMP assay was conducted by mixing the substrate with inoculum and nutrients, and incubating the mixture in a sealed bottl e at predetermined temperature for a period

PAGE 22

22 of time to allow complete degr adation of the substrate. In th is research the BMP of each of the feedstock was first determined. Configurations of Anaerobic Digesters A typical anaerobic digester contains four components a closed vessel, a heating system, a mixing system and a gas -liquid-solid s eparation system ( Grady, Daigger et al. 1999), but mixi ng is not always required. A closed vessel is used to exclude dissolved oxygen and ensure the dev elopment of anaerobic conditions. Heating is usually required in anaerobic digestion to maintain the optimal temperature for microbial activities, typica lly ranging from mesophilic (around 37 C) to thermophilic (around 55 C) temperatures. In many case s, mixing is provided to improve the homogeneity of the digester contents and reduce the resistance to mass and heat transfer. Several methods are used to accomp lish mixing, including use of mechanical mixers, slurry recirculation or biogas recirculation ( Karim, Hoffmann et al. 2005). Many anaerobic digesters are cylindric al concre te tanks with a cone shaped bottom and steel or concrete covers, though ot her materials and configurations can be used. In general, there are seven types of anaerobic digesters: 1) continuous stirred tank reactor (CSTR), 2) anaerobic contactor, 3) upflow anaerobic sludge blanket (U ASB), 4) anaerobic filter (AF), 5) hybrid UASB and anaerobic filter (U ASB/AF), 6) downflow st ationary fixed film (DSFF) and 7) fluidized bed (FB) digesters ( Grady, Daigger et al. 1999). Among these, a CSTR re actor is well mixed with no solid and liquid separati on and its hydraulic retention time (HRT) and solids ret ention time (SRT) are identical (Malina 1992) The rest of reactor configurations provide signific ant retention of active biomass, resulting in large differences between SRT and HRT ( Hall 1992; Defour, Derycke et al. 1994). Long SRTs maintain accumulation of active biom ass and allows higher organic load ing rate.

PAGE 23

23 Therefore, reactor types 2 to 7 are comm only considered as high rate anaerobic processes. Table 1-1 summarizes the prim ary benefits and drawback of different types of anaerobic digesters. All high rate anaerobic processes share certain properties. High biomass concentration are maintained in the digester and thereby allow long SRTs to be achieved while keep HRT relatively s hort. High biomass concentrations makes high organic loading rate possible, resulting in sm all reactor volumes. Choice of digester design is influenced by the solids content of waste. Based on the characteristics of feedstocks studied here, a downflow stationary fixed film digester was used to digest GBP, a continuous mixed digester was used for stillage, and batch high solids mixed and unmixed digesters were used for tailings. Factors Affecting Anaerobic Digestion Anaerobic digestion is mainly affected by temperature, pH, retention time, chemical composition of substrat es and the presence of toxins ( Bitton 2005). Mesophilic anaerobic digestion operates at temperatures from 25 C to 40 C with an optimum at 37 C, while thermophilic anaerobic digestion oper ates at temperatures ranges of 50 C to 60 C. Due to the higher temperature, t he thermophilic condition provides a faster digestion rate and allows for a higher subs trate loading rate. High temperature also leads to greater inactivation of pathogens ( Koster 1988). Most methanogens function optimally at a pH of 6.7 to 7.5 ( Bitton 2005). The acidic environment caused by the organic acidogenic bacteria could lead to failure of methanogens is. Under normal conditions, the pH reduction is buffered by bicarbonate that is added externally or produced by methanogens. However, the buffering capacity can be upset, especially when organic overload ing happens, stopping CH4 production. An increase in volatile organic acid concentration serves as an early indicator of system upset and monitoring

PAGE 24

24 of volatile organic acid level is suggested to ensure system stability. HRT is the average time spent by anaerobic microorganisms in the anaerobic digester. HRT must be long enough to allow metabolism of those micr oorganisms. The usable substrates for anaerobic microbial communities include carbohy drates, proteins, lipids etc., but a few compounds such as lignin ar e non-degradable for anaerobic microorganism. Providing nutritionally balanced substrates plays a role in maintaining an adequate anaerobic digestion. The carbon to nitrogen to phosphorus ration for anaerobic bacteria is reported to be 700 to 5 to 1 ( Sahm 1984). Sulfur is also required by methanogens, but it is toxic at levels exceeding 150-200 mg/L( Speece 1983). Trace elem ent as iron, nickel and cobalt are also neces sary for anaerobic mi croorganisms to grow. The presence of toxicants is responsible for the occasional fa ilure of anaerobic digestion. Inhibition of methanogenisis is usually accompanied by reduced methane production and accumulation of volatile organic acids. So me of the common inhibitors are oxygen, ammonia and volatile organic acids. Early studies indicate that trac e level of oxygen can adversely affect methanogens which are obligate anaerobes ( Roberton and Wolfe 1970), but later findings tend to agree that methanogens can tolerate oxygen due to the protection of the sludge aggregates ( Kato, Field et al. 1993). Unionized ammonia is toxic to methanogens and the inhibitory level is round 1500 to 3000 mg/L However, since the formation of unionized ammonia is pH dependent, little toxic ity is observed at neutral pH. Volatile organic acids such as acetic acid and butyric acid are also toxic to methanogens and propionic acid display toxi city to both acidogenic bacteria and methanogens. The toxicity of volatile acids is minimal if the pH is maintained near neutrality ( Bitton 2005).

PAGE 25

25 Methods for Investigation of Microbial Communities During the past decade, progress had been made in the development of molecular techniques for determining the numbers and activity of microorganism in anaerobic digestion. These technologies in clude 16S rRNA gene fingerprinting and sequencing, and fluorescent in situ hybridization ( Manes, West et al. 2011). In this thesis, 16s rRNA gene sequence analys is was employed to investigate the microbial community structure and divers ity. There are 16S rRNA ta rgeted oligonucelotide probes for bacteria or archaea in general as well as for individual species in particular. Current Technologies for 16S rDNA Sequencing There are several options of sequencing technology. Sanger sequencing, als o known as capillary sequencing, produces long reads (up to 800 bases) which are helpful characterizing gene functions and taxonomic composition of microbial communities ( Wommack, Bhavsar et al. 2008). Barcoded pyrosequencing, an innovativ e sequencing technology initially developed in the 1980s, makes microbial community study orders of magnitude more efficient by being able to generate a much greater number of sequenc es in a single run ( Shoemaker, Lashkari et al. 1996; Hamady, Walker et al. 2008). It also eliminates the laborious step of preparing clone libraries ( Ronaghi, Karamohamed et al. 1996) Barcoded pyrosequencing uses molecular barcode techniques to add a uni que tag to each primer before PCR amplificatio n ( Binladen, Gilbert et al. 2007; Parameswaran, Jalili et al. 2007). Because each sample is tagged with a known primer an equimolar mixture of samples can be amplified and sequenced and sequences can be assigned to samples based on the unique barcodes. Though shorter sequences are produced in pyrosequencing, the benefit of a large number of short reads outweighs the dr awbacks of short read lengths

PAGE 26

26 for many kinds of rRNA-based community analysis: sequence fragm ents as short as 100 bases, covering only 8% length of fulllength 16S rRNA gene, provide results with comparable resolution to 70% as t hose obtained using full-length sequences ( Liu, Lozupone et al. 2007). Given that S anger sequencing is more expensive than pyrosequencing, requires cloning of DNA frag ments into a common host cell (typically E. coli ), and has remarkably lower output, the latter is a more cost-effective option for investigating the distribution of microbial diversity among samples at this point. 16S rDNA Sequencing Data Analysis In general, analyses of microbial diversit y can be divided in three dimensions ( Hamady and Knight 2009). First, an analysis c an examine either alpha div ersity (how many kinds of taxa or lineages are in one sa mple) or beta diversity (how taxa or lineages are shared among samp les, e.g., along a gradient). Second, an analysis can be either qualitative, examin ing only presence-absence, or quantitative, also taking into account relative abundance (qualitative analyses and quantitative analyses are also called analyses of community membership and community structure, respectively). Third, an analysis can be either phylogenetic, building a phylogenetic tree to relate the sequences, or taxon based, treating all taxa at a given rank (e.g., orders rank by aboudance). Taxon-based and phylogenetic methods provide different but equally useful insights. In this study, a taxonbased analysis was performed, in which query sequences was matched by similarity to an existing sequence in the 16S rRNA database by BLAST algorithm.

PAGE 27

27 Table 1-1. Comparison of Anaerobic Digesters Process Benefit Drawbacks CSTR Suitable for a wide range of wastewaters Large bioreactor volume required Easy to mix Effluent quality can be poor if large concentration of anaerobic organisms is generated Efficiently handles high suspended solid wastewater Poor performance at short SRT Large bioreactor volume to dilute inhibitors Requires separate mechanical mixing AC Performance not dependent on sludge settleability Biomass settleability critical to digestion performance Suitable for concentrated wastewaters Only suitable for waste with low to moderate levels of suspended solids Easy to mix Mechanically complex system Relatively high effluent quality achievable Reduced bioreactor volume compared to CSTR UASB High biomass concentrations and long SRT achievable Performance dependent on development of dense, settable solids Small bioreactor volumes Little process control possible High quality effluent achievable Low process loading required if wastewater contains suspended solids Mechanically simple Compact system AF High biomass concentrations and long SRT achievable Suspended solids accumulation may negatively impact digestion performance Small bioreactor volumes Not suitable for high suspended solids wastewater High quality effluent achievable Little process control possible Mechanically simple High cost for media and support Compact system Performance not dependent on development of dense, settleable solids

PAGE 28

28 Table 1-1. Continued. Process Benefit Drawbacks Hybrid UASB/AF High biomass concentrations and long SRT achievable Lower process loading required if wastewater contains suspended solids Small bioreactor volumes Little process control possible High quality effluent achievable Mechanically simple Compact system Performance partially dependent on development of dense, settleable solids Reduced media cost DSFF High biomass concentrations and long SRT achievable Biodegradable suspended solids not generally degraded Small bioreactor volumes High cost for media and support High quality effluent achievable Organic removal rate generally lower than other high rate processes Mechanically simple Little process control possible Compact system Performance not significantly impacted by wastewaters with suspended solids Performance not dependent on development of dense, settleable solids FB High biomass concentrations and long SRT achievable Long startup period required Small bioreactor volumes High power requirement for bed fluidization and expansion Excellent mass transfer characteristics Not suitable for high suspended solids wastewater High quality effluent achievable Mechanically complex system Most compact of all high rate processes High cost for carrier media Performance not dependent on development of dense, settleable solids Adapted from Grady ( Grady, Daigger et al. 1999)

PAGE 29

29 Figure 1-1. Transesterification of triglycerid

PAGE 30

30 Figure 1-2. Ethanol producton process from sugarcane bagasse

PAGE 31

31 CHAPTER 2 BIOCHEMICAL METHANE POTENTIAL OF GLYCEROL BYPRODUCT FROM BIODIESEL PRODUCTION USING WASTE VEGETABLE OIL Summary Biochem ical Methane Potential (BMP) assays were performed at mesophilic temperature on the crude glycero l byproduct (GBP) obtained from the settling phase of biodiesel production unit. GBP contained gl ycerol, methanol, oil residues and some water. The average methane yield of GBP was 455 ml CH4 at STP / ml GBP. This yield was higher than that obtained fr om pure glycerol, indicating that in addition to glycerol, GBP contained other compounds from which methane was pr oduced. However, the duration of digestion to achieve this yield was long, requiring 100 days of incubation and the amount of glycerol loaded into the assa y had an effect on methane production. The average methane yield of GBP at 2 ml loadi ng after the same incubation period was significantly lower. It was determined that the lower production rate at higher loading was not due to the presence of inhibitory compounds in GBP but due to an organic overloading exceeding the gl ycerol-degradation capacity of the inoculum. Volatile organic acid (VOA) accumulated to much hi gher concentrations in overloaded assays, with propionic acid being dominant. Pr esence of a co-substrate with GBP showed improved methane production. Continuous diges tion studies were then carried out over 40 days.

PAGE 32

32 Background Biodiesel refers to monoalky l transesterfi ed ester of fatty acid from vegetable oil or animal fat. It is a non-toxic and a renewable diesel fuel which can be used neat or as a blend with petroleum diesel fuel. Compar ed to fossil fuel, biodiesel holds many advantages like lower carbon monoxide and particulate emissions and higher cetane number ( Alptekin and Canakci 2008). Due to conc erns of global warming and increases in the price of petroleum in recent years, there is considerable inter est in renewable fuels. Biodiesel production is expandi ng throughout Asia, Europe and America. Several large scale plants, about 120 in European Union and 165 plants in USA, annually produce 14 billion liters of biodiesel combined ( Alptekin and Canakci 2008). On the other hand, the relative ease in making biodiesel als o encouraged the growth of small scale production facilities, which does not require large investment and complex operation but is able to make sufficient fuel for self-use in households, farms or in communities. Biodiesel production process involves r eacting vegetable oil with methanol in the presence of a catalyst like potassium hydr oxide. The process also generates a crude glycerol byproduct (GBP). GBP contains unr eacted vegetable oil, methanol, catalyst and other side products like soap and requires pur ification before it can be marketed. Only large scale plants are able to economic ally refine the crude glycerol. Alternate options need to be considered for small scale producers. GBP may be anaerobically digested to bi ogas consisting of 60-70% methane. Biogas is a good energy source, and can be us ed in the production process to reduce the fossil energy inputs. Large scale plants can also integrate the anaerobic digestion

PAGE 33

33 process with the biodiesel production facility to improve overall energy efficiency and reduce carbon footprin t of the process. For feeding GBP to anaerobic digesters, it is necessary to evaluate the characteristics of its degradation, namely degradability, methane produc tion potential, degradation rate and inhibitory effect. Anaerobi c digestion of glycerol has been reported in the literature. In th ese studies, either analytical grade pure glycerol ( Holm-Nielsen, Lomborg et al. 2008) or refined glycerol from biodies el production ( Chen, Romano et al. 2008) was co-digested with manure. The fo cus of those studies was to improve volumetric methane productivity of manur e di gesters by adding glycerol as a supplement to manure digesters In this study, crude GBP as obtained from the settling tank of a small-scale biodiese l unit at the University of Florida was used as a sole substrate in biochemical methane potential assa ys. In addition to the methane potential of the substrate, the assay also provided information on the rate of degradation, extent of degradation and the presence of any inhibitory compounds or toxins in the substrate. These batch assays were performed at mesophilic (37 C) temperature. The feasibility of using a one-stage, continuous digester for anaerobic digestion of GBP was also investigated. Material and Method Feedstock Crude GB P drained from the fatty acid meth yl ester settling tank of a small scale biodiesel facility at University of Florida ca mpus was used as feedstock. This facility processes waste cooking oil in a batch reac tor using methanol as reactant and KOH as catalyst.

PAGE 34

34 BMP Assay BMP assays measure the meth ane potentia l of feedstocks at optimal conditions for culturing anaerobic digester microbial consortia ( Owen, Stuckey et al. 1979; Shelton and Tiedje 1984; Owens and Chynoweth 1993). It was conducted by mixing the substrate with inoculum and nutri ents, and inc ubating the mixture in a sealed bottle at predetermined temperature for a period of time to allow complete degradation of the substrate. The entire set of assays consisted of 18 bottles, and each assay was conducted in 250 ml serum vials at mesophilic temperature of 37oC. Assays contained substrate (GBP or pure glycerol), inoc ulum and other supportive solutions (micronutrient, macronutrients, co-substrate and pH buffer liquid) with total volume of 154 ml. The inoculum used was collected fr om a laboratory scale anaerobic digester that had been digesting various biomass feedstock and wastewater at mesophilic temperature for three mont hs. Nutrients stock solution was prepared according to Owens et al. ( Owens and Chynoweth 1993). The amount of the substrate added was varied and in some assays a co-substrate was added. The reason being crude GBP may contain unreacted or partially reacted oil and it has been shown that lipids and long chain fatty acids are degraded more efficient ly in the presence of co-substrates especially carbohydrates ( Kuang, Lepesteur et al. 2002). The co-substrate used in the study was a water extract of sugar beet taili ng (tailing water extract). Sugar beet tailing contains a large fraction of readily solubl e or ganic matter and these are extracted easily upon addition of water. The organic compone nts of this extract are primarily carbohydrates and have been shown to degr ade readily in an anaerobic digester ( Liu, Pullammanappallil et al. 2008). H owever, t he amount of extract added was in small quantities ensuring the extract did not contri bute significantly to methane production.

PAGE 35

35 Table 2-1 lists the assays along with its contents and quantit ies loaded. Assays contained 1 ml or 2ml of GBP. 50 ml of tailings water ex tract was also added in an assay containing 2 ml GBP. Each assay was duplicated. Three types of control assays each in duplicate, were also analyzed. The first type of control assay contained only inoculum (C-I), the second type contained 1 ml (C-II, 1) or 2 ml (C-II, 2) of analytical grade glycerol (99.5%) with inoculum and t he third (C-III) type contained 50 ml of tailings water extract with inoculum. The first type of control assays was set up to provide data on methane production from residual substrates and endogenous metabolism in inoculum. Data using second type of control assay provided information on the nature of degradation of glycerol, and when compared to data from GBP assays was able to provide information on presence of inhibitory compounds in GBP and potential of excess methane production from GBP. The third type of control assay provided information on methane potential of tailings extract. Semi-continouous Digestion of GBP The feasibility of anaerobic digestion of GBP was further evaluated in a laboratory scale 5 L semi-continuously fed st irred anaerobic digester. The digester was constructed by modifying a Pyrex glass ja r. The height and inner diameter of the digesters were 0.406 m and 0.061 m, respecti vely. A glass flange fitted with a rubber O-ring was used to seal the top of the di gester. The flange was clamped using stainless steel clamps for gas and liquid tightness. The digester was provided with several ports for gas outl et, sample withdrawal and digeste r liquor circulation. Gas production from the digesters was measured by a positive disp lacement gas meter which consisted of a clear PVC Utube filled with anti-freeze solution, a solid stat e time delay relay (Dayton Off Delay Model 6X153E), a Grainger float switch, a Redington counter and Fabco Air

PAGE 36

36 solenoid valve. More details of the digester and gasmeter can be found in Koppar and Pullammanappallil ( Koppar and Pullammanappallil 2008). The gas meter was calibrated in line to determine the volume of biogas per count. An incubator was used to provide 37 C environment for the digester. The digest er was first inoculated with 3 L inoculum and added with 100 grams of sugar beet tailings to activate the inoculum. The digester was allowed to stabilize for 2 days before the feeding wa s started. GBP was added at 2 ml per day from day 3 to day 18 and increased to 3 ml per day from day 18 to day 40. Analysis Methane production, chemical oxygen demand and volatile organic acid concentration were monitored in the BMP a ssays. Biogas that accumulated in the serum vials was withdrawn period ically using a syringe. Biogas was allowed to fill the syringe until the pressure in vial head s pace equilibrated with that in the syringe. Methane and carbon dioxide composition was m easured using a Fisher Gas Partitioner (model 1200). The gas chromatograph was calibrated with an external standard containing N2, CH4, and CO2 in 25:45:30 volume ratio. Liquid samples were withdrawn from the serum bottles fo r soluble Chemical Oxygen Demand (SCOD) and Volatile Organic Acid (VOA) measurement. Liquid samp les were centrifuged at 8000 RPM for 10 minutes (Fisher Marathon micro H centrifuge), filtered using Mil lipore filter paper (pore size 1.2 um) for SCOD and VOA analys is. SCOD liquid samples were added in Hach COD vials (range 20-150 ppm) and baked in Hach COD reactor for 2 hours. sCOD was measured using Hach DR/890 Co lorimeter. VOA liquid samples were measured using Shimadzu gas chroma tograph (GC-9AM equipped with a flame

PAGE 37

37 ionization detector) for acetic, propionic, isobutyric, butyric, isovaleric and valeric acid concentrations. Daily methane production and pH was m easured for the semi-continuous digester. Biogas volume was measured by th e displacement gas meter, and methane content of biogas was analyzed using t he same gas chromatograph as described above. The pH of digester liquor was measured with an Accumet pH meter. The performance of the batch biochemical methane potential assays was evaluated by fitting the cumulative methane production data to the modified Gompertz equation ( Lay, Li et al. 1998). The modified e quation describes cumulative methane production from batch digesters assuming t hat methane production is a function of bacterial growth and is presented below: 1 expexp t P eR PMm (2.1) Where M is the cumulative methane production, L at any time t, P is the methane yield potential, L, Rm is the specific maximum methane production rate, L d-1, is the duration of lag phase, d, and t is the time at which cumulative methane production M is calculated, d. The parameters P, and Rm were estimated for each of the 15 data sets by using the Solver feat ure in MS-Excel. The value of parameters which minimized the sum of the square of errors between fit and experimental data were determined. The model as applied to the anaerobic digestion process assumes that the rate limiting step is me thanogenesis and then the parameter Rm is related to the specific growth rate of the methanogens ( ) through the product yield coefficient, YXP (g biomass /L methane STP), i.e. Rm = / YXP

PAGE 38

38 Results Characterization of the GBP The densit y of GBP and pure glycerol (99. 5% grade) was determined to be 1.16 g(ml)-1 and 1.20 g(ml)-1, respectively. GBP was also determined to lose 29.6% weight after it had been dried at 104C for 24 hour s. The lost mass was considered as methanol. In biodiesel production, glycerol phase was reported to contain 62.9% to 76.6% glycerol, based on weight, depending on what feedstock was used. The rest in glycerol phase usually contained methanol, cata lyst and oil residues. Specially, for the situation that waste vegetabl e oil (WVO) was used, methanol content in glycerol phase (GBP) was report ed to be 21.6% ( Thompson and He 2006). The reported methanol content fell into a close range as the methanol content determined by this study. Using this as a reference, GBP was as sumed to contain 29.6% methanol and the rest 70.6% are glycerol. Methane Potential The profiles of the average cumulative methane yield of assay A-1 (containing 1 ml GBP), A -3 (containing 2 ml GBP) and CII, 1 (containing 1 ml pure glycerol) are shown in Figure 2-1. Sample standard deviation is shown as error bars. The methane potential of GBP and pure glycerol over 100 days of incubation was 456 ml CH4 at STP (ml sample)-1 and 372 ml CH4 at STP (ml sample)-1 from assays A-1 and C-II, 1, respectively. The degradation rate of GBP a nd glycerol indicated by the slope of curve, initially increased, reached a maximum and then decreased to a minimum as substrate was utilized. The rates of evolution of me thane were similar for the first 40 days. Methane production continued in a ssay A-1 whereas it tailed o ff in C-II, 1. Eighty five percent of ultimate methane yield of G BP was produced in 60 days. The higher

PAGE 39

39 methane yield from GBP could be because GBP contained a mixture of substrates such as long chain fatty acids, residual oil and me thanol that are better energy sources than glycerol itself. Assay A-3 contained 2 ml GBP and the methane potential was determined to be 832 ml CH4 at STP (2 ml sample)-1 after 200 days. It took twice as long as assay A-1 did to achieve the ultimate methane yield. The Gompertz equation was used to fit the cumulative methane production data for assay A-1, C-II, 1 and A-3. Equation parameters, P, and Rm were calculated and their values were provided in table 2-2. The simulation results showed a bi-phase tr end (fraction 1 and fraction 2 in Figure 21) for assays A-1 and A-3, i.e., the slope initially increased then decreased by day 20 (both assays) but then t he methane production began to increase again from day 30 (day 40 for assay A-3) t hen decrease until methane production completely ceased. Fraction 1 of assays A-1 and A-3 completed methane production in a short period of time, suggesting its easy degradability. Intere stingly, both methane yield and methane production rate of fraction 1 doub led in assay A-3 that contained twice as much GBP as in assay A-1. This indicated the prolonged di gestion time seen in assay A-3 was not due to the doubled amount of fraction 1. In contra st, fraction 2 of assay A-3 shared similar methane production rate to assay A-1 and took additional 100 days to produce a comparable methane yield. It appeared fracti on 2 was the limiting factor in GBP digestion from assay A-3 and caused overl oading issue as its quantity increased. The cumulative methane plot from assay C-II, 1 did not show the bi-phasic trend. As digestion of any pure substrate the met hane production rate increases at first then

PAGE 40

40 decreases as substrate is consumed until it completely ceases when all substrate is used up. Fraction 2 in assays A-1 and A-3 had very similar methane production profiles to assay, C-II, 1. Therefore, it was hypothesized that frac tion 2 of GBP is glycerol, and fraction 1 is methanol. Table 2-2 indicated 1 ml (1.2 g) pure glycerol capable of producing 394 ml CH4; while glycerol frac tion of 1 ml (1.16g) GBP capable of producing 320 ml and 280 ml CH4 from a ssay A-1 and A-3, respectively. This suggested that glycerol content in GBP aver aged 77.1%, based on weight. With the fact that methanol and glycerol fraction m ade 32% and 68% of tota l methane production, respectively, methanol content was calcul ated to be 28.1% based on weight. The two calculated values generally agreed the assumption m ade about GBP composition. The methane yield of assay A-2 (containing 2 ml GBP) over 100 days of incubation was considerably less than that of assay A1. The reason for the low yield was doubted to toxicity of GBP. Since volume of GBP doubled, the amount of toxic compound in assay, if any, could have increased and inhibited methane production. However, methane yield of assay C-II, 2 (containing 2 ml pure glycero l) was also much less than the yield of assay C-II, 1 (containing 1 ml pure glycerol). Therefore, it appeared that the low methanogenic activity in assay A-2 was due to organic overloading instead of the presence of inhibitory compounds. Low me thanogenisis caused by organic overload in anaerobic digestion had been reported in some studies ( Amon, Amon et al. 2006; HolmNielsen, Lomborg et al 2008; Ruiz, Blazquez et al. 2009). No tably, in Holm-Nielsen et al. ( Holm-Nielsen, Lomborg et al. 2008), organi c overloading was reported to have occurred in anaerobic digesters when gl ycer ol concentration exceed 5 g L-1 7 g L-1.

PAGE 41

41 Assay A-2 and C-II, 2 contained 2 ml GBP and 2 ml glycerol respectively dissolved in 150 ml solution. The corresponding concentration was 12 g L-1 GBP for assay A-2 and 16.8 g L-1 glycerol for assay C-II, 2. These si gnificantly exceed the suggested loading amount resulting in organic overloading. Assay A-3 (containing 2 ml GBP and co-substrate) exhibited higher methane production and the methane yield over 200 days incubation was 415 ml CH4 at STP (ml sample)-1. The methane yield obtained from a ssay A-1 and A-3 was consistent and verified the methane potential of GBP bei ng around 400 ml CH4 at STP (ml GBP)-1. However, assay A-3 took twice as long to r each the yield as assay A-1. This suggested assay A-3 experienced organic overloading as well which possibly caused the lag phase. Assay C-III indicated the methane yiel d of tailing extract was very small (data now shown). This excluded the possibility that the extra methane production of assay A3 compared to assay A-2 was from the tailing extract. If longer period of incubation was allowed, assay A-2 and C-II, 2 might have overcome the lag phase and have produced comparable methane yields to those of assay A-1 and A-3. This was not shown in the study presented here, but im proved methane production had been shown in assay A-3 with carbohydrate containing co-substrat e. Similar results had been reported in anaerobic digestion with co-substrate of oleate in inhibited digesters ( Kuang, Lepesteur et al. 2002; Kuang, Pullammanappallil et al. 2006). The anaerobic co-digestio n of different organic compounds is a common pr actice to improve the performance of anaerobic digesters ( Bolzonella, Pavan et al. 2006; Siles, Martin et al. 2010). For this reason, GBP was co-digested wit h tailings extract. The improved methane production from assay A-3 may be that the adding tailings water ex tract to the sample balanced

PAGE 42

42 nutrients ratio (e.g., car bon to nitrogen ratio) and prov ided readily degradable organic matters to compensate for low degradability and stimulate activity of microorganisms and degradation of GBP so t hat the organic overloading pr oblem could be overcome faster ( Kuang, Lepesteur et al. 2002; Kim, Han et al. 2003; Krupp, Schubert et al. 2005; Li, Chen et al. 2009). Soluble Chemical Ox y gen Demand Profiles sCOD was measured periodically in the assa ys. The sCOD profiles in assay A-1 is shown in Figure 2-1A. Along with measure sCOD values, t he expected theoretical COD values is also plotted. The expected values were calculated using stoichiometry that 1 g L-1 of COD being capable of producing 0.35 L CH4 at STP. Using the conversion, it is possible to predict the rema ining sCOD concentration in the assays by knowing how much methane has been produced. Through comparing the expected sCOD and measured sCOD, consistency of the results can be verified. T he initial concentration of sCOD for calculation of expected sCOD was assumed same as the measured COD concentration. As seen in Figure 2-2, t he initial sCOD in A-1 was 8,600 mg L-1. This value dropped to 3900 mg L-1 on day 45. The expected values showed a faster initial drop compared to the measur ed and thereafter followed the measured trend closely until day 45. Initially in addition to sCOD some insoluble matters may have also solubilized and converted to methane, yi elding more methane than predicted by dropping measured sCOD. The same method wa s used to derive the expected sCOD profile for assay A-2. Both expected and measured sCOD profiles showed that sCOD concentration in assay A-2 maintai ned at the range from 12,000 mg L-1 to 14,000 mg L-1 for the duration of day 0 to day 100 (Figure 2-2C). Non-degrading sCOD explained the low methane production in assay A-2. Measured sCOD data after day 100 was not

PAGE 43

43 available for assay A-2. Expected and m easured sCOD values were compared for assay C-II,1 and C-II,2 as well as shown in Figure 2-2B and 2-2D. sCOD concentration in assay C-II, 1 degraded as methane produc ed but barely changed in assay C-II, 2 where methane production was low. Volatile Organic Acid Profiles Concentration of VOAs, i.e., acetic acid propionic acid, isobutyric acid, butyric acid, isovaleric acid and valeric acid wa s measured in assays A-1 and A-2. Propionic acid was found to be the most important cont ributor to total concentration of V OAs. Acetic acid was the second mo st important and its concentra tion was about half of that of propionic acid. Concentration of all other VOAs was negligible. Pr ofiles of propionic acid concentration in assay A-1 and A-2 are shown in Figure 2-3. Propionic acid concentration in assay A-2 reached 1200 mg L-1 and was not degraded up to day 80 whereas its concentration in assay A-1 started to decrease after peaking at 500 mg L-1. Accumulation of higher VOAs concentration in assay A-2 could be responsible for the low met hane production. Amon et al. ( Amon, Amon et al. 2006) reported that when VO As exceeded 5000 mg L-1, the anaerobic digestion process was no longer stable and overloading was likely. Pr opionic acid concentration in assays A-2 was only 1200 mg L-1 and less than the reported possible inhibitory threshold. This probably indicated that a sma ll system as used in this study had low tolerance on VOA concentration. Methane Production from Semi-continuous Digestion of GBP Semi-continuous digestion was conducted by manually feeding GBP per day to a mesophilic, one-stage lab scale digester. T he digestion was start ed with 100 grams of sugar beet tailin gs and GBP feeding began at 0.67 ml (day)-1(L) -1 from day 3. Due to

PAGE 44

44 that the digester was initially provided wi th highly degradable substrates, daily methane production increased substantiall y and started decreasing after 5 days as the substrates were quickly consumed. The daily methane pr oduction stabilized at 450 ml CH4 at STP (ml substrate)-1 after 11 days. This yield was very similar to the GBP methane yield obtained in BMP assay A-1. G BP loading rate was further increased to 1 0.67 ml (day)1(L) -1 from day 18. The daily methane produ ction dropped immediat ely and stabilized at 0.25 ml CH4 at STP (ml substrate)-1 after 7 days. Profiles of daily methane production are shown in Figure 2-4. It appeared that the digester performance wa s inhibited when GBP loading rate increased from 0. 67 to 1 ml (day)-1(L) -1. A study ( Chen, Romano et al. 2008) reported a similar result. In their study, a mixture (84% manure and 16% glycerol) was fed to a continuous digester of 18 L. The digestion in hibition was observed when the substrate loading rate increased from 42 g day-1 to 82 g day-1. The corresponding glycerol loading rates were calculated to be 0. 29 and 0. 57 ml (day)-1(L)-1, respectively. Sugar beet tailings were considered as one of the nitrogen sources because they contained 7.5% crude protein on a Volatile Solid basis ( Liu, Pullammanappallil et al. 2008). Nitrogen may also come from digester inoculums that were reported to contain 145 mg L-1 total NH4 as N (Koppar and Pullammanappalli l 2008). Using the assumptions that GBP contai ns 70.6% glycerol and 21.6% methanol on a weight basis the average C/N ratio of semi-continuous GBP digestion over 40 days was calculated to be 58. Sinc e N sources were not provi ded continuously, real-time C/N ratio was expected to start with a much lower value an d slowly increase as N depleted and more C was fed.

PAGE 45

45 Discussion The theoretical methane yield of glycero l at 99.5% grade is 534 ml CH4 at ST P (ml glycerol)-1. The practical methane yield of the glycerol is 372 ml CH4 at STP (ml glycerol)-1 as seen in Figure 2-1 for assay C-II, 1. Low COD and VOA concentration found in assay C-II, 1 at the end of incubation indicated most of the organic contents had been degraded and the differ ence between the theoretical and practical yield was not due to incomplete substrate degradation. The practical yield of 372 ml recovers 69.6% of the theoretical yi eld, probably suggesting the c onfidence of those BMP assays is around 70%. Thompson and He ( Thompson and He 2006) repor ted that the heat of combustion of the crude glycerol ranged from 18600 kJ /kg to 25200 kJ /kg. It had shown 1 ml GBP was able to yield 450 ml CH4 at STP .Using the heat of combustion of methane (37 kJ L-1), the energy that can be gained fr om biogasifying GBP to methane was calculated to be 18300 kJ /kg. There was no big difference in directly combusting and anaerobically digesting GBP to obtain energy. However, glycerol combustion will produce highly toxic acrylonitirle if operat ed inappropriately, and cause server health concern. The advantage of anaerobi c digestion is obvious for it is environmental-friendly and conserve most of the energy. An emergy analysis was conducted for biodiesel production process at small scale (please refer to Appendix A for deta iled information). It revealed that 104 KJ electricity and 495 KJ fuel energy (natural gas) is consumed to produce 1 kg biodiesel ( Sheehan, Camobreco et al. 1998). Electricity is mainly used in equipment operations and fuel energy is used for heating purpose. Ty pically 0.1 kg GBP is generated per 1 kg biodiesel pr oduced. According to the BMP assa y results, 0.1 kg GBP is able to produce

PAGE 46

46 38.8 L CH4 at STP. Given CH4 to electricity efficiency of 25%, it would generate 378 kJ electricity (kg biodiesel)-1. Electricity consumes 28% of the produced biogas energy, leaving in excess of electricity energy that can be sold to make profits. If biogas were applied in boilers to produce steam, it w ould produce an energy equivalent of 1058 KJ assuming a combustion efficiency of 70%. Th is is sufficient to cover fuel energy consumptions as well, sugges ting anaerobic digestion has the potential to improve the energy efficiency of biodiesel production. In full scale biodiesel plants, energy requirement is more intensive because byproducts (GBP) purification is usually carried out. In a biodiesel plant with a capacity of 100,000 ton per year, 3064 KJ fuel energy is consumed to produce 1 kg biodiesel (Emiliani and Pistocchi 2006). The produced methane from GBP is able to provide appr oximately 30% energy consumed. If GBP were digest ed, the purification would not be needed and the energy consumption would be lower. Chen et al. ( Chen, Romano et al. 2008) report ed higher methane production rate of glycerol digestion. Their study obtai ned methane yield of 390 ml CH4 at STP (ml glycerol)-1 using a similar BMP method in 15 days, whereas it took 60 and 90 days for GBP and pure glycerol, respectively, to ac hieve a comparable methane yield in the experiment conducted in this study. This could be explained as smaller loading quantity leading to faster substrate breakdown ra te. In Chens study, the batch reactors contained 2.8 ml (L) -1 glycerol, less than half of the substrate concentration of 6.7 ml GBP (L)-1 and 6.7 ml glycerol (L)-1in assay A-1 and C-II respectively. Figure 2-1 and 2-2 showed 32% and 40% of the ultimate methane yield were produced in the first 20 days for assay A-1 and C-II, 1. This indicated t hat assay A-1 and C-II, 1 degraded 2.1 ml (L)-

PAGE 47

471 and 2.68 ml (L)-1 glycerol in 20 days. It appeared GBP or glycerol at 2 to 3 ml (L)-1 could be degraded as fast as in 15 to 20 days as Chen et al. concluded, but digestion of GBP or glycerol with higher concentration required much longer time. Among those assays, glycerol used in assay A-3 (99.5% grade) was believed to have characteristics closest to the refined glycerol used in Chens study. Thus, 2.68 ml (L)-1 glycerol degraded by assay A-3 within 20 days was closest to glycerol concentration applied in Chens study. Chens study also carried out continuous digestion of glycerol along with dairy manures. In trial 1, mixture 1 (c ontaining 16% (%wt) glycerol and 84% manure) was fed and microbial inhabitation was obser ved when glycerol feeding rate increased from 0.29 ml glycerol (L)-1 (day)-1 to 0.57 ml glycerol (L)-1 (day)-1. Further digestion was then carried out using mixture 2 containing less glycerol (9% glycerol and 91% manure) and the methane yield was good for feeding glycerol at 0.43 ml glycerol (L)-1 (day)-1 and incrementally increasing to 2.57 ml glycerol (L)-1 (day)-1. The methane yield from digesting mixture 2 was much higher t han the combined methane yield of both substrates digested separatel y in Chens batch reactors In order to investigate feasibility of the methane yiel d, theoretical methane yield of mixture 2 was calculated by proportional combining theoretical methane yield of glycerol and methane yield of manure derived in using the batch reactors. Surp risingly, the practica l yield of mixture 2 was so high that it exceeded the calculated theoretical yield by 43% to 72%, depending on mixture 2 loading rate. This may imply that the glycerol used in Chens study contained other compositions that have hi gher methane potential than glycerol itself. However, the practical of mixture 1 only accounted for 86% to 105% (depending on mixture 1 loading rate) of the corresponding theoretical methane yield. The low methane

PAGE 48

48 yield from digesting mixture 1 was attributed to high C/N ra tio of 20 by the authors. Average C/N ratio was found as high as 58 in the semi-continuous digestion of GBP conducted in this study. Actual C/N ratio at late stage of the digestion (3 ml GBP day-1 L1 feeding) was expected even higher, where hampered methane yield was observed. It was hypothesized that low methane yield of semi-continuous digestion of GBP could be due to 1) organic loading problems as found in BMP assays and continuous digestion of mixture 1 conducted in Chens study; and/or 2) high C/N ratio resulting from N depletion. Closing Remarks The BMP assays performed at mes ophilic co ndition showed the methane potential of GBP was 450 ml CH4 at STP (ml GBP)-1, if organic overloading was avoided. GBP digestion exhibited a bi-phase trend, where di gestion of methanol completed fast in phase 1 followed by slow digestion of glycerol in phase 2. Organic overloading problem was seen in assay A2, A-3 and C-II, 2 as their substrate concentration (GBP or glycerol) considerab ly exceed the suggested concentration (5 g L-1 7 g L-1) to use glycerol to produce biogas ( Holm-Nielsen, Lomborg et al. 2008). Due to co-digesting with ta iling extracts, assay A-3 overcame the lag phase a nd made a decent amount of methane, whereas assa y A-2 and C-II, 2 failed to produce adequate methane at the end of the study. The correlation between the remaining COD concentration and methane production was prov ed strong as the predicted COD profile was consistent with the measured COD pr ofile. VOA analysis showed degradation of VOAs in the assay A-1 and C-II, 1 and accumula tion of VOAs in the assay A-2 and C-II, 2 where organic overloading occurred. Prop ionic acid had the highest concentration among VOAs monitored, implying the ferment ation pathway of GBP could be through

PAGE 49

49 propionic acid, but further studies are needed before drawing any conclusion. Semicontinuous digestion of GBP determined daily methane yield stabilized at 450 ml CH4 at STP (ml substrate)-1 at loading rate of 0.67 ml GBP (day)-1(L) -1. Digestion performance was inhibited as GBP loading ra te increased to 1 ml (day)-1(L) -1.

PAGE 50

50 Table 2-1. Contents and c onstituents of the assays Assays Number of Replicates GBP Inoculum Nutrient Solution Tailings water extract DI water ml ml ml ml ml A-1 2 1994.28 050 A-2 2 2984.28 050 A-3 2 0994.28 051 Control assays Number of Replicates Glycerol Inoculum Nutrient Solution Tailings water extract DI water ml ml ml ml ml C-I 2 0994.28 051 C-II,1 2 1994.28 050 C-II,2 2 2984.28 050 C-III 2 01004.28 500 Table 2-2. Parameters of Gompertz Equation P Rm Overall methane yield ml CH4@STP ml CH4@STP(days)-1 days ml CH4@STP (ml)-1 Assay C-II,1 394.156.440.52 394.15 Fraction 1 146.5817.560.00 466.39 Assay A-1 Fraction 2 319.818.870.64 Fraction 1 268.9530.691.61 828.25 Assay A-3 Fraction 2 559.3010.372.14

PAGE 51

51 Figure 2-1. Cumulative methane yiel d from assay A-1, C-II, 1 and A-3

PAGE 52

52 Figure 2-2. sCOD profile in assays A-1, A-2, C-II, 1 and C-II, 2 along with expected sCOD concentrations

PAGE 53

53 Figure 2-3. Propionic acid pr ofiles from assay A-1 and A-2

PAGE 54

54 Figure 2-4. Daily methane yields fr om semi-continuous digestion of GBP

PAGE 55

55 CHAPTER 3 ANAEROBIC DIGESTION FOR TREATMENT OF STIL LAGE FROM CELLULOSIC BIOETHANOL PRODUCTION Summary Thermophilic an aerobic digestion of st illage from cellulosic ethanol production process was investigated. Methane potent ial of bagasse and hardwood stillage was determined by conducting Biochemical Met hane Potential (BMP) assays. A methane potential of about10 ml CH4 at STP (ml stillage)-1 or 200 ml CH4 at STP (g VS)-1 was obtained for bagasse stillage at concentration of 14.2 and 17.8 g VS (L)-1; while 6.5 ml CH4 at (ml stillage)-1 or 100 ml CH4 (g VS)-1 was obtained for hardwood stillage at concentration of 17.2 and 21.5 g VS (L)-1. Coarsely separated bagasse stillage (fraction < 0.5 mm size) was successfully anaerobically digested in a laboratory scale (15 L) continuously fed digester which was operat ed for three months. The digester was operated at HRTs of 21 and 14 days and organic loading rate (OLR) of 1.85 and 2.39 g COD L-1d-1. The methane yield from the stillage fr om the digester was about 90% of the yieldfrom the BMP assays. The influent soluble COD (sCOD) was reduced from between 35.4 38.8 gCOD (L-1) to between 7.5 8 gCOD (L-1). The soluble chemical oxygen demand (sCOD) removal efficiency wa s 7580% with the sCOD of effluent. The success in anaerobic digestion of bagasse still age without any dilution was attributed application of long HRTs and low OLRs. A mass and energy balance was developed for an integrated process of cellulosic ethanol production and stilling treatment using anaerobic digestion. It showed the anaerobic di gestion being able to produce 70% of the energy required by the ethanol production process.

PAGE 56

56 Background Currently, commercial bioethanol production processes focus on utilizing classic crops as feedstock lik e sugar cane and wheat t hat require high quality agricultural land for growth ( Balat 2011). Given that serious problems face the world food supply today, diverting farmland or crops for fuel producti on is detrimental to the food supply on a global scale ( Demirbas 2011). The issue is more pressing in developing countries where food scarcity may happen. An alter nate option is to produce ethanol from nonfood biomass resources like forestry and agricultural residues and urban wastes. Ethanol produced from these feedstocks is usually referred as cellulosic ethanol, because in these feedstocks the predominant carbonhydrates converted to ethanol are the polymers cellulose and hemicellulose. Cellulosic materials are first subjected to biological and/or ther mochemical pretreatment. Enzymes and/or high temperature along with an acid/base is introduced to hydrolyze cellulose and hemicellulose to sugars. Pretreatment is followed by fermentation of the sugars. For example, at the Biofuels Pilot Plant in University of Florida, a re combinant E. Coli KO11 strain developed in the Microbiology and Cell Science Department is employed to ferment both the hexose and pentose sugars to ethanol (Yomano, York et al. 1998; Yomano, York et al. 2008; Yomano, York et al. 2009).The fermentation is preceded by a dilute acid hydrolysis using phos phoric acid as catalyst and an enzymatic hydrolysis using commercial cellulase enzymes. After fermentation, t he broth is distilled to produce ethanol. Distillation generates a by product called stilla ge, which is generally a high volume, high strength acidic waste that pr esents significant disposal or treatment problems. Stillage from corn ethanol production is used as anima l feed as either distillers dried grains (DDG) or distillers dried grains with so lubles (DDGS). However, DDGS production

PAGE 57

57 consumes large amount of energy due to th e evaporation and drying process (Kaparaju et al., 2010).Stillage produced from nonfood s ources as sugarcane bagases usually has poor nutritive quality and is not suitable to use as animal feed. Utilization of stillage as a f uel or source of biofuel can improve energy efficiency and reduce carbon footprint of ethanol production process. A possible option to add value to the stillage would be to use it as a fuel in bioethanol production process. However, due to its high moisture content, direct combustion would not provide net energy. Anaerobic digestion was found to be a viable and sustainable scheme for simultaneous waste treatment and energy production from stillage ( Wilkie, Riedesel et al. 2000). The biogas produced from anaerobi c digestion can be used to supplement the energy consumed in the pretreatment and distillation operat ions of ethanol production process. Several studies have been carried out on anaerobic digestion of ethanol stillage from conventional sugar or starch based feedstocks such as grain and sugar cane molasses ( Sheehan and Greenfield 1980; Shin, Bae et al. 1992; Wilkie, Riedesel et al. 2000; Tang, Fujimura et al. 2007). However, few studies were conducted on digesting ethanol stillage produc ed from cellulosic materi als. Unlike processing of sugar or corn for ethanol, cellulosic materials have to be s ubjected to harsher processing steps to depolymerize the structural polysaccharides. These processes result in side reaction products that could be potentially inhibitory to microbial growth. So anaerobic digestion of cellulosic ethanol stillage may be fraught with problems. In this study, stillage obtained from the Biofuel Pilot Plant at the University of Florida was used as a sole substrate in biochemical methane potential assa ys. In addition to the methane potential

PAGE 58

58 of the substrate, the assay also provided information on the rate of degradation, extent of degradation and the presence of any inhibitory compounds or toxins in the substrate. These batch assays were performed at a thermophilic (55 C) temperature. The feasibility of using a one-stage, continuous digester for anaer obic digestion of stillage was then investigated. Materials and Methods Feedstock Stillage c ollected from the Biofuel Pilot Pl ant at University of Florida was used as feedstock for anaerobic digestion. The pilot plant produces ethanol from lignocellulosic biomass, mainly sugarcane bagasse and hardwood chips. The process for ethanol production at the Biofuel Pilot Pl ant was as follow: The raw material was soaked in 1% phosphoric acid for 4 hours. The excess acid was removed in a screw press. The acid soaked fibers were l oaded into a hydrolyser. Dilute-acid steam pretreatment was carried out at 180C fo r 10 minutes to solubilize lignocellulosic biomass from its native form, in which it is recalcitrant to cellulase enzyme converting systems, into a form for which en zymatic hydrolysis is effective (Lynd, Weimer et al. 2002; Kumar, Chandra et al. 2010). The pret reated feedstock was hydrolyzed with cellulase. Fermentation of the enzyme-treated slurry was then carried out using a recombina nt E.coli strain. The stillage was collected after the distillation process in which ethanol was separated from fermentation broth at 60 C. For the experiments reported in this c hapter, some of the stillage was also filtered through a sieve with pore size of 0.5 mm for coarse separation of solid and liquid. The unfiltered stillage, filtered stillage, and the fracti on retained by the sieve are hereby referred as whole stillage, stillage filtrate and stillage residue, respectively. Two

PAGE 59

59 types of stillages were used in the bioc hemical methane potential assays namely stillage produced from sugarcane bagasse as feedstock and stillage produced from hardwood as feedstock. In cellulosic ethanol fermentation process, the lignin in the feedstock passes unconverted through the process and remains in the stillage ( Ojeda, vila et al. 2011). For value add ition to stilla ge and thereby to improve the economics of the process it has been sugges ted that the lignin frac tion be used as a fuel ( Larsen, Petersen et al. 2008) or as a raw material for biocompos ites ( Gonzlez, Santos et al. 2011). Lignin is not degraded in an anaerob ic digester and moreover microbial hydrolysis of fiber is a slow step. Therefore, by separ ating the fibers only the quickly degradab le portion of the stillage was fermented. If fibers are removed, it is also necessary to quantify the methane potential that would be lost with the fibers by not digesting it. Biochemical Methane Potential Assay (BMP) BMP assays measure the meth ane potentia l of feedstocks at optimal conditions for culturing anaerobic digesti on microbial consortia ( Owen, Stuckey et al. 1979; Shelton and Tiedje 1984; Owens and Chynoweth 1993). It was conducted by mixing the substrate (the material for which the methane potential is being determined) with inoculum and nutrients, and incubating the mixt ure in a sealed bottle at predetermined temperature for a peri od of ti me to allow complete degradat ion of the substrate. Assays contained substrate, inoculum and other supportive solutions (micronutrient, macronutrients, and pH buffer). The inocul um used was collected from a laboratory scale anaerobic digester that had been di gesting various biomass feedstock and wastewater at ther ophilic temperature for over three y ears. Nutrient stock solution was prepared according to Owens et al. ( Owens and Chynoweth 1993). Four types of

PAGE 60

60 assays were carried out, each in duplicat e, containing, hardwood whole stillage, bagasse whole stillage, bagasse stillage filtrate or bagasse stillage residue. A control assay containing deionized water was set up to provide data on methane production from residual substrates and endogenous met abolism in inoculum. The Biofuel Pilot Plant used hardwood chips once fo r a trial operation. As a result, hardwood stillage was available in small quantity that was only suffi cient for the BMP assa ys but not for the lab scale anaerobic digestion, which was later carried out using only the bagasse stillage. Table 3-1 lists the assays along with its contents and quant ities loaded. The methane yield was calculated by subtracting the me thane produced in the control assays in order to take into account the methane contribution from the inoculum. Anaerobic digestion process can be carried out at mesophilic (32-38 oC) or thermophilic (50 57 oC) temperatures. A therm ophilic temperature of 55oC was chosen here for the following reason. In industry, distillation is usually conducted at 60 -70 oC to separate ethanol from the ferm entation mixture. Therefore, stillage would be produced at these temperatures. To take advantage of the hot stillage that is available, an onsite anaerobic digester can be operated at a thermophilic temperat ure. It has been shown that at thermophilic temperatures the rate of degradation is higher and consequently the throughput for a given digester volume can be increased or the volume of digester required for a specified throughput can be decreased ( Wang, Ma et al. 2011). Continuous Digestion of Bagasse Stillage Filtrate The feasibility of anaerobic digestion of bagasse stillage filtrate was further evaluated in a laboratory scale 15 L semi -continuously fed anaerobic digester. The digester was constructed by modifying a Pyre x glass jar. The height and inner diameter of the digesters were 0.8 m and 0.3 m, respectively. The bottom of the glass bottle was

PAGE 61

61 cut and replaced by a stainless plate with ports This served as the digester lid and was attached to the bottle with screws and a rubber gasket for air-tight seal. A glass flange fitted with a rubber O-ring was used to seal the top of the bo ttle. The flange was clamped using stainless steel clamps for gas and liquid tightness. The digester was provided with several ports for gas outle t, sample withdrawal and digester liquor circulation. The bottle was placed upside do wn on a tripod stand with influent entering through the bottom. Since the digester c ontent was not mechanically mixed, solids settled down and accumulated at the bottom, a llowing significant separation of SRT and HRT. This makes the digester resemble a SOLCON reactor to some extent ( Srivastava, Biljetina et al. 1989). Figure 31 and 3-2 provides schematics of the tripod stand and the continuous digester, respectively. Gas produc tion from the digester s was measured by a positive displacement gas meter which consisted of a clear PVC Utube filled with anti-freeze solution, a solid state time delay relay (Dayton Off Delay Model 6X153E), a Grainger float switch, a Redington counter and Fabco Air solenoid valve. The gas meter was calibrated in line to determine the volume of biogas per count. The biogas was vented to the building exhaus t piping from the exit of the volumetric gas meter. The digester was operated in a semi-cont inuous mode with feeding occurring 20 times per day. Feeding was controlled by a two headed MasterFlex pump that was operated by a timer. The timer cycle operated such that each pumping cycle lasted from 5 minutes to 10 minutes to give HTRs of 21 and 14 days. A second timer was used to control the influent substrate mi xer so that it would turn on for 10 minutes before influent pumping began and did not turn off until af ter the pumping cycle was finished. The pump was calibrated to deliver 28 mL still age per 5 minutes. A second MasterFlex

PAGE 62

62 pump was provided for digester liquor circ ulation to provide mixing and improve temperature homogeneity and contact of subsra te to microbes. The circulation rate was set at 50 ml per minute. The digester was started up by in oculating with 12 L inoculum and allowed to stabilize for 1 day before the f eeding was started. Stillage was filtered through a sieve with 0. 5 mm mesh and the filt rate was fed as the feed. Like the BMP assays, the lab scale semi-continuous digester was maintained at 55 oC and the stable temperature was achieved by use of a heating tape controlled by a Campell Scientific datalogger (Model CR10). The tem perature was controlled by on-off protocol, i.e., heating tape was turned off when temperature reached 57 oC and turned on when temperature dropped below 53 oC. Temperature was monitored using a T-type thermocouple located halfway between center and wall of the digester. The digester was also covered by a layer of insu lation material to reduce heat loss. Monitoring and Analysis Unfiltered stillag e, filtrate and residue samp les were analyzed for total solids (TS), volatile solids (VS), soluble COD (sCO D), ammonia-N, phosphate-P, ethanol and forage nutritional compositions. The BMP assays were monitored daily for biogas production and methane composition of biogas, and periodically for soluble COD and volatile organic acid concentration. The pH of the mixed liquor at the start and end of the assay was also measured. Biogas that accumulated in the serum vials was withdrawn using a syringe. Biogas was allowed to fill the syringe unt il the pressure in the vial head space equilibrated with that in the syringe. Volume of biogas in the syringe was noted and then used as a sample for compositional analysis.

PAGE 63

63 Biogas production from the continuously fed anaerobic digester was monitored in line using a positive displacement gas meter. The biogas was analyzed for methane content daily. The soluble COD, volatile organic acid, ammonia-N and phosphate-P concentrations of the effluent was analyzed onc e weekly. A forage nutritional analysis of a composite effluent sample was also carried out. TS and VS contents were measured using an oven (Fisher Scientific Isotemp model 350G) and a muffle furnace (Fisher Scientific Isotemp), respectively. Samples were centrifuged at 8000 RPM for 10 minutes (Fisher Marathon micro H centrifuge) and filtered using Millipore filt er paper (pore size 1.2 m) for sCOD, ammonia and phosphorus analysis. For sCOD measurement, samples were added in Hach COD vials after appropriate dilution and baked in Hach COD reactor for 2 hours. SCOD was measured using Hach DR/890 Colorimeter. Ammonia and phosphorus concentration was measured using ammonia-selective electrode and ascorbic acid method, respectively ( American Public Health, American Water Works et al. 1999) Ethanol content was measured using an Agilent gas chromatograph (Agilent Technologies, 6890N). Methane and carbon dioxide compos ition was measured using a Fis her Gas Partitioner (model 1200). The gas chroma tograph was calibrated with an external standard containing N2, CH4, and CO2 in 25:45:30 volume ratio. Volatile organic acids were measured using Shimadzu gas chroma tograph (GC-9AM equipped with a flame ionization detector). The volatile organi c acids detected were acetic, propionic, isobutyric, butyric, isovaleric and valeric acid. Stillage samples and effluent from diges ter were also analyzed for forage nutritional compositions, including crude protein, soluble protein, Acid Detergent Fibers

PAGE 64

64 (ADF), Neutral Detergent Fibers (NDF),lig nin, Ethanol Soluble Carbohydrates (ESC, including glucose, fructose, sucrose, maltos e, and short fructose chains ), Digestible Energy (DE), Metabolizable Energy (ME) an d Total Digestible Nutrients (TDN). ADF isolates cellulose and lignin and NDF isolat es cellulose, hemicellulose and lignin. Knowing the lignin content and by difference, cellulose and hemicellulose contents was determined. These analyses were conducted by a commercial forage testing laboratory (Dairy One, Inc, Ithaca, New York). Results Characterization of the Stillage Selected characteristics of stilla ge and digested stillage were determined and listed in Table 3-2. This includes TS and VS content, sCOD concentration, cellulose, hemicellulose, lignin, ESC and TDN cont ents for bagasse whole stillage, bagasse stillage filtrate, bagasse stillage residue and the effluent from the continuous digester. TS and VS content, and sCOD concentration for hardwood whole stillage is also listed. The bagasse and the hardwood stillages had a dr y matter content of 6.89 and 8.86 % respectively. The lower solid content is as a result of operating the ethanol fermentor at less than 10% solids content to maintain g ood mixing. When 100 ml of whole bagasse stillage was filtered in a 0.5 mm sieve, it yielded 19 g residue (retained on the sieve) and 69 ml of filtrate. The so luble COD of bagasse stillage was more than that in the hardwood stillage. The bagasse stillage filtra te had an ethanol content 0f 0.75 g (L)1.DE, ME and TDN values (dry matter basis) of bagasse stillage (bagasse whole stillage, bagasse stillage filtrate and bagasse stillage residue) are in the range of 2500 kcal (kg)-1, 2000 kcal (kg)-1 and 55%, respectively.

PAGE 65

65 Methane Potential Profiles of the average methane yiel d from assay BW-1 (containing40 ml bagasse whole stillage), BW-2 (containing 50 ml bagasse whole stillage), BR (containingbagasse stillage resi due), BF (containing stillage filtrate), HW-1 (containing 40 ml hardwood wh ole stllage) and HW-2 (c ontaining 50 ml hardwood whole stillage) are shown in Figure 3-3. Sample standard deviations are shown as error bars.The methane potential of bagasse whole stillage over 80 days of incubation was 10.25 ml CH4 at STP (ml stillage)-1 and 10.95 ml CH4 at STP (ml stillage)-1 from assays BW-1 and BW-2, respectively. The degradation rate of st illage as indicated by the slope of the curves, initially increased, reached a maximum and then decreased to a minimum as substrate was utilized. The rates of evolution of methane were very similar for the first 20 days. A slightly higher methane production rate was observed a fter that for assay BW-2, leading to a higher methane yield in the end. In general, the methane yields and evolution rates of assay BW-1 and BW-2agreed with each other quite well. For both assays, 85% of ultimate methane yield was produced in 50 days. The methane yields from assay BW-1 and BW-2 can be also be expressed per gram VS in substrate and this corresponds to 191 ml CH4 (g VS)-1 and 204 ml CH4 (g VS)-1, respectively. The methane potential of hardwood whole stillage was determined to be 6.72 CH4 at STP (ml stillage)-1and 6.18 CH4 at STP (ml stillage)-1 for assay HW-1 and HW-2, respectively. Assay BR and BF contained bagasse stillage residue and bagasse stillage filtrate, respectively. These assays contained residue and filtrate obtained by coarse filtering 100 ml of whole st illage, i.e., 19 g and 69 ml respectively. The methane yields for assay BR and BF normalized per ml of whole stillage, were calculated as

PAGE 66

66 ml) (100 Volume Stillage Whole filtrate)or (residue n Productio CH Individual4 (3.1) The methane yield from assay BR and BF were 3.05 ml CH4 at STP (ml stillage)1 and 6.91 ml CH4 at STP (ml stillage)-1, respectively, after 80 days of incubation. Assay BF and BR had similar methane production profiles for the first 40 days. Methane production continued and increa sed in assay BF whereas it tailed off in BR. Soluble COD and Volatile Organic Acids in BMP Assays Soluble COD was measured periodica lly in the assays. The soluble COD profiles in assay BW-1, BW-2 and HW-1 and HW-2 are shown in Figure 3-4 and 3-5, respectively. The initial measured SCOD in BW-1 and BW-2 were 11 and 13 gL-1 in the assay.These values dropped to around 5 g L-1 after 25 days and slowly decreased to 3.6 g L-1 in BW-1 and 2 g L-1 in BW-2. The maximum SCOD drop rate was seen from day 13 to day 27.This was consistent with the methane yield profiles, which achieved highest methane evolution rate during a simila r period (day 17 to day 35). Soluble COD in HW-1 assay dropped from 7 g L-1 to 2 g L-1 during the assay and in HW-2 it dropped from 10 g L-1 to 3.1 g L-1. Concentration of VOAs, i.e., acetic acid propionic acid, isobutyric acid, butyric acid, isovaleric acid and valeric acid wa s also measured in assays BW-1, BW-2, HW-1 and HW-2 and the profiles of acetic acid, but yric acid and propionic acid concentration in assay BW-1, BW-2, HW-1 and HW-2 are shown in Figure 3-4 and 3-5. Acetic acid was found to be the most abundant contributor to total concentration of VOAs. Butyric acid was the second most abundant and its conc entration was about half of that of

PAGE 67

67 acetic acid. Concentration of all other VO As was negligible. The VOA profiles for both assays were similar. For assay BW-1 and BW-2, concentration of acetic acid and butyric acid increased shortly after the assay was initiated. Acetic acid and butyric acid started to degrade after peaking at around 1.3 g/L and 0.7 g/L, respectively at day 10. Accumulation of propionic acid persisted fo r a longer time and its concentration did not decrease until day 35. The VOA profiles for assay HW-1 and HW-2 were similar, but butyric acid was in much less concentration than in assays BW-1 and BW-2. This could result from different compositions of bagasse and hardwood stillage. In general, degradation of COD and VOAs was evident in assays BW-1 BW-2, HW-1 and HW-2. Continuous Digestion of Bag asse Stillage Filtrate (BF) The bagasse stillage filtrate digestion was conducted in 2 stages and in each stage a different HRT was utilized to operate the digester. The digester was operated for 91 days. Table 3-3 summarizes the steady state results. Steady state occurred when the daily biogas production and biogas composition was stable unless the feed (flow rate or composition was changed).Figur e 3-6 and 3-7 illustrates profiles of CH4 production, pH, effluent sCOD, VOA, phos phorus and ammonia concentration with operation time for both stages. During stage 1 (day 1 to day 22), BF was fed at 571 ml/day (hydraulic retention time, HRT =21 days). Daily methane production increased and stabilized at 9.48 L CH4 (L BF) -1 after 12 days. This accounted 95% of CH4 potential of stillage filtrate as determined by the BMP assay, suggesting conversion of stillage to CH4 was nearly complete at HRT of 21 days HRT. The sCOD in the digested effluent was 7.8 g L-1. The VOA analysis of digested effluent showed acetic acid and propionic acid were the dominant acids. A sharp increase was seen for acetic acid and propionic acid concentration from day 4 to day 10, and its concentration stabilized at

PAGE 68

68 around 1 gL-1 and 0.8 g L-1 (acetic acid and propionic acid, receptively) thereafter when the steady state was reached. Phosphate-P concentration in the effluent steadily increased as the digester was operated and by day 21 it had reached 200 mg L-1. Ammonia-N concentration in the effluent fluctuated and was dropping below 200 mg L-1 by day 21. During stage 2 (day 23 to day 90), the feed rate was increased to 857 ml/day (14 days of HRT). The digester acclim ated to the change quickly and the CH4 production reached a new stable state after only one day (day 24). The average CH4 yield was 10.50 L CH4 (L stillage) -1, slightly increased from the CH4 yield at stage 1. This indicated that the digester stability was not affected by the increased loading rate. Effluent sCOD and VOA profiles during stage 2 kept low levels, once again indicating a stable digestion perform ance. Phosphate-P and NH3N concentration stabilized around 350 mg L-1 and 100 mg L-1, respectively. Discussion A forage analys is on bagasse stillage was conducted to determine if it has value as animal feed. Previously, extensiv e compositional anal ysis was conducted on distillers dry grains with solubles (DDGS) produced in co rn ethanol plants by several researchers. Sophiehs et al. ( Spiehs, Whitney et al. 2002) collected 118 DDGS samples from 10 ethanol plants in t he Minnesota-South Dakota-region and reported the average DE and ME values (dry ma tter basis) were 3,990 and 3, 749 kcal/kg, respectively. A cellu losic biomass analysis and forage nutritio nal analysis for DDGS and wet distillers grains (wet cake produced from centrifuging whole stillage) produced from Big River Resources, LLC, a dry-grind ethanol facility at West Burlington found that DDGS and wet distillers grain have sim ilar nutritional composition wit h high crude protein contents

PAGE 69

69 of 30% and TDN of 90% on a dry matter basis ( Kim, Mosier et al. 2008). Compared to these value s the nutritional value of bagasse stillage was substantially lower. For example DE, ME and TDN were 2,470 kcal/ kg, 2060 kcal/kg, and 53% respectively compared to values close to 4000 kcal/kg and 90%. Moreover, crude protein was only 9.9%. So bagasse stillage is nutritionally po or and not suitable for use as animal feed unless supplemented wit h other components. When the whole stillage was filtered, some of the fibe rs also passed through the sieve resulting in 4.8% solids content in the filtrate. The total mass of dry matter retained on the sieve and that whic h passed through the sieve was 6.8 g (048.0691858.019 ) in 100 ml which was in close agreement with t he solids content of 6.89 % for bagasse whole stillage. Since the methane yields for assay BR and BF were calculated per whole stillage volume, they were simply added to determine the whole stillage methane potential and then compared to that obtained in assay BW-1 and BW-2. The combined methane production profile of assay BR and BF was shown as dotted line in Figure 3-3. The combined yield was calculated to be 9.97 ml CH4 at STP (ml stillage)-1, which was consistent with the yield obtained from assa y BW-1 and BW-2. This confirmed that the methane potential of the stillage was approximately 10 ml CH4 at STP (ml stillage)-1. The stillage filtrate gener ated approximately 70% CH4 potential out of the whole stillage. On a dry matter basis the methane yield fr om BF was 212 ml methane (BF dry matter)-1 was much greater than t hat from BR which was 85m l methane (BR dry matter)-1. Since BR dry matter was mainly composed of su spended residual fibers, it can be concluded that very little of the residual fibers in t he stillage were digested an d that most of the

PAGE 70

70 methane was produced from dissolved organic matter. As discussed later, this was further confirmed from soluble COD measurements in the assay. It has been reported that the economics of the lignocellulosic ethanol process is highly dependent on the income of co-products ( Sassner, Galbe et al. 2008). During the downstream process, stillage stream can be treated with anaer obic digestion and the produced biogas can be incinerated to provide heat and improve overall energy efficiency ( Wilkie, Riedesel et al. 2000; Wingren, Galbe et al. 2008). A technoeconomic model of anaerobic digestion of stilla g e for a spruce to ethanol process has been developed ( Barta, Reczey et al. 2010) and it concluded that the difference in the production cost of ethanol bet ween using whole stillage and the liquid fraction in anaerobic digestion for producing biogas wa s negligible (0.4-0.5 euro/liter). The average methane yield of hardwood stillage was 6.5 ml (ml-1) about 35% less than that of bagasse stillage. This coul d be attributed to the lower sCOD (measure of dissolved organic matter) of HW stillage On a VS content basis, BW produced 195 ml CH4 (g VS)-1 whereas HW produced 98 ml CH4 (g VS)-1. Stillage methane yields from this study was lower than 324 ml (g VS)-1 obtained from wheat straw stillage ( Kaparaju, Serrano et al. 2010). The low CH4 yield was probably due to that the wheat stillage contained much higher sCOD (61 g/L) than that of the stillage used in this study. The bagasse stillage still contained about 0.075% ethanol which is less than reported 0.23% ethanol in wheat straw stillage ( Kaparaju, Serrano et al. 2010). Ethanol in the stilla ge will also cont ribute to the methane potential of stillage. Theoretically, about 0.73 L methane at STP wo uld be produced per g of et hanol, therefore, methane yield from stillage from the ethanol c oncentration measured above would be 550 ml CH4

PAGE 71

71 (L stillage)-1 or 0.55 ml (ml-1), which is about 5% of the methane yield obtained from bagasse stillage filtrate. Along with measured sCOD values, the ex pected theoretical sCOD values were also plotted. The expected values were calculated using stoichiometry that 1 g L-1 of sCOD upon mineralization in an anaerobic digester produces 0.35 L CH4 at STP. Using the conversion, it is possible to predict t he remaining sCOD concentration in the assays by knowing how much methane has been produc ed. The initial values of expected sCOD were assumed the same as the initia l measured sCOD in the assay. For all assays, the expected values followed the meas ured trend closely. Nevertheless, slightly lower values were seen at the end of assa y for expected than measured SCOD. If only the soluble COD was minera lized, then the expected value should match the measured value. The lower expected value could mean that some fibers contained in the whole stillage could have been digested to methane as well. Based on the difference between expected and measured it was determined that 2.3 g COD (L-1) in BW-1 and 3.1 g COD (L-1) in BW-2 of fibers was mineralized. A ssuming the fibers are cellulosic (therefore 1.0667 gCOD/g) this corresponds to 2.2 g (L-1) and 2.9 g (L-1). In other words, only an additional 3-4% of the dry ma tter would have been converted to methane. This justified the later use of only bagasse st illage filtrate for continuous di gestion. For hardwood, the SCOD profiles (measured and expected) fo r assay HW-1 and HW-2 showed similar behaviors with sCOD decreasing steadily as methane was produced and difference between measured and expected at the end of the assay was less than that in the BW assays.

PAGE 72

72 The bagasse filtrate stillage was succe ssfully treated in a continuous single stage digester operated for about 70 days at an HRT of 14 da ys. The bagasse filtrate stillage was fed as received after coarse separ ation of fibers. The pH in the digester was maintained at around 7.5 without any pH control. The methane yield obtained from the digester was comparable to that obtai ned from the BMP assays. The phosphate concentration decreased from 530 mg L-1 to 350 mgL-1 in the effluent and the ammonia from 300 mgL-1 to 100 mg L-1. CH4 yield can be also estima ted by knowing the amount of SCOD that has been degraded At the state where CH4 yield was constant, sCOD of the digester effluent and stillage filtrate was measured to be 8.0 g/L and 38.8 g/L, respectively. The expected CH4 yield was 8.96 L CH4 (L stillage) -1, which generally agreed to the actual CH4 yield. Anaerobic digestion has been used for treat ing various types of ethanol stillage (or distillery wastewater) including sugar beet stillage, potato stillage, wheat stillage and shochu stillage, and has resulted in COD remova l efficiencies of 75-95% using different reactor configurations ( Weiland and Thomsen 1990; Nagano, Arikawa et al. 1992; Wilkie, Riedesel et al. 2000; Hutnan, Hornak et al. 2003; Schaefer and Sung 2008). It should be noted that the stillag e was produced from fermentation of sugar or starch based feedstocks and hence contains a higher content of degradabl e organic carbon. The sCODs of some these stillage was 100 gL-1 or higher. In c ontrast the stillage obtained was from the fermentation of a pretreated cellulosic feedstock with low soluble organic matter. Stillage obtained from such f eedstocks is expected to contain lignin and lignin related compounds as phenols ( Kaparaju, Serrano et al. 2010), which were reported inhibitory for anaerobic digestion ( Sierraalvarez, Field et al. 1994; Torry-Smith,

PAGE 73

73 Sommer et al. 2003). By separating the coarse fibers from the stillage, lignin content of the digester feed in this case was 1.37% (wet weight basis) which was lower than the 7.5% lignin content report ed by Kaparaju et al ( Kaparaju, Serrano et al. 2010). Successful anaerobic digestio n of cellulosic stillage (from wheat straw) at short HRTs (1-2 days) and high OLR (10 20 g COD (l d-1) has been reported ( Torry-Smith, Sommer et al. 2003; Kaparaju, Serrano et al. 2010). Howe ver in these studies, stillag e was either diluted by a readily degradable s ubstrate such as animal manure or subject to a series of pretreatm ent (centrifuge, filtration, autoclaving) before loading to digesters. Digestion of stillage alone as receiv ed with very little pretreatment has not been reported. The results of this study demonstrated anaer obic digestion of stillage alone was feasible. The organic loading ra te could be increased by lowering HRT if specialized digester designs were used. Ho wever, designs like anaerobic filter would be susceptible to clogging due to the fiber content even in the filtered stillage. The significance of implementing an anaerobic digester for treating cellulosic ethanol stillage was ascertained by determining t he potential utilization of biogas as fuel in the ethanol plant. For example, biogas c ould be used to generate steam for feedstock pretreatment and/or as fuel for distillation. To evaluate energy efficiency of the overall process, a mass balance was developed. Assumption was made that 1 g raw bagasse underwent a series of processes (dilute ac id pretreatment, en zymatic hydrolysis, fermentation and distillation) to produce ethanol; the byproduct stillage was anaerobically digested and produced CH4. The raw bagasse contains 50% moisture and on dry matter basis 48% cellulose, 23% hem icellulose and 27% lignin. A total 4.11 g of 180oC steam, water and cellulase was appli ed to 1 g raw bagasse and produced

PAGE 74

74 0.11 g of ethanol and 5 g of stillage. T he ethanol yield was calculated based on degradation in cellulose and hemicellulose contents between bagasse and stillage. Raw bagasse compositions and quantities of steam enzyme and water used in the process were provided by the Biofuel Pilot Plant at Un iversity of Florida. Analysis of cellulose, hemicellulose and lignin in stillage was c onducted by Dairy One. The mass balance illustration is shown in Figure 3-5. In the integrated process, steam s were introduced in pretreatment and distillation which were considered as energy intensive steps. For a typical commercialized distillation proc ess, 15 lbs steams are needed per gallon anhydrous ethanol to separate ethanol from 10% ethanol containing mixture ( Madson 2003). Using this as a reference, 0.27 g steam was required for 0.11 g ethanol. Assuming the steam was at 250 oC, it would consume 0.71 KJ energy. In bagasse pretreatment, 1.34 KJ heat was required to generate 0.5 g 180oC steam. As proposed before, energy consumed can be supplemented by the energy generated from anaerobic digestion. The anaerobic digestion process produced 37 ml CH4, equivalent to 1.47 KJ heat given CH4s heat of combustion being 891 KJ/mol. The energy calculation suggested the energy efficiency of cellulosic ethanol production could be improved significantly by integrating anaerobic digestion t hat produces biogas energy to cover 70% of energy consumed by steams gener ation. Since only stillage filtrate was digested, a strategy needs to be proposed to manage the leftover stillage residue which has been shown to have relatively low CH4 pot ential. The stillage residue was produced from fermentation of lignocellulosic bi omass and contained relatively high solid contents. The lignin fraction has been subject to pretreatment and fermentation process in the bioethanol production and has been reduced to small particles mechanically and

PAGE 75

75 biologically. This would make lignin a better fuel than traditional biomass and well suited for energy production by combustion ( Larsen, Petersen et al. 2008). In fact, lignin is used by some lignite fired power plants as solid biofuels in either dried and wet forms ( Kaparaju, Serrano et al. 2009). Closing Remarks BMP assayswere conducted on bagasse w hole stillag e, bagasse stillage filtrate, bagasse stillage residue and hardwood wh ole stillage. It showed that the CH4potential of bagasse whole stillage being around 10 ml CH4 at (ml stillage)-1 or 200ml CH4 (g VS)1, while bagasse stillage filtrate contributed the major CH4potential of 70%. CH4 potential of hardwood whol e stillage was determined to be around 6.5 ml CH4 at (ml stillage)-1, about 40% less than that of bagasse whole stillage due to less organic matters contained in hardwood whole st illage. Degradation of sCOD and VOA concentration in the assays was evident as CH4 being produced. The continuous digestion of stillage filtra teoperating at HRTs of 21 and 14 days achieved similar CH4 yields as obtained in the BMP assays. This suggested the digestion practically obtained complete conversion of stillage to CH4. sCOD and VOA profiles indicated the digester stability was well maintained at HRTs of 21 and 14 days. During the state where the CH4 production was stable, the COD remova l efficiency was around 80%. The success in the cellulosic stillage digestion can be a ttributed to the operations of long HRTs and low OLRs. Ammonia and phosphorus analysis, however, determined the digestion effluent contained relatively high concentra tion of phosphorus. A further process would be necessary to remove the nutrients befor e discharging the effluent to environment. The mass and energy balance of the integrat ed process indicated anaerobic digestion

PAGE 76

76 has the potential to conserve most of t he energy consumed in steam generation in ethanol production process.

PAGE 77

77 Table 3-1. Contents and cons tituents of the BMP assays Assays Number of Replicates Bagasse whole stillage Bagasse stillage filtrate Bagasse stillage residue Hardwood whole stillage Inoculum Nutrient solution DI water ml ml g ml ml ml ml BW-1 2 40 0 0 01004.2910 BW-2 2 50 0 0 01004.290 BF 2 0 69 0 01004.830 BR 2 0 0 19 0100 3.40 HW-1 2 0 0 0 401004.2910 HW-2 2 0 0 0 501004.290 Control 2 0 0 0 01004.2950

PAGE 78

78 Table 3-2. Characteristics of bagasse whol e stillage, stillage filtra te, stillage residue and hardwood whole stillage. Bagasse whole stillage Bagasse stillage residue Bagasse stillage filtrate Effluent Hardwood whole stillage TS % (wwa) 6.89.3318.58.064. 80.172.40.028.86.12 VS % (ww) 5.36.3116.53 0.133.11.12ND6.46.15 SCOD g/L 38.6NDb38.8 21.8 Ethanol g/L ND ND0.75NDND Compositional Forage Analysis (dry matter basis ) Crude Protein % 9.908.9010.6010.50ND Soluble Protein % 4.162.235.196.41ND Cellulose % 23.0028.909.402.20ND Hemicellulose % 4.308.300.400.40ND Lignin % 20.0022.0017.903.90ND ESC % 4.504.50%5.801.60ND TDN % 53.0046.0065.0085.00ND DE Mcal/kg 2.472.182.953.73ND ME Mcal/kg 2.061.762.533.32ND aww = wet weight based. bND: Not determined.

PAGE 79

79 Table 3-3. Monitored par ameters of anaerobic digestion HRT, days 21 (day 15-22) 14 (day 22-day 91) SCOD, g/L Influent 38.8 38.4 Effluent 7.5 8 P, g/L Influent 0.53 0.53 Effluent 0.19 0.35 NH3 as N, g/L Influent 0.3 0.3 Effluent 0.24 0.1 pH 7.76 7.55 OLR, g COD (L d)-1 1.85 2.39 VOA (acetic and proponic acid), g COD /L 2.23 1.44 CH4 Yield L (L substrate)-1 9.48 10.43 CH4 Yield, L (g CODremoved)-1 0.30 0.38

PAGE 80

80 Figure 3-1. Custom-built tripod stand for anaerobic digesters. Adapted from Polematidis (Polematidis, 2007)

PAGE 81

81 Figure 3-2. Continuous anaerobic reactor schematic. Adapted from Polematidis ( Polematidis 2007)

PAGE 82

82 Figure 3-3. Methane yields of Assay BW-1, BW-2, BF, BR, HW-1 and HW-2

PAGE 83

83 Figure 3-4. COD and VOA profiles of assay BW-1 and BW-2, data represents measured (solid line) and estima ted from model (dashed line)

PAGE 84

84 Figure 3-5. COD and VOA profiles of assay HW-1 and HW-2, data represents measured (solid line) and estima ted from model (dashed line)

PAGE 85

85 Figure 3-6. Daily methane yield fr om continuous digestion of stillage

PAGE 86

86 Figure 3-7. Profiles of SCOD, VOA, phosphorus and amm onia as N of the semicontinuous anaerobic digester

PAGE 87

87 Figure 3-8. Mass balance of the int egrated process of integr ating anaerobic digestion with ethanol production.

PAGE 88

88 CHAPTER 4 COMPARISON OF BATCH ANAEROBIC D IGESTION OF SUGARBEET TAILINGS AT MIXED AND NONMIXED CONDITIONS Summary Thermophilic an aerobic digestion of sugar beet tailings at non-mixed and mixed conditions was conducted. Six experiments each were carried out in two identical digesters with a working volume of 3 L. The tailings in a non-mixed digester was bulked with lava rocks while the mixed digester contents were continuously mixed at 100 RPM. The mixing intensity of 180 RPM was normalized to a unit power input of 6.42 W/m3. Though it was not regarded as a high intensity reported to interrupt syntrophic interactions, the mixed digester showed lower CH4 yield and slower CH4 production rate. The average methane yields from the non-mixed and the mixed digester were 0.34 L CH4 at STP (g VS)-1 and 0.25 L CH4 at STP (g VS)-1, respectively. Higher sCOD and VOA accumulation was detected in the mixed digester which confirmed its depressed digestion performance. The 16s rRNA clone library analysis revealed a diverse microbial community for the non-mixed digester with phylotypes Methanoculleus and Methanosarcina being dominant methanogens. However, Methanosaeta was identified as the only methanogens at a very low abundance (2%) and a hydrogen-producing bacterial phylotype Petrotoga was detected with high abundance (70%) in the mixed digester. Dominance of Petrotoga was speculated to limit the substrate supply for aceticlastic Methanosaeta resulting in inadequate growth. The mixed digester showed higher methane production rate when inoculated with the non-mixed diges ter mixed liquor though the overall methane yield was not improved. A COD balance calculation suggested that a fraction of substrate may have been minera lized to hydrogen along with methane.

PAGE 89

89 Background Among various technologies that ar e available for anaerobic digestion, continuous ly stirred tank reactors (CSTR) are typically used to process high solid wastewater. Thorough mixing of digester contents helps particle size reduction and evolution of biogas, distributes micr oorganism uniformly and improves mass and heat transfer, and therefore is regarded as essential in efficient anaerobic digestion ( Sawyer and Grumbling 1960; Meynell 1978). Mixing is usually accomplished by mechanical mixers, slurry recircul ation or biogas recirculation ( Karim, Hoffmann et al. 2005). The signific ance of mixing in anaerob ic digestion has been reported in many studies (Smith, Elliot et al. 1996; Kim, Kim et al. 2000; McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001; Vavilin and Angelidaki 2005; Hoffmann, Garcia et al. 2008; Suryawanshi, Chaudhari et al. 2010). Fact ors affecting digesters mixing include mixing strategy (continuous or inte rmittent), duration and intensity, among which, the effect of mi xing intensity on anaerobic digestion are found contradictory. While some studies showed mixing improved biogas production for anaerobic digestion ( Karim, Hoffmann et al. 2005; Vavilin and Angelidaki 2005; Kaparaju, Buendia et al. 2008), opposite resu lts were reported by others ( Ghaly and Benhass an 1989; Chen, Chynoweth et al. 1990; Vedrenne, Beline et al. 2008). Mixing at high intensity was shown to result in delayed methane ( Vavilin and Angelidak i 2005) production whereas no signif icant difference was found in methane production for digesters with a br oad range of mixing intensity ( Hoffmann, Garcia et al. 2008). In spite of these disagreement s, most studies tended to agree that anaerobic digestion could be int erfered by exce ssive mixing while it may benefit from moderate mixing. The negative effect caused by intense mixing was interpreted as high shear forces disrupting microbial fl ock structures and disturbing syntrophic

PAGE 90

90 relationship between bacteria and methanogenes (McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001; Vavilin and Angelidaki 2005; Kaparaju, Buendia et al. 2008). Anaerobic digestion and mixing studies was mostly conducted on animal manure (in which most of the degradable portion is present as soluble organic matter), no studies were found investigati ng effect of mixing on anaerobic digest ion of solid feedstocks. On a per hectare basis, sugar beet is one of the most efficient sources of ethanol and is widely used in Europe for ethanol fermentation ( Kaffka 2009; Vries, Ven et al. 2010). However, no firm in United States has begun prodcuing ethanol from sugar beet ( Outlaw, Ribera et al. 2007). Effe ctive anaerobic digestion of sugar beet tailings using non-mixed system has been reported ( Liu, Pullammanappallil et al. 2008). The result showed a n et savi ng of treating sugar beet tailings with anaerobic digestion was around 4000 USD per day for the East Grand Forks plant, American Crystal Sugar Company which spends 1,000,000 USD annually to dispose the tailings ( Teixeira, Chynoweth et al. 2005) Non-mixed anaerobic digestion conserves energy and does not require fine shredding of the substrates. It can be conducted at ambient pressure and at both mesophilic and thermophilic temperatures. ( Pullammanappallil, Clarke et al. 2005) Howev er, when carrying out anaerobic digestion of sugar beet tailings, Polematidis ( Polematidis 2007) observed substrate compaction and/or flotation at non-mixed conditions and this negativ ely impacted digestion performance. A solution was found by providing bulking materials along with tailings. This approach elimi nated substrate compac tion and improved methane production rate substantially ( Polematidis 2007). Substrate compaction and flotation could also have been eliminated by providing gentle agit ation of digester contents. The objective of this st udy was to evaluate and compare anaerobic

PAGE 91

91 digestion of sugar beet tailings at non-mi xed and mixed conditions. The experiments were carried out in two single-stage batch systems operated at thermophilic temperature (55C). The non-mixed digester was added with bulking agents whereas the mixed digester was provided with lo w-speed continuous mixing. Digestion performance at both conditions, including CH4 yield, CH4 production rate and sCOD concentration were measured and compared. To further explore the digestion at mixed and non-mixed conditions, the microbial community structure was studied. Comparative analysis of bacterial and archael 16s rRNA was carried out for the two digesters and the major microbial phylotypes were identified. Material and Method Feedstock The sugar beet tailing s was provided by America Crystal Sugar Company, Minnesota and stored at 4 C before the experiment. The tailings were washed two times with tap water to soluble organic matter mainly sugars. Wash water was discarded. Anaerobic Digesters Two identical anaerobic digester s (d igester 1 and 2) were constructed by modifying 5 L Pyrex glass bottles. The digester was sealed on top with a glass lid fitted with a rubber ring. The lid was clamped using a stainless steel clamp for gas tightness. The height and inner diameter of the digester were 0.406 m and 0.061 m, respectively. The digester was provided with several ports for gas and liquid sample withdrawal and biogas exhaust. Biogas production was measured by a positive displacement gas meter, which consists of a clear PVC U-shaped tube filled with anti-freeze solution, a solid state time delay relay (Dayton Off Delay Model 6X153E),

PAGE 92

92 a Grainger float switch, a Redington counter and a Fabco Air solenoid valve. The gas meter was calibrated in line to determine the volume of biogas per count. Experiment Description The study was carried out in two sets consisting of six experimental runs which were conducted in succession with di gester liquors from the previous run being used to initiate the next run, as show n in Figure 4-3. Experim ent set 1 involved mixing digester liquors and exchanging betw een digester 1 and 2 while set 2 kept the digester liquor unchanged for the individual digester. In each run, digester 1 and 2 were operated at non-mixed and mixed condition, respecti vely. At beginning of a run, digester 1 was added 0.3 kg washed sugar beet tailings on a wet weight basis along with 200 g bulking materials (lava rocks from a landscaping supplier, 0.025 m in diameter averagely) to pr event substrate compaction and/or floatation. The tailings and bulking materials were added in the fo llowing alternative manner with 3 layers: tailing-rock-tailing-rock-tailing rock. The same amount of washed sugar beet tailings (0.3 kg on a wet weight basis) was added in digester 2. A 50.8 mm x 9.5 mm PTFE coated polygon bar was placed in the digester and the mixing of digester content was achieve by using a large volume magnetic stirrer (Siceneware Cool Stirrer). The mixing intensity was set as 100 revolutions per minutes (RPM). No bulking materials were used for digester 2. Both digesters we re kept in a 55 C incubator throughout the experiments. Three experimental runs were conducted in set 1. In run 1, each digester was inoculated with 3 L inoculumn taken fr om an anaerobic digester that has been fed with desugarized molasses at thermophilic temperature for 2 years. Run 1 was ended when the gas production was low and the digesters were emptied and washed thoroughly. Residual sugar beet tailings were discarded, and digester liquor

PAGE 93

93 was saved for run 2. Digester 1 and 2 liquors from run 1 were mixed and divided into equal amount and used as the inoculum for run 2. As for run 1, run 2 was initialed by adding 0.3 kg fresh washed tailings and inoculated with 3 L inoculumn. When the biogas production stopped run 2 was ended and the digesters were again unloaded and washed. Residue tailings were discarded wh ile digester liquors were kept for the next run. In run 3, new fresh substrates were added and digester liquors from run 2 were exchanged and used as the inoculum, t hat is, digester 1 was inoculated with digester 2 liquor and digester 2 was inoculated from digester 1 li quor from the last run. Run 3 was considered complete when the biogas production was low. Additionally, three experi mental runs were conducted in set 2. Digester 1 liquors was recovered from run 3 and diluted by an equal volume of deionized water. The diluted digester 1 liquor wa s then divided equally and used to inoculate digester 1 and 2, respectively. In run 5 and 6, diges ter 1 and digester 2 were added with 0.3 kg fresh washed tailings and inoculated with 3 L inoculum as before. Digester 1 liquors from run 4 and 5 were used to inoculate digester 1 to start run 5 and 6, respectively. Digester 2 was inoculated in the same m anner. Each experimental run was operated for 15 to 20 days. Temperature Monitor A large-volume magnetic stirrer was used to provide mixing for digester 2. According to the manufacturer, the stirrer motor is insulated to prevent the heating from being transmitted to thermolabile solu tion. To demonstrate th at digester 2 liquor was not overheated by the us e of the stirrer, a simp le temperature monitoring experiment was carried out. Th re e mercury thermometers were used to measure the temperature in the incubator, digester 1 and digester 2. Temperatures were read manually. Observations were made periodi cally, ranging from minutes to hours.

PAGE 94

94 Chemical Analysis Total Solids (TS) and Volatile Solids (VS) contents were determined for sugar beet tailing. Tailing samples were dr ied at 105C for 24 hours using a Fisher Scientific Isotemp Oven (Model 350G) and co mbusted at 550 C for 3 hours using a muffle furnace (Fisher Scientific Isotemp) for TS and VS measurement, respectively. Daily biogas production, biogas composit ion, pH and soluble chemical oxygen demand (sCOD) concentration were monitored for digester 1 and 2. Biogas volume was measured by the displacement gas me ter and biogas composition was analyzed using Fisher Gas Partitioner (Model 1200) The gas chromatograph was calibrated with an external standard containing N2, CH4, and CO2 in 25:45:30 volume ratio. Digester 1 and 2 liquor samples were collect ed periodically for pH, sCOD. pH was measured using an Accument pH meter. Fo r sCOD analysis, liquid samples were centrifuged at 8000 RPM for 10 minutes (F isher Marathon micro H centrifuge) and filtered using Millipore filter paper (pore size 1.2 um). Prepared samples were added in Hach COD vials (range 20-150 ppm), bak ed in Hach COD reactor for 2 hours and measured using Hach DR/890 Colorimeter for sCOD concentration. Volatile organic acid (VOA) concentration was monitored from run 1 through run 3. The analysis was conducted using Shimadzu gas chromatograph (GC-9AM equipped with a flame ionization detector) for acet ic, propionic, isobutyric, but yric, isovaleric and valeric acid concentrations. Microbial Community Analysis Microbial communities were investigat ed for the original incolumn (run 1) and digester 1 and digest er 2 liquors in run 6. Digester liquors were sampled for both digesters at day 3 when the methane produc tion was active. DNA was extracted and purified by using FastDNA Kit and PowerCl ean DNA kit, respectively. The quality of

PAGE 95

95 DNA was verified by agar gel electrophores is running at 100V for approximately 50 minutes. Results were visualized in UV li ght. Extracted DNA was stored at -20 C before using as a template for Polymerase Chain Reaction (PCR ) amplification.PCR reaction was triplicated for each sample and was done in a mixture of 28 ul of RNase free water (provided by QIAGEN), 0.5 ul of forward primer, 0. 5 ul of reverse primer 20 ul of HotMaster solution (provided by 5 Prime) and 0.5 to 2 ul of template depending on its concentration. The total am ount of template added was below 20 ng. Primers Archaea A1 (5GCCTTGCCAGCCCGGCTCAGAAGCCGTT TCATTAGATACCCA-3) and Archaea Rev (5-GCCTCCCTCGCGCGATCAGTCTTM GGGGCATTCNKACCT-3) were used for archaeal 16S rDNA amplif ication. Primers 454B 27F (GCCTTGCCAGCCCGCTCAGTCAGAGTT TGATCCTGGCTCAG-3) and 454A 338R (5-GCCTCCCTCGCG CCATCAGCATGCTGCCTCCCG TAGGAGT-3) were used for bacterial 16S rDNA amplification. All primers were provided by Invitrogen. The PCR reaction was conducted using Eppendorf Mastercycler at the following program: an initial denaturat ion at 94 C for 2 mininutes; 30 cycles of denaturing at 94 C for 20 seconds, annealing at 53 C for 20 sec onds, extension at 65 C for 1 min; end of extension at 65 C for 6 min. Amplifi ed products were purified using QIAquick PCR purification kit and t he quality was verified by agar gel electrophoresis. The purified fresh PCR products were ligated into p CR4-TOPO vectors and then transformed into chemically com petent E.coli DH5 -T1 cells by using Invitrogen TOPO TA Cloning Kit. E.coli colonies were screened on Lur ia Broth agar plates with 50 ug/ml kanamycin. 96 colonies for each sample we re randomly selected into a 96 welled plate preloaded with 200 ul LB medium (50 ug/ml kanamycin). Archaea population was expected to have less diversity and only half as many as clones as for bacteria

PAGE 96

96 were developed for archaea. Samples were s ent to Interdisciplinary Center for Biotechnology Research (ICBR) at University of Flori da for 16s rRNA sequencing. Sequence similarity searches were perfo rmed using BLAST in the RDP (Ribosomal Database Project) database. Phylotype was defined as a group of cloned sequences with 90% identity (taxonomic order level). The percentage of coverage was calculated using the equation [1-(n/N)], w here n is the number of phylotypes represented by a single clone and N is the total number of clones obtained ( Good 1953). Results Characterization of Substrates TS and VS contents of sugar beet tailings were determined for 6 runs and shown in ta ble 4-1. The volatile matter conten t of tailings used in run 1, run 2, run 3, run 4, run 5 and run 6 were 10.68%, 9.23% 8.19%, 10. 53%, 10.17% and 10.26% in run 1, run 2, run 3, run 4, run 5 and run 6 respective ly. Substrate loading quantities and digester packing density on wet weight and dry matter basis were calculated for each run and also listed in Table 4-1. Temperature Profiles The temperature profiles of incubator digester 1 and digester 2 are illustrated in Figure 4-3 There was no noticeable differe nce between digester 1 and digester 2 temperature. Temperatures of incubat or, digester 1 and digester 2 dropped when measurements were taken withi n a short time interval of 10 to 20 minutes. This was because incubator had to be opened frequently to make measurements which cooled the chamber, thus lowering the temp erature.. Both digester 1 and digester 2 showed slightly lower temperatures than that of incubator possible due to the slow

PAGE 97

97 evaporation of digesters liquor s. Generally, temperatures in digester 1 and 2 were maintained well within thmermophilic range. CH4 Production The profiles of cumulative methane yield for digester 1 and 2 in six runs are shown in Figure 4-4. In run 1, digester 1 achieved CH4 yield of 0.36 CH4 at STP kg VS-1 and digester 2 achieved 0.23 m3 CH4 at STP kg VS-1. The maximum CH4 production rate of digester 1 and 2 was 1.3 m3 m-3 d-1 and 0.61 m3 m-3 d-1, respectively. The maximum rate was achiev ed sooner in digester 1 (at day 3) than in digester 2 (at day 4). Run 1 was ended when daily CH4 production rate was low (0.035 m3 m-3 d-1 for digester 1 and 0.030 m3 m-3 d-1 for digester 2). Run 2 was started by flooding digester 1 and 2 with mixed digester 1 and 2 liquors from run 1. The CH4 yields obtained were similar to those obtained in run 1. In this run digester 1 and 2 achieved CH4 yields of 0.35 m3 CH4 at STP kg VS-1 and 0.27 m3 CH4 at STP kg VS-1, respectively. Daily methane production rate peaked at 1.24 m3 m-3 d-1 on day 3 for digester 1 and 0.48 m3 m-3 d-1 on day 5 for digester 2. In run 3, the CH4 yields of digester 1 and 2 were 0.34 m3 and 0.25 m3 CH4 at STP kg VS-1, respectively. Daily CH4 production rate reached maximum on day 3 for digester 1 (0.90 m3 m-3 d-1) and day 4 for digester 2 (0.47 m3 m-3 d-1). For run 4, 5 and 6, digester obtained CH4 yields round 0.36 STP kg VS-1 and maintained the maximum CH4 production rate at the level from 0.8 to 1.1 m3 m-3 d-1. On the contrary, digester 2 showed lower CH4 production rates than in first 3 r uns, though the cumulative yields CH4 were similar. Its maximu m production rate at run 4, 5 and 6 were calculated around 0.25 to 0.32 m3 m-3 d-1. In all six runs digester 1 exhibited noticeably higher CH4 yield and CH4 production rate than digester 2.

PAGE 98

98 Profiles of sCOD Degradation Comparison The profiles of sCOD for digester 1 and 2 are shown in Figure 4-5. sCOD comparisons were made between digester 1 and 2 within a certain run and also for a digester in different runs. In general, sCOD concentration of both digesters initially increased, reached a maximum and decreased to a minimum. The initial increase of sCOD was due to the solubilization of sugar beet tailings. Digester 1 exhibited similar sCOD profiles in all 6 runs whereas in digester 2 much higher sCOD accumulations was seen in run 4, 5 and 6 than in run 1, 2 and 3. In the first 3 runs, sCOD concentration in digester 2 increased to the range of 5 to 6 g L-1. However, for run 4, 5 and 6, the sCOD concentration in digester 2 reached a maximum of 6 to 10 g L-1 and seemed to persist for longer time bef ore noticeable degradatio n. Within a run, digester 2 had accumulation of higher sCOD concentration and slower sCOD degradation than digester 1. sCOD accu mulation of digester 1 was around 4 g L-1, and degradation became evident after 2 to 3 days. Profiles of Volatile Organic Acid The profiles of VOA for digester 1 and 2 in run 1, 2 and 3 are shown in Figure 4-6 Acetic acid, propionic acid and butyric acid were dominant among the organic acids detected. The concentration of isobutyric acid, isovaleric acid and valeric acid were negligible and are not sh own in Figure 4-5. Acetic acid had the highest concentration among VOAs in digeste r 1. The concentration reached around 1 g L-1 and then quickly decreased as CH4 was produced. High concentration of acetic acid was also observed in digester 2 and the degradation appeared slower than in digester 1. Profiles of propionic acid were distinct between digester 1 and digester 2. Much higher propionic acid accumulation (maximum around 0.8 ) was seen in digester 2 than in digester 1. T he persistence of propionic acid probably

PAGE 99

99 indicated the hindered digestion. Concentration of propionic acid of digester 1 in run was initially high, because the digeste r was inoculated with digester 2 liquor recovered from run 2. The pr opionic acid started degrading shortly after run 3 began. Microbial Communities Structure Bacterial and archaeal 16S rRNA gene clone libraries were constructed for the original incolum and digester 1 and 2 liq uors at run 6 and pres ented in Table 4-3 and 4-4. A total of 215 bacterial phylo types were obtained from 271 16S rRNA cloned sequence and 125 archaeal phylotype s were obtained from 128 clones with the criterion greater than 90% identity The comparative 16S rRNA gene analysis revealed that microbiological communities were distinct among the inoculum, digester 1 liquors and digester 2 liquors. Bacteria phyla Actinobacteria, Firmcutes Synergistales and Thermotogae were abundant in the inoculum with Thermotogae being dominant (52%), which has been identifi ed in thermophilic anaerobic digesters ( Riviere, Desvignes et al. 2009; Weiss, Jerome et al. 2009). The phylum Ther motogae was also dominant in digester 2 (70%) and abundant in digester 1 (22%). Its closest rela tive was identified as Petrotoga olearia with a low similarity of 90%. In digester 1, bacterial phylum Firmcutes was found to be dominant and genera Caloramator Clostridium and Sporobacterium were identified within the phylum. Firmcutes have been commonly observed in anaerobic processes treating organic wastes ( Hatamoto, Imachi et al. 2007; Narihiro, Terada et al. 2009; Sasaki, Hori et al. 2011). Some phylotypes detect ed in digester 1 were either not found (classified in the phylum Proteobacteria) or detected with lo w abundance (clas sified in the phylum Synergistets) in digester 2. As expect ed, the identified archaeal phylotypes were less diverse than bacteria, and all the archaea clones are member in phyla Crenarchaeota and Euryarchaeota The dominant archaeal phylotypes for

PAGE 100

100 inoculum, digester 1 and digeser 2 was surprisingly found to be Crenarchaeota to which methanogens does not belong. Phylotypes.within Euryarchaeota were Pyrobaculum -like phylotypes, which were reported to be facultative anaerobes and grow at thermophilic tem perature and neutral pH. However, they have not been reported found in anaerobic reactors. In spite of Crenarchaeota being dominant, methanogenic phylotypes were still found abundant in the inoculumn (37%) and digester 1 (46%). Methanogenic phylotypes in digester 2 were identified with high identity (95%-99%) and classified in order Methanobacteriales Methanomicrobiales and Methanosarchinales while phylotypes in the inoculum were all classified in the order Methanomicrobiales In contrast, methanogenic phylotypes in digester 2 were detected at much less abundance of 2% They were assigned to the order Methanobacteriales Overall, digester 1 showed more diverse microbial communities than digester 2. This may explain the low methanogenic activity observed in digester 2. The rank abundances of bacterial and ar chaeal phylotypes are shown in Figure 47. Discussion Determination of Mixing Intensity Utilization of mixing at high intens ity has been shown to be detrimental to anaerobic digestion ( McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001; Vavilin and Angelidaki 2005; Kaparaju, Buendia et al. 2008). The mixing rate used in digester 2 was carefully chosen based on recommendations in literature. Hoffman ( Hoffmann, Garcia et al. 2008) in vestigated the effect of mixi ng at different intensities of 50, 250, 500 and 1500 RPM on anaerobic di gesters (4.5 L working volume) fed with 5% VS based cow manure slurry. Differences in the CH4 yields at four mixing intensities were statistically insignific ant. Computer automated radioactive particle

PAGE 101

101 tracking (CARPT) in conjunction with computational fluid dynamics (CFD) was utilized to map shear dist ribution throughout the digesters and estimate local velocities. As expected, the digester mi xed at 250 RPM showed lowest vertical, horizontal and azimuthal velocities (CAR PT was not successful for the 50 RPM digester due to the low mixing intensity). Shear stresses were also lowest and uniformly distributed in the 250 RPM digester. High local shear stresses were present near the mixer regi on for all analyzed digesters. Using these results as references, the mixing rate employed in this study was chosen to be 100 RPM. The rate only developed a small vortex and dispensed the tailings well (no floatation or settlement) in digester 2. It was expected to aid in particle size reduction and mass transfer but avoid excessive mixing. However, digester 2 showed lower CH4 yield and CH4 production rate than the non-mixed par allel (digester 1), though the 100 RPM was close to the low mixing intensity used in Hoffmans study. Since digester 2 has a working volume smaller than the diges ter used by Hoffman, it is possible that 100 RPM could still introduce intense mixi ng to the smaller system used here. To compare the mixing intensities used in differ ent studies, it is necessary to develop a method for normalization because mixing in anaerobic digestion at different scales has been achieved by various mechanical wa ys such as impeller mixing and shaking (using a shaking table). The comparison was made between this study and other four studies ( Karim, Hoffmann et al. 2005; Vavilin and Angelidaki 2005; Hoffmann, Garcia et al. 2008; Kaparaju, Buendia et al. 2008). The mixing intensities was normalized to power supply per unit volume, W/m3 according to equation 4.1 and 4.2, which are used for impeller mixing and shaking, respectively. 53' iiPDNNP (4.1) where PN = Power number, dimensionless

PAGE 102

102 = density of digester liquor, kg/m3 iN =Rotational speed, s-1 iD =Impeller diameter, m smadsP0 (4.2) where m= mass, kg a= acceleration, m/s2 s= displacement, m 22 1 mvE (4.3) where E=kinetic energy m= mass v = velocity The power numbers were assumed to be 0.35 (paddle impeller) for the impeller used in Hoffman and Karims studies ( Karim, Hoffmann et al. 2005; Hoffmann, Garcia et al. 2008) and also for the mixer (a polygon bar) used in this study. This assumption was made as the power number i s independent of Reynolds number and only depends on impeller geometry when turbulence was achieved. Reynolds numbers were calculated and show ed turbulence was achieved, except for the 50 RPM-mixed digester used by Hoffman. Although the 50 RPM digester was determined being at transition state, Reynol ds number of 0.35 wa s still assigned. All digester liquors were assumed to have the same density and viscosity as water for Reynolds number calculation. Equation 4.2 was used to normalize mixing intensity reported in Kaparajus study ( Kaparaju, Buendia et al. 2008). Kaparaju carried out experiments in 1 L serum bottles (450 ml working volum e) and the bottle contents

PAGE 103

103 were mixed by using a shaker table. Mi xing was provided in three intensities, vigorous (100 times per minute with a 3.5 cm stroke), gently (35 times per minute with a 1.2 cm stroke) and minimum (manually shaking the bottles every time biogas sample were withdrawn). Normalization was calculated for vigorous and gentle mixing only. Shaking bottles were assum ed to have a symmetric parabola shaped velocity profile against time, where velocity is zero at beginning and end and reaches maximum at midpoint of a vibrating cycle. Using this assumption, acceleration of serum bottles can be calculated. The exper iment (digestion of diluted municipal household solid wastes) carried out in Vavilins study ( Vavilin and Angelidaki 2005) was very similar to Kaparajus method. Shak er tables were used to provide vigorous (105 times per min with a 5cm length) and gentle mixing (58 round per minutes with a 17 cm radius) while minimal mixing wa s c onducted by hand shaking for 1 minute every 1 to 2 days. The only difference was t hat bottles at gentle mixing were shaking in a pattern of approximate circular moti on instead of linear vibr ation. Equation 4.2 and the same assumption were first used to ca lculate power input per unit volume for bottles at both gentle and vigorous mixing in Vavilins study. The results showed gentle mixing provided orders of magnit ude more power than vigorous mixing. It clearly overestimated the power supply of gent le mixing. It is true that bottles at gentle mixing moved much longer distance (a circle with 0.17 cm radius ) that the bottles at vigorous mixing (0.05 m) in a shak ing cycle, but velocity of circular motion was expected to be relatively constant which would consume much less energy as required by linear vibration to change the ve locity frequently. Therefore, a different assumption was made that the bottles at gent le mixing in Vavilins study underwent uniform circular motion, and the kinetic energy can be simply calculated using Equation 4.3. To calculate the corresponding power inpu t, the kinetic energy was

PAGE 104

104 assumed to be consumed in 3 shaking cycles (3 rounds or 2.9 seconds). The normalization results are shown in Table 4-4. In Karims study ( Karim, Hoffmann et al. 2005), the power supply per unit volume has been provided for mixed dige sters. The equation P=TA (where T is torque and A is angular speed) was used to calculate the power input per unit volume (8 W/m3) for the digester that was provided with impeller mixing. Though a different equati on was used in this study to calculate the unit power input (9.12 W/m3) for the same digester, th e result was in accordance with Karims calculation. The power i nput per unit volume of digester 2 was calculated to be 1.10 W/m3, which was not regarded as a high value compared to other results. In Karims study, the im peller mixed digester produced 22% more CH4 than the unmixed one when 10% (TS based) manure slurry was fed. This suggested that the low CH4 production seen in digester 2 was not due to the excessive mixing as reported in some studies ( McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001; Vavilin and Angelidaki 2005; Kaparaju, Buendia et al 2008). The unit power input of 1.10 W/m3 was also close to the unit power i nput at gentle mixing in Vavilins study. However, in Kaparajus studies, t he gentle mixing showed a very low unit power input of 0.06 W/m3. In comparison to that, 100 RPM used in digester 2 exceeded degree of gentle mixi ng though it was regarded as low-intense mixing according to Karim, Hoffman and Vavilin. Contradiction was found in comparing the mixing intensities between shaking and impe ller mixing. Kaparaju reported a digester mixed at 12.67 W/m3 showed low and delayed CH4 production compared to digestion with less intense mixing (0.06 W/m3 or no mixing); while Karim reported improved CH4 production in digestion mixed at 9.12 W/m3 compared to the nonmixed counterpart. It appeared that shak ing and impeller-like mixing could have opposite effects on anaerobic digestion ev en though the unit power inputs were

PAGE 105

105 comparable. A pilot scale anaerobic digestion of cow manure was further carried out using an impeller equipped reactor in Kaparajus study. The mixing was provided at unit power input of 312.5 w/m3, significantly exceeding t he unit power input for the vigorous mixing (12.67 and 28.69 w/m3), which was suggested to be avoided in anaerobic digestion by Kaparaju and Vavilin Though there was a minor discrepancy in determining the mixing used in digester 2 whether being high-intense or lowintense, many studies agreed that mi xing around 100-200 RPM was not viewed as excessive mixing and was typically applie d to facilitate anaerobic digestion ( Karim, Hoffmann et al. 2005; Hoffmann, Garcia et al. 2008; Alkaya and Demirer 2011; Alkaya and Demirer 2011; Penteado, Santana et al. 2011; Yu, Ma et al. 2011). In fact mixing power provided by the magnetic mixer for digester 2 (1.10 w/m3) is well within the range of typical volumetric power input (1 to 10 w/m3) recommended for anaerobic digesters ( Grady, Daigger et al. 1999; Stroot, McMahon et al. 2001). The mixing intensity applied in digester 2 wa s therefore expected not to cause the disruption of spatial asso ciations between microorganisms and ineffective transfer of hydrogen and acetate from syntrophic ac etogens to neighboring methanogens as reported in vigorously mixed systems ( Conrad, Phelps et al. 1985; Whitmore, Lloyd et al. 1987; Stroot, McMahon et al. 2001; Hoffmann, Garcia et al. 2008; Kaparaju, Buendia et al. 2008). Another explanation for the hindered per formance at the mixed c ondition could relate to the difference in substr ate distribution. Mixi ng may promote rapid hydrolysis and fermentation, while methanogens and syntrophis may not have been able to turn over fermentation products at t he rate of formation due to the inhibitory effect of some VOAs, resulting in poorer digest ion performance (McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001). Digeste r 2 was observed to accumulate

PAGE 106

106 propionic acid at concentrations which has been reported to inhibit methanogenesis ( Barredo and Evison 1991). In addition to mixing intensity, mixing frequen cy was also expected to affect anaerobic digestion. Continuous mixing (mi xing using a shaker table) has been shown to be detrimental for anaerobic r eactors digesting municipal waste (Grady, Daigger et al. 1999; McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001). Similar to digester 2, continuous ly mix ed digesters in these studies exhibited persistent accumulation of propionic acid and less abundant of methanogens compared to the minimally mixed counterparts (manual shaking for 2 min everyday). McMahon and Stroot conclud ed continuous mixing appear ed to disrupt spatial juxtaposition of syntrophic bacteria and t heir aceticlasic and hydrogenotrophic partners and resulted in unstable digestion performance. In comparison with this study, persistence of propionic acid was much more severe in McMahon and Stroots study. This can be attributed to that they used a significantly higher mixing power of 1500 W/m3. However, the effect of mixing in tensity was not discussed in their studies. While several studies has showed anaerobic digestion at non-mixed or less mixed conditions (i.e., inte rmittent mixing instead of c ontinuous mixing ) exhibited higher CH4 yield than the continuously mixed digestion others has found no significant difference in performance between completely and minimally mixed digestion ( Dague, McKinney et al. 1970; Chen, Chynoweth et al. 1990). Furthermore, mixing has showed to hinder industrial scale anaerobic digestion as well ( Pfeffer 1987). These findings seemed to support the results obtained in this study, suggesting practice of continuous mixing fo r high solid digesters may not be always feasible since the possible inhibitory effect and operational problems.

PAGE 107

107 Digestion Performance Comparison In all six runs, digester 1 showed consistent CH4 yield, CH4 production rate and sCOD profiles. Digester 2 produced lower CH4 yield and CH4 production rate compared to that of digeste r 1. In run 4, 5 and 6, CH4 production rate of digester 2 was further decreased though the final yield CH4 was kept similar. It seemed that digester 2 performance deteriorated in the la st 3 runs where the digester liquor was reused as the inoculum for the following r un. This was confirmed by the high sCOD accumulation and long persistence period be fore degradation (6-8 days) observed for digester 2 in run 4, 5 and 6. Digester 2 appeared in capable of degrading sCOD to the extent achieved in digester 1. During anaerobic digestion, solid subs trate (sugar beet tailings) is first hydrolyzed to soluble forms by extrac ellular enzyme before cellular uptake. Fermentation of soluble organic matter produces a mixture of volatile organic acids. sCOD measurements include both the hydrolyzed organic matter as well as organic acids. Digester 2 was only able to convert a fraction of sCOD (large difference seen between initial and final sCOD) and the rest remained in the digester liquor. Because in run 4, 5 and 6, the digeste r liquor was used for inoculation for the next run, the unconverted sCOD from the previous run was detected as init ial sCOD in the following run. Approximately 2 g L-1 of sCOD was left non-degraded in each run and accounted for the initial sCOD, resulting in higher and higher sCOD accumulation from run 4 through run 6. This phenomena was not seen for digester 2 in run 1, 2 and 3, because digester 1 and 2 liquors were either mixed or exchanged before inoculation. Providing digester 2 with digeste r 1 liquor (partial or all) also seemed to improve the rate of sCOD degradation, whic h started after about 5 days in run 1, 2 and 3. This agreed with generally with higher CH4 production rate in run 1, 2 and 3

PAGE 108

108 than in run 4, 5 and 6 for digester 2. In r un 3, despite the fact that digester 1 was inoculated with digester 2 liquor, it neither showed delayed CH4 production rate nor poor sCOD degradation. Speculation was made that microorganisms involved in sCOD degradation and methanogensis could quickly adapt to a stagnant environment and help to recover CH4 production quickly in spite of having been exposed to the mixing conditions. A st udy has showed anaerobic digestion can be improved by providing a quiesc ent environment for bacteria ( Stroot, McMahon et al. 2001). Addition of the bulking age nts in diges ter 1 was also expected to facilitate the development of microbes by preventing compaction and providing surface for concentrated growth. Microbial Community In the development of 16s rRNA library, t he number of si ngletons (phylotypes represented by a single clone) were found few, leading to high coverage values ( Good 1953). Bacteria phyla Proteobacteria and Synergistets were identified in digester 1. Species in these phyla are known as syntrophic bacteria that play critical roles in methanogens is. For instance, Syntrophobacter classified in Proteobacteria was identified as a key player in syntrophic propi onate degradation ( Botsch and Conrad 2011). While the function of Synergistets in anaerobic ecos ystem remains largely unknown, they are suspected to be involved in amino acid degradation and function as a biological electron acceptor in syntrophy with hydrogenotrophic methanogens ( Godon, Moriniere et al. 2005). More than 50% bacterial phyloty pes in digester 1 were members of phylum Firmcute in which, phytotypes Caloramator Clostridium and Sporobacterium were found with high abundance. Studies have reported that some Clostridium species were capable of syn trophic long chain fatty acid degradation ( Hatamoto, Imachi et al. 2007) while Sporobacterium olearium (the

PAGE 109

109 close representative of phylotype Sporobacterium ) utilized aromatic compounds to produce acetate and butyrate ( Mechichi, Labat et al. 1999). This may indicated those microbes played an intermediate role in anaerobic digest ion, providing substrates for small-molecule degraders. Caloramator fervidus (the close representative of phylotype Caloramator ) was found as one of the dominant hydrogen-producing bacteria in anaerobic digesters (Kongjan, O-Thong et al. 2011). It is likely to provide hydrogen to hydrogetrophic methanogens as genera Methanoculleus and Methanosarcina (both aceticlastic and hydrogenrophic) found in digester 1. Some Clostridium species were also reported to ut ilize acetate in syntrophic association with hydrogenotrophi c methanogens ( Schnurer, Schink et al. 1996; Hattori, Kamagata et al. 2000). Phylotype Thermotogales were found dominant (70%) in digester 2 with its closest representative being Petrotoga which was also found high abunda nce(53%) in a lab-scale thermophilic continuous-flow stirred-tank (CFTR) reactor fed with a commercial dog food ( Sasaki, Hori et al. 2011). Phylotype Petrotoga as well as another Thermotogales member Ther motoga have been known as fermentative and sulfur reducing bacteria. Some Petrotoga and Thermotoga species were reported to produce H2 through sugar fermentation ( Lien, Madsen et al. 1998; Balk, Weijma et al. 2002). Inhibition of H2 at high partial pressure has been showed inhibitory on growth of a Petrotoga species. The inhibition was alleviated by adding elem ent sulfur or thiosulfate because the bacterium was able to use them as elec tron acceptors to produce hydrogen sulfide (H2S) and reduce H2 concentration. The presence of abundant Petrotoga -like bacteria in both digester 1 and 2 could be a ttributed to the inoculum, in which the bacteria were already found dominant. The in column was taken from an anaerobic reactor digesting desugarized molasses that has been reported to contain relatively

PAGE 110

110 high sulfate concentration ( Fang, Boe et al. 2011). This was speculated to favor the bacteria growth in the inoculumn. Interestingly, a Petrotoga relative, Thermotoga lettinga was reported to grow through H2 fermentation in the absence of electron acceptors (as sulfur) by taking advant age of the syntrophic relation with a hydrogenotrophic methanogen Methanothermobacter (Balk, Weijma et al. 2002). Though Petrotoga was remotely related to (belong to the same order) Thermotoga it could shar e a similar syntrophic relation with methanogens as well. However, hydrogenotrophic methanogens were not i dentified in digester 2. The only methanogenic phylotype was Methanosaet a, which has been determined obligately aceticlastic ( Patel and Sprott 1990). This indicated in digeser 2 Petrotoga was unable to gain metabolic advanta ges in the community because Methanosaet a cannot maintain a low H2 partial pressure. Therefore, syntrophic relation between Petrotoga and Methanosaet a was not expected for digester 2. As a result, Petrotoga had to grow through H2 fermentaion pathway. The high H2 generation ability of Petrotoga has been reported in several studies ( Lien, Madsen et al. 1998; L'Haridon, Miroshnichenko et al. 2002; Miranda-Tello, Fardeai et al. 2007; Kano, Mukaidani et al. 2009). This could af fect VOA degr adation which requires very low H2 concentration ( Bitton 2005). It was possible that acetate formation was reduced in digester, leading to ins ufficient growth (3% relative abundance to all archaea) of Methanosaet a and low CH4 yield and production rate. Some Petrotoga related bacteria ( Thermotoga neapolitana ) exhibited stimulated growth and greatly improved hydrogen production in a agitated envir onment (75 rpm, 50 ml medium) ( Van Ooteghem, Jones et al. 2004), which coul d exacerbate the acetate scarcity in digester 2. In contrast, digester 1 can benefit from Petrotoga since hydrogentrophic methanogens use H2 as the substrate. As discussed above, sulfur reducing bacteria

PAGE 111

111 are also autotrophic and they could compete with methanogens for H2 utilization. Dar et al reported methanogens wa s outcompeted by sulfur r educing bacteria in a coculture with sulfate concentrati oin at 1.5 g/L. The compet itive yet syntrophic relation between sulfur reducing bacteria and met hanogens makes their interaction quite complex and not well understood ( Dar, Kleerebezem et al. 2008) Discovery of sulfur reducing bacteria in anaerobic digestion was confirmed in a similar study, in which sugar beet pulp was ut ilized as feedstock ( Labat and Garcia 1986). Archaeal phylotype Pyrobaculum was found dominant in digeste r 2 with a relative abundance of 97%. Amo et al. isolated a Pyrobaculum speices ( Pyrobaculum calidifontis ) from a hot spring and demonstrated significantly st imulated growth and in a 200 ml medium with 120 rpm shaking ( Amo, Paje et al. 2002). This probably suggested the high abundance of Pyrobaculum was due to the mixing provided by diges ter 2. Pyrobaculum species were reported to grow heterotrophically via sulfur/sulfate oxidation ( Amo, Paje et al. 2002). This c ould explain the high abundance of Pyrobaculum in the inoculum and digester 1 as well, since the inc oculum was from a digester fed substrate contai ning high sulfate concentration COD Balance During ana erobic digestion, the degr adable components of tailings were solubilized (hydrolyzed) and then converted to CH4. sCOD was measured in the experiments. CH4-COD was calculated using the c onversion factor that 1 L CH4 at STP has a COD equivalent of 2.86 g. A previous study determined the degradable COD of 0.3 kg sugar beet tailings (a verage 10% VS content) to be around 31.67 g using BMP assays ( Polematidis 2007). This value wa s used as the in itial degradable COD in the solid substrates for all 6 runs. The remaining degradable COD in the solid at a specified time was the diffe rence between initial degradable COD and sum

PAGE 112

112 of measured CH4-COD and sCOD in the same peri od. By plotting, solid degradable COD, sCOD and CH4-COD against time, the COD balance profiles of digester 1 and 2 were developed for experiment set 1 and 2, respectively (F igure 4-8). The tailings used in run 3 were determined to have rela tively low VS contents and produced an approximately 20% less CH4 (yield times VS) than in run 1 and 2. COD balance for experiment set 1 was calculated using the data from run 1 and 2 only to avoid large errors. Digesters from run 1 were inoculated with the liquor taken from a digester fed with different substrates (desugarized molass es). When run 1 liquors were used as inoculum for run 2, they had been adapted to degrade the tailings. Therefore, higher initial CH4 production rates (day 2 to day 4) were seen for digesters in run 2 than in run 1, especially for digester 1. The variation in initial CH4 productions mainly resulted in errors seen at day 2 and day 4 for experiment set 1. In both experiment sets (run 1, 2 and run 4, 5, 6), hydrolysi s seemed to proceed well for both digester 1 and 2, since most of the degradable COD was converted (either as sCOD or CH4COD) at the end. While degradable COD left in digester 1 was negligible, 2-3 g of degradable COD stayed in digester 2. The comparative 16s rRNA library analysis has showed presence of a hydrogen-producing bacterial phylotype Petrotoga with high abundance (73%) in digester 2. The production of hydrogen could account for the discrepancy since hydrogen content of biogas was not measured. Hydrogen oxidization stoichiometry shows 1 L hydrogen consumes 0.72 g oxygen. To compensate for the missing COD of 2-3 g, digester 2 would produce 3-4 L hydrogen gas, or around 30% of biogas would be hydrogen (volume/volume, CO2 excluded). Petrotoga was reported to have a doubling time of 12 hours ( Lien, Madsen et al. 1998) and had the potential to outpace the growth of Methanosaet a (the only methanogen in digester 2) with doubling time being 3 to 5 days ( Janssen 2003;

PAGE 113

113 Yoochatchaval, Ohashi et al. 2008). Theref ore, hydrogen production would be faster when usable substrates as glucose became available to Petrotoga This agreed with the observation of a rapid H2 production during an anaer obic digestion of waste lactose ( Banks, Zotova et al. 2010). As seen in the sCOD profiles, COD balance also showed high sCOD accumulation in diges ter 2. VOA made up approximately 2.5 g COD equiv alent at the end of the experiment. The rest of sCOD was expected to be other fermentation intermediates such sugars. It appeared the presence of H2 not only affected VOA degradation but also sugar fermentation. Closing Remarks Anaerob ic digestions of sugar beet tailings were performed at non-mixed and mixed conditions for 6 runs using two batch anaerobic reactors. The mixed condition caused delayed CH4 production rate and overall low CH4 yield in mixed digester (digester 2). The comparative 16s rDNA library analysis revealed Petrotoga and Pyrobaculum dominated in bacteria and archaea community, respectively, in digester 2. The abundance of methanogens ( Methanosaeta as the only identified phylotype) was very low. In contrast, digester 1 was detected with a diverse group of methanogens including hydrogenotrophic Methanoculleus and acetocalstic Methanosarcina (also hydrogenotrophic). The mixing intensities in digester 2 were not considered as high intensity. The hindered digestion performance was attributed to the inhibition by H2. Digester 1 and 2 were added the same quantities of substrate at the beginning of each r un, but the COD balance ca lculation showed 2 3 g substrate COD missing in digester 2 ( neither converted to sCOD nor CH4). It was hypothesized that digester 2 produced hydr ogen which was not measured in this study, resulting in the discrepancy.

PAGE 114

114 Table 4-1. Substrate characteristi cs and loading amounts for digester 1 and 2 Sugar beet tailings Unit Run 1 Run 2 Run 3 TS % (wt/wt) 11.50%.39%10.22%.23% 8.97%.43% VS % (wt/wt) 10.68%.56% 9.23%.19% 8.19%.46% Digestion loading Unit Wet weight kg 0.3 0.3 0.3 Dry weight kg 0.034 0.031 0.027 Volatile matter kg 0.032 0.029 0.025 Inoculumn added L 3 3 3 Packing density kg/m3, wet weight basis 100100 100 Packing density kg/m3, dry weight basis 11.4210.42 8.97 Sugar beet tailings Unit Run 4Run 5 Run 6 TS % (wt/wt) 11.22%.49%10.88%.72% 11.00%.17% VS % (wt/wt) 10.53%.45%10.17%.66% 10.26%.14% Digestion loading Unit Wet weight kg 0.3 0.3 0.3 Dry weight kg 0.034 0.033 0.033 Volatile matter kg 0.032 0.031 0.031 Inoculumn added L 3 3 3 Packing density kg/m3, wet weight basis 100100 100 Packing density kg/m3, dry weight basis 11.2210.88 11

PAGE 115

115 Table 4-2. Power input per unit volume Digester Working volume Temperature Mixing method Rotational speed Normalized intensity L C RPM W/m3 aKaparaju et al. Batch 0.4 55Shaking 10012.76 350.06bVavilin et al. Batch 0.5 37Shaking 10528.69 581.28cHoffman et al. Continuous 4.5 35Impeller mixing 500.041 25011.13 50091.74 15001109dKarima et al. Continuous 3.37 35Impeller mixing 2759.12eThis study Batch 3 55Magnetic mixing 1001.10 a Substrate, fresh cow manure; su bstrate to inoculum ratio, 0.11 b Substrate, municipal household solid wate; substrate to inoculum ratio, 0.125 c Feed, cow manure sl urry 5.8% TS based; d Feed, cow manure slurry 10% TS based; e Substrate, sugar beet tailings; su bstrate to inoculum ratio, 0.10

PAGE 116

116 Table 4-3. Bacterial 16S rRNA gene lib rary of the inoculum, digester 1 and digester 2 liquors. Incoulumn Digester 1 Digester 2 Clone Analyzed 94 90 87 Phylotypes 90 58 76 Coverage 90.42%93.33%95.40% Bacteria aActinobacteria b1.11 1.72 -cRubrobacterales d1.11 Coriobacteriales -1 7 2Bacteroidetes 1.11 1.32 Bacteroidales 1.11 1.32 Firmcutes 32.22 51.7223.68 Bacillales -6 9 0Clostridiales 23.33 44.8318.42 Thermoanaerobacteriales 8.89 5.26 Proteobacteria -3 4 5Betaproteobacteria -1 7 2Hydrogenophilales -1 7 2Synergistetes 13.33 20.69 3.95 Synergistales 13.33 20.69 3.95 Tenericutes --1 3 2 Acholeplasmatales --1 3 2 Thermotogae 52.22 22.4169.74 Thermotogales 52.22 22.4169.74 a Bacterial phyla are named according to National Center for Biotechnology Information (NCBI) taxonomy database b Phylum frequencies are calculated as a phylogenenic group in percentage of the total sequences analyzed c Bacterial orders are named according to National Center for Biotechnology Information (NCBI) taxonomy database d Order frequencies are calculated as a phylogenenic group in percentage of the total sequences analyzed

PAGE 117

117 Table 4-4. Archaeal 16S rRNA gene library of the inoculum, digester 1 and digester 2 liquors Incoulumn Digester 1 Digester 2 Clone Analyzed 44 41 43 Phylotypes 42 41 42 Coverage 100.00% 97.73% 97.67% Archaea aCrenarchaeota b62.8 53.66 97.62 cThermoproteales d62.8 53.66 97.62 Euryarchaeota 38.1 46.36 2.38 Methanobacteriales 7.32 2.38 Methanomicrobiales 37.2 24.39 Methanosarchinales 14.63 a Archaeal phyla are named according to National Center for Biotechnology Information (NCBI)t axonomy database b Phylum frequencies are calculated as a phylogenenic group in percentage of the total sequences analyzed c Archaeal orders are named according to National Center for Biotechnology Information (NCBI) taxonomy database d Order frequencies are calculated as a phylogenenic group in percentage of the total sequences analyzed

PAGE 118

118 Figure 4-1. Digester and gas meter configuration

PAGE 119

119 Figure 4-2. Inoculum used for experimental set 1 and 2

PAGE 120

120 Figure 4-3. Temperature profiles of incubator, digester 1 and digester 2

PAGE 121

121 Figure 4-4. Methane yield in di gester 1 and 2 from all 6 runs

PAGE 122

122 Figure 4-5. SCOD concentration in digester 1 and 2 from 6 runs

PAGE 123

123 Figure 4-6. VOA concentration in digester 1 and 2 fr om three runs

PAGE 124

124 Figure 4-7. The rank abundances of bacterial and archaeal phylotypes in the original inoculum, and digester 1 and digester 2 liquors

PAGE 125

125 Figure 4-8. Average COD balance of digesters 1 and 2

PAGE 126

126 CHAPTER 5 MICROBIAL POPULATION IDENTIFICAT ION AND DY NAMICS DURING BATCH THERMOPHILIC ANAEROBIC DIGESTION AT MIXED AND NON-MIXED CONDITIONS Summary Two trials of anaerobic digestion of sugar beet tailings at mixed and non-mixed condition were carried out in addition to those reported in Chapter 4. Digestion conditions were similar to that conducted in Chapter 4. The results confirmed digestion performance was affected by mi xing, where lower methane yield and delayed methane production rate were observed. This was further verified by the accumulation of high sCOD and persistence of VOAs (particularly propionic acid) in the mixed digester. To elucidate the di fferent digestion performance, microbial community analysis was conducted using 454 pyrosequencing. The microbial community of the mixed and non-mi xed digester shifted signifi cantly from that of the inoculum and adapted to a stable structure swiftly. Methanogens were more abundant and diverse in the non-mixed digester, including Methanobacterium Methanoculleus Methanosarcina and Methanothermobacter Syntrophic bacteria such as Desulfotomaculum Pelotomaculum and Syntrophomonas were also identified. They were known to play a crucial role in propionate and butyrate degradation. In comparison, relati ve abundance of syntrophic bacteria and methanogens, particularly hydrogenotrophic me thanogens, were noticeably lower in the mixed digester. This agreed with the observation of low methane yield and accumulation of propionic acid. Hydrogen producing bacteria we re identified in relatively abundant numbers in th e mixed digester. Among them, Ruminococcus species were of special interest, whic h were reported to produce hydrogen in the presence of hydrogen scavengers. It was hy pothesized that presence of hydrogen inhibited VOA degradation and resulted in lower methane yield in the mixed diges ter.

PAGE 127

127 Background Class ically, environmental microbial communities are analyzed by construction of 16S rRNA clone libraries and subsequent sequencing of individual clones. The approach, termed Sanger s equencing, has been applied for many biogas-producing microbial communities ( Huang, Zhou et al. 2002; Huang, Zhu et al. 2005; Klocke, Mahnert et al. 2007; Klocke, Nettmann et al. 2008) and was used in Chapter 4. Traditional Sanger sequencing technology has its limitations in revealing the whole c omplexity of microbial communi ty because a relatively small amount of clones are sequenced. New development of high-throug hput sequencing technologies such as 454 barcoded pyro sequencing not only eliminates the laborious step of preparing clone libraries, but also makes large scale environmental sequencing cost effective and keeps the bias small ( Ronaghi, Karamohamed et al. 1996). However, phylogenetic ass ignments based on barcoded pyrosequencing could be less precise due to the short read lengths compar ed to relatively larger sequence lengths resulting from classical 16S rRNA sequencing. Barcoded pyrosequecing has been reported to be applied in microbial population identification in anaerobic digestion and one of t hose studies revealed that Clostridia is the most prevalent taxonomic class, and Methanomicrobiales is dominant among methanogenic Archaea ( Krause, Diaz et al. 2008). 16s rDNA library analysis has revealed distinct microbial community structures between mixed and non-mixe d anaerobic digesters digesting sugar beet taili ngs. However, due to the limitation of generating small number of sequencing result s, the analysis was carried out once for each digester and did not provide the info rmation of the changing of microbial communities with time. Thus, 454 pyrosequencing was utilized as the approach for further exploration of the microbial communi ty structure and its dynamics. Similar to

PAGE 128

128 Chapter 4, two runs of anaerobic digestion of sugar beet tailings was carried out at mixed and non-mixed condition. Diges tion performances such as CH4 production, SCOD and VOA concentrations were m onitored and compared. The microbial communities were identified and characte rized by analyzing sequence data obtained from using 454 barcod ed pyrosequencing. Method and Materials Feedstock The sugar beet tailing s was provided by America Crystal Sugar Company, Minnesota and stored at 4 C before using. The tailings were washed using tap water before loading to the digesters. Wash water was discarded. Anaerobic Reactors and Operation Two digest ers (digester 1 and 2) of 5 L were constructed by modifying Pyrex glass jars. Gas production from the di gesters was measured by a positive displacement gas meter. Refer to Chapter 4 for det ailed information about the anaerobic digesters and gas meters. Digester 1 was operated at non-mixed condition and added with a bulking agent. Digester 2 content was continuously mixed. A 50.8 mm x 9.5 mm PTFE coat ed polygon bar was placed in the digester and the mixing of digester content was ac hieve by using a large volume magnetic stirrer (Siceneware Cool Sti rrer). Digester 1 and 2 were placed in a 55C incubator throughout the experiments. Tw o experimental trials were carried out. In trial 1, digester 1 and 2 were loaded with 0.3 kg (wet weight) of washed tailings and maintained at 55 C. 2 kg of bulking materi als (lava rocks from landscaping supplier, 0.025 m in average size) were added into diges ter 1 along with substr ates to prevent substrate compaction and floatation. Digester 2 was not added bulking materials and the digester content was continuously mixing at 100 RPM. Each digester was

PAGE 129

129 inoculated with 3 L inoculumn taken from an anaerobic digester that has been digesting with sugar beet tailings for months. Trial 1 was ended when the gas production from both digesters was low. Digester 1 and 2 were then emptied and washed thoroughly. Residual substrates we re discarded while the digester liquor from trial 1 were saved and used as inoculums to initiate trial 2. About 3 L digester 1 and 2 liquors were recovered and used to inoculate digester 1 and 2 in trial 2, respectively. Additionally 3 runs were carried out in the mixed digester (digester 2). In the 3 runs, digester 2 was fed with only propionic acid (analytical grade 99.5%). This was to demonstrate whether propionic acid degradation was affect ed by the mixing intensity used digester 2. Refer to Appendix B for detail. Chemical Analysis Total Solids (TS) and Volatile Solids (VS) contents were determined for the feedstock sugar beet tailings. Gas compos ition (CH4 and CO2) was analyzed using Fisher Gas Partitioner (Model 1200). sC OD was determined using Hach COD vials and measured using Hach DR/890 Colorime t er. VOA analysis was conducted using Shimadzu gas chromatograph (GC-9AM equipp ed with a flame ionization detector) for acetic, propionic, isobutyric, butyric, is ovaleric and valeric acid concentrations. For analyses details, refer to analysis section in Chapter 4. Molecular Biological Analysis Microbial c ommunity analysis was conduc ted for the original inoculums and digester 1 and 2 liquors. Digester liquors were sampled at day 3, 15 and 18 for trial 1 and day 0, 4, 8, 11, 12 and 14 for trial 2. A total 17 samples were analyzed.

PAGE 130

130 DNA Extraction and Purification Total DNA was extracted and purified by using FastDNA Kit and PowerClean DNA kit, respectively, according to the manufacturess instruction The qualit y of DNA was verified by agar gel electrophoresis running at 100V for approximately 50 minutes. Results were visualized in UV ligh t. Extracted DNA wa s stored at -20 C before using as a template for Polymera se Chain Reaction (PCR) amplification. Polymerase Chain Reaction Amplification For each sample, 16S rRNA gene was am plified using a composite forward primer and a reverse primer containing a unique 8-base barcode us ed to tag each PCR product ( Hamady, Walker et al. 2008). Thr ee independent PCR reactions were carried out for each of 17 samples T he reactions were performed in a 50 l volume, containing 20 to 30 ng of DNA template and 20L of HotMasterMix [0.5U Taq DNA Polymerase, 45 mM KCl, 2.5 mM Mg2+, and 200 M of dNTP (5Prime GmbH)] and 1L of barcoded primer (100 pmol es). The amplification protocol was as follows: one initial denaturation cycle at 94oC for 3 minutes, 30 denaturation cycles at 94oC for 45 seconds, annealing at 50oC for 30 seconds, and extension at 65oC for 90 seconds, and one final extension cycle at 65oC for 10 minutes. The Prokaryote primer information is provided as follows.

PAGE 131

131 Original primers (yellow tubes) 515 GTGTGCCAGCMGCCGCGGTAA 806R GGGGACTACVSGGGTATCTAAT 454 Primers (blue tubes) F515 GCCTTGCCAGCCCGCTCAG GTGTGCCAGCMGCCGCGGTAA A1-806R GCCTCCCTCGCGCCATCAG AACGCACGCTAG GGGGACTACVSGGGTATCTAAT A2-806R GCCTCCCTCGCGCCATCAG AACTCGTCGATG GGGGACTACVSGGGTATCTAAT A3-806R GCCTCCCTCGCGCCATCAG AACTGTGCGTAC GGGGACTACVSGGGTATCTAAT A4-806R GCCTCCCTCGCGCCATCAG AAGAGATGTCGA GGGGACTACVSGGGTATCTAAT A5-806R GCCTCCCTCGCGCCATCAG AAGCTGCAGTCG GGGGACTACVSGGGTATCTAAT A6-806R GCCTCCCTCGCGCCATCAG AATCAGTCTCGT GGGGACTACVSGGGTATCTAAT A7-806R GCCTCCCTCGCGCCATCAG AATCGTGACTCG GGGGACTACVSGGGTATCTAAT A8-806R GCCTCCCTCGCGCCATCAG ACACACTATGGC GGGGACTACVSGGGTATCTAAT A9-806R GCCTCCCTCGCGCCATCAG ACACATGTCTAC GGGGACTACVSGGGTATCTAAT A10-806R GCCTCCCTCGCGCCATCAG ACACGAGCCACA GGGGACTACVSGGGTATCTAAT A11-806R GCCTCCCTCGCGCCATCAG ACACGGTGTCTAGGGGACTACVSGGGTATCTAAT A12-806R GCCTCCCTCGCGCCATCAG ACACTAGATCCG GGGGACTACVSGGGTATCTAAT underline = 454 primers in red = barcodes Because each sample was amplif ied with a known tagged primer, an equimolar mixture of the PCR-amplified DN A from each sample can be sequenced and sequence can be assigned to samples based on the unique barcodes. Since there were 12 pairs of primer s available, 17 samples were amplified in two batches (set 1 and 2) as specified in Table 5-1. Samples (1 inoculum sample, 6 trial 1 samples and 2 trial 2 samples) in Set 1 were tagged and amplif ied with primers A1 through A9. Samples (8 trial 2 samples) in Set 2 were tagged and amplified with primers A1 though A5 and pr imes A10, A11 and A12 Three replicated PCR products were combined for each sample and purified using the QIAquick PCR Purification Kit (Qiagen, CA) and quantified using on-chip gel electrophoresis with Agilent 2100 Bioanalyzer and DNA Lab Chip Kit 7500. The quantity of each PCR product was made equal within a set. Set 1 contained 3888 ng DNA in total with

PAGE 132

132 each PCR product being 432 ng. Set 2 c ontained 3840 ng DNA with each PCR product being 480 ng. Set 1 and set 2 were then sent to Interdisciplinary Center for Biotechnology Research (ICB R) at University of Flor ida for pyrosequencing. Total 11536 sequence reads with av erage alignment length of 239 bp and 3237 sequence reads with average alignment length of 238 bp were produced for set 1 and set 2, respectively. To identify 16 rDNA reads in formation, a homology search to the RDP (Ribosomal Database Projec t) database was conducted by means of BLAST. Search results were grouped into Operational Taxono mic Units (OTUs) at different levels (Phylum, Class, Order, Family and Genus) according to the nearest neighbor (best matching BLAST-hit) in the RDP database. Ph ylotype was defined as a group of 16s rDNA sequences with 95% identity (taxonomic genuslevel). A non-parametric estimation of maximum species richness was derived using Chao1 estimate ( Schloss, Westcott et al. 2009). The samp ling coverage was then defined as number of OTUs by Chao1 estimated richness. The following dissimilarity cut-offs were used to represent OTUs at different taxonomic ranks, 5%, genus; 10%, family, order and classes, and 20% phylum ( Hong, Bunge et al. 2006). Results Characteristics of Feed Substrates TS and VS contents of sugar beet tailings, lo adedquantities and packing density were determined for 2 trials and pres ented in Table 5-1. The average TS and VS contents of sugar beet tailings were 10.90% (wt/wt) and 9.68% (wt/wt), respectively. Methane Production Cumulative CH4 yield for digester 1(non-mixed) and 2 (mixed) in two trials are shown in Figure 5-1. CH4 production rate for digester 1 peaked at 0.70 m3 d-1 (kg

PAGE 133

133 VS)-1 on day 5, and 0.34 m3 d-1 (kg VS)-1 on day 11 for digester 2. Digester 1 achieved CH4 yield of 0.37 m3 CH4 at STP (kg VS)-1 and digester 2 achieved CH4 yield of 0.24 m3 CH4 at STP (kg VS)-1 at the end of trial 1. Trial 2 was started by flooding digester 1 and 2 with digester liquor left in trial 1. Digester 1 showed higher CH4 production rate than in run 1 whereas di gester 2 showed similar production rate. Maximal CH4 production rate for digester 1 and 2 were 0.94 m3d-1(kg VS)-1 on day 4 and 0.35 m3d-1(kg VS)-1 on day 7, respectively. The CH4 yield was 0.35 m3 CH4 at STP kg VS-1 for digester 1 and 0.26 m3 CH4 at STP kg VS-1 for digester 2 and the end of run 2. Degradation of Organic Matters Profiles of soluble COD (sCOD) and VOA concentration of digester 1 and 2 were shown in Figure 5-2. sCOD concentration of both di gesters in itially accumulated, reached a maximum and decr eased to a minimum. Digester 1 showed less sCOD accumulation and faster degradation rate than digester 2 in both trials. This was consistent with the higher CH4 production rate of digester 1 seen in Figure 5-1. Acetic acid was detected with the highest concentrations among VOAs in digester 1. The concentration decreased rapidly to a negligible amount as methane was produced. Butyric and propionic acid were detected with the second and third highest concentrations, respectively. As for acetic acid, degradation of butyric acid was rapid and started from day 3. Though accumulation of propionic acid was low in trial 1, the degradatio n of was delayed and did not begin until near the end of the trail 1. In trial 2, propionic acid degraded al ong with acetic and butyric acid which occurred shortly after day 4. In digester 2, acetic acid, propionic acid and butyric acid have dominant concentrations over other VOAs. The initial increase and degradation of the three acids we re noticeably slower in digester 2 than in digester 1.

PAGE 134

134 Concentration of acetic acid and butyric acid peaked between day 8 to day 11 and it dropped thereafter. This time frame coin cided with the occurrence of maximal CH4 production rate in digester 2. Distinct fr om other VOAs, there was no evidence of degradation of propionic acid throughout trail 1 and 2 for digester 2. Accumulation of propionic acid was around 400 mg/L in tria l 1 and was further increased to above 1000 mg/L in trial 2 Microbial Community Analysis All 16 rDNA reads were taxonomically classified by means of BLAST search to RDP reference database. Sampling cov erage at dissimilarity cut-off of 5%, 10% and 20% (corresponding to genus, family/order/class and phylum, respectively) was compared and showed in Figure 5-3. At hi gher taxonomic ranks (phylum, class and order), a defined OTU tended to contains a cluster of 16s rDNA reads whereas number of singleton OTUs increa sed significantly at the low rank (genus), resulting in a low Chao1 richness. Therefore, the mi crobial community was analyzed at order rank (sequence identity of 90% or mo re) the relative abundance (appearance frequency) of detected OTUs were showed in Table 5-3. Chao1 estimation at order rank gave coverage of approximately 70% or more for most samples. The structure of microbial community in digester 1 and di gester 2 shifted significantly from the inoculum community. Bacterial phyla Proteobacteria and Firmcutes were found dominant in inoculum microbial communi ty with relative abundance being 41% and 33%, respectively. Numbers of OTUs at 10% cut-off belonged to Proteobacteria were much more than that of Firmcutes, making Proteobacteria the most diverse phylum. Phylotypes Hydrogenophaga Arcobacter and Pseudomonas were found abundant within phyla Proteobacteria while phylotypes Bacillus Clostridium Coprothermobacter Cryptanaerobacter and Proteiniborus were abundant within

PAGE 135

135 phylum Firmcutes. As the digestion began, the relative abundance of Proteobacteria quickly diminished in both diges ters. In contrast, the abundance of Firmcutes was maintained and was identified with similar ph ylotypes as in the inoculum. Bacteria phyla Bacteroidetes and Thermotogae were seen increasing abundance in both digesters as the digestion was progressing. Typical phylotypes identified in phylum Bacteroidetes included Bacteroides and Petrimonas (for both digesters). No defined phylotypes (95% identity, genus leve l) were not detected in phylum Thermotogae The closest relative was identified as Thermotogales (order level). Although the microbial community of digester 1 and 2 shared aforementioned similarities at phylum rank (e.g., both abundant Firmcutes), they were distinct at genus rank (defined phlotype).Phylotypes Acetanaerobacterium and Ruminococcus in phylum and phylotype Ruminofilibacter in phylum Bacteroidetes were detected in digester 2 and exhibited increasing abundance with the development of the digestion. However, these phylotypes were either not detected or detected at a very low abundance in digester 1. Phylotype Bacteroides was detected in both digesters, but the abundance was significantly greater in digester 2 ( 40% versus 10%). Likewise, its relative abundance low at beginning (inoculum) showed an increase. Furthermore, digester 1 and digester 2 showed different distribution of phylum Synergistales While Synergistales gradually developed and reached the relatively high abundance of 15% in digester 1 (after 18 days), it was detected with constantly low abundance in digester 2. Anaerobaculum was identified as the dominant phylotype in Synergistales Compared to bacteria, archaea was identified at low abundance for inoculum digester 1 and digester 2 samples. As expected, methanogens were found abundant within phylum Euryarchaeota and the main phylotypes included Methanobacterium Methanoculleus Methanosarcina and Methanothermobacter

PAGE 136

136 Abundance of methanogens increased with the digestion proceeding. The abundance was found highest at day 5 and day 4 (cumulative time day 32) for digester 1 in trial 1 (abundance 6%) and trial 2 (11%), respectively. Development of methanogens seemed delayed in digester 2 with the highest abundance (4%) seen at day 18 in trial 1. In trial 2, the abundance of methanogens continuously reduced and no significant detections wa s obtained near the trial end. Discussion Digestion Performance Comparison As concluded in Chapter 4, diges ter 1 (nom-mixed) achieved higher CH4 yield and CH4 production rate than digester 2 (mixed). Digester 1 in trial 2 was inoculated with digester 1 liquor recovered from trial 1, which had been adapted to tailings decomposition. Theref ore, digester 1 showed higher CH4 production rate in trial 2 than in trial 1. This was consist ent with the higher sCOD degradation rate of digester 1 in trial 2 though the initial increas e of sCOD in both trials was both around 8000 mg/L. The adapted digester 1 has a maximum daily CH4 production rate of 0.94 m3d-1(kg VS)-1, which agreed well with the maximum daily CH4 production rate of the nonmixed digester in C hapter 4. The initial sCOD accumulation of digester 2 was around 10000 to 12000 mg/L, and a lagged per iod of 8 to 10 days was observed before degradation. As for digester 1, digeste r 2 inoculum was adapted in trial 2, but neither CH4 production nor substrate decompositi on showed accelerated rate. This suggested the mixing affected di gester 2 performance adversely. The VOA profiles revealed that the initial accumulati on of acetic acid of both digesters (trial 1 and trial 2) was in a close range of 1.3 g/L. Howeve r, acetic acid in digester 2 persisted for 10 to 12 days befor e degradation whereas acetic acid in digester 1 degraded substantially. C onversion of propionic acid appeared

PAGE 137

137 problematic in digester 2 as high a ccumulation was observed. In anaerobic digestion, acetogenic bacteria ferment propi onate to acetate which was then utilized by methanogens to CH4. The pathway plays an important role in methanogenesis ( Fukuzaki, Nishio et al. 1990). The high accumu lation of propionic ac id in digester 2 probably s uggested propionate d egradation was inhibited. The accumulation was further increased in trial 2, indicating dige stion inhibition could be exacerbated if the digester liquor was exposed to mixing and used for inoculation repeatedly. It was well accepted that accumulation of propionic acid indicated an anaerobic process instability ( Hill, Cobb et al. 1987; Fox, Suidan et al. 1988; Nielsen, Uellendahl et al. 2007) but it can also be considered as t he cause of the process failure as its accumulation had been reported inhibitory for methanogens ( Barredo and Evison 1991). Microbial Community 16s rDNA sequences reads were compared to the entries of RDP database, and grouped to OTUs at 5 taxonomic ran ks according to their nearest neighbors (Best BLAST hits). Most reads could on ly be assigned to higher taxonomic ranks. Number of BLAST hits obtai ned at genus rank was significantly reduced, suggesting most members of microbial community in anaerobic digestion are unexplored. Microbial community shift from the inoc ulum (day 0) was observed for digester 1 and 2. The shifting occurred swiftly. At day 3 the community structure was significantly different from the original while the change s were small thereafter for digester 1 and 2, respectively. The swift change was expected as the result of quick adaption to the new environment of the microbial consortium in the inoulum. The newly adapted consortium managed to ma intain their community structure throughout the experiment period.

PAGE 138

138 Though the microbial community between digester 1 and 2 were mostly distinct, they shared some similarity. Proteobacteria phylotypes Hydrogenophaga and Pseudomonas were identified in the inoculum but reduced to low abundance in both digesters shortly after digestion began. Hydrogenophaga and Pseudomonas are recognized for their degradative abili ties in the presence of oxygen ( Willems, Busse et al. 1989; Juwarkar, Singh et al. 2010). In trial 1, the digesters were inoculated with stored incolum that may have been exposed to aerobic environ ment and thus favored the growth of some aerobes. Hydrogenophaga and Pseudomonas were taken over quickly by anaerobes as anaerobic digestion became active. Relative abundance of some facultative bacteria as Acrobacter ( Vandamme, Falsen et al. 1991) also decreased significantly in digester 1 and 2. Firm cutes Phylotypes Bacillus Clostridium Coprothermobacter and Proteiniborus were also abundant in the original inoculum. Their relative abundance maintained throughout the digestion for digester 1 and 2. Abundance of Clostridium was increased at early stage of the digestion (day 3 to day 5) and was the dominant species in the period. These bacteria are commonly identifi ed in anaerobic digesters and involved in degradations of a wide range of carbohydrates and prot eins and producing acetic acid, hydrogen and carbon dioxide as typical fermentation products ( Etchebehere, Pavan et al. 1998; Niu, Song et al. 2008; Cardinali-Rezende, Mo raes et al. 2011; Kim, Shin et al. 2011). Hydrolysis is usually cons idered as the first step in anaerobic digestion and this could explain the increase of hydrolytic bacteria like Clostridium at beginning of the experiment. Phylotypes Bacteroides and Petrimonas within phylum Bacteroidetes were seen in greatly increased abundance after digestion started. These bacteria were reported to play impor tant roles in hydrolysis and acidogenesis as well among which, Bacteroides are known to be cellulolytic and were expected

PAGE 139

139 to help into degrading cellulose which wa s detected as one of the main components in sugarbeet tailings ( Grabowski, Tindall et al. 2005; Liu, Pullammanappallil et al. 2008; Russell, Muck et al. 2009; Ziganshin, Schmidt et al. 2011). Bacteria order Synergistetes showed differences in dynamics between digester 1 from digester 2. Only one bacteria phylotype Anaerobac ulum was identified within the order Its relative abundance gradu ally increased, reached and maintained at 15% in digester 1. Diges ter 2 had a constantly lower abundance of Anaerobaculum compared to that of digester 1 at the same perio d. The abundance exhibited a decreasing pattern in digester 2 from trial 1 to trial 2, and in trial 2 it was kept overall low. Identification of Anaerobaculum in anaerobic digestion were reported and they were known to degrade peptide and a limited number of carbohydrates ( Rees, Patel et al. 1997; Baena, Fardeau et al. 1998; Manes, Fernandez et al. 2001; Menes and Muxi 2002; Sousa, Smidt et al. 2007; Weiss, Jerome et al. 2008). Sugihara et al st udied the propionate degr ading ability of a microbial c onsortium by subjecting t he enriched culture to sequencing fed batch cultivation with a periodica l impulse of propionate ( Sugihara, Shiratori et al. 2007). The consortium population analy sis revealed an Anaerobaculum related species being the major bacterial constituent. Though Anaerobaculum was not recognized as a syntrophic propionate utilizing bacterium, Sugiharas study seemed to imply they play a role in propionate degradation. Theref ore, it was postulated that the fewer growth of Anaerobaculum probably resulted in accumulation of propionic acid in digester 2. Anaerobaculum sp. related species were also reported to persist in a LCFA (long chain fatty acid) enriched cultures, suggesting a possible LCFA degrading ability ( Manes, Fernandez et al. 2001; Hatamoto, Imachi et al. 2007; Palatsi, Vias et al. 20 11). Furthermore, hydrogen (H2) inhibition was reported on

PAGE 140

140 glucose utilization by Anaerobaculum mobile ( Menes and Muxi 2002). This could explain the low growth of in digester 2 where free H2 were speculated producing (discussed later). Phylotype Ruminofilibacter was identified in digester 2. The relative abundance increased greatly after the diges tion began (for both trials) and then depressed slowly after reaching a maxi mum. The phylotype was observed with constantly low abundance in digester 1. Ruminofilibacter was detected in anaerobic digesters and shown to have pronounced xylanolytic activity (Krber, Bekel et al. 2009; Wei, Zankel et al. 2011). It might s uggest that xylan being released from tailin gs in digester 2 as a result of hydrolysis. Bacterial phylotypes Desulfotomaculum Pelotomaculum and Syntrophomonas were detected in both digest er 1 and 2 at low abundance (below 2%). Species in these genera are well known as syntrophic bacteria that play crucial role degradation of short chain fatty acid such as propionate and butyrate ( Plugge, Balk et al. 2002; de Bok, Harmsen et al. 2005; Mller, Worm et al. 2010). The presence of syntrophic bacteria at low leve ls agreed wit h the study of McMahon et al. ( McMahon, Zheng et al. 2004). This might indicate effective propionate or butyrate degradation does not require high abundanc e of syntrophic bacteria. Similar to syntrophic bacteria, methanogens appeared at low abundance throughout the digestion (compared to the total bacteri a). This generally agreed with the reported methanogen abundance ranging from 0.1% to 15% (relative abundance of the total microbial population) (Solera, Romero et al. 2001; Huang, Chen et al. 2003; Dar, Kleerebezem et al. 2008; Shin, Lee et al. 2010). Identified methanogens were closely related to Methanobacterium Methanoculleus Methanothermobacter Methanosarcina and an uncharac terized methanogenic archaeon CH1 270. Among

PAGE 141

141 them, members in Methanobacterium Methanoculleus Methanothermobacter are hydrogentrophic (Xun, Boone et al. 1989; Maestrojuan, Boone et al. 1990; Elberson and Sowers 1997; Valentine, Blanton et al. 2000; Shcherbakova, Rivkina et al. 2011). Methanosarcina species are mostly aceticlasic b ut are also able to use H2 ( Kotsyurbenko, Chin et al. 2004). Metabo lic pathway of the unchara cterized methanogen is unknown, which was det ected in adult chicken ceca ( Saengkerdsub, Anderson et al. 2007). Methanogens popul ation dynamics exhibited changing abundance over time. Increased abundance was usually associated with acetic acid degradation. For instance, a marked incr eas e in methanogen abundance was seen from day 3 to day 5 (trial 1) for digester 1 and during the same period (day 4) acetic acid was decreased sharply. This indicated CH4 was mainly produced through aceticlastic way. In average, the abundance of methanogens and syntrophic bacteria in digester 2 was lower than in digester 1, specially hydrogenotrophic methanogens. The phenomenon could be the consequence of disrupted spatial juxtapositions of syntrophic bacteria and methanogens resulting from continuous mixing ( Conrad, Phelps et al. 1985; Dolfing 1992; McMahon, Stroot et al. 2 001). However, the mixing utilize d for digester 2 was not regarded in tense as discussed in Chapter 4. The disruption was expected to be limited. Therefore, an al ternative hypothesis of H2 inhibition was made and discussed as follows. Bacteria phylotypes Acetanaerobacterium and Ruminococcus were found to be relatively abundant in digester 2. They are recognized to degrade cellulose and produce hydrogen as one of the fermentation products ( Ren, Xing et al. 2007; Jindou, Brulc et al. 2008; Saraphirom and Reungsang 2010; Mosoni, Martin et al. 2011). Application of species in Rum inococcus has been widely used in hydrogen production on a variety of feedstock ( Antonopoulou, Ntaikou et al. 2007; Ntaikou,

PAGE 142

142 Gavala et al. 2008; Ivanova, Rkhely et al. 2009; Ntaikou, Koutros et al. 2009; Zhang, Banaszak et al. 2009; Ntaikou, Gavala et al. 2010). Though Ruminococ cus species were reported to be mesophilic their appearance has been confirmed in theromphilic anaerobic digesters as well ( Yang, Tsukahara et al. 2008). This supports the detection of Ruminococcus in thermophilic digesters in this study. In anaerobic digestion, hydrogen could be generated through fermentation of intermediate products as sugars or VOAs ( Lettinga, Field et al. 1997; Grady, Daigger et al. 1999; Angenent, Karim et al. 2004). Unfortunately, hydrogen (H2) production from organic substrate is energetically unfavorable due to proton being a poor electron acceptor. The low midpoint redox potential of redox couple H+/H2 of -414 mV suggests the energetic difficulty in reducing H+. Therefore, production of H2 requires a strong reducing agent ( Angenent, Karim et al. 2004). Common redox mediators involved in ferm entation include NADH and F ADH2, and the redox potential of redox couples NAD+/NADH and FADH/FADH2 is -320 and -220 mV, respectively. The high redox potentials re expected to cause an energetic problem in reducing H+ to H2 ( Stams and Plugge 2009). To make NADH oxidation and FADH2 oxidation coupled to H+ reduction energetically feasible the partial pressure of H2 has to be very low. Assuming the intrace llular concentrations of the oxidized and reduced forms of NADH (or other electron ca rriers) are about equa l, the maximum H2 partial pressure that allow the proc ess of hydrogen producing is determined by: RT EEF PxH H)(2 exp00 max2 2 ( Angenent, Karim et al. 2004) (5.1)

PAGE 143

143 wheremax2HPis the allowed maximum H2 partial pressure,0 xE is the redox potential of the electron carrier,02HEis the redox potentia l of redox couple H+/H2,F is Faraday's constant, R is the ideal gas cons tant and T is the Kelv in temperature. The low H2 partial pressure can be achieved in the presence of H2 scavengers as homoacetogens and hydrogenotrophic methanogens. This is a syntrophic relation in which bacteria produce H2 that is consumed as t he substrate by methanogens, which in turn keep low H2 partial pressure for the bacteria to grow through H2 producing pathway. This process is refe rred to interspecies hydrogen transfer ( Thauer, Jungermann et al. 1977). The dev elopment of syntrophic communities allo ws H2 production to become energetically fa vorable and sustain degradation of organic compounds and production of CH4. Bacteria that are only able to oxidize NADH coupled to hydrogen formation ar e considered obligate syntrophs. Production H2 through NADH oxidation plays a crucial role in their energy metabolism as this is the only manner in which they can dis pose the electron deriv ed from substrate oxidation ( Reddy, Wolin et al. 1972). Some well known obligate sy ntrophs include species belonging to genera Syntrophomonas (butyrate degrading), Syntrophus (benzoate degrading), Syntrophobacter (propionate degrading) and Peloto maculum (propionate degrading) (Harmsen, Van Kuijk et al. 1998; Jackson, Bhupathiraju et al. 1999; Imachi, Sekiguchi et al. 2002; McInerney, Rohlin et al. 2007). Presence of H2 scavengers as hydrogenotrophic methanogens typically existing in anaerobic digestion is essential for their survival. However, many fermenting bacteria can also oxidize NADH by reducing intracellular met abolites and their syntrophic relation with H2 scavengers is facultative. A classic exam ple is the study of fermentation product in a rumen bacterium pure culture Ruminococcus albus which is known to have two alternative pathways for NADH oxidization: either by reducing acetyl-coenzyme A to

PAGE 144

144 ethanol or by reducing H+( Iannotti, Kafkewit.D et al. 1973). Since NADH oxidiz ation through proton reduction is energetically difficult, energy metabolism of the bacterium mainly proceeds through ethanol fermentation pathway in the absence of H2 scavengers. H2 is still produced through the decarboxylation of the fermentation intermediate pyruvate. Ferrodoxin (Fd) in t he key redox mediator in the conversion. The redox potential of the redox couple Fd (ox)/Fd (red) is -400 mV or lower depending on the source ( Stams and Plugge 2009). The relative low potential suggests H2 production through oxidation of reduced from of ferrodoxin is energetically easier. For this reason, anaerobic processes were developed for H2 production using fermenting bacteria with sim ilar metabolic pathways. In fact, most H2 production observed in biol ogical system can be attribut ed to decarboxylation of pyruvate (Angenent, Karim et al. 2004). Mixed cult ures are usually preferred over pure cultures for H2 production because process with mixed culture are simpler to operate and control, and may have a boarder choice of feedstocks ( Valdez-Vazquez, Ros-Leal et al. 2005). Ye t in a mixed culture H2 produced from fermentative bacteria could be readily converted CH4 to due to the energetically favorable syntrophic pathway. A pretreatment process is needed to suppress the activity of methanogens, including heating, acidic treatment, adding inhibiting chemicals and so forth ( Danko, Pinheiro et al. 2008). pH control is crucial and readily achievable for H2 production because methanogenic activity drops sharply in an acidic environment ( Chen, Lin et al. 2002). Though H2 production is mostly carried out at pH 4 to 6 for H2 production, H2 was also detected in anaerobic proce sses operated at neural or near neutral pH ( Fang and Liu 2002; Valdez-Vazquez, Ros-Leal et al. 2005; Zhu, Parker et al. 2009). In these studies a mixed culture was used to produce H2 at either thermophilic or mesophilic temperature without applying a pr etreatment to inhibit H2

PAGE 145

145 consuming microorganism. H2 production was recorded at pH 7 (or slightly higher), though this pH was not determined optimal. Fang and Zhu reported a lower H2 production (30% H2 of biogas ) and yield of 30 ml (g VS) -1, respectively; whereas Valdez-Vazquez derived a much higher H2 yield of 100 and 300 ml (g VS) -1 from an anaerobic digester operated at mesophilic and the rmophilic temperature, respectively. This contradicts with the syntrophic metabolism of H2 producing bacteria. At neutral pH, methanogenic activity was expected (since methanogens was not suppressed) and the produced H2 would be a substrate for CH4 production. However, a recent study seemed to support the observation of free H2 production in the presence of methanogens. Rychlik and May studied the effect of Methanobrevibacter smithii on growth rate, organic acid production and specific ATP activity of Ruminococcus albus in a co-ulture ( Rychlik and May 2000). The result indic ated no increase in the growth rate acetate or ATP production, suggesting Ruminococcus albus did not receive energetic advantage from co-culturing with the methanogen. This disagreed with the exper iment of Iannotti et al, in which Ruminococcus albus was shown to gain more ATP in the presence of a different H2 consuming species Vibrio succinogenes The discrepancy probably suggests the syntrophic interaction between H2 producers and scavengers could be species dependent. Another indirect evidence came fr om Zhou et al, who investigated effect of methanogenic inhibitors on methanogens and three rumen bacteria Fibrobacter succinogenes, Ruminococcus albus and Ruminococcus flavefaciens (Zhou, Meng et al. 2011). While the anti-methanogen compounds effectively reduced the population of methanogens, the inhibiting effect was insignific ant on none of the bacterial population, suggesting the syntrophic relation was weak or did not exist. Bacteria phylotype Ruminococcus along with Acetanaerobacterium were detected only in

PAGE 146

146 digester 2 with high abundance. They were speculated to produce H2 even at neutral pH (average pH 7.22 for digester 2, pH profile not shown), particularly for Ruminococcus because it had been specifically proven to lack energetic advantage growing with methanogens. These bacteria could then choose ethanol fermentation for energy metabolism and produce H2 through ferredoxin oxidation. Presence of free H2 was expected to have adverse e ffect on anaerobic digestion. Obligate syntrophic bacteria are extremely sensitive to H2. Even at low partial pressure, H2 can inhibit syntrophic metabolism and thereby limit the substrate supply to methanogens. This could explain t he low abundant methanogens detected in digester 2. This inhibition seemed to be cumulative as digester 2 community dynamics showed decreasing methanogen abundance with time. Inhibition of syntrophic interactions could also result in accumulation of VOAs because many syntrophs are known to degrade VOAs. This was verified by the high accumulation of propionic acid observed for digester 2, particularly in trial 2. Interestingly, some species of Ruminococcus were reported to produce propi onate other than ethanol as fermentation products, which could also lead to propionic acid accumulation ( Ren, Xing et al. 2005). The discussions abov e appeared to justify the possible H2 production from digester 2. A fraction of substrate COD was converted to H2 and led to a lower CH4 yield. Presence of free H2 inhibited growth of syntrophic bacteria, limiting acetate production and methanogensis. The lowered CH4 yield could be attributed to both sCOD loss (H2 production was not monitor ed in the study) and the inhibiting effect of H2. Further investigations are requi red to verify these hypotheses. Support for the hypotheses were found in a similar case, where simultaneous H2 and CH4 production was observed in a si ngle-stage anaerobic digestion though H2 was produced under slightly ac idic pH ( 6.3-6.6) ( Banks, Zotova et al. 2010). The

PAGE 147

147 produced gas mixture (CO2 excluded) consis ted of 20% of H2 and 80% of CH4. The estimate of H2 content of biogas produced in t he mixed digestion conducted in Chapter 4 is close to Banks results. The microbial community analysis conducted in Chapter 4 using Sanger sequencing showed less diversity. Some bacterial phylotypes at low abundance were detected by 454 pyrosequencing but were not seen in the classic identification method. This indicated 454 sequencing prov ided a better coverage of the microbial community ( Huang, Chen et al. 2003). For in stance, low abundant syntrophic bacteria ( D esulfotomaculum Pelotomaculum and Syntrophomonas ) were not identified by the 16s rDNA clone library anal ysis. Nevertheless, detection of bacteria with high abundance was generally consistent (F or the same period, i.e., day 3 after digestion began). Bacteria order Clostridiales Bacillales and Thermotogales were detected abundant by both methods. Species in Clostridiales and Bacillales are mostly known to be involved in ca rbohydrates and prot ein degradation ( Kim, Song et al. 2010). Both methods revealed low abundance of bacterial order Synergistales in digester 2 while marked higher abundanc e in di gester 1. The Identified phylotypes of Synergistales were reported in associati on with syntrophic pr opionate-degrading community ( Sugihara, Shiratori et al. 2007). Closing Remarks This study was conducted to further investigate the effect of mi xing on anaerobic digestion, focusing on the microbial community dynamics. Two anaerobic digestion trials were carried out at mixed and non-mixed condition. The mixed digester 2 exhibited lower CH4 yield, delayed CH4 production and persistence of propionic acid. This confirmed the results obtained in Chapter 4. The microbial community analysis revealed the abundanc e of methanogens, syntrophic bacteria

PAGE 148

148 and an Anaerobaculum species in the non-mixed digester but were markedly lower in the mixed digester. Their abundance showed a declining pattern over time, suggesting the mixed digestion could deteriorate if the di gester liquor were reused for inoculation. Some bacteria known to produce H2 through sugar fermentation were detected at in digester 2, among which, phylotype Ruminococcus were abundant. Ruminococcus species were found in many anaerobic processes for H2 production and reported to not change fermentati on pathways when co-cultured with methanogens. Therefore, it is possible that that methanogens is was affected in the mixed digester due to the pr esence of minor amount of H2 which could have inhibited syntrophic interactions and VOA degradation.

PAGE 149

149 Table 5-1. PCR amplifications of 16s rRNA genes using 12 barcoded primer s (an uniform forward primer F515 and 12 different reverse primers A1 through A12) Cumulative time Day 0 Day 3 Day 15 Day 18 Trial 1 Day 0 (inoculum)Day 3 Day 15 Day 18 Digester 1 Primer A1 Primer A2 Primer A4 Primer A6 Digester 2 N/A Primer A3 Primer A5 Primer A7 Cumulative time Day 28 Day 32 Day 36 Day 39 Day 40 Day 42 Trial 2 Day 0 Day 4 Day 8 Day 11 Day 12 Day 14 Digester 1 Primer A8 Primer A10 Primer A1 2N/A Primer A3 N/A Digester 2 Primer A9 Prime A11 Primer A1 Primer A2 Primer A4 Primer A5 Set 1 Set 2

PAGE 150

150 Table 5-2. Substrate charac teristics and loaded quantities for digesters 1 and 2 Sugar beet tailings Unit Run 1 Run 2 TS % (wt/wt) 10.91%.20%10.90%.35% VS % (wt/wt) 9.39%.75%9.94%.31% Digestion loading Unit Wet weight kg 0.3 0.3 Dry weight kg 0.033 0.033 Volatile matter kg 0.028 0.030 Inoculumn added L 3 3 Packing density kg/m3, wet weight basis 100100 Packing density kg/m3, dry weight basis 10.9110.9

PAGE 151

151 Table 5-3. Analysis of 16S-rDNA sequence at taxon rank of order. Time Day 0 Day 3 Day 5 Day 18 Day 28 Day 32 Day 36 Day 39 Day 40 Day 42 aI bD1 cD2 D1 D2 D1 D2 D1 D2 D1 D2 D1 D2 D2 D1 D2 D2 dBacteria g99.66 99.58 100.00 95.43 99.46 98.54 98.30 98.89 99.01 92.45 99.39 96.48 100.00 100.00 98.40 100.00 100.00 eAcidobacteria 0.27 0.00 0.00 0.00 0.00 0.15 0.00 0.37 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 fAcidobacteriales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Actinomycetales 0.20 0.00 0.00 0.00 0.00 0.15 0.00 0.19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Bifidobacteriales 0.00 0.00 0.00 1.83 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Chloroflexi 0.00 0.00 0.00 0.00 0.00 0.15 0.00 0.00 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Anaerolinaeles 0.00 0.00 0.00 0.00 0.00 0.15 0.00 0.00 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Bacteroidetes 4.53 2.12 33.64 7.47 46.48 17.72 52.30 28.39 53.96 23.27 48.78 24.62 40.74 41.37 34.80 45.23 42.20 Bacteroidales 3.92 2.12 33.64 7.32 46.48 17.13 51.62 27.64 53.63 23.27 48.78 23.62 40.74 41.37 34.80 45.23 42.20 Flavobacteriales 0.14 0.00 0.00 0.15 0.00 0.59 0.00 0.74 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Sphingobacteriales 0.47 0.00 0.00 0.00 0.00 0.00 0.68 0.00 0.33 0.00 0.00 1.01 0.00 0.00 0.00 0.00 0.00 Firmcutes 32.81 80.04 59.29 58.38 42.14 42.90 24.19 29.31 17.82 31.76 20.12 37.19 37.04 26.10 20.00 22.11 21.97 Bacillales 6.22 17.62 25.23 17.99 6.23 8.05 8.18 2.60 0.83 5.97 2.44 8.04 0.00 1.20 2.00 0.50 0.58 Clostridiales 20.09 52.87 33.33 30.49 35.37 27.82 14.82 21.71 15.51 24.21 17.68 26.13 36.11 24.10 17.60 20.10 19.65 Erysipelotrichales 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Halanaerobiales 0.54 0.85 0.10 1.37 0.00 0.44 0.00 0.93 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Lactobacillales 0.00 0.42 0.31 0.15 0.00 0.15 0.00 0.00 0.00 0.31 0.00 0.50 0.00 0.00 0.00 0.50 0.00 Natranaerobiales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Thermoanaerobacterales 5.82 8.28 0.31 8.38 0.54 6.44 1.19 3.90 1.49 0.94 0.00 2.51 0.93 0.80 0.40 1.01 1.73 Thermolithobacterales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Planctomycetes 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.56 0.00 0.00 0.00 0.00 0.00 0.00 0.80 0.00 0.00 Planctomycetales 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.56 0.00 0.00 0.00 0.00 0.00 0.00 0.80 0.00 0.00 Proteobacteria 40.80 11.46 5.92 9.15 2.85 11.42 6.98 18.00 7.92 5.03 0.61 6.53 0.93 2.81 6.80 2.01 1.73 Alteromonadales 0.07 0.00 0.00 0.15 0.00 0.00 0.17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Burkholderiales 32.81 8.92 4.88 6.40 1.76 8.78 3.75 8.72 5.45 2.52 0.61 4.02 0.93 2.41 2.80 2.01 1.73 Campylobacterales 1.56 1.27 0.00 0.30 0.00 0.15 0.00 0.56 0.00 0.00 0.00 0.00 0.00 0.00 0.40 0.00 0.00 Caulobacterales 0.00 0.00 0.00 0.15 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Chromatiales 0.41 0.00 0.00 0.00 0.14 0.15 0.00 0.74 0.17 0.31 0.00 0.50 0.00 0.00 0.40 0.00 0.00 Desulfovibrionales 0.41 0.21 0.00 0.00 0.00 0.00 0.00 0.19 0.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Desulfuromonadales 0.20 0.00 0.00 0.00 0.00 0.00 0.68 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

PAGE 152

152 Table 5-3. Continued. Time Day 0 Day 3 Day 5 Day 18 Day 28 Day 32 Day 36 Day 39 Day 40 Day 42 aI bD1 cD2 D1 D2 D1 D2 D1 D2 D1 D2 D1 D2 D2 D1 D2 D2 Enterobacteriales 0.07 0.00 0.00 0.15 0.00 0.00 0.00 0.19 0.00 0.00 0.00 0.00 0.00 0.40 0.40 0.00 0.00 Hydrogenophilales 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.19 0.33 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Methylococcales 0.07 0.00 0.00 0.00 0.14 0.15 0.00 0.19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Myxococcales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.40 0.00 0.00 Pseudomonadales 1.76 0.21 0.73 0.00 0.41 0.15 0.00 0.37 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Rhizobiales 0.07 0.00 0.00 0.46 0.00 0.15 0.17 0.37 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Rhodobacterales 1.76 0.00 0.10 0.15 0.00 0.44 0.17 2.41 0.33 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Rhodocyclales 0.27 0.00 0.21 1.22 0.14 0.73 1.87 0.37 1.16 1.26 0.00 0.50 0.00 0.00 0.00 0.00 0.00 Rhodospirillales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Sneathiellales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Sphingomonadales 0.07 0.21 0.00 0.00 0.00 0.00 0.00 2.60 0.00 0.94 0.00 1.51 0.00 0.00 2.40 0.00 0.00 Syntrophobacterales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Thiotrichales 0.00 0.00 0.00 0.00 0.00 0.00 0.17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Xanthomonadales 0.81 0.64 0.00 0.15 0.27 0.73 0.00 1.11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Spirochaetes 3.25 1.49 0.10 5.64 0.00 4.69 0.17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Spirochaetales 3.25 1.49 0.10 5.64 0.00 4.69 0.17 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Synergistetes 13.19 1.70 0.31 2.90 1.49 12.45 8.69 7.98 6.27 12.89 1.22 12.56 0.93 3.21 18.80 3.02 0.58 Synergistales 13.19 1.70 0.31 2.90 1.49 12.45 8.69 7.98 6.27 12.89 1.22 12.56 0.93 3.21 18.80 3.02 0.58 Tenericutes 0.07 0.64 0.10 0.76 0.00 0.29 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Acholeplasmatales 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Haloplasmatales 0.00 0.64 0.10 0.76 0.00 0.29 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Thermotogae 4.53 2.12 0.62 11.13 6.50 8.78 5.96 14.29 13.04 19.18 28.66 15.58 20.37 26.51 17.20 27.64 33.53 Thermotogales 4.53 2.12 0.62 11.13 6.50 8.78 5.96 14.29 13.04 19.18 28.66 15.58 20.37 26.51 17.20 27.64 33.53 4Archeae 0.34 0.42 0.00 2.74 0.54 1.46 1.70 1.11 0.99 7.55 0.61 3.52 0.00 0.00 1.60 0.00 0.00 5Euryarchaeota 0.34 0.42 0.00 2.74 0.54 1.46 1.70 1.11 0.99 7.55 0.61 3.52 0.00 0.00 1.60 0.00 0.00 6Methanobacteriales 0.27 0.21 0.00 0.00 0.54 0.15 0.00 0.56 0.00 0.63 0.61 0.00 0.00 0.00 0.00 0.00 0.00 Methanomicrobiales 0.07 0.21 0.00 0.46 0.00 0.88 0.00 0.56 0.17 2.83 0.00 3.02 0.00 0.00 1.60 0.00 0.00 Methanosarcinales 0.00 0.00 0.00 2.29 0.00 0.44 1.70 0.00 0.83 4.09 0.00 0.50 0.00 0.00 0.00 0.00 0.00 a Inoculum b Digester 1, c Digester 2, d Taxonomic domain e Bacterial and archaeal phyla are assigned according to National Center for Biotechnology Inform atin (NCBI) Taxonomy database f Bacterial and archaeal orders are assigned according to National Center for Bi otechnology Informatin (NCBI) Taxonomy database g Relative abundance are calculated as a phylogenenic group in percentage of the 16s rDNA reads analyzed

PAGE 153

153 Figure 5-1. Methane yield from digesters 1 and 2

PAGE 154

154 Figure 5-2. SOCD and VOA prof iles of digester 1 and digester 2

PAGE 155

155 A Figure 5-3. Estimate of Chao1 coverage at taxonomic ranks of phy lum, class, order, family and genus: A) digester 1 B) digester 2

PAGE 156

156 B Figure 5-3. Continued

PAGE 157

157 Figure 5-4. Microbial community dynamics of digester 1 and 2

PAGE 158

158 CHAPTER 6 CONCLUSIONS AND FUTURE WORK Conclusions The conclusions from this research work are presented in two parts. Part one focuses on the value addition through anaerobic digestio n of byproducts from common liquid biofuel production process. The objec tive to develop a process for simultaneous waste treatment and energy production was fulfilled. The BMP assays determined high methane producing potential of the byproducts and the following long term laboratory scale anaerobic digestion studies provided the feasibility and economics to apply a full scale process to biodiesel and ce llulosic ethanol industry. The glycerol byproduct was co llected from a pilot biodiese l plant at University of Florida. Mesophilic BMP assays were c onducted on glycerol byproduct and compared with BMP of 99.5% grade glycerol. GBP me thane potential was 456 ml CH4 at STP (ml sample)-1 compared to 372 ml CH4 at STP (ml sample)-1 for pure glycerol. The higher methane yield from glycerol product was attri buted to the mixture containing long chain fatty acids, residual oil and methanol that are better energy sources than glycerol itself. The energy that could be recovered from bi ogasifying glycerol byproduct was calculated to be 18300 kJ /kg using the heat of combustion of methane of 37 kJ L-1. This is comparable to the energy that can be obtained from directly combusting glycerol. The Gompertz equation was used to fit the cu mulative methane production data for the glycerol byproduct. The simulation results exhibited a biphasic trend (fraction 1 and fraction 2 in Figure 2-1). Fraction 1 comple ted methane production in a short period of time, suggesting its ease of degradability; wh ile fraction 2 has a similar methane production profile and methane yiel d to that of 99.5% grade glycerol. Fraction 2 is

PAGE 159

159 glycerol and fraction 1 is methanol. Overloading issues were found in some BMP assays containing 2 ml glycerol byproduct as seen from low methane production and accumulation of volatile organic acids. Co-di gestion of 2 ml of glycerol byproduct with a readily degradable substrate appeared to alleviate the overloading issue, but it still took twice as long to reach a similar yield achieved in assays loaded with 1 ml of glycerol byproduct. Continuous digestion of glycero l byproduct was further evaluated in a laboratory scale 5 L anaerobic digester using G BP as the sole substrate. Methane yield at a feed rate of 0.67 ml (L) -1 (day)-1 was around 450 ml CH4 at STP (ml substrate)-1, which was comparable to the yield obtai ned from the BMP assays. However, when feeding rate was then increased to 1 ml (L) -1(day)-1 the yield decreased to 200 ml CH4 at STP (ml substrate)-1. However, feeding glycerol by product as sole substrate to anaerobic digesters may easily cause an organi c overloading problem. Therefore, dilution or co-digestion with anot her substrate is recommended. A similar approach was used to investi gate the anaerobic digestion of stillage produced from cellulosic ethanol production. BMP assays were first carried out to determine the methane potential and later stillage was fed to a laboratory scale (15 L) continuous anaerobic digester for ov er two months to observe long term digestibility of stillage. In industry, distillation is usually conducted at 60-70oC to separate ethanol from the fermentation mixture. To take advantage of the hi gh temperature, an onsite anaerobic digester can be operated at thermophilic temperature for high substrate degradation rates. Thus, the BMP assays and anaerobic digestion of stillage were conducted at 55oC to simulate this situation. Sinc e fibers contained in the stillage were expected to be difficult to degrade and would cause a clogging problem for anaerobic

PAGE 160

160 digestion. The stillage was filtered through a sieve with 0.5 mm mesh. The filtered stillage and the fraction retained by sieve are referred as stillage filtrate and stillage residue. The BMP assays showed stillage filtrate and residue produced 70% and 30% of the methane potential of the whole stillage, respectively. The stillage residue contains only 18% solids, and calculation showed on ly 50% of methane was produced from the solid fraction. In combination, only 15% of total methane was pr oduced from the solid fraction. In a continuous digester the metha ne yield of stillage filtrate was between 9 to 10 L CH4 (L substrate) -1(day)-1. The sCOD removal efficiency was around 80%. Nutrient analysis showed ammonia and phosphate was not limiting in the digester. A mass and energy balance was developed for in tegrating anaerobic digestion to the cellulosic ethanol production proce ss in the Biofuel Pilot Plant at University of Florida. It showed that the biogas can cover 70% of the energy consumed for process heat and steam. The second part of this research work studied the effect of mixing on anaerobic digestion under thermophilic conditions. Digestion performance at mixed and non-mixed conditions was compared, including methane yi eld, methane production rate, sCOD and VOA degradation. To elucidate the differences, the microbial communities were investigated using 16S rDNA analysis. A tota l of eight experimental runs were carried out at laboratory scale for mixed and non-mixed digestion, respectively. Sugar beet tailings were used as the feedstock. The non-mixed digester was added with a bulking agent to prevent substrate floatation and compaction. In all eight runs, the mixed digester consistently showed lower methane yield, delayed methane production and higher accumulation of sCOD and VOA than the non-mixed counter part. More sCOD

PAGE 161

161 and VOA accumulated in the mixed digester as the digester liquor recovered from the last run was used for inoculation in the next run. This suggested constant exposure to mixing might result in loss of activity of anaerobic bacteria. The microbial community structure was then investigated using cl assical 16s rDNA clone analysis and 454 pyrosequencing. Both techniques showed that methanogens are in lower relative abundance in the mixed digester, explaining the low methane production. The relative abundance of bacteria Anaerobaculum was also found to be lower in the mixed digester but high in the non-mixed digester. The bac terium was reportedly associated with a propionate degrading microbial consortium. Lack of Anaerobaculum could be one of the reasons for proipionic acid accumulati on in the mixed digester. The microbial communities analyzed using 454 pyrosequenc ing showed higher diversity. Some bacteria that were not detected with 16s rDNA clone analysis were detected by 454 pyrosequencing, such as Pelotomaculum The bacterium is known to syntrophically grow with methanogens and play an important role in propionate degradation. The relative abundance of syntrophic bacteria an d methanogens decreased with time in the mixed digester and was overall lower than in the non-mixed digest er. It was generally agreed that mixing may inhibit methane producti on by disrupting the syntrophic relation even at low intensities of 1.1 W/m3.Therefore, an alternative hypothesis was made that hydrogen was producing during the mix ed digestion and hindered the methane production. This was verified by the i dentification of high abundant hydrogen producing bacteria as Ruminococcus and Acetanaerobacterium in the mixed digester. To further demonstrate the syntrophic pr opionate degradation was not af fected under the mixing condition used in the study, propionic acid wa s provided as the only substrate. The VOA

PAGE 162

162 analysis showed steady propionic acid degra dation within 15 days for consecutive 3 runs. The study of effect of mixing sugges ted practice of continuous mixing for high solid digesters may not be always feasible. Th is is supported by some studies that reported high methane yield at non-mixed or less mixed condition than in continuously mixed digesters. Future Work This research work attempted to i nve stigate the microbiol ogical nature of anaerobic digestion and included t he application of the process in treating wastes from production of common biofuels. The study presented here opens some areas for further research. The following topics of interest can be addressed: Understanding the role of co-subst rate in promoting GBP degradation Exploring an effective strategy to m anage stillage residue containing high content of fibers and lignin that are difficult to degrade in anaerobic digestion. Possible options for example are using the fiber re sidues (mainly lignin) as fuel or as feedstock for bioproducts. Develop a secondary process to treat the effluent after the anaerobic digestion of stillage to further reduce the sCOD conc entration and remove nutrients (primary phosphate) to be able to recycle the water within the biorefinery. Investigating the effect of mixing mode in addition to mixing intensity on anaerobic digestion, i.e., comparing digestion perfo rmance at continuous mixing, intermittent mixing and no-mixing condition. Investigating the metabolism of hydrogen producing bacteria (e.g., Ruminococcus) in co-culturing with methanogens and syntrophic bacteria to provide a better understanding of hydrogen in hibition on methanogensis.

PAGE 163

163 APPENDIX A EMERGY ANALYSIS OF BIODIESEL PRODCTION FROM WASTE VEGETABLE OIL To evaluat e the energy economics of a bi odiesel production process in a pilot plant at University of Florida, emergy analysis was conducted. Waste vegetable oil was used as the feedstock to produce biodie sel. The results are showed in Table A1. Table A-1. Emergy evaluation of biodiesel production Item QuantityUnit Unit Emergy Values Emergy /batch (sej/unit) (E10 sej/batch) Renewable inputs None Non-renewable storage None Sum of free inputs 0 Purchased inputs Infrastructure Storage tank 1 1.11E+15 212.32 Steel tank 3 2.40E+14 138.18 Steel drum 5 1.21E+14 115.83 Tote tank 1 4.42E+14 84.76 Pump 3 3.15E+14 181.52 PVC pipe 15 2.84E+13 81.77 Heating tape 11 6.48E+13 136.91 Pickup truck 1 2.27E+16 4360.84 Drum insulator 1 1.85E+14 35.43 Sum of infrastructure input 5347.58 Operational inputs Waste vegetable oil 284L 1.01E+11 2854.66 Methanol 62.5L 5.63E+11 3517.17 KOH 1590g 5.68E+10 9031.20 Electricity 33458KJ 2.00E+08 669.16 Natural gas 159237KJ 2.86E+07 454.96 Water 151.4L 3.39E+08 5.13 Transport 2.58tkm 1.10E+12 283.80 Labor 4hour 1.46E+10 5.84 Sum of operational cost 15822.41 Total emergy 21169.99 Output 284 L 7.45E+11 21169.99 Biodiesel UEV w/labor 284L 7.45E+11 UEV w/o labor 284L 7.45E+11

PAGE 164

164 APPENDIX B DEGRADATION OF PROPION IC ACID IN M IXED ANAEROBIC DIGESTION Chapter 4 and Chapter 5 presented result s from several (eight) trials of sugarbeet tailings digestion. The results consistently showed lower methane yield and slower methane production ra te from the mixed digest er, compared to the nonmixed digester. The mixing speed of 100 RP M provided a volumetric power input at 1.1 W/m3, which was within the typical range of mixing intensity used in anaerobic digestion ( Grady, Daigger et al. 1999). Different hypotheses have been proposed to interpret the adverse effect of mi xing. Mo st studies tend to agree mixing force may interfere with the syntrophic associations between bacteria and methanogens by disrupting their spat ial juxtaposition ( Conrad, Phelps et al. 1985; Whitmore, Lloyd et al. 1987; Dolfing 1992; McMahon, Stroot et al. 2001; Stroot, McMahon et al. 2001; Vavilin and Angelidaki 2005; Hoffmann, Garcia et al. 2008; Kaparaju, Buendia et al. 2008). The microbial c ommunity analysis identified abundant hydrogen producing bacteria in the mixed digester. Based on th is result, an alternative hypothesis was postulated that hydrogen was produced in the mixed digester and inhibited the intermediate fermentation, particularly vo latile organic acid (VOA) degradation. Acetic acid and propionic acid had pronounced concentration among VOAs monitored. While most of accumulated acetic acid eventually degraded propionic acid still persisted after 20 days of digesti on. To investigate whether the propionic acid degradation was hindered by the mixing, further experiments were conducted by providing propionic acid as the sole s ubstrate. Since the main substrate (sugar) for hydrogen producing bacteria was not ava ilable, the inhibi ting effect from hydrogen was limited. Therefore, if the propionic acid still degraded, it could be attributed to the syntrophic associations that survived under the mixing condition.

PAGE 165

165 The sugar beet tailings was provided by America Crystal Sugar Company, Minnesota and stored at 4 C before using. The tailings were washed using tap water before loading to the digesters. Wash wate r was discarded. Analytical grade (99.5% pure) propionic acid was purchased from Fisher Scientific. The digester is the same as that used in Chapters 4 and 5. It was constructed by modifying a 3 L Pyrex glass jar. Gas production from the digesters was measured by a positive displacement gas meter. Re fer to Chapter 4 for detailed information about the anaerobic digesters and gas meters. The digester contents werecontinuously mixed at 80 RPM. A 50.8 mm x 9.5 mm PTFE coated polygon stirrer bar was placed in the digester and th e digester was mixed by a large volume magnetic stirrer (Scienceware Cool Sti rrer). The digester was placed in a 55C incubator throughout the experiments. The digester was initially loaded with 0.3 kg sugar beet tailings and inoculated with 3 L leachate taken from a pilot scale anaerobic digester digesting food wastes. When the digestion of sugarbeet tailings was completed, 1.3 L of digester liquor wa s recovered and used as inoculum in the following experiments. Three experimental trials were carried out. A solution was prepared by mixing propionic acid in 100 ml deionized water and dosed to the digester. The solution pH was adjust ed between 5 and 6 by adding potassium hydroxide. In trial 1, 1. 3 g propionic acid was dosed at the beginning. The trial was ended when propionic acid was degraded to lo w concentration. The digester liquor was kept and 1.4 g propionic acid was dosed again to start trial 2. Likewise, when propionic acid degraded, trial 2 was consi dered complete and the digester was again dosed with 2 g propionic acid to initiate tria l 3. Each trial was operated for 12 to 20 days.

PAGE 166

166 The digester liquor was sampled periodicall y for pH, volatile organic acid (VOA) analysis. pH was measured using an Accument pH meter. Samples were centrifuged at 8000 RPM for 10 minutes (Fisher Marathon micro H centrifuge) and filtered using Millipore filter paper (por e size 1.2 um). VOA analysis was conducted using Shimadzu gas chromatograph (GC-9AM equipp ed with a flame ionization detector) for acetic, propionic, isobutyric, butyric, is ovaleric and valeric acid concentrations. The digester content was mixed at 80 RPM. The speed was reduced (compared to 100 RPM used in Chapter 4 and 5) due to the decreased volume. Eighty RPM gave the volumetric power input of 1.3 w/m3 in 1.3 L digester liquor. This is comparable to the volumetric power input (1.1 w/m3) used in Chapter 4 and 5. For calculation of volumetric pow er input of the digester, please refer to Chapter 4. The VOA and pH profiles of the digeste r are shown in Figures 6-1 and 6-2. Three experimental trials were carried out ov er 53 days. pH of all trials was kept above 7.5. The dosing quantities were det ermined to provide a close concentration as observed in the mixed digester in Chapter 4 and 5. VOA analysis revealed that initial concentration of propioni c acid in trial 1, 2 and 3 wa s 1 g/L, 1 g/L and 1.6 g/L, respectively consistent with the amount dos ed. In trial 3, the in itial concentration of propionic acid was higher than in tria l 1 and 2 because more propionic acid was dosed. In all trials, propionic acid ex hibited a steady decrease, indicating the syntrophic community could survive at the mixing intensity of 1.3 w/m3. In trial 1, propionic acid spiked at 1 g/ L and decreased to 0.4 g/L after 4 days. Degradation of propionic acid was delayed in trials 2 and 3. Degradation was not noticeable until 8 days. Acetic acid was produced from the degradation of propionic acid. Though the digester in trial 3 was pulsed a higher dos e, most propionic acid degraded in 14 days, suggesting methanogensis was not inhibi ted at propionic acid concentration of

PAGE 167

167 1.6 g/L. In Chapter 5, the mixed digester accumula ted higher concentration of propionic acid (1.4 g/L) than the non-mixed digester. This was speculated to result in low methane production in the mixed digester because the propionic acid accumulation being reported inhibitory to methanogens. However, the observed propionic acid degradation in trial 3 may exclude the speculation In summary, the three experimental tria ls showed consistent propionic acid degradation and provided the evidence t hat syntrophic relation between methanogens and propionate utilizing bacteria was not disrupted by the mixing used in Chapter 4 and 5.

PAGE 168

168 Figure B-2. pH profile

PAGE 169

169 Figure B-3. VOA profile

PAGE 170

170 LIST OF REFERENCES Alkaya, E. and G. N. Demirer (2011). "Anaerobic acidification of sugar-beet processing wastes: Effect of operational parameters." Bioma ss & Bioenergy 35(1): 32-3 9. Alkaya, E. and G. N. Demirer (2011). "Anaerobic mesophilic co -digestion of sugar-beet processing wastewater and beet-pulp in batch reactors." Renewable Energy 36(3): 971-975. Alptekin, E. and M. Canakci (2008). "Determi nation of the density and the viscosities of biodiesel-diesel fuel bl ends." Renewable Energy 33(12): 2623-2630. American Public Health, A., A. Americ an Water Works, et al. (1999). Inorganic nonmetallic consituents. Standard methods for the examination of water and wastewater. L. S. Clesceri, A. E. Gr eenberg and A. D. Eaton. Washington, DC, American Public Health Association : Part 4000 4075-4117. Amo, T., M. L. F. Paje, et al. (2002). "Pyrobaculum ca lidifontis sp. nov., a novel hyperthermophilic archaeon that grows in atmospheric air." Archaea 1 (2): 113121. Amon, T., B. Amon, et al. ( 2006). "Optimising methane yield fr om anaerobic digestion of manure: Effects of dairy systems and of glycerine supplem entation." International Congress Series 1293 : 217-220. Angenent, L. T., K. Karim, et al. (2004). "Production of bioen ergy and biochemicals from industrial and agricultural wastew ater." Trends in Biotechnology 22(9): 477-485. Antoni, D., V. V. Zverlov, et al. (2007). "B iofuels from microbes." Applied Microbiology and Biotechnology 77 : 23-35. Antonopoulou, G., I. Ntaikou, et al. ( 2007). "Biohydrogen production from sweet sorghum biomass using mixed acidogen ic cultures and pure cultures of Ruminococcus albus." Global Nest Journal 9 (2): 144-151. Baena, S., M. L. Fardeau, et al. (1998). "Aminobacterium colombiense gen. nov. sp. nov., an amino acid-degrading anaerobe isolated from anaerobic sludge." Anaerobe 4 (5): 241-250. Balat, M. (2011). "Production of bioethanol from lignocellulosic materials via the biochemical pathway: A review." E nergy Conversion and Management 52(2): 858-875. Balk, M., J. Weijma, et al. (2002). "Thermo toga lettingae sp nov., a novel thermophilic, methanol-degrading bacterium isolated from a thermophilic anaerobic reactor." International Journal of Systematic and Evolutionary Microbiology 52: 1361-1368.

PAGE 171

171 Banks, C. J., E. A. Zotova, et al. (2010) "Biphasic production of hydrogen and methane from waste lactose in cyclic-batch r eactors." Journal of Cleaner Production 18(Supplement 1): S95-S104. Barredo, M. S. and L. M. Evison (1991). "Eff ect of Propionate Toxicity on Methanogenenriched Sludge, Methanobrevibacter-smithii and Methanospirillm-hungatii at Different pH Values.." Applied and Environmental Microbiology 57(6): 1764-1769. Barta, Z., K. Reczey, et al. (2010). "Tec hno-economic evaluation of stillage treatment with anaerobic digestion in a softwood-to -ethanol process." Biotechnology for Biofuels 3 (21): (15 September 2010). Binladen, J., M. T. P. Gil bert, et al. (2007). "The Use of Coded PCR Primers Enables High-Throughput Sequencing of Multiple Homolog Amplification Products by 454 Parallel Sequencing." PLoS One 2 (2): Article No.: e197. Bitton, G. (2005). Anaerobic digestion of wastewater and biosolids. Wastewater Microbiology. Hoboken, N.J., Wiley-Liss : 346-360. Bolzonella, D., P. Pavan, et al. (2006). "Anaerobic co-digestion of sludge with other organic wastes and phosphorus re clamation in wastewater treatment plants for biological nutrients removal." Water Science and Technology 53(12): 177-186. Botsch, K. C. and R. Conrad (2011). "Fractionation of stable carbon isotopes during anaerobic production and degrad ation of propionate in defined microbial cultures." Organic Geochemistry 42(3): 289-295. Cardinali-Rezende, J., A. M. M. Moraes, et al. (2011). "Phylogenetic and physiological characterization of organic waste-degrading bacterial communities." World Journal of Microbiology & Biotechnology 27(2): 245-252. Chen, C.-C., C.-Y. Lin, et al. (2002). "Acidbase enrichment enhances anaerobic hydrogen production process." Applied Microbiology and Biotechnology 58(2): 224-228. Chen, T., P. Chynoweth, et al. (1990). "Anaerobi c digestion of munici pal solid waste in a nonmixed solids concentrating digestor." Applied Biochemistry and Biotechnology 24-25 (1): 533-544. Chen, T. H., P. Chynoweth, et al. ( 1990). "ANAEROBIC-DIGESTION OF MUNICIPAL SOLID-WASTE IN A NONMIXED SOLI DS CONCENTRATING DIGESTER." Applied Biochemistry and Biotechnology 24-5 : 533-544. Chen, X., R. T. Romano, et al. (2008). Anaerobic Co-Di gestion of Dairy Manure and Glycerin. ASABE Annual International Meeting Providence, Rhode Island. Conrad, R., T. J. Phelps, et al. (1985). "Gas Metabolism Evidence in Support of the Juxtaposition of HydrogenProducing and Methanogenic Bacteria in Sewage

PAGE 172

172 Sludge and Lake Sediments." Applied and Environmental Microbiology 50(3): 595-601. Dague, R. R., R. E. McKinney, et al. ( 1970). "Solids retention in anaerobic waste treatment systems." Journal Wa ter Pollution Control Federation 42 (2 PART 2): R29-R46. Danko, A. S., F. Pinheiro, et al. (2008). "Effect of methanogenic inhibitors, inoccula type, and temperature on biohydrogen pr oduction from food components." Environmental Engineer ing and Management Journal 7 (5): 531-536. Dar, S. A., R. Kleerebezem et al. (2008). "Competition and coexistence of sulfatereducing bacteria, acetogens and me thanogens in a lab-scale anaerobic bioreactor as affected by changing s ubstrate to sulfat e ratio." Applied Microbiology and Biotechnology 78(6): 1045-1055. de Bok, F. A. M., H. J. M. Harmsen, et al. (2005). "The first true obligately syntrophic propionate-oxidizing bacterium, Pelotomaculum schinkii sp nov., co-cultured with Methanospirillum hungatei, and emended description of the genus Pelotomaculum." International Journal of Systematic and Evolutionary Microbiology 55: 1697-1703. Defour, D., D. Derycke, et al. (1994). "F ield experience with different systems for biomass accumulation in anaerobic r eactor technology." Water Science and Technology 30(12): 181-191. Demirbas, A. (2008). "Biofuels sources, biofuel policy, biofuel economy and global biofuel projections." Ener gy Conversion and Management 49(8): 2106-2116. Demirbas, A. (2011). "Competitive liquid bi ofuels from biomass." Applied Energy 88(1): 17-28. Dolfing, J. (1992). "The ener getic consequences of hydrogen gradients in methanogenic ecosystems." Fems Microbiology Letters 101 (3): 183-187. Elberson, M. A. and K. R. Sowers (1997). "Isolation of an aceticlastic strain of Methanosarcina siciliae from marine c anyon sediments and emendation of the species description for Methanosarcina si ciliae." International Journal of Systematic Bacteriology 47(4): 1258-1261. Emiliani, D. and M. Pistocchi (2006). "Biodi esel production from vegetable oils." Industria Saccarifera Italiana 99(4): 91-98. Etchebehere, C., M. E. Pavan, et al. ( 1998). "Coprothermobacter platensis sp. nov., a new anaerobic proteolytic thermophilic bac terium isolated from an anaerobic mesophilic sludge." International Jour nal of Systematic Bacteriology 48: 12971304.

PAGE 173

173 Fang, C., K. Boe, et al. (2011). "Anaerobi c co-digestion of desugared molasses with cow manure; focusing on sodium and pot assium inhibition." Bioresource Technology 102 (2): 1005-1011. Fang, H. H. P. and H. Liu (2002). "Effect of pH on hydrogen production from glucose by a mixed culture." Bioresource Technology 82 (1): 87-93. Felde, A. v. (2008). "Trends and developm ents in energy plant breeding special features of sugarbeet." Sugar Industry / Zuckerindustrie 133 (5): 342-345. Fox, P., M. T. Suidan, et al. (1988). "ANAEROBIC TREATMENT OF A BIOLOGICALLY INHIBITORY WASTE-WATER." Journal Water Pollution Control Federation 60(1): 86-92. Fukuzaki, S., N. Nishio, et al. (1990). "Inhi bition of the fermentat ion of propionate of methane by hydrogen, acetate, and propi onate." Applied and Environmental Microbiology 56(3): 719-723. Ghaly, A. E. and R. M. Benhassan (1989). "Cont inuous production of biogas from dairy manure using an innovative no-mix reactor." Applied Biochemistry and Biotechnology 20-1 : 541-559. Godon, J. J., J. Moriniere, et al. (2005). "Rarity associat ed with specific ecological niches in the bacterial world: the 'Synergistes' exampl e." Environmental Microbiology 7 (2): 213-224. Gonzlez, D., V. Santos, et al. (2011). "M anufacture of fibrous reinforcements for biocomposites and hemicellulosic o ligomers from bamboo." Chemical Engineering Journal 167 (1): 278-287. Good, I. J. (1953). "The population frequencies of species and the estimation of population parameter s." Biometrika 40(3-4): 237-264. Grabowski, A., B. J. Tindall, et al. (2005). "Petrimonas sulfur iphila gen. nov., sp nov., a mesophilic fermentative bacterium isolated from a biodegraded oil reservoir." International Journal of Systematic and Evolutionary Microbiology 55: 1113-1121. Grady, C. P. L., G. T. Da igger, et al. (1999). Anaerob ic processes. Biological wastewater treatment. Ne w York, Marcel Dekker. 2 rev a expa: 645. Grady, C. P. L., G. T. Da igger, et al. (1999). Anaerobic processes. Biological wastewater treatment. Ne w York, Marcel Dekker. 2 rev a expa: 601-620. Grady, C. P. L., G. T. Daig ger, et al. (1999). Fundamentals of biochemical operations. Biological wastewater treatment New York, Marcel Dekker. 2 rev a expa: 3134.

PAGE 174

174 Hall, E. R. (1992). Anaerobic treatment of wastewaters in suspended growth and fixed film processes. Design of Anaerobic Pr ocesses for the Treatment of Industrial and Municipal Wastes. Boca Raton, CRC Press. 7: 50-71. Hamady, M. and R. Knight (2009). "Microbial community profiling for human microbiome projects: Tools, techniques, and challenges." Genome Research 19 (7): 11411152. Hamady, M., J. J. Walker, et al. (2008) "Error-correcting barcoded primers for pyrosequencing hundreds of samples in multiplex." Nature Methods 5 (3): 235237. Harmsen, H. J. M., B. L. M. Van Kuijk, et al. (1998). "S yntrophobacter fumaroxidans sp. nov., a syntrophic propionate-degrading sulfate-reducing bacterium." International Journal of Systematic Bacteriology 48: 1383-1387. Hatamoto, M., H. Imachi, et al. (2007). "Diversity of anaerobic microorganisms involved in long-chain fatty acid degradation in methanogenic sludges as revealed by RNA-based stable isotope probing." Applied and Environmental Microbiology 73(13): 4119-4127. Hattori, S., Y. Kamagata, et al. (2000). "Thermacetogeniu m phaeum gen. nov., sp nov., a strictly anaerobic, thermophilic, syn trophic acetate-oxidizing bacterium." International Journal of Systematic and Evolutionary Microbiology 50: 1601-1609. Hill, D. T., S. A. C obb, et al. (1987). "Using volatile fa tty-acid relationships to predict anaerobic digester failure." Transactions of the Asae 30 (2): 496-501. Hoffmann, R. A., M. L. Garc ia, et al. (2008). "Effect of shear on performance and microbial ecology of continuously sti rred anaerobic digest ers treating animal manure." Biotechnology and Bioengineering 100 (1): 38-48. Holm-Nielsen, J. B., C. J. Lomborg, et al (2008). "On-line near infrared monitoring of glycerol-boosted anaerobic digestion processe s: Evaluation of process analytical technologies." Biotechnology and Bioengineering 99(2): 302-313. Hong, S. H., J. Bunge, et al. (2006). "P redicting microbial species richness." Proceedings of the Na tional Academy of Sciences of the United States of America 103 (1): 117-122. Hu, H. B. and T. K. Wood (2010). "An evol ved Escherichia coli strain for producing hydrogen and ethanol from gl ycerol." Biochemical and Biophysical Research Communications 391 (1): 1033-1038. Huang, L.-N., Y.-Q. Chen, et al. (2003). "C haracterization of methanogenic Archaea in the leachate of a closed municipal solid waste landfill." Fems Microbiology Ecology 46(2): 171-177.

PAGE 175

175 Huang, L. N., H. Zhou, et al. (2002). "Diver sity and structure of the archaeal community in the leachate of a full-scale recircul ating landfill as examined by direct 16S rRNA gene sequence retrieval." Fems Microbiology Letters 214 (2): 235-240. Huang, L. N., S. Zhu, et al. (2005). "Mol ecular phylogenetic diversity of bacteria associated with the leachate of a closed municipal solid waste landfill." Fems Microbiology Letters 242 (2): 297-303. Hutnan, M., M. Hornak, et al. (2003). "Anaer obic treatment of wheat stillage." Chemical and Biochemical Engineering Quarterly 17(3): 233-241. Iannotti, E. L., Kafkewit.D et al. (1973). "Glucose fermentation products of ruminococcus-albus grown in continuous culture with vibrio-succinogenes changes caused by interspecies transfer of h2." Journal of Bacteriology 114 (3): 1231-1240. Imachi, H., Y. Sekiguchi, et al. (2002). "P elotomaculum thermopropionicum gen. nov., sp. nov., an anaerobic, thermophilic, syntr ophic propionate-oxidizing bacterium." Int J Syst Evol Microbiol 52(5): 1729-1735. Ivanova, G., G. Rkhely, et al. (2009) "Thermophilic biohydr ogen production from energy plants by Caldicellulosiruptor saccharolyticus and comparison with related studies." International J ournal of Hydrogen Energy 34(9): 3659-3670. Jackson, B. E., V. K. Bhupathiraju, et al. (1999). "Syntrophus aciditrophicus sp. nov., a new anaerobic bacterium that degrades fa tty acids and benzoate in syntrophic association with hydrogen-using microor ganisms." Archives of Microbiology 171 (2): 107-114. Janssen, P. H. (2003). "S elective enrichment and purif ication of cultures of Methanosaeta spp." Journal of Microbiological Methods 52(2): 239-244. Jindou, S., J. M. Brulc, et al. (2008). "Ce llulosome gene cluster analysis for gauging the diversity of the ruminal cellulolytic bacterium Ruminococcus flavefaciens." Fems Microbiology Letters 285 (2): 188-194. Johnson, D. T. and K. A. Taconi (2007). "The glycerin glut: Options for the value-added conversion of crude glycerol resulting fr om biodiesel production." Environmental Progress 26(4): 338-348. Juwarkar, A. A., S. K. Singh, et al. (2010). "A comprehensive overview of elements in bioremediation." Reviews in Envir onmental Science and Bio-Technology 9 (3): 215-288. Kaffka, S. R. (2009). "Can feedstock production for biofuels be sustainable in California?" California Agriculture 63(4): 202-207.

PAGE 176

176 Kano, S., T. Mukaidani, et al. (2009). "Div ersity of Indigenous Anaerobes and Methane Conversion System from Reservoir Oil by Indigenous Anaerobes in Depleted Oil Fields." Journal of the Japan Petroleum Institute 52(6): 297-306. Kaparaju, P., I. Buendia, et al. (2008). "E ffects of mixing on methane production during thermophilic anaerobic digesti on of manure: Lab-scale and pilot-scale studies." Bioresource Technology 99(11): 4919-4928. Kaparaju, P., M. Serrano, et al. (2010). "Opt imization of biogas production from wheat straw stillage in UASB r eactor." Applied Energy 87(12): 3779-3783. Kaparaju, P., M. Serrano, et al. (2009). "Bioethanol, biohydrogen and biogas production from wheat straw in a biorefiner y concept." Bioresource Technology 100 (9): 2562-2568. Karakashev, D., D. J. Batstone, et al. ( 2005). "Influence of envir onmental conditions on methanogenic compositions in anaerobi c biogas reactors." Applied and Environmental Microbiology 71(1): 331-338. Karim, K., R. Hoffmann, et al. (2005). "Anaerobic digestion of animal waste: Effect of mode of mixing." Water Research 39(15): 3597-3606. Kato, M. T., J. A. Field, et al. (1993). "High tolerance of methanogens in granular sludge to oxygen." Biotechno logy and Bioengineering 42(11): 1360-1366. Kim, H. W., S. K. Han, et al. (2003). "The optimisation of food waste addition as a cosubstrate in anaerobic digestion of sewage sludge." Waste Management & Research 21 (6): 515-526. Kim, I. S., D. H. Kim, et al. (2000). "Effect of particle size and sodium ion concentration on anaerobic thermophilic food waste diges tion." Water Science and Technology 41(3): 67-73. Kim, J., S. G. Shin, et al (2011). "Common key acidog en populations in anaerobic reactors treating different wastewaters: Molecular identification and quantitative monitoring." Water Research 45(8): 2539-2549. Kim, M., M. Song, et al. (2010). "Gro wth condition and bacteri al community for maximum hydrolysis of suspended organic materials in anaerobic digestion of food waste-recycling wastewater." App lied Microbiology and Biotechnology 85(5): 1611-1618. Kim, S. and B. E. Dale (2004). "Global pot ential bioethanol production from wasted crops and crop residues." Biomass & Bioenergy 26(4): 361-375. Kim, Y., N. S. Mosier, et al (2008). "Composition of corn dry-grind ethanol by-products: DDGS, wet cake, and thin stillage." Bioresource Technology 99(12): 5165-5176.

PAGE 177

177 Klocke, M., P. Mahnert, et al. (2007). "M icrobial community analysis of a biogasproducing completely stirred tank reacto r fed continuously wit h fodder beet silage as mono-substrate." Systemat ic and Applied Microbiology 30(2): 139-151. Klocke, M., P. Mhnert, et al. (2007). "M icrobial community analysis of a biogasproducing completely stirred tank reacto r fed continuously wit h fodder beet silage as mono-substrate." Systemat ic and Applied Microbiology 30(2): 139-151. Klocke, M., E. Nettmann, et al. (2008). "C haracterization of t he methanogenic Archaea within two-phase biogas reactor system s operated with plant biomass." Systematic and Appl ied Microbiology 31(3): 190-205. Kongjan, P., S. O-Thong, et al. (2011). "P erformance and microbial community analysis of two-stage process with extreme thermophilic hydrogen and thermophilic methane production from hydrolysate in UASB reactors." Bioresource Technology 102 (5). Koppar, A. and P. Pullamm anappallil (2008). "Single-st age, batch, leach-bed, thermophilic anaerobic digestion of s pent sugar beet pulp." Bioresource Technology 99(8): 2831-2839. Koster, I. W. (1988). Microbial, chemical and technological aspects of the anaerobic dedgradation of organic pollutant s. Biotreatment System s. D. L. Wise. Boca Raton, Fla., CRC Press. 1: 280-300. Kotsyurbenko, O. R., K. J. Chin, et al. (2004). "A cetoclastic and hydrogenotrophic methane production and methanogenic populati ons in an acidic West-Siberian peat bog." Environmental Microbiology 6 (11): 1159-1173. Krause, L., N. N. Diaz, et al. (2008). "Taxonomic compos ition and gene content of a methane-producing microbial community isolated from a biogas reactor." Journal of Biotechnology 136 (1-2): 91-101. Krber, M., T. Bekel, et al. (2009). "Phylogenetic characterization of a biogas plant microbial community integrating clone library 16S-rDNA sequences and metagenome sequence data obtained by 454-pyrosequencing." Journal of Biotechnology 142 (1): 38-49. Krupp, M., J. Schubert, et al. (2005). "Feasib ility study for co-digestion of sewage sludge with OFMSW on two wastewater treatment plants in Germany." Waste Management 25(4): 393-399. Kuang, Y., M. Lepesteur, et al (2002). "Influence of co-s ubstrates on structure of microbial aggregates in long-chain fatty acid-fed anaerobic digesters." Letters in Applied Microbiology 35(3): 190-194.

PAGE 178

178 Kuang, Y., P. Pullammanappallil, et al. (2006). "Recovery of oleate-inhibited anaerobic digestion by addition of simple substrates ." Journal of Chemical Technology and Biotechnology 81(6): 1057-1063. Kumar, K., C. J. Rosen, et al. (2002). "Kinetics of nitr ogen mineralization in soils amended with sugar beet processing by-pr oducts." Communications in Soil Science and Plant Analysis 33(19-20): 3635-3651. Kumar, L., R. Chandra, et al. (2010). "Can the same steam pretreatment conditions be used for most softwoods to achieve good, enzymatic hydrolysis and sugar yields?" Bioresource Technology 101 (20). Labat, M. and J. L. Garcia (1986). "St udy on the development of methanogenic microflora during anaerobicdigestion of sugar-beet pul p." Applied Microbiology and Biotechnology 25 (2): 163-168. Larsen, J., M. O. Petersen, et al. (2008). "The IBUS proce ss Lignocellulosic bioethanol close to a commercial reality." Chemical Engineering & Technology 31(5): 765772. Lay, J. J., Y. Y. Li, et al. (1998). "Mathematical model for methane production from landfill bioreactor." Journal of Environmental Engineering-Asce 124 (8): 730-736. Leclerc, M., J. P. Delgenes, et al. (2004). Diversity of the arc haeal community in 44 anaerobic digesters as determined by si ngle strand conforma tion polymorphism analysis and 16S rDNA sequencing." Environmental Microbiology 6 (8): 809-819. Lele, S. S., P. J. Rajadhyaksha, et al. (1989) "Effluent treatment fo r alcohol distillery thermal pretreatment with energy recovery." Environmental Progress 8 (4): 245252. Lettinga, G., J. Field, et al. (1997). "Adv anced anaerobic wastewater treatment in the near future." Water Science and Technology 35(10): 5-12. L'Haridon, S., M. L. Miroshnichenko, et al. (2002). "Petrotoga olearia sp nov and Petrotoga sibirica sp nov., two thermophilic bacteria isolated from a continental petroleum reservoir in Western Siberia." International Journal of Systematic and Evolutionary Microbiology 52: 1715-1722. Li, R. P., S. L. Chen, et al. (2009). "A naerobic Codigestion of Ki tchen Waste with Cattle Manure for Biogas Production." Energy & Fuels 23: 2225-2228. Lien, T., M. Madsen, et al. (1998). "Petro toga mobilis sp. nov., from a North Sea oilproduction well." International Jour nal of Systematic Bacteriology 48: 1007-1013. Liu, W., P. C. Pullammanappallil, et al. (2008). "Thermophilic ana erobic digestion of sugar beet tailings." Transactions of the Asabe 51(2): 615-621.

PAGE 179

179 Liu, Z., C. Lozupone, et al (2007). "Short pyrosequencing reads suffice for accurate microbial community analysis." Nucleic Acids Research 35(18): Article No.: e120. Liu.Y. and W. B. Whitman (2008). "Metabolic, phylogenetic, and ecological diversity of the methanogenic archaea." Annals of t he New York Academy of Sciences 1125 : 171-189. Lynd, L. R., P. J. Weimer, et al. (2002). "M icrobial cellulose utilization: Fundamentals and biotechnology (vol 66, pg 506, 2002)." Microbiology and Molecular Biology Reviews 66 (4): 739-739. Ma, F. R. and M. A. Hanna (1999). "Biodi esel production: a review." Bioresource Technology 70(1): 1-15. Mackie, R. I. and M. P. Bryant (1981). "Metabolic-activity of fatty acid-oxidizing bacteria and the contribution of acetate, propionat e, butyrate, and c o2 to methanogenesis in cattle waste at 40-degrees-c and 60degrees-c." Applied and Environmental Microbiology 41(6): 1363-1373. Madson, P. W. (2003). Ethanol distillation: the fundamentals. The Alcohol Textbook. K. A. Jacques, T. P. Lyons and D. R. Kelsall. Nottingham, Nottingham University Press: 319-327. Maestrojuan, G. M., D. R. Boone, et al. (1990). "Trans fer of methanogenium-bourgense, methanogenium-marisnigri, methanogen ium-olentangyi, and methanogeniumthermophilicum to the genus methanoculleus gen-nov, emendation of methanoculleus-marisnigri and methanogenium, and description of new strains of methanoculleus-bourgense and methanoculleus-marisnigri." International Journal of Systematic Bacteriology 40(2): 117-122. Malina, J. F., Jr. (1992). A narobic sludge digestion. Design of Anaerobic Processes for the Treatment of Industria l and Municipal Wastes. J. F. Malina, Jr. and F. G. Pohland. Boca Raton, CRC Press. 7: 170-192. Manes, C. L. D., N. West, et al. (2011). "Dynamic bacterial communities on reverseosmosis membranes in a full-scale desalination plant." Biofouling 27(1): 47-58. Manes, R. J., A. Fernandez, et al. (2001). "Ph ysiological and molecular characterisation of an anaerobic thermophilic oleate-degr ading enrichment culture." Anaerobe 7 (1): 17-24. McInerney, M. J., M. P. Bryant, et al. (1981). "Syntrophomonas-wolfei gen-nov sp-nov, an anaerobic, syntrophic, fatty-acid ox idizing bacterium." Applied and Environmental Microbiology 41(4): 1029-1039. McInerney, M. J., L. Rohlin, et al. (2007) "The genome of Syntrophus aciditrophicus: Life at the thermodynamic limit of microbial growth." Proceedings of the National Academy of Sciences of the United States of America 104 (18): 7600-7605.

PAGE 180

180 McMahon, K. D., P. G. Stroot, et al. (2001). "Anaerobic codi gestion of municipal solid waste and biosolids under various mixi ng conditions II: Microbial population dynamics." Water Research 35(7): 1817-1827. McMahon, K. D., D. D. Zheng, et al. (2004). "Microbial population dynamics during startup and overload conditions of anaerobic dige sters treating municipal solid waste and sewage sludge." Biotec hnology and Bioengineering 87(7): 823-834. Mechichi, T., M. Labat, et al. (1999). "Spor obacterium olearium gen, nov., sp, nov., a new methanethiol-producing bacterium that degrades aromatic compounds, isolated from an olive mill wa stewater treatment digester. International Journal of Systematic Bacteriology 49: 1741-1748. Meher, L. C., D. V. Sagar, et al. (2006). "Technical aspects of biodiesel production by transesterification a review." Renew able & Sustainabl e Energy Reviews 10(3): 248-268. Menes, R. J. and L. Muxi (2002). "Anaerobaculum mobile sp nov., a novel anaerobic, moderately thermophilic, pepti de-fermenting bacterium that uses crotonate as an electron acceptor, and emended descrip tion of the genus Anaerobaculum." International Journal of Systematic and Evolutionary Microbiology 52: 157-164. Meynell, P. J. (1978). Me thane : planning a digester. New York, Schocken Books. Miranda-Tello, E., M. L. Fardeai, et al (2007). "Petrotoga halophila sp nov., a thermophilic, moderately halophilic, fe rmentative bacterium isolated from an offshore oil well in Congo." International Journal of Systematic and Evolutionary Microbiology 57: 40-44. Mosoni, P., C. Martin, et al. (2011). Long-term defaunation increases the abundance of cellulolytic ruminococci and methanogens bu t does not affect the bacterial and methanogen diversity in the rumen of sheep." Journal of Animal Science 89(3): 783-791. Mller, N., P. Worm, et al. (2010). "Syn trophic butyrate and pr opionate oxidation processes: from genomes to reaction mechanisms." Environmental Microbiology Reports 2 (4): 489-499. Nagano, A., E. Arikawa, et al. (1992). "The treatment of liquor waste-water containing high-strength suspended-solids by membra ne bioreactor system." Water Science and Technology 26(3-4): 887-895. Narihiro, T., T. Terada, et al. (2009). "C omparative Analysis of Bacterial and Archaeal Communities in Methanogenic Sludge Granul es from Upflow Anaerobic Sludge Blanket Reactors Treating Various F ood-Processing, High-Strength Organic Wastewaters." Micr obes and Environments 24(2): 88-96.

PAGE 181

181 Nielsen, H. B., H. Uellendahl, et al. (2007). "Regulation and optimization of the biogas process: Propionate as a key parameter." Biomass & Bioenergy 31 (11-12): 820830. Niu, L. L., L. Song, et al. (2008). "Prote iniborus ethanoliaenes gen. nov., sp nov., an anaerobic protein-utilizing bacte rium." Internati onal Journal of Systematic and Evolutionary Microbiology 58: 12-16. Ntaikou, I., H. N. Gavala, et al. (2008). "Hydrogen product ion from sugars and sweet sorghum biomass using Ruminococcus albu s." International Journal of Hydrogen Energy 33(4): 1153-1163. Ntaikou, I., H. N. Gavala, et al. (2010). "Application of a modified Anaerobic Digestion Model 1 version for fermentative hy drogen production from sweet sorghum extract by Ruminococcus albus." Inte rnational Journal of Hydrogen Energy 35(8): 3423-3432. Ntaikou, I., E. Koutros, et al. (2009). "Valorisat ion of wastepaper using the fibrolytic/hydrogen producing bacterium Ruminococcus albus." Bioresource Technology 100 (23): 5928-5933. Ojeda, K., O. vila, et al. (2011) "Evaluation of technologica l alternatives for process integration of sugarcane bagasse for sust ainable biofuels production--Part 1." Chemical Engineering Research and Design 89(3): 270-279. Owen, W. F., D. C. Stucke y, et al. (1979). "Bioassay for monitoring biochemical methane potential and anaerobic toxicity." Water Research 13(6): 485-492. Owens, J. M. and D. P. Chy noweth (1993). "Biochemical me thane potential of municipal solid-waste (msw) components." Water Science and Technology 27(2): 1-14. Palatsi, J., M. Vias, et al. (2011). "Anaer obic digestion of slaughterhouse waste: Main process limitations and microbial co mmunity interactions." Bioresource Technology 102 (3): 2219-2227. Panella, L. (2010). "Sugar beet as an energy crop." Sugar Tech 12(3/4): 288-293. Parameswaran, P., R. Jalili, et al. ( 2007). "A pyrosequenci ng-tailored nucleotide barcode design unveils opportunities for large-scale sample multiplexing." Nucleic Acids Research 35(19). Parawira, W. (2009). "Biotech nological production of biodiesel fuel using biocatalysed transesterification: A review." Critical Reviews in Biotechnology 29 (2): 82-93. Parkin, G. F. and W. F. Owen (1986). "Fundamentals of Anaerobic Digestion of Wastewater Sludges." Journal of Environmental Engineering 112 (5): 867-920.

PAGE 182

182 Patel, G. B. and G. D. Sprott (1990) "Methanosaeta-concilii gen-nov, sp-nov (methanothrix-concilii) and methanosaeta-t hermoacetophila nom-rev, comb-nov." International Journal of Systematic Bacteriology 40(1): 79-82. Penteado, T. Z., R. S. S. S antana, et al. (2011). "Effect of agitation on the performance of an anaerobic sequencing bat ch biofilm reactor in the treatment of dairy effluents." Water Science and Technology 63(5): 995-1003. Pfeffer, J. T. (1987). "Evolution of the Re fCOM system." Biotechnological advances in processing municipal wastes fo r fuels and chemicals.: 331-341. Plugge, C. M., M. Balk, et al. (2002). "Desulfotomaculum thermobenzoicum subsp thermosyntrophicum subsp nov., a thermoph ilic, syntrophic, propionate-oxidizing, spore-forming bacterium." International Journal of Systematic and Evolutionary Microbiology 52: 391-399. Polematidis, I. M. (2007). T hermophilic, batch, high-solids bi ogasification of sugar beet tailings [electronic resource]. Gainesvi lle, Fla., University of Florida. Pullammanappallil, P., W. Cla rke, et al. (2005). High-so lids, leach-bed anaerobic digestion of organic fraction of munici pal solid waste. 4th International Symposium on Anaerobic Digestion of Solid Waste. Copenhagen, Denmark, International Waste Association. Reddy, C. A., M. J. Wolin, et al. (1972) "Ferredoxinamide and nicotinamide adenine dinucleotide-dependent h2 produc tion from ethanol and forma te in extracts of s organism isolated from methanobacillus-omelianskii." Journal of Bacteriology 110 (1): 126-&. Rees, G. N., B. K. C. Patel, et al. ( 1997). "Anaerobaculum thermoterrenum gen nov, sp nov, a novel, thermophilic bacterium which ferments citrate." In ternational Journal of Systematic Bacteriology 47(1): 150-154. Ren, N. Q., D. F. Xing, et al. (2005). "Microbial community dynamics during start-up of bio-hydrogen producing reactor." Future of Urban Wastewater Systems Decentralisation and Reuse: 161-168. Ren, N. Q., D. F. Xing, et al. (2007). "Microbial communi ty structure of ethanol type fermentation in bio-hydrogen producti on." Environmental Microbiology 9 (5): 1112-1125. Riviere, D., V. Desvignes, et al. (2009). "Towards the definition of a core of microorganisms involved in anaerobic di gestion of sludge." Isme Journal 3 (6): 700-714. Roberton, A. M. and R. S. Wolfe (1970). "Adenosine triphosphate pools in methanobacterium." Journal of Bacteriology 102 (1): 43-&.

PAGE 183

183 Romano, R. T. and R. H. Zhang (2008). "Co-digestion of onion juice and wastewater sludge using an anaerobic mixed biofilm reactor." Bioresource Technology 99(3): 631-637. Ronaghi, M., S. Karamohamed, et al. (1996). "Real-time DNA sequencing using detection of pyrophosphate release." Analytical Biochemistry 242 (1): 84-89. Ruiz, I., R. Blazquez, et al. (2009). "M ethanogenic toxicity in anaerobic digesters treating municipal wastewater ." Bioresource Technology 100 (1): 97-103. Russell, J. B., R. E. Muck, et al. (2009). "Quantitative analysis of cellulose degradation and growth of cellulolytic bacteria in the rumen." Fems Microbiology Ecology 67(2): 183-197. Rychlik, J. L. and T. May (2000). "The effect of a meth anogen, Methanobrevibacter smithii, on the growth rate, organic acid production, and specific ATP activity of three predominant ruminal cellulolyti c bacteria." Current Microbiology 40(3): 176180. Sabourin-Provost, G. and P. C. Hallenbeck ( 2009). "High yield conversion of a crude glycerol fraction from biodiesel produc tion to hydrogen by photofermentation." Bioresource Technology 100 (14): 3513-3517. Saengkerdsub, S., R. C. Ander son, et al. (2007). "Identific ation and quantification of methanogenic archaea in adult chicken ceca." Applied and Environmental Microbiology 73(1): 353-356. Sahm, H. (1984). "Anaerobic wastewater tr eatment." Adv Biochem Eng Biotechnol 29: 83-115. Saraphirom, P. and A. Reungs ang (2010). "Optimization of biohydrogen production from sweet sorghum syrup using statistical methods." International Journal of Hydrogen Energy 35(24): 13435-13444. Sasaki, D., T. Hori, et al (2011). "Methanogenic pathway and community structure in a thermophilic anaerobic digestion process of organic solid waste." Journal of Bioscience and Bi oengineering 111 (1): 41-46. Sassner, P., M. Galbe, et al. (2008). "Techno-economic evaluation of bioethanol production from three differ ent lignocellulosic materials." Biomass & Bioenergy 32(5): 422-430. Sawyer, C. N. and A. M. Grumbling (1960). "Fundamental c onsideration in high-rate digestion." Journal of the Sanitary Engineering Division 86(2): 49-64. Schaefer, S. H. and S. H. Sung (2008). "Ret ooling the ethanol industry: thermophilic anaerobic digestion of thin stillage for methane production and pollution prevention." Water Environment Research 80 (2): 101-108.

PAGE 184

184 Schloss, P. D., S. L. Westcott, et al (2009). "Introducing mothur: Open-Source, Platform-Independent, Community-Supported Software for Describing and Comparing Microbial Communities." A pplied and Environmental Microbiology 75(23). Schnurer, A., B. Schink, et al. (1996). "Clo stridium ultunense sp nov, a mesophilic bacterium oxidizing acetate in syntrophic association with a hydrogenotrophic methanogenic bacterium." International J ournal of Systematic Bacteriology 46 (4): 1145-1152. Shcherbakova, V., E. Rivkina, et al. ( 2011). "Methanobacterium arcticum sp. nov., a methanogenic archaeon from Holocene Arctic permafrost." International Journal of Systematic and Evol utionary Microbiology 61: 144-147. Sheehan, G. J. and P. F. Greenfield (1980). "UTILIZATION, TREATMENT AND DISPOSAL OF DISTILLERY WASTE-WATER." Water Research 14(3): 257-277. Sheehan, J., V. Camobreco, et al. (1998). Life Cycle Inventory of Biodiesel and Petroleum Diesel for Use in an Urban Bus. Golden, National Renewable Energy Laboratory : 168-187. Shelton, D. R. and J. M. Tiedje (1984). "General-met hod for determining anaerobic biodegradation potential." Applied and Environmental Microbiology 47(4): 850857. Shin, H. S., B. U. Bae, et al. (1992). "Anaerobic-digestion of distillery waste-water in a 2phase uasb system." Water Science and Technology 25(7): 361-371. Shin, S. G., S. Lee, et al (2010). "Qualitative and quantitat ive assessment of microbial community in batch anaerobic digesti on of secondary sludge." Bioresource Technology 101 (24): 9461-9470. Shoemaker, D. D., D. A. Lashkari, et al (1996). "Quantitative phenotypic analysis of yeast deletion mutants using a highly par allel molecular bar-coding strategy." Nature Genetics 14(4): 450-456. Sierraalvarez, R., J. A. Field, et al. (1994) "Overview of the anaerobic toxicity caused by organic forest industry waste-water pollutants." Water Science and Technology 29(5-6): 353-363. Siles, J. A., M. A. Martin, et al. (2010). "Anaerobic co-digestion of glycerol and wastewater derived from biodiesel m anufacturing." Bioresource Technology 101 (16): 6315-6321. Smith, L. C., D. J. Elliot, et al. (1996). "Mixing in upflow anaer obic filters and its influence on performance and scale-up." Water Research 30(12): 3061-3073.

PAGE 185

185 Solera, R., L. I. Romero, et al. (2001). "Det ermination of the Microbial Population in Thermophilic Anaerobic Reactor: Compar ative Analysis by Different Counting Methods." Anaerobe 7 (2): 79-86. Sousa, D. Z., H. Smidt, et al. (2007). "Syntrophomonas zehnderi sp nov., an anaerobe that degrades long-chain fatty acids in co-culture with Methanobacterium formicicum." International Journal of S ystematic and Evolutionary Microbiology 57: 609-615. Speece, R. E. (1983). "Anaerobic biotechnology for industrial wastewater-treatment." Environmental Science & Technology 17(9): A416-A427. Spiehs, M. J., M. H. Whitney, et al. (2002). "Nutrient database for distiller's dried grains with solubles produced from new ethanol plants in Minnesota and South Dakota." Journal of Animal Science 80(10): 2639-2645. Srivastava, V. J., R. Biljetina, et al. (1989). "Biogasification of community-derived biomass and solid-wastes in a pilot-scale solcon reactor." Applied Biochemistry and Biotechnology 20-1 : 587-602. Stams, A. J. M. and C. M. Plugge (2009). "Electron transfer in syntrophic communities of anaerobic bacteria and archaea." Nature Reviews Microbiology 7 (8): 568-577. Stroot, P. G., K. D. McMahon, et al. (2001) "Anaerobic codigestion of municipal solid waste and biosolids under various mixi ng conditions--I. digester performance." Water Research 35(7): 1804-1816. 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. Sugihara, T., H. Shiratori, et al. (2007). "Unique diversity content of a thermophilic anaerobic microbial consortium that utiliz es propionate in a synthetic medium." Journal of General an d Applied Microbiology 53(6): 363-369. Suryawanshi, P. C., A. B. Chaudhari, et al. (2010). "Thermophilic anaerobic digestion: the best option for waste treatment." Critical Reviews in Biotechnology 30(1): 3140. Takara, D., S. Nitayavardahana, et al. (2010) "Sustainable Bioenergy from Biofuel Residues and Wastes." Wa ter Environment Research 82 (10): 1694-1719. Tang, Y. Q., Y. Fujimura, et al. (2007). "Anaerobic treatment performance and microbial population of thermophilic upflow anaerobic filter r eactor treating awamori distillery wastewater." Journal of Bioscience and Bioengineering 104(4): 281-287. Teixeira, A. A., D. P. Chynoweth, et al. (2005). Space-based SEBAC-II solid waste management technology for commercial ap plication to beet sugar industry.

PAGE 186

186 International Conference on Environmental Systems. Rome, Italy, Society of Automobile Engineering. Thauer, R. K., K. Jungerm ann, et al. (1977). "Energy-c onservation in chemotropic anaerobic bacteria." Bacteriological Reviews 41(1): 100-180. Thompson, J. C. and B. B. He (2006). "Characterization of crude glycerol from biodiesel production from multiple feedstocks." A pplied Engineering in Agriculture 22(2): 261-265. Torry-Smith, M., P. Sommer, et al. (2003). "Purification of bioethanol effluent in an UASB reactor system with simultaneous biogas formation." Biotechnology and Bioengineering 84(1): 7-12. Valdez-Vazquez, I., E. Ros-Leal, et al. (2005). "Semi-continuo us solid substrate anaerobic reactors for H2 production from organic waste: Mesophilic versus thermophilic regime." Internati onal Journal of Hydrogen Energy 30(13-14): 13831391. Valentine, D. L., D. C. Blanton, et al. (2000). "Hyd rogen production by methanogens under low-hydrogen conditions." Archives of Microbiology 174 (6): 415-421. Van Ooteghem, S. A., A. Jones, et al. (2004) "H-2 production and carbon utilization by Thermotoga neapolitana under anaerobic and microaerobic growth conditions." Biotechnology Letters 26(15): 1223-1232. Vandamme, P., E. Falsen, et al. (1991). "Rev ision of campylobacter, helicobacter, and wolinella taxonomy emendation of generic descriptions and proposal of arcobacter gen-nov." International Jour nal of Systematic Bacteriology 41(1): 88103. Vavilin, V. A. and I. Angelidaki (2005) "Anaerobic degradation of solid material: Importance of initiation centers fo r methanogenesis, mixing intensity, and 2D distributed model." Biotec hnology and Bioengineering 89 (1): 113-122. Vedrenne, F., F. Beline, et al. (2008). "The effect of incubation conditions on the laboratory measurement of the methane producing capacity of livestock measurement wastes." Bioresource Technology 99: 146-155. Vries, S. C. d., G. W. J. v. d. Ven, et al. (2010). "Resource use efficiency and environmental performance of nine major biofuel crops, processed by firstgeneration conversion techniques ." Biomass and Bioenergy 34(5): 588-601. Wang, W., W. Ma, et al. (2011). "Therm ophilic anaerobic digestion of Lurgi coal gasification wastewater in a UASB reactor." Bioresource Technology 102 (3): 2441-2447.

PAGE 187

187 Weiland, P. (2003). "Production and energetic use of biogas from energy crops and wastes in Germany." Applied Biochemistry and Biotechnology 109 (1): 263-274. Weiland, P. and H. Thomsen (1990). "Operational behavior of an industrial fixed-bed reactor for biomethanation of alcohol slops from differe nt crops." Water Science and Technology 22(1-2): 385-394. Weiss, A., V. Jerome, et al. (2009). "Investigation of factors influencing biogas production in a large-scale thermophilic municipal biogas plant." Applied Microbiology and Biotechnology 84(5): 987-1001. Weiss, A., V. Jerome, et al. (2008). "Diversity of t he resident microbiota in a thermophilic municipal biogas plant." Applied Microbiology and Biotechnology 81(1): 163-173. Wei, S., A. Zankel, et al. (2011). "Investigat ion of mircroorganisms colonising activated zeolites during anaerobic biogas producti on from grass silage." Bioresource Technology 102 (6): 4353-4359. Whitmore, T. N., D. Lloyd, et al. (1987). "Hydrogen-dependent control of the continuous anaerobic-digestion process." Applie d Microbiology and Biotechnology 26(4): 383-388. Wilkie, A. C., K. J. Riedesel, et al. (2000). "Stillage characterization and anaerobic treatment of ethanol stillage from c onventional and cellulosic feedstocks." Biomass & Bioenergy 19(2): 63-102. Willems, A., J. Busse, et al. (1989) "Hydrogenophaga, a new genus of hydrogenoxidizing bacteria that includes hydrogenophaga-flava comb-nov (formerly pseudomonas-flava), hy drogenophaga-palleronii (f ormerly pseudomonaspalleronii), hydrogenophaga-pseudoflava (formerly pseudomonas-pseudoflava and pseudomonas-carboxydoflava), and hydrogenophaga-taeniospiralis (formerly pseudomonas-taeniospiralis)." In ternational Journal of Systematic Bacteriology 39(3): 319-333. Wingren, A., M. Galbe, et al. (2008). "Energy considerations for a SSF-based softwood ethanol plant." Bioresource Technology 99(7): 2121-2131. Wommack, K. E., J. Bhavsar, et al. (2008). "Metagenomics: R ead length matters." Applied and Environm ental Microbiology 74(5): 1453-1463. Xun, L. Y., D. R. Boone, et al. (1989). "Deoxyribonucleic-acid hybridization study of methanogenium and methanocorpusculum species, emendation of the genus methanocorpusculum, and transfer of methanogenium-aggregans to the genus methanocorpusculum as methanocorpusculum-aggregans comb-nov." International Journal of Systematic Bacteriology 39(2): 109-111.

PAGE 188

188 Yang, Y., K. Tsukahara, et al. (2008). "B iodegradation and methane production from glycerol-containing synthetic wastes with fixed-bed bioreactor under mesophilic and thermophilic anaerobic conditi ons." Process Biochemistry 43(4): 362-367. Yeoh, B. G. (1997). "Twophase anaerobic treatment of c ane-molasses alcohol stillage." Water Science and Technology 36(6-7): 441-448. Yomano, L. P., S. W. York et al. (1998). "Isolation a nd characterization of ethanoltolerant mutants of Escherichia coli KO11 for fuel ethanol production." Journal of Industrial Microbiology & Biotechnology 20(2): 132-138. Yomano, L. P., S. W. York et al. (2009). "Deletion of methylglyoxal synthase gene (mgsA) increased sugar co-metabolism in ethanol-producing Escherichia coli." Biotechnology Letters 31(9): 1389-1398. Yomano, L. P., S. W. York, et al. (2008). "Re-engineering E scherichia coli for ethanol production." Biotechnology Letters 30(12): 2097-2103. Yoochatchaval, W., A. Ohashi, et al. (2008). "Characteristics of granular sludge in an EGSB reactor for treating lo w strength wastewater." In ternational Journal of Environmental Research 2 (4): 319-328. Yu, L. A., J. W. Ma, et al (2011). "Numerical simulation of mechanical mixing in high solid anaerobic digester." Bioresource Technology 102 (2): 1012-1018. Zhang, H., J. E. Banaszak, et al. ( 2009). "Focused-Pulsed sludge pre-treatment increases the bacterial diversity and relative abundance of acetoclastic methanogens in a full-scale anaer obic digester." Water Research 43 (18): 45174526. Zhou, Z., Q. Meng, et al (2011). "Effects of Methanoge nic Inhibitors on Methane Production and Abundances of Methanogens and Cellulolytic Bacteria in In Vitro Ruminal Cultures." Appl. Environ. Microbiol. 77(8): 2634-2639. Zhu, H., W. Parker, et al. (2009). "B uffer requirements for enhanced hydrogen production in acidogenic digestion of food wastes." Bioresource Technology 100 (21): 5097-5102. Ziganshin, A. M., T. Schmidt, et al. (2011) "Bacteria and archaea involved in anaerobic digestion of distillers grains with solubles." Applied Microbiology and Biotechnology 89(6).

PAGE 189

189 BIOGRAPHICAL SKETCH Zhuoli Tian was born in Guizhou, China. He received his bachelors degree in microbiology from Central South University, China in 2006. Thereafter, he was enrolled in the Graduate School at University of Florida as a masters student. He worked as a research assistant at Biopr ocess Laboratory under Dr. Pull ammanappallil. He moved on to the Ph.D program in August 2008. After graduat ion, he plans to work in the field of biofuel and environmental engineering.