Effect of Mechanical Pretreatment on Solubilization and Biomethanation of Food Waste

MISSING IMAGE

Material Information

Title:
Effect of Mechanical Pretreatment on Solubilization and Biomethanation of Food Waste
Physical Description:
1 online resource (105 p.)
Language:
english
Creator:
Graunke,Ryan E
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Interdisciplinary Ecology
Committee Chair:
Wilkie, Ann C
Committee Members:
Hochmuth, George J
Moore, Kimberly A

Subjects

Subjects / Keywords:
anaerobic -- biodigestion -- bioenergy -- biofuel -- biogas -- biomethanation -- digestion -- enzymes -- food -- hydrolysis -- mechanical -- methanogenesis -- pretreatment -- solubilization -- sustainability -- waste
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre:
Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The current disposal of food waste in landfills is an unsustainable practice that causes many environmental, social, and economic problems. Diverting food waste from landfills for anaerobic digestion presents the opportunity for three positive outcomes: sustainable waste utilization, production of renewable bioenergy, and recovery of biofertilizer. Currently, food waste digestion does not exist at the commercial scale outside of co-digestion with manure or municipal sludge. The anaerobic digestion of food waste is considered to be limited by the rate of hydrolysis, which impedes the use of food waste as a feedstock in high-rate digestion. Food waste pretreatment has been examined to increase the hydrolytic rate of food waste. Current literature reports pretreatment methods that are not economically feasible for commercial-scale digestion. The present study examined the use of practical mechanical pretreatment (food disposer and meat grinder) for increasing the solubilization and biomethanation of food waste. Three solubilization assays were conducted to compare the solubilization kinetics of intact food waste to those of pretreated food waste. The three assays measured 1) the release of endogenous soluble organic material, 2) enzymatic hydrolysis with excess commercial hydrolytic enzymes, and 3) enzymatic hydrolysis using a microbial inoculum. In all three assays, pretreated food waste exhibited significantly higher solubilization kinetics than intact food waste. Pretreatment of food waste released the full complement of soluble organic material (25% solubilization) immediately, while intact food waste was 4.1% solubilized initially and 21% solubilized at 8 h through the leaching of endogenous soluble organics at a 1st order rate (k) of 0.36 h-1. With excess commercial hydrolytic enzymes, pretreated and intact food waste was 60% and 49%, respectively, solubilized in 8h with enzymatic hydrolysis rates of 0.84 to 1.19 h-1 and 0.25 h-1, respectively. When incubated with a microbial inoculum, pretreated and intact food waste was 52% and 37% solubilized in 24 h, which indicated that, for pretreated food waste, a microbial inoculum is nearly as effective in 24 h as excess commercial enzymes in 8h. In an active anaerobic digester, the microbial hydrolysis rate would be even greater than in the assay. The results showed that mechanical pretreatment increased the solubilization kinetics of food waste, which increased the available substrate for biomethanation. Two biochemical methane potential assays were performed on intact and pretreated (meat grinder with 0.5 cm plate openings) food waste. The first assay was loaded at 3.5 g COD/L, while the second assay was loaded at 2 g COD/L. The first assay at the moderate loading rate showed slower methanogenic kinetics (k=0.12 d-1) than the second assay at the reduced loading rate (k=0.20 d-1). Pretreated food waste at the moderate loading rate showed a lower cumulative methane yield than the intact food wastes in either assay (282 and 320 mL/g COD, respectively). These results were due to acidification at the moderate loading rate, which was exacerbated through pretreatment. The experiments show that in the BMP assay, methanogenesis and not hydrolysis was the rate-limiting step, and at moderate loading rates, increased solubilization can inhibit methanogenesis. Therefore, using a high-rate digester, such as a fixed-film reactor with retained methanogenic biomass, is necessary to accommodate increased solubilization and acidogenesis of pretreated food waste.
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 Ryan E Graunke.
Thesis:
Thesis (M.S.)--University of Florida, 2011.
Local:
Adviser: Wilkie, Ann C.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-02-29

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043048:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 EFFECT OF MECHANICAL PRETREATMENT ON SOLUBILIZATION AND BIOMETHANATION OF FOOD WASTE By RYAN E. GRAUNKE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Ryan E. Graunke

PAGE 3

3 To my parents and Kim thank you for all your support and encouragement

PAGE 4

4 ACKNOWLEDGMENTS First and foremost, I would like to thank my advisor, Dr. Ann C. Wilkie in the Soil and Water Science Department. Without her dedicated guidance, this thesis would not have been possible. When I was an undergraduate, she inspired me to pursue my interest in sustainability and food waste and opened my eyes to the world of anaerobic digestion. She has guided me from an undergraduate with little practical, laboratory experien ce to a fully lab competent graduate researcher. Our shared passion for food waste reduction, anaerobic digestion, and sustainability has led us down many interesting and rewarding paths and will hopefully lead to an even brighter, sustainable future. I w ould also like to thank my thesis committee members, Dr. George Hochmuth and Dr. Kimberly Moore for their many helpful comments during my research. I would like to thank my lab mate, Scott Edmundson for his assistance and support during the research proce ss as well as for his microscopy expertise. I also thank Camilo Cornejo for his help with various experiments and John Owens for advice on data analysis. Finally, I would like to thank my parents, Robert and Barbara Graunke, and my girlfriend, Kimberly G orski, for personal support and encouragement during my graduate experience. This research was supported by funding from the Hinkley Center for Solid and Hazardous Waste Management.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ....................... 10 ABSTRACT ................................ ................................ ................................ ................................ ... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 14 Anaerobic Digestion ................................ ................................ ................................ ............... 17 Anaerobic Digesters and Operating Conditions ................................ .............................. 18 Anaerobic Microbial Consortia ................................ ................................ ....................... 20 Hydrolysis ................................ ................................ ................................ ................ 20 Acidogenesis ................................ ................................ ................................ ............ 21 Acetogenesis ................................ ................................ ................................ ............. 21 Methanogenesis ................................ ................................ ................................ ........ 21 Hydrolysis and Extracellular Enzymes ................................ ................................ ................... 22 Cellulases ................................ ................................ ................................ ......................... 24 Amylases ................................ ................................ ................................ ......................... 24 Lipases ................................ ................................ ................................ ............................. 25 Proteases ................................ ................................ ................................ .......................... 25 Factors Influencing Fo od Waste Hydrolysis ................................ ................................ .......... 26 Pretreatment of Food Waste ................................ ................................ ................................ ... 27 Thermal and Freezing/Thawing Pretreatment ................................ ................................ 29 Enzymatic Pretreatment ................................ ................................ ................................ ... 31 Mechanical Pretreatment ................................ ................................ ................................ 32 Thesis Rationale ................................ ................................ ................................ ...................... 35 Hypothesis ................................ ................................ ................................ .............................. 36 Objectives ................................ ................................ ................................ ............................... 36 2 MATERIALS AND METHODS ................................ ................................ ........................... 37 Standard Food Waste ................................ ................................ ................................ .............. 37 Composition ................................ ................................ ................................ .................... 37 Analysis ................................ ................................ ................................ ........................... 38 Food Waste Microscopy ................................ ................................ ................................ ......... 38 Solubilization Assay ................................ ................................ ................................ ............... 39 Mechanical Pretreatment ................................ ................................ ................................ 40 Commerci al Enzyme ................................ ................................ ................................ ....... 40 Microbial Inoculum ................................ ................................ ................................ ......... 41 Sampling Technique ................................ ................................ ................................ ........ 42

PAGE 6

6 Statistical An alysis ................................ ................................ ................................ .......... 42 Biochemical Methane Potential Assay ................................ ................................ ................... 42 Simulated BMP Assay for pH Measurements ................................ ................................ 43 Methanogenic Inoculum ................................ ................................ ................................ .. 44 Methane Production Measurement ................................ ................................ .................. 44 Statistical Analysis ................................ ................................ ................................ .......... 46 Physiochemic al Parameters ................................ ................................ ................................ .... 46 Total Chemical Oxygen Demand ................................ ................................ .................... 46 Soluble Chemical Oxygen Demand ................................ ................................ ................ 47 To tal Solids and Volatile Solids ................................ ................................ ...................... 48 pH ................................ ................................ ................................ ................................ .... 48 Total Nitrogen and Total Phosphorus ................................ ................................ .............. 49 Alkalinity ................................ ................................ ................................ ......................... 49 Conductivity ................................ ................................ ................................ .................... 50 Organic Acid and Sugar Analysis ................................ ................................ ................... 50 3 EFFECT OF MECHANICAL PRETREATMENT ON SOLUBILIZATION OF FOOD WASTE ................................ ................................ ................................ ................................ ... 51 Microscopy of Pretreated Food Waste ................................ ................................ ................... 52 Endogenous Solubilization Assay ................................ ................................ .......................... 57 Twenty four Hour Solubilization ................................ ................................ .................... 57 Endogenous Solubilization Kinetics ................................ ................................ ................ 60 Estimates of parameters ................................ ................................ ........................... 60 Fitted curves. ................................ ................................ ................................ ............ 61 Commercial Enzyme Assay ................................ ................................ ................................ .... 62 Twenty four Hour Solubilization ................................ ................................ .................... 62 Enzymatic Hydrolysis Kinetics ................................ ................................ ....................... 67 Estimates of parameters ................................ ................................ ........................... 67 Fitted curves ................................ ................................ ................................ ............. 67 Microbial Inoculum Assay ................................ ................................ ................................ ..... 68 Twenty four Hour Solubilization ................................ ................................ .................... 69 Enzymatic Hydrolysis Kinetics ................................ ................................ ....................... 71 Es timates of parameters ................................ ................................ ........................... 72 Fitted curves ................................ ................................ ................................ ............. 73 Discussion ................................ ................................ ................................ ............................... 74 Summary and Conclusion s ................................ ................................ ................................ ..... 79 4 EFFECT OF MECHANICAL PRETREATMENT ON BIOMETHANATION OF FOOD WASTE ................................ ................................ ................................ ....................... 81 Moderate Loading Rate Biochemical Methane Potential Assay ................................ ........... 82 Cumulative Methane Production ................................ ................................ ..................... 83 Loading Rate and pH ................................ ................................ ................................ .............. 86 Reduced Loading Rate Biochemical Methane Potential Assay ................................ ............. 87 Cumulative Methane Production ................................ ................................ ..................... 87 Discussion ................................ ................................ ................................ ............................... 90

PAGE 7

7 Summary and Conclusions ................................ ................................ ................................ ..... 95 5 CONCLU SIONS ................................ ................................ ................................ .................... 97 Solubilization Kinetics ................................ ................................ ................................ ............ 98 Biomethanation Kinetics ................................ ................................ ................................ ........ 99 Practical Implication of Research ................................ ................................ ......................... 100 LIST OF REFERENCES ................................ ................................ ................................ ............. 101 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 105

PAGE 8

8 LIST OF TABLES Table page 1 1 Enzyme classes and target substrates found in food waste ................................ ................ 23 2 1 Range of physiochemical properties of food waste ................................ ........................... 37 2 2 Composition of standard food waste ................................ ................................ .................. 38 2 3 Physiochemical properties of standard food waste ................................ ............................ 38 2 4 Treatment regimes for solubilization assays ................................ ................................ ...... 39 2 5 Commercial enzyme cocktail formulation ................................ ................................ ......... 41 2 6 Physiochemical properties of microbial inoculum used in solubilization assay ................ 41 2 7 Biochemical methane potential assay regime ................................ ................................ .... 43 2 8 Physiochemical properties of inocula for the biochemical methane potential assays ....... 4 4 3 1 Estimated parameters and 8 h endogenous solubilization in endogenous solubilization assay ................................ ................................ ................................ ............ 60 3 2 Estimated parameters and 8 h hydrolyzed COD for commercial enzyme assay ............... 67 3 3 Estimated parameters and 6 h hydrolyzed COD for initial hydrolysis in the microbial inoculum assay ................................ ................................ ................................ ................... 72 3 4 Estimated parameters and 24 h hydrolysis for secondary hydrolysis in the microbial inoculum assay ................................ ................................ ................................ ................... 73 3 5 Maximum extent of solubilization for intact and pretreated (grinder 0.5 cm plate) food waste in the three solubilization assays ................................ ................................ ..... 74 3 6 Initial solubilization of food waste in pretreatment studies ................................ ............... 78 3 7 One day solubilization of food waste in pretreatment studies. ................................ .......... 78 4 1 Estimates of parameters for methane production kinetics in the moderate loading rate BMP assay. ................................ ................................ ................................ ................. 85 4 2 Calculated COD removal at 5, 10, 20 and 30 days in the moderate loading rate BMP assay. ................................ ................................ ................................ ........................ 85 4 3 Total COD, SCOD, pH, and conductivity after 30 days of digestion in the moderate loading rate BMP assay. ................................ ................................ .................... 85

PAGE 9

9 4 4 Estimates of parameters for methane production kinetics in the reduced loading rate BMP assay. ................................ ................................ ................................ ........................ 89 4 5 Calculated COD removal at 5, 10, 20 and 30 days in the reduced loading rate BMP assay. ................................ ................................ ................................ ........................ 89 4 6 Total COD, soluble COD, pH, conductivity, and alkalinity after 30 days of digestion in the reduced loading rate BMP assay. ................................ ................................ ............ 90 4 7 Ten day cumulative methane production of food waste in pretreatment studies. .............. 94

PAGE 10

10 LIST OF FIGURES Figure page 1 1 Food waste recovery hierarchy. ................................ ................................ ......................... 16 1 2 Closed loop cycle created by the anaerobic digestion of food waste. ............................... 17 1 3 Sequential metabolic phases in anaerobic digestion. ................................ ......................... 20 1 4 The hydrolysis of organic macromolecules into their constituent monomers ................... 23 1 5 The variable composition of food waste.. ................................ ................................ .......... 23 1 6 Food waste composition by degradability. ................................ ................................ ........ 28 2 1 Apparatus for measuring methane production through hydraulic displacement. .............. 45 3 1 Acidification of food waste under unbuffered, undiluted conditions ................................ 52 3 2 Photomicrographs of apple. ................................ ................................ ............................... 53 3 3 Photomicrographs of bean. ................................ ................................ ................................ 54 3 4 Photomicrographs of broccoli. ................................ ................................ ........................... 55 3 5 Photomicrographs of potato. ................................ ................................ .............................. 56 3 6 Soluble chemical oxygen demand in the endogenous solubilization assay. ...................... 58 3 7 Mean pH for the en dogenous solubilization assay ................................ ........................... 58 3 8 Sugars as a percent of SCOD for pretreated food waste (grinder 0.5 cm plate) in the endogenous solubilization assay ................................ ................................ ........................ 59 3 9 Sugars as a percent of SCOD for intact food waste in the endogenous solubilization assay ................................ ................................ ................................ ................................ ... 59 3 10 Mean solubilization of food waste in the en dogenous solubilization assay ..................... 60 3 11 Fitted endogenous solubilization curves for the endogenous solubilization assay ............ 61 3 12 Soluble chemical oxygen demand in the commercial enzyme assay. ............................... 63 3 13 Mean pH in the commercial enzyme assay ................................ ................................ ........ 64 3 14 Sugars as a percent of SCOD for pretreated food waste (grinder 0.5 cm plate) in the commercial enzyme solubilization assay. ................................ ................................ .......... 64

PAGE 11

11 3 15 Sugars as a percent of SCOD for intact food waste in the commerc ial enzyme solubilization assay ................................ ................................ ................................ ........... 65 3 16 Organic acids and ethanol as a percent of SCOD for pretreated food waste (grinder 0.5 cm plate) in the commerc ial enzyme solubilization assay ................................ .......... 65 3 17 Organic acids and ethanol as a percent of SCOD for intact food waste in the commercia l enzyme solubilization assay ................................ ................................ ........... 66 3 18 Mean solubilization of food waste in the commercial enzyme assay ............................... 66 3 19 Fitted enzymatic hydrolysis curves for the commercial enzyme assay ............................. 68 3 20 Soluble chemical oxygen demand of food waste in the microbial inoculum assay.. ......... 70 3 21 Mean pH f or the microbial inoculum assay ................................ ................................ ...... 70 3 22 Mean solubilization in microbial inoculum assay. ................................ ............................ 71 3 23 Mean hydrolysis in the microbial inoculum assay ................................ ............................ 72 3 24 Fitted solubilization curves for the microbial inoculum assay ................................ .......... 73 3 25 Six h solubilization of intact and pretreated (grinder 0.5 cm plate) food waste in the three solubilization assays. ................................ ................................ ................................ 74 4 1 Cumulative methane production of intact and pretreated food waste in the moderate loading rate BMP assay. ................................ ................................ .................... 84 4 2 Cumulative methane production from glucose and cellulose controls in the moderate loading rate BMP assay. ................................ ................................ .................... 84 4 3 Mean pH in simulated BMP assays loaded at 2, 4, and 8 g COD/L with intact and pretreated food waste. ................................ ................................ ................................ ........ 86 4 4 Cumulative methane production of intact and pretreated food waste in the reduced loading rate BMP assay ................................ ................................ ....................... 88 4 5 Cumulative methane production of glucose and cellulose controls in the reduced loading rate BMP assay. ................................ ................................ ...................... 89 4 6 Cumulative methane production of pretreated and intact food waste in both BMP assays. ................................ ................................ ................................ ....................... 91

PAGE 12

12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science EFFECT OF MECHANICAL PRETREATMENT ON SOLUBILIZATION AND BIOMETHANATION OF FOOD WASTE By Ryan E. Graunke August 2011 Chair: Ann C. Wilkie Major: Interdisciplinary Ecology The current disposal of food waste in landfills is an unsustainable practice that causes many environmental, social, and economic problems. Diverting food waste from landfills for anaerobic digestion presents the opportunity for three positive outcomes: s ustainable waste utilization production of renewable bioenergy, and recovery of biofertilizer Currently food waste digestion does not exist at the commercial scale outside of co digestion with manure or municipal sludge The anaerobic digestion of foo d waste is considered to be limited by the rate of hydrolysis, which impede s the use of food waste as a feedstock in high rate digestion. F ood waste pretreatment has been examined to increas e the hydrolytic rate of food waste. C urrent literature reports pretreatment methods that are not economically feasible for commercial scale digestion. T he present study examine d the use of practical mechanical pretreatment ( food disposer and meat grinder ) for increasin g the solubilization and biomethanation of food waste. Three solubilization assays were conducted to compare the solubilization kinetics of intact food waste to those of pretreated food waste The three assays measured 1) the release of endogenous soluble organic material 2) e nzymatic hydrolysis with excess commercial hydrolytic enzymes, and 3) enzymatic hydrolysis using a microbial inoculum. In all three assays pretreated food waste exhibited significantly higher solubilization kinetics than intact

PAGE 13

13 food waste. Pretreatment of food waste released th e full complement of soluble organic material (25% solubilization) immediately while intact food waste was 4.1% solubilized initially and 21% solubilized at 8 h through the leaching of endogenous soluble organics at a 1 st order rate (k) of 0.36 h 1 With excess commercial hydrolytic enzymes, pretreated and intact food waste was 60% and 49%, respectively solubiliz ed in 8h with enzymatic hydrolysis rates of 0.84 to 1.19 h 1 and 0.25 h 1 respectively. When incubated with a microbial inoculum, pretreated and intact food waste was 5 2 % and 37% solubilized in 24 h, which indicate d that, for pretreated food waste, a microbial inoculum is nearly as effective in 24 h as excess commercial enzymes in 8h. In an active anaerobic digester the microbial hydrolysis rate would be even greater than in the assay. The results show ed that mechanical pretreatment increase d the solubilization kinetics of food waste which increase d the available substrate for biomethanation Two b iochemical methane p otential assays were pe rformed on intact and pretreated ( meat grinder with 0.5 c m plate openings ) food waste. The first assay was loaded at 3.5 g COD/L, while the second assay was loaded at 2 g COD/L The first assay at the moderate loading rate showed slower methanogenic kine tics (k=0. 12 d 1 ) than the second assay at the reduced loading rate (k=0. 20 d 1 ). Pretreated food waste at the moderate loading rate showed a lower cumu lative methane yield than the intact food wastes in either assay (282 and 320 mL/g COD, respectively). These results were due to acidification at the moderate loading rate, which was exacerbated through pretreatment. The experiments show that in the BMP assay, methanogenesis and not hydrolysis was the rate limiting st ep and at moderate loading rate s increased solubilization can inhibit methanogenesis. Therefore using a high rate digester, such as a fixed film reactor with retained m ethanogenic biomass, is necessary to accommodate increased solubilization and acidog enesis of pretreated food waste.

PAGE 14

14 CHAPTER 1 INTRODUCTION The current disposal of food waste is a global problem that is garnering increased public attention. Food waste represents a significant proportion of our municipal solid waste (MSW). The United States Environmental Protection Agency (U S EPA) estimate s that in 2008, the U S generated 32 million tons of food waste ( U S EPA 20 11 ) or 12.7% of total U S MSW generation. Florida alone generated 1.8 million tons in 2008 ( FDEP 20 11 ), which accounts for only 3% of this food waste was recycled, while in Florida only 1% was recycled; the remaining 99% of food waste was predominantly sent to landfills. A substantial amount of food waste is also sent to sewage treatment plants via food waste disposers in homes, restaurants and grocery stores. Food waste, which has a high organic content, can significantly increase the energy required to aerobically treat the waste at a wastewater treatment plant The amount of food waste sent to the sewer is not accounted for in Florida but this food waste is in addition to the 1.8 million tons sent to landfills There is enormous potential in Florida to capture unrecycled food waste and utilize it for beneficial reuse. Florida currently has a goal to recycle 75% of MSW by 2020 ( FDEP 2 01 1 ). In order to meet this goal Diverting food waste from landfills has already been made mandatory in some areas. In 1993, Germany banned the landfilling of solid wast e with a total organic carbon content of more than 3% ( EEA 2009 ). In 2009, the City of San Francisco passed an ordinance ( San Francisco Environment Code, Chapter 19 ) requiring all households and businesses to collect food waste and organics separately for recycling at composting facilities. The United Kingdom is currently considering banning food waste disposal in landfills.

PAGE 15

15 Disposal of food waste in landfills has many negative consequences. One of the most serious problems is the emission of harmful g reenhouse gases from food waste decomposition in the landfill. When food waste is placed in the anaerobic conditions of a landfill, methane is generated as a byproduct of bac terial degradation. Methane, a potent greenhouse gas, has a 20 year global warmi ng potential 74 times that of carbon dioxide (IPCC 2007). Methane from landfills known as landfill gas, has historically been emitted to the atmosphere via evolution through the uncapped landfi ll surface. Current practices include capping the landfill w ith a liner and collecting the landfill gas for flaring or, in some cases, combustion for electrical generation. Despite these practices, landfills are the third largest anthropogenic source of U.S. methane emissions O f the methane that is generated wit hin landfills 44% is emitted to the atmosphere (U S EPA 2011). Food waste is rapidly degradable and can produce methane prior to the landfill capping while the land fill is still being filled, which results in food waste contributing largely to the methane emissions from landfills Diverting food waste from landfills would significantly reduce landfill methane emissions. Another major problem resulting from the disposal of food waste in landfills is the generation of landfill leachate. Landfill lea chate contains many harmful pollutants from various wastes in the landfills. One of the most critical pollutants is ammonium ( Kjeldsen et al., 2002 ) A mmonium results from the decomposition of proteinaceous materials Food waste is therefore a significant source of ammoni um in leachate. Leachate management is an enormous burden for landfill operators. Groundwater must be continually monitored and leachate must be treated (either on site or by transportation to a wastewater treatmen t plant) Diverting food waste from landfills can significantly reduce the burden of treating ammoni um in leachate.

PAGE 16

16 Wasting food, regardless of its disposition, can have inherent negative consequences. Food that is not eaten still requires the same amou nt of water, energy, labor, fertilizer, pesticide, and land to grow as food that is consumed Energy is needed to transport, process, and cook the food, which contributes to the embodied energy lost when food is wasted ( Cuellar & Webber, 2010 ) The collection of food waste results in odor and vermin problems, which are particularly proble matic in urban areas. Wasting food also carries with it many ethical dilemma s, such as the issue of global hunger and malnourishment. While there will always be food waste, reducing food waste and developing beneficial reuse can greatly reduce the problem of food waste. Figure 1 1. Food waste recovery hierarchy (U S EPA 2006) Note: Industrial uses include anaerobic digestion. There are many options for diverting food waste from landfills and creating beneficial reuse, including source reduction donation to food banks, use as animal feed, and composting. The U S EPA has developed a food waste recovery hierarchy outlining various recovery and disposal options for food waste (Figure 1 1). Food waste can be used as a source of sustainable bioene rgy and biofertilizer. The process of anaerobic digestion can convert food waste into methane and soluble nutrients through microbial degradation. Anaerobic digestion has typical ly been utilized in the U S for animal manure and municipal sludge. Food w aste, however, makes

PAGE 17

17 an excellent feedstock due to its high organic content, and the diversion of food waste for anaerobic digestion re duce s the negative environmental, social, and economic consequences of food waste while creatin g two beneficial end produ cts ( Graunke & Wilkie, 2008 ) Anaerobic Digestion Anaerobic digestion is a microbial process that utilizes anaerobic fermentation and anaerobic respiration to metabolize organic materials into methane and carbon dioxide, which collectively are called biogas. During the process, nutrients within the organic material are released into soluble forms and contained within th e digester effluent. The nutrient rich effluent (k nown as biofertilizer) can be used as a replacement for fossil fuel based synthetic fertilizers. Reusing the nutrients from food waste prevents those nutrients from entering landfills where they become po llutants in leachate and possibly groundwater Instead, biofertilizers from food waste promote organic and sustainable agriculture by using nutrients captured fro m waste. Anaerobic digestion of food waste creates a closed loop, sustainable cycle for the beneficia l reuse of food waste (Figure 1 2) Figure 1 2. Closed loop cycle created by the anaerobic digestion of food waste.

PAGE 18

18 Anaerobic Digesters and Operating Conditions Anaerobic diges ter is a term used to describe a wide variety of reactors that contain the microbial ecology necessary for anaerobic digestion to occur. Some common types of anaerobic digesters include batch reactors, continuously stirred tank reactor s (CSTR), up flow anaerobic sludge blanket (UASB) reactors plug flow reactors cov ered lagoon s two phase digester s and fixed film digester s The performance and operating conditions depend upon the digester type and the feedstock used in the process Important operational parameters for anaerobic digesters are temperature hydraulic retention time (HRT), pH, organic loading rate (OLR) methane production rate, and methane yield. Anaerobic digesters can be general ly classified by operating temperature as mesophilic (25 45C) or thermophilic (45 60 C) digesters. The microbial ecology that exists in a digester is dependent on the temperature at which it is operated. For example, mesophi lic digesters contain mesophilic microbial consortia which are comprised of different organisms than those adapted to thermophilic conditions In gene ral thermophilic digesters have faster kinetics and methane production rates than mesophilic digester s due to increased microbial metabolic rates Digesters are also classified by HRT. Hydraulic retention time is the amount of time the feed stock is held within a digester for the desired extent of degradation and methane yield The HRT is typically in units of days and is calculated by dividing the total digester volume by the volume fed per day. Two critical factors that contribute to the HRT of a diges ter are the feedstock and microbial population. Less degradable feedstock s such as those with high particulate matter require more time in a digester due to the time required to solubilize the material. Continuously stirred tank reactor digesters for example, generally require longer HRTs due to low microbial populations H igh rate digesters, such as fixed film reactors which contain media for attached growth have low HRTs due to increased microbial populations.

PAGE 19

19 One of the most important parameters controlling digestion is pH. The methanogens that are responsible for methane production in the digester have a specific optimum pH range ( 6.5 8.2 ) in which they can thrive At a pH below this range, the activity of methanogens is sharply reduced ( Speece, 2008 ) This reduces the conversion of organic acids to methane, an d the digester become s acidified or stuck A drop in pH can occur through overloading the system beyond the maximum OLR. The OLR is the rate at which the feedstock is loaded in the digester which is based upon the rate at which the microbial consortia c an metabolize the feedstock to methane In anaerobic digestion, bacteria ferment organic matter into organic acids prior to methanogenesis, and the rate of acid production typically exceeds the rate of methane production. By exceedin g the maximum OLR of a digester, increased acidogenesis causes accumulation of organic acids, which can decrease the pH and inhibit methanogenesis. This situation is a positive feedstock loop in which methanogenic inhibition further reduce s organic acid conversion to methane and cause further acidification. Acidification is particularly problematic for food waste digestion due to the high org anic concentration of food waste Th e tendency of food waste to acidify is one reason why food waste digestion does not exist on the commercial scale with food waste as the sole feedstock. Existing facilities co digest food waste with manure or municipal sludge Manure and sludge have a hig her alkalinity than food waste and act as a pH buffer for the high organic acid concentrations from food waste While the process may be acceptable in some situation s co d igestion with manure or sludge is not always an option for food waste digestion. I n urban areas where food waste is prevalent, these materials might not always be available, and there is limited space for large co digestion facilities The addition of manure or sludge to food waste digesters presents oth er issues, such as pathogens, a ntibiotics and heavy metals in the feedstock and

PAGE 20

20 resulting biofertilizer. Optimizing the microbial kinetics of food waste digestion is imperative for implementing commercial scale food waste dige stion without the use of manure or municipal sludge. Anaero bic Microbial Consortia The operating condition s of a digester are dependent on the microbial consortia in the digester. Anaerobic digestion operates through the sequential metabolic process es of mixed microbial consortia and is generally divided into fou r pha s es each with a unique microbial consortium : hydrolysis, acidogenesis, acetogenesi s, and methanogenesis (Figure 1 3 ). Figure 1 3 S equential metabolic phases in anaerobic digestion. Hydrolysis In order for microbial assimilation to occur, large organic polymers (i.e. carbohydrates, proteins, and lipids) must first be hydrolyzed into their constituent monomeric compounds (i.e. glucose, amino acids, and fatty acids). The hydrolyzed molecules are released into the soluble phase in the digester wher e they are assimilated by the microbial biomass. In an anaerobic digester, hydrolysis is primari ly facilitated by extracellular enzymes (including both cell free and

PAGE 21

21 cell bound enzymes) produced by the hydrolytic consortium. Several genera of bacteria h ave been identified for the production of h ydrolytic enzymes in a digester; some of the more common genera include Clostridium Eubacteria and Pseudomonas ( Lynd et al., 2002 ; McInerney, 1988 ) Anaerobic fungi present in the digester can also produce hydrolytic enzymes. In many cases the hydrolytic bacteria a re also acidogenic bacteria and use the hydrolyzed monomers in their own metabolism. Acidogenesis Acidogenic bacteria consume the monomeric products of hydrolysis and, through fermentation, produce long c hain and volatile fatty acids (VFAs) Acid fermenta tion is a complex phase in which a number of different acidogens produce several different organic acids. Other acidogens can consume these acids and ferment them into different organic acids ( Kim & Gadd, 2008 ) Acidogenesis is generally the fastest phase in the digestion process and in an unbalanced digester, increased acidogenesis can cause acidification. Acetogenesis There are two primary types of acetogenic bacteria: obligate hydrogen producing acetogens (OHPA) and homoacetogens ( Kim & Gadd, 2008 ) The OPHA are the more dominant group of acetogens in anaerobic digestion and consume the fatty acid intermediates from acidogenesis to produce acetate through fermentation The fermentation process also produces CO 2 and H 2 If the concentration of H 2 is too great, OPHA are inh ibited from producing acetate. Homoacetogens produce acetate by consuming CO 2 and H 2 which helps to rebalance the metabolism for the OPHA Methan ogenesis Methanogenesis occurs through anaerobic respiration by using electron s from acetate and hydrogen produced in acetogenesis to reduce CO 2 to methane ( Wilkie, 2008 ) Specialized

PAGE 22

22 archaebacteria called methanogens comprise the methanogenic consortium in anaerobic digestion. There are two primary types of methanogens: acetoclastic methanogens and hydrogenotrophic methanogens. The acet o clastic methanogens consume acetate to prod uce methane and carbon d ioxide. By consuming the acetate which is the end product of the acid fermentation process, the acet o clastic methanogens maintain pH balance in the digester. Acet o clastic methanogenesis is responsi ble for the majority of methanogenesis in most anaerobic digester s ( Speece, 2008 ) H ydrogenotrophic methanogens produc e methane from carbon dioxide and hydrogen They often live in a mutualistic relationship, called syntropy, with the OPHA, by consuming the H 2 from the OPHA This reduces the concentrat ion of H 2 so the OPHA can metaboliz e organic acids into acetate which is then consumed by acetoclastic methanogens and the pH in the digester is kept in balance The syntropy between acetogens and methanogens is the driving force for functional anaerobi c digestion. The inevitable result of the sequential metabolism of anaerobic digestion is th at the overall rate of anaerobic digestion is dependent upon the slowest of the individual phase s. It is widely cited that for the anaerobic digesti on of high particulate feedstocks, such as food waste, hydrolysis is generally considered the rate limiting step ( Eastman & Ferguso n, 1981 ; Izumi et al., 2010 ; Palmowski & Muller, 2003 ; Wang et al., 2006 ) Therefore, to enhance the anaerobic digestion of food waste, the process of hydrolysis is of critical importance. Hydrolysis and Extracellular Enzymes Hydrolysis of solid organic material in an anaerobic digester is largely performed thr ough the production of extracellular enzymes by the hydrolytic microbial consortium These enzymes break bonds within the organ ic molecules and release the constituent monomers, which are then assimilated by the microorganisms (Figure 1 4 ) While a large number of different hydrolytic enzymes function in a digester, four important classes of hydrolytic enzymes are cellulases,

PAGE 23

23 amylases, lipases, and proteases. These four classes of enzymes hydrolyze some of the most prevalent component s of food waste (Table 1 1 ) The relative importan ce of each of these enzym e classes will vary depending on the comp osition of food waste (Figure 1 5 ) Figure 1 4 The hydrolysis of organic macromolecules into their constituent monomers Table 1 1 Enzyme cla sses and target substrates found in food waste Enzyme class Target food waste substrates Cellulases Broccoli cabbage corn cobs, fruit and vegetable trimmings Amylases Bread, rice, pasta, potatoes, Lipases Oil, dairy products, meat trimmings Proteases Meat, legumes, dairy products A. Restaurant food waste B. Butcher shop food waste C. Bakery food waste Figure 1 5 The variable composition of food waste. Note: These figures are for illustration only and do not represent measured values. Carbohydrates Proteins Lipids Lipids Proteins Carbohydrates Lipids Proteins Carbohydrates

PAGE 24

24 Cellulases Depending upon the source, certain food wastes c an have a high content of cellulose, particularly kitchen scraps or food processing wastes, which contain the inedible portion of produce, i.e. fruit and vegetable trimmings Cellulose is a long chain carbohydrate composed of crystalline glucose polymers In cellulosic plant material, crystalline cellulose is associated with amorphous regions of hemicelluloses a nd pectin. Due to its large size and crystalline structure, cellulose tends to degrade slower than more labile materials, such as starch Cellulases are hydrolytic enzymes produced by cellulolytic bacteria that break the bonds holding the glucose molecul es together. In anaerobic conditions, bacteria require close proximity to the cellulose molecule, and cell bound forms of cellulases are generally produce d ( Lynd et al., 2002 ) The bacteria produce a protuberance called a cellulosome that binds to both the bacterial cell wall and the cellulose surface. The cellulosome is a scaffolding protein structure that holds the active sites of the cellulases. Because the cellulosome must attach to the substrate surface, i ncreasing the substrate surface area can increase t he opportunities for these attachments to occur. Amylases Starch is one of the most prevalent macromolecules found in food waste Starch is a branched polymer of glucose monomers. Starch has two subunits, amylose, a straight chain molecule and amylopectin, side chain branches. Unlike cellulose which has a crystalline structure the branched structure of starch allows for faster hydrolysis. The hydrolysis of starch is facilitated by amylases, which break the bonds between the glucose monomers. Because of the abundanc e of starch in food waste, amylases are critical for the anaerobic digestion of food waste Wang et al. ( 2009 ) found that high lactic acid levels produced during acid fermentation inhibit amylases specifically glucoamylases The enzymes can become denature d through

PAGE 25

25 prolonged lactic acid inhibition Maintaining the organic acid balance, particularly lactic acid, is important during food waste digestion for proper starch hydrolysis to occur. Lipases Lipids are important mac romolecule s in anaerobic digestion because of the ir high methane production potential compared with carbohydrates or proteins ( Wilkie, 2008 ) Lipids however can be problematic in digester s due to clogging or flotation and interference with the degradation of other substrates ( Cirne et al., 2007 ) Therefore, the hydrolysis and digestion of lipids are important for the optimization of food waste anaerobic digestion. Lipases are extracellular lipolytic enzymes that break the lipid molecule into free fatty acids and glycerol. Lipases function at the lipid water surface interface by adhering to the lipid molecules ( Sharma et al., 2001 ) in a phenomenon termed interfacial activation. Lipases contain a surface loop structure that acts as a lid to cover its hydrophobic active site. When the lipase binds to the lipid surface the lid opens and exposes the lipase active site to the lipid. The hydrophobicity of the active site helps to bind the lipase to the lipid surface. Because lipases need direct contact with the lipid molecule increasing the surface area of food waste increases the opportunities for lipase/lipid interactions. Proteases Proteins are important in an anaerobic digester because they have higher methane potential than carbohydrates ( Wilkie, 2008 ) and because proteins are nitrogenous compounds, their presence in the feedstock increases the fertilizer value of the effluent through the re lease of ammoniacal nitrogen. Proteases are hydrolytic enzymes that break the peptide bonds between the amino acids of a protein. Protein molecules have a wide variety of structural morphologies, which can facilitate or impede access to the peptide bonds in the protein Collagen found in meat, for example has a tightly bound, triple helix tertiary structure which greatly reduces the

PAGE 26

26 degradability of the protein. Disrupting the tissue and exposing increased surface area of more recalcitrant proteins like collagen, can increase access of the proteases to the peptide bonds and facilitate protein hydrolysis. Factors Influencing Food Waste Hydrolysis While the presence of hydrolytic bacteria and enzymes is necessary for hydrolysis to occur during anaerobi c digestion, several physiochemical factors influence the performance of enzymatic hydrolysis. Three critical factors are pH inhi bition, product inhibition and temperature As with the microorganisms in a digester, the extracellular enzymes they produce also have an optimum pH range, and these ranges can vary depending on the enzyme. In the sequential metabolism of the anaerobic consortia the products of hydrolysis are fermented into organic acids. If the rate of acidogenesis exceeds the rate of metha nogenesis, organic acids may accumulate and cause a drop in pH, which results in enzyme inhibition and denaturing. In addition, enzymes can be inhibited through product inhibition from accumulation of the hydrolytic and acidogenic products, which cause s t he product/substrate equilibrium to favor unhydrolyzed substrate. He et al. ( 2006 ) studied the ef fects of pH and acetate inhibition on potato starch. They concluded that carbohydrate hydrolysis was inhibited by low pH, while protein hydrolysis was only inhibited by a low pH in the prese nce of acetate. Veeken et al. ( 2000 ) performed a similar study examining the impact of pH and VFA concentration o n the hydrolysis of the organic fraction of municipal solid waste and found the rate of hydrolysis was pH dependent. Temperature also has a significant impact on the activity of hydrolytic enzymes which have an optimum temperature range. High temperatures can denature enzymes while low temperatures can halt enzym e activity. Veeken and Hamelers ( 1999 ) examine d the hydrolytic rate of the organic fraction of municipal solid waste at 20C and 40C. A first order rate

PAGE 27

27 constant of 0.03 0.15/day was found at 20C, while a rate constant of 0.24 0.47/ day was found at 40C. T he solubilization of food waste was found to be great er at 35C and 45C than at 25C or at thermophilic temperatures (55C and 65C) ( Komemoto et al., 2009 ) High temperature c an denature enzymes, which may explain why a higher solubilization rate was found at 3 5C and 4 5C These studies suggest that mesophilic temperatures are ideal for the solubilization of food waste. Pretreatment of Food Waste An important common feature o f hydrolytic enzymes is that they require contact with the exposed surface area of the organic substrate Therefore, one limiting factor for the rate of hydrolysis is the surface area available for enzymatic interaction. Because hydrolysis is considered the rate limiting step in the anaerobic digestion o f high particulate feedstocks, increasing the substrate surface area and availability to hydrolytic enzymes can enhance the overall kinetics of anaerobic digestion. In order to increase the substrate surface area, numerous pretreatment techniques have been studied. While there is exhaustive literature studying the pretreatment of waste activated sludge or the org anic fraction of municipal solid waste literature address ing the pretreatment of food waste for anaerobic digestion is comparatively sparse. An important measure when studying hydrolysis is the fractionation of the chemical oxygen demand (COD) of a substr ate. Chemical oxygen demand is a surrogate measurement for organic matter content and is used to determine the stoichiometric methane potential of a substrate (1 g COD stoichiometrically produce s 0.35 L methane at STP). The total COD (TCOD) of a substrat e can be fractionated between soluble COD (S COD or CODsf as the soluble fraction of TCOD ) and particulate COD (PCOD). The endogenous CODsf of a substrate is the soluble material contained within cells and tissues, for example, glucose. Particulate COD c an be segregated into labile particulate matter and resistant particulate matter (Figure 1 6 )

PAGE 28

28 Examples of labile particulate matter include starches and oligosaccharides and examples of resistant particulate material s include cellulose, proteins, and tr iglycerides As PCOD solubilizes by enzymatic hydrolysis, t he CODsf of the substrate increases. Organic materials also contain recalcitrant material that has very low degradability. Recalcitrant organic matter can be either soluble (e.g. tannins) or particulate (e.g. lignin). Figure 1 6 Food waste composition by degradability. Note: Figure i s for illustration only and not based on measured values. Because the anaerobic consortia can only assimilate soluble material determining the rate at which the endogenous SCOD is released and the rate at which PCOD is hydrolyzed in to SCOD is critical in determining the anaerobic dige stion kinetics of the substrate. To illustrate the effects of COD s f on methane kinetics, Prashanth et al. ( 2006 ) conducted a study digesting a synthetic substrate. The substrate consisted of cellulose (PCOD) and sucrose and peptone (SCOD). These were loaded at different initial COD s f (100, 75, 50, 25, and 0%) and were loaded at varying feed/inoculums ratios. Afte r 60 days of digestion, the cumulative methane ( at Endogenous Soluble Labile Particulate Resistant Particulate Recalcitrant Organics and Mineral Ash

PAGE 29

29 STP) produced by 0% CODsf (pure cellulose) ranged from 92 248 mL/g COD, depending on inoculum loading rate. This is significantly below the stoichiometric maximum of 350 mL/g COD. When COD s f was 100% 60 day cumulative methane production was 264 35 0 mL/g COD. A COD s f of 7 5% resulted in 87% 91% of the methane produced by 100% COD s f. Below COD s f of 7 5% methane production dropped significantly. The authors concluded that at a COD s f less than 75%, hydrolys is is considered to be rate limiting. There fore by increasing the COD sf of food waste, the kinetics of anaerobic digestion are more favorable. This can be accomplished by two means, increasing the release of endogenous SCOD from the material and increas ing the substrates availability to hydrolytic enzymes. In order to increase the COD sf a variety of pretreatment methods, including thermal freezing/thawing enzymatic and mechanical have been reported in the literature. Thermal and Freezing/Thawing Pretreatment One method of increasing food waste solubilization is through thermal pretreatment. Heating food waste disrupts tissue and ruptures cells which releases SCOD and increases surface area for enzymatic hydrolysis Wang et al. ( 2006 ) found that th ermal pretreatment increased solubilization and biomethanation of food waste in a two phase digester (a two phase digester separates the acidogenic phase f rom the methanogenic phase in two reactors). Heating shredded food waste to 70C for 2 h or 150C for 1 h increased the initial COD sf, through endogenous SCOD release, from 17.2% for control (shredded food waste) to 20.3% and 31.7% respectively, for 70 C and 150 C pretreatments. The study also found increased methane production with thermal pretreatment; the digestion time to prod uce the same volume of methane decreased 33.3% with 70C pretreatment and 50% with 150C pretreatment compared with control.

PAGE 30

30 Liu et al. ( 2008 ) studied the effects of heating and freezing/thawing food waste prior to digestion in a two phased anaerobic digester. When food waste freezes ice crystals form that rupture cells and tissues W hen the food waste thaw s endogenous SCOD is released. They found that thermally pretreated, shredded food waste (150 C for 1 h) and frozen/thawed, shredded food waste (24 h at 20C then 12 h at 25C) both showed (via SEM micrographs of the food waste) increased porosity and looser cell structure, which increased surface area and availability for hydrol ytic enzymes. Both pretreatments also showed an increased release of endogenous SCOD. Freezing/thaw ing increased the initial CODsf from 14.9% (control, shredded food waste) to 25.4%. Thermal pretreatment increased initial CODsf from 7.8% (control, shredded food waste) to 16.9%. The SCOD was measured daily on the leachate from the acidogenic reactor. Cumulative SCOD production over the first 6 d was 13% and 39% greater than control for thermally pretreated food waste and frozen/ thawed food waste, respectively. This indicated th at pretreatment also enhanced the enzymatic hydrolysis of food waste. The increased solubilization kinetics showed corresponding increases in biomethanation kinetics. Total methane produced over 12 d increased by 29 % over control with thermal pretreatmen t and 11 % with freezing/thawing. To produce the same volume of methane, thermal pretreatment decreased operational time by 48%, while freezing / thawing decreased operational time by 42%. Both of these studies indicated that thermal pretreatment and freez ing/thawing can significantly increase the release of endogenous SCOD and substrate availability for hydrolytic enzymes, resulting in increased kinetics for solubilization and methane generation. However, thermal pretreatment and freezing / thawing are both energy inten sive methods of pretreatment. Liu et al. ( 2008 ) performed an energy balance between the energy required for pretreatment and

PAGE 31

31 the energy gained through increased methane kinetics, which showed the energy used offset the energy gained. Therefore, the net benefit of therma l and freezing/thawing pretreatment is null ified Another problem with these pretreatment methods is that the endogenous enzymes and microorganisms in the food waste which contribute to hydrolytic enzymes and the microbial consortia of the digester, can be denatured and k illed during the pretreatment. Enzymatic Pretreatment Enzymatic pretreatment has also been studied to increase solubilization of food waste. Because solubilization in anaerobic digestion is facilitated by hydrolytic enzymes, the addition of these enzymes as a pretreatment step can enhance food waste solubilization. Kim et al. ( 2005 ) used an enzyme cocktail of commercial carbohyd rases, lipases, and proteases for enzymatic pretreatment of food waste They found that blended food waste pretreated with 0.4% (v/v) enzyme cocktail resulted in a maximum CODsf of 68.5%, while an untreated control (blended fresh food waste) had a CODsf maximum of 37.5%. The enzyme treatment reached maximum solubilization within 4 d, with 61% of SCOD produced within the firs t 12 h ; the control, however, plateau ed within the first few hours. Enzymatic pretreatment may increase food waste solubilization ; however the cost of enzy mes may be prohibitive for commercial scale implementation of food waste anaerobic digestion. B eca use fo od waste composition varies, a uniqu e enzyme cocktail may be required for each type of food waste. If necessary, e nzyme pretreatment of food waste is most suited for specialized or unique food waste as a boutique pretreatment method. One example whe re enzymatic pretreatment may be useful is for high lipid food waste streams. Because of their hydrophobic nature, lipids can bind to other substrates and microbial biomass in the digester. This action inhibits digestion of the feedstock reduces methane production and can cause clogging problems, as was demonstrated by Neves et al. ( 2008 ) The study examined a standard food waste with an excess

PAGE 32

32 of lipids, cellulose, proteins, or carbohydrates added in different formulations The food waste with an excess of lipids showed a much slower b iomethanation rate. To reach 50% degradation with excess lipids, 14.8 d were requir ed, while all other food waste formulations required 3.0 5.9 days. To reach 85% degradation with excess lipids, 57 d were required, while other food waste formulations req uired 23 32 d. Because of the slow kinetics of lipid hydrolysis and digestion, as well as their potential to inhibit digestion of other substrates the authors recommend ed employing special measures with high lipid food wastes. One option would be the application of lipases. In lab scale experiments, t he application of pancreatic lipases to slaughterhouse waste has a positive effect on reducing particle size of fat molecules prior to anaerobic digestion ( Masse et al., 2001 ) Mendes et al. ( 2006 ) applied pancreatic lip ases to dairy processing waste with a high lipid concentration, and found positive results in hydroly tic and methanogenic kinetics. Within 12 h of lipase treatment, 39.5% of lipids were hydrolyzed. The 15 d cumulative methane pr oduction of lipase treated waste was over twice that of the treatment control For the anaerobic digestion of high lipid food waste, lipase addition may be a beneficial pretreatment me thod However, for gene ric food waste (i.e. restaurant, grocery store, and kitchen waste), enzym atic pretreatment may be a costly and unnecessary step when appropriate mechanical pretreatment and particle size reduction is applied. Mechanical Pretreatment Mechanical pretreatment functions through particle size reduction and tissue and cell maceration by various mechanical devices. Because hydrolysis is mediated by extracellular enzymes and their interaction with substrate surface area, increasing the surface area through mechanical pretreatment can increase food waste solubilization. Studies have shown the importance of particle size on the kinetics of anaerobic digestion of food waste Sanders et al. ( 2000 ) measured the solubilization of potato starch particles in batch experiments They found

PAGE 33

33 that the rate of solubilization of starch particles is directly related to the surface area. Palmowski and Mller ( 2003 ) tested the effect of communition on the ana erobic digestion of various organic materials including apple and rice in 1 L batch assays They found that through communition, there was a n increase in cumulative biogas production from the anaerobic digestion of apple and rice and a nearly 40% and 3 0%, respective decrease in digestion time over control Analysis of the specific surface area through nitrogen adsorption, of rice showed that there was a positive correlation between substrate surface area and endogenous SCOD releas ed. B ecause particle size and surface area are c ritically important in the anaerobic digestion of potato starch, apples and rice the use of mechanical pretreatment for food waste should show similar results. Izumi et al. ( 2010 ) examined the effects of mechanical pretreatment on a s tandard food waste. The food waste was pretreated using a household food disposer (control) and then further pretreated using a ball mill at different number s of revolutions. The mean particle size after disposer treatment was 0.888 mm, while additional ball milling for 300 revolutions resulted in mean particle size of 0.843 mm. The particle size decreased with increased ball milling, to reach a minimum of 0.393 mm with 40,000 revolutions. Ball milling also increased the release of endogenous SCOD of the food waste over disposer pretreatment alone. Food waste pretreated with the disposer alone had an initial solubilization of 28.1%; additional pretreatment with the ball mill at 300 revolutions increased solubilization to 39.8%. However the decrease in particle size with increased ball milling revolution s did not seem to increase the release of endogenous SCOD ; with 40,000 revolutions, the initial solubilization was 40.3%. This suggested that the entirety of the endogenous SCOD was released with minimal ball milling and was not correlated with decreased particle size by further pretreatment.

PAGE 34

34 The authors also explored the ef fect of mechanical pretreatment on methane production. The pretreated food waste was digested for 16 days under mesophilic conditions at a loading rate of 10 g COD/L in lab scale batch anaerobic digest ers Biogas production and pH were measured over the course of the digestion period. The decrease in particle size with the additional pretreatment of ball milling had a positive effect on methane production. Cumulative methane production for disposer treatment alone was 251 mL/g COD (at STP), while dispos er and ball mill ing for 1000 revolutions had a cumulative methane production of 322 mL/g COD (at STP). With increased ball milling above 1000 revolutions, however, methane production decreased. At 40,000 revolutions cumulative methane was only 254 mL/g C OD (at STP). The 40,000 revolutions treatment also showed the most significant initial drop in pH and the slowest recovery rate. The authors measured VFA concentration and found that the 40,000 revolutions treatment had a total VFA (acetic, propionic, n butyric, i butyric, n valeric, i valeric) concentration of 5,600 mg/L. This indicates that with excessive pretreatment, acidification through VFA accumulation can occur as the rate of hydrolysis and acidogenesis is greater than the methanogenic rate. Thi s study suggests that moderate mechanical pretreatment can enhance the anaerobic digestion of food waste by overcoming the rate limitation of hydrolysis; however, excessive pretreatment can increase the hydrolytic and acidogenic rates to a point where they exceed the methanogenic rate and inhibit methan ogen e sis The practical application s of this study are questionable because ball milling, while potentially increasing digestion kinetics, is not feasible for pretreatment in commercial scale food waste dige stion. Therefore, there is a need to study how practical pretreatment methods can increase the solubilization and biomethanation kinetics of food waste.

PAGE 35

35 Thesis Rationale The existing literature suggests that for the anaerobic digestion of food waste, hydrolysis is the rate limiting step. Food waste has a high particulate fraction that must be hydrolyzed and solubilized prior to microbial fermentation and methanogenesis. Because hydrolysis of particulate matter is facilitated by extracellular hydrolyt ic enzymes, increasing the availability of the food waste to these enzymes can enhance solubilization and consequently methanogenesis Additionally, disrupting tissue and rupturing cells can increase the immediate release of endogenous soluble material. Literature suggests that various means of food waste pretreatment can have a positive effect o n solubi lization and methane generation One of the major weaknesse s in the methods studied is that the pretreatment methods are often not practical for commerci al scale implementation. For example thermal pretreatment requires substantial energy consumption and ball milling of food waste is not suitable for digestion at other than lab scale. Another issue with the previous studies is that none examine d intact food waste as a control In all the studies, the control food waste was itself pretreated (i.e. shredded or ground) prior to the pretreatment methods examined. Therefore, the true impact of the pretreatment method tested was obscured because the food waste was in fact pretreated prior to the pretreatment experiment Considering these issues this current effort sets forth to examine the enhanceme nt of food waste solubilization and biomethanation through low tech, practical mechanical pretreatment methods. The pretreatment methods were selected with implementation in mind so that any resulting enhancement of anaerobic digestion could be expected on site at a restaurant, school or grocery store. In many cases these location s already have pretreatment apparatuses on site (i.e. food disposer or meat grinder) so pretreatment for a food waste digester could fit within existing infrastructure Additionally th e current work compares the effects of mechanical pretreatment

PAGE 36

36 against intact, fresh food waste. By comparing pretreat ed food waste to intact food waste, the true i mpact of pretreatment on solubilization and biomethanation can be measured Hypothesis M echanical pretreatme nt of food waste will disrupt tissues and rupture cells, which will enhance food waste solubilization through increased release of endogenous soluble material and greater availability to hydrolytic enzymes Enhanced solubilization kinetics can increase biomethanation through increased generation of soluble materials for metabolism by the anaerobic consortia Objectives The goal of this study is to examine to the effects of mechanical pretreatment on the anaerobic digestion of food waste. There are two primary objectives that will be addressed: 1) a ssess the effects of pretreatment on solubilization kinetics of food waste and 2) a ssess the effects of pretreatment on biomethanation kinetics of food waste.

PAGE 37

37 CHAPTER 2 MATERIALS AND METHOD S Standard Food Waste Empirical evi dence has indicated that food waste is a highly variable substrate Food waste audits were conducted at local schools and restaurants over a two week period T he daily generated food waste was analyzed for total solids (TS), volatile solids (VS), total chemical oxygen demand (TCOD), total nitrogen (TN), and total phosphorus (TP) (Table 2 1) in order to develop a parameter range of food waste collected in the community. Due to the variability of food was te, it was deemed necessary to develop a standard food waste a s a substrate for repeatable experimentation. The standard food waste was developed to be within the range of the physiochemical p ropertie s of food waste collected during the waste audits. The standard food waste also contained components representing various constituent macromolecules of food waste (i.e. carbohydrates, lipids, and fats). For each experiment, the s tandard food waste was formulated immediately prior to the experiment in order t o fully capture the degradability of fresh food waste. P recautions were tak en (e.g. produce was washed, gloves were worn when handling ) to reduce microbial contamination that could impact the integrity of the food waste. Table 2 1. Range of p hysiochemica l properties of food waste P roperty Schools (n=30) Restaurants (n=50) T otal solids (TS) (%) 25.45 42.57 10.64 41.55 V olatile solids (% TS) 84.71 95.09 85.63 97.73 T otal chemical oxygen demand (g/kg w et weight (ww) ) 342.04 511.96 127.67 565.94 T otal nitrogen (% TS) 2.03 4.44 2.17 3.95 T otal phosphorus (% TS) 0.32 0.75 0.23 0.44 Composition The standard food waste used in the experiments consisted of six components: apple, potato, bread, broccoli, beans, and cheese (Table 2 2). These ingredients were proportioned

PAGE 38

38 empirically to simulate actual food waste. Each ingredient was selected to represent a range of different types of macromolecules found in food waste. Table 2 2. Composition of standard food waste Components Per cent composition ( ww ) Percent composition (TCOD) Macromolecule represented Apple, Red Delicious 24 11.9 Carbohydrate (sugar, pectin) Potato, Russet 24 13.7 Carbohydrate (starch) Beans, red kidne y (canned, drained and rinsed) 20 21.2 Protein Broccoli (florets) 12 4.28 Carbohydrate (cellulose) Bread, white hamburger bun 12 29.7 Carbohydrate (starch) Cheese, sharp Cheddar 8 19.2 Protein, lipid Analysis Physiochemical p roperties were measured on the standard food waste to confirm that it was analogous to the food waste collected in the waste audits. Total solids, VS, TCOD, SCOD, conductivity, alkalinity, TN, and TP were measured on t he standard food waste (Table 2 3). Table 2 3. Physiochemical properties of standard food waste P roperty Mean Standard Deviation T otal Solids (TS) (%) 29.94 0.20 V olatile solids (% TS) 95.41 0.001 T otal chemical oxygen demand (g/kg ww ) 348.89 22.87 Soluble chemical oxygen demand ( g/kg ww ) 87.9 2.7 Conductivity (mS/cm) 4.37 0.16 Alkalinity (mg CaCO 3 eq. /kg ww ) 1344.8 3.67 T otal nitrogen (% TS) 3.08 0.12 T otal phosphorus (% TS) 0.32 0.0008 Note: All parameters were measured in triplicate Food Waste Microscopy Intact and pretreated food waste was examined micro scopically to observe the structural changes in the cells and tissues through pretreatment. Four components of the standard food waste (apple, bean, broccoli, and potato) were individually observed. For intact food waste, small cross sections were delica tely cut using a scalpel to preserve cell integrity. For pretreated

PAGE 39

39 food waste, each component was ground using a manual meat grinder with 0.5 cm plate openings (Grizzly H7778 No 32, Grizzly Industrial, Inc., Bellingham, WA ). A small sample was placed on a microscope slide with glass coverslip. Photos were taken under darkfield illumination on a Nikon Labophot compound microscope (Nikon Corporation Tokyo, Japan) using a Spot Insight color mosaic digital camera (Diagnostic Instruments Inc., Sterling Heights, MI ) Solubilization Assay To study the impact of pretreatment on food waste solubilization and hydrolysis, t hree solubilization assays were conducted: a n endogenou s solubilization assay, commercial enzyme assay and m icrobial inoculum assay. The full trea tment regime is shown in Table 2 4 Each treatment was replicated in triplicate A phosphate buffer ( 0.5 M at pH 6.5) was used to prevent acidification, which can inactivate and denature hydrolytic enzymes. Food waste was combined with the other components of the assay s (i.e. buffer, DI water, enzymes, and/or inoculum), and placed in a 1 L gl ass beaker. Beakers were covered with plastic film (Glad Cling Wrap) held with a rubber band to reduce evaporation and po tential contamination. Ass ays were maintained at 35C in a static water bath (Versa Bath S Model 236, Thermo Fisher Scientific, Waltham, MA ) Experimental blanks of commercial enzyme and microbial inoculum were included in the assay to account for non fo od waste SCOD Table 2 4 Treatment regimes for solubilization assays Endogenous solubilization Commercial e nzyme Microbial inoculum Food waste 10 g (ww) 10 g (ww) 10 g (ww) Phosphate buffer 250 mL 250 mL N/A Commercial enzyme N/A 1 g N/A Microbial inoculum N/A N/A 990 mL Total volume 1 L (640 mL DI water) 1 L (640 mL DI water) 1 L Loading Rate 3.5 g COD/L 3.5 g COD/L 3.5 g COD/L Treatments Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposer Intact Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposer Intact Grinder (0.5 cm plate) Intact

PAGE 40

40 Mechanical P retreatment Three mechanical pretreatment methods were studied in the endogenous solubilization and commercial enzyme solubilization assays (Table 2 5 ) : an in sink food disposer (Badger 9 model Insinkerator, Racine, WI), and a manual meat grinder with two different plate s (0.5 cm and 1.0 cm plate openings) (Grizzly H7778 No 32, Grizzly Industrial, Inc., Bellingham, WA). Based on results from the first two assays, the grinder with the 0 .5 cm plate was selected as the representative pretreatment method for the microbial inoculum assay To ensure representativeness 500 g of the standard food waste was formulated and ground separately for each pretreatment m ethod. Triplicate 10 g (ww) mixed, representative sample s from each 500 g batch of pretreated food waste were used in the assay. Pretreatment occurred just prior to the assay in order to measure the immediate solubilization kinetics of the food was te. T he nominal time required for pretreatment was 0.1 h. For intact food waste, a kitchen knife was used to cut a single, cubic piece of each component needed to formulate 10 g (ww) of the standard food waste, for example, a 2.4 g cube of p otato and 0.8 g cub e of cheese Any v ariation in TCOD was measured in triplicate on each formulation of the standard food waste and accounted for when calculating percent solubilization (CODsf). Commercial Enzyme To measure enzymatic hydrolysis in the solubilization assa y a commercial digestive enzyme powder (Digest, Enzyme d ica Inc., Port Charlotte, FL) was employed. The formula include d a variety of hydrolytic enzymes including cellulases, amylases, proteases, and lipases. The entire enzyme formulation is shown in Tabl e 2 5 According to a customer representative from Enzyme d ica Inc., the enzyme sources were Aspergillus niger and Aspergillus orzay e Prior to use in the assay, the requisite amount of enzyme capsules were emptied and enzyme powder within was completely mixed to ensure representativeness

PAGE 41

41 Table 2 5 Commercial enzyme cocktail formulation Enzyme Concentration (per 0.25 g) Unit definition Amylase 12,000 DU Dextrinizing units Protease 42,000 HUT Hemoglobin units on a tyrosine base Lipase 500 FCC LU Food Chemical Codex lipase units Cellulase 200 CU Cellulase units Lactase 850 ALU Acid lactase units Alpha Galactosidase 75 GALU Alpha galactosidase units Maltase 200 DP Diastic power units Invertase 175 INVU Invertase units Pectinase 50 Endo PGU Endo polygalactur onidase units Microbial Inoculum The microbial inoculum used in the solubilization assay was formulated by mixing 10 g (ww)/L of standard food waste (ground with 0.5 cm plate meat grinder), 100 mL/L of flushed dairy manure (UF Dairy Uni t, Hague, FL) 250 mL/L of phosphate buffer (0.5 M at 6.5 pH), and 640 mL/L of DI water The inoculum was placed in crimped capped serum bottle s (Wheaton, Millville, NJ) and incubated at 35C in a water bath (Versa Bath S Model 236, Thermo Fisher Scientific, Waltham, MA ) for 30 days to allow microorganisms to acclimate to the food waste and to produce extracellular hydrolytic enzymes. Physiochemical p roperties of the inoculum are shown in Table 2 6 One day prior to assay, serum bottles were open ed and inoculum was screened using a 20 mesh wire sieve (850 m opening) (Newark Wire Cloth Co., Newark, NJ) to remove large particulates Screened inoculum was placed into two 4 L glass beakers, covered with plastic film (Glad Cling Wrap) and held at 35C overnight prior to the assay Inoculum was continuously mixed using a magnetic stir bar while adding to food waste in the assay. Table 2 6 Physiochemical p roperties of microbial inoculum used in solubilization assay P roperty Microbial Inoculum Source 10 % flushed dairy manure with food waste TCOD (mg/L) 2932 77.5 SCOD (mg/L) 1685 41.9 pH 6.46

PAGE 42

42 Sampling Technique At 0, 1, 2, 4, 6, 8, 12, and 24 h SCOD and pH were measured. Each replica te was removed from the water bath, one at a time, and t he plastic film cover was removed The replicat e was thoroughly m ixed using a magnetic stir bar, and a 2 0 mL aliquot was sampled for SCOD analysis. The aliquot was immediately placed into a refrigerat ed water bath (Refrigerated Bath Model 90, Thermo Fisher Scientific, Waltham, MA ) at 4 C to quench enzyme activity. Following aliquot sampling, pH was measured on the replicat e The plastic film cover was replaced and replicat e was returned to the water bath. Statistical Analysis Solubilization and hydrolysis were fit to first order kinetic models using non linear fitting. Where: CODsf= COD soluble fraction at time t (%) CODsf f =ultimate CODsf (%) k=kinetic rate constant (h 1 ) The solver application in Microsoft Excel 200 7 was utilized to solve for CODsf f and k by minimizing the residual sum of squares of triplicate data points. Standard error of each te used to determine significan t differences between mean values of parameters. Biochemical Methane Potential Assay Biochemical methane potential (BMP) assays were conducted to determine the effect of pretreatment on the methane production kin etics of food waste. The BMP assay used in t his study is a modification of the methods developed in Owen et al. ( 1979 ) Two BMP assays were conducted on pretreated and intact food wa ste: a moderate loading rate assay and a re duced loading rate assay. Table 2 7 shows the full assay regime for each BMP assay. The moderate

PAGE 43

43 loading rate and reduced loading rate BMP assay s were loaded with the standard food waste at nominal organic loading rates of 3.5 g COD/L and 2 g COD/L, respectively. The assays were conducted using glass serum bottles with a nominal volume of 534 mL (Wheaton, Millville, NJ). Food waste was either in tact or pretreated using a meat grinder with 0.5 cm plate openings per the methodology used for the solubilization assays. For the moderate loading rate assay, bottles were inoculated w ith flushed dairy manure. For the reduced loading rate assay, the dig estate from the moderate loading rate assay was conserved and reused as an inoculum. While mixing, 400 mL of the inoculum was poured into each bottle. Positive controls were run using glucose and cellulose at a stoichiometric loading rate of 2.13 g COD/L and 2 g COD/L for the moderate loading rate and reduced loading rate assays, respectively, with 400 mL of inoculum. An experimental blank was included in each assay to account for methane production from the inoculum All treatment s control s and blanks were conducted in triplicate. Once all bottles were loaded, they were sealed with a rubber septum stopper and an aluminum crimp cap (Wheaton, Millville, NJ) Bottles were inverted to prevent potential gas leakage and placed into a 35C incub ator. Bottles were manually shaken once daily. Table 2 7. Bi ochemical methane potential assay r egime Moderate loading rate BMP Reduced loading rate BMP Pretreated l oading rate (g COD/L measured ) 3.55 2.05 Intact loading rate (g COD/L measured ) 3.48 2.18 Glucose and cellulose loading rate (g COD/L) 2.13 2.00 Inoculum added to each bottle (mL) 400 400 Simulated BMP Assay for pH Measurements Foll o wing the moderate loading rate BMP assay, a simulated BMP assay was conducted in which subsamples were taken for 20 days from each bottle for measurement of pH. Methane production was not measured in the simulated assay. The assay was conducted in triplicate with

PAGE 44

44 p r etreated and intac t food waste loaded at three different loading rates: 2, 4, and 8 g COD/L. Bottles were manually mixed during sampling and a 10 mL aliquot was taken by piercing the septum with a stainless steel needle (19 gauge) and 12 mL syringe. Aliquot pH was measure d immediately with limited agitation or mixing to reduce CO 2 evolution. Bottles were manually degassed to prevent acidification through CO 2 accumulation. The inoculum in the simulated BMP assay was flushed dairy manure. Methanogenic Inoculum The in o culum used in the moderate loading rate assay was flushed dairy manure obtained from the University of Florida Dairy Unit (Hague, FL). In the reduced loa ding rate assay, the digestate from the moderate loading rate assay was conserved and euse d as an inoculum Prior to use in each assay, inoculum was thoroughly mixed and screened using a 20 mesh wire sieve (850 m opening) (Newark Wire Cloth Co., Newark, NJ) to remove large particulates. A 400 mL mixed, representative aliquot of screened inoculum was used in each replicate of the assay. Total COD, SCOD, and pH of each BMP i nocul um are presented in Table 2 8 Table 2 8 Physiochemical p roperties of inocula f or the biochemical methane potential assays Property Moderate loading rate BMP Reduced loading rate BMP Source 100% Flushed dairy manure Digestate of moderate loading rate BMP TCOD (mg/L) 4108 80.8 2582 64.2 SCOD (mg/L) 620 9.9 487 6.3 pH 7.54 7.82 Methane Production Measurement Methane production was measured volumetrically through hydraulic displacement using a modification to methods described in Wilkie et al. ( 2004 ) Bottles were removed from the incubator and measured individually. The apparatus for methane measurements ( Figure 2 1) consisted of a short, clamped tube (B) with a needle on either end (C) connected between the BMP s erum bottle (A) and an inverted 250 mL serum bottle filled with 5M KOH (D). The tube

PAGE 45

45 was opened allowing the biogas to flow from the BMP bottle and bubble through the 5M KOH. By passing through 5M KOH, carbon dioxide was stripped out of the biogas into solution leaving only methane in the headspace of the KOH bottle. Alizarin was added to the KOH solution at a rate of 1g/L as a pH indicator for carbon dioxide saturation. As pressure built in the KOH bottle, the solution was displaced through another tu be in the septum (F) which was connected to a 50 mL pipette with 0.5 mL graduations (G). Once gas ceased bubbling from the BMP bottle and the KOH solution was equilibrated between the bottle and pipette, the mL of KOH displaced into the pipette was measur ed, which corresponded to the methane produced from the BMP bottle. The BMP bottle was then disconnected and the pipette was zeroed by placing KOH bottle upright and draining the displaced KOH solution back into the bottle while venting the headspace with a needle through the KOH bottle septum The BMP bottle was then returned to the incubator. Figure 2 1. Apparatus for measuring methane production through hydraulic displacement. Components are: A) BMP Bottle, B) tube for biogas flow, C) needles in septa, D), 5M KOH with alizarin in 250 mL serum bottle, E) roller clamps for tube closure, F) tube for displace d KOH solution, G) 50mL pipette with 0.5 mL graduations, H) stand.

PAGE 46

46 Gas mea surements were taken for 30 days in both BMP assays All methane production values were standardized to standard temperature and pressure (STP) (0C and 1 atm) and calculated per g COD loaded basis after subtracting methane production from the inoculum blanks The percent of substrate COD converted to methane (% COD removal) was calculated based on the stoichiometric COD equivalence of methane (2.86 g COD/L CH 4 @ STP) as a fraction of TCOD loaded. Total COD, SCOD, pH, and conductivity were measured on the digestate from each bottle in both BMP assay s after 3 0 days. Alkalinity was also measured on th e digestate from each bottle in the reduced loading rate assay after 30 days Statistical Analysis Cumulative methane production w as fit to a first order kinetic model using non linear fitting. Where: CH 4 = cumulative methane production at time t (mL/g COD loaded @STP) CH 4f =ultimate cumulative methane production (mL/g COD loaded @STP) k=kinetic rate constant (day 1 ) The solver application in Microsoft Excel 200 7 was utilized to solve for CH 4 f and k by minimizing the residual sum of squares of triplicate data points. Standard error of each 05) was used to determine significan t differences between mean values of parameters. Physiochemical Parameters Total Chemical Oxygen Demand Total COD was measured using a modification to standard methods ( APHA, 2005 ) Triplicate 25 g (wet weight (ww)) mixed, representative samples of food waste were obtained. The samples were fully homogenized using a 360 mL stainless steel blender (Waring,

PAGE 47

47 Torrington, CT) for 1 minute on high speed Deionized water was added during blend ing to facilitate homogenization. The entire homogenized samples were diluted to 500 mL with DI wate r using volumetric flasks. The diluted samples were poured into 600 mL glass beak ers and fully homogenized with magnetic stir bars Using a transfer pipette and volumetric flask, a 25 mL aliquot was obtained from each sample and further diluted to 250 mL with DI water in a 250 mL volumetric flask for a final dilution of 1 g food waste/200 mL. A 1 mL aliquot of the 1/200 diluted samples and 1 mL DI water were added to Hach COD reagent tubes (HR 20 1500 mg COD/L). Tubes were digested for 2 h at 150C in a Hach Model 45600 COD reactor (Hach Company, Loveland, CO). Digested tubes were read on a Hach 890 colorimeter (Hach Company, Loveland, CO). Total COD on inocula and digestate following BMP assays was measured by diluting a 5 mL mixed aliquot to the req uisite dilution within the range of the COD reagent tubes. Two mL of diluted aliquot were then placed in the COD tube and analyzed as described above. Soluble Chemical Oxygen Demand To measure s oluble chemical oxygen demand (SCOD) samples were completely mixed with a magnetic stir bar to obtain representative aliquot s. A 20 mL aliquot was transferred into a 25 mL glass centrifuge tube using a transfer pipette. Tubes were promptly placed into a refrigerated water bath (Refrigerated Bath Model 90, Thermo Fisher Scientific, Waltham, MA ) at 4 C to quench enzyme activity and prevent further metabolism Aliquots were then centrifuged at 12,000 rpm for 30 minutes in a refrigerated centrifuge (Sorvall RC 5B Refrigerated Superspeed Centrifuge, DuPont Instrumen ts, Wilmington, DE) at 4 C. After centrifuging, the supernatant of each aliquot was filtered through a 0.45 m nylon syringe filter ( Thermo Fisher Scientific, Waltham, MA ) using a 3 mL syringe. Filtered supernatant was diluted to required concentration w ith DI water. Two mL of diluted supernatant were placed into

PAGE 48

48 a Hach COD reagent tube (HR 20 1500 mg COD/L) and digested at 150C for 2 h in a Hach Model 45600 COD reactor (Hach Company, Loveland, CO). Digested tubes were read on a Hach 890 colorimeter (H ach Company, Loveland, CO). To measure endogenous SCOD of food waste, triplicate 25 g (ww) mixed, representative samples of food waste were obtained. The samples were fully homogenized using a 360 mL stainless steel blender (Waring, Torrington, CT) for 1 minute on high speed Deionized water was added during blend ing to facilitate homogenization. Blended food waste was diluted to 250 mL with DI water. A 20 mL aliquot was measured for SCOD as described above. Total Solids and Volatile Solids Total solids (TS) and volatile solids (VS) were measured according to standard methods ( APHA, 2005 ) Triplicate 100 g mixed, representative samples of food waste were weighed into pre ashed, pre weighed 200 mL disposable aluminum dishes. Samples were dried at 103C in a drying oven (Precision Model STG 80 Thermo Fisher Scientific, Waltham, MA) for 24 h. Dried samples were placed in a desiccator to cool to room temperature, then weighed to record TS. To measure volatile solids (VS), dried samples were ashed for 2 h in a n ashing furnace (Thermolyne 30400 Thermo Fisher S cientific, Waltham, MA) at 550C. Ashed samples were placed in a dessicator to cool to room temperature and then weighed. Ash weight was subtracted from TS to calculate VS. pH An Accumet Model 10 pH meter (Thermo Fisher Scientific, Waltham, MA) usi ng a Ross Sure Flow combination electrode was used for measuring pH Samples were gently mixed with a magnetic stir bar during measurement to reduce CO 2 evolution Sample temperature was accounted for when measuring pH.

PAGE 49

49 Total Ni trogen and Total Phosphor us Tot al nitrogen (TN) and total phosphorus (TP) were measured by blending a 25 g representative sample of food waste in a 360 mL stainless steel blender for 1 minute. Triplicate 0.25 g representative subsamples of the blended food waste were weighed onto 1 Kim Wipe tissue (1 ply, 11 x 21 cm). Samples, including KimWipe, were digested using a modification of the aluminum block digestion procedure of Gallaher et al. ( 1975 ) Catalyst used was 1.5 g of 9:1 K 2 SO 4 :CuSO 4 and digestion was conducted for at least 4 h at 375C using 6 ml of H 2 SO 4 and 2 ml H 2 O 2 Nitrogen and phosphorus in the digestate were measured by semiautomated colorimetry ( Hambleton, 1977 ) A blank KimWipe was also measured to correct for the TN and TP in the tissue. Al kalinity Alkalinity was measured using a modification to standard methods ( APHA, 2005 ) A 25 g (ww) sample of food waste was mixed with 25 mL DI water. The food waste solution was then titrated with 0.12 N H 2 SO 4 (standardized with 0.05 N NaCO 3 ) while mixing with a magnetic stir bar until a pH of 4.5 was reached. Alkalinity was calculated using the following formula. Alkalinity on digestate from BMP assay s was measured by titrating a 50 mL mixed aliquot using the procedure described above. Alkalinity was calculated using the following formula.

PAGE 50

50 Conductivity Conductivity was measured on a n Accumet Model 30 conductivity meter (Thermo Fisher Scientific, Waltham, MA). The meter was standardized with 0.01M KCl. To measure conductivity of food waste, 25 g (ww) of food waste was diluted with 25 mL DI and mixed using a magnetic stir bar. Organic Acid and S ug ar A nalysis Organic acid s and sugar s were measured by high pressure liquid chromatography (HPLC) using an HP 1090 Series II chromatograph (Hewlett Packard, Palo Alto, CA) equipped with a Bio Rad Aminex HPX 87H ion exclusion column (45C; solvent phase, 4 mM H 2 SO 4 ; flow rate, 0.5 ml min detector at 210 nm). One mL of supernatant samples from SCOD analysis was filtered using a Ten were added to 1 mL samples Samples were analyzed with standards for glucose, xylose, lactic acid, succinic acid, formic acid, acetic acid, propionic acid, butyric acid, and ethanol.

PAGE 51

51 CHAPTER 3 EFFECT OF MECHANICAL PRETREATMENT ON SOLUBILIZATION OF FOOD WASTE The anaerobic digestion of high particulate feedstock, such as food waste, is considered to be l imited by the rate of solubilization (Eastman and Ferguson 1981, Palmowski and Mulle r 2003, Wang et al. 2006, Izumi et al. 2010). Increasing the solubilization of food waste through pretreatment is proposed as a solution to increase the rate and extent of food waste anaerobic digestion. Pretreatment methods examined in literature includ e thermal freezing/ thawing, enzymatic and mechanical However the methods proposed in literature are energy a nd resource intensive and are im practical for commercial scale utilization Therefore, it is necessary to study the enhancement of food waste solubilization kinetics through implementing low tech, practical pretreatment methods. T hree mechanical pretreatment methods were examined in this study: a manual meat grinder (with two different plate sizes) and an in sink food disposer Different pretr eatment methods were tested to discern any measureable differences between these methods. I n order to study the effect of pretreatment on food waste solubilization, a series of solubilization assays were conducted comparing pretreated food waste to intact food waste. Solubilization is defined in the present study as the sum of the release of endogenous soluble organic material and the enzymatic hydrolysis of particulate material By placing food waste in water, the release of the endogenous soluble organ ic material was able to be measured The assay measured the rate and extent that the soluble material is released into the aqueous solution. This release can be a result of both physical cell rupture and tissue disruption or hydraulic leaching. Enzym atic hydrolysis was measured by adding either a commercial hydrolytic enzyme powder or a microbial inoculum to the food waste Hydrolysis was then calculat ed by subtracting the release of endogenous soluble organic material from the total solubilization Solubilization in the assays was measured as soluble chemical oxygen demand (SCOD) and was

PAGE 52

52 normalized as a fraction of total COD (TCOD) to account for natural variation in the standard food waste. Solubilization is presented as the COD soluble fraction (CODsf) and is divided into solubilization through endogenous SCOD release (CODesf) and solubilization though enzymatic hydrolysis (CODhsf) In the assays, buffering and dilution were required because preliminary solubilization assays without buff ering o r dilution resulted in sharp decrease s in pH (Figure 3 1 ), which inhibited enzymatic and microbial activity. This chapter discusses the results of three solubilization assays: 1) a n endogenous solubilization experiment 2) a commercial enzymes experiment and 3) a microbial inoculum experiment Figure 3 1 Acidification of food waste under unbuffered, undiluted conditions Note : food waste loading rate was 500 g (ww)/L (175 g COD/L) Microscopy of Pretreated Food Waste Intact and pretreated food waste was observed micros copically to eluc idate structural changes in food waste through mechanical pretreatment that would favor increased solubilization. Figure s 3 2 to 3 5 show photo micrographs of intact (A) and pretreated (B) standard food waste components. The photo micrographs show that pretreatment leads to significant structural changes that result in both cell rupture, which release s in tra cellular content, and tissue disruption, which causes looser structure to facilitate enzymatic hydrolysis. The 3 3.5 4 4.5 5 5.5 6 0 12 24 36 48 60 72 pH time (h) Grinder without added enzymes Grinder with added enzymes Intact without added enzymes Intact with added enzymes

PAGE 53

53 solubilization assays were conducted to quantify the effects of this structure change on solubilization. A B Figure 3 2. Photom icrographs of apple. A) Intact apple with intact skin tissue (1) and intact cells (2). B) Pretreated apple with released int ra cellular material (3) and cell wall debris (4). 1 2 3 4

PAGE 54

54 A B Figure 3 3 Photom icrographs of bean A) Intact bean with intact skin tissue (1) and intact cells (2). B) Pretreated bean with fragmented s kin tissue (3) and released intra cellular material (4). 2 1 3 4

PAGE 55

55 A B Figure 3 4 Photom icrographs of broccoli. A) Transverse section of i ntact broccoli stem with intact vascular tissue and (1) and intact cells (2). B) Pretreated broccoli with fragmented stem tissue (3) and released int ra cellular material (4). 3 4 2 1

PAGE 56

56 A B Figure 3 5 Photom icrographs of potato. A) Intact potato with intact cells (1) and int ra cellular starch (2). B) Pretreated potato with cell debris (3) and free starch (4). 1 2 3 4

PAGE 57

57 Endogenous Solubilization Assay A n endogenous solubilization assay was conducted by adding intact and pretreated food waste to water The purpose of this assay was to determine release of the endogenous SCOD of pretreated and intact food waste. When measuring the physiochemical parameter of the standard food waste, it was determined that the endogenous soluble chemical oxygen demand (SCOD) was 87.9 mg/kg (ww) or 25% soluble fraction of COD (CODsf) (Table 2 3) This is the approximate maximum of endogenous SCOD release. Twenty four Hour Solubilization The assay was conducted for 24 h with SCOD and pH measured at 0 1, 2, 4, 6, 8, 12 a nd 24 h ( for data analysis purposes 0.1 h is the nominal time allocated for pretreatment) Figure 3 6 present s the triplicate SCOD measurements of intact and pretreated food waste In all treatments, except intact food waste, solubilization appeared to pl ateau prior to 8 h and solubilization decreased at 12 and 24 h This decrease was likely due to mi crobial assimilation of SCOD. Microbial assimilation was suggested by the decrease in pH at 12 and 24 h (Figure 3 7 ), which was a result of bacterial fer mentation and acidogenesis Additionally, o rganic acid and sugar analysis of the samples confirm ed that sugars were al most entirely consumed by 12 h (Figure s 3 8 and 3 9 ) At 1 h, 1 5% of SCOD of pretreated food waste was glucose but was below d etection by 12 h For further data analysis, the 12 and 24 h data points were disregarded because the bacterial assimilation of SCOD confounds any further increases in solubilization Figure 3 10 shows the mean solubilization of food waste over the first 8 h. B ecause there was minimal enzymatic hydrolysis in this assay, the total solubilization (CODsf) represent ed solubilization through the release of endogenous SCOD (CODesf) from the food waste.

PAGE 58

58 Figure 3 6 Soluble chemical oxygen demand in the endogenous solubilization assay. Note: Total COD of gr inder (0.5 cm plate), gr inder (1.0 cm plate), disposer and intact food waste is 3284, 3368, 3556, 3688 mg/L, respectively. Figure 3 7 Mean pH for the endogenous solubilization assay. Error bars represent stan dard deviation. 0 200 400 600 800 1000 1200 0 4 8 12 16 20 24 mg SCOD/L time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposal Intact 6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7 0 4 8 12 16 20 24 pH time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposer Intact

PAGE 59

59 Figur e 3 8 Sugars as a percent of SCOD for pretreated food waste (grinder 0.5 cm plate) in the endogenous solubilization assay Error bars represent standard deviation. Figur e 3 9 Sugars as a percent of SCOD for intact food waste in the endogenous solubilization assay. Error bars represent standard deviation. 0 0.5 1 1.5 2 2.5 0 5 10 15 20 25 30 0 4 8 12 16 20 24 % SCOD (Xylose) % SCOD (Glucose) time (h) Glucose (Grinder 0.5 cm) Xylose (Grinder 0.5 cm) 0 0.5 1 1.5 2 2.5 3 0 5 10 15 20 25 30 0 4 8 12 16 20 24 % SCOD (Xylose) % SCOD (Glucose) time (h) Glucose (Intact) Xylose (Intact)

PAGE 60

60 Figure 3 10 Mean solubilization of food waste in the endogenous solubilization assay. Error bars represent standard deviation. Endogenous Solubilization Kinetics The release of endogenous SCOD was fit to a first order kinetic equation with the kinetic rate constant (k) and final endogenous soluble fraction (CODesf f ) a s fitted parameters. Estimates of parameters Table 3 1. Estimated parameters and 8 h endogenous solubil ization in endogenous solubilization assay b: Standard error T able 3 1 shows estimates of the fitted parameters and 8 h CODesf for each treatment in the assay. All three pretreatment methods showed significantly higher kinetic rate constants (k) and COD esf f than intact food waste The meat grinder with the 0.5 cm plate showed a 0 5 10 15 20 25 30 35 0 1 2 3 4 5 6 7 8 CODsf (%) time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposal Intact k (h 1 ) CODesf f (%) CODesf 8 (%) Treatment Estimate a SE b Estimate a SE b Grinder (0.5 cm plate) 24.328 X 2.564 27.840 X 0.260 27.840 Grinder (1.0 cm plate) 12.654 Y 1.356 27.146 XY 0.442 27.146 Disposer 14.907 Y 0.912 27.222 Y 0.236 27.222 Intact 0.364 Z 0.028 21.089 Z 0.609 19.943

PAGE 61

61 significantly higher k than other two pretreatmen ts and a significantly higher CODesf f than the disposer f Fitted curves. Figure 3 11 Fitted endogenous solubilization curves for the endogenous solubilization assay The fitted curves with measured data p oints are presented in Figure 3 11 All three pretreatment methods show ed much faster solubilization than intact food waste. T he pretreatment immediately release d the en dogenous SCOD of the food waste. T he meat grinder (0.5 cm plate) show ed more immediate release than e ither other pretreatment; however this difference was negligible as the differen ce was only within the first hour. All three pretrea tments released the full complement of endogenous SCOD ( measured at approximately 25%) (Table 2 3) immediately through the macerating action of pretreatment alone Intact food waste, however, show ed a much slower release of endogenous SCOD (k=0.36 h 1 ). The release of endogenous SCOD from intact food waste occu r r ed through hydraulic leaching, which was a 0 5 10 15 20 25 30 35 0 1 2 3 4 5 6 7 8 CODesf (%) time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposal Intact CODesf=CODesf f (1 e ( k*t) )

PAGE 62

62 slow process compared to the immediate expulsion through pretreatment. A fter 8 h the full complement of endogenous SCOD was not releas ed from the intact food waste. Commercial Enzyme Assay For particulate feedstoc ks, the majority of solubilization occurs as a result of enzymatic hydrolysis (Eastman and Ferguson 1981). Pretreatment can increase the availability of food waste to these enzymes through decreased particle size and looser physical structure. To measure the enzymatic availability of food waste, hydrolytic enzymes were includ ed in a solubilization assay A commercial hydrolytic enzyme powder was selected to represent the hydrolytic enzymes produced in an anaerobic digester. The commercial enzymes were a pplied in an excess amount in the assay (0.1 g/g food waste (ww)) so that the enzyme quantity would not be limited. This allows the difference s in substrate availability to be distinguished Twenty four Hour Solubilization The assay was conducted for 24 h with SCOD and pH measured at 0 1, 2, 4, 6, 8, 12 and 24 h (for data analysis purposes 0.1 h is the nominal time allocated for pretreatment ) Figure 3 12 present s the SCOD of intact food waste and food waste pretreated wit h meat grinder (0.5 cm plate and 1.0 cm plate) and food waste disposer. As in the endogenous solubilization assay, solubilization appeare d to plateau within 6 8 h and decrease at 12 and 24 h. The decrease was likely due to microbial assimilation of SCOD, which was suggest ed by the drop in pH (Figure 3 1 3 ), c onsumption of sugars (Figure s 3 1 4 and 3 15 ), and producti on of organic acids (Figure s 3 1 6 and 3 17 ). For the first 8 hours in the assay, glucose was approximately 50% of the SCOD of pretreated food waste, and was 10% at 12h. Formic, acet ic, and succinic acids and ethanol were at detectable ranges by 24 h, which indicated that SCOD had been consumed for acidogenesis. For further data analysis, the 12 and 24 h data points were disregarded because the bacterial assimilation of SCOD confounds any further increases in solubilization Figure 3 1 8 shows the

PAGE 63

63 mean solubilization of food waste in the commercial enzyme assay. All pretreated food waste was approximately 50% solubilized within 2 h while intact food waste was approximately 23% solubilized at 2 h By 6 h pretreated food waste reached 60% solubilization and intact food waste reached a mean solubilization of 45% by 6 h. These percentages represent ed total solubilization, which include d both endogenous SCOD released and enzymatic hydrolysis. To assess the kinetics of enzymatic hydrolysis the endogenous solubilization (Figure 3 6) from the previous experiment was subtract ed from the total solubilization in this experiment Figure 3 12 Soluble chemical oxygen demand in the commercial enzyme assay. Note: Total COD of gr inder (0.5 cm plate), gr inder (1.0 cm plate), disposer and intact food waste is 3284, 3368, 3556, 3688 mg/L, respectively. Soluble COD from commercial enzyme has been subtracted. 0 500 1000 1500 2000 2500 0 4 8 12 16 20 24 mg SCOD/L time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposal Intact

PAGE 64

64 Figure 3 1 3 Mean pH in the commercial enzyme assay. Error bars represent standard deviation. Figure 3 1 4 Sugars as a percent of SCOD for pretreated food waste (grinder 0.5 cm plate) in the c o mmercial enzyme solubilization assay Sugars from commercial enzyme have been subtracted. 6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7 0 4 8 12 16 20 24 pH time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposer Intact 0 0.5 1 1.5 2 2.5 3 0 10 20 30 40 50 60 70 80 90 100 0 4 8 12 16 20 24 % SCOD (Xylose) % SCOD (Glucose) time (h) Glucose (Grinder 0.5 cm) Xylose (Grinder 0.5 cm)

PAGE 65

65 Figure 3 1 5 Sugars as a percent of SCOD for intact food waste in the c ommercial enzyme solubilization assay Sugars from commercial enzyme have been subtracted. Figure 3 1 6 O rganic acids and ethanol as a percent of SCOD for pretreated food waste (grinder 0.5 cm plate) in the c ommercial enzyme solubilization assay Compounds from commercial enzyme have been subtracted. 0 0.5 1 1.5 2 2.5 3 0 10 20 30 40 50 60 70 80 90 100 0 4 8 12 16 20 24 % SCOD (Xylose) % SCOD (Glucose) time (h) Glucose (Intact) Xylose (Intact) 0 0.5 1 1.5 2 2.5 3 0 4 8 12 16 20 24 % SCOD time (h) Formic Acid (Grinder 0.5 cm) Acetic Acid (Grinder 0.5 cm) Succinic Acid (Grinder 0.5 cm) Ethanol (Grinder 0.5 cm)

PAGE 66

66 Figure 3 1 7 Organic acids and ethanol as a percent of SCOD for intact food waste in the c ommercial enzyme solubilization assay. Compounds from commercial enzyme have been subtracted. Fig ure 3 1 8 Mean solubilization of food waste in the commercial enzyme assay. E rror bars represent standard deviation. 0 0.5 1 1.5 2 2.5 3 0 4 8 12 16 20 24 % SCOD time (h) Formic Acid (Intact) Acetic Acid (Intact) Succinic Acid (Intact) Ethanol (Intact) 0 10 20 30 40 50 60 70 0 1 2 3 4 5 6 7 8 CODsf (%) time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposal Intact

PAGE 67

67 Enzymatic Hydrolysis Kinetics Enzymatic hydrolysis was fit to a first order kinetic equation with the kinetic rate constant (k) and final hydrolyzed soluble fraction (CODhsf f ) as fitted parameters. Estimates of parameters Table 3 2 shows estimates of the fitted parameters and CODhsf at 8 h for each treatment in the assay. All three pretreatment methods showed significantly higher rate constants than intact food waste. The grinder (0.5 cm plate) had a statistica lly lower k than either of the other two 1, all three pretreatments had a similar k. The grinder (0.5 cm plate) pretreatment and intact food waste had statistically higher CODhsf f than either of the other two pretr eatments a higher CODhsf f than the grinder (1.0 cm plate). Table 3 2. Estimated parameters and 8 h hydrolyzed COD for commercial enzyme assay a: b: Standard error Fitted curves The fitted curves with measured data p oints are presented in Figure 3 1 9 All three pretreatments show ed much greater hydrolysis rate s than intact food waste At 2 h, enzymatic hydrolysis solubilize d over 25 % of pretreated food waste and less than 15 % of intact food waste. However by 8 hours, approximately 30% of both int act and pretreated food waste w as enzymatically hydrolyzed. The increased hydrolysis rates indicated that pretreatment improve d the immediate availability for hydrolytic enzymes due to looser structure and increased surface k (h 1 ) CODhsf f (%) CODhsf 8 (%) Treatment Estimate a SE b Estimate a SE b Grinder (0.5 cm plate) 0.839 X 0.052 32.180 X 0.485 32.141 Grinder (1.0 cm plate) 1.194 Y 0.219 29.228 Y 1.024 29.226 Disposer 1.054 Y 0.130 30.548 Y 0.765 30.541 In tact 0.252 Z 0.027 32.685 X 1.648 28.331

PAGE 68

68 are a through tissue disruption and cell rupture. The pretreatment of food waste is critical i n improving the short term solubilization of food waste, from both the release of endogenous soluble material and immediate availability to hydrolytic enzymes. Figure 3 1 9 Fitted enzymatic hydrolysis curves for the commercial enzyme assay Microbial Inoc ulum Assay The commercial enzyme assay demonstrated food waste hydrolysis kinetics under conditions with excess commercial hydrolytic enzymes. To measure the hydrolysis kinetics of food waste in the presence of microorganisms and enzymes produced by these microorganisms an experiment was conducted using a microbial inoculum The purpose of this assay was to simulate the solubilization that would occur in an anaerobic digester by using an inoculum containing a hydrolytic microbial consortium. The microbi al inoculum was derived from flushed dairy manur e because it contains the microbial consortia, including hydrolytic micro organisms, necessary for anaerobic digestion. Flushed dairy manure was loaded with 10g (ww)/L of pretreated (grinder 0.5 cm plate) foo d waste and buffered with phosphate buffer (0.5 M at 6.5 0 5 10 15 20 25 30 35 40 0 1 2 3 4 5 6 7 8 CODhsf (%) time (h) Grinder (0.5 cm plate) Grinder (1.0 cm plate) Disposal Intact CODhsf=CODhsf f (1 e ( k*t) )

PAGE 69

69 pH) The i noculum was incubated for 30 days prior to the assay to allow the hydrolytic consortium to adapt to food waste as a substrate and to the conditions of the assay In the microbial inoculum assay, pretreatment was represented by the meat grinder with 0.5 cm pla te openings. The previous two assays showed there was little difference between pretreatment methods; however, t he meat grinder with the 0.5 cm plate did show the most immediate relea se of endogenous SCOD (Figure 3 10 ), which represents greate r cell and tissue destruction. Twenty four Hour Solubilization The assay was conducted for 24 h with SCOD and pH measured at 0 1, 2, 4, 6, 8, 12 and 24 h (for data analysis purposes 0.1 h is the nominal time allocated for pretreatment ) Figure 3 20 presents the SCOD of intact and pretreated food waste using the microbial inoculum Unlike the endogenous solubilization and commercial enzyme assays, solubilizat ion did not plateau or decrease at 8 h The continued increase in solubilization may have be en a result of the micro organisms assimilating less SCOD than in the other two assays. In the first two assays, which started at a low microbial population, microorganisms consumed more SCOD as they b egan to grow and reproduce. However, with the microbial inoculum there was likely less growth and reproduction, so that there was less SCOD assimilated. The presence of a microbial population throughout the entirety of the microbial inoculum assay was i ndicated by the s teady decrease in pH (Figure 3 21 ) in the first 8 h followed by pH stabilization from 8 to 24 h. In contrast, the first two assays show ed a stable pH from 0 to 8 h fo llowed by a d ecrease in pH at 12 and 24 h (Figures 3 7 and 3 13 ) Due to the decreased microbial assimilation of SCOD, solubilization was able to be determined for the full 24 h. As in the commercial enzyme assay, the tot al solubilization rep resent ed both endogenous SCOD release and e nzymatic hydrolysis (Figure 3 22 ). In t he microbial inoculum assay, total solubilization reached 52% and 3 7 % for pretreated and intact food waste, respectively. By subtracting the endogenous SCOD release

PAGE 70

70 (Figure 3 6 ) from total solubilization in this assay, the kinetic s of enzymatic hydrolysis were able to be assessed. Figure 3 20 Soluble chemical oxygen demand of food waste in the microbial inoculum assay Note: Total COD of pretreated and intact food waste is 3405 and 3705 mg/L respectively. Soluble COD of microbial inoculum has been subtracted. Figure 3 21 Mean pH for the microbial inoculum assay. Error bars represent standard deviation. 0 200 400 600 800 1000 1200 1400 1600 1800 2000 0 4 8 12 16 20 24 mg SCOD/L time (h) Grinder (0.5 cm plate) Intact 6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7 0 4 8 12 16 20 24 pH time (h) Grinder (0.5 cm plate) Intact

PAGE 71

71 Figure 3 22 Mean solubilization in microbial inoculum assay. Error bars represent standard deviation. Enzymatic Hydrolysis Kinetics A noticeable difference between the microbial inoculum assay and the commercial enzyme assay was that enzymatic hydrolysis appear ed to have biphasic kinetics For both intact and pretreated food waste, the initial hydrolysis appear ed to plateau a t 6 hours at which point secondary hydrolysis oc cur red from 6 to 24 h (Figure 3 23 ) The biphasic hydrolysis may have be en due to the microbial nature of the inoculum The first hydrolysis kinetic curve may have been a result of hydrolytic enzymes consti tutively produced by the microorganisms and were present in the inoculum at the start of the assay When the fresh food waste was added to the inoculum, the microorganisms must adapt to the new feedstock and induce further hydrolytic enzyme production This adaptation period result ed in the second kinetic curve. To determine kinetic rates initial (0 6 h) and secondary (6 24 h) hydrolysis were fit to individual first order kinetic equations with the kinetic rate constant (k) and final hydrolyzed soluble fraction (CODhsf f ) as fitted parameters. 0 10 20 30 40 50 60 0 4 8 12 16 20 24 CODsf (%) time (h) Grinder (0.5 cm plate) Intact

PAGE 72

72 Figure 3 23 Mean hydrolysis in the microbial inoculum assay. Error bars represent standard deviation. Estimates of parameters Table 3 3 shows estimates of the fitted parameters and CODhsf at 6 h for the initial hydrolysis of pretreat ed and intact food waste The rates of both food wastes were statistically similar, while pretreated food waste showed a significantly higher CODhsf f Table 3 3 Estimated parameters and 6 h hydrolyzed COD for initial hy drolysis in the microbial inoculum assay (t=0 to 6 h) a: Different letters in the same column b: Standard error Table 3 4 shows estimates of the fitted parameters and CODhsf at 24 h for the secondary hydrolysis of pretreated and intact food waste. The rates of both food wastes were stat istically similar, while pretreated food waste showed a significantly higher CODhsf f 0 5 10 15 20 25 30 0 4 8 12 16 20 24 CODhsf (%) time (h) Grinder (0.5 cm plate) Intact k (h 1 ) COD h sf f (%) COD h sf 6 Treatment Estimate a SE b Estimate a SE b Grinder (0.5 cm plate) 1.989 X 0.395 10.167 X 0.365 10.167 In tact 8.681 X 6.009 3.090 Y 0.376 3.087

PAGE 73

73 Table 3 4 Estimated parameters and 24 h hydrolysis for secondary hydrolysis in the microbial inoculum assay (t=6 to 24 h) a: b: Standard error Fitted curve s Figure 3 24 shows the fitted curves for the initial and secondary hydrolysis with measured data points Although the initial and secondary hydrolysis rates were similar, pretreated food waste show ed significantly higher extent s of hydrolysis over both hydrolysis periods. Within the first 2 h 10% of the TCOD of pretreated food waste was hydrolyzed, while only 3% of intact food waste was hydrolyzed. After 24 h, 25% of pretreated food was hydrolyzed, and 17% of intact food waste was hydrolyzed. The increased hydrolysis of pretreated food waste indicate d that pretreatment increase d the availability of the substrate to hydrolytic enzymes both in the initial hydrolysis and secondary hydrolysis. Figure 3 24 Fitted solubilization c urves for the microbial inoculum assay 0 5 10 15 20 25 30 0 4 8 12 16 20 24 CODhsf (%) time (h) Grinder (0.5 cm plate) Intact CODhsf=CODhsf f (1 e ( k*t) ) k (h 1 ) COD h sf f (%) COD h sf 24 Treatment Estimate a SE b Estimate a SE b Grinder (0.5 cm plate) 0.184 X 0.037 25.418 X 1.179 24.861 In tact 0.168 X 0.048 17.854 Y 1.802 17.131

PAGE 74

74 Discussion Figure 3 2 5 Six h solubilization of intact and pretreated (grinder 0.5 cm plate) food waste in the three solubilization assays. Table 3 5 Maximum extent of solubilization for i ntact and pretreated (grinder 0.5 cm plate) food waste in the three solubilization assays Assay Food Waste Maximum solubilization (CODsf) Time at which maximized (h) 1 st order r ate (k) (h 1 ) Endogenous Solubilization Pretreated 26.7 0.65 < 1 24.3 a Endogenous Solubilization Intact 21.0 0.22 8 0.36 a Commercial Enzyme Pretreated 60.9 1.11 6 0.8 4 b Commercial Enzyme Intact 49.1 5.52 8 0.25 4 b Microbial Inoculum Pretreated 51.9 3.30 24 0.18 4 c Microbial Inoculum Intact 36.6 1.17 24 0.1 7 c a: Kinetic rate for release of endogenous SCOD b: Kinetic rate for enzymatic hydrolysis c: Kinetic rate for secondary enzymatic hydrolysis Results from the solubilization assays indicate d that mechanical pretreatment was very effective at increasing the s olubilization kinetics of food waste. Figure 3 2 5 shows the combined solubilization curves for pretreated (grinder with 0.5 cm plate) and intact food waste in the first 6 h in all three assays Table 3 5 shows the maximum extent of solubilization measured in the solubilization assays, time at which the solubilization maximized, and the estimated first order 0 10 20 30 40 50 60 70 0 1 2 3 4 5 6 CODsf (%) time (h) Grinder (0.5 cm plate) Intact Endogenous Release Endogenous Release Commercial Enzyme Commercial Enzyme Microbial Inoculum Microbial Inoculum

PAGE 75

75 kinetic rates. Mechanical pretreatment cause d immediate release of the endogenous solub le material within the food waste and increase d food waste availability for hydrolytic enzymes. At 6 h, 28% of pretreated food waste is solubilized through the release of endogenous SCOD alone, while the addition of commercial enzymes increased solubiliza tion to 60% (32% solubilization through enzymatic hydrolysis) Intact food waste was only 19% solubilized at 6 h th r ough endogenous SCOD release and this release did not occur immediately as with pretreated food waste. The addition of commercial enzymes to intact food waste increased solubilization to 44% (25% solubilization through enzymatic hydrolysis) Using a microbial inoculum pretreated and intact food waste reached 52% and 3 7 % solubilization by 24 h (Table 3 5) U sing the natural microbial cons ortia p resent in an anaerobic digester, pretreated food waste is nearly as solubilized within 24 hours as it is within 6 h using excess commercial hydrolytic enzymes. Additionally the inoculum required an adaptation period due to the experimental conditio ns of the assay ; this period would be negligible in an active, continuously fed anaerobic digester. Therefore, with practical mechanical pretreatment, the hydrolytic enzymes produced by the microbial consortia are sufficiently effective to nullify any g ain in solubilization by the addition of commercial hydrolytic enzymes. A critical difference between intact and pretreated was the release of endogenous SCOD immediately following pretreatment (0.1 h in present study). Pretreatment immediately release d the entire endogenous SCOD of the food waste (25% CODsf), while intact foo d waste was only 4% solubilized (Figure 3 10 ). The initial solubilization of pretreated food waste in the present study was comparable to literature values using more intensive pre treatment methods. Table 3 6 shows the values for the initial food waste solubilization calculated from literature studies. It is important to note that all the literature studies used pretreated food waste (i.e. shredded or

PAGE 76

76 ground) as a control and that a significant amount of endogenous SCOD had already been released prior to the pretreatment used in the studies. Izumi et al. (2010) which found an immediate solubilization of 28% using a food disposer, increased initial solubilization to 3 9 % with 300 revol utions (1 minute at 300 rpm) of a ball mill. However, using a ball at 40,000 revolutions (20 minutes at 2000 rpm) initial solubilization was still only 40% solubilized. This showed that minimal ball milling released the full complement of endogeno us SCOD and the majority of which was released through the control pretreatment using a food disposer. Other s tudies showed similar initial solubilization of pretreated food waste 17% to 3 2 % through thermal and freezing/thawing pretreatment while contr ol (shredded) food waste was 8 % to 17 % solubilized ( Liu et al., 2008 ; Wang et al., 2006 ) However, some of this increase may ha ve been due to thermal hydrolysis when heating food waste in addition to released endogenous SCOD. D ifferent types of food wastes have different endogenous SCOD; therefore it is important to consider the increase of pretreated food waste over controls. Because the literature values were using shredded or ground food waste as a control the gain in initial solubilization through additional pretreatment is much less than the gain achieved over intact food waste. The additional pretreatment using energy in tensive methods only show ed marginal gains in endogenous SCOD release compared to the large gain using p ractical pretreatment ( shredding, disposer, or grinder ). Also, the endogenous SCOD of the food waste in the present study was measured at 25%, a ll of w hich is released immediately through pretreatment with the grinder (0.5 cm plate). Therefore, additional intensive pretreatment would not release additional endogenous SCOD. to hydrolytic enzymes when incubated with a microbial inoculum for 24 h. P retreated and intact

PAGE 77

77 food waste was 5 2 % and 37%, respectively solubilized within 24 h. These values were within the range found in literature that measured solubilization of pretreated food waste in acidogenic reactor s Table 3 7 shows the calculated literature values for 24 h solubilization in an acidogenic reactor. Wang et al. (2006) thermally pretreated food waste by heating to 70C for 2 h or 150C for 1 h. In 24 h co ntrol food waste (shredded) 70C pretreatment and 150C pretreatment w ere 31 %, 37 %, and 57 % solubilized respectively Liu et al. (2008) a l s o examined thermally pretreated food waste (150C for 1 h), as well as frozen/thawed food waste. Control and thermally pretreated food waste were 17% and 31% respectively, solubilized within 24 h, and control and frozen/thawed food waste were, 24% and 33% respectively solubilized by 24h. Kim et al. (2005) used a hydrolytic enzyme cocktail at a rate o f 0.2 % (v /v), for food waste pretreatment in an acidogenic reactor. Control and pretreated food waste showed a 24 h CODsf of 37.5% and 52.5%, respectively. However, the present study found a similar 24 h solubilization (51.9%) without the addition of commercial e nzymes using practical pretreatment and the microbial consortia present in an anaerobic digester. The values in the present studies, while similar to literature values, were achieved without the use of expensive or energy intensive pretreatments. Therefo re, it can be concluded that practical mechanical pretreatment methods are as effective as the intensive pretreatment methods in literature at increasing the release of endogenous soluble organic material and the enzymatic hydrolysis of food waste.

PAGE 78

78 Table 3 6 Initial solubilization of food waste in pretreatment studies Control Pretreatment Control Pretreat ed Increase in Source of food waste treatment method CODsf (%) CODsf (%) CODsf (%) a Source University dining shredded 2 h @ 70 o C 17.2 20.3 3.1 Wang et al. 2006 University dining shredded 1 h @ 150 o C 17.2 31.7 1 4.5 Wang et al. 2006 University dining shredded 1 h @ 150 o C 7.8 16.9 9.1 Liu et al. 2008 University dining shredded frozen 24 hr @ 20 o C 14.9 25.4 10.5 Liu et al. 2008 S tandard food waste food disposer ball mill (300 revolutions) 28.1 39.0 10.9 Izumi et al.2010 S tandard food waste food disposer ball mill (40,000 revolutions) 28.1 40.3 12.3 Izumi et al.2010 S tandard food waste i ntact meat grinder 0.5 cm plate 4.3 25.3 21.0 Current study S tandard food waste i ntact meat g r inder 1.0 cm plate 4.3 19.6 15.3 Current study S tandard food waste i ntact food disposer 4.3 21.1 16.8 Current study a: Increase is di fference between pretreated CODsf and control COD sf Note: Values are calculated from the respective literature Table 3 7 One d ay solubilization of food waste in pretreatment studies Control Pretreatment Loading rate Solubilization 1 d CODsf (%) Increase in Source Source of food waste treatment method (g COD/L) a temperature Control Pretreated CODsf (%) b University dining shredded 2 h @ 70C 42.1 35C 30.9 36.9 6.0 Wang et al. 2006 University dining shredded 1 h @ 150C 42.1 35C 30.9 57.1 26.2 Wang et al. 2006 University dining shredded 1 h @ 150C 45.1 35C 16.6 31.1 7.2 Liu et al. 2008 University dining shredded f rozen 24 h @ 20C 54.3 35C 23.9 33.1 9.2 Liu et al. 2008 University dining blended e nzyme cocktail 20.0 35C 37.5 52.5 15.0 Kim et al. 2005 S tandard food waste i ntact meat grinder 0.5 cm plate 3.5 35C 3 6.6 5 1.9 15. 3 Current study a: Loading rates in literature studies are for the acidogenic reactor in a two phase system b: Increase is difference between pr etreated COD sf and control CODsf Note: Values are calculated from the respective literature

PAGE 79

79 Summary and Conclusions Three solubilization assays were conducted to compare the solubilization kinetics of intact food waste to mechanically pretreated food waste The three assays measured 1) the release of endoge nous SCOD 2) enzymatic hydrolysis using commercial enzymes and 3) a microbial inoculum Pretreatment of food waste using an in sink food disposer and meat grinder (with either 1.0 or 0.5 cm plate openings) resulted in significantly higher solubilization kinetics than intact food waste. The release of endogenous SCOD occurred immediately in pretreated food waste, while endogenous SCOD from intact food waste was released at a rate of 0.36 h 1 Intact food waste was only 4.1% solubilized initially, while pretreated food waste was 19% to 25% immediately following pretreatment The application of c ommercial enzyme s allowed the enzymatic hydrolysis of food waste to be measured. Within 8 h, 61% and 49%, respectively, of intact and pretreated food waste w ere solubilize d, which corresponded to 30 to 32% enzymatic hydrolysis for both food waste s Although the extent of enzymatic hydrolysis between intact and pretreated food waste were similar within 8 h, the rates of hydrolysis were significantly higher for pretreat ed fo od waste (0.8 1.2 h 1 ) than intact food waste (0.3 h 1 ). This indicated an increase in the immediate availability of the pretreated food waste to hydrolytic enzymes due to looser physical structure and reduced particle size A microbial inoculum also showed increased solubilization of pretreated food waste over intact food waste. Within 24 h of incubatio n with the microbial inoculum 52 % of pretreated food waste and 37% of in tact food waste was solubilized. Pretreated f ood waste was nearly as s olubilized in 24 h with a microbial inoculum, as in 8 h with excess commercial hydrolytic enzymes In an active, continuously fed anaerobic digester the microbial hydrolysis would show more rapid kinetics due to the lack of the adaptation period that was present in the assay. This indicated that the hydrolytic enzymes produced by microbial consortia could sufficiently solubilize pretreated food waste without the

PAGE 80

80 addition of expensive, commercial enzymes The three solubilization assays showed that the u se of practical mechanical pretreatment methods, through disrupting tissue and rupturing cells, greatly enhance d the short term solubilization kinetics of food waste.

PAGE 81

81 CHAPTER 4 EFFECT OF MECHANICAL PRETREATMENT ON BIO METHAN ATION OF FOOD WASTE Anaerobic digestion is a sequential metabolism that is performed by mixed microbial consortia Particulate feedstocks, such as food waste, must first be solubilized through enzymatic hydrolysis The soluble organic material is then fermented into organic acids, which are ultimately metabolized into acetic acid. M ethanogens consume acetic acid and produce met hane as a metabolic byproduct. For the anaerobic digestion of high particulate feedstocks the rate limiting step is considered to be hydrolysis ( Eastman & Ferguson, 1981 ; Izumi et al., 2010 ; Palmowski & Muller, 2003 ; Wang et al., 2006 ) As the first step in the process, rate limitation by hydrolysis can hinder the entire digestion process which lead s to reduced biogas production and an increased reactor size needed to treat a given volume of waste. Therefore, increasing the rate of hydrolysis and solubilization through pretreatment of food waste is proposed as a method to increase the overall efficiency of food waste digestion. Literature studies have indicated that various pretreatment meth ods enhance the solubilization and subsequent methane production kinetics of food waste ( Izumi et al., 2010 ; Liu et al., 2008 ; Wang et al., 2006 ) The pretreatment methods reported in literature, however, are energy and resource intensive and are not economically feasible for widespread implementation of food waste anaerobic digestion. The present study assesse d the enhancement of anaerobic digestion using low tech, practical pretreatment methods. Pretreating food waste with a m anual meat grinder and in sink dispo ser was shown to significant ly increase food waste solubilization compared with intact food waste. Therefore, it was hypothesized that increased solubilization of pretreated food waste provides increased substrate avail able for methanogenesis which can potentially increase biomethanation kinetics. To test this hypothesis, biochemical methane potential (BMP) assay s were conducted on pretreated and intact food waste.

PAGE 82

82 The BMP assay is used as a laboratory scale method for measuring the methane production kinetics of a feedstock under batch loading conditions ( Owen et al., 1979 ) The assay utilizes a mixed microbial inoculum from a methanogenic source, which contains the required consortia of microorganisms necessary for hydrolysis, acidogenesis, acetogenesis and methanogenesis to occur. As the microbial consortia digest the feedstock methane production is measured over time, which is used to calculate the degradation rate of the feedstock. This measurement allows both the rate and extent of feedstock degradation to be determined under simulated anaerobic digester conditions. A load ing rate of 2 to 4 g COD/L is typically used in the BMP assay because at higher loading rates, acidification can occur which lowers the pH in the assay and inhibits methanogenesis. Because the BMP assay is a batch digestion process, proper inoculum select ion is critical for optimum results. The assay is designed with excess nutrients, alkalinity, and inoculum so that under proper loading rates the only limiting factor is the degradation rate of the feedstock itself. For this reason, t he BMP assay was uti lized in the present study to compare the methane production kinetics of intact food waste to pretreated food waste. Moderate Loading Rate Biochemical Methane Potential Assay Results from three solubilization assay indicated that pretreated food waste had significantly higher solubilization kinetics than intact food wa ste (Chapter 3). Therefore a BMP assay was conducted at the same loading rate as the solubilization assays, 10 g (ww)/L, which corresponded to a nominal COD loading rate of 3.5 g COD/L a mo derate loading rate for the typical range of the BMP assay In the BMP assay, pretreated food waste was represen ted by the meat grinder (0.5 cm plate) pretreatment method In the solubilization assays (Chapter 3), all three practical pretreatment methods showed approximately similar increases in food waste solubilization. The meat grinder with 0.5 cm plate openings was selected as the pretreatment method for the BMP assay b ecause this method produced more homogenous pretreated food

PAGE 83

83 waste for obtaining representative samples, which is critical for the BMP assay due to the small mass of substrate used ( Wilkie et al., 2004 ) Flushed dairy manure (100% v/v) was used as an inoculum for the assay because it contains the mixed microbial consortia and sufficient nutrients and alkalinity needed for anaerobic digestion ( Wilkie, 2005 ) The assay included a glucose and cellulose control loaded at 2 g/L (2.13 g COD/L) to measure the activity of the inoculum. An inoculum blank was also measured in the assay to account for any meth ane production from the COD of t he inoculum. Cumulative Methane Production Cumulative methane production was measured for 30 d in the assay. M ethan e production measurements were normalized to per g COD loaded and STP conditions after subtracting methane production from the inoculum blank. Cumulative methane production was modeled to a first order kinetic equation with the kinetic rate constant (k) and ultimate cumulative methane yield (CH 4f ) as f itted parameters. Figures 4 1 and 4 2 show the fitted model against measured data points for food waste and controls, respectively. Table 4 1 shows estimates of fitted parameters for the kinetic models for intact and pretreated food waste and glucose and cellulose controls. Intact food waste showed a significantly higher rate (k) and extent (CH 4f ) of methane production than pretreated food waste at Table 4 2 shows t he percent COD removal at 5, 10, 20, and 30 d based on the stoichiometric COD equivalent of methane ( 2.86 g COD/L CH 4 @ STP ). The assays showed that by 20 d, over 90% of intact food waste was converted to methane while 81% of pretreated waste was converted at 2 0 d and by 30 days only 84% of pretreated food waste had been converted Table 4 3 shows the total COD, soluble COD, pH and conductivity measured on the digestate at the

PAGE 84

84 end of the assay. After 3 0 days, there was more unconverted TCOD fro m pretreated foo d waste than intact food waste, as there was less methane produced from the pretreated food waste. Figure 4 1. Cumulative methane production of intact and pretreated food waste in the moderate loading rate BMP assay Pretreatment is meat grinder (0.5 cm plate) Figure 4 2. Cumulative methane production from glucose and cellulose controls in the moderate loading rate BMP assay 0 50 100 150 200 250 300 350 400 0 5 10 15 20 25 30 Cumulative CH 4 (mL/g COD @ STP) day Grinder (0.5 cm plate) Intact CH 4 = CH 4f (1 e ( k*t) ) 0 50 100 150 200 250 300 350 0 5 10 15 20 25 30 Cumulative CH 4 (mL/g COD @ STP) day Glucose Cellulose CH 4 = CH 4f (1 e ( k*t) )

PAGE 85

85 Table 4 1. Estimates of parameters for methane production kinetics in the moderate loadi ng rate BMP assay. k (day 1 ) CH 4f (mL/g COD added) Mean a S tandard error Mean a S tandard error Pretreated 0.122 W 0.0022 312.32 W 2.23 Intact 0.128 X 0.0022 346.25 X 2.28 Glucose control 0.221 Y 0.0036 291.44 Y 1.40 Cellulose control 0.182 Z 0.0002 270.60 Z 0.25 Table 4 2 Calculated COD removal at 5 1 0 20 and 30 days in the moderate loading rate BMP assay % COD removal a Day 5 Day 10 Day 20 Day 30 Pretreated 36.6 1.57 64.9 1.57 81.2 2.83 83.8 3.19 Intact 41.6 0.47 74.2 1.28 91.1 0.38 92.8 0.11 Glucose control 54.2 1.11 78.7 1.08 80.6 0.56 81.1 0.37 Cellulose control 20.4 0.71 56.0 1.57 72.5 0.92 72.5 0.96 a: based on COD equivalent of cumulative methane production Note: M ean values one standard deviation Table 4 3 Total COD, SCOD, pH, and conductivity after 3 0 days of digestion in the moderate loading rate BMP assay TCOD SCOD pH Conductivity (g/L) (g/L) (m S /cm) Pretreated 2.64 0.09 0.488 0.005 6.96 0.01 3.50 0.10 Intact 2.40 0.06 0.482 0.004 6.94 0.01 2.93 0.27 Glucose c ontrol 2.65 0.05 0.488 0.006 6.86 0.01 3.14 0.04 Cellulose c ontrol 2.93 0.17 0.596 0.015 6.84 0.02 3.15 0.32 Inoculum b lank 2.29 0.05 0.379 0.009 7.04 0.02 3.00 0.03 Note: M ean values one standard deviation The results of this assay indicated that, at a loading rate of 3.5 g COD/L, pretreated food waste did not have higher methanogenic kinetics than intact food waste It is likely that the lo ading rate exce eded the capacity of the BMP assay Exceeding the loading rate would result in increased acidogenesis, which would lower the pH and inhibit methanogenesis. The high solubili zation o f pretreated food waste can exacerbate the problem due to more rapid

PAGE 86

86 acidogenesis than intact food waste. Despite a neutral pH by 30 d (Table 4 3 ), it is likely that a period of reduced pH at inhibitory levels occurred in the assay for pretreated food was te. Loading Rate and pH Results from t he BMP assay indicated that a moderate loading rate o f 3.5 g COD/L may have caused inhibited biomethanation of pretreated food waste due to acidification Therefore for further study of food waste biomethanation it was critical to determine the effect of loading rate on pH. To assess this effect, a simulated BMP assay was conducted at loading rates of 2 4 and 8 g COD/L for intact and pretreated food waste. To measure pH, 10 mL subsamples w ere tak en from each bottle over 20 d. Figure 4 3. Mean pH in simulated BMP assay s loaded at 2, 4, and 8 g COD/L with intact and pretreated food waste. Error bars represent standard deviation. Figure 4 3 shows the mea n pH for intact and pretreated food waste at each loading rate. Within 2 days a t a loading rate of 4 g COD/L pH drop ped below 6.5 which is considered the minimum op timum pH for methanogenesis ( Speece, 2008 ) for both intact and pretreated food waste. Pretreated food waste, however, show ed a faster pH decrease and remain ed below the pH 5 5.5 6 6.5 7 7.5 8 0 2 4 6 8 10 12 14 16 18 20 pH day Intact 2g/L Pretreated 2g/L Intact 4g/L Pretreated 4g/L Intact 8g/L Pretreated 8g/L

PAGE 87

87 of intact food waste for most of the 20 d period which suggest ed that the enhanced solubilization of pretreated food waste resu lt ed in increased acidogenesis. Increased acidogenesis can inhibit methanogenesis if the methanogens cannot consume the increased organic acids and the acids ac cumulate. This is clearly exacerbated at a loading rate of 8 g COD/L which is far above the capacity of the BMP assay, where pH dropped to nearly 5.5 and recovered to only 6.5 within 20 d. At a loading rate of 2 g COD/L pH did not drop below 6.5 and intact and pretreated food waste followed approximately the same pH drop and recovery. Therefore acidification would not becom e inhibitory for methanogenesis at this reduced loading rate. Reduced Loading Rate Biochemical Methane Potential Assay To lessen the impact of acidification on methanogenesis in the BMP assay, the loading rate was reduced to 2 g COD/L. The assay compared pretreated (grinder 0.5 cm plate) against intact food waste with glucose and cellulose controls. The inoculum used in this ass ay was the conserved digestate from the moderate loading rate BMP. The conserved inoculum had lower total and soluble COD compared with fresh flushed dairy manure, which further reduce d the potential for acidification through COD overloading. Cumulative Methane Production Methane production was measured for 30 days in the assay. All methane production measurements were normalized to per g COD loaded and STP conditions after subtracting methane production from the inoculum blank. Cumulative methane produ ction was modeled to a first order kinetic equation with the kinetic rate constant (k) and ultimate cumulative methane yield (CH 4f ) as fitted parameters. Figures 4 4 and 4 5 show the fitted model against measured data points for food waste and controls, respectively. Table 4 4 shows estimates of fitted parameters for the kinetic models for intact and pretreated food waste and glucose and cellulose controls. With the reduced loading

PAGE 88

88 rate, pretreated food waste had a significantly higher kinetic rates (0. 21 d 1 ) than intact food waste (0.20 d 1 ) at similar waste also showed a statistically higher COD ; however, this may be an effect of the model because the measured cumulative methane yield s at 30 d are the same for both food wastes (Figure 4 5) Tables 4 5 shows the percent COD removal at 5, 10, 20, and 30 d based on the stoichiometric COD equivalent of methane ( 2.86 g COD /L CH 4 @ STP). Both food waste s are approximately 60% convert ed to methane by 5 d and approximately 90% convert ed by 20 d. Table 4 6 shows the total COD, soluble COD, pH, conductivity and alkalinity measured on the digestate at the end of the assay. After 30 days of digestion, digestate from intact and pretreated food waste ha d similar TCOD, SCOD, pH, and alkalinity. Therefore a t a loading rate of 2 g COD/L, both intact and pretreated food waste ha d similar methane production kinetics under the conditions of the BMP assay. Figure 4 4 Cumulative methane production of intact and pretreated food waste in the reduced loading rate BMP assay Pretreatment is meat grinder (0.5 cm plate) 0 50 100 150 200 250 300 350 0 5 10 15 20 25 30 Cumulative CH 4 (mL/g COD @ STP) day Pretreated Intact CH 4 = CH 4f (1 e ( k*t) )

PAGE 89

89 Figure 4 5 Cumulative methane production of glucose and cellulose controls in the reduced loading rate BMP a ssay Table 4 4 Estimates of parameters for methane production kinetics in the reduced loading rate BMP assay. k (day 1 ) CH 4f (mL/g COD added) Mean a Standard error Mean a Standard error Pretreated 0.205 W 0.0039 332.119 X 2.53 Intact 0.197 X 0.0037 323.028 Y 2.46 Glucose control 0.218 Y 0.0044 304.899 Z 2.31 Cellulose control 0.177 Z 0.0040 319.076 Y 3.05 Table 4 5 Calculated COD removal at 5 1 0 20 and 30 days in the reduced loading rate BMP assay COD removal (%) a Day 5 Day 10 Day 20 Day 30 Pretreated 61.7 0.18 84.0 1.07 90. 6 1.89 90.6 1.89 Intact 58.2 0.79 80.6 1.88 89.4 2.60 91.4 2.06 Glucose control 56.4 1.84 81.0 1.78 83. 6 1.61 83.6 1.61 Cellulose control 21.7 0.95 67.6 0.62 84.4 0.48 86.6 1.45 a: based on COD equivalent of cumulative methane production Note: M ean values one standard deviation 0 50 100 150 200 250 300 350 0 5 10 15 20 25 30 Cumulative CH 4 (mL/g COD @ STP) day Glucose Cellulose CH 4 = CH 4f (1 e ( k*t) )

PAGE 90

90 Table 4 6 T otal COD, soluble COD, pH, conductivity and alkalinity after 3 0 days of digestion in the reduced loading rate BMP assay T otal COD S oluble COD pH Conductivity Alkalinity (g/L) (g/L) (m S /cm) (mg CaCO 3 eq./L) Pretreated 2.63 0.06 0.388 0.017 6.97 0.02 3.81 0.09 2600 25.0 Intact 2.56 0.06 0.385 0.002 6.96 0.01 3.65 0.04 2568 12.0 Glucose c ontrol 2.63 0.04 0.393 0.012 6.90 0.01 3.37 0.02 2344 13.9 Cellulose c ontrol 2.80 0.15 0.513 0.013 6.88 0.01 3.50 0.04 2312 79.9 Inoculum b lank 2.33 0.09 0.370 0.024 7.22 0.01 3.37 0.02 2384 6.9 Note: M ean values one standard deviation Discussion Compared to the moderate loading rate BMP assay, the methane production kinetics of both intact and pretreated food waste were greater at the reduced loading rate Figure 4 6 compares the cumulative methane production of food waste in the two assays. Over the first 10 d, methan e production was greater at the reduced l oading than in the moderate loading rate assay The decreased methane pr oduction correlate d to decreased pH at a moderate loading rate (Figure 4 3 ). In the simulated BMP assay, pH remained below 7 for the first 10 d at a loading rate of 4 g COD/L. After 10 d, acid ification from the intact food was overcome and cumulative methane production was approximately equal to that of intact and pretreated food waste loaded at 2 g COD/L. However methane production and pH of pretreated food waste remain ed below that of intact food waste after 10 days at the moderate loading rate These two assays demonstrate d the importance of maintaining proper pH in an anaerobic digester especially within the initial stages of digestion In fact in a low rate digester, such as in the BMP assay, the pH appear ed to be mor e critical to efficient anaerobic digestion than increas ing solubilization through pretreatment. At a high loading rate, pretreatment can reduce methane production, while at a low loading rate, pretreatment does not appear to increase methane kinetics. T herefore, to increase the efficiency of methane production from pretreated food waste, it is necessary to utilize high rate anaerobic digestion, such as a fixed film reactor, to accommodate the higher rate of

PAGE 91

91 solubilization and acidogenesis by facilitating a greater metabolic flux of the acidogenic products A high rate digester can thus maintain neutral pH conditions due to more rapid consumption of organic acids by the methanogenic consortia. Figure 4 6 Cumulative methane production of pretreated and intact food waste in both BMP assay s The moderate loading rate and reduced loading rate BMP assays in this study indicate d that food waste is a readily degradable substrate for anaerobic digestion. At a proper loading rate (2 g COD/L for the BMP assay), food waste is more than 80% converted to methane with in 10 d and 90% convert ed within 20 d (Table 4 5) Table 4 7 lists literature values for 10 d cumulative methane yield in studies examining food waste pretreatment. The present study found the highest 10 d methane yields for control food waste at 282 mL/g COD and the second highest for pretreated food waste, at 294 mL/g COD. Because the 10 d methane yield of intact food waste is so great, the BMP assay is not able to show an increase for pretr eated food waste. The methanogenic rate of the BMP assay cannot accommodate the increased solubilization of pretreated food waste. 0 50 100 150 200 250 300 350 400 0 5 10 15 20 25 30 Cumulative CH 4 (mL/g COD @ STP) day Moderate Loading Rate BMP Reduced Loading Rate BMP Pretreated Pretreated Intact Intact

PAGE 92

92 The methane yield from the control is particularly interesting because the intact control in the present study showed greater methanogenesis than pretreated (shredded/ ground) controls in the literature studies. Izumi et al. ( 2010 ) used a loading rate of 10 g COD/L and produced a 10 d methane production of 234 mL /g COD for the control food waste. The lower methane production than the present study may be an effect of the food waste or inoculum used in the experiment as the pH remained neutral throughout the 1 0 d even at a loading rate of 10 g COD/L. The authors were able to increas e 10 d methane production to 320 mL /g COD using a ba ll mill pretreatment for 1000 revolutions (2 minutes at 500 rpm) However with further ball milling for 40,000 revolutions (20 minutes at 2000 rpm) methane production was reduced to 252 mL /g COD which was concluded to be caused by acidification and inhibitory VFA levels due to excessive pretreatment and solubilization. These results support the findi ngs in the present study that pretreatment of food waste can cause acidification through excessive solubilization at high loading rates. To overcome the dilemma of acidificat ion through pretreatment, an anaerobic digester design that can handle high concen trations of organic acids is necessary. Wang et al ( 2006 ) and Liu et al. ( 2008 ) examined the biomethanation kinetics of pretreated food waste using a hybrid anaerobic solid liquid (HASL) system. The HASL system is a two phase digester th at employs an acidogenic reactor and a modified upflow anaerobic sludge blanket as a methanogenic reactor ( Liu et al., 2010 ) It was designed for more optimum digestion of food waste. The ability to handle a higher loading rate, allows the HASL system to show improved biomethanation kinetics of pretreated food waste. Using the HASL system, Wang et al ( 2006 ) increaed 10 d methane yields by 16 and 40 mL/g COD for thermally pretreated food waste over control and Liu et al. ( 2008 ) increased 10 d methane y ields by 48 and 33 mL/g COD, respectively, for thermally

PAGE 93

93 pretreated and frozen/thawed food waste over control. These studies suggest that by using a high rate anaerobic digester, such as the HASL or a fixed film reactor, pretreatment of food waste can inc rease biomethanation kinetics.

PAGE 94

94 Table 4 7 Ten day cumulative methane production of food waste in pretreatment studies. Control Pretreatment Loading rate Digestion 10 day cumulative CH 4 Increase b Source Food waste source treatment method (g COD/L) temperature (mL/g COD ) (mL/g COD ) Control Pretreated University dining shredded 2 h @ 70C <10 a 35C 198.1 214.0 15.9 Wang et al. 2006 University dining shredded 1 h @ 150C <10 a 35C 198.1 237.8 39.7 Wang et al. 2006 University dining shredded 1 h @ 150C <10 a 35C 206.0 253.5 47.5 Liu et al. 2008 University dining shredded f rozen 24 h @ 20C <10 a 35C 197.3 230.2 32.9 Liu et al. 2008 S tandard food waste food disposer ball mill (1000 revolutions) 10 37C 234.2 320.0 85.8 Izumi et al.2010 Standard food waste food disposer ball mill (40,000 revolutions) 10 37C 234.2 251.6 17.4 Izumi et al.2010 Standard food waste intact grinder 0.5 cm plate 2 35C 282.1 293.9 1 1 8 Current study a: The dige ster used was a two phase system Effluent from the methanogenic reactor was used to dilute the leachate from the acidogenic reactor to maintain proper organic loading rate. b: Increase is difference between pretreated cumulative CH 4 and control cumulative CH 4 Note: Values are calculated from the respective literature

PAGE 95

95 Summary and Conclusions Two b iochemical methane production assays were performed on intact and pretreated (grinder with 0.5 c m plate) food waste. The first assay was loaded at a moderate loading rate ( 3.5 g COD/L ) using flushed dairy manure as an inoculum, while the second assay was loaded at a reduced loading rate ( 2 g COD/L ) using the conserved digestate from the initial assay as an inoculum. The extent of degradation and first order rate constants for the assays were estimated by non linear fitting of parameter estimates on triplicate data sets. In both assays pretreated and intact food waste showed approximately similar first order rate constants: 0.12 d 1 at 3.5 g COD/L and 0.20 d 1 at 2 g COD/L. At a loading rate of 3.5 g COD/L, intact food waste showed a greate r 30 d methane yields (325 mL/g COD) than pretreated food waste (294 mL/g COD). While at 2 g COD/L, both intact and pretreated showed approximately similar 30 d met hane yields (320 g COD/L). T he decreased kinetics at the higher loading rate were attributed to acidification through inc reased acidogenesis. This was confirme d by measuring pH in a simulated BMP assay, where intact food waste at 4 g/L decreased in pH to below 6.5 in the first 3 d. Pretreated food waste showed a more depressed pH at the moderate loading rate, which suggests that pretreatment at excessive loading rates can inhibit methanogenesis through acidification. While the reduced loading rate (2 g COD/L) BMP assay showed increased methane production over the moderate loading rate (3.5 g COD/L) assay, pretreated food waste did not show increased methane production over intact food waste. Therefore, it can be co ncluded that under the conditions of the BMP assay, pretreatment of food waste does not enhance the rate of biomethanation of food waste. In fact, at higher loading rates, pretreatment can be detrimental to methane production due to increased acidificatio n. These results indicated that in the BMP assay s methanogenesis was the rate limiting step rather than hydrolysis. This is elucidated

PAGE 96

96 when comparing the solubilization assays to the BMP assays. Within 24 h using a microbial inoculum over 50% of pretr eated food waste is solubilized; however in the reduced loading rate BMP assay, 4 days are required for 50% COD removal through methanogenesis Increasing the methanogenic rate of a digester is, therefore, necessary in order to accommodate the increased s olubilization of pretreated food waste through a greater metabolic flux of acidogenic products Using a digester with a high methanogenic rate, such as a fixed film reactor can facilitate optimum anaerobic digestion of food waste The retained microbial biomass in a fixed film digester has a much greater methanogenic population, which can more quickly metabolize organic acids into methane. In a fully optimized system, methanogenesis would consume organics acids as they are produced. This would also all ow the digester to operate at a higher loading rate because orga nic acids would not accumulate and cause acidification, as they did at moderate loading rates in the BMP assay. Experimentation with high rate anaerobic digest ion is necessary to verify the h ypothesis that the increased solubilization from practical food waste pretreatment can result in enhanced biomethanation kinetics

PAGE 97

97 CHAPTER 5 CONCLUSIONS The anaerobic digestion of food waste represents an opportunity for the diversion of food waste from landfills for the production of renewable energy and biofertilizer. Currently food waste is not anaerobically digested as a sole feedstock in any commercial scal e anaerobic digestion facility One limitation is the rate of hydrolysis of food waste which is considered the slowest step in the anaerobic digestion of s olid material. Hydrolysis in anaerobic digestion is largely facilitated by extracellular hydrolytic enzymes produced by the microbial consortia. These extracellular enzymes function th rough surface area contact with the particulate substrate Food waste, due to its high particulate nature has a low surface area to volume ratio which impedes the rate of hydrolysis of the material. Therefore, by increasing the surface area and enhancin g substrate availability for hydrolytic enzymes, the solubilization rate of food waste can be increase d Through increasing the surface area of food waste, tissue disruption and cell rupture occur, which release s endogenous soluble organic material and th ereby enhance s solubilization. Increased solubilization can increase the overall rate of food wa ste digestion by overcoming the rate limitation of hydrolysis To increase the solubilization rate of food waste, several methods of food waste pretreatment h ave been reported in the literature and have shown a positive correlation between pretreatment and increased solubilization and biomethanation kinetics. However, as these studies are at laboratory scale, the pretreatment methods employed were not practical for commercial scale digestion. Therefore, the current study examined the effect s on anaerobic digestion kinetics of practical pretreatment methods (food disposer and meat grinder) that could be employed commercially. The two primary objec tives of this study were to examine the impact s of pretreatment on the 1) solubilization kinetics and 2) biomethanation kinetics of food waste

PAGE 98

98 Solubilization Kinetics A series of solubilization assays were conducted comparing the solubilization rates of pretreated and intact food waste. These experiments demonstrated that pretreatment has two principal impacts on the solubilization of food waste First, endogenous SCOD was released immediately through pretreatment while endogenous SCOD was released at a first order kinetic rate of 0.36 h 1 from intact food waste Second, pretreat ment increased the rate of hydrolysis of food waste through increasing the substrate availability to hydrolytic enzymes Using a commercial enzyme, the first order kinetic rate constant were 0.84 to 1.19 h 1 for pretreated food waste hydrolysis and 0.25 h 1 for intact food waste hydrolysis By 8 h pretre ated food waste was 60% solubilized (28% through endogenous SCOD release and 32% through enzymatic hydrolysis). Intact fo od waste was 49% solubilized (21% though endogenous SCOD release and 28% through enzyme hydrolysis) in 8 h. A microbial inoculum was used in the solubilization assay as a surrogate for the hydrolytic microbial consor tium in an anaerobic digester. Within 24 h incubation with the inoculum, 52% and 37% respectively, of pretreated and intact food waste was solubilized. The solubilization kinetics using the inoculum also appeared to be biphasic with a se cond kinetic curve at 6 h. The biphasic kinetics may b e due to an adaptation period required for the microbial inoculum to adapt to the fresh food waste in the assay. In both the initial and secondary hydrolysis phases, pretreated food waste showed a greater extent of enzymatic hydrolysis. This assay indica ted that pretreated food waste was nearly as solubilized in 24 h with a microbial inoculum, as in 8 h with excess commercial hydrolytic enzymes. In an active, continuously fed anaerobic digester microbial enzymatic hydrolysis would not have the adaptatio n period observed in the assay and would show more rapid hydrolysis kinetics with the effect that hydrolytic enzymes produced by the microbial consortia could sufficiently solubilize pretreated food waste without the addition of expensive, commercial enzym es The

PAGE 99

99 solubilization assays in the present study indicate d that pretreated food waste, using low tech, practical methods, is a readily soluble substrate. The results of the assays support the hypothesis that practical, mechanical pretreatment can enhan ce food waste solubilization Enhanced solubilization was accomplished through the disruption of tissue and rupture of cells, which both release d endogen ous soluble material and exposed surface area for hydrolytic enzymes. Biomethanation Kinetics Two biochemical methane potential (BMP) assays were conducted to compare the biomethanation kinetics of pretreated food waste with intact food waste. The first assay was conducted at a moderate loading rate of 3.5 g COD/L and the second at a reduced loadi ng rate of 2 g COD/L. Both assays showed similar first order rate constants for pretreated and intact food waste : 0.12 d 1 i n the first BMP and 0. 20 d 1 in the second BMP Intact and pretreated food waste at the reduced loading rate and intact food waste at the moderate loading rate showed similar 30 d extents of methane production at approximately 320 mL CH 4 /g COD at STP or 91% COD removal. Pretreated food waste at the moderate rate, however, showed reduced 30 d methane production at 295 mL CH 4 /g COD at STP or 83% COD removal. The reduced methanogenic kinetic rates at the moderate loading rate w ere due to acidification that overwhelmed the methanogenic population in the BMP assay. This was measured through conducting a simulated BMP assay at 2 4, and 8 g COD/L from which samples were taken throughout 20 d for pH measurements At the moderate loading rate, pH quickly dropped to below 6.5 within the first 3 days and was depressed below neutral for the first 10 d. Pretreated food waste a t the m oderate loading rate showed a greater pH depression, which correlates to the reduced 30 d methane yield from pretreated food waste in the first BMP assay. Therefore it is concluded that in the BMP assays, methanogenesis was the rate limiting step and not hydrolysis and at a higher loading rate, increased solubilization and acidogenesis can inhibit

PAGE 100

100 methanogenesis though acidification. Using a high rate anaerobic digester, such as a fixed film reactor, which contains a higher methanogenic population via r etained microbial biomass is necessary to accommodate the increased solubilization of pretreated food waste. Further experimentation is necessary to study the effect s of using high rate digestion on food waste pretreated using practical pretreatment meth ods. Practical I mplication of Research One of the primary goals of this study was to study t he impact of pretreatment on the anaerobic digestion of food w aste without using expensive, impractical pretreatment methods that have been reported in the literature. Rather, the pretreatment methods studied herein employ ed equipment that is currently available throughout the community where food waste is generated. The use of such pretreatment equipment facilitates the commercialization of food waste dig estion since the results found with these methods could be replicated on pilot or commercial scale digesters. In addition an ethical argument could be made for the use of p ractical pretreatment methods, as t he access to food waste digestion should not b e limited by access to expensive or proprietary pretreatment methods. To take full advantage of the increased solubilization kinetics of pretreated food waste, a high rate anaerobic digester is ideal A fixed film digester with retained microbial biomass has a much higher concentration of m ethanogens and hence a much higher methanogenic rate. Because the solubilization of pretreated food waste is so rapid, high methanogen populations are required to quickly convert the soluble organic material to methane The increased solubilization of pretreated food waste also increase s the acidogenic rate; therefore a high methanogenic population that can rapidly consume the acids is necessary to reduce the potential for acidification in the digester. By coupling the benefits of practical pretreatment of food waste with high rate anaerobic digesters, commercial scale food waste digestion can become a reality.

PAGE 101

101 LIST OF REFERENCES APHA. 2005. Standard Methods for the Examination of Water and Wastewater 20th ed American Public Health Association, Washington D.C. Cirne, D.G., Paloumet, X., Bjornsson, L., M.M., A., Mattiasson, B. 2007. Anaerobic digestio n of lipid rich waste Effects of lipid concentration. Renewable Energy 32 (6), 965 975. Cuellar, A., Webber, M. 2010. Wasted food, wasted energy: The embedded energy in food waste in the United States. Environmental Science Technology 44 (16), 6464 6469. Eastman, J.A., Ferguson, J.F. 1981. Solubilization of particulate organic carbon during the acid phase of anaerobic digestion. Journal (Water Pollution Control Federation) 53 (3), 352 366. EEA. 2009. Diverting waste from landfill: Effectiveness of waste management policies in the European Union. European Environmental Agency. Copenhagen. FDEP. 2011. Florida 75% Recycling Goal http://www.dep.state.fl.us/waste/recyclinggoal75/ Last accessed : May 4, 2011. FDEP. 2011. Solid Waste Management in Florida 2008 Annual Report. Florida Department of Environmental Protection. http://www.dep.state.fl.us/waste /categories/recycling/SWreportdata/08_data.htm Last accessed: May 4, 2011. Gallaher, R.N., Weldon, C.O., Futral, J.G. 1975. An aluminum block digester for plant and soil analysis. Soil Science Society of America Journal 39 (4), 803 806. Graunke, R.E., W ilkie, A.C. 2008. Converting Food Waste to Biogas: Sustainable Gator Dining. Sustainability: the Journal of Record 1 (6), 391 394. Hambleton, L.G. 1977. Semiautomated method for simultaneous determination of phosphorus, calcium and crude protein in animal feeds. Journal of the Association of Official Analytical Chemists 60 (4), 845 852. He, P.J., Lu, F., Shao, L.M., Pan, X.J., Lee, D .J. 2006. Enzymatic hydrolysis of polysaccharide rich particulate organic waste. Biotechnology and Bioengineering 93 (6), 1145 1151. IPCC. 2007. Fourth Assessemnet Report: Climate Change 2007: Working Group I: The Physical Science Basis. Intergovernmenta l Panel on Climate Change. http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch2s2 10 2.html Last accessed May 4, 2011. Izumi, K., Okishio, Y. k., Nagao, N., Niwa, C ., Yamamoto, S., Toda, T. 2010. Effects of particle size on anaerobic digestion of food waste. International Biodeterioration & Biodegradation 64 (7), 601 608.

PAGE 102

10 2 Kim, H.J., Choi, Y.G., Kim, G.D., Kim, S.H., Chung, T.H. 2005. Effect of enzymatic pretreatment on solubilization and volatile fatty acid production in fermentation of food waste. Water Science & Technology 52 (10 11), 51 59. Kim, H.K., Gadd, G.M. 2008. B acterial Physiology and Metabolism Cambridge University Press, Cambrige. Kjeldsen, P., Barlaz, M.A., Rooker, A.P., Baun, A., Ledin, A., Christensen, T.H. 2002. Present and long term composition of MSW landfill leachate: A review. Critical Reviews in Envi ronmental Science and Technology 32 (4), 297 336. and biogas production in anaerobic solubilization of food waste. Waste Management 29 (12), 2950 2955. Liu, X.Y., Ding, H.B., Srerramachandran, S., Stabnikova, O., Wang, J.Y. 2008. Enhancement of food waste digestion in the hybrid anaerobic solid liquid system. Water Science & Technology 57 (9), 1369 1373. Liu, X.Y., Ding, H.B., Wang, J.Y. 2010. Food Waste to Bioene rgy. in: Bioenergy and Biofuels from Biowastes and Biomass (Eds.) S.K. Khanal, R.Y. Surampalli, T.C. Zhang, B.P. Lamsal, R.D. Tyagi, C.M. Kao, American Society of Civil Engineers. Reston, VA. Lynd, L.R., Weimer, P.J., van Zyl, W.H., Pretorius, I.S. 2002. Microbial cellulose utilization: Fundamentals and biotechnology. Microbiology and Molecular Biology Reviews 66 (3), 506 577. Masse, L., Kennedy, K.J., Chou, S.P. 2001. The effect of an enzymatic pretreatment on the hydrolysis and size reduction of fat pa rticles in slaughterhouse wastewater. Journal of Chemical Technology and Biotechnology 76 (6), 629 635. McInerney, M.J. 1988. Anaerobic hydrolysis and fermentation of fats and proteins. in: Biology of Anaerobic Microorganisms (Ed.) A.J.B. Zehnder, John W iley and Sons. New York. Mendes, A.A., Pereira, E.B., de Castro, H.F. 2006. Effect of the enzymatic hydrolysis pretreatment of lipids rich wastewater on the anaerobic biodigestion. Biochemical Engineering Journal 32 (3), 185 190. Neves, L., Goncalo, E., Oliveira, R., Alves, M.M. 2008. Influence of composition on the biomethanation potential of restaurant waste at mesophilic temperatures. Waste Management 28 (6), 965 972. Owen, W.F., Stuckey, D.C., Healy, J.B., Young, L.Y., McCarty P.L. 1979. Bioassay for monitoring biochemical methane potential and anaerobic toxicity. Water Research 13 (6), 485 492.

PAGE 103

103 Palmowski, L.M., Muller, J.A. 2003. Anaerobic degradation of organic materials significance of the substrate surface area. Water S cience & Technology 47 (12), 231 238. Prashanth, S., Kumar, P., Mehrotra, I. 2006. Anaerobic Degradabilty: Effect of Particulate COD. Journal of Environmental Engineering 132 (4), 488 496. Sanders, W.T.M., Geerink, M., Zeeman, G., Lettinga, G. 2000. Anaerobic hydrolysis kinetics of particulate substrates. Water Science & Technology 41 (3), 17 24. Sharma, R., Chisti, Y., Banjerjee, U.C. 2001. Production, purification, characterization, and a pplications of lipases. Biotechnology Advances 19 (8), 627 662. Speece, R.E. 2008. Anaerobic Biotechnology and Odor/Corrosion Control for Municipalities and Industries Archae Press, Nashville, TN. U.S. EPA. 2006. Putting Surplus Food to Good Use: A How to Guide for Food Service Providers. United States Environmental Protection Agency. http://www.epa.gov/osw/conserve/materials/organics/pubs/food guide.pdf Last accessed: May 4, 2011. U.S. EPA 2011. 2011 US Greenhouse Gas Inventory Report: Chapter 8 Waste. United States Environmental Protection Agency. http://epa.gov /climatechange/emissions/downloads11/US GHG Inventory 2011 Chapter 8 Waste.pdf Last accessed: May 4, 2011. U.S. EPA. 2011. Municipal Solid Waste Generation, Recylcing, and Disposal in the United States: Facts and Figures for 2009. United States Environme ntal Protection Agency. http://www.epa.gov/epawaste/nonhaz/municipal/pubs/msw2009 fs.pdf Last accessed: May 4, 2011. Veeken, A., Hamelers, B. 1999. Effect of temperature on hy drolysis rates of selected biowaste components. Bioresource Technology 69 (3), 249 254. Veeken, A., Kalyuzhnyi, S., Scharff, H., Hamelers, B. 2000. Effect of pH and VFA on hydrolysis of organic solid waste. Journal of Environmental Engineering 126 (12), 1 076 1081. Wang, J. Y., Liu, X. Y., Kao, J., Stabnikova, O. 2006. Digestion of pre treated food waste in a hybrid anaerobic solid liquid (HASL) system. Journal of Chemical Technology and Biotechnology 81 (3), 345 351. Wang, X.Q., Wang, H.Q., Liu, Y.Y., Ma, H.Z., Wang, X.M. 2009. Kinetics and thermodynamics of glucoamylase inhibition by lactate during fermentable sugar production from food waste. Journal of Chemical Technology and Biotechnology 85 (5), 687 692.

PAGE 104

104 Wilkie, A.C. 2005. Anaerobic Digestion: Biology and Benefits. in: Dairy Maure Management: Treatment, Handling, and Community Relations Natural Resource, Agriculture, and Engineering Service. Ithaca, NY, pp. 63 72. Wilkie, A.C. 2008. Biomethane from biomass, biow aste, and biofuels. in: Bioenergy (Eds.) J.D. Wall, C.S. Harwood, A. Demain, American Society for Microbiology. Washington D.C., pp. 195 205. Wilkie, A.C., Smith, P.H., Bordeaux, F.M. 2004. An economical bioreactor for evaluating biogas potential of part iculate biomass. Bioresource Technology 92 (1), 103 109.

PAGE 105

105 BIOGRAPHICAL SKETCH Ryan Graunke was born in Stuart, Florida to Robert and Barbara Graunke. In 2008, he graduated Summa Cum Laude from the University of Florida with a Bachelor of Science in environmental science. His undergraduate thesis Broward Dining Association for the Advancement of Sustainability in Higher Education. Ryan graduated in 2011 Science in interdisciplinary ecology.