anaerobic phased solids digester pilot demonstration project · 2013-09-09 · anaerobic phased...

Energy Research and Development Div is ion FINAL PROJECT REPORT

ANAEROBIC PHASED SOLIDS DIGESTER PILOT DEMONSTRATION PROJECT

MARCH 2011

CEC ‐500 ‐2013 ‐077

Prepared for: California Energy Commission

Prepared by: University of California at Davis and Onsite Power Systems, Inc.

PREPARED BY: Primary Author(s): Ruihong Zhang, UC Davis Josh Rapport, UC Davis

Contributing Author(s): Dave Konwinski, OPS Scott Archibald, OPS Hamed El-Mashed, UC Davis Iain Clark, UC Davis Richard von Langen, RvL Associates

Contract Number: 500-02-004 UC MR-029 Prepared for: California Energy Commission Abolghasem Edalati Project Manager Linda Spiegel Office Manager Energy Generation Research Office Laurie ten Hope Deputy Director ENERGY RESEARCH AND DEVELOPMENT DIVISION Robert P. Oglesby Executive Director

DISCLAIMER This report was prepared as the result of work sponsored by the California Energy Commission. It does not necessarily represent the views of the Energy Commission, its employees or the State of California. The Energy Commission, the State of California, its employees, contractors and subcontractors make no warrant, express or implied, and assume no legal liability for the information in this report; nor does any party represent that the uses of this information will not infringe upon privately owned rights. This report has not been approved or disapproved by the California Energy Commission nor has the California Energy Commission passed upon the accuracy or adequacy of the information in this report.

i

ACKNOWLEDGEMENTS

This research was made possible by funding from the California Energy Commission, Propane Research Council, University of California, Davis, and Onsite Power Systems, Inc. Intellectual contributions were made by the following individuals:

• University of California, Davis

Ruihong Zhang, Joshua Rapport, Iain Clark, Shumei Jiang, Xiguang Chen, Karl Hartman, Hamed M. El‐Mashad, Robert Williams, and Gary Matteson

• Onsite Power Systems, Inc.

Dave Konwinski, Scott Archibald, Todd Winters

• Sacramento Municipal Utility District

Valentino Tiangco

• Advanced Food Technologies

Dave Bartell

• TGP West

Mitch Torp

• Hal Murphree and Associates

Hal Murphree

• California Institute of Energy Efficiency

Ken Krich

The following companies contributed to the design, development and construction of the digester project:

• Brown and Caldwell (Computer controls design)

• Rockwell Engineering (Material processing and hydraulic mixing)

• Mazzie Injectors (Hydraulic mixing)

• Kouba Engineering (Engineering design review)

• Graybar (Instrumentation and controls components)

• Advanced Food Technologies (Material processing equipment and design)

• Wonderware (Computer control and data collection programs)

ii

PREFACE

The California Energy Commission Energy Research and Development Division supports public interest energy research and development that will help improve the quality of life in California by bringing environmentally safe, affordable, and reliable energy services and products to the marketplace.

The Energy Research and Development Division conducts public interest research, development, and demonstration (RD&D) projects to benefit California.

The Energy Research and Development Division strives to conduct the most promising public interest energy research by partnering with RD&D entities, including individuals, businesses, utilities, and public or private research institutions.

Energy Research and Development Division funding efforts are focused on the following RD&D program areas:

• Buildings End‐Use Energy Efficiency

• Energy Innovations Small Grants

• Energy‐Related Environmental Research

• Energy Systems Integration

• Environmentally Preferred Advanced Generation

• Industrial/Agricultural/Water End‐Use Energy Efficiency

• Renewable Energy Technologies

• Transportation

This report is the final report for the Anaerobic Phased Solids Digester Pilot Demonstration project (Contract Number 500‐02‐004) conducted by University of California, Davis, and Onsite Power Systems, Inc. The information from this project contributes to Energy Research and Development Division’s Renewable Energy Technologies Program.

For more information about the Energy Research and Development Division, please visit the Energy Commission’s website at www.energy.ca.gov/research/ or contact the Energy Commission at 916‐327‐1551.

iii

ABSTRACT

The anaerobic phased solids digester system is a two‐stage, high‐solids, waste‐to‐energy conversion technology developed at the University of California at Davis. A 30,000‐gallon pilot system was built and operated by researchers at the University of California at Davis and Onsite Power Systems, Inc. during 2008‐2009. The pilot plant successfully treated up to three tons per day of commercial food waste from the Campbell’s® processing facility in Sacramento, California. The system was operated at 50 degrees Celsius and loaded once per week in phased batches with a 14‐ day solids retention time. Minimal mechanical difficulties were encountered, primarily centered on instrumentation and heat transfer. Modifications made during the operational period resolved these issues and provided invaluable experience for future generations of the technology. The technical feasibility of converting solid waste to biogas was proven. Performance improvements can be made based on the data from this study.

The digesters were fed with food waste for six months, resulting in an efficient conversion to biogas before excessive acidity in the second‐stage reactor halted biogas production. Adding fresh inoculum (microorganisms used to start a new culture) recovered biogas production within a month and led to steady‐state biogas production after two months. The average biogas and methane yields were 500 cubic meters per ton volatile solids and 275 cubic meters per ton volatile solids, respectively during the steady‐state period. Eighty to 90 percent of the volatile solids in the food waste were converted to biogas. High hydrogen concentrations were seen in biogas from the first‐stage reactors. The digester effluent had an average total solid content of about five percent; was rich in nutrients like nitrogen, phosphorus, and potassium; and had relatively low salt concentrations, making it a potentially valuable liquid fertilizer. The effluent may require some treatment prior to land application.

Keywords: Anaerobic phased solids digester, anaerobic digestion, biogas, food waste, pilot plant, operational report, renewable energy, waste treatment, high solids digestion

Please use the following citation for this report:

Zhang, Ruihong and Joshua Rapport. (UC Davis). 2011. Anaerobic Phased Solids Digester Pilot Demonstration Project. California Energy Commission. Publication Number: CEC‐500‐2013‐077.

iv

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ....................................................................................................................... i

PREFACE ....................................................................................................................................................ii

ABSTRACT .............................................................................................................................................. iii

TABLE OF CONTENTS .......................................................................................................................... iv

LIST OF FIGURES .................................................................................................................................... x

LIST OF TABLES .................................................................................................................................... xv

EXECUTIVE SUMMARY ........................................................................................................................ 1

Introduction ............................................................................................................................................ 1

Project Purpose ....................................................................................................................................... 1

Project Results ......................................................................................................................................... 1

Project Benefits ....................................................................................................................................... 3

CHAPTER 1: Pilot Digester Design and Test Plan ............................................................................. 5

1.1 Introduction ................................................................................................................................ 5

1.2 Project Overview ........................................................................................................................ 5

1.2.1 UC Davis Biogas Energy Demonstration Project .......................................................... 5

1.2.2 APS‐Digester System Overview....................................................................................... 6

1.2.3 Feedstock Characteristics .................................................................................................. 8

1.3 Pilot Digester System Design Considerations ........................................................................ 8

1.3.1 Mass Balance of the Pilot Digester System ..................................................................... 9

1.3.2 Energy Balance of the Pilot Digester System ............................................................... 10

1.4 Pilot Digester Component Design Specifications ................................................................ 11

1.4.1 Concrete Pad ..................................................................................................................... 11

1.4.2 Reactor Vessels ................................................................................................................. 11

1.4.3 Heat Load and Heat Exchangers ................................................................................... 12

v

1.5 Pilot Digester System Operational Design ........................................................................... 15

1.5.1 Hydrolysis Reactor Loading ........................................................................................... 15

1.5.2 Liquid Circulation Process .............................................................................................. 16

1.5.3 Hydraulic Mixing System ............................................................................................... 16

1.5.4 Solids Residue and Water Recovery .............................................................................. 17

1.5.5 Tank Pressure/Vacuum Equalization ............................................................................ 18

1.5.6 Safety Pressure Relief Valve ........................................................................................... 18

1.5.7 Limiting Introduction of Air into the System .............................................................. 18

1.5.8 Computer Control System .............................................................................................. 18

1.5.9 OPS Proprietary Computer Controls Software and Process Protocol ...................... 19

1.6 Biogas Collection and Processing System ............................................................................. 20

1.6.1 Biogas Collection .............................................................................................................. 20

1.6.2 Biogas Upgrading ............................................................................................................ 21

1.6.3 Biogas Measurement ........................................................................................................ 21

1.6.4 Biogas Compression ........................................................................................................ 21

1.6.5 Electrical generator .......................................................................................................... 21

1.6.6 Safety Biogas Flare ........................................................................................................... 22

1.7 Safety and Environmental Considerations ........................................................................... 22

1.7.1 Safety Considerations ...................................................................................................... 22

1.7.2 Environmental Considerations ...................................................................................... 23

1.8 Pilot Digester Plant Testing Program and Modification Guidelines ................................ 24

1.8.1 Engineering Design Modifications ................................................................................ 24

1.8.2 Material Handling Testing .............................................................................................. 26

1.8.3 Expanded Feedstock Testing, Protocol and Procedures ............................................. 27

1.8.4 UCD‐RT Development of CDS Design Modifications ................................................ 27

1.8.5 Final As‐Built Engineering Design Review .................................................................. 27

1.9 Pilot Digester System Performance Test Plan ...................................................................... 28

1.9.1 Biogas Production and Stability of Anaerobic Reactors ............................................. 28

vi

1.9.2 Energy Conversion Performance and Air Emissions of Engine‐Generator ............. 28

1.10 Conclusions ............................................................................................................................... 31

CHAPTER 2: Pilot Digester Construction Report ............................................................................. 32

2.1 Introduction .............................................................................................................................. 32

2.2 Pilot Digester Construction Specifications ........................................................................... 32

2.2.1 Concrete Pad ..................................................................................................................... 33

2.2.2 Reactor Vessels ................................................................................................................. 33

2.2.3 Boilers and Heat Exchangers .......................................................................................... 35

2.3 Material Handling and Digester Operational Equipment ................................................. 37

2.3.1 Hydrolysis Reactor Loading ........................................................................................... 37

2.3.2 Liquid Circulation Process .............................................................................................. 41

2.3.3 Internal Hydraulic Mixing System – Hydrolysis Tanks 1 and 2 ............................... 43

2.3.4 Hydraulic Internal Mixing System – Hydrolysis Tanks 3 and 4 ............................... 43

2.3.5 Hydraulic Internal Mixing System – Biogasification Tank ........................................ 44

2.3.6 Solids Residue and Water Recovery .............................................................................. 45

2.3.7 Tank Pressure/Vacuum Equalization ............................................................................ 46

2.3.8 Safety Pressure Relief Valve ........................................................................................... 46

2.3.9 Limiting Introduction of Air Into the System .............................................................. 47

2.3.10 Computer control system ................................................................................................ 47

2.3.11 OPS Proprietary Computer Controls Software and Process Protocol ...................... 51

2.4 Biogas Collection and Processing System ............................................................................. 51

2.4.1 Biogas Collection .............................................................................................................. 52

2.4.2 Biogas Compression ........................................................................................................ 54

2.4.3 Electrical Generator ......................................................................................................... 55

2.4.4 Safety Biogas Flare ........................................................................................................... 57

2.5 Safety and Environmental Considerations ........................................................................... 57

2.5.1 Safety Considerations ...................................................................................................... 57

2.5.2 Environmental Considerations ...................................................................................... 58

vii

2.6 Equipment and Cost of UC Davis Biogas Energy Plant ..................................................... 58

CHAPTER 3: Operation and Testing of Pilot Digester System with Solid Food Waste .......... 62

3.1 Modifications Made Prior to Operation and Testing .......................................................... 62

3.2 Operational Experience ........................................................................................................... 63

3.2.1 Provision and Loading of Feedstock ............................................................................. 63

3.2.2 Start‐up and Development of Steady‐State Operations ............................................. 72

3.2.3 Maintenance, Repairs and Upgrades ............................................................................ 73

3.3 APS Digester Performance Results ........................................................................................ 79

3.3.1 Data Collection and Analytical Methods ...................................................................... 79

3.3.2 Steady‐State Biological Stability and Performance ..................................................... 87

3.4 Conclusions ............................................................................................................................... 95

CHAPTER 4: Engineering, Economic and Environmental Analysis of APS Digester ............... 97

4.1 Review of Current Technologies Used For the Anaerobic Digestion of Municipal Solid Waste ........................................................................................................................ 97

4.1.1 Introduction to the Treatment of the Organic Fraction of Municipal Solid Waste ……………………………………………………………………………………………97

4.1.2 Overview of Anaerobic Digestion Technologies for Treating Municipal Solid Waste ……………………………………………………………………………………………97

4.1.3 Examples of Commercial Anaerobic Digester Systems Used for Treating Municipal Solid Waste .................................................................................................................. 100

4.1.4 Biogas and Methane Production and Digester Performance ................................... 107

4.1.5 Economics of Anaerobic Digestion of MSW .............................................................. 110

4.1.6 Conclusions from the Review of Commercial Anaerobic Digestion Technologies for the Treatment of Municipal Solid Waste .............................................................................. 112

4.2 Background Information Used For Modeling the Mass and Energy Balance and Financial Performance of the APS Digester System ...................................................................... 113

4.2.1 Description of the APS Digester System Modeling Project ...................................... 113

4.2.2 Objectives of the APS Digester System Modeling Project ........................................ 114

4.2.3 Comparison with and Evaluation of Previous Models of the APS Digester System ……………………………………………………………………………………………115

viii

4.2.4 Review of Technologies Used to Convert Biogas to Heat, Electricity, and Natural Gas ……………………………………………………………………………………………117

4.3 Development of a Model for Calculating the Mass and Energy Balance and Financial Performance of the APS Digester System ...................................................................... 121

4.3.1 Calculation of the Mass and Energy Balance and Financial Performance of the APS Digester System .............................................................................................................................. 122

4.3.2 Assumptions Made When Modeling The APS Digester System ............................ 148

4.3.3 Validation of the Present Model by Comparing Results with Previously Published Studies... .......................................................................................................................................... 167

4.4 Results and Conclusions of the Economic and Financial Analysis ................................. 170

4.4.1 Results of the Mass and Energy Balance Calculations .............................................. 170

4.4.2 Results of the Financial Analysis ................................................................................. 179

4.4.3 Conclusions from the Model of the Financial and Engineering Performance of the APS Digester System ..................................................................................................................... 197

4.4.4 Recommendations for Continued Development of the APS Digester System Model ……………………………………………………………………………………………201

CHAPTER 5: Waste‐to‐Energy Project for a Food Processing Plant ........................................... 202

5.1 Introduction ............................................................................................................................ 202

5.2 Project Setting ......................................................................................................................... 202

5.2.1 Location and Environmental Setting ........................................................................... 202

5.2.2 Regulatory Requirements ............................................................................................. 203

5.2.3 Waste Characteristics ..................................................................................................... 204

5.2.4 General Process Description ......................................................................................... 205

5.2.5 Project Assumptions ...................................................................................................... 205

5.3 Project Description ................................................................................................................. 206

5.3.1 Waste Preprocessing and Equalization ....................................................................... 207

5.3.2 Anaerobic Phased Solids (APS) Digestion .................................................................. 209

5.3.3 Digestate and AD Effluent Processing ........................................................................ 209

5.3.4 Biogas ............................................................................................................................... 210

5.3.5 Mass and Energy Balance ............................................................................................. 210

ix

5.4 Cost Options ........................................................................................................................... 214

5.4.1 Capital Cost Opinion ..................................................................................................... 215

5.4.2 Operating Cost Opinion ................................................................................................ 216

CHAPTER 6: Characterization of Food and Green Wastes as Feedstock for Anaerobic Digesters ................................................................................................................................................. 218

6.1 Introduction ............................................................................................................................ 218

6.2 Materials and Methods .......................................................................................................... 220

6.2.1 Characterization of Fresh and Green Wastes ............................................................. 220

6.2.2 Anaerobic Digestion Tests ............................................................................................ 221

6.3 Results and Discussion .......................................................................................................... 222

6.3.1 Characteristics of Food Wastes .................................................................................... 222

6.3.2 Characteristics of Green Wastes ................................................................................... 225

6.3.3 Anaerobic Digestion Tests ............................................................................................ 226

6.4 Conclusions ............................................................................................................................. 233

CHAPTER 7: Thermophilic Digestion of Green and Food Wastes with an Anaerobic Phased Solids Digester System .......................................................................................................... 235

7.1 Introduction ............................................................................................................................ 235

7.2 Background and Research Objectives ................................................................................. 235


7.3.1 Experimental Design ..................................................................................................... 236

7.3.2 Feedstock Collection and Preparation ........................................................................ 237

7.3.3 APS‐Digester System Description ............................................................................... 238

7.3.4 Analysis and Measurements ........................................................................................ 240

7.3.5 APS‐ Digester System Startup and Operational Procedures ................................... 241

7.3.6 Data Analysis .................................................................................................................. 242


7.4.1 Anaerobic Digester System Startup ............................................................................. 242

7.4.2 Digestion of Green and Food Waste Mixture ............................................................ 243

x

7.4.3 Digestion of Green Waste .............................................................................................. 244

7.4.4 Digestion of Food Waste ............................................................................................... 247

7.5 Conclusions ............................................................................................................................. 249

CHAPTER 8: Biodegradability and Soil Amendment Potential of Anaerobically Digested Residues................................................................................................................................. 251

8.1 Introduction ............................................................................................................................ 251

8.2 Background and Research Objectives ................................................................................. 251


8.3.1 Food and Green Wastes ................................................................................................ 253

8.3.2 Digestion System ............................................................................................................ 253

8.3.3 Particle Size Distribution ............................................................................................... 253

8.3.4 Carbon Dioxide and Oxygen Consumption Measurements .................................... 254

8.3.5 Data Analysis .................................................................................................................. 256


8.4.1 Element Contents ........................................................................................................... 257

8.4.2 Compost Suitability ....................................................................................................... 258

8.4.3 Respirometry Experiments ........................................................................................... 260

8.5 Conclusions ............................................................................................................................. 264

APPENDIX A: List of Patents, Publications, Journal Articles and Presentations .................... A‐1

APPENDIX B: Digester System Illustrations .................................................................................. B‐1

LIST OF FIGURES Figure 1. A schematic of APS‐Digester System ..................................................................................... 7 Figure 2. Mass balance of UC Davis Pilot Demonstration Digester System ...................................... 9 Figure 3. Energy Balance for Summer (A) and Winter (B) Seasons .................................................. 10 Figure 4. Schematic of Biogas Fueled CHP System ............................................................................ 30 Figure 5. APS Digester Pilot Plant. Orientation of Reactors and Gas Processing Equipment at the

APS Digester Pilot Plant at UC Davis. .......................................................................................... 32 Figure 6. Main Frontal View of Digester System with Loading Dock ............................................. 33

xi

Figure 7. Five Anaerobic Reactors Including Four Hydrolysis Reactors and One Biogasification Reactor .............................................................................................................................................. 34

Figure 8. Buffer Tank Located Between Hydrolysis and Biogasification Reactors ........................ 35 Figure 9. Dual Boiler System ................................................................................................................. 35 Figure 10. Propane Operated Boiler for System Start‐Up .................................................................. 36 Figure 11. GT External Heat Exchanger System .................................................................................. 36 Figure 12. Air Fin Heat Exchanger System ........................................................................................... 37 Figure 13. Delivery of Bins Containing Feedstock Materials ............................................................. 38 Figure 14. Bins of Waste Materials Delivered for Loading ................................................................ 38 Figure 15. Forklift Dumps Bin into Receiving Hopper ....................................................................... 39 Figure 16. Receiving Hopper Metering Feedstock into Conveyor System ...................................... 39 Figure 17. Conveyor Delivers Feedstock to Hydraulic Piston Pumps ............................................. 40 Figure 18. Hydraulic Piston Pump Located at HT 3 and HT 4 .......................................................... 40 Figure 19. Hydraulic Piston Pump Located at HT 1 and HT 2 .......................................................... 41 Figure 20. Valve System for HT Loading.............................................................................................. 41 Figure 21 Screen Filter, Self Cleaning System ...................................................................................... 42 Figure 22. pH Meter and Circulation Buffer Tank .............................................................................. 42 Figure 23. In‐line Chopper Pump and Pipe System ............................................................................ 43 Figure 24 Hydraulic Mixing Pump HT 3 and HT 4 ............................................................................ 44 Figure 25 GT Hydraulic Mixing Pump and Pipe System ................................................................... 44 Figure 26. HT Drain Valve and Connection to Drain System ............................................................ 45 Figure 27. Drain Section to be Connected to Pump and Screw Press Unit ...................................... 46 Figure 28. Pressure Relieve Units and Flame Arrester ....................................................................... 47 Figure 29 Process Flow Monitoring Screen .......................................................................................... 48 Figure 30. Spread Spectrum Wireless Data Transmission Boxes ...................................................... 49 Figure 31. Biogas Flow, Biohydrogen Flow and pH Meters .............................................................. 50 Figure 32. Motor Control Cabinet with HMI Interface Panel ............................................................ 50 Figure 33. One of Three Magnetic Flow Meters................................................................................... 51 Figure 34. Biogas Collection Manifold .................................................................................................. 52 Figure 35. Hydrogen Sulfide Filter with SulfaTreat ............................................................................ 53 Figure 36. Coalescent Filter, Expansion Tank and Drain Traps (Located on Ground) ................. 53 Figure 37. Biogas Pressure Regulator System ..................................................................................... 54 Figure 38. Biogas Blower Unit ................................................................................................................ 55 Figure 39. Engine‐Generator System with Control Switch Box......................................................... 56 Figure 40. Engine‐Generator System with Control Switch Box......................................................... 56 Figure 41. Schematic of APS Digester System ..................................................................................... 63 Figure 42. Timeline of APS Digester Operations ................................................................................. 64 Figure 44. Feedstock Sampling. Sampling Food‐Waste From a Municipal Food‐Waste

Composting Facility. ....................................................................................................................... 66 Figure 45. Transfer to Chopper Pump. Transfer from Feed Hopper to Chopper Pump Hopper

of Food Waste From a Municipal Solid Waste Composting Facility. ...................................... 67 Figure 46. Campbell’s® Processing Waste. Food‐Waste From Campbell’s® Processing Plant

(Sacramento, CA) as Delivered to the Pilot Plant at UC Davis ................................................. 67

xii

Figure 47. Campbell’s® Feedstock Loading. Feedstock loading. Clockwise from Top Left: Bin Receiving, Sampling, Transfer, and Tipping ............................................................................... 69

Figure 48. Material Handling Equipment Food‐Waste Hydraulic Bin Tipping Unit, Loading Hopper, and Chopper‐Pump Hopper (Left), and Hydraulic Ram Pump (Right) ................. 70

Figure 49. Biogas Production During Startup of the APS Digester System ..................................... 73 Figure 50. Total Working Volume and Contribution of Each Tank to the Overall Volume ......... 80 Figure 51. Typical Gas Production Profile ............................................................................................ 81 Figure 52. Gas Flow‐Rate Adjustment Chart for Sage Thermal Mass Flow Meters Calibrated on a

Mix Of 65 Percent CH4 and 35 Percent CO2 ................................................................................. 82 Figure 53. Effect on Measured Biogas Flow Rate (●) of Adding/Removing Water (▲) From Gas

Flow Line .......................................................................................................................................... 83 Figure 54. Relation Between Water Added to Gas Pipe and Change in Measured Biogas Flow

Rate .................................................................................................................................................... 83 Figure 55. Example of Typical Sampling Schedule, TK300 = Biogasification Tank, TK101 =

Hydrolysis Tank 1, and TK102 = Hydrolysis Tank 2 .................................................................. 85 Figure 56. Hydrogen Content of Biogas from the Two Hydrolysis Tanks and the Biogasification

Tank ................................................................................................................................................... 88 Figure 57. Methane and Hydrogen Content of Biogas During the Second Operational Period... 89 Figure 58. Cumulative Volatile Solids Loading (t) Overlaid with Biogas and Methane Yields

(m3/t VS) Over the Duration of the Testing Period ..................................................................... 90 Figure 59. Running Organic Loading Rate ........................................................................................... 90 Figure 60. pH Levels in the Hydrolysis and Biogasification Tanks .................................................. 91 Figure 61. Mass of Volatile Solids and Volatile/Total Solids Ratio for the APS Digester System 93 Figure 62. Volatile Solids Reduction ..................................................................................................... 94 Figure 63. Schematic Diagram of the APS Digester System. ........................................................... 105 Figure 64. Pilot Demonstration Plant for the APS Digester System Technology at UC Davis,

Front View (top) and Rear View (Bottom) ................................................................................. 106 Figure 65. Laboratory, Pilot and Full‐Scale Anaerobic Digester Biogas Production Rate When

Treating a Variety of Feedstocks [43] ......................................................................................... 109 Figure 66. Capital and Operating Cost Curves for MSW Digesters from Two Studies [46, 50],

Adjusted for Inflation to 2007 U.S. Dollars ................................................................................ 111 Figure 67. Reported Electrical Conversion Efficiency and Capital Cost Curves for Reciprocating

Engine Cogeneration Systems (Adapted from Firestone [62]) ............................................... 118 Figure 68. Conceptual Flow Chart for the Design of an Integrated Mass/Energy Balance and

Financial Calculator for the Anaerobic Phased Solids Digester ............................................. 122 Figure 69. Mass Flow Diagram for Calculating the Mass Balance of the Anaerobic Digester,

Including All Possible Biogas Uses. Dashed Lines Indicate Biogas Flow ............................. 127 Figure 70. State Change as Biogas Cools from the Reactor Temperature (Tr) to Ambient

Temperature (Tamb) and Water Condenses (ncondense = Moles of Water Condensed, nBG = Moles of Biogas) ............................................................................................................................. 132

Figure 71. Flow Chart for Determining the Distribution of Biogas in the CG Scenario............... 141 Figure 72. Flow Chart for Determining the Distribution of Biogas When Converting to Heat,

Electricity and CNG (CG+CNG Scenario) ................................................................................. 145

xiii

Figure 73. Theoretical Maximum Biogas Yield Potential at 298 K,101.325 Kpa as a Function of Methane Content. .......................................................................................................................... 152

Figure 74. Water Vapor Density and Biogas Vapor Content as a Function of Temperature. Curves Were Fit to Published Data [70, 76]. .............................................................................. 154

Figure 75. Capital and Maintenance Cost Data and Curve Fitting for Combined Heat and Power Engine Generators [62] ................................................................................................................. 160

Figure 76. Ten‐Year Trend in the Annual Escalation Rate for Electricity in the U.S. ................... 165 Figure 77. Effect of Setting the Minimum Attractive Rate of return (MARR) Equal to the After‐

Tax Interest Rate on Debt. Data Generated By the Present Model ......................................... 167 Figure 78. Overall Mass and Water Balance for the APS Digester System. Percentages Given as a

Fraction of the Mass of Feed Input to the Digester May Not Sum to 100 Percent Due to Rounding ........................................................................................................................................ 172

Figure 79. Mass of Trace Gases Calculated Based on the Mass Balance over a Range of Biodegradabilities, with the Biogas Yield Fixed at 526.7 m3 t‐1 VS Fed .................................. 173

Figure 80. Effect of Mean Annual Ambient Temperature on the Amount of Heat Required Annually By the APS Digester System ....................................................................................... 176

Figure 81. Effect on the Amount of CNG Produced of Changing the Boiler Recovery Efficiency (for the CNG Scenario) ................................................................................................................. 177

Figure 82. Effect of the Fraction of Heat Captured from the Generator on the Amount of CNG and Heat Produced (for the CG+CNG Scenario) ...................................................................... 178

Figure 83. Capital Cost Breakdown for the Three Scenarios ........................................................... 179 Figure 84. Ordinary Expenses Breakdown for the Three Scenarios ............................................... 180 Figure 85. Revenue breakdown for the Three Scenarios (Not Including Tax Credits) ................ 182 Figure 86. 20‐Year Cash Flows for the Three Scenarios Analyzed .................................................. 182 Figure 87. Effect of Financing on Levelized Cost of Energy ($ kWh‐1) for the CG Scenario, Where

Energy = Electricity (HHV, 30 Percent Efficiency), and the CNG and CG+CNG Scenarios, Where Energy = Natural Gas (HHV) .......................................................................................... 184

Figure 88. Effect of Financing a 15 Percent MARR on the Levelized Cost of Treatment ($ t‐1 Wet Waste), with the Base Case (Zero Debt, 4.42 Percent MARR) Included for Comparison ... 185

Figure 89. Maximization of NPW for the CNG Producing Case over a Range of Biogas Diversion Schemes. Biogas Not Diverted to the Generator o Boiler Was Converted to CNG. ............ 188

Figure 90. Effect of C‐Location (Demand Charge) Vs. Independent Location (No Demand Charge) of APS Digester on NPW (Assuming Equal Capital Costs). Retail = $0.0857 kWh‐1, Wholesale = $0.067 kWh‐1, Feed‐In = $0.09572 kWh‐1 ............................................................... 189

Figure 91. Effect of Co‐Location (Demand Charge) Vs. Independent Location (No Demand Charge) of APS Digester on Revenues and Expenses (Assuming Equal Capital Costs). Retail = $0.0857 kWh‐1, Wholesale = $0.067 kWh‐1, Feed‐In = $0.09572 kWh‐1 .................. 189

Figure 92. Sensitivity of the CG Scenario’s Financial Performance to Changes in the Assumptions (Top) and a Comparison of the Slopes of the Linear Sensitivity Curves (Bottom) .......................................................................................................................................... 190

Figure 93. Sensitivity of the CNG Scenario’s Financial Performance to Changes in the Assumptions (Top) and a Comparison of the Slopes of the Linear Sensitivity Curves (Bottom) .......................................................................................................................................... 191

xiv

Figure 94. Sensitivity of the CG+CNG Scenario’s Financial Performance to Changes in the Assumptions (Top) and a Comparison of the Slopes of the Linear Sensitivity Curves (Bottom) .......................................................................................................................................... 192

Figure 95. NPW of the CG Scenario as a Function of Annual Ordinary Expenses ...................... 193 Figure 96. NPW and IRR of the CG Scenario as a Function of the Feed MC (All Other Feed

Characteristics Held Constant) .................................................................................................... 194 Figure 97. Effect on NPW of the CG and CG+CNG Scenario When Including Heat Sales at a Rate

of $0.0273 kWh‐1. ............................................................................................................................ 195 Figure 98. Effect of Mean Annual Ambient Temperature on the NPW of All Three Scenarios,

Including the CG Scenario with All Usable Heat Sold for $0.0273 kWh‐1 ............................. 196 Figure 99. Effect of the R‐Value of the Insulation on the Digester Heat Requirement and NPW of

All Three Scenarios and the CG Scenario When Heat Sales Were Included ........................ 197 Figure 100. Main Components of a Specific Waste: Not In a Scale ................................................. 221 Figure 101. Daily Average MC, VS (Weight Basis) and VS/TS of Food Wastes with the Standard

Deviations as Indicated By Y Error Bars .................................................................................... 223 Figure 102. Daily Average MC, VS (Weight Basis) and VS/TS of Food Wastes with the Standard

Deviations as Indicated By Y Error Bars .................................................................................... 224 Figure 103. Weekly Average MC, VS (Wet Weight Basis) and VS/TS of Green Wastes. Y Error

Bars Show the Standard Deviations ........................................................................................... 226 Figure 104. Methane yield of Food Waste During Anaerobic Digestion at 50°C .......................... 227 Figure 105. Daily Methane Production During Digestion of Food Wastes ................................... 228 Figure 106. Methane Yield During the Digestion of Green Waste Placed Loosely in the Reactor

.......................................................................................................................................................... 229 Figure 107. Methane Yield During the Digestion of Green Waste Placed in Nylon Bags

Submerged in the Reactor. ........................................................................................................... 229 Figure 108. Daily Methane Production During Digestion of Green Waste Placed Loosely in the

Reactor ............................................................................................................................................ 230 Figure 109. Daily Methane Production During Digestion of Green Waste Placed in Nylon Bags

Submerged in the Reactor ............................................................................................................ 231 Figure 110. Biogas Composition during Food Wastes Digestion at Two Different Initial

Loadings (6.8 and 10.5 gVS/L) ..................................................................................................... 232 Figure 111. Biogas Composition During Green Waste Digestion ................................................... 232 Figure 112. Schematic of Laboratory Set‐Up of APS Digester System ........................................... 239 Figure 113. Photo of APS‐Digester System in the Laboratory ......................................................... 240 Figure 114. pH of the Five Reactors in the APS‐Digester System During the Digestion Of Green

Waste in the APS‐Digester System During the 72‐Day Startup Period ................................. 243 Figure 115. Cumulative Biogas and Methane Production in the APS‐Digester System at BR/HRs

of 0.93 and 0.55 ............................................................................................................................... 243 Figure 116. Daily Biogas Production in Five Reactors of the APS‐Digester System During

Digestion of Green Waste at a BR/HR of 0.55 ........................................................................... 245 Figure 118. Methane and Carbon Dioxide Contents of Biogas Produced in the APS‐Digester

System during Digestion of Green Waste at Two Different BR/HR Ratios .......................... 246

xv

Figure 119. Cumulative Biogas and Methane Production in the APS‐Digester System During Digestion of Green Waste at BR/HR of 0.55 and 0.25. .............................................................. 247

Figure 120. Daily Biogas Production of Five Reactors in the APS‐Digester System During Digestion of Food Waste at a BR/HR of 0.55 ............................................................................. 248

Figure 121. Composition of Biogas Produced From One of the Hydrolysis Reactors During Digestion of Food Waste in APS‐Digester System at a BR/HR of 0.55 .................................. 248

Figure 122. Cumulative Production of Biogas, Methane and Hydrogen During Digestion of Food Waste in APS‐Digester System at a BR/HR of 0.55 (3rd Experimental Run) ................ 249

Figure 123. Particle Size Distribution for Different Materials. ........................................................ 254 Figure 124. Schematic of the Set‐Up for the Respirometry Experiment. ....................................... 256 Figure 125. Cumulative CO2 Produced For Different Materials .................................................... 261 Figure 126. Average Rate of Carbon Dioxide Evolution (Mg CO2 C/Hr/G V.S.) Over the Two

Week Experiment .......................................................................................................................... 261 Figure 127. Cumulative Oxygen Consumption (mg O2/g V.S) Over the Two Week Experiment

for Different Materials .................................................................................................................. 262 Figure 128. Average Rate of Oxygen Consumption of the Three Reactors for Different Materials

.......................................................................................................................................................... 262 Figure 129. Maximum Carbon Dioxide Evolution Index (MCI) for Different Materials ............ 263 Figure 130. Dynamic Respiration Index (DRI) for Different Materials. ........................................ 264 Figure 131. Pilot Digester System Site Plan ....................................................................................... B‐2 Figure 132. Hydrolysis Reactor Vessel Detail ................................................................................... B‐3 Figure 133. Material Handling System for Pilot Digester System – Plan View ............................ B‐4 Figure 134. Material Handling System for Pilot Digester System – Elevation View A ............... B‐5 Figure 135. Material Handling System for Pilot Digester System – Elevation View B ............... B‐6 Figure 136. Layout Diagram of UC Davis Biogas Energy Demonstration Plant ......................... B‐7

LIST OF TABLES

Table 1. Feedstock Characterization Summary ..................................................................................... 8 Table 2. Heat load summary of UC Davis Pilot Digester System ..................................................... 14 Table 3. Exhaust Gas Analyses and Reference Methods .................................................................... 29 Table 4. Engine CHP Measurement Locations ..................................................................................... 31 Table 5. Chemical Composition of Food‐Waste Samples (Carbon, Nitrogen, Carbon: Nitrogen

Ratio, Protein, Crude Fat, Phosphorus, and Potassium) on a Dry‐Mass Basis ....................... 68 Table 6. Solids Content Analysis for all Campbell’s® Food Waste Loaded into the APS Digester

System (MC = Moisture Content, TS = Total Solids, VS = Volatile Solids) .............................. 71 Table 7. Summary of Commercial Anaerobic Digester Technologies with Large Scale Reference

Plants. Data from Company Websites as of February, 2008 And Adapted from Nichols [13]. .......................................................................................................................................................... 100

Table 8. Published biogas and Methane Yields for Full‐Scale Digesters Treating OFMSW from Various Sources ............................................................................................................................. 108

xvi

Table 9. Mass Balance Calculations for Several Full‐Scale Digesters in the Literature ................ 110 Table 10. Assumptions Used in the Literature for Modeling the Economic Performance of an

APS Digester System ..................................................................................................................... 116 Table 11. Standard Composition of Natural Gas as Compared with Untreated Biogas. Adapted

from Krich Et al [64] ...................................................................................................................... 119 Table 12. Pressure and Energy Requirements for Storage of Purified Biogas. Adaptedfrom Ross,

et al [59] ........................................................................................................................................... 121 Table 13. Scenarios Evaluated and Unit Processes Compared Using the Financial and

Mass/Energy Balance Model ........................................................................................................ 124 Table 14. Cost Categories Considered for the APS Digester System Cost Calculator.................. 146 Table 15. List of Pre‐Construction Cost Items Used for Estimating the Pre‐Construction Cost of a

New Stand‐Alone APS Digester System Project ....................................................................... 146 Table 16. Design Parameters Used to Size The APS Digester System and Compute the Amount

and Volume of Waste Treated ..................................................................................................... 149 Table 17. Characteristics and Anaerobic Degradability of Municipal Food Waste (FW) and

Green Waste (GW) Collected from Tte Bay Area in California .............................................. 150 Table 18. Base Assumptions Used for Calculating Mass Balance onNon‐Combustion Biogas

Processing Equipment .................................................................................................................. 155 Table 19. Base Values Assumed for Modeling the Heat Production and Consumption of the APS

Digester System ............................................................................................................................. 157 Table 20. Base Values Assumed for Modeling the Electricity Production and Consumption for

the Digester and Biogas Upgrade Unit ...................................................................................... 158 Table 21. Assumed Costs Used to Calculate the Capital Costs for the APS Digester System .... 159 Table 22. Assumed Costs Used t Calculate the Running Ordinary Expenses for the APS Digester

System ............................................................................................................................................. 162 Table 23. Prices (in 2007 Dollars) Used for Calculating Revenues for the APS Digester System164 Table 24. Financial Assumptions Used for Calculating Cash Flows .............................................. 166 Table 25. Comparison of the Current Model with Previously Published Results of the APS

Digester System ............................................................................................................................. 169 Table 26. Overall Digester Performance Results ............................................................................... 170 Table 27. Compositional Breakdown of Biogas a Several Points in the Digester Process ........... 174 Table 28. Fate of Biogas, Methane, and Carbon Dioxide Generated by the APS Digester System,

and the Overall Greenhouse Gas Balance for the System ....................................................... 175 Table 29. Energy Balance for Each Scenario (Percentages Calculated Per Total Energy Available

in the Biogas) .................................................................................................................................. 176 Table 30. Comparison of Overall Financial Performance of the Three Scenarios Analyzed (for a

Discount Rate of 4.42 Percent Y‐1) ............................................................................................... 179 Table 31. Results of the Incremental IRR Analysis ( Percent y‐1) ..................................................... 185 Table 32. Effect of the Renewable Electricity Producer (REP) Credit, the Nonconventional Fuel

(NCF) Credit, and the Alternative Vehicle Fuel (AVF) Credit on the Net PresentWorth (NPW) and Cost of Waste Treatment (COT) ............................................................................. 187

Table 33. Waste Quantities and Characteristics ................................................................................ 204 Table 34. Digester Heat Transfer Calculations ................................................................................... 210

xvii

Table 35 Overall Thermal Energy Requirements .............................................................................. 211 Table 36 Maintenance Heat Exchanger Requirements ..................................................................... 212 Table 37 Warm‐Up Heat Exchange Requirements ............................................................................ 212 Table 38 Mass Balance Calculations .................................................................................................... 213 Table 39 Energy Balance Calculations ................................................................................................. 214 Table 40. FPP Waste‐to‐Energy Capital Cost Opinion ..................................................................... 216 Table 41. FPP Waste‐to‐Energy Annual Operating Cost Opinion ................................................. 217 Table 42 Characteristics of Green and Food Wastes Reported in the Literature .......................... 219 Table 43. Elemental Composition of Food and Green Wastes ......................................................... 224 Table 44. Characteristics of the Feedstock Used in the Experiments with Standard Deviation

Between Brackets ........................................................................................................................... 226 Table 45. Average Biogas and Methane Yields and Biogas Composition for Food and Green

Waste. The Numbers Between Brackets are Standard Deviations ......................................... 233 Table 46. Experimental Design and Substrate Characteristics Used (GW = Green Waste, FW =

Food Waste) .................................................................................................................................... 238 Table 47. Characteristics of the Thermophilic Seed Sludge for Biogasification Reactor .............. 241 Table 48. Experimental Design and Performance of the APS‐Digester Under Different

Conditions (GW = Green Waste, FW = Food Waste) ................................................................ 249 Table 49. Initial and Final Moisture Content of the 1:1 Mix, Final pHof the 1:1 Mix, the Final

Bulk Density of the Mix, and the Initial VS/TS Ratio of Each Type of Waste With Potting Soil ................................................................................................................................................... 255

Table 50. Elemental Content (Percent of Dry Solids) of the Fresh and Digested Food and Green Waste Is Shown .............................................................................................................................. 257

Table 51. Heavy Metal Content (mg/L) of the Fresh and Digested Food and Green Waste ...... 258 Table 52. Compost Suitability Parameters For The Digested And Fresh Food And Green Waste

.......................................................................................................................................................... 260

1

EXECUTIVE SUMMARY

Introduction Over 2.6 million tons of green waste and 5.9 million tons of food waste are generated each year in California. Converting these waste materials into bio‐energy and other valuable products will provide renewable energy as well as environmental and economic benefits. Anaerobic digesters can convert a variety of organic materials into a biogas that contains 50 to 80 percent methane, which can be readily used as a fuel for electrical power and heat generation, or as raw material for a number of industrial products. Well‐designed and functional anaerobic digesters would make significant contributions to the development of a renewable, distributed energy supply in California. However, advanced anaerobic digestion systems that are reliable, highly effective and simple to operate and maintain have yet to be developed and demonstrated for the conversion of high solid wastes.

Project Purpose This goal of this project was to scale‐up, test, and demonstrate a new anaerobic digestion technology called the Anaerobic Phased Solids Digester system (APS‐Digester). The APS‐Digester is an advanced technology with innovative design features that optimize the bacterial degradation of organic wastes, provide efficient material handling solutions, and combine the favorable features of both batch and continuous operations in a single biological system. The APS‐Digester system had been proven to be reliable and stable at the laboratory level for converting various types of organic wastes to biogas. However, material handling, actual waste stream characteristics, digester performance, residual material characteristics and net value at a larger scale had yet to be shown. The testing and analysis of the APS‐Digester at a scale greater than bench‐scale was needed prior to full‐scale commercialization.

Project Results UC Davis developed, constructed, and tested a pilot‐scale anaerobic digester system based on the anaerobic phased solids digester technology with funding from the California Energy Commission. UC Davis also conducted research, demonstration, and technology transfer activities. The anaerobic phased solids digester technology has two issued patents (U.S. Patent 6,342,378 and 7,556,737) and one patent pending. It has innovative design features that provide excellent conditions for bacterial degradation of organic wastes and efficient material handling solutions, and also combines the favorable features of both batch and continuous operations in a single biological system.

The project commenced in mid‐2004. An engineering design and test plan of the pilot scale APS‐Digester system were completed in March 2005 and a research report was submitted to the Energy Commission. The primary construction of the pilot digester system was completed in September of 2006. The start‐up ceremony was held on October 24, 2006, and more than 300 people attended, including representatives from UC Davis, state and federal agencies, and various companies. More than 10 news media organizations also attended the ceremony and

2

reported widely on this event and the UC Davis Biogas Energy Project. The construction of the pilot digester facility continued as material handling tests were performed with different types of biomass feedstock (food waste, dairy manure, and straw) and was completed in early 2008. Another research report was submitted to the Energy Commission. The construction costs for the pilot digester facility exceeded the planned budget, and funds were raised from UC Davis, Onsite Power Systems, Inc., and other partners. The total cost of the pilot digester facility was more than $1.5 million. The pilot digester system was housed in the newly developed UC Davis Biogas Energy Plant and had the capacity to treat three to eight tons of organic waste per day with an expected biogas production of 11,400‐22,900 cubic feet of biogas per day. The formal testing of the pilot digester system was conducted between June 2008 and May 2009 using food processing waste from the Campbell’s® processing facility in Sacramento, California. The digester system achieved steady‐state biogas and methane yields of 500 and 275 cubic meters per ton of volatile solids in the food waste, respectively, at an organic loading rate of 2.0 grams of volatile solids per liter reactor volume per day (gVS/L/d). High solids reduction rates of 80‐95 percent were sustained throughout one year of continuous digestion.

The research and demonstration of the anaerobic phased solids digester system at a pilot scale was highly successful. UC Davis licensed the anaerobic phased solids digester technology to Onsite Power Systems, Inc., and Clean World Partners, LLC for commercial development. More than 3,000 domestic and international visitors toured the pilot digester facility during this project. The UC Davis Biogas Energy Project was used by the university as a showcase for innovation, technology transfer, and public education and was widely reported in various news media and magazines (such as Biomass and BioCycle). It was featured in the UC Davis Centennial Exhibition at the 2008 California State Fair and was recognized as one of 100 ways that UC Davis has transformed the world as part of the university’s centennial celebration (From Trash to Gas, http://centennial.ucdavis.edu/).

Commercial development of the anaerobic phased solids digester system was carried out with several commercial partners located in California, including Grand Central Recycling and Transfer Station in City of Industry, Campbell’s® in Sacramento, and Garden Highway Foods in Sacramento. An anaerobic phased solids digester system capable of treating 25 tons of food waste per day was designed for Campbell’s® soup processing plant and the design report was submitted to the Energy Commission. Due to the time required for acquiring the capital and permit for constructing this facility, the development of Campbell’s® Soup Digester could not be completed before the end of this project contract.

In addition to the design, construction, and testing of the pilot anaerobic phased solids digester system, laboratory research and testing were conducted to characterize the food waste and green waste that were intended for treatment at the pilot digester facility. Two reports were submitted on the laboratory research. An engineering design model was developed and used to scale up the anaerobic phased solids digester system from lab and pilot scale (1–3000 kilograms of wet waste per day) to full scale (115 tons per day) based on laboratory and pilot‐scale research and development.

3

A cash‐flow model was integrated for calculating net present worth and internal rate of return. The incremental rate of return was also calculated for three biogas usage scenarios. A revenue‐requirements model was integrated to calculate the constant‐dollar levelized annual cost of energy and cost of waste treatment for the system. The model was used to compare three biogas usage scenarios: (1) converting as much biogas to electricity and heat as possible via an engine generator with heat recovery (cogeneration) while providing the heat required to keep the system at 55 degrees Celsius (°C); (2) converting enough biogas directly to heat using a boiler to keep the system at 55°C while converting the remainder to renewable natural gas via a pressure swing adsorption unit; and (3) converting enough biogas to electricity and heat using cogeneration to provide the heating and electrical requirements of the system (with additional biogas converted to heat via a boiler if needed) while converting the remainder to natural gas via a pressure swing adsorption unit. This engineering design and economic analysis model will be a very useful tool for developing future commercial projects.

The research reports submitted to the Energy Commission as well as a list of patents and scientific publications that resulted from this project is presented in Appendix A. These reports and publications contain the details of research, development, and demonstration activities.

Project Benefits The successful demonstration of the APS‐Digester proved the technical feasibility of converting solid waste to biogas. This technology benefits California by providing a cost‐effective source for renewable energy as well as contributing to reduced greenhouse gas emissions.

5

CHAPTER 1: Pilot Digester Design and Test Plan 1.1 Introduction Over 2.6 million tons of green waste and 5.9 million tons of food waste are generated each year in California. Converting these waste materials into bio‐energy and other valuable products will provide renewable energy as well as environmental and economic benefits. Anaerobic digesters can convert a variety of organic materials into a biogas that contains 50 to 80 percent methane, which can be readily used as a fuel for electrical power and heat generation, or raw material for a number of industrial products. Well‐designed and functional anaerobic digesters would make significant contributions to the development of a renewable, distributed energy supply in the State. However, advanced anaerobic digestion systems that are reliable, high‐rate and simple to operate and maintain are yet to be developed and demonstrated for the conversion of high solid wastes.

This project will scale‐up, test, and demonstrate a new anaerobic digestion technology called the Anaerobic Phased Solids Digester system (APS‐Digester) (U.S. patent 6,342,378). The APS‐Digester technology was developed at the University of California, Davis and is an advanced technology with innovative design features that optimize the bacterial degradation of organic wastes, provide efficient material handling solutions, and combine the favorable features of both batch and continuous operations in a single biological system. The APS‐Digester system has been proven to be reliable and stable at the laboratory level for converting various types of organic wastes to biogas. However, material handling, actual waste stream characteristics, digester performance, residual material characteristics and net value at a larger scale have yet to be shown. The testing and analysis of the APS‐Digester at a scale greater than bench‐scale is needed prior to full‐scale commercialization. A pilot scale APS‐Digester of 3 tons per day capacity will be constructed and tested at the University of California, Davis, using recycled green and food wastes supplied by a commercial waste collection company.

1.2 Project Overview

1.2.1 UC Davis Biogas Energy Demonstration Project The Onsite Power Systems design team (the TEAM) has researched and evaluated the construction and operational design requirements for a 3 ton per day APS‐digester and biogas energy Commercial Demonstration System (CDS) for the UC Davis Biogas Energy Demonstration project. Feedstock parameters were based on a comprehensive feedstock characterization study. The design, material handling and operational specifications were based on the information provided by the UC Davis research team (UCD RT) and data collected from recent laboratory test results. Additional design criteria were developed based on the information gathered from site visits to the Hay Road Compost Facility of Norcal Waste Systems, Inc. in Dixon, CA.

6

During the engineering design stage of this project several tasks were accomplished. A summary of the tasks completed to determine the operational parameters is as follows:

• Initial CDS engineering design and modifications were completed and reviewed for potential operational and economic enhancements.

• CDS computer control system was developed and modifications made to facilitate data collection criteria as determined by the UC Davis research team (UCD‐RT).

• CDS modifications were identified to enhance public education and commercial demonstration programs.

• Additional UC Davis Laboratory analysis was conducted on alternative feed stocks and design modifications were based on the supplemental test results.

• Thorough engineering review of final CDS design with all modifications was conducted to ensure the highest probability of a successful design.

1.2.2 APS-Digester System Overview The APS‐Digester developed at UC Davis is a multiple reactor, two‐phased, sequenced, batch‐fed solids digester capable of producing a relatively constant biogas production rate. Normal (commercial scale) system operation will be described, followed by the operating differences for the pilot scale plant.

The APS‐Digester is a high solids, high rate digester capable of handling most organic materials regardless of moisture content or physical characteristics. Figure 1 shows the schematic of the APS‐Digester System. The system is divided into two phases with a semi‐continuous circulation of liquid between the vessels. Feed material is loaded into the hydrolysis reactor, acted on by extra‐cellular enzymes and acidogenic bacteria, thereby liquefied and converted to simple organic acids. These acids are transferred to the bio‐gasification reactor, where they are reduced further into biogas by methanogenic bacteria. The biogas is a flammable mixture of methane (CH4) and carbon dioxide (CO2). Multiple hydrolysis reactors allow for a time separation between the beginnings of the batch hydrolysis reactions. This time separation contributes to a level biogas production rate despite the batch loading and operational schedule.

The APS‐Digester system is designed for the organic material to be loaded into one of several hydrolysis reactors over a period of days. The nominal system operation designed for a 12‐day retention time in each hydrolysis reactor is for one of four hydrolysis reactors to be loaded and sealed every three days. The material is saturated with pre‐heated water and warmed to the operating temperature, digestion begins, and the reactor liquid is circulated to the bio‐gasification reactor. A hydrolysis reactor for a commercial system would be loaded daily, for three days, and then sealed after the final load. Depending on the feedstock and operating conditions, methane production can occur in both the hydrolysis and bio‐gasification reactors. This may be due to transferring of methanogenic bacteria in the liquid re‐circulated to the hydrolysis reactors. The feedstock material is retained in the hydrolysis reactor for a

7

predetermined period of time (12 days nominal) and then after digestion, the remaining material is removed through a screw press for residual solids dewatering and water recovery.

Figure 1. A schematic of APS-Digester System

Water Recirculation from Gasification Tank

Buffer Tank

Soil Amendment

Nutrient Enriched WaterSolids Separator Residue Material

& Reclaimed Process Water

Tank Drainage System Hydrolysis

3

BiogasificationReactor

BiogasWater Gravity Fed toBuffer Tank

Hydrolysis 2

Hydrolysis 1

Hydrolysis 4

The APS‐Digester pilot plant at UC Davis will be operated similarly to a commercial size system with only minor differences due to reactor scale. To load a 3‐ton per day system daily would require extensive labor and shipment of partial truckloads of material. To mitigate these difficulties, the pilot system will be loaded once every three days, with the liquid recycled and loaded along with the new material for direct heat transfer. The loading sequence begins with the oldest hydrolysis reactor being drained through a screw press. The solids will be placed in an empty dumpster for return to the compost facility for further handling while the liquid is transferred to a holding tank. This water will be heated over the next 16 hours to an elevated temperature sufficient to raise the temperature of the feed material to the system operating temperature. The hot water will then be used to transfer the required warm‐up heat to the cold feed while helping carry it into the reactor. After the determined digestion time, the tank is drained, solids removed and the cycle repeated. Each hydrolysis reactor will be similarly cycled with a 3‐day separation between loadings.

The following pilot plant engineering design is based on the operational sequence cited above. All effort has been made to minimize differences in the design and operation of the pilot plant from the commercial scale concept. The largest difference will be in the reactor heat loads. The surface area to volume ratio improves with larger tanks and will result in decreased heat loss per reactor volume. Also, as the pilot digester will be fed once every three days, vs. daily for

8

three days, water pre‐heating will be required to be completed in 16 hours vs. 3 days. These differences are not expected to materially affect the overall plant performance.

1.2.3 Feedstock Characteristics In order to properly design a digester, the physical and chemical characteristics of the feedstock must be known. Critical parameters such as moisture and volatile solids content, as well as bulk density, digestibility, and biogas yield significantly affect digester design and operating procedures. Green waste and food waste studies were completed on representative feedstock supplies, and the overall variability of the material was determined. The final assessed average values are summarized in Table 1. A separate report will be prepared on the complete feedstock characterization study.

Table 1. Feedstock Characterization Summary

Parameter Green Waste Food Waste

Moisture content, (% wet base) 70 75

VS, (% wet base) 25 23

VS/TS, (%) 85 86

Bulk density (yd3/ton) 4 2.5

1.3 Pilot Digester System Design Considerations The feedstock characteristics listed in Table 1, along with previous laboratory results, were used to determine the pilot plant design criteria. As the hydrolysis reactor volume limits the total system capacity, all subsequent design parameters are based on the available tank capacities. Four, 10,000 gallon used stainless steel wine tanks were identified and purchased for the project. This volume established the overall loading capacity of the system as defined by the bulk density of the feedstock. As green waste materials allow for very stable digestion, they were chosen as the initial loading material. The first trials will be conducted on 100 percent green waste material, and the initial capacity designations are based on this figure. From previous studies, the average bulk density of green waste is 4‐yd3/ton. Allowing for a 15 percent headspace, the reactors will have a 42‐yd3 capacity, or 10.6 tons of green waste per loading. Based on a 14‐day cycle (no weekend loading), with 1 day for loading and 1 day for unloading, the material will be held in the hydrolysis reactor for 12 days. As one reactor will be fed every 3.5 days (Monday or Thursday loading), the total plant loading would be 3.0 tons per day. The bio‐gasification reactor is equal in size to the hydrolysis reactors, 10,000 gallons (38,000 liters). To ensure this phase of the digester will not be overloaded, the organic loading rate (OLR) was calculated. In the laboratory scale studies of green waste, more than 50 percent of the biogas generation was from the hydrolysis reactors. Therefore, it is assumed that less than 50 percent of the daily volatile solids (VS) will be introduced to the bio‐gasification reactor. The VS content

9

of the green waste is assumed to be 25 percent of the wet weight of the feedstock, or 0.75 tons per day. Therefore, the bio‐gasification reactor will be fed with less than 340 kg [VS] per day for a total OLR of 9.0 g [VS]/l/day.

1.3.1 Mass Balance of the Pilot Digester System A total account of the mass in and out of the APS‐Digester was conducted to estimate output streams. Figure 2 shows the mass balance performed for a 10 ton‐batch of green waste. The assumptions used included an initial feed moisture content of 70 percent, with an 80 percent VS/TS ratio and a 70 percent VS destruction rate, and are set from previous laboratory research with a similar feedstock. For 10 tons of wet material fed per batch, the output streams are estimated to be 5.0 tons of recoverable water (1210 gallons), 3.3 tons of residual solids at 60 percent moisture content, and the destroyed VS (1.7 tons) being converted to biogas.

An estimate was made for the biogas density to evaluate the total mass balance. Assuming 6 ft3 of methane per pound VS fed, a 65 percent methane content biogas saturated with water vapor at 135 °F, approximately 44,308 ft3 (1.7 tons) of biogas will be produced per 10 tons of wet material. This mass is equal to the VS destruction and therefore equates to a total accounting of the mass in and out. If the initial moisture content of the feedstock is adjusted to 90 percent prior to the start of digestion, 2 tons of recovered water will be needed for this purpose, leaving 3 tons of recovered water as effluent from the digester system.

Figure 2. Mass balance of UC Davis Pilot Demonstration Digester System

Wat

er In

(7

tons

)

VS

In

(2.4

t)

Tota

l Mas

s In

(1

0 to

ns) Residual

Solids3.3 tons

(2t water)(1.3t TS)

Biogas(1.7 ton VS)

Recoverable Water(5.0 tons)

10

1.3.2 Energy Balance of the Pilot Digester System An energy balance was performed for summer and winter seasons. Figure 3 (a and b) shows the energy balance for a 10 ton‐batch of green waste during summer and winter seasons, respectively. Based on the high heating value of the VS (for example 18.75 MJ/kg [VS]), a total energy input of 388 therms could be calculated. The total energy content in the biogas produced is estimated to be 288 therms which represents 74 percent of the total energy input. The net electricity produced will be 69 therms, which represents about 24 percent of the biogas energy. During the summer season (Figure 3a); net recoverable heat represents about 28 percent of the biogas energy. So, an overall net conversion efficiency of 52 percent of the biogas energy could be expected. Unrecoverable heat losses and backloads of the digester system account for about 24 percent of the biogas energy in each.

The overall net conversion efficiency for winter seasons (Figure. 3b) is estimated to be 47 percent of the biogas energy, lower than the efficiency for summer seasons. This is due to the fact that the initial temperature of the feedstock was assumed to be 40°F in winter compared to 60°F in summer season. Thus, more energy will be required during winter to raise the initial temperature of the feedstock to the digestion temperature (135°F). Moreover, higher energy losses to the environment could also be calculated as the outside air temperature was assumed to be 32°F in winter compared to 60°F in summer. The unrecoverable heat losses and the net recoverable heat represent about 29 percent and 24 percent of the biogas energy, respectively.

Figure 3. Energy Balance for Summer (A) and Winter (B) Seasons

Bio

gas

Ene

rgy

(288

ther

ms)

Tota

l Ene

rgy

In

(388

ther

ms)

Bacterial growth and undigested residuals (100 therms)

Net Electricity (69 therms)

Net Recoverable Heat

(81 therms)

Unrecoverable Heat Losses (69 therms)

Gro

ss(2

19 th

erm

s)

Backload(69 therms)

Bio

gas

Ene

rgy

(288

ther

ms)

Tota

l Ene

rgy

In

(388

ther

ms)

Bacterial growth and undigested residuals (100 therms)

Net Electricity (69 therms)

Net Recoverable Heat

(68 therms)

Unrecoverable Heat Losses (82 therms)

Gro

ss(2

06 th

erm

s)

Backload(69 therms)

A B

11

1.4 Pilot Digester Component Design Specifications The pilot plant will be constructed on a concrete pad previously installed for a different pilot design. The new design is a scale version of a commercial plant and will require additional tankage over the previous system. Therefore, additional concrete will be installed with allowances made for material handling and sufficient room for process equipment. Each sub‐system is described below with critical design criteria and operating parameters.

1.4.1 Concrete Pad A concrete pad will be constructed to accommodate the pilot digester system. This area will consist of two sections, one that is 35 feet x 58 feet and an adjoining section that is 47 feet x 65 feet. The larger section will be used to hold the material receiving, processing and residue recovery equipment. A section of this area contains a sloped area that will allow the floor at the back end of the roll‐off to be level with the rest of the pad and allow a skid‐steer to enter the bin without a ramp. Typical roll‐off bin floors are 8 inches off ground level. The designated area is 21‐feet wide x 20‐feet deep. The sloped area begins 5 feet south of the north access road and slopes down to the south for 20 feet to a drop of 5 inches. The sloped area begins 5 feet from the east side of the pad and continues 21 feet to the west.

The South 14 foot by 35 foot area will hold the biogas storage tank, clean‐up and processing equipment and generator system. The generator will be installed within a metal screen enclosure for personnel safety and equipment security. An isolated sidewalk section will be installed to the east of the existing site. The walk will provide an entryway to the CDS that will be clear of all equipment and operations. This walkway will be used for tours of the facility and will connect to the existing site in front of the controls building. An additional area from the controls building to the viewing stairs and catwalk will be created. The sidewalk will begin 21 feet to the east of the north concrete pad addition. The sidewalk will be 4 feet wide and run 30 feet to the south at which point it will turn west at a 45 degree angle and extend 40 feet to the concrete pad.

The area located between the sidewalk and the concrete pad will be landscaped with project and sponsorship display boards installed. The panels will be 4 feet high x 8 feet long, mounted on a metal support frame, and set in a raised planter area 5 feet x 14 feet x 12 inches above grade. The display panels will have a brief overview of the APS process and list in detail the various sponsors of the project through their funding or donations of equipment and services.

1.4.2 Reactor Vessels The physical design of the system is based on using four Hydrolysis Tanks (HT), each having a working volume of 8,500 gallons. The selected tanks are 10 feet‐8 inches in diameter and 16 feet high and set the total system capacity at 3 tons per day for green wastes. The bio‐gasification reactor volume is equivalent to one of the hydrolysis reactors and a suitable used vessel has been identified and procured. All five tanks are installed on the west side of the pad. The hydrolysis reactors (Appendix B) have sloped bottoms and each contains a full‐height water

12

jacket. Each tank will be insulated with ceramic insulating paint, as will all pipes and fittings that will contain the warm digester liquids.

The tanks are mounted on steel stands that allow a minimum of two feet of clearance under the tanks for construction and future modifications. Two pipe‐fittings will be installed in the bottom of the tank to be used to connect the hydrolytic mixing system. Each pipe connection will be located on a centerline halfway between the tank center and side, perpendicular to the drain line.

The existing 8,400‐gallon tank will be converted to an effluent water tank (RWT) and will be used to store reclaimed water. This tank will be insulated and connected to the heat exchanger system for water pre‐heating before being fed to the newly loaded hydrolysis reactor. The existing steel buffer tank will be replaced with an 11‐foot high, 900‐gallon stainless steel tank, installed toward the west, closer to the HTs.

1.4.3 Heat Load and Heat Exchangers Operating an anaerobic digester system at elevated temperatures increases the rate of material degradation and biogas production. However, heat is required to be added to the reactor vessels to maintain the operating temperature, as well as a significant amount of heat is required to be added for the initial heating up of the feedstock from ambient to operating temperature. These two heat loads are referred to as the maintenance and warm‐up heat, respectively. Specific calculations are included in the appendix and summarized in Table 2.

1.4.3.1 Maintenance Heat The reactors for the pilot plant are jacketed vessels constructed of stainless steel. The circulation of hot water through the jacket eliminates the need for external heat exchangers to meet the maintenance heat requirements to offset the normal heat loss from tanks to the environment Individual vessel temperature control will be accomplished through the jacket water flow rate. Water from a constant temperature source will be circulated through pipes connecting the vessel jackets in parallel. Each reactor will have a thermal control system to regulate the flow of hot water through the jacket. The higher the flow rate, the higher the heat transfer rate.

The critical variable regulating the operating heat loss is the temperature difference between the liquid inside the vessel and the outside surface temperature of the vessel (for example driving force of heat transfer). As the heat loss will occur from the jacket water, the maintenance heat load is based on a jacket water inlet temperature of 140 °F, and its difference from the external tank surface temperature, via conduction through any insulation. Spray‐on insulation with an R‐value of 12 hr.ft2.oF/Btu will be applied to all vessels, valves and piping. As a conservative, maximum, heat loss estimate, the tank surface is assumed to be at ambient temperature. This estimate assumes no convective resistance to heat transfer, and any radiant losses resulting in a surface temperature below ambient being countered by convective heat gain. For the worst case and to further determine the maximum maintenance heat load, the coldest winter night temperatures are used for the surface temperature. Temperature data was collected from

13

historic meteorological records, with a 32 °F minimum temperature used for the winter season. The maintenance heat requirements were determined for seasonal averages to indicate the range of heat requirements throughout the year. The maintenance heat (Qm) can be calculated as follows:

)( sjm TTAUQ −=

Where U is the overall heat transfer coefficient in Btu/hr.ft2.°F, A is the surface area of the reactor in ft2, Tj is the jacket temperature (140°F) and Ts is the temperature of the reactor surface (32°F).

1.4.3.2 Warm-Up Heat To address the initial material warm up efficiency, the following engineering operational parameters will be designed into the digester demonstration system. When a hydrolysis reactor is unloaded, the existing heated water is reclaimed and will be stored in the reclaimed water tank. This reclaimed water will be heated to approximately 155°F prior to the reloading of the hydrolysis reactor. The heated reclaimed water will mixed with the new material during the loading process. This will raise the temperature of the feedstock to approximately 125°F to 130°F. In addition, the reclaimed will contain residual bacteria that will help to inoculate the new feedstock.

The principle thermal loads are for initial material warm up and digester maintenance heat requirements. The greatest thermal load is the initial heating up of the feedstock material from ambient to the operating temperature (Appendix B). The heat required to raise the temperature of the material to the digestion temperature (Tf) is described by

)( if TTCmQ −=

Where Q is the heat required to raise the temperature of feedstock in Btu/hr, m is the mass flow rate of the feedstock to be heated in lb/hr, C is heat capacity of feedstock material in Btu/lb‐F, and Ti is the final and initial temperature of feedstock in °F. As the feed material has reasonably high moisture content (62 percent), the heat capacity (C) of the material is assumed to be equal to that of water, 1 Btu/lb.°F. As the system will be operated as a thermophilic reactor, the digestion temperature (Tf) to be achieved is 135°F. To determine the maximum heat rate required, and therefore heat exchanger capacity, the assumption is made to use the coldest average ambient temperature during the year. In the Sacramento/Davis area, the minimum average daily ambient temperature for December and January, as recorded by local weather stations, is taken as 40 °F. This temperature is used as the initial average temperature (Ti) of the feedstock material.

14

Table 2. Heat load summary of UC Davis Pilot Digester System

Heat Load Winter Spring/Fall Summer unit Description

Maintenance 60067 52281 44494 Btu/hr Operating Heat Loss (through jacket)

Heat-Up 118750 106250 93750 Btu/hr Material Warm-Up (Mon/Thur night only)

Total Required 178817 158531 138244 Btu/hr Total Heat Required (Mon/Thur)

Gen Duty 39 percent 58 percent 77% Allowable generator load

Operating 9.3 13.8 18.4 hrs/day Allowable Operating Hours (full load)

Gas Heat 135048 93202 51354 Btu/hr Heat from direct biogas combustion

Waste Heat 43769 65329 86890 Btu/hr Waste heat from generator

Available Heat 178817 158531 138244 Btu/hr Total available

Make-up Heat 0 0 0 Btu/hr Supplemental Heat Required

The above calculations include material warm‐up requirements (Mon and Thur night only). The maximum maintenance heat requirement is 60,067 Btu/hr (winter nights). From the table, the maximum anticipated heat load would be approximately 180,000 Btu/hr during the 16 hours prior to each loading, every 3 days (Monday and Thursday nights only). The recoverable waste heat from the onsite generator will not be sufficient for this demand and hence some biogas will be required to be burned directly for the additional heat requirement. Therefore, the generator will not be able to be operated at full load for the entire warm up period. The above table suggests the maximum load for continuous operation or number of hours of full load operation the generator can be run on the warm up days (Monday and Thursday nights) by season. Unrestricted generator operations are allowed all other times (136 hours per week or 80 percent of total available hours).

1.4.3.3 Heat Exchangers Maintaining a system temperature of 135 °F is critical to the production of biogas and life of the thermophilic bacteria. A 125,000 Btu per hour liquid‐to‐liquid heat exchanger will be used to maintain the temperature. To avoid the potential destruction of bacteria, reactor fluids should not be exposed to heat exchanger surfaces that exceed 150 °F. The gasification tank (GT) will have an independent shell and tube heat exchanger system, which will use a standard hot water heater as the heat source. The designed heat exchanger will be mounted vertically on the GT, with the existing circulation pump used to connect the heat exchanger and heat source piping systems.

15

All other tanks will share the HT heat exchanger. Each hydrolysis tank has a full water‐jacket to serve as the tank heat exchanger component. A shell and tube heat exchanger operating off generator waste heat will provide the hot water source to the tank water jackets. The heat exchanger will be installed next to the source of heat and will utilize an existing pump for circulation. A separate electric heating element will be included for back up heating.

1.5 Pilot Digester System Operational Design

1.5.1 Hydrolysis Reactor Loading The demonstration plant is designed to use grass and food waste with various combinations of each as the primary feedstock. The operational plan is to load one tank every three days and provide for a predetermined solids retention time. Feedstock will be available every 3‐4 days from Norcal Waste Systems, Inc. who will deliver a roll‐off bin with fresh material on Monday and Thursday. Potential testing modifications will include the delivery of ground and un‐ground material. Based on information gathered from site visits to the composting operation in Dixon, CA, Norcal Waste Systems, Inc. will process material through a Trommel‐screen to remove gross contamination prior to delivery to the pilot plant.

A new feedstock roll‐off bin will be left on site and the dewatered residual materials will be returned to the compost operation for further conditioning, if necessary. The plant operator will have one day to load the material in the hydrolysis reactor. The empty roll‐off bin will then be used to hold the residue material recovered during the next unloading process 2 days later. Once a HT is unloaded, Norcal will deliver a second roll‐off of green materials the next day. The same driver will pick up the first roll‐off with the residue materials and the cycle continued, limiting feed delivery to only two times per week.

Assuming a ¼ cubic yard bucket, unloading the 40‐yard bin will require 160 trip‐cycles of the skid‐steer. Assuming the bin will be unloaded in 4 hours, each cycle would be limited to 1.5 minutes. The material receiving and processing system will be required to handle 4.5 cubic feet per min. of feedstock.

The operational schedule is as follows:

Monday Empty selected HT, residue and process water recovery

Tuesday Load empty HT, 40 cubic yards or 10 tons

Wednesday Monitoring operations

Thursday Empty selected HT, residue and process water recovery

Friday Load empty HT, 40 cubic yards or 10 tons

Saturday Monitoring operations

Sunday Monitoring operations

16

Modifications to the material receiving and processing system may be required as the feed mixture is shifted over to greater percentages of food. On an equal wet weight basis, food waste is close to half the volume of grass. These modifications, if any, will be determined as current operational performance dictates. Loading the feedstock materials into the tanks will be accomplished by a metered feeding system as follows:

• The material will be transferred from the roll-off bin to a receiving hopper using a skid-steer tractor with a front loader. The receiving hopper will be a metering auger with hopper. The auger will moderate the material flow into an elevating auger.

• Elevating auger feeds material to a water‐float separator system allowing metals, rocks and other heavy materials to drop out.

• The water‐float system will transfer material onto a belt conveyor system that will serve as an observation line to remove plastics and other inorganic materials as desired. The belt conveyor will deliver the material to a hydraulic piston pump that will transfer the material through an 8 inch feed line into the lower section of a specific hydrolysis tank selected for loading.

• At the north end of the belt conveyor, a separating vibrator bed will be installed. This allows material to spread‐out over the bed to provide random physical observation of the materials and inspection for non‐organics materials.

To avoid overfilling the tank with make‐up water it will be necessary to maintain a control mechanism that will draw down fluid from the tank being filled or limit the liquid being introduced into the feedstock hopper/hydraulic piston pump to maintain tank level and pressure.

1.5.2 Liquid Circulation Process Two screen filter systems with the design ability to be a self‐cleaning filter system will be installed on each HT. The circulation system will be constructed with stainless steel piping, 150 # SS flanges and fittings. Circulation valves selected are 2 inch 150 # flanged ball valves 316 SS with electric actuator and position switch. Two existing electric pumps will be utilized for the circulation system. Each valve and pump will be connected to the main computer control system and will operate according to circulation operational protocol.

1.5.3 Hydraulic Mixing System Each HT, and the GT, will include a hydraulic mixing system and a Mazzie injector mixing system. To accommodate the hydraulic system, two fittings will be installed in the tank bottoms. All feed line pipes will be either 1 inch or 2 inch stainless steel (as required) and a 2 inch 150 # flanged ball valve, 316 SS with electric actuator and position switch will be installed on each pipe. The Moyno Series 500 pump and valves will be connected to, and operated by, the main computer control system.

17

The GT will have an independent mixing system that will be identical to the HT mixing systems, with the exception of the pump and chopper system. The GT system will utilize an existing circulation pump. A Moyno Series 500 pump system will be installed between HT1 and HT2 with a second unit between HT3 and HT4. All pipe installations will be located between the two tank pairs. This greatly reduces pipe requirements and allows one Moyno pump to provide the mixing functions for two HT. Each circulation pump will have a variable speed drive installed to be operated through the main computer control system. Each valve will be connected to the main computer control system and will be operated according to the designated operational protocol.

1.5.4 Solids Residue and Water Recovery At the end of each retention cycle, the selected HT will be drained and the solids extracted from the working fluid using a mechanical screw press. The press will be installed on the north end of the main drain line providing a straight line to the press. Water extracted from this process will be sent to the existing 8,400‐gallon reclaimed water tank (RWT). The solids will be transferred to the empty roll‐off bin, which will be picked during the next transfer of feedstock cycle. The existing 8,400‐gallon reclaimed water tank will provide enough storage capacity for the reclaimed water recovery process.

Each HT and the GT will be connected to the 8‐inch drain line through an 8‐inch manual gate valve (150 # RF gate valve cast steel). A 150 # lb flange will be installed on the bottom section of each tank, next to existing 18‐inch man‐way. An 8‐inch cast steel pipe will tie‐into the main 8‐inch cast steel pipe drain line. The main drain line will exhibit a required slope to the north section on digester site. The drain line will have a trash pump installed near the north end of the line.

Just after the trash‐pump, the drain line will connect to a screw press. It was determined that the press will provide the solids in a moisture content that will not drip liquid when transferred to the roll‐off box. The water recovery drain from the screw press will be routed to the RWT through a stainless steel pipe system. A secondary water line from the RWT will be connected to the south end of the main 8‐inch drain line to provide the ability to flush out the feed/drain system into the press in a closed loop water transfer cycle.

Residue solids recovered from the screw press will be discharged into a receiving hopper. The receiving hopper will distribute the residue materials into an elevating auger system. The elevating auger system will feed into a horizontal auger system over the roll‐off bin area. A discharge chute will be installed in each auger over the two roll‐off bin locations allowing for a specific bin to be loaded.

Each auger system will be constructed with reversible operation capability. Each drive motor in the residue recovery system, including the screw press will have a variable speed drive connection with operation either the equipment location or in the main computer control center.

18

1.5.5 Tank Pressure/Vacuum Equalization To assure that all tanks remain at their design pressure or vacuum, their head pressures are equalized when fluid is transferred from one tank to another. To accomplish this, an additional manifold, designated the pressure balance manifold (PBM), is incorporated into the system. Each tank has an electronic control valve connected to the PBM located behind the check valve. When fluid transfer from one tank to the other, the two involved tanks will have their gas outlets opened to one another. After the fluid transfer is complete, both valves are closed to allow separation of their biogas streams. Flow readings will not take place during fluid transfers to reduce pressure effects on the measurements.

1.5.6 Safety Pressure Relief Valve Over‐pressure and vacuum limitations are designed through the installation of a spring‐loaded valve in the top of each tank. These specially designed valves do not allow the tank to exceed safe operating pressures or vacuums. Venting of the tank occurs if the pressure exceeds 6‐inches of water column. Conversely, if the tank reaches a vacuum greater than 2‐inches of water column, the valve allows the tank to draw in biogas from the biogas manifold to equalize the tank without drawing in oxygen from the air and creating a potentially explosive mixture of methane and oxygen.

1.5.7 Limiting Introduction of Air into the System Every effort is made to avoid introducing air into the tanks and manifolds that contain methane. During initial filling some air is introduced with the feedstock in the tank. Filling the tank as full as possible with water after introducing the feedstock is recommended to minimize the head of air at the onset of digestion. A natural safety feature in the digestion process is that carbon dioxide is prevalent at the beginning of the digestion and explosive methane and air mixtures are more difficult to form at the beginning of the digestion cycle. The level of each tank is monitored by the central supervisory control and data acquisition SCADA computer, which in turn controls the specific levels through the control of the water supply valves.

1.5.8 Computer Control System A comprehensive computer monitoring and control system was designed for the CDS project. A centralized computer control center will be constructed within the main operations building. The computerized control system will monitor all mechanical components and various processes in the CDS operation from this location. Lighted control panels will provide indication of the operational status of all mechanical, operational and process components. The CDS operator(s) will control and oversee the various processes from this location and make required adjustments and changes as needed. The computer control software and process protocol is proprietary to OPS and OPS will retain ownership of all computer software and process protocols.

The overall digestion process is designed to operate almost entirely automatically by a computer control system and automated controls components. Various circulation and

19

hydraulic mixing cycles are scheduled for every two to four hours through a preset sequence of tanks and processes. Various process protocols identified earlier will be incorporated into the main computer control system. Each of the circulation and hydraulic mixing cycles are programmed for automatic performance at required intervals. Computer control components are designed into all of the valve assemblies, pump and motor systems and piping sub‐assemblies with signal and data connections going to a main computer control center located in the CDS operations building.

The control suite monitors are standard LCD displays and the software allows standard, non‐animated flow diagrams and control functions. A large plasma display screen with animated flow diagrams for demonstration and educational purposes may be installed in the main controls building later in the project.

In addition to the control components, a comprehensive sensor and monitoring system has been designed into numerous components of the CDS tanks, assemblies and sub‐assemblies. This will enable continuous monitoring of the entire CDS and system processes. Monitoring sensor components have been incorporated into the CDS design to monitor:

• Solid and liquid levels in all CDS tanks • Temperature sensors in all CDS tanks • Process flows through circulation piping systems and assemblies • Position indicators on all valve assemblies • Operational performance of hydraulic mixing processes • Operation performance of heat exchanger systems • Gas pressure in all CDS tanks • Sensors for gas leak detection • Sensor and monitoring components on other various CDS assemblies, sub‐systems,

mechanical and operational components • CDS component, process and equipment performance and failure notification system • Designed computer controls components include, but are not limited to: • Level sensor systems • Motor starters and variable speed motor controls • Valve assembly relays and position indicators • Air operational controls components • ADS processes control and operational components • Safety system valves, pressure relief and emergency operational sensors and

components

1.5.9 OPS Proprietary Computer Controls Software and Process Protocol Operational components and computer protocol will be installed to facilitate remote monitoring and control capabilities. Dedicated data and/or phone lines will be installed for remote access to the computer control system. This monitoring and sensor system will send signals to the

20

computer control system to take any appropriate action, such as emergency shutdown, changes in operational schedules, diversion of liquid and material flows, gas flow diversion and as well as notifications to appropriate personnel of the location of any problem.

This computer monitoring and control system will have the capability to be monitored and operated remotely by authorized personnel.

Dr. Ruihong Zhang, and her research staff at UCD, will monitor the CDS for process and operational performance and biological processes and will have to ability to make adjustments or corrections to the operation remotely.

Onsite Power Systems will monitor and modify the CDS for optimal mechanical and operational performance. This monitoring system will also provide early detection of potential mechanical problems within the system.

1.6 Biogas Collection and Processing System As a project requirement, the daily production of biogas will be measured and recorded. The chemical composition of the biogas will be measured periodically by taking gas samples from ports installed in the hydrolysis and gasification reactor biogas collection lines and analyzing them with a gas chromatograph (GC). A detailed gas analysis will be used to verify measured readings and gas composition through the periodic collection and analysis of samples. Additionally, a gas flow meter will continually monitor the gas production and record the values at a central supervisory control and data acquisition (SCADA) computer. The generated biogas will be used to fuel an engine‐generator connected to the local power grid.

1.6.1 Biogas Collection Each reactor vessel in the system will generate biogas. The biogas quantity and composition will vary as the digestion process progresses though each cycle. In the early stages of decomposition, primarily carbon dioxide is generated in the hydrolysis reactors. As time progresses, the methane content of the biogas increases to a maximum of approximately 60 percent to 70 percent by volume. As the hydrolysis reactors are sequentially batch loaded, each will exhibit a different gas composition, while the biogas generated in the bio‐gasification reactor remains relatively constant.

To assure the proper operation of the engine fuelled by the biogas, it is necessary to achieve a relatively constant volume and quality of biogas throughout the process. This is achieved by mixing the biogas output from each reactor into a common manifold.

Each reactor will have a 2‐inch black iron collection pipe installed at the apex of its conical lid with a check valve to prevent the backflow of biogas through the common gas manifold. The pipe is directed down the side of each tank to approximately 4 feet above the concrete footing where a T‐fitting will turn it 90 degrees and continue the run to the gas clean‐up system at the south end of the facility. A manual drain trap valve will be placed at the bottom of each pipe to allow the removal of condensate from the line. A valve will be located just above the drain trap

21

for periodic gas sampling. Immediately below the sample port and above the drain trap, each line is directed toward the south end of the plant maintaining a slight slope to assure that any excess condensate is collected near the gas chiller where it will be removed from the line.

1.6.2 Biogas Upgrading Biogas is pre‐cooled through a gas‐to‐gas heat exchanger, and then routed through a modified refrigerator to condense the water. The heat exchanger recovers a portion of the refrigeration energy by re‐warming the dehumidified gas to ambient conditions before passing to the gas filtration system.

To reduce sulfur emissions and extend the operational life of the generator, the biogas will be passed through an iron/steel filter for removal of hydrogen sulfide. Sensors located prior to, and after, the filter send data to the SCADA computer to analyze the effectiveness of the filter and log the sulfur content of the gas stream sent to the engine‐generator. One aspect of the plant operational requirements consists of evaluating the effectiveness of different filter systems. The first filter will consist of a container filled with steel wool. Iron in the wool combines with the hydrogen sulfide to form iron sulfate, thereby removing the sulfur from the biogas. Other forms of filters will be tested, one being an iron sponge material, for sulfur removal effectiveness.

1.6.3 Biogas Measurement The quantity and quality of biogas produced is a key factor in determining the economic viability of the digester system. To properly determine the total energy produced by the APS‐Digester, the biogas must be accurately measured for total gas production and methane content. A gas flow meter will be incorporated into the system.

1.6.4 Biogas Compression To assure the best performance from the engine‐generator, biogas must be supplied at a relatively constant pressure and flow. Pipe losses and filter backpressures may reduce the supply pressure below the engine’s required value. To assure proper gas flow through the system, and proper line pressure to the engine, a blower will be installed after the filtration system to raise the biogas pressure to approximately 5 psig. An inline regulator then reduces the pressure to the working value dictated by the engine load.

1.6.5 Electrical generator One of the primary goals of the project is to deliver energy to the grid. It will be considered a design goal to deliver about 22 kW (528 kWh per day) of electric energy to the grid. The electrical energy delivered to the grid will be measured by the instrumentation and metering installed on the equipment. TGP West, Inc. will supply the generator package consisting of:

• Natural Gas Generator, 42.5 kW Series Star, 480 Volt, 3 Phase, installed. • Fuel system, including all parts and materials, installed. • 110 inches x 110 inches x 8 feet expanded metal enclosure, installed.

22

• Critical Silence muffler package, installed. • Electronics system governor control with shut downs

The initial design for the electrical interconnection equipment is based on commercially available power conditioning devices used in the wind power industry (Burgey 10 kW inter‐tie inverter). This equipment is designed to convert 3 phase AC power from a wind turbine, at constantly varying voltages and frequencies, to a set DC power and then invert to line quality power with a grid inter‐tie connection. Two devices operating in parallel (20 kW total) will receive input power from the biogas generator and allow the engine to be operated according to gas availability and heat load requirements without complicated monitoring of power output. TGP West’s Mitch Thorp will coordinate with UC Davis electricians regarding the specifications and installation of the equipment to meet the project requirements. This will meet the requirement to demonstrate the ability to deliver power to the campus grid. The generator system will be operated on a full‐time basis when feasible, but not continuously providing power to the grid. During any demonstration, or daily operations, the material processing and loading equipment may be operated from the generator and / or the electricity provided by the grid. The nominal electrical generating capacity of the system is anticipated to be 528 kWh per day, with some reduction as required to meet thermal loads during peak heat requirements. A 500‐gallon horizontal propane tank will be installed to provide a back‐up fuel supply for thermal requirements as needed.

1.6.6 Safety Biogas Flare All of the ADG will be used to run the engine/generator, or provide supplemental heating through a standard gas fired hot water heater. The generator will supply energy to the local grid at the UC Davis wastewater treatment plant through a meter recording the generated electricity. The generator simultaneously provides utility load to the pilot plant. In case of shut‐down of the engine‐generator and/or the cessation of biogas consumption occurs, a pressure relief valve on the inlet manifold to the engine will send the biogas to a flare to preclude raw biogas from being released to the atmosphere. The flare will have a continuous pilot light fueled from an external propane storage tank to assure that any low Btu biogas is combusted. A flame arrester check valve is placed in the main gas line going to the flare and the engine‐generator. It is specifically designed to handle low to medium Btu biogas and extinguish any flame that might back‐flash from either source.

1.7 Safety and Environmental Considerations

1.7.1 Safety Considerations Operator safety in feed stream loading involves operating mechanical material handling systems, and exposure to waste material that is undergoing decomposition. The Pilot Digester Final report will define the appropriate mechanical guards for the equipment, and personal protective clothing including equipment designed to protect operators from common respiratory hazards.

23

Operator safety in tank operation and maintenance involves handling high temperature water (140 oF), explosive gases (methane), and noxious gases (hydrogen sulfide, and carbon dioxide). The report describes the engineering of the safety flare, tank pressure/vacuum equalization hardware; limiting introduction of air into the system; biogas compression system; and the computer control system in the section entitled Tank Pressure/Vacuum Monitoring and Equalization. The Pilot Digester Final Report will define the appropriate operating procedures, which are consistent with California Code of Regulations, Title 8 Sections 5156 and 5158 for Confined Space Entry Program.

Operator safety in the plant operations must be insured throughout the 3‐ton per day digester pilot plant. This report has defined the location of installed safety relief valves in the Tank Pressure/Vacuum Monitoring and Equalization section. This report also describes the fencing that encloses the engine generator in existing pad expansion of the Component System Specification section. The Pilot Digester Final Report will define appropriate Lock‐out ‐ Tag Out procedures.

An operator‐training program for loading and residue recovery will be developed as a component of the Pilot Digester Final Report. The UC Davis Campus Office of Environmental Health and Safety will review this program as well as the entire safety program described above. This office requires an Injury Illness Prevention Program Manual to be on site at all times. This manual has been developed and is currently located in the computer room at the existing pilot digester plant.

The UC Davis Campus, Facilities Services Department, will perform oversight of all operations at the pilot digester plant. All personnel working at the digester location will be required to review the Injury Illness Prevention Program Manual and be aware of the location of the posted Emergency Contact Information.

1.7.2 Environmental Considerations Air emission and/odors are potential concerns when the feedstock is loaded into the digestion reactor vessels and the spent liquid and solids are taken out. This report describes the engineering for Hydrolysis Reactor Loading, Gas Collection and Processing Systems, Biogas Collection and Tank Isolation, and Biogas Measurement, in the Component Design Specification section. The prompt placement of the feedstock into a water stream which is pumped into the Hydrolysis Tank is to minimize the potential for air emission. The engineering described in tank pressure/vacuum equalization, safety relief valve, safety flare, and limiting introduction of air into the system in Tank Pressure/Vacuum Monitoring and Equalization section shows a design that keeps the system closed.

The air emissions from the engine generator are controlled in two areas of the report. The engineering of the Biogas Clean‐up and Electric Generator is described in the Component Design Specification section. The engine generator is to be operated under the Solano‐Yolo Air Quality Management Districtʹs Rule 412 for Stationary Internal Combustion Engines.

24

The air emission from the Safety Flare is described in two areas of this report. The engineering of the Biogas Clean‐up is described in the Component Design Specification section. The engineering of the Safety Flare is described in the Tank Pressure/Vacuum Monitoring and Equalization section. The Pilot Digester Final report will describe the operating practices that will be applied to be in compliance with Solano‐Yolo Air Quality Municipal Districtʹs rules.

The liquid effluent from the pilot digester is to be directed to the UC Davis Campus Waste Water Treatment Plant. The engineering for water recovery is described in the Solids Residue and Water Recovery section of the Component Design Specifications section.

The solids residues from the pilot digester are to be transported to the Hay Road Compost Facility of Norcal Waste System, Inc. in Dixon, California. The engineering for the sludge recovery is described in the Solids Residues and Water Recovery section of the Component Design Specifications section.

1.8 Pilot Digester Plant Testing Program and Modification Guidelines

1.8.1 Engineering Design Modifications Differences between the CDS as‐built operational results and the final UC Davis operational performance parameters will be identified. This will result in the identification of instances where CDS modifications may be required to satisfy the UC Davis testing requirements. The TEAM will perform a design evaluation of the required modifications and a design modification will be developed that will be used to address the specific design issues and identify the required modifications.

During this engineering design modifications phase the TEAM will assign a Project Manager (PM) that will manage the engineering design tasks for the CDS and identified modifications. This PM will become the primary contact person for all TEAM members, engineers, suppliers and UCD‐RT who have functional responsibility for specific module development. Modification changes go through an engineering review process prior to incorporation into the CDS engineered design.

The following list is a summary of tasks and responsibilities included in CDS engineering design modifications phase:

• UCD‐RT review of modifications for CDS functionality. • Engineering design team review and acceptance of modifications or alternatives as to

overall system performance and operation. • Modification cost and trade‐off analysis. • A verification analysis of the CDS and performance parameters achieved through

modifications will be performed. • Final modifications design review with UC Davis team members. • Incorporation of modifications into engineering design final drawings and specifications

as agreed upon.

25

• The development teams will use design review meetings as needed to provide design integrity throughout the development process. A design modifications integration analysis will be performed to insure that there is no negative effect to other systems and processes functionality within the CDS. After the integration analysis is completed, The TEAM will conduct a design over‐sight engineering review meeting to review the final engineered changes and modifications. Upon completion of an over‐sight review, all CDS design drawings including all design modifications will be released.

• Mechanical and Process Components Testing • Engineering design changes resulting in modifications to the current CDS mechanical

design and process flow may result in the use of alternative components to achieve modified performance specifications. In addition, during the engineering design phase alternative components will be identified that either will benefit the CDS project through cost savings or improved system performance. A cost‐to‐benefit analysis will be completed on potential alternative components.

• Certain mechanical components and process modifications will require advanced engineering design testing to validate the components functionality in the CDS engineering design or for the enhancement of the CDS performance. Modification and alternative components may include, but are not limited to:

• Biomass feedstock handling and processing equipment • Material circulation or chopper pump assemblies • Circulation piping valve assemblies and filters • Residue handling and processing equipment • Wastewater handling and processing equipment or systems • Computer controls components • CDS Start‐up procedures and protocol

During this project phase, the TEAM and the UCD‐RT will test various components and processes of the CDS. Some CDS functionality cannot be fully validated at the design level and will require onsite verification once the CDS System is installed at the UC Davis campus. The engineering design testing tasks are a means to verify that all design modifications have been correctly applied to achieve the required results. The method used for testing will be developed by the TEAM and will define the CDS mechanical or process components testing to be performed, the expected results and determine alternative components if required. Various mechanical and process components will be connected to or installed on the CDS by OPS or UCD‐RT members. A System Test will be performed for each mechanical or process component after its set‐up process has been completed.

The System Test is an iterative process and involves the TEAM and UCD‐RT. Design problems are corrected by the TEAM, and re‐tested. Any CDS problems identified by the UCD‐RT are referred back to the TEAM for resolution. Documentation of all issues will be maintained from creation to resolution. As‐Built Design Drawings will be created as necessary upon correction of

26

any identified problems and modification to the CDS. During the Mechanical and Process Components Testing, performance files will be created electronically and either output to diskette or transmitted electronically to the TEAM and UCD‐RT and is to be used for comparison to engineering design expectations.

1.8.2 Material Handling Testing This will be a comprehensive analysis and testing program to determine CDS specific engineering designs and operational parameters for various feedstock‐handling components. This comprehensive testing will also provide the operational and process protocols for the CDS and commercial projects on various feedstock supplies.

The Biomass feedstock handling testing will begin at the CDS during the start‐up and full operation period as specified in the Project Schedule. The testing period is scheduled for 150 days and the testing and analysis work will be completed in two phases.

Phase One is scheduled for 60 days during start‐up period and will consist of feedstock material handling and analysis directed toward defining the CDS engineering design parameters including:

• CDS assessment of the material handling capabilities and the rate of system loading using various loading parameters.

• Determination of volume of materials received to the re‐compaction rate of materials loaded into the digester tanks.

• Engineering design assessment for processing and loading of raw waste material into the CDS and identify potential operational alternatives.

• System analysis for the removal of non‐organic materials • Determination of operational conditions that will be implemented into final CDS

operational protocol. Phase Two is scheduled for 90 days during full CDS operation and will consist of comprehensive analysis of CDS loading process and operations parameters. Included in Phase Two will be a comprehensive analysis of the CDS material handling process and remaining residue solids materials handling. Performance data collected will be incorporated into full‐scale system design. Phase Two analysis and evaluations will include:

• Evaluation of digestion retention time and loading rates of select Biomass feedstock and potential enhancements to improve biogas production.

• Evaluation of residue material recovery system and material handling • Analysis and evaluation of CDS effluent water properties, storage, reuse and disposal. • Operational controls design and modifications based on systems evaluations. • Process computer control specifications and operational protocol.

27

1.8.3 Expanded Feedstock Testing, Protocol and Procedures The feedstock handling system testing will consist of running test procedures, verification of the test results and visual inspection of CDS functionality of designated percentages of various green waste feedstocks. A testing program will be created by the TEAM and the UCD‐RT and performed at the CDS as required. The research team will provide the testing protocol and procedures used for various types and percentages of green waste to food waste feedstock.

The TEAM will be responsible for providing the various required system adjustments requested by UCD‐RT to run the defined testing procedures. The TEAM will monitor performance, problems and fixes that occur during the feedstock‐handling program and will provide this information to the UCD‐RT. Design modifications will be developed based on operational performance of various CDS systems and processes for:

• Material processing and loading • Residue processing for identified percentages • CDS operational criteria as determined by modifications • Computer controls and software modifications

Based on CDS operational performance of various feedstock percentages, the TEAM will generate design modifications by determining specific operational and design requirements for each percentage based blend of green waste and food waste. CDS engineering design changes will be incorporated to meet operational requirements.

1.8.4 UCD-RT Development of CDS Design Modifications The TEAM and UCD‐RT will review engineering design change recommendations based on feedstock handling results. The TEAM will determine if any mechanical, processing equipment or components identified in this design change recommendations require engineering testing at the CDS. Upon final engineering review by the TEAM and UCD‐RT, and based on results from required engineering components testing, the TEAM may incorporate these design changes into the CDS design.

UCD‐RT will submit CDS design modifications recommendations based on advanced laboratory testing results that warrant these modifications. The CDS modifications recommendations will be submitted to the TEAM during a scheduled design meeting. The TEAM will review design changes and modifications for engineering functionality and will deliver an engineering design modification based on value‐added benefits to the CDS system and/or process. The TEAM and UCD‐RT will then mutually accept, conditionally accept, or reject the CDS design modifications.

1.8.5 Final As-Built Engineering Design Review The TEAM and UCD‐RT will perform a CDS design and system performance evaluation and will mutually accept or conditionally accept the final CDS as‐built design. The final review of all

28

CDS as‐built engineering design drawings by the TEAM and UCD‐RT will take place after the first 60 days of full CDS operation and will occur at a date mutually agreed to by the TEAM and UCD‐RT. The purpose of the review is to review all design modifications and the verification of the mechanical and operational functionality of the CDS based on final system design. This design review and analysis will form the basis for the Acceptance Criteria for commercial digester engineering design work.

1.9 Pilot Digester System Performance Test Plan

1.9.1 Biogas Production and Stability of Anaerobic Reactors The performance of anaerobic digester system will be evaluated based on the daily biogas production rate, biogas composition, total solids and volatile solids destruction on the feedstock after the digestion, and pH of reactors. During the full CDS operation, biogas production rate from all reactors (hydrolysis and biogasification) will be measured daily using in‐line gas flow meters and recorded on computer stations. The methane and carbon dioxide contents will be measured weekly by taking gas samples from each reactor through biogas collection lines and analyzing them with GC/TCD. In addition, the biogas samples will be taken before and after moisture and hydrogen sulfide removing devices and analyzed for the contents of hydrogen sulfide and other sulfur compounds and ammonia. An online monitoring pH probe will be installed in the buffer tank to measure continuously the pH of effluent from the hydrolysis reactors, which will also be the influent for the biogasification reactor. The readings from the pH probe will be recorded on the computer station and evaluated on daily basis. In addition, liquid samples will be taken from all the hydrolysis and biogasification reactors on weekly basis to measure pH and ammonia concentration. Temperature of each reactor will be monitored via the thermo‐sensors installed on the liquid recirculation lines of each reactor and the sensor readings will also be recorded on the computer station and evaluated on daily basis.

Destruction of total solids and volatile solids in the feedstock after digestion and characteristics of digester effluent (solids and liquid fractions) will be determined by measuring three batches of each testing feedstock after the digester reaches steady‐state. The destruction of solids will be measured by performing the mass balances of each batch before and after digestion. In addition, fate of nutrients (nitrogen, phosphorus, potassium) and metals (Zn, Cu, Na, Ca, Mg, and so forth) will also be measured.

1.9.2 Energy Conversion Performance and Air Emissions of Engine-Generator Engine emissions will be measured using standard CARB and/or US EPA stationary source measurement protocols. Table 3 lists the items to be measured in the exhaust gas with the corresponding reference methods. Measurements should be made for engine operating at full load (or highest load allowed by biogas production rate), 75 percent load and 50 percent load.

29

Engine performance and efficiency will be determined using guidelines in ASME Performance Test Code for Reciprocating Internal Combustion Engines, PTC‐17,1 or similar. Energy and mass balances for the engine‐generator system will be determined as part of the engine and CHP system performance (Figure 4 and Table 3).

Table 3. Exhaust Gas Analyses and Reference Methods

Emissions Testing Reference Methods Analyze CARB*2 US EPA3 NOx 100 7E CO 100 10 Total Hydrocarbon (THC) 100 25A Non‐methane Hydrocarbons (NMHC) 25 SO2 100 or 6 6C Total Reduced Sulfur (TRS) 15 or 16A 16A Total PM 5 5 NH3 CO2 100 3A Exhaust Gas Velocity and Mass Flow‐rate

2 or 2A

Moisture in Exhaust gas 4 Oxygen 100 Units will be reported as; ppmvd (@15% O2, dry), lb/hr, and lb/kWh electrical output

* CARB = California Air Resources Board

1 American Society of Mechanical Engineers, Performance Test Code for Reciprocating Internal

Combustion Engines, ASTM PTC‐17, New York, NY. 1997.

2 CARB Source Test Methods are available at; http://www.arb.ca.gov/testmeth/testmeth.htm

3 US EPA Source Test Methods are available at, http://www.epa.gov/ttn/emc/promgate.html

30

Figure 4. Schematic of Biogas Fueled CHP System

Engine

Electrical Power Out

Biogas Input

Air Intake

A BA

C

D

E

Heat Exchanger

Exhaust

G

F

Cooled Engine Exhaust

H

Liquid from

Digester

Radiator

I

J

K

L

Liquid to

Digester

M

Excess Heat

Cool Liquid

Hot Liquid

31

Table 4. Engine CHP Measurement Locations

Location Measurement

A Fuel gas pressure

B Fuel gas flow

C Fuel gas temperature

D Grab sample for biogas analysis (methane, ammonia, sulfur, moisture)

E Air temperature, pressure, humidity and potentially flow rate

F Electrical power output measurement

G Emissions testing plus flow‐rate and temperature

H Digester liquid temperature ‐ cool

I Engine out exhaust temperature

J Engine out water temperature

K Engine water inlet temperature and flow rate

L Digester liquid temperature ‐ hot

M Digester liquid flow rate

A fully documented report on emissions and energy performance of the engine CHP system will be produced including recommendations for improvements and further development (other feedstock, digester operating parameters, different prime movers, as well as other biogas uses).

1.10 Conclusions Final construction specifications regarding the design of the physical plumbing, material handling, residue recovery, hydraulic mixing, process circulation, computer control system and operational protocol will be based on the fundamental designs presented within this report.

Upon completion of the Energy Commission kick‐off meeting, the goal is to conduct an engineering design meeting with all design team members. The agenda for this meeting will be to perform a final review of all the design specifications and drawings, operational planning and design modifications program. The desired outcome of this meeting is to have all team members agree to agree on the final construction specifications and operational parameters of the CDS project. This meeting prior to construction of the CDS and will be held at UC Davis.

32

CHAPTER 2: Pilot Digester Construction Report 2.1 Introduction The pilot‐scale APS‐Digester system designed as described above was constructed on the campus of UC Davis. This report documents the equipment and components, and construction process that were built into the pilot digester system at UC Davis Biogas Energy Plant. Attempts were made to adhere to the original design concept. However, changes to the design were made as needed to preserve system integrity, minimize cost, and enhance system safety. Changes made to the original concept were noted as far as possible.

2.2 Pilot Digester Construction Specifications The pilot digester plant was constructed and installed adjacent to the UC Davis wastewater treatment plant as pictured below. Each sub‐system’s critical design criteria and operating parameters are described in the following sections.

Figure 5. APS Digester Pilot Plant. Orientation of Reactors and Gas Processing Equipment at the APS Digester Pilot Plant at UC Davis.

Photo Credit: UC Davis.

33

2.2.1 Concrete Pad A concrete pad was constructed to accommodate the pilot digester system. This area consists of two sections, one that is 35 feet x 58 feet and an adjoining section that is 47 feet x 70 feet. The larger section will be used to hold the material receiving, processing and residue recovery equipment. A section of this area contains a sloped area that will allow the floor at the back end of the roll‐off to be level with the rest of the pad and allow a skid‐steer to enter the bin without a ramp. Typical roll‐off bin floors are 8 inches off ground level. The designated area is 21 feet x 20 feet. The sloped area begins 5 feet south of the north access road and slopes down to the south for 20 feet to a drop of 5 inches. The sloped area begins 5 feet from the east side of the pad and continues 21 feet to the west.

The third section located at the South section is 17 feet by 26 feet area houses the biogas storage tank, clean‐up and processing equipment and generator system. The generator was installed within a metal screen enclosure for personnel safety and equipment security.

2.2.2 Reactor Vessels

Figure 6. Main Frontal View of Digester System with Loading Dock

The physical design of the system is based on using four Hydrolysis Reactors (HT); each with a working volume of 8,500 gallons. The selected tanks are 10 feet‐8 inches in diameter and 16 feet high and set the total system capacity at 3 tons per day for green wastes. The bio‐biogasification reactor volume is equivalent to one of the HT reactors

34

and a suitable used vessel was identified and procured. All five tanks are installed on the west side of the pad.

The hydrolysis reactors have sloped bottoms and each contains a full‐height water jacket. All the tanks are insulated with foam insulation and all the pipes and fittings that will contain the warm digester liquids are insulated with ceramic paint. The tanks are mounted on steel stands that allow a minimum of two feet of clearance under the tanks for construction and future modifications.

Figure 7. Five Anaerobic Reactors Including Four Hydrolysis Reactors and One Biogasification Reactor

To accommodate the blending of water containing organic acids from each hydrolysis reactor and to control the loading of organic acids into the biogasification reactor, a 900 gallon stainless steel buffer tank was installed and is located between the reactor and the first hydrolysis reactor.

35

Figure 8. Buffer Tank Located Between Hydrolysis and Biogasification Reactors

2.2.3 Boilers and Heat Exchangers The installed boiler system includes two boiler units. The first is a 260,000 Btu boiler that will operate on propane and will be used in the digester start-up. The second boiler system is a 700,000 Btu boiler that is designed to operate on the biogas produced. Both boilers are interconnected to the heat exchanger loop

Figure 9. Dual Boiler System

36

Figure 10. Propane Operated Boiler for System Start-Up

2.2.3.1 Heat Exchangers

Maintaining a system temperature of 135 °F is critical to the production of biogas and life of the thermophilic bacteria. To avoid the potential destruction of bacteria, reactor fluids should not be exposed to heat exchanger surfaces that exceed 150 °F. An independent shell and tube heat exchanger system was installed for the biogasification reactor and is connected to the main hot water loop. The designed heat exchanger is installed at the GT and a dedicated circulation pump is used to connect the heat exchanger and heat source piping systems.

Figure 11. GT External Heat Exchanger System

37

The remaining HT are connected to the main hot water loop and each HT has a full water‐jacket to serve as the tank heat exchanger component. The boiler system heat provides the hot water source to the tank water jackets. Heating of each HT is controlled through the computer system and was designed with the ability of heating one or more HT simultaneously. When digester system heating is not required and the engine‐generator is off line for maintenance and /or repairs, excess biogas will continue to be burned in the biogas boiler. To allow continuous operation of the biogas boiler during this period, an air fan heat exchanger was installed to allow full operation of the biogas boiler system.

Figure 12. Air Fin Heat Exchanger System

2.3 Material Handling and Digester Operational Equipment

2.3.1 Hydrolysis Reactor Loading The demonstration plant is designed to use different proportions of grass and food waste as the primary feedstock. The operational plan is to load one HT reactor either daily, or every three days based on feedstock materials and provide for a predetermined solids retention time.

38

Figure 13. Delivery of Bins Containing Feedstock Materials

The feedstock materials will be collected and delivered to the digester in plastic bins, each bin is 48 inches x 48 inches and 28 inches high and bins have a snap on cover. This allows for clean delivery of feedstock materials and provides the ability to examine materials, take samples and monitor material weights and volumes to be loaded.

Figure 14. Bins of Waste Materials Delivered for Loading

Once the material is delivered to the site, the bins will be dumped into a receiving hopper that allows for metering of the materials into either an enclosed drag chain

39

conveyor or a Vaughan chopper pump system. Loading options will be determined based on feedstock characteristics.

Figure 15. Forklift Dumps Bin into Receiving Hopper

Figure 16. Receiving Hopper Metering Feedstock into Conveyor System

The conveyor system carries the material to two hydraulic piston pumps. One pump is located between HT 1 and HT 2, the second pump is located between HT 3 and HT 4. The conveyor discharges material into the selected hydraulic pump based on which HT

40

is to be loaded. Once the material is delivered to the piston pump; it is then loaded into the selected HT.

Figure 17. Conveyor Delivers Feedstock to Hydraulic Piston Pumps

Figure 18. Hydraulic Piston Pump Located at HT 3 and HT 4

41

Figure 19. Hydraulic Piston Pump Located at HT 1 and HT 2

Figure 20. Valve System for HT Loading

2.3.2 Liquid Circulation Process Two self‐cleaning screen filter systems were installed on both HT 3 and HT 4, and one self‐cleaning screen filter system was installed on both HT 1 and HT 2. Two types of circulation pumps were installed to provide the ability to test various circulation processes. Each valve and pump will be connected to the main computer control system and will operate according to circulation operational protocol.

42

Figure 21 Screen Filter, Self Cleaning System

During the circulation process, water containing organic acids is extracted from each HT and delivered to a buffer tank. The pH of water collected from each HT is monitored using instrumentation that provides real time readouts and sends monitoring data to the SCADA system through wireless transmission. All readings may be monitored from an instrument panel while SCADA is performing control operations and data collection in the control room. This provides information on HT performance and allows controlled circulation of the organic acids to the GT, allowing for pH control.

Figure 22. pH Meter and Circulation Buffer Tank

43

2.3.3 Internal Hydraulic Mixing System – Hydrolysis Tanks 1 and 2 During the circulation process, water and feedstock material will be drawn from the HT through a stainless steel pine line and fed to a Vaughan in‐line chopper pump. The material will then be returned to the same HT through a feed line to mixing nozzles installed inside each HT. Mixing nozzles are installed that can be positioned to provide total mixing of the HT. This ensures suspension of feedstock materials and prevents crust formation on the liquid surface.

Figure 23. In-line Chopper Pump and Pipe System

2.3.4 Hydraulic Internal Mixing System – Hydrolysis Tanks 3 and 4 During the circulation process, water will be decanted through screen units through a stainless steel line and fed to a pump. Only water will be extracted and circulated while mixing HT 3 and HT 4.

44

Figure 24 Hydraulic Mixing Pump HT 3 and HT 4

2.3.5 Hydraulic Internal Mixing System – Biogasification Tank

An independent mixing system identical to the HT 3 and HT 4 mixing systems was constructed for the GT. The GT system will utilize a dedicated pump and nozzles located at the bottom level of the tank.

Figure 25 GT Hydraulic Mixing Pump and Pipe System

Each circulation pump has a dedicated drive installed that is either manually operated or operated through the main computer control system. Each air actuated valve has been

45

connected to the main computer control system and will be operated according to the designated operational protocol. A main air controller system was installed inside the computer control building.

2.3.6 Solids Residue and Water Recovery At the end of each retention cycle, the selected HT will be drained and the solids extracted from the working fluid using a mechanical screw press. The press will be installed on the north end of the main drain line providing a straight line to the press. Water extracted from this process will be sent to the existing reclaimed water tank (RWT). The solids will be transferred to the empty roll‐off bin, which will be picked during the next transfer of feedstock cycle. The existing reclaimed water tank will provide enough storage capacity for the reclaimed water recovery process.

Each HT and the GT will be connected to the drain line through a manual gate valve. The main drain line will exhibit a required slope to the north section on digester site.

Figure 26. HT Drain Valve and Connection to Drain System

46

Figure 27. Drain Section to be Connected to Pump and Screw Press Unit

A trash pump will be installed at the end of the drain system to meter the flow of water and materials into the material recovery system. Just after the trash‐pump, the drain line will connect to a screw press. It was determined that the press will provide solids with a moisture content that will not drip liquid when transferred to the roll‐off box. The water recovery drain from the screw press will be routed to the RWT through a stainless steel pipe system. A secondary water line from the RWT will be connected to the south end of the main 8‐inch drain line to provide the ability to flush out the feed/drain system into the press in a closed loop water transfer cycle.

2.3.7 Tank Pressure/Vacuum Equalization To assure that all tanks remain at their design pressure or vacuum, their head pressures are equalized when fluid is transferred from one tank to another. To accomplish this, each tank’s head space is connected via the biogas collection system piping. Pressure transducers have been installed on each digester tank and will be monitored by the computer control system.

2.3.8 Safety Pressure Relief Valve Over‐pressure and vacuum limitations are designed through the installation of a spring‐loaded valve in the top of each tank. These specially designed valves do not allow the tank to exceed safe operating pressures or vacuums. Venting of the tank occurs if the pressure exceeds 54‐inches of water column. As an additional safety factor, a flame arrester has been installed on each digester reactor.

47

Figure 28. Pressure Relieve Units and Flame Arrester

2.3.9 Limiting Introduction of Air Into the System Every effort is made to avoid introducing air into the tanks and manifolds that contain methane. During initial filling some air is introduced with the feedstock in the tank. Filling the tank as full as possible with water after introducing the feedstock is recommended to minimize the head of air at the onset of digestion. A natural safety feature in the digestion process is that carbon dioxide is prevalent at the beginning of the digestion cycle and consequently, explosive methane and air mixtures are more difficult to form. The level of each tank is monitored by the SCADA computer, which in turn controls the specific levels through the control of the water supply valves.

2.3.10 Computer control system A comprehensive computer monitoring and control system was designed for the CDS project. A centralized computer control center was constructed within the main operations building. The computerized control system is designed to monitor all mechanical components and various processes in the CDS operation from this location. Lighted and animated control panels will provide indication of the operational status of all mechanical, operational and process components. The CDS operator(s) will control and oversee the various processes from this location and make required adjustments and changes as needed. The computer control software and process protocol is proprietary to OPS and OPS will retain ownership of all computer software and process protocols.

48

Figure 29 Process Flow Monitoring Screen

The overall digestion process is designed to operate almost entirely automatically by a computer control system and automated controls components. Various circulations and hydraulic mixing cycles will be scheduled based on various types of feedstock through a preset sequence of tanks and processes. Various process protocols identified earlier have been incorporated into the main computer control system. Each of the circulation and hydraulic mixing cycles are programmed for automatic performance at required intervals. Computer control components are designed into all of the valve assemblies, pump and motor systems and piping sub‐assemblies with signal and data connections going to a main computer control center located in the CDS operations building.

In addition to the control components, a comprehensive sensor and monitoring system has been designed into numerous components of the CDS tanks, assemblies and sub‐assemblies. This will enable continuous monitoring of the entire CDS and system processes. Monitoring sensor components have been incorporated into the CDS design to monitor:

• Solid and liquid levels in all CDS tanks • Temperature sensors in all CDS tanks • Process flows through circulation piping systems and assemblies • Position indicators on all valve assemblies • Operational performance of hydraulic mixing processes • Operation performance of heat exchanger systems • Gas pressure in all CDS tanks • Sensor and monitoring components on other various CDS assemblies, sub‐

systems, mechanical and operational components

49

• CDS component, process and equipment performance and failure notification system

• Designed computer controls components include, but are not limited to: • Level sensor systems • Motor starters and variable speed motor controls • Valve assembly relays and position indicators • Air operational controls components • ADS processes control and operational components • Safety system valves, pressure relief and emergency operational sensors and

components

Figure 30. Spread Spectrum Wireless Data Transmission Boxes

50

Figure 31. Biogas Flow, Biohydrogen Flow and pH Meters

Figure 32. Motor Control Cabinet with HMI Interface Panel

51

Figure 33. One of Three Magnetic Flow Meters

2.3.11 OPS Proprietary Computer Controls Software and Process Protocol Operational components and computer protocol were installed to facilitate remote monitoring and control capabilities. A USB wireless air card will be used for remote access to the computer control system. This monitoring and sensor system will send signals to the computer control system to take any appropriate action, such as emergency shutdown, changes in operational schedules, diversions of liquid and material flows, gas flow diversions and as well as notifications to appropriate personnel of the location of any problem. This computer monitoring and control system will have the capability to be monitored and operated remotely by authorized personnel with various levels of passwords and access codes. Dr. Ruihong Zhang, and her research staff at UCD, will monitor the CDS for process and operational performance and biological processes and will have to ability to make adjustments or corrections to the operation remotely. Onsite Power Systems will monitor and modify the CDS for optimal mechanical and operational performance. This monitoring system will also provide early detection of potential mechanical problems within the system.

2.4 Biogas Collection and Processing System As a project requirement, the daily production of biogas will be measured and recorded. The chemical composition of the biogas will be measured periodically by taking gas samples from ports installed in the hydrolysis and biogasification reactor biogas collection lines and analyzing them with a gas chromatograph (GC). A detailed gas analysis will be used to verify measured readings and gas composition through the periodic collection and analysis of samples. Additionally, a gas flow meter will continually monitor the gas production and record the values at a central supervisory control and data acquisition (SCADA) computer.

52

2.4.1 Biogas Collection Each reactor vessel in the system will generate biogas. The biogas quantity and composition will vary as the digestion process progresses though each cycle. In the early stages of decomposition, primarily carbon dioxide is generated in the hydrolysis reactors. As time progresses, the methane content of the biogas increases to approximately 60 to 70 percent by volume. As the hydrolysis reactors are sequentially batch loaded, each will exhibit a different gas composition, while the biogas generated in the bio‐gasification reactor remains relatively constant.

To assure the proper operation of the engine fuelled by the biogas, it is necessary to achieve a relatively constant volume and quality of biogas throughout the process. This is achieved by mixing the biogas output from each reactor into a common manifold.

Each reactor has a 2‐inch black iron collection pipe installed at the apex of its conical lid, which is connected to a common gas manifold linking all digester tanks. The biogas collection manifold is located along the tops of the digester tanks and runs to the gas clean‐up system at the south end of the facility. A manual valve has been installed at each HT and the GT to allow for gas sample collection.

Figure 34. Biogas Collection Manifold

Each hydrolysis reactor has valves controlled by SCADA to isolate reactors from biohydrogen (blue pipe) and biomethane (red pipe). Any reactor may be isolated from other reactors. All reactors are normally open to each other allowing pressure equalization during filling and emptying operations.

53

The biogas collected from the HT reactor is passed through and blended with the biogas produced in the GT. The biogas is then directed from the GT through a water separator with drain and into a hydrogen sulfide scavenging filter containing iron SulfaTreat material.

Figure 35. Hydrogen Sulfide Filter with SulfaTreat

Several moisture drain traps were installed throughout the biogas collection system to allow for the removal of condensate. Biogas from the hydrogen sulfide filter passes through a coalescent filter trapping small particles and removing more water vapor from biogas stream.

Figure 36. Coalescent Filter, Expansion Tank and Drain Traps (Located on Ground)

54

The biogas is then sent through an expansion tank to a pressure regulator , which allows for 14 inches of water column output pressure

Figure 37. Biogas Pressure Regulator System

2.4.2 Biogas Compression To assure the best engine‐generator performance, biogas must be supplied at a relatively constant pressure and flow. Pipe losses and filter backpressures may reduce the supply pressure below the engine’s required value. To assure proper gas flow through the system, and proper line pressure to the engine, a blower was installed after the filtration system to raise the biogas pressure to approximately 5 psig. An inline regulator then reduces the pressure to the working value dictated by the engine load

55

Figure 38. Biogas Blower Unit

2.4.3 Electrical Generator One of the primary goals of the project is to generate electricity from the produced biogas. It will be considered a design goal to deliver about 22 kW (528 kWh per day) of electric energy to the grid. The electrical energy delivered to the grid will be measured by the instrumentation and metering equipment installed on the Cummins generator system.

The generator system will be operated on a full‐time basis when feasible, but may not continuously provide power to the project. During demonstrations or daily operations, the material processing and loading equipment may be operated using power from the generator and/or the electric grid. The nominal electrical generating capacity of the system is anticipated to be 528 kWh per day, with some reduction as required to meet thermal loads during peak heat requirements.

56

Figure 39. Engine-Generator System with Control Switch Box

Figure 40. Engine-Generator System with Control Switch Box

57

2.4.4 Safety Biogas Flare

All of the ADG will be used to run the engine/generator, or provide supplemental heating through a standard gas fired hot water heater. Biogas not used through the generator will be used in the boiler system. In case of shut‐down of both the engine‐generator and boiler and/or the cessation of biogas consumption occurs, a pressure relief valve on the inlet manifold to the engine will send the biogas to a flare. This will prevent raw biogas from being released into the atmosphere. A flame arrester check valve is placed in the main gas line going to the flare and the engine‐generator. It is specifically designed to handle low to medium Btu biogas and extinguish any flame that might back‐flash from either source.

2.5 Safety and Environmental Considerations

2.5.1 Safety Considerations Operation and maintenance of the digester system may require handling high temperature water (140 oF), explosive gases (methane), and noxious gases (hydrogen sulfide, and carbon dioxide). The report describes the engineering of the Safety Flare, Tank Pressure/Vacuum Equalization hardware; limiting introduction of air into the system; biogas compression system; and the computer control system in the section entitled: Tank Pressure/Vacuum Monitoring and Equalization. The Pilot Digester Final Report will define the appropriate operating procedures, which are consistent with California Code of Regulations, Title 8 Sections 5156 and 5158 for Confined Space Entry Program.

Operator safety during plant operation must be insured throughout the 3‐ton per day digester pilot plant. This report has defined the location of installed Safety Relief Valves in the Tank Pressure/Vacuum Monitoring and Equalization section. This report also describes the fencing that encloses the engine generator in Existing Pad Expansion of the Component System Specification section. The Pilot Digester Final Report will define appropriate Lock‐out ‐ Tag Out procedures.

An operator‐training program for loading and residue recovery will be developed as a component of the Pilot Digester Final Report. The UC Davis Campus Office of Environmental Health and Safety has inspected the plant and will review the operational program as well as the entire safety program described as required. This Office requires an Injury Illness Prevention Program Manual to be on site at all times. This manual has been developed and is currently located in the Computer Room at the existing pilot digester plant.

The UC Davis Campus, Facilities Services Department, will perform oversight of all operations at the pilot digester plant. All personnel working at the digester location will be required to review the Injury Illness Prevention Program Manual and be aware of the location of the posted Emergency Contact Information.

58

2.5.2 Environmental Considerations Air emission and/odors are potential concerns when the feedstock is loaded into the digestion reactor vessels and the effluent liquids and solids are removed. This report describes the engineering for Hydrolysis Reactor Loading, Gas Collection and Processing Systems, Biogas Collection and Tank Isolation, and Biogas Measurement, in the Component Design Specification section. The prompt placement of the feedstock into a water stream which is pumped into the Hydrolysis Tank will minimize the potential for air emission. The engineering of the tank pressure/vacuum equalization system, safety relief valves, safety flares, and air introduction reduction system is described in the Tank Pressure/Vacuum Monitoring and Equalization section, which shows a closed system design.

Air emissions from the engine generator are discussed in two sections of the report. The engineering of the Biogas Clean‐up and Electric Generator is described in the Component Design Specification section. The engine generator is to be operated under the Solano‐Yolo Air Quality Management Districtʹs Rule 412 for Stationary Internal Combustion Engines.

The air emission from the Safety Flare is described in two sections of this report. The engineering of the Biogas Clean‐up is described in the Component Design Specification section. The engineering of the Safety Flare is described in the Tank Pressure/Vacuum Monitoring and Equalization section. Biogas Clean‐up and Safety Flare design comply with Solano‐Yolo Air Quality Municipal Districtʹs emissions rules.

The liquid effluent from the pilot digester will be directed to the UC Davis Campus Waste Water Treatment Plant. The engineering for water recovery is described in the Solids Residue and Water Recovery section of the Component Design Specifications section.

The solids residues from the pilot digester will be transported to the Hay Road Compost Facility of Norcal Waste System, Inc. in Dixon, California. The engineering for the sludge recovery is described in the Solids Residues and Water Recovery section of the Component Design Specifications section.

2.6 Equipment and Cost of UC Davis Biogas Energy Plant This document provides a list of equipment and structures that have been built into the UC Davis Biogas Energy Plant with a total cost of $1,713,747. The items listed under Equipment Purchased for UC Davis and Equipment Donated to UC Davis have ownership to UC Davis and have a total cost of $793,084. These include the equipment and services purchased with UC Davis fund of $192,199. The items listed under Equipment Purchased and Owned by Onsite Power Systems have ownership to Onsite Power Systems and have a total estimated cost of $920,663.

59

Summary -Cost

Equipment Purchased for UC Davis by Onsite Power Systems with CEC Grant 294,999$ Equipment Donated to UC Davis by Onsite Power Systems and other Collaborators 305,886$ Equipment Purchased and Owned by Onsite Power Systems 920,663$ Equipment and services purchased with UC Davis Funds 192,199$ TOTAL: Construction Costs - Project Value 1,713,747$ Equipment purchased for UC Davis by Onsite Power Systems using CEC Grant AmountEngineering & design services 25,000$ 3 - 10,000 gallon digester tanks 18,909$ Concrete work on esxisting pad expansion 11,462$ Engineering & design services 37,085$ Four steel digester tank stands 8,673$ Used 2" stainless steel pipe and fittings 1,618$ CEC meeting 463$ Concrete work on esxisting pad expansion 9,417$ Contracted services for seting digester tanks 6,640$ Material processing equipment fabracation 3,400$ Structural fabracation on drain line system 2,762$ Structural fabracation on system piping 3,643$ Structural steel and fabracation 84,432$ Piping systems fabracation, rental equipment 5,063$ Contracted welding services 30,092$ Circulation fabracation services 1,341$

Equipment purchased by UC Davis using CEC GrantChopper Pumps 45,000$

TOTAL: Purchased through CEC Grant funds 294,999$ Equipment purchased for UC Davis with UC Davis Fund Chopper pumps 3,125$ Liquid Flowmeters 15,315$ Gas flare 3,017$ computer software 4,609$ Supplies for piping, air valves 27,701$ Bin Dumper 9,068$ Welding service 26,160$ Digester Insulation 49,938$ Computer Programing Service 38,049$ Scale 1,811$ Waste bins 4,936$ Grainger Supplies 1,620$ Auger Seals 524$ Ducts 2,926$ Control Panel 3,000$ pH Probes 400$

Total Equipment and Services funded by UC Davis 192,199$

60

Equpment donated to UC Davis by Onsite Power Systems AmountDigester tank stands modifications including mounting plates 2,318$ 8 tube x 9" Diameter - 10' heat exchanger 2,155$ 8" x 10' long incline auger 862$ 3' wide x 9.5' long Commercial inspection table 4,315$ 9" x 15' Screw conveyor 1,618$ 10,000 gallon Reclaimed water tank (sandblasted & painted) 4,826$ Stainless steel material receiving hopper 1,894$ Additional concrete installation 10,690$ Catwalk & structure sandblasting & painting 6,055$ 1/2" x 8" steel mounting pads & metal crossover stairs 891$ 5 Clarkson 10" pneumatic knife gate valves 2,896$ 1 ½” into 2 ½”, 20 feet long stainless steel heat-exchanger 5,393$ 100' - 3" and 100' 6" Stainless steel pipe 4,200$ Refurbish 30 valves - clean & paint 1,650$ 125 Gallon open top tank (water float system) 647$ 9" x 25' stainless steel auger 2,157$ Structural primer and paint 847$ Contracted welding - drain line piping and valve modifications 9,688$ Contracted welding - structural and catwalk modifications 11,061$ Equipment rentals for component delivery & use in modifications 3,930$ Contracted fabrication & modification services (AFT) 12,800$ Power transformers, electrical panels, components, and site wiring 28,600$ Motor controllers and computer control components 6,500$ Various 1" to 4" Electric pumps & motors 6,700$ Valves air, electric & manual, pipe fittings, connectors, supplies 3,400$ Computer control building & storage container 5,600$ 2620 Square foot concrete pad 27,400$ Kobold 2' liquid flow meter, instrumentation and connections 5,220$ Ashcroft Pressure Transducer (3) 5,342$ Pressure transducer instrumentation and connections 9,840$ Sage gas flow meters (2) instrumentation and connections 14,800$ Bristol pressure transducers (2) 3,890$ Bristol pressure transducer instrumentation and connections 6,280$ Kobold 3" Circulation flow meter (3) instrumentation and connections 8,720$

TOTAL - Donated by Onsite Power Systems 223,186$

Equipment Donated by others2 - 10,000 gallon stainless steel tanks 20,000$ 4" Inlet trash pump 7,600$ 1 1/2" Stainless steel pipe, stainless steel flange fittings 2,950$ 8" x 14' - 1/2 Pitch auger 850$ 150' of 48", 30" and 24" catwalk structure 3,400$ Four stairways with railings and landings 2,300$ Four high intensity light fixtures 2,000$ 8" steel pipe for drain system & structural steel 2,300$ 30 Various 4", 6" and 8" valves 24,800$ 9" x 30' - 1/2 Pitch stainless steel incline auger 5,000$ Commercial De-rock system 11,500$

TOTAL - Donated by others 82,700$

TOTAL Equipment Provided to UC Davis as In kind 305,886$

61

Equipment purchased and owned by Onsite Power Systems AmountEngineering design review services - Brown & Caldwell 26,870$ Outside engineering consulting services 13,240$ Reclaimed water tank inc. stand and modifications 11,480$ Boiler / heat exchanger circulation systems inc. pumps, piping, pneumatic and manual valves, fabrication & assembly 18,847$ Biogas boiler system inc: pumps, piping, valves, fabracation and assemble 10,485$ Biogas heat exchanger - venting excess heat 14,840$ Larrs 320,000 Btu boiler 3,217$ Heat exchanger system - Hydrolysis tanks inc: pumps, pneumatic and manual valves, pumps, fabrication and assembly 16,500$ Heat exchanger system - Gasification & buffer tanks inc: pumps, pneumatic and manual valves, pumps, fabrication and assembly 12,500$ Comoso air control system including controllers, valves, panels and fabrication 10,600$ Groth - 3 Pressure - vacuum relief units 4,300$ Control room electrical panels and components 3,567$ Jaeger - Biomedia disks for gasification tank 2,973$ 2 - 9" X 14' augers - stainless steel 2,977$ H2S tank - 900 gallon unit with modifications and fabrication 5,300$ Precision Filtration system 1,928$ Zabel - Incline conveyor system 8,881$ Incline conveyor feed system inc.: stand and modifications 3,400$ Zabel - 10" Hydraulic pump system 10,348$ Hydraulic pump feed system inc.: piping, valves and fabrication 21,180$ Material loading systems support components, installation and modifications 19,980$ Structural steel, fabrication and installation 6,480$

Decant auger systems HT1 and HT2 inc.: valve, fi tting and fabrication 13,084$ Biogas pressure and vacum controls system inc.: piping, valves, control components, instramentation and fabrication 17,842$ Biogas clean-up and processing system including water knock-outs, check valve, flame arrestor, valves, piping., controls, instramentation and fabrication 23,847$ Biogas collection system inc.: piping, valves, control components and fabrication 21,892$ Sage Biogas flow meter 3,900$ Biohyddrogen collection system inc.: piping, valves, control components and fabrication 23,480$ Hydrolyic mixing and circulation systems including piping, valves, pumps, flow meter, Mazzi injectors 16,420$ Gasification and buffer tank circulation systems including piping, valves, pumps, flow meter, Mazzi injectors and fabrication 28,640$ 10 X 20 Expanded steel equipment unit 8,300$ GM 42 hp engine and generator system 16,370$ Computer control instramentation, components and installation materials 114,870$ Instramentaion tesing and calibration - Brown & Caldwell 15,789$ Computer controls programing, screen development and test Brown & Caldwell 78,606$ Air controls systems inc: lines, panels, controls, fabracation and installation 18,170$ Site electrical - computer controls and instramentation 29,843$

Site electrical - Pumps, motors, equipment and Motor Control Center construction and installation 32,487$ Computer control building remodel, electrical, panels and equipment 9,852$ General construction and project operational overhead 34,659$ Rental equipment, trucking and crane services 28,641$ Material processing fabracation, welding and consturction 28,292$ Stainless stell valves, fittings, pipe and materials 42,823$ Hydraulic mixing systems components 25,874$ Hydraulic mixing and circulation systems fabraction, welding and construction 26,487$

$

62

CHAPTER 3: Operation and Testing of Pilot Digester System with Solid Food Waste 3.1 Modifications Made Prior to Operation and Testing The Anaerobic Phased Solids (APS) Digester system was designed and built at the University of California at Davis as described in the previous chapters. The original pilot plant design included four hydrolysis tanks, but for digestion of food waste only two hydrolysis tanks were needed, as shown in the figure below. Initial testing revealed the superior mixing and material handling capabilities of the Vaughan in‐line chopper pump. Therefore, hydrolysis tanks 1 and 2 were utilized.

The feedstock loading system was modified slightly from the previous description after testing with various feedstocks (for example manure and source‐separated municipal solid waste). A hydraulic bin‐dumping station was installed to reduce the workload for the forklift operator as well as the cost of the forklift, since a simple lift without bin‐dumping capabilities could be utilized. A chopped pump replaced the conveyer system for transferring feedstock to the hydraulic piston pumps. This also allowed for size reduction of the feedstock. A recirculation line allowed for addition of process water to the feedstock receiving hopper to dilute feedstock to pumpable consistency without adding clean water.

The gas collection system remained as described previously. However, the thermal mass‐flow gas meter upstream of the hydrogen sulfide scrubber was removed due to technical issues. Hydrogen sulfide concentrations were monitored manually using sampling tubes. In addition, only one set of gas collection lines were utilized to allow for interconnection of the head spaces to maintain pressure balance.

The biogas generator could not ultimately be connected to the utility grid due to financial and regulatory issues. This limited the ability to operate the generator. Furthermore, the biogas fueled boiler was unable to operate due to fluctuating biogas flow rates. A biogas storage system would have buffered these fluctuations, but was not financially feasible for testing and is recommended for future research. The electrical generator was tested on biogas to prove operability, but biogas was primarily flared during the testing period.

Effluent from the digester was discharged directly to the adjacent wastewater treatment plant during testing. The high level of solids reduction and solids retention eliminated the need for solid/liquid separation for the effluent stream which was 97 percent water.

63

Figure 41. Schematic of APS Digester System

BUFFERTANK

BIOGASIFICATION REACTOR

HYDROLYSISHR 1

HYDROLYSISHR 2

Liquid

LIQUID FED TO BUFFER TANK

BIOGAS(CH4+CO2)

FEED

BIOGAS(H2+CO2))

Effluent

3.2 Operational Experience

3.2.1 Provision and Loading of Feedstock The APS Digester system was designed to treat solid organic wastes from a variety of agricultural, industrial and municipal sources. In practice, cost scaling curves are steep for digesters, providing financial incentives for larger systems. Waste streams from individual sources are often too small to justify anaerobic digestion. Therefore, larger centralized systems are more financially feasible than small distributed systems. In such systems, waste streams will need to be transported to the digester as they become available. To minimize trucking expenses, waste streams should be dewatered if possible and trucks should be loaded to their maximum capacity. The APS Digester system was designed with these constraints in mind.

The phased‐batch loading of the APS Digester system allows for feedstock to be loaded as it arrives at the facility, without the need for equalization tanks. The hydrolysis tanks of the APS Digester system act simultaneously as equalization and acidification tanks. However, loading waste in large batches results in uneven gas production, with large spikes in organic acid production occurring immediately after loading. These spikes are due to the natural progression of the decomposition process which begins with acid forming hydrolysis of the organic wastes. Typical continuous digester systems may fail when large quantities of acid are produced, but the APS Digester configuration shields the methanogenic microorganisms from sudden pH changes. As a result, the APS Digester system was able to sustain steady operation despite sporadic batch loading and

64

spikes in acid and gas production. If more consistent gas production rates are required, wastes can be loaded on a more frequent, consistent basis.

For this operational testing period, food waste was trucked to the digester and loaded weekly. This helped reduce shipping costs and allowed for proof‐of‐concept testing over the course of a sustained 8‐month period. The quantity of food waste required to maintain steady operation of the system and achieve performance targets was one to three tons‐per‐day of solid wastes with an average moisture contents of 75‐85 percent. Therefore, arrangements were made to have solid wastes delivered weekly by truck with up to ten tons per week loaded into the digester in a single batch. Although attempts were made to have batches delivered regularly, scheduling conflicts often resulted in batches being loaded on different days each week. However, on average, batches were loaded every seven days, except during the start‐up period when loading occurred less frequently as the loading rate was ramped up to the target loading rate. A timeline of the APS digester operation is shown below.

Figure 42. Timeline of APS Digester Operations

Food waste feedstock was delivered in stackable plastic bins as shown below. Each bin measured 1.0 cubic yards and held approximately one‐half ton of wet waste. However, the exact quantity of organic matter in the organic‐waste batch depended on the extent of filling of the bins, the moisture content of the waste, and the organic fraction of the solids (determined using the volatile solids fraction as a proxy). The number of bins delivered was determined based on the target loading rate, but the actual loading rate was unknown until after the food waste samples were analyzed for total and volatile solids content. The analysis required two days and waste batches needed to be loaded as they arrived to reduce odors and waste degradation. As a result, the loading rate

65

fluctuated, just as it would in a full‐scale setting where organic wastes would be loaded as they are generated.

Figure 43. Feedstock Bins, Stackable Plastic Bins Used to Deliver Batches of Organic Waste to the APS Digester System


In order to test the pilot APS Digester system, a consistent organic waste stream of sufficient quantity and quality was required. The UC Davis campus food‐service waste stream was determined to be of insufficient quantity at the time that testing began in June of 2008. Two outside solid waste providers were contacted to determine the optimal waste stream for testing the APS Digester system. Initially, a municipal solid waste composting facility was contacted to determine whether they could provide waste of consistent quantity and quality. When the waste quality was determined to be too variable, waste from a food processing facility was tested. This waste stream was found to be of consistent quality (for example particle size, moisture content, and general composition) and sufficient quantity. Furthermore, the proximity of the food processing facility reduced shipping cost and provided for the flexibility in feedstock provision necessary for an effective scientific study.

3.2.1.1 Organic-Fraction of Municipal Solid Waste

Initially, the Jepson Prairie (Dixon, CA) composting facility provided food waste mechanically separated from restaurant waste. Approximately ten tons of food waste were delivered to UC Davis in three separate batches.

66

Figure 44. Feedstock Sampling. Sampling Food-Waste From a Municipal Food-Waste Composting Facility.


The food waste was loaded into the digesters. However, during loading, large fragments of non‐digestible materials (for example wooden boards, metals, and plastics) were found in the waste stream. These lead to mechanical failures in the material handling system which delayed feeding and added expenses to the operation. In order to test the food waste from Jepson Prairie in the digester system, it was determined that additional waste pre‐processing (such as screening) and debris‐handling equipment would be needed either at the source or on‐site ahead of the loading auger. The solid waste provider was unable to provide the additional processing required, and project funds were limited. Therefore, an alternate source of food waste was sought.

67

Figure 45. Transfer to Chopper Pump. Transfer from Feed Hopper to Chopper Pump Hopper of Food Waste From a Municipal Solid Waste Composting Facility.

Photo credit: Onsite UC Davis.

3.2.1.2 Food-Processing Waste

The Campbell’s® processing plant (Sacramento, CA) ultimately provided the food waste required for this study. The waste stream consisted almost entirely of organic materials (such as vegetables, rice, noodles, and meat) with minimal contaminants (with the exception of sporadic tin cans). Food waste was delivered and loaded into the digester system without difficulty. Arrangements were made to secure weekly deliveries of food waste, transported via a flat‐bed truck in up to 24 lidded, stackable plastic bins.

Figure 46. Campbell’s® Processing Waste. Food-Waste From Campbell’s® Processing Plant (Sacramento, CA) as Delivered to the Pilot Plant at UC Davis

Photo credit: UC Davis.

68

Bins were weighed upon receipt, and samples of the food waste were collected for analysis. The carbon/nitrogen ratio was generally between 10 to15, but occasionally ranged higher. Protein content was inferred from the nitrogen and ash content. The crude fat content ranged from 7to– 22 percent, and most likely varied depending on how much meat was being processed at the Campbell’s® factory.

Table 5. Chemical Composition of Food-Waste Samples (Carbon, Nitrogen, Carbon: Nitrogen Ratio, Protein, Crude Fat, Phosphorus, and Potassium) on a Dry-Mass Basis

Date C (%) N (%) C:N Protein (%) Crude Fat (%) P (%) K (%)

06/16/08 58.1 2.04 28.5 12.7 NA 0.139 0.34

11/19/08 50.8 2.87 17.7 17.9 22.05 0.208 0.35

02/19/09 46.8 4.15 11.3 25.9 7.62 0.272 0.42

02/25/09 47.9 3.74 12.8 23.4 12.21 0.216 0.47

03/03/09 48.5 3.46 14.0 21.7 15.39 0.215 0.66

03/12/09 49.3 3.31 14.9 20.7 18.36 0.215 0.42

Average 50.23 3.26 16.53 20.38 15.13 0.211 0.44

Source: UC Davis ANR Analytical Laboratory.

A forklift transferred the bins from the truck to the loading dock, lids were removed manually and bins were individually lifted onto the tipping station. A hydraulic arm tipped the bin into the loading hopper. Bins were rinsed into the loading hopper using a minimal amount of reclaimed water from the adjacent wastewater treatment station. Less than 250 gallons of water (less than 1 percent of reactor volume) were added per loading event.

69

Figure 47. Campbell’s® Feedstock Loading. Feedstock loading. Clockwise from Top Left: Bin Receiving, Sampling, Transfer, and Tipping


Inside the loading hopper, a rotating auger transferred waste into the chopper‐pump hopper. Effluent from the digester was flushed through the loading hopper to facilitate the transfer. The chopper pump simultaneously chopped and pumped the food waste into a hydraulic‐ram pump for loading into the digester. A pair of pneumatic gate valves directed the food waste from the hydraulic‐ram pump into one of two 10,000 gallon hydrolysis tanks. Each batch was loaded entirely into one tank, and loading alternated between the two tanks. Thus, if a batch were loaded into one hydrolysis tank, the opposite tank would be loaded the following week. This allowed hydrolysis to occur for two weeks per batch.

70

Figure 48. Material Handling Equipment Food-Waste Hydraulic Bin Tipping Unit, Loading Hopper, and Chopper-Pump Hopper (Left), and Hydraulic Ram Pump (Right)


The loading station was controlled manually. Loading 10 tons of waste required approximately two hours in the absence of delays. In a full‐scale operation, the loading station could be designed to accommodate larger quantities of waste in a shorter amount of time by increasing the hopper size, adding extra pumping stations, and/or increasing the piping diameter.

Samples were collected from each bin individually during every loading event and analyzed. During the first run, ten bins at a time were loaded into the digester every two weeks for the first eleven weeks. For the next seven weeks, twenty bins were loaded once per week, at which point the biogasification tank became acidic and loading was halted. During this initial operational period, the total and volatile solids contents of the food waste deliveries ranged from 13to 27 percent and 95‐98 percent on a wet and dry basis, respectively. The large variation in solids content led to a faster increase in organic loading rate than intended, especially in the last three weeks of the first run, which may have contributed to the acidification of the biogasification tank.

During the second operational period, loading was ramped up more slowly. Initially, five bins per week were loaded for three weeks, then ten bins per week were loaded for the next five weeks, and fifteen bins per week were loaded for the final twelve weeks of the operational period. Plans had been put in place to raise the loading rate to twenty bins per week, but the average total and volatile solids contents of the food waste during the last four weeks of operations increased, most likely due to a change at the Campbell’s® processing facility. This naturally raised the organic loading rate to the target level without the need to increase the wet mass of waste loaded. Additionally, the methane yield began to decline, indicating that the health of the microorganisms was deteriorating. Thus the number of bins loaded remained constant until the digester

71

became biologically stable. When methane yield continued to decline without any noticeable variation in pH, the ammonia concentration of the biogasification reactor was tested and found to be higher than desirable. At this point, loading ceased.

Table 6. Solids Content Analysis for all Campbell’s® Food Waste Loaded into the APS Digester System (MC = Moisture Content, TS = Total Solids, VS = Volatile Solids)

Run Date Time Interval

(days) Wet Mass (lbs) MC (%) TS (%) VS/TS (%) VS (%)

1 5/28/2008 ‐ 5,077 81.8 18.2 96.0 17.5

6/13/2008 16 12,130 72.9 27.1 95.1 25.8

6/23/2008 10 13,841 74.8 25.2 97.1 24.5

7/11/2008 18 11,709 76.4 23.6 97.3 22.9

7/30/2008 19 22,828 87.5 12.5 95.8 12.0

8/14/2008 15 13,760 84.1 15.9 96.4 15.4

8/21/2008 7 24,459 79.6 20.4 97.1 19.8

8/27/2008 6 22,965 82.7 17.3 97.3 16.8

9/5/2008 9 25,580 78.8 21.2 97.1 20.6

9/10/2008 5 23,579 80.9 19.1 96.4 18.4

9/17/2008 7 26,526 78.3 21.7 97.2 21.1

9/25/2008 8 26,238 77.3 22.7 97.5 22.1

Subtotal 228,692 81.0 19.0 96.5 18.3

2 11/13/2008 ‐ 4,975 77.6 22.4 97.3 21.8

11/19/2008 6 13,034 78.5 21.5 96.9 20.8

12/3/2008 14 12,005 79.3 20.7 97.1 20.1

12/11/2008 8 11,734 73.8 26.2 96.7 25.3

12/17/2008 6 11,353 77.4 22.6 96.9 21.9

12/22/2008 5 12,585 76.7 23.3 97.4 22.7

12/30/2008 8 12,620 82.2 17.8 96.9 17.2

72

3.2.2 Start-up and Development of Steady-State Operations Starting a digester requires balancing reactor conditions to support both faster growing acidifying bacteria and slower growing methanogens. Acidogens and methanogens are codependent for survival. Methanogens consume the metabolites of acetogens thereby removing inhibitory hydrogen and volatile fatty acids and keeping acid levels in a range favorable to the growth of both organisms. Overloading a digester during startup can result in fatty acid accumulation that overwhelms the methanogenic community, leading to decreased pH and reactor failure. For this reason, loading was gradually increased during startup to promote the development of a stable microbial community. The start‐up period for digesters typically lasts one to three months. Reducing the start‐up time requires starting with a healthy inoculum and carefully managing the initial feeding schedule to provide sufficient time for microbial growth while maintaining a steady supply of food for the microbes.

Initially, bacterial inoculum was added to the digester in the form of 3,000 gallons of sludge from an operational thermophilic digester at the East Bay Municipal Utility District wastewater treatment plant. The sludge was immediately brought to a temperature of 45°C, which was within 10°C of the digester operating temperature (55°C). Gas production was recorded and feeding began within one week of adding the inoculum. Initially, plans were made to operate all four hydrolysis tanks. Two of the tanks would be mixed using a sewage pump, and the other two would be mixed using a

1/7/2009 8 12,120 72.7 27.3 97.3 26.6

1/15/2009 8 18,150 77.7 22.3 97.0 21.6

1/27/2009 12 18,060 82.3 17.7 96.0 17.0

2/2/2009 6 19,140 80.4 19.6 97.1 19.0

2/19/2009 17 16,414 76.1 23.9 93.9 22.5

2/25/2009 6 18,140 80.0 20.0 92.2 18.4

3/4/2009 7 14,600 82.5 17.5 92.5 16.2

3/12/2009 8 18,860 76.9 23.1 96.6 22.3

3/18/2009 6 21,160 72.5 27.5 97.3 26.7

3/26/2009 8 21,140 77.9 22.1 95.2 21.0

4/1/2009 6 20,960 78.3 21.7 96.7 21.0

Subtotal 277,050 78.0 22.0 96.1 21.2

Grand Total 505,742 79.5% 20.5 96.3 19.8

73

recently installed chopper pump. This would have allowed for comparison of the two pumps. However, shortly after loading it became apparent that the chopper pump was more efficient and reliable. Therefore, the sewage pump was removed and the remaining tanks connected to the chopper pump were brought online for full operation. Although an additional chopper pump was purchased, it quickly became clear that only two hydrolysis reactors were required, and operations would be simplified by using two tanks instead of four. The batch of food waste initially loaded into the now unused hydrolysis tanks was transferred into the two operational hydrolysis tanks. This allowed for gradual feeding of the biogasification tank for several weeks while arrangements were made for trucking new batches of food waste to the pilot plant.

The inoculum was allowed to stabilize in the biogasification tank for a week before mixing commenced. The generation of biogas without Food waste addition (Figure 49) was probably a result of the degradation of residual organic material in the inoculum. This signaled that the methanogenic community was active and that feeding could begin. Food waste was loaded eight days after inoculation, resulting in a steady increase in biogas production.

Figure 49. Biogas Production During Startup of the APS Digester System

3.2.3 Maintenance, Repairs and Upgrades During the operational and testing period, the system performed as designed. There were some areas where design improvements could be incorporated to increase

74

performance and/or efficiency. Equipment performance was also monitored and allowed us to identify equipment that integrated well into the system, or needed to be replaced in future designs. Additionally, during the operational period some minor maintenance and repairs were completed.

Material Receiving and Chopper Pump Systems Food waste was delivered in lidded bins from the Campbell’s® facility in Sacramento where the material was passed over a screening unit prior to delivery. In a full‐scale operation, food waste could be delivered by tanker trucks and pumped directly into the digester. On one occasion this was attempted at the pilot plant. Food waste became stuck in the tanker because the truck’s pump was not designed to handle high‐solids. It was also very difficult to collect reliable samples for solids analysis. In a full‐scale operation, process water could be used to assist in removal of solids from a tanker truck.

Even with the pre‐screening, non‐biodegradable items were regularly present in the bins including soup cans, plastic lids, water bottles, and small metal pieces including various nuts, bolts and other metals. Most of the heavy inorganic pieces were removed through the hydro‐float unit installed on lower section of the material receiving hopper. The hydro‐float unit allowed heavy inorganic material to settle during loading so that they were not transferred to the chopper pump. Evaluation of the hydro‐float unit determined that if the length of the float were extended by two to three feet to obtain a higher float time, additional heavy contaminants would be removed.

Soup cans, plastic lids and water bottles and other light inorganic materials passed through the receiving hopper and into the chopper pump auger hopper. This auger system pushed the plastic and light inorganic material through the chopper pump, which shredded these materials and pumped them into the hydrolysis reactor. Some cans and plastics could be removed in the clean‐out area between the chopper pump auger and chopper pump unit. The original design required about thirty minutes to open the clean‐out cover. A quick‐release latch was installed for opening the clean‐out cover to reduce the clean‐out time to about five minutes.

Hydraulic Piston Pumps – Loading Hydrolysis Reactors Once the material passed through the material‐receiving and chopper‐pump auger systems, the chopper pump delivered the material to the hydraulic‐piston‐pump hopper. The hydraulic piston pump is located between the two hydrolysis reactors. A pneumatic valve on the reactor to be loaded was opened from the control panel prior to loading; the hydraulic piston pump would then pump the material into the selected reactor. A clear cover was installed atop the hydraulic‐piston‐pump hopper which stopped any splashing of material out of the hopper and also helped to contain odors generated from material left in the hopper between loadings.

Reactor Hydraulic Mixing System Hydraulic reactor mixing systems were redesigned based on operational information collected from mechanical operation tests conducted in late 2007 and early 2008. These

75

changes included the installation of a chopper pump used as the hydraulic mixing pump which provided continuous particle size reduction in addition to mixing. Liquid and solids drawn from the hydrolysis reactor were pumped through the chopper pump and returned to the reactor through two mixing nozzles placed at angles designed to mix the reactor contents. This unit performed as anticipated and required no maintenance or modifications during the operational and testing period. Prior to startup, the hydrolysis tanks were filled with water and mixed while a large port atop the tank was opened for visual inspection of the mixing behavior. Suspended materials in the water were observed to be pulled down into the mixing vortex. Furthermore, during the operational period, samples collected before and after mixing revealed differences in solids content and pH that indicated mixing was occurring.

Decant Units Two decant units with screens were installed on each hydrolysis reactor allowing water to be removed from the reactor while keeping the larger solid particles in the reactor. During the operational and testing period, the decant units were back‐flushed on a regular basis to prevent the screens from becoming caked with solids. After about seven months of operation, it was noted that the external seals on the decant auger shaft were starting to leak. The decant units were designed to be isolated from the reactor, allowing maintenance work to be completed without interrupting the digester processes. Each of the decant units were serviced after eight months of operation; the seals were replaced without interrupting the digestion process.

Biogasification Reactor Mixing System During the operational and testing period, the biogasification reactor’s hydraulic mixing system was evaluated, and although it mixed the reactor and provided leachate recirculation for at least 10 months, it was determined that improvements could be made. After reviewing operational flow data, it was suspected that the relatively small sewage pump powering the biogasification‐reactor mixing system was not providing adequate mixing of the reactor. The installed unit pulled liquid out of the reactor through a 10‐horsepower pump and returned the liquid through three , two‐inch nozzles. This did not provide for enough liquid flow for adequate mixing. It is recommended that the hydraulic mixing system on the biogasification reactor should be replaced with the same mixing system installed on the hydrolysis reactors (Vaughan pump and mixing nozzles) without a chopper unit on the mixing pump.

Biogas Collection and Processing The biogas collection and clean up systems performed as anticipated. Biogas generated in the hydrolysis reactors was directed into the biogasification reactor. Then, the biogas was collected from the biogasification reactor passed through a hydrogen sulfide removal system to a flow meter and finally to a flare installed on site. During the operational period, it was noted that the biogas flows were higher than expected. Upon inspection, it was noted that moisture in the line that housed the biogas flow meter was

76

affecting the performance of the flow meter, and that inaccurate measurements were being logged by the computer.

During operation, the biogas line was isolated, purged, and the mounting connection for the flow meter was modified to keep the flow meter probe out of the moisture that would build up. Additionally, a self‐draining moisture collection unit was installed on the line housing the biogas flow meter to allow automatic drainage of moisture from the line. After installing the moisture removal unit, the flow meter resumed normal operations.

Funds were only available for installation of one gas flow meter, and the headspaces of the three reactors had to be connected to allow for equalization of the liquid volumes. This meant that the gas flow rate from each tank could not be measured separately. In order to make these measurements and to separately collect the gas from the hydrolysis and biogasification reactors, the gas collection system would have to be modified. This would involve installing check valves and separate flow control valves with individual gas pressure sensors, and revising the control program to allow for separation of the gas streams.

Boiler System A boiler system (LAARS Heating Systems, Co., Rochester, NH) was installed and modified to accept propane instead of natural gas. The propane conversion unit failed numerous times during the operational and testing period. A service company made frequent modifications to both the boiler and the propane conversion components; however, none of these modifications adequately repaired the unit. Repeated failure of the boiler system resulted in downtimes ranging from several hours to over one week which in turn made heating of the digesters somewhat inconsistent. It was also determined that the propane flow through the boiler was not properly designed by the subcontractor. Modifications made to the computer control program to adjust the propane flows based on outlet and inlet water temperatures improved the boiler’s performance by allowing the boiler to cycle on and off more efficiently. However, it did not eliminate the boiler problems.

It is recommended that the existing boiler system should be replaced with a boiler designed to operate strictly on propane at higher efficiency. Almost all potential commercial project sites have access to natural gas, but the location of the pilot digester facility on campus does not have access to a natural gas line. Thus propane will have to remain the start‐up and maintenance heating fuel source. On full‐scale commercial digester projects, a natural gas/biogas boiler can be integrated into the design along with the option to utilize recoverable heat from power generation, biogas compression units, or existing recoverable heat at the installation location.

In addition to providing heat with a more efficient and reliable boiler, improvements in the insulation of the heat‐transfer pipes would also improve the energy efficiency of the plant. During mixing, the temperature of the reactors declined faster than when

77

unmixed, suggesting that heat loss through the uninsulated piping system was significant. An attempt was made to insulate the largest liquid transfer pipes with foam wrap. However, this was only a temporary fix; a more permanent insulation should be installed on the heat‐transfer pipes at the pilot plant. For a full‐scale installation, all pipes should be insulated with polyurethane foam, especially for thermophilic reactors and in locations with low winter and nighttime temperatures. Polyurethane foam was successfully used to insulate the reactors. Heat loss from the reactors in the absence of mixing or heating was minimal, demonstrating the effectiveness of the insulation.

Computer Controls and Operational Program During the operational and testing period, the computer control system performed adequately, allowing the plant operators to change the mixing, heating, liquid recirculation, and gas collection, schedules as needed. The ability to manually open and close valve and start and stop pumps was crucial for tailoring plant operation. In addition, sufficient automation was incorporated to reduce operator errors. Numerous meetings were held to discuss the flexibility of the control program relating to operational ease, data collection, and process control with the goal of improving overall operational efficiency. Numerous control program improvements were identified including the ability to make real time process control changes, set alarms, change data collection settings, and manage security access. Some of the identified improvements were incorporated into the existing control program. Others were documented and will be incorporated into the redesigned computer control program in the future.

Data on system operation and performance was stored in a database and retrieved using software (ActiveFactory, Wonderware, Inc.) that allowed for extensive data analysis. The time investment for learning to operate the data collection system was significant, and needs to be taken into account when designing training programs for future plant operators. In addition, it will be helpful for plant operators to have training in programming SQL for data acquisition.

The lack of secure remote data access also added substantial time to the operation of the plant and data collection efforts. Adding wireless remote access to the control system and data collection system will greatly improve plant safety as well as reduce the cost of managing such a complex facility. Future upgrades to the pilot plant should include extension of wireless connectivity, integration of alerts and system status reports into an automated reporting protocol, and regular data backup to a secure remote server. The latter would also enable operators to access plant data without adding any load to the CPU of the primary operating computers.

Several times, the operating computer had to be restarted after a software crash in the midst of plant operations. The plant continued to operate successfully during these incidents, but data was not collected during the system reset and operators temporarily lost the ability to control plant operations. At a full‐scale facility, redundancy and backup systems will be incorporated to prevent any possible problems due to computer failures.

78

Instrumentation During the operational period, all of the various instrumentation components including liquid and gas flow meters, pressure sensors (tank and biogas), level sensors, and temperature sensors performed as anticipated. Moisture in the biogas lines did affect biogas flow meter measurements, but the modifications made corrected these problems.

The instrumentation components installed to measure the pH did not perform and were considered as nonfunctional. Numerous attempts were made to clean, re‐calibrate and manually test the two pH sensors; however none of these efforts had any positive results. Upon review of manufacturer information and other commercial installations, it was determined that if the inline pH sensors were to be used, they would need to be cleaned and recalibrated daily using a combination of manual methods and a re‐circulating water wash. Daily manual pH testing is rapid and sufficient for control of plant operations, and it may prove to be less costly than installing a working inline pH testing system.

Reactor volumes were measured by a pressure transducer at the bottom of the biogas reactor. This meant that the estimated tank volume fluctuated with the gas pressure in the headspace, which had to be subtracted programmatically from the tank volume using computer software. As such, the tank volume measurement was less accurate than desired, but accurate enough for well‐maintained digester operations.

Temperature sensors were placed in the liquid transfer pipes rather than within the reactors themselves. This meant that accurate reactor temperatures were only measured during mixing operations. While this was sufficient for monitoring reactor temperature, it made data analysis difficult, since the temperature fluctuated between mixing events. In the future, either a change needs to be made to the computer control system instructing the system only to record temperature data several minutes after mixing has begun, or additional temperature sensors need to be placed in the reactors such that accurate resting temperature measurements can be made.

Overall Digester System Mechanical Review From an overall system mechanical, operational, and performance review, the digester system performed well within design expectations. There were several mechanical and design issues; however, they were expected for this first full‐scale operational run. Based on all of the operational and performance data collected, only minimal modifications to the existing pilot digester facility will be needed before resuming operations. These modifications will improve digester performance and efficiency resulting in a more robust system. The immediately planned modifications include:

• Replacement of the biogasification mixing pump with a more powerful and efficient model

• Replacement of the existing boiler with a more efficient and robust model

• Extension of wireless internet access and remote operation capabilities

79

• Modification of the control software to include alarms and additional controls

• Connectivity of the biogas‐powered engine generator for self‐sustaining system operation

3.3 APS Digester Performance Results

3.3.1 Data Collection and Analytical Methods Instrument Calibration

Liquid Level Meters Each tank at the APS Pilot Plant was built with a designed capacity of 10,000 gallons. During the operation described herein, three reactors were run for a total capacity of 30,000 gallons. The average working volume was 25,642 gallons (± 2,355 gallons) or 85 percent (± 8 percent) of design capacity. Tank volumes were held relatively stable throughout the operational period, except when the tanks were drained by about 3,000 gallons in November, 2008 to make room for additional inoculum after the acidification event. Recirculation of leachate between the hydrolysis and biogasification reactor was controlled by measuring the change in each tank’s volumes during recirculation. Because the volume measurement was found to drift, possibly due to changes in gas pressure, the control software was designed to measure all three tanks’ volumes just prior to recirculation. However, a bug intermittently prevented the software from taking the prescribed volume measurements. Therefore, tank volumes were checked and manually adjusted during loading operations. This was responsible for the variation in tank volume seen in Figure 50.

80

Figure 50. Total Working Volume and Contribution of Each Tank to the Overall Volume

Gas Flow Meters

Gas flow was measured with a Sage thermal mass flow meter located near the exit of the hydrogen sulfide scrubber. No external pressure or blower was applied to the gas. Gas transport and flow rate was established entirely by the buildup of gas pressure in all three reactors’ headspaces. In order to provide sufficient gas pressure for the gas conversion systems, a control valve was installed adjacent to the flare. The control software was programmed to open and close the flare control valve at manually adjustable set points. As a result, gas flow was periodic, as seen in Figure 51, although the flare could be fixed in the open position if needed or when gas production was high enough to maintain a gas pressure consistently above the closing set point.

81

Figure 51. Typical Gas Production Profile

Gas flow rate

Tank Pressure

Time

Because only one flow meter was available due to its high cost, only the combined gas flow from all three tanks was measured. Sage thermal mass flow meters were calibrated using a mix of 65 percent methane and 35 percent carbon dioxide. A model was provided by Sage Metering, Inc. for adjusting the gas flow rate reading for different gas mixes. This model revealed that the flow meter would under‐estimate the gas flow rate when the methane content was less than 65 percent, and it would over‐estimate the flow rate when the methane content was over 65 percent (see Figure 52), although the flow meter would measure the biogas flow‐rate with ±5 percent accuracy when the methane content was 53‐77 percent. Nonetheless, the measured gas flow rate was adjusted programmatically based on the measured methane content to more accurately represent the actual gas production rate.

82

Figure 52. Gas Flow-Rate Adjustment Chart for Sage Thermal Mass Flow Meters Calibrated on a Mix Of 65 Percent CH4 and 35 Percent CO2

Biogas Methane Content (%)

0 10 20 30 40 50 60 70 80Gas

Flo

w-r

ate

Adj

ustm

ent F

acto

r (%

)

90

100

110

120

130

140

+5%

-5%

y = 0.2418x2 - 0.7163x + 1.3613R² = 0.9999

Furthermore, thermal gas flow meters depend heavily on the size and shape of the gas pipe in which the sensor is placed. Condensation was discovered in the pipe used for measuring gas‐flow rate. Upon removing the water, the gas flow meter reading immediately dropped proportionally to the amount of water removed (see Figure 53 ). This was most likely due to the alteration of the pipe depth. After discovering the condensation, an automatic drainage valve was installed in the gas pipeline. However, the majority of the gas flow data were collected prior to installing the new valve. Therefore, different quantities of water were added and removed from the pipeline to determine the magnitude of the effect (see Figure 53). The volume of water in the gas pipe at any given point was unknown. However, 1.5 liters were withdrawn the first time water was removed from the pipe on February 6, 2009. It can be assumed that the amount of water in the pipe would be a function of the ambient temperature, which was at its lowest when water was first removed. During the rest of the year, less water should have been in the pipeline, since the biogas would have been more highly saturated at higher ambient temperatures. To be conservative, gas flow rates were reduced by 60 percent prior to February 6, 2009 for this report. At this level, the gas flow measurements immediately after February 6, 2009 were within 5 percent of those before this date.

83

Figure 53. Effect on Measured Biogas Flow Rate (●) of Adding/Removing Water (▲) From Gas Flow Line

Date

2/2/09 2/9/09 2/16/09 2/23/09 3/2/09

Bio

gas F

low

Rat

e (s

cf./h

)

0

200

400

600

800

1000

Wat

er A

dded

to G

as L

ine

(L)

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

Figure 54. Relation Between Water Added to Gas Pipe and Change in Measured Biogas Flow Rate

-69.1%

-47.3%

41.9%

y = 0.4782xR² = 0.8562

-80.0%

-60.0%

-40.0%

-20.0%

0.0%

20.0%

40.0%

60.0%

-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0

Cha

nge i

n A

vg G

as F

low

Rat

e (%

)

Volume of Water Added (L)

84

pH Meters

Throughout the operational period, pH was measured using a laboratory grade pH meter (Accumet, Inc.). The probe was located at the pilot facility to reduce the time between sampling and pH measurement. The probe was calibrated using three calibration standards at pH 4.1, 7.0, and 10.1 (Thermo Fisher Scientific, Inc.) whenever the calibration was off by more than 1 percent when checked against the pH 7 standard.

Mixing, Heating And Recirculation Scheduling

A complex loading, mixing, heating, and leachate recirculation schedule was developed to balance the needs of the digester system. The biogasification tank was heated using a crossflow heat exchanger, such that the tank contents were pumped out of the tank, through a heat exchanger, and back into the tank, while hot water was simultaneously circulated from the boiler, through the heat exchanger, and back to the boiler. The same hot water pump was used to circulate hot hater through the two water jackets of the hydrolysis tanks. The boiler was sized to heat only one tank at a time, therefore each tank was heated in series, and each tank had to be mixed during heating to transfer heat from the water jacket to the reactor contents. This meant that the total time available for heating each tank was limited, and when one tank required more heating, the mixing times all had to be adjusted accordingly. Therefore, the temperatures of the tanks had to be monitored regularly and the mixing schedule adjusted.

Leachate recirculation typically occurred four times per day, although the number of recirculation events could be adjusted by the plant operator. The contents of the hydrolysis tanks were allowed to settle for at least one hour prior to recirculation. Leachate was transferred between the tanks as follows:

1. A fixed volume (V1) was transferred from Hydrolysis Tank 1 to the Buffer Tank.

2. A fixed volume (V2) was transferred from Hydrolysis Tank 2 to the Buffer Tank.

3. A volume equal to V1 was transferred from the Biogasification Tank into Hydrolysis Tank 1.

4. A volume equal to V2 was transferred from the Biogasification Tank into Hydrolysis Tank 2.

5. The contents of the Buffer Tank were transferred into the Biogasification Tank.

Each recirculation event required between 30 minutes and an hour, except when the decant augers became clogged. Decant augers were backwashed regularly prior to recirculation to prevent clogging, but as the organic loading rate increased the augers had a tendency to drain more slowly. This usually remedied itself after a couple days of hydrolysis, at which point the solids disintegrated enough to avoid clogging the auger.

Loading new batches of food waste into the tanks was performed manually, and recirculation was manually halted during the loading event. Batches were typically loaded in less than six hours. Thus, the loading did not significantly interrupt the

85

mixing, heating, and recirculation schedule. A sample timing schedule for loading, mixing, recirculation, and sampling is shown below.

Figure 55. Example of Typical Sampling Schedule, TK300 = Biogasification Tank, TK101 = Hydrolysis Tank 1, and TK102 = Hydrolysis Tank 2

Sampling Methodology

Gas, liquid, and solid samples were collected regularly. Food waste for solids analysis was always sampled immediately prior to loading, and shipments of food waste were typically loaded into the digester within 48 hours of delivery. Food waste samples were collected from each bin separately, and bins were weighed upon receipt. Initially, the solids contents of the samples were analyzed individually, to check for variability between the bins. When it was discovered that the solids content varied significantly between bins, composite samples were assembled in proportion to the relative mass of each bin prior to performing solids analysis. Food waste samples were transferred to the lab and stored at 4°C within 3 hours of collection. Solids analysis was performed within several days of sample collection. Moisture, total solids, volatile solids, and ash content

86

were analyzed using a standard combustion method (Standard Methods for the Examination of Water and Wastewater). The remaining composite samples were stored at ‐20°C. A small subsample of food wasteloaded throughout the operational period was analyzed for carbon, nitrogen, protein, fat, phosphorus, and potassium content at the end of the operational period.

Gas and liquid samples were typically taken from the tanks three times per week. Liquid samples from the biogasification tank, and both hydrolysis tanks were taken after each reactor had finished mixing. Liquid samples were also collected during transfer of liquid between the tanks during recirculation. Leachate was transferred from each of the hydrolysis reactors into the buffer tank prior to being transferred into the biogasification tank. Therefore, liquid samples being transferred into the biogasification tank represented a sample of leachate from both hydrolysis tanks after allowing solids to settle out. The pH of each liquid sample was measured at the pilot plant immediately after the sample was drawn. Liquid samples were then transferred to the lab within three hours of collection and stored at 4°C until being analyzed for solids content as described above for food waste samples. Liquid samples were also occasionally analyzed for ammonia and BOD. Liquid samples were stored at ‐20°C after solids analysis had been performed.

Biogas samples were collected either from a port located between the hydrogen sulfide scrubber and the flare, or from ports atop each of the tanks. For mixed biogas representative of what would typically enter the boiler, generator, or flare, the former port was used. For samples of the biogas coming from each of the separate tanks, the ports on top of the tanks were used, and the tank headspace was isolated from the rest of the biogas collection system using a manually operated valve. This was meant to reduce the influence of biogas composition from neighboring tanks on the biogas sample from a specific tank. However, because the headspaces were connected, the biogas samples may still have been contaminated by biogas from neighboring tanks. Therefore, in addition to isolating the tank, biogas samples were collected shortly after the tank had begun to be mixed, in the hopes that entrained biogas would be released during mixing. Each of the tanks was mixed in series, and rapid gas pressure increases had been seen repeatedly during mixing, indicating that entrained gas was being released from the tank. Biogas samples collected from each individual tank on the same day, varied significantly in terms of hydrogen and methane content, indicating that the collection scheme was capturing differences between the different tanks. Separate samples were only collected for a period of about three months, after which only composite samples were collected.

Biogas samples were collected by allowing the gas to flush through a glass sampling tube for at least thirty seconds prior to sealing the tube with a pair of valves on opposite ends of the tube. A rubber septum was fitted to a sampling port midway along the sampling tube. This septum was changed regularly to avoid leakage. The gas sampling methodology was tested extensively at the beginning of the operational period to determine the best methods for ensuring adequate data quality. Flushing times were

87

compared. Time between sampling and measuring biogas content was evaluated, and it was determined that measuring biogas content within several hours of sample collection did not significantly affect the outcome.

The biogas composition (H2, CH4 and CO2) of each reactor was measured daily using a gas chromatograph (GC) (Agilent® GC6890 N) equipped with a thermal conductivity detector (TCD) and packed column (Alltech® C‐9000) of length 3.05 m, outside diameter 3.18 mm, internal diameter 2.16 mm and with 80/100 mesh carbosphere. The carrier gas was argon at a flow rate of 30.8 mL min‐1. The injector, oven and detector temperatures were 120, 100 and 120 °C, respectively.

Biogas flow rates were recorded by the automated data collection system, and data were downloaded weekly from the primary server. Biogas flow meter data points were retrieved in six‐second intervals, subject to the condition that the flow meter reading was above 0.1 cubic foot per min to filter out flow meter noise. If the time interval between consecutive data points was greater than one minute, it was assumed that the gas flow control valve was shut during this interval, and the total gas volume produced was assumed to be zero. For all other intervals, the average biogas production during that time interval was calculated using a trapezoidal integration algorithm as follows:

Where V = volume of biogas produced during the interval (scf), Rt = biogas flow rate (scf/min) at the beginning of the interval, Rt+1 = biogas flow rate (scf/min) at the end of the interval, and ∆t = interval length (min). Daily biogas production totals were calculated by summing the individual intervals for that day.

Methane and hydrogen contents were measured three times per week. The methane and hydrogen content of the biogas between measurements was interpolated using a linear interpolation method to calculate running methane and hydrogen production totals.

3.3.2 Steady-State Biological Stability and Performance Hydrogen Production Potential

One distinct advantage of the APS Digester system is the ability to capture hydrogen‐rich biogas separately from methane‐rich biogas. The system was designed with separate biogas collection lines specifically for this purpose. However, during this study, biogas streams were not collected separately. Due to the design of the APS Digester system and the need for pressure balance between the different reactors, the headspaces of the three reactors were connected via a contiguous two‐inch gas pipe that transported the gas through a hydrogen sulfide scrubber, past a thermal mass flow meter, and ultimately to a flare. Because the headspaces were connected, the flow rate of gas from the individual tanks was not measured. Therefore, the overall hydrogen yield could not be calculated. However, an attempt was made to analyze the hydrogen content of the biogas from each tank separately.

88

As shown in the figure below, the hydrogen content of the biogas in the hydrolysis tanks increased sharply within a day or two of each loading event. The hydrogen content was consistently higher in the hydrolysis tank that had just been loaded, and it was much lower in the biogasification tank. Hydrolysis usually led to large bursts of hydrogen‐rich biogas production, as evidenced by the spikes in gas production that followed each loading event. Therefore, even though the hydrogen yield could not be calculated, it may be expected to be quite high especially given the high hydrogen content of the gas samples.

Figure 56. Hydrogen Content of Biogas from the Two Hydrolysis Tanks and the Biogasification Tank

Jul '08 Aug '08 Sep '08 Oct '08 Nov '08

H2

Con

tent

(%)

0

5

10

15

20

25

30

35

40

Hydrolysis Tank 1Hydrolysis Tank 2Biogasification TankLoading Event

These bursts of biogas production also may explain why the biogasification tank appeared to have high hydrogen content. Hydrogen‐rich biogas from the hydrolysis tanks had infiltrated the biogasification tank. Later, when only composite biogas samples were collected, the biogas hydrogen content remained less than 5 percent with minor fluctuations as shown in the Figure 57..

Methane Content

The average methane content of the biogas was 54 percent during steady‐state biogas production. However, because batches were loaded once per week, methane content fluctuated. The methane content declined sharply after each loading event as carbon dioxide was produced during hydrolysis, but it consistently remained combustible

89

(above 45 percent) during steady‐state operation. More frequent loading of food waste would reduce the variability in methane content and overall biogas production.

Figure 57. Methane and Hydrogen Content of Biogas During the Second Operational Period

Dec '08 Jan '09 Feb '09 Mar '09 Apr '09 May '09

Gas

con

tent

(% b

y vo

l)

0

20

40

60

80

100

Mean Methane

Mean Hydrogen

MethaneHydrogenLoading Event

Biogas and Methane Production and Yield

Overall, the APS Digester system produced biogas, methane, and hydrogen as expected based on previous laboratory studies. Batch digestibility tests of source‐separated food waste resulted in a 30‐day methane yield of 440 m3/t VS at 50°C with an average methane content of 73 percent (Zhang et al. 2007). Laboratory tests of food waste in a bench‐scale APS Digester system also resulted in methane yields of 320 – 430 m3/t VS (unpublished data, UC Davis). Throughout the operational period in this study, the methane yield reached a maximum of 437 m3/t VS and averaged 244 m3/t VS, as shown in the figure below. During the steady‐state biogas production period from January through April 2009, the average methane yield was 275 m3/t VS.

90

Figure 58. Cumulative Volatile Solids Loading (t) Overlaid with Biogas and Methane Yields (m3/t VS) Over the Duration of the Testing Period

May '08 Jul '08 Sep '08 Nov '08 Jan '09 Mar '09 May '09

Cum

ulat

ive

Load

ing

(t V

S)

0

5

10

15

20

25

30G

as y

ield

(m3 /t

VS)

0

100

200

300

400

500

600

700

Mean steady-state CH4 yield

Cumulative Volatile Solids LoadingMethane yieldBiogas yield

During the first run (June – November 2008), methane yield increased quickly, but acid accumulated in the biogasification tank caused a sharp decline in methane yield before steady‐state was attained. The acid accumulation was caused by an overly rapid increase in the organic loading rate. As seen in the figure below, food waste was loaded bi‐monthly between June and mid‐August, 2008 at an organic loading rate equivalent to 1.0 kg VS/m3/d. An equivalent amount of food waste was then loaded weekly, which raised the organic loading rate to 2.0 kg VS/m3/d within several weeks, as shown in the Figure 59..

Figure 59. Running Organic Loading Rate

Jun '08 Aug '08 Oct '08 Dec '08 Feb '09 Apr '09

Org

anic

load

ing

rate

(kg

VS/

m3 /

d)

0.0

0.5

1.0

1.5

2.0

2.5

91

The biogas and methane yields quickly declined, but although the pH dropped from 8.2 to 7.2 at the same time, it hovered at close to 7 before dropping precipitously on September 30, 2009, as seen in the figure below. In hindsight, the initial decline in pH and steady decline in methane yield foreshadowed the abrupt biological failure, but it was not recognized at the time due at least in part to the high buffering capacity of the biogasification tank.

Figure 60. pH Levels in the Hydrolysis and Biogasification Tanks

May '08 Jul '08 Sep '08 Nov '08 Jan '09 Mar '09 May '09

pH

3

4

5

6

7

8

9

Biogasification TankHydrolysis Tank 1Hydrolysis Tank 2

The organic loading rate was raised more gradually for the second run (Nov 2008 to Apr 2009). Food waste was loaded weekly from the beginning, and the organic loading rate was raised from 1.0 to 2.0 kg/m3/d over the course of four months. This resulted in the development of steady‐state gas yield within six weeks of re‐starting loading, and the steady‐state period lasted over 3 months. During this period, the methane yield ranged from 220 – 330 m3/t VS. The fluctuation seen in the methane yield was due partially to the batch loading of food waste which caused the organic loading rate to fluctuate. Eventually, unsustainably high ammonia concentrations in the biogasification reactor forced plant operators to cease loading. The high ammonia concentration resulted from the high protein content of food waste (20‐25 percent on dry matter basis). The high ammonia concentration may have also been inhibiting the microorganisms, resulting in a slightly lower methane yield than expected for food waste .

Effluent Quality and Solids Reduction

Biogasification and Hydrolysis pH

92

Typically, anaerobic digester operators monitor pH as an indicator of digester health. In a single‐stage anaerobic digester, the pH has to be maintained close to neutral to avoid inhibiting the methanogenic microorganisms and triggering self‐perpetuated acid accumulation. In a two‐stage digester, acids can accumulate in the first stage without affecting the methanogenic microbes housed in the second‐stage. In this study, the pH of the first‐stage hydrolysis tanks fluctuated sometimes as low as 4.5 while the second‐stage biogasification tank remained stable between pH 7 and 8. Because the APS Digester system re‐circulated microorganisms from the biogasification tank into the hydrolysis tank, the pH of the hydrolysis tanks quickly returned to neutral between batch loading events. This allowed the hydrolysis tanks to behave temporarily as biogasification tanks. If the pH were to remain more consistently low in the hydrolysis tanks, as would be expected at full‐scale with more frequent loading of fresh Food waste, the biogasification reactor may need to be increased in size.

The pH in the biogasification tank was monitored as an indicator of digester health. However, an important lesson learned in this study was that the pH can remain neutral even as the biological activity declines when the buffering capacity is high. The alkalinity of the biogasification tank was measured four times between March 25 and April 1, 2009, and it was found to range from 10,600 to 14,200 mg/L as calcium carbonate with an average of 12,300 mg/L. Because of the high alkalinity (which was supported in part by a high ammonia concentration), the pH remained stable even as the methanogens were inhibited, as evidenced by the declining methane yield in September 2008. From September 25 to September 30, the pH in the biogasification tank suddenly declined from 7.2 to 5.5, at which point loading was discontinued. Ammonia concentration was not measured at that point, but it is likely that high ammonia concentrations may have also contributed to the microbial inhibition. In May 2009, the methane yield began to decline despite relatively stable, near‐neutral pH in the biogasification tank. It was not clear whether acid or ammonia was responsible for the decline, as loading was halted due to the potential impact of high ammonia in the effluent on the adjacent wastewater treatment facility.

Volatile Solids

Effluent samples collected regularly were analyzed for solids content. The resulting analysis demonstrated that both total and volatile solids were accumulating within all three tanks, but the accumulation became more pronounced during the second run (after December 2008). The ratio of volatile to total solids remained relatively stable between 75 – 85 percent, especially during the second run. However, there was a gradual upward trend in VS/TS, indicating that the solids‐reduction efficiency may have been reduced, perhaps by the accumulation of ammonia in the system.

93

Figure 61. Mass of Volatile Solids and Volatile/Total Solids Ratio for the APS Digester System


Mas

s of V

S in

the

Tank

s (kg

)

0

1000

2000

3000

4000

5000

VS/

TS R

atio

(%)

0

20

40

60

80

100

Hydrolysis tank 1Hydrolysis tank 2Biogasification tankCombined total Total VS/TS

The reduction of volatile solids is typically calculated for a continuously stirred tank reactor as the ratio of volatile solids in the feed to that of the effluent. This calculation assumes that the contents of the reactor are well mixed, and solids do not accumulate. Solids did accumulate in the digester system; therefore, the volatile solids reduction was also calculated based on the total mass balance of VS remaining in the system. The latter was calculated by totaling the mass of volatile solids within the three tanks based on the tanks’ volatile solids concentration and volume. Because of the accumulation of volatile solids in the system, the actual volatile solids reduction was 80‐85 percent, while the apparent volatile solids reduction (effluent based) was 90‐95 percent). In practice, the effluent discharged would contain very low volatile solids concentration (3‐5 percent), but the biogasification tank would eventually have to have solids removed (for example via an auger) to avoid excessive solids accumulation in the tank. Since the accumulation occurred gradually, solids could be removed as part of an annual or bi‐annual cleaning process. The gradual increase in volatile solids reduction seen in the figure below for the internal‐VS based calculation between December 2008 and February 2009 is due to the presence of volatile solids in the tanks from the first run.

94

Figure 62. Volatile Solids Reduction


VS

Red

uctio

n (%

)

0

20

40

60

80

100

Total VS reduction (effluent VS basis)Total VS reduction (internal VS basis)

As seen in the VS reduction figure, a larger fraction of the VS reduction was occurring in the biogasification tank during the first run (June – November 2008) as compared with the second run (November 2008 – May 2009). This may help explain the higher methane and biogas yield observed during this period. However, the lack of stability in methane production is echoed here in VS reduction in the biogasification reactor. During both runs, the total volatile solids reduction was very high (approximately 80 percent). With additional operational experience, it may be possible to improve performance by altering the frequency of loading, the solids retention time, and the recirculation rate.

BOD, Ammonia, and Nutrients in the Effluent

The biological oxygen demand (BOD) of effluent from the biogasification tank was measured twice during steady‐state biogas production periods. The BOD ranged from 32,875 to 30,391 mg/L. This is relatively high, indicating the need for some post‐treatment at a full‐scale operation.

The ammonia concentration in the tanks was measured in April 2009, and was found to range from 2,160 to 2,650 mg/L in the biogasification tank and from 2,380 to 3,520 ppm in the hydrolysis tank on three different days over a one week period. Almost one year later, the ammonia concentration was measured again and was found to be in the same range.

Given an average nitrogen content in the feedstock of 3.26 percent (dry basis) and solids content of 20.5 percent (wet basis), the theoretical maximum ammonia concentration (at 100 percent conversion of nitrogen to ammonia) would be 6,680 ppm as ammonia nitrogen. Therefore, the conversion rate of nitrogen in the food waste was approximately

95

40‐50 percent. Since the volatile solids conversion rate was 80‐90 percent, there may have been a relatively large fraction of residual organic nitrogen in the effluent.

Effluent samples collected from the biogasification reactor on five separate days during the steady‐state period (3/27/09 – 4/3/09) were combined and analyzed for mineral and nutrient composition. The effluent was found to contain 27.3 and 5.09 g/L carbon and nitrogen, respectively. Thus, only half of the residual nitrogen was from ammonia, confirming the theoretical prediction that significant quantities of non‐ammonia nitrogen remained. This may indicate relatively high fertilizer value for the effluent. However, other minerals were also measured.

The crop nutrients phosphorous and potassium were present in the effluent at relatively high concentrations: 571 and 870 mg/L, respectively. At this level, without further treatment the effluent contained the equivalent of 42.4, 4.8, and 7.3 pounds of nitrogen, phosphorous, and potassium per thousand gallons of effluent. Sodium was present at a concentration of 2.5 g/L, or 20.9 lbs per thousand gallons. Although this concentration should not inhibit bacteria, it may limit use of the effluent for land application without further treatment. Calcium and magnesium concentrations in the effluent were negligible at 40.0 and 1.6 mg/L, respectively. The average solids content during this period was 52 g/L, of which 41 g/L were volatile. The combined minerals analyzed made up 36 g/L or 70 percent of the total solids. Effluent treatment to reduce BOD and concentrate nutrients while removing sodium would be recommended if applying the effluent as a liquid fertilizer.

3.4 Conclusions The operation of the APS Digester system over the course of one year provided invaluable experience that will help further the development of this technology. The system performed very well, especially considering the scale‐up factor. Mechanical problems were minimal and easily repaired. The majority of repairs were made to the boiler. Some repairs were made to the feedstock loading system. The mixing and material transfer systems worked without incident. The mechanical problems that did occur led to design improvements that will be incorporated into the system for future testing. Improvements made at this scale will translate directly to large‐scale anaerobic digestion systems.

Biologically, the APS Digester system mirrored lab‐scale systems. Although there were some biological instabilities, food waste is typically a very difficult substrate to digest due to the speed of hydrolysis and high nitrogen content. Despite the instabilities, the system converted 80‐90 percent of the volatile solids in the food waste, yielding high levels of hydrogen and methane gas for sustained lengths of time. Furthermore, by observing the system’s behavior during periods of instability, future plant operators will be able to predict when the system is unstable, which will allow for appropriate corrections to be made. The opportunity to re‐inoculate the digester system and change the startup procedure provided an opportunity to test hypotheses and make adjustments that reduced the startup time and improved system stability. Through this

96

experience, future food waste digesters can be designed to avoid limitations imposed by high nitrogen content and loading rate.

The APS Digester system proved to be an effective and flexible solid‐waste anaerobic digestion system. The pilot plant at UC Davis has the potential to provide continued experience with high‐solids waste digestion, and can be used as an excellent training tool for engineers designing and operating anaerobic digesters.

97

CHAPTER 4: Engineering, Economic and Environmental Analysis of APS Digester 4.1 Review of Current Technologies Used For the Anaerobic Digestion of Municipal Solid Waste

4.1.1 Introduction to the Treatment of the Organic Fraction of Municipal Solid Waste Estimates for per capita generation of municipal solid waste (MSW) in the U.S. range from 752 to 1200 kg y‐1, not including construction and demolition debris, wastewater sludge, or non‐hazardous industrial waste [1, 2]. This includes 12 percent food scraps, 13 percent yard trimmings, and 34 percent paper and paperboard. Nearly 225 Mt y‐1 of MSW is landfilled and 30 Mt y‐1 is burned for energy in the U.S. [1, 2]. Composting operations recovered 62 percent of the yard trimmings and 2 percent of the food scraps, while 50 percent of the paper and paperboard were recycled [1]. Currently in the U.S., several landfill bioreactors are used to enhance the biodegradation and biogasification of MSW. There are no commercial scale anaerobic digestion (AD) systems processing source‐ or mechanically‐separated organic fractions of MSW (OFMSW). However, the U.S. has had extensive research and commercial experience with AD in the treatment of municipal, industrial and agricultural wastewater and the existence of several commercial facilities outside the U.S.

By contrast, there is nearly 4 Mt y‐1 of organic MSW treatment capacity in large scale commercial digesters in Europe [3]. This MSW AD treatment capacity has developed over the past 20 years largely due to policies enacted by the European Union aimed at reducing disposal of biodegradable materials [4]. Here the current and emerging AD technologies and their uses for treating OFMSW are reviewed from a technical and economic perspective.

4.1.2 Overview of Anaerobic Digestion Technologies for Treating Municipal Solid Waste

AD has been used successfully at farms and municipal wastewater treatment facilities in the U.S. for many years. Applying AD, a traditionally liquid‐treatment technology, to high solids feedstock, such as MSW, requires solving difficult material handling problems. Successful MSW digesters use extensive pre‐ and post‐digestion processing units, regardless of the waste source or digester type. Pre‐sorting (source‐separated waste is optimal) facilitates handling and material flow and reduces volume of inert materials introduced to the digester. A typical sorting line includes visual (often manual) sorting and removal of bulky items, followed by mechanical, hydraulic and/or biological mechanisms that comminute and dissolve organic components while attempting to remove inorganic components. The screened and sorted material can then be transferred to the digester.

98

In general, the optimal conditions for anaerobic digestion of organic matter are neutral pH, constant temperature, and consistent feeding rate. Imbalances among the different microorganisms can develop if conditions are not maintained near optimum. The most common result of imbalance is the build‐up of organic acids which suppress the methanogenic organisms. Acid accumulation is usually controlled naturally by inherent chemical buffers and by the methanogens as they consume organic acids to produce methane and carbon dioxide. These natural controls can breakdown if substrate is fed too quickly, if inhibitory compounds accumulate, or if the feed stream lacks alkalinity to buffer pH. Solid concentrations higher than about 40 percent total solids (TS) can also result in process inhibition. The TS content of OFMSW typically ranges from 30‐60 percent, thus some water may need to be added prior to anaerobic digestion. Process water can be used, but care must be taken to avoid the accumulation of soluble inhibitory compounds such as ammonia and salt. Thus, digesters that significantly dilute the incoming waste require addition of fresh water (or treated process water). The microorganisms involved in AD are adapted to relatively narrow temperature ranges; roughly 30‐40 °C and 50‐60 °C for mesophilic and thermophilic microorganisms respectively. Operation in the thermophilic range results in faster reaction kinetics which reduces the size of digester required but may require a heat source and associated insulation. Commercial digesters (with the exception of covered manure lagoons) do not typically operate at temperatures below the mesophilic range, although digestion in the psychrophilic temperature range is technically feasible.

Digesters can be designed to treat OFMSW in a variety of configurations. The sorted waste can be diluted and treated similarly to low‐solids wastewater in a single tank; the so‐called single‐stage wet digester. However, MSW slurry behaves differently from wastewater. Because of the heterogeneous nature of MSW, the slurry tends to separate and form a scum layer which can evade and clog pump outlets and pipes [5]. To prevent this, removal of inert solids and thorough homogenization is desirable. Solids can also short circuit to the effluent pipe before they have broken down completely; therefore design modifications have been made in some digesters to allow longer contact time between bacteria and dense, recalcitrant material [5]. Furthermore, MSW often contains compounds that are toxic to the AD microorganisms such as battery acids and heavy metals, especially when the waste is not source‐separated. In diluted slurry, these compounds diffuse throughout the reactor, and, in high enough concentrations, can inhibit the microorganisms.

High‐solids or dry digesters that do not require large quantities of water and avoid some of the difficulties associated with wet digestion have been developed. This digester type is prevalent in Europe making up 60 percent of the single‐stage digester capacity installed to date [3]. Dry digesters treat waste streams with 20‐40 percent total solids without adding dilution water [5]. However, these systems may retain some process water or add some fresh water either as liquid or in the form of steam used to heat the incoming feedstock. Heavy duty pumps, conveyors, and augers are required for handling the low‐moisture feedstock, which adds to the capital costs. The reduction in

99

pre‐treatment equipment required (such as hydrocyclones and pulpers) compensates for some of this additional cost. Most high‐solids digesters operate as plug‐flow digesters, which usually requires inoculation of the incoming OFMSW with some of the exiting digestate prior to loading [5].

Digesters can also be designed to treat waste in two or more stages. Single‐stage digesters are simpler to design, build, and operate and are generally less expensive than multiple‐stage digesters. However, the organic loading rate (OLR) of single‐stage digesters is limited by the ability of methanogenic organisms to tolerate the sudden decline in pH that results from rapid acid production during hydrolysis. Two‐stage digesters separate the initial hydrolysis and acid‐producing fermentation from methanogenesis and allow for higher loading rates and potentially higher throughput and biogas production.

Multi‐stage systems take advantage of the fact that the microorganisms involved in hydrolysis and fermentation have different optimal conditions than those involved in methanogenesis. By optimizing each stage separately, the overall biogas production and waste treatment rate can be increased [6]. Since methanogenic archaea prefer pH in the range of 7 – 8.5 while acidogenic bacteria can tolerate lower pH, organic acids produced in the first stage are diluted into the second stage at a controlled rate. Some multi‐stage systems apply a microaerophilic process for the first stage in an attempt to increase the oxidation of lignin and make more cellulose available for hydrolysis [7, 8]. Although adding oxygen to an anaerobic environment seems counterintuitive, sludge granules can shield the obligate anaerobes from oxygen poisoning and the practice has been shown to increase biogas yield in some situations [7‐10]. However, if the first stage produces hydrogen and/or methane, hazardous conditions could be created with oxygen addition. The process flexibility of multi‐stage systems increases cost and complexity; requiring careful consideration during system design and selection. In Europe, about 90 percent of the installed AD capacity is composed of single‐stage systems and about 10 percent is composed of two‐stage systems [3, 11].

The systems described above are continuous flow designs. Systems that operate in batch and sequential batch mode have reduced complexity and material handling requirements. With properly designed conveyance and porting systems, feedstock does not need to be carefully metered into a batch reactor, which eliminates the need for complex material handling equipment. Some of the first high‐solids digesters were envisioned as modified landfills [12]. This resulted in the creation of batch systems that recycled leachate in a manner similar to landfill bioreactors, though a carefully controlled batch digester can yield higher biogas production rates and lower retention times [5]. The primary disadvantage of batch digesters is uneven biogas production and lack of microbial population stability. To surmount these issues, batch systems can also be combined with multi‐stage configurations to form hybrid systems with batch loading and continuous waste treatment.

100

4.1.3 Examples of Commercial Anaerobic Digester Systems Used for Treating Municipal Solid Waste A range of digesters are commercially available globally that span the full range of design options (Table 7). The operation and performance of some of the most popular systems are discussed below.

Table 7. Summary of Commercial Anaerobic Digester Technologies with Large Scale Reference Plants. Data from Company Websites as of February, 2008 And Adapted from

Nichols [13].

Process Name No. of Plants

Capacity (103 t y‐1)

No. of Stages

Digester TS Operating Temp

1 2 <20% >20% 35°C 55°C

AAT 8 3 – 55 x x x

ArrowBio 4 90 – 180 x x x

BTA 23 1 – 150 x x x x x

Biocel 1 35 x x x

Biopercolat 1 100 x x x

Biostab 13 10 – 90 x x x

DBA‐Wabio 4 6 – 60 x x x

DRANCO 17 3 – 120 x x x

Entec 2 40 – 150 x x x

Haase 4 50 – 200 x x x x

Kompogas 38 1 – 110 x x x

Linde‐KCA/BRV 8 15 – 150 x x x x x x

Schwarting‐Uhde

3 25 – 87.6 x x x

Valorga 22 10 – 270 x x x x

Waasa 10+ 3 – 230 x x x x

Note: Includes operational or planned plants that accept any of the following: MSW, kitchen waste, food waste, yard waste, or green waste. Does not include food processing waste or wastewater. May include co-digestion with other organics such as bio-waste or sewage sludge. Pilots and demonstrations were excluded. The list is not exhaustive and system names may change as companies acquire and alter technologies.

101

Waasa

The Waasa digester is a single‐stage, wet digester and was one of the original designs used to treat MSW. A vertical pulper/hydrocyclone homogenizes diluted MSW while removing floating debris and sunken grit. Density‐fractionated OFMSW is then pumped to the pre‐chamber of a continuously stirred tank reactor to alleviate short circuiting and inoculate the incoming feedstock. The largest Waasa plant is located in Groningen, Netherlands where four 2,740 m3 tanks treat 92,000 t y‐1 of OFMSW out of an initial 250,000 t y‐1 of mixed MSW [13]. Many wet digesters and systems that incorporate OFMSW into existing wastewater treatment digesters utilize similar pre‐treatment processes [14].

Dry Anaerobic Composting (Dranco)

The Dranco digester is a dry, single‐stage thermophilic anaerobic digestion system [15]. Feed is introduced into the top of the digester and flows to the conical bottom where an auger removes the digestate. A fraction of the digestate is used to inoculate incoming feed and steam is added to increase the temperature to thermophilic range. The rest of the digestate is dewatered to produce wastewater and press cake. The press cake contains active bacteria, some ammonia, and undigested solids that are aerobically stabilized for use as compost. Source‐separated waste is preferred to maintain the quality of the compost. There is no mixing within the reactor, other than some bubbling of biogas against the downward plug‐flow of the substrate.

The Dranco digester at Brecht, Belgium was reported to have maintained a high average OLR (15 kg VS m‐3 d‐1) over a one year period [5]. The digester was treating OFMSW at 35 percent TS with a 14 d hydraulic retention time (HRT) while reducing the VS by 65 percent and producing biogas with a yield of 468 m3 t‐1 VS [11]. The waste consisted of 15 percent kitchen waste, 75 percent garden waste, and 10 percent paper. The Dranco system in Salzburg, Austria treating 80 percent kitchen waste and 20 percent garden waste achieved a biogas yield of 622 m3 t‐1 VS [11]. The OLR was not reported, but elsewhere it has been stated that the typical Dranco system is designed for 12 kg VS m‐3 d‐1 [5].

The Dranco digester in Brecht consists of two single‐stage vertical digesters. The first (Dranco I) was built for $6.1 x 106 in 1991 with a volume of 808 m3 and a design capacity of 12,000 t y‐1 [15, 16]. Improvements made over the subsequent 15 years allowed the operators to increase the loading rate to its current level of 20,000 t y‐1. In 2000, a second digester was added to the Brecht site (Dranco II). The new digester accommodates 50,000 wet t y‐1 in a 3,100 m3 tank. MSW is loaded into a high‐speed sieve‐lined drum with a 4‐hour residence time for mechanical particle‐size reduction and sorting of the organic fraction. OFMSW is mixed with digestate and steam heated to 50°C prior to loading via cement pump into the digester. Keeping the temperature slightly lower than optimal (55°C) limits production of inhibitory ammonia and reduces the amount of heat required. The digesters are fed continuously 16 hrs d‐1 during the week and 12 hrs d‐1 on the weekends.

102

Valorga

The Valorga digester is a dry, single‐stage digester that treats organic solid waste with 25‐30 percent TS [13]. Unlike other plug‐flow digesters, the Valorga design uses pressurized biogas for mixing. This eliminates the need for an inoculation loop. The vertical cylindrical digester contains a partition extending across two‐thirds of the digester’s diameter. This forces material entering at the bottom to flow around the wall before exiting [17]. According to Nichols (2004) feedstocks with less than 20 percent TS do not perform well in the Valorga system because grit particles settle and clog the biogas injection ports. The retention time is on the order of 21 days and the biogas yields are reported to be 220 ‐ 270 m3 t‐1 VS [13].

The Valorga plant at Varenne‐Jarcy, France uses a 50 m rotating drum with a 2 to 3 day retention time to provide pre‐treatment and mechanically separate the organic fraction of mixed MSW. Although the drum is not aerated, the temperature of the MSW increases indicating that biological activity helps break down the organic fraction into smaller particles which are screened out. These fines are sent to a dosing unit for storage and steam heating prior to being pumped into one of the three 4,000 m3 digesters. Source separated OFMSW and mechanically separated OFMSW are loaded into separate digesters, allowing plant operators to control the quality of the compost produced.

Kompogas

The Kompogas digester is a high-solids plug-flow design. The cylindrical reactor is oriented horizontally and contains internal rotors that assist in degassing and homogenization [13, 18]. The system is prefabricated in 15,000 or 25,000 t y‐1 sizes. The internal solids content has to be carefully maintained at 23‐28 percent for the system to flow properly; therefore some process water and digestate is mixed with incoming OFMSW which also provides inoculation [18]. Retention time is 15‐20 days under thermophilic conditions. Biogas yield data were not found in the literature; however, this is a widely used system, particularly in Switzerland.

BTA

The BTA system is one of the oldest and most successful AD systems in terms of the number of existing operational digesters [13]. The majority of the BTA digesters are large (>100,000 t y‐1) two‐stage units [19, 20]. A pulper and a hydrocyclone are used to fractionate the MSW. A solid/liquid separation unit passes leachate directly to a low‐solids methanogenesis digester, while the solid extract is mixed with process water to 25 percent TS and then pumped into a hydrolysis digester with a 4 d HRT [20]. Hydrolysis leachate is transferred into the methanogenesis digester which has a 2‐d HRT. Dewatered digestate is then either treated aerobically or disposed. Installations with a designed capacity of less than 100,000 t y‐1 often utilize the pulper as the hydrolysis tank, eliminating one step in the process.

103

Super Blue Box Recycling (SUBBOR)

The SUBBOR digester is an experimental two‐stage batch digester. Both digesters are operated in batch configuration at thermophilic conditions with landfill leachate added to the first stage to bring the solids content to 25 percent TS [21]. An interim steam treatment between the two stages is applied to more completely degrade the MSW [22]. In laboratory studies, this enhanced the biogas yield by 40 percent and resulted in fine, peat‐like mass of residual solids [21, 22].

Biopercolat

The Biopercolat process is a two‐stage process with a dry, partially aerobic first‐stage and a wet second stage [5, 15]. Process water is continually percolated through the first‐stage: a horizontal tunnel that slowly rotates to aerate the mixture and prevent clogging and channeling [5]. The leachate passes on to the second stage, an attached‐growth, plug‐flow digester operating in the mesophilic temperature range. After two to three days in the percolator, the solids are separated and transferred to an enclosed tunnel composter. The liquid fraction is transferred to the second‐stage digester, and displaced liquid is partly recirculated back through the percolator and partly aerated for disposal as wastewater. The system has been reported to reduce TS by 45 percent and produce 70‐80 m3 t‐1 wet MSW [23, 24]. More detailed performance measures are unavailable.

Biocel

The full‐scale (50,000 t y‐1) Biocel system is comprised of fourteen 720 m3 leach bed digesters, each loaded in 480 m3 batches to keep the pile short and avoid compaction [25]. Digester temperature is kept at 35‐40 °C by heating leachate which is sprayed over the pile. The digester retention time is 21 days and the post‐treatment aeration bed retention time is 7‐21 days [12, 25, 26]. Fresh MSW is sorted manually and loaded by shovel without size reduction or screening. A system of vacuums flushes oxygen out of the headspace and capture odors when loading and unloading the digester [25]. For each ton of MSW loaded, the system produces 70 kg of biogas, 120 kg of water vapor, 500 kg of compost, and 230 kg of wastewater [25]. At the pilot scale the maximum achievable OLR was 7 kg VS m‐3 d‐1 [26] which is similar to continuous digesters, but at full‐scale the average OLR was 3.6 kg VS m‐3 d‐1. The Biocel system requires 40 percent less capital investment than continuously fed digesters, but requires ten times as much space [5].

Sequential Batch Anaerobic Composting (SEBAC)

The SEBAC system consists of two or three batch, leach‐bed digesters loaded in sequence such that leachate can be transferred between digesters by sprayer [27‐30]. Roughly chopped OFMSW is placed in a batch digester. Leachate from a mature digester is sprayed onto the fresh material as an inoculant, while leachate is recycled to the top of the mature pile until methanogenesis stabilizes. The digester is then switched over to internal recirculation until methane production slows as the batch matures. In laboratory trials the SEBAC process had difficulty starting when loaded with pure food

104

waste [30]. Bulking agents were required to prevent compaction and allow leachate to drain. An early pilot study reported methane yields of 160 and 190 m3 t‐1 VS when the retention time was 21 and 42 days respectively [28]. The waste stream contained 60 percent paper and cardboard, 10 percent plastic, and 6 percent yard waste, and the authors reported that the yields represented 80 ‐ 90 percent of the ultimate methane potential.

Anaerobic Phased Solids (APS) Digester

The APS Digester system is a two‐stage hybrid system with a dry, sequentially‐loaded, batch first stage and a wet, attached‐growth second stage (methanogenic) with leachate recycling between the stages (Figure 63) [31, 32]. Leachate recirculation prevents solids from fouling the wet methanogenesis digester and because the batches are phased, the leachate contains a relatively constant concentration of organic acids.

105

Figure 1. Schematic Diagram of the APS Digester System

A demonstration APS Digester system plant was developed at the University of California, Davis in 2003‐7 with a 3‐5 t d‐1 treatment capacity of (Figure 64) The pilot plant consists of five 38 m3 tanks kept at thermophilic temperatures and mixed using liquid recirculation. The pilot facility is undergoing commissioning. In laboratory studies, the system was able to digest rice straw with a lignocellulosic content (lignin, cellulose and hemicellulose) of 85 percent while reducing 40 – 60 percent of the TS and producing 400 ‐ 500 m3 biogas t‐1 VS which is comparable to yields seen for more easily degradable substrates [31]. Other substrates tested on the APS design at laboratory scale include food waste, OFMSW, food processing wastes and animal manure. Biogas yields from restaurant food waste and green waste (grass clippings) were 600 and 440 m3 t‐1 VS, respectively with a 12 d retention time [33].

LIQ

UID

RE

CIR

CU

LA

TIO

N F

RO

M B

IOG

ASI

FIC

AT

ION

RE

AC

TO

R

BU

FFE

RT

AN

K

HY

DR

OL

YSI

SR

4

BIO

GA

SIFI

CA

TIO

N

REA

CTO

R

HY

DR

OL

YSI

SR

2H

YD

RO

LY

SIS

R 3

HY

DR

OL

YSI

SR

1

Solid

s

Liq

uid

RE

SID

UA

L

MA

TE

RIA

L

DR

AIN

ED

FR

OM

R

EA

CT

OR

S

SOLI

DS

SEPA

RA

TER

LIQ

UID

FED

TO

BU

FFER

TA

NK

BIO

GA

S(C

H4+

CO

2)

FEED

BIO

GA

S(H

2+C

O2)

106

Figure 64. Pilot Demonstration Plant for the APS Digester System Technology at UC Davis, Front View (top) and Rear View (Bottom)

107

BioConverter The BioConverter digester is a single‐stage continuous system. However, in its full‐scale application, an equalization tank used for pulping and metering feed into the batch digesters may serve as a first‐stage hydrolysis digester [34]. The original pilot system consisted of eight 380 m3 attached‐growth digesters with simultaneous biogas and liquid recycling [34]. Performance and operational details were unavailable for the system, but it was one of the first full scale digesters treating municipal food waste in the U.S. It was shut down in March 1999 due to odor control problems [34].

4.1.4 Biogas and Methane Production and Digester Performance Digester performance is usually measured based on biogas and methane yields and production rates. Biogas production is reported on a volumetric basis at standard conditions (273 K, 101.325 kPa). Generally, reported biogas volumes are not adjusted for moisture content, although methane content when measured by gas chromatography assumes dry biogas. This may bias methane yields depending on the digester temperature, biogas moisture saturation level, and type of gas meter used. The biogas yields can be normalized and reported per unit of wet waste fed, TS fed, VS fed, and/or VS destroyed to compare different feed streams and reactor systems. Biodegradability varies among feed streams as the VS component is often not entirely biodegradable. Biogas yields per unit VS fed are most often reported in the literature. When comparing AD systems treating different MSW streams, compositional differences between the feed sources must be considered [35, 36] Digestion and biogas production rates are also important. Biogas production rate is often standardized per unit of digester volume per day, making this a good description of digester efficiency. The volumetric biogas production rate can be calculated as the product of the OLR and the specific biogas yield. Biogas yields of several full‐scale plants treating OFMSW ranged from 60 ‐ 180 m3 t‐1 wet feed and 250 – 710 m3 t‐1 VS Table 8). Methane yields ranged from 140 – 400 m3 t‐1 VS, but the majority of the systems yielded 200 ‐ 300 m3 t‐1 VS. The methane content of biogas ranged from 52 – 65 percent.

108

Table 8. Published biogas and Methane Yields for Full-Scale Digesters Treating OFMSW from Various Sources

Technology Location Feed Type

Biogas Yield

CH4 Yield (m3 t-1 VS) Ref

(m3 t-1 wet)

(m3 t-1 VS)

Valorga France Mixed MS, paper 145 383 204 [37]

Netherlands SS-OFMSW (Kitchen and garden waste)

92 428 226

Germany 126 530 277

Valorga Italy SS-OFMSW 180 714 400 [38]

Italy Grey waste, MS-OFMSW, sludge

60 255 140

Dranco Germany Grey waste 147 506 NA [11, 38] Belgium Garden, kitchen,

and paper waste 103 468 260

Austria Kitchen and garden waste

135 622 NA

BTA (wet process)

Germany OFMSW 92 511 332 [39]

Kompogas Switzerland SS-OFMSW 90 350 NA [40, 41]

BTA (wet process)

Germany OFMSW, manure, and food waste

76 405 255 [42]

In an extensive review of reported biogas yields and OLR from laboratory, pilot, and full scale MSW digesters, Hartmann and Ahring [43] found that source‐separated MSW yielded more biogas than mechanically separated MSW. Using their data, biogas production rates were also evaluated (Figure 65)). As expected, thermophilic digesters had higher biogas production rates than mesophilic digesters. It also appeared that dry digesters had higher biogas production rates than wet digesters, although this trend was more pronounced in the thermophilic temperature range. Most of the digesters studied exhibited biogas production rates in the range of 2 – 4 m3 m‐3 d‐1.

109

Figure 65. Laboratory, Pilot and Full-Scale Anaerobic Digester Biogas Production Rate When Treating a Variety of Feedstocks [43]

0.00

1.00

2.00

3.00

4.00

5.00

6.00

OFM

SW

SS-O

FMSW

MS-

OFM

SW

SS-O

FMSW

Food

Was

te

MS-

OFM

SW

MS-

OFM

SW/p

aper

OFM

SW

Kitc

hen

Was

te/P

aper

OFM

SW

Yard

+ F

ruit

Was

te

MS-

OFM

SW

MS-

OFM

SW

SS-O

FMSW

Dry Semi-dry Wet Dry Semi-dry

Wet

Mesophilic Thermophilic

Bio

gas P

rodu

ctio

n Rat

e (m

3 m

-3 d-1

)

Another review found that for single‐stage digesters, HRT ranged from 9 – 30 days, with the average for thermophilic reactors being 66 percent that of mesophilic digesters (10‐16 vs. 15‐25 days) [44]. No clear difference was found between wet and dry digesters in terms of HRT, but the obtainable OLR was about three times higher for dry digesters than wet digesters and two times higher for thermophilic digesters than for mesophilic digesters. The achievable OLR for source separated OFMSW (SS‐OFMSW) was about 65 percent that of mechanically sorted OFMSW (MS‐OFMSW), presumably due to the higher digestibility of SS‐OFMSW. One German full‐scale wet digester was reported to reach OLR comparable to those of dry systems while maintaining an average biogas productivity of 4.5 m3 m‐3 d‐1 [39]. The OFMSW treatment capacity of a full‐scale, single‐stage wet digester was raised from 7,200 t y‐1 to 12,000 t y‐1 by increasing the OLR from 8 to 15 kg COD m‐3 d‐1 stepwise over two months. This suggests that existing wastewater sludge digesters may be able to accommodate a significant amount of OFMSW.

The overall mass balance for anaerobic digestion can be difficult to calculate. The mass of biogas can either be inferred from its volume and composition, which may not include trace elements, or it can be calculated from the mass balance. The accuracy of the mass balance calculation, however, depends on the accuracy with which the residual stream is measured. At full scale, small errors in the measurements should not contribute as much to the accuracy of the overall measurement. However, several full‐

110

scale mass balance calculations resulted in widely varying mass balance distributions (Table 9). These studies did not divulge the method used for calculating the mass balance, nor did they break down the mass of the biogas into its constituents.

Table 9. Mass Balance Calculations for Several Full-Scale Digesters in the Literature

Digester Feed Input (kt y‐1)

Biogas (%)

Wastewater (%)

Residual Solids (%)

Ref

Valorga 16 (SS‐OFMSW) 23 45 33 [38]

Valorga 34.5 (Grey, MS‐OFMSW, sludge)

17 32 50 [38]

Biocel 50 (SS‐OFMSW) 22 23 55 [25]

4.1.5 Economics of Anaerobic Digestion of MSW Several difficulties constrain the discussion of the economics of MSW AD, including lack of real cost data, dynamic market fluctuations across time and geography, and inconsistency in system boundary definition. The capital, operating costs, and revenues for most commercial anaerobic digestion projects have not been made public. Many AD systems have been built in stages and re‐designed along the way so that the overall cost may not represent the cost of the next installation. Labor, land, transportation, taxes, material, fabrication, and administrative expenses affect costs differently in different countries.

Despite these difficulties, several studies have attempted to estimate digester costs [45‐50]. Tsilemou and Panagiotakopoulos [46] extracted cost data on 16 different facilities in the literature. The capital costs included all predevelopment and construction costs. Operating costs included labor, maintenance, materials, testing, insurance, overheads, and training costs, but not the costs of transporting residuals to disposal sites. No revenues were included in the cost figures. Clarke [50] also published capital and operating costs for six European digesters. The cost data reported by Tsilemou and Clarke were converted from Euro to U.S. dollars using purchasing power parity and adjusted for inflation to 2007 U.S. dollars using the consumer price index (Figure 66).

Although separated by 10 years, the capital cost curves from the two studies were similar. Based on these cost curves, an economy of scale of about 0.5 was apparent for both studies. The operating cost curves were different for the two studies; however the earlier data from Clarke did not fit the curve well indicating that the costs compared may not have included the same items. For the data that fit the cost curve, an economy of scale of 0.6 was calculated.

111

Figure 66. Capital and Operating Cost Curves for MSW Digesters from Two Studies [46, 50], Adjusted for Inflation to 2007 U.S. Dollars

y = 1.8109x0.5581

R² = 0.9091

y = 1.6049x0.5099

R² = 0.8796

y = 297.59x-0.617

R² = 0.9376

y = 260.22x-0.176

R² = 0.5384

0

50

100

150

200

250

0

10

20

30

40

50

60

0 50 100 150 200 250 300 350 400

Ann

ual O

rdin

ary

Expe

nses

($ t-1

y-1ca

paci

ty)

Cap

ital C

ost (

106 $)

Digester Capacity (103 t y-1)

Tsilemou - Capital Cost Clarke - Capital CostTsilemou - Operating Cost Clarke - Operating Cost

Revenues for anaerobic digesters can come from energy (gas, heat, and electricity), tipping or service fees (landfill disposal offset), secondary products (compost, water, liquid fertilizer, and feedstock for downstream processes), carbon offset credits, and government incentives (renewable energy tax credits and price supports). Most digesters convert biogas to electricity, but natural gas could be a viable alternative use, either directly injected into the natural gas grid or used as a vehicle fuel as is currently being tested in Sweden, Switzerland, and Germany [51]. Biogas contains 55‐65 percent methane with the remainder being primarily CO2, water vapor, and trace amount of H2S. If the methane content is increased to over 95 percent by removing the CO2, water, and trace contaminants, the biogas can be injected into natural gas pipelines. The cost of biogas purification equipment would be partially offset by the reduction in spending on electrical generation equipment. Furthermore, natural gas markets may be more accessible than electricity markets since purified methane can be compressed for storage, allowing the energy to be supplied as needed.

Revenues from tipping fees are crucial for the financial viability of AD systems in the U.S. [52]. The average tipping fee for waste hauled to landfills in the U.S. in 2004 was $37.79 t‐1 [53]. Between 1985 and 1998 tipping fees increased at a relatively steady rate of $2 per year, but tipping fees stagnated after 1998. They were also found to vary widely by region [53]. In 2004 tipping fees ranged from $26 t‐1 in the central U.S. to $77 t‐1 in the northeastern U.S. Because AD facilities only treat OFMSW, the tipping fees charged

112

could be different from those charged by landfills and material recovery facilities. Tipping fees charged by incineration facilities have historically been slightly higher than landfill tipping fees [53]. The price of tipping fees received by AD facilities could be influenced by transportation costs, environmental restrictions and land pressures, as well as competition between facilities accepting OFMSW, especially as the sector expands.

The total amortized annual cost of waste treatment for anaerobic digestion, including revenues, was found to range from 70 – 130 $ t‐1, in 2007 U.S. dollars [46, 49, 50]. Although this is higher than the average tipping fee received at U.S. landfills, the lifecycle cost of landfilling mixed waste in the U.S. (including revenues from recycled material and energy) was found to range from 60 – 135 $ t‐1, depending on the amount of recycling, biogas collection, and transportation involved in the scenario [54]. These costs should be compared cautiously, however, since they may not include the same considerations and were performed on systems in different countries. The anaerobic digestion costs also did not include the value of pollution avoidance. An Australian study valued the environmental cost avoidance for AD at 4.3 $ t‐1 in 2007 USD [50].

4.1.6 Conclusions from the Review of Commercial Anaerobic Digestion Technologies for the Treatment of Municipal Solid Waste

Anaerobic digestion has the potential to treat a large fraction of the waste currently being landfilled in the U.S. Diversion of food scraps alone would reduce the landfill disposal rate by 10 percent in the U.S. This feedstock has been shown to be the most highly and efficiently digestible and the most easily separable from mixed MSW streams and may be the most appropriate target for the first AD treatment plants to develop in the U.S. Alternatively, OFMSW may be integrated with existing AD treatment systems. In Italy, two wastewater treatment plants incorporate sorted OFMSW into pre‐existing digesters, the additional organic material increased the OLR by 20 percent (from 1.0 to 1.2 kg VS m‐3 d‐1) while increasing the biogas production rate by 60 percent (from 600 to 950 m3 d‐1). The other plant digested 60 percent sludge and 40 percent OFMSW (VS basis) which augmented the biogas production rate five‐fold. In the U.S., integration of MSW treatment with wastewater treatment may provide a first step toward full‐scale AD of MSW.

No single commercial AD system stands out as superior to others though single stage designs comprise 90 percent of installed capacity in Europe. Realistic digester performance should fall within the range seen in the literature: roughly 300 – 600 m3 t‐1 VS for biogas yield, 200 – 300 m3 t‐1 VS for methane yield, and 2 – 4 m3 m‐3 d‐1 for biogas production rate. The actual performance of a digester will depend heavily on the composition of the OFMSW stream. The most common problems faced by digester operators are related to material handling. Material handling equipment for a solid waste digester is more complex than that used in wastewater digesters. Source separating the OFMSW alleviates some of the material handling requirements and can

113

improve digestate quality as well as biogas production, but often source separated waste contains contaminants that need to be removed prior to digestion.

The existing cost curves indicate relatively large economies of scale for AD, which explains the increasing sizes of newly installed plants throughout Europe [3]. The overall cost of AD in Europe has tended to exceed the tipping fees received at U.S. landfills, which may be one reason decision makers have not yet applied AD to MSW in the U.S. However, a lifecycle assessment of the true cost of anaerobically digesting MSW, including emissions reductions and landfill cost avoidance, has not been performed for the U.S. The data for landfills in the U.S. and AD systems in Europe indicated that such an analysis may result in a favorable benefit –cost ratio for AD of MSW. Therefore, the true cost of commercial AD technologies, especially as compared with other management strategies, merits further study.

4.2 Background Information Used For Modeling the Mass and Energy Balance and Financial Performance of the APS Digester System

4.2.1 Description of the APS Digester System Modeling Project UC Davis researchers recently developed and patented a two‐stage anaerobic digestion system known as the Anaerobic Phased Solids (APS) Digester which is capable of degrading organic matter with low moisture content [55‐57]. The APS Digester system was designed to collect the biogas that is produced during the degradation process for use as an energy source, and to stabilize the digested solids and effluent liquid for use as an agricultural or horticultural nutrient source. A pilot plant based on the APS Digester design was developed on the campus of UC Davis with funding support from California Energy Commission, California Integrated Waste Management Board, UC Davis, Onsite Power Systems, Inc. and other companies. It began operating in 2007. The completion of this plant provided the economic data for performing a financial analysis. However, questions remained as to how to best use the biogas produced by the digester system. The aim of the current research was to develop a computer model for evaluating the performance of the pilot digester system and analyzing the financial implications of using the biogas in various capacities. Questions that needed to be addressed were:

• How much of the energy in the biogas produced would be needed to run the digester system?

• How much biogas, residual solids, and water would be produced by the digester system?

• How sensitive was the financial performance to changes in the physical and financial assumptions?

Norcal Waste Systems, Inc., a waste collection agency in The Bay Area in Northern California processing OFMSW at a composting facility in Dixon, CA was interested in applying the APS technology to their waste stream. The system was also being

114

evaluated for full‐scale use with several other feedstocks including UC Davis campus waste.

Specifications of the pilot-scale APS Digester system built at the University of California at Davis

The APS Digester system consists of five anaerobic reactors. Four of the five reactors (hydrolysis reactors) are used to mainly hydrolyze and convert the organic solids into organic acids with concurrent production of biogas containing hydrogen, methane and carbon dioxide. The fifth reactor (biogasification reactor) is used to convert organic acids collected from the four hydrolysis reactors into biogas containing methane and carbon dioxide. The pilot plant developed at UC Davis used five 38m3 vertical insulated steel cylindrical reactors. The hydrolysis reactors were spaced about 1.5 m apart and suspended above ground on 1 m high steel legs. Each reactor had a hot water jacket for heating the contents and measured 3.3 m in diameter and 4.6 m in height. The biogasification reactor was set on a cement foundation and was heated via an external heat exchanger powered by a boiler which could be fired with either natural gas or biogas. The temperature of the heat exchanger was kept at or below 80°C to avoid killing the microbes. The same pump that circulated liquid through the heat exchanger also powered the jets used to mix the reactor contents. Therefore, the biogasification reactor could be mixed without heating the fluid but could not be heated without also mixing the contents. The liquid from the hydrolysis reactors was pumped to a buffer reactor and then into the biogasification reactor. Liquid was then circulated directly from the biogasification reactor back into the hydrolysis reactors. All the reactors were insulated with polyurethane foam. The hydrolysis reactors were fed with solid waste in 12‐day batches. Feeding was staggered such that a different reactor was loaded with a new batch every three days. Process water was used to maintain the working volume at 85 percent of the total reactor volume, leaving 15 percent for biogas in the headspace.

The pilot plant had a throughput capacity of approximately 3 t d‐1 (assuming an average solid waste density of 240 kg m‐3). The actual throughput rate would depend on the type of material being loaded. The plant was designed to operate at 55‐60 °C but could be set to lower temperatures as well. Processing controls were automated with computerized feedback sensors [58].

4.2.2 Objectives of the APS Digester System Modeling Project The overall aim of this research was to model and evaluate the financial feasibility and mass and energy balance of the APS Digester system under a variety of process configurations and economic parameters.

Research Objectives

• Create a computer model that integrates mass and energy balance computations with financial parameters to calculate the engineering design needs, environmental impact, and overall financial performance of the APS Digester system in terms of net present worth (NPW), internal rate of return (IRR),

115

levelized annual cost of energy (COE), and levelized annual cost of waste treatment (COT).

• Populate the model with pilot and lab‐scale APS Digester system data to compare the results with published studies on similar systems and test the sensitivity of the model to changes in the underlying assumptions.

• Compare the financial and engineering performance of a system that converts biogas to electricity with one that converts biogas to commercial‐grade natural gas.

4.2.3 Comparison with and Evaluation of Previous Models of the APS Digester System Two investigators have reported on the economic performance of a theoretical APS Digester system [32, 52]. The base assumptions for both studies were compared with each other as a point of departure for the current model (Table 10). Hartman [32] used a cash flow model to compare projects based on NPW, IRR, cost/benefit ratio, and uniform cash flows. His analysis was based on the planned pilot scale APS Digester system to be built at the University of California at Davis in which the biogas would be used for combined heat and electricity generation. He found the project to be economically viable and reported that present worth could more than double if high‐end assumptions were made on operating conditions. The financial viability was highly sensitive to changes in the tipping fee charged for treating the waste. He also assumed no value for the residual solids, thus he concluded that increasing compost sales could greatly increase the present worth. He did not explicitly calculate the parasitic load, but instead used a lumped operating and maintenance cost that was assumed to include the cost of the parasitic electrical demand.

Williams [52] based his analysis on plans for a commercial scale APS Digester system at CSU Channel Islands with nine times as much feedstock capacity in which all the biogas would be converted to electricity but some of the electricity was consumed by the system. His model was a revenue requirements model which assumed fluctuating annual revenues equal to the annual expenses. This simplified the calculation of the levelized annual COE required to achieve zero NPW. His model assumed that 80 percent of the capital was paid for with debt at an interest rate of 5 percent but equity required a 15 percent minimum attractive rate of return. It then amortized revenues at a weighted average of these two rates.

The APS Digester system and other municipal solid waste digesters had a wider range of levelized COE compared with other biomass energy plants. The COE produced by the APS Digester system was higher than that of most other waste‐to‐energy conversion technologies, but the COE was highly sensitive to the tipping fees, capital cost, and operation and maintenance (OM) costs. The report postulated that this was because APS

116

Digester system profitability depended primarily on non‐energy income. He also noted that OM costs might have been unrealistically high in his calculations.

Based on these studies it was determined that a precise estimation of the capital and operating costs was required. A cash‐flow model should be used to provide flexibility, but the model should be able to calculate the COE as well as NPW and IRR. The debt ratio, interest rate, and minimum attractive rate of return should be chosen carefully.

Table 10. Assumptions Used in the Literature for Modeling the Economic Performance of an APS Digester System

Hartman Williams

Performance Characteristics

Capacity (t d‐1) 28.2 125

Biogas Use Co‐generation Electricity only

Economic Life (yrs) 20 20

Reactor Volume (m3) 1,893 NA

Methane Yield (m3 kgVS‐1) 0.240 0.577

Biodegradability (% VS reduction) 85 NA

Generator Efficiency (% HHV) 32 34

Gross Electricity Potential (kW) 275 1564

Parasitic Load (% of gross) NA 20

Net Conversion Efficiency (kW t‐1) 0.0357 0.0226

Costs

Total Capital Cost ($) 2,207,208 16,908,278

Specific Capital Cost ($ t ‐1 capacity) 286.7 244.5

Generator Capital Cost (% of total) 19 9.8

Digester O&M Cost (% of digester capital) 4 3

Specific O&M Cost ($ t‐1) 13.5 18.9

Revenues

Tipping Fee ($ t‐1) 13.25 18

Sale Price of Electricity ($ kWh‐1) 0.076 NA

117

Hartman Williams

Sale Price of Heat ($ kWh‐1) 0.024 NA

Sale Price of Residual Solids ($ t‐1) 0 13.20

Financial Parameters

Debt Ratio (%) 100 80

Tax Rate (%)a 35 40.34

Interest Rate on Debt (%) 5 6

Inflation Rate (%) Not specified 2.1

Minimum Attractive Rate of Return (%) 5 7.8

Outcomes

Required Levelized Cost of Electricity ($ kWh‐1) 0.035 0.080

a Williams included state and federal taxes and a 0.009$/kWh tax credit, whereas Hartman included only federal taxes.

4.2.4 Review of Technologies Used to Convert Biogas to Heat, Electricity, and Natural Gas Biogas consists of primarily methane and carbon dioxide in ratios from 1:1 up to 3:1. The energy content of the biogas depends on the methane content, but biogas will easily ignite even at a 1:1 ratio of methane to carbon dioxide. The combustion heat can be either used directly or converted to electricity. Alternatively, biogas can be refined to form a variety of products, including natural gas, methanol, hydrogen, and gasoline [59, 60]. Most anaerobic digesters employ electrical conversion, and although some systems use microturbine generators or fuel cells, the majority use internal combustion (IC) engine generators.

Combined Heat and Power Generation from Biogas

The electrical conversion efficiency of IC engine generators was found to range from 30‐40% (HHV) which was slightly higher than the average efficiency for microturbine generators [61, 62]. Microturbines had slightly higher heat recovery efficiencies (70 percent vs. 55 percent) and higher quality heat than IC engines [61]. They also tended to be smaller, lighter, and quieter, but the technology was relatively young. Stirling engines were another emerging technology with the potential to offer advantages over reciprocating Otto‐cycle IC engines. They have shown electrical conversion efficiencies of 40‐50 percent with reduced maintenance requirements and lower emissions [61].

118

Stirling engines also accept a wide variety of fuels. However, they cost about twice as much as reciprocating IC engines, although they became more cost competitive at capacities of less than 20kW. The installed capital cost of IC engine generators with heat capture systems was well studied and cost curves were available for use in this study (Figure 67 [2].

Figure 67. Reported Electrical Conversion Efficiency and Capital Cost Curves for Reciprocating Engine Cogeneration Systems (Adapted from Firestone [62])

3031323334353637383940

500

700

900

1100

1300

1500

1700

0 1000 2000 3000 4000 5000 6000

Con

vers

ion

Effic

ienc

y (%

)

Cap

ital C

ost (

$ kW

-1) 2

007

Maximum Capacity (kW)

Capital Cost Conversion Efficiency

Anaerobic digestion has been used primarily as a waste water treatment technology. Anaerobic digesters have been sized to handle the waste that an individual farm or industrial operation needs to treat. In such cases, the electricity produced from biogas was only used on site, and often the electrical demand was too low to consume all the electricity produced. As a result, much of the biogas was flared. In California, biogas producers have experienced difficulty acquiring purchasing agreements from the utility industry [63]. Because electricity cannot be stored as easily as natural gas, electricity prices have been more volatile than those of natural gas [64]. Given these considerations, it may be more feasible to produce and sell natural gas than electricity. The natural gas pipeline may be more accessible than the public electricity grid, and natural gas can be supplied as demand fluctuates. However, adding bio‐based natural gas to the public pipeline has not been a standard practice and may face technical and administrative obstacles as well.

Processing Needed for Using Biogas as Natural Gas

It has not been investigated whether, given the additional expenses involved in purifying methane, an anaerobic digester that converts biogas to natural gas would be economically feasible. In order to determine whether the benefits of upgrading biogas to natural gas justify the costs, these costs and benefits must be accurately measured.

119

Petroleum based natural gas was found to have stringent requirements in terms of composition (Table 11). In order to sell and use biogas as natural gas the contaminants in the biogas must be removed. Methane contains 890 kJ mol‐1 of energy, according to the higher‐heating value (HHV) for pure methane reported by the National Institute for Science and Technology [65]. The required energy content of natural gas was within 2 percent of the energy content of methane; therefore almost all carbon dioxide would have to be removed to upgrade biogas to pipeline‐grade natural gas. Furthermore, biogas could contain up to 3,000 ppm hydrogen sulfide depending on the sulfur content of the feedstock. Natural gas required a hydrogen sulfide concentration of less than 4 ppm and less than 2 percent of the saturated MC at 25°C. Therefore, hydrogen sulfide would have to be removed and the biogas would have to be dried. Furthermore, hydrogen sulfide in the biogas will corrode the storage system if not removed. The nitrogen and oxygen content of biogas from anaerobic digesters is typically close to zero as long as air does not leak into the biogas handling system.

Table 11. Standard Composition of Natural Gas as Compared with Untreated Biogas. Adapted from Krich Et al [64]

PG&E Standards for Pipeline Natural Gas

Raw Biogas

Energy Content (kJ m-3)

> 36,000 20,000 – 30,000

H2S Content (ppm) < 4 50 – 3,000

MC (g m-3) < 0.1 3 – 15

CO2 + N2 Content (%)

< 1 35 – 45

O2 Content (%) < 0.1 0 – 0.2

Many methods are available for removing the carbon dioxide, hydrogen sulfide, water vapor, and gaseous impurities from biogas, such as water scrubbing, chemical scrubbing, membrane separation, biofiltration, and carbon molecular sieving [64, 66]. Water stripping was the most commonly used commercial technology, followed by chemical scrubbing with polyethylene glycol and pressure swing adsorption (PSA) which employs a molecular sieve [66].

The petroleum industry commonly uses PSA technology to remove impurities from raw fossil‐based natural gas. The technology can simultaneously remove carbon dioxide and low levels of other impurities including hydrogen sulfide [67]. The technology differentially separates gas molecules based on size by pressurizing the biogas in the presence of a molecular sieve. Pores in the sieve exclude methane but allow other

120

components such as hydrogen sulfide, carbon dioxide, carbon monoxide and nitrogen to be adsorbed. A valve allows purified methane to leave the unit while retaining the contaminants. The valve then switches to a different outflow and the pressure is decreased, allowing the adsorption particles to release the contaminants. The rate at which pressure is cycled and the pressure difference determines the size of the unit. Many commercial units use multiple compressors for operating the PSA unit and pressurizing the purified biogas into its final storage state.

Several landfills were found to use the PSA system to convert landfill gas to natural gas [67]. Landfill gas tends to have a relatively high O2 content due to leakage, and may be contaminated with siloxanes, volatile organic compounds, and halocarbons due to the variety of materials deposited into landfills. Oxygen molecules are about the same size as methane molecules, making it difficult to separate the oxygen. These contaminants are not typically present in the biogas produced by anaerobic digesters; therefore the PSA unit should produce natural gas more efficiently from the biogas from anaerobic digesters than from landfill gas [64]. However, biogas tends to contain higher levels of hydrogen sulfide than landfill gas. Therefore, most of the hydrogen sulfide should be removed via a separate unit.

Although in‐situ techniques for removing hydrogen sulfide such as injecting air into the biogas or iron chloride into the feed stream would eliminate the need for separate unit processes, these techniques were reported to be difficult to control and to have undesirable side‐effects [64]. Stand‐alone hydrogen sulfide removal units or scrubbers could be wet or dry, with NaOH, polyethylene glycol or water being used as the medium for wet scrubbers and iron‐oxide or hydroxide for dry scrubbers [64]. Dry scrubbers were found to be the most efficient, but the medium was typically more expensive since the reactive compounds had to be bound to a medium such as steel, wood chips or carbon compounds.

After upgrading, the biogas can be compressed for storage and injected into the pipeline. Storage pressure ranges and the energy required to achieve those pressures for natural gas are listed in Table 12. Natural gas stored at high pressure is known as compressed natural gas (CNG) and is often used as a low‐emissions transportation fuel. The CNG fueling stations store the gas at 24,800 kPa (3,600 psi) which requires thick‐walled holding vessels. For transportation natural gas is often liquefied to make liquefied natural gas (LNG), but this requires energy and capital intensive cooling equipment as well as specialized cryogenic storage systems and handling procedures [64]. Therefore, LNG was not considered in the current modeling effort.

121

Table 12. Pressure and Energy Requirements for Storage of Purified Biogas. Adaptedfrom Ross, et al [59]

Storage Pressure

(kPa)

Compression Energy Requirement

(kWh m-3)

Low Pressure Storage 14 no compression needed

Medium Pressure Storage 1,380 0.2

High Pressure Storage 13,800 0.5

4.3 Development of a Model for Calculating the Mass and Energy Balance and Financial Performance of the APS Digester System A model was developed to use the biogas production and waste reduction results of a laboratory‐scale APS Digester system to design a full‐scale system. The mass balance was calculated to predict the quantities of gaseous, solid, and liquid products, including carbon emissions from the energy conversion units, as well as biogas, compost and wastewater. The energy balance was calculated to predict the net energy production for the system, including heat and electricity. The results were then entered into a financial simulator to predict the 20‐year financial performance of the system. Although several high quality software packages were available for modeling the mass and energy balance and financial performance of industrial processes, the present model was built using Microsoft Excel® spreadsheet software which allowed the model to be tailored to the APS Digester system. Due to the novelty of the APS Digester system, the spreadsheet included a cost estimator for predicting the system’s capital costs and ordinary expenses as well as mass and energy balance calculator. The model used experimental data for a specific feedstock to predict biological performance. This was combined with the heat and electricity requirements for each unit process to predict the mass and energy balances of the digester system. The capital costs of the materials and operating expenses were calculated and revenues were computed based on the prevailing market prices. These were escalated at the 10‐year average Consumer Price Index (CPI) inflation rate, making this a current dollar analysis. A current dollar analysis was used because taxes were included in the cash flows, and using a constant dollar analysis would have resulted in errors due to the accelerated depreciation of capital expenses used for calculating taxes [68, 69]. Current‐dollar annual cash flows were computed and discounted at the minimum attractive rate of return (MARR) which was set equal to the after‐tax November 2007 Federal Reserve Bank prime interest rate. The NPW was calculated from the discounted cash flows. The IRR was calculated from the undiscounted cash flows. The incremental IRR was calculated on the difference between undiscounted cash flows for each pair of scenarios possible. The energy revenue required (from electricity in the CG scenario and natural gas in the CNG and CG+CNG

122

scenarios) to generate the assumed MARR was calculated using the revenue requirements method [68, 69]. This was levelized, discounted, and standardized per kWh of energy (using the HHV for methane in the natural gas producing scenarios) to generate the COE. Similarly, the tipping fee revenue requirement was calculated, levelized, discounted, and standardized per ton of OFMSW treated to generate the COT for each scenario.

Figure 68. Conceptual Flow Chart for the Design of an Integrated Mass/Energy Balance and Financial Calculator for the Anaerobic Phased Solids Digester

4.3.1 Calculation of the Mass and Energy Balance and Financial Performance of the APS Digester System The model included the following components: Frontend Inputs:

• Performance Parameters

• Financial Parameters

• Income and Expenses (cost calculator and prices)

• Internal Calculations

• Digester Performance

• Financial Parameters

• Cost Calculator

• Capital Costs

123

• Ordinary expenses

• Revenues

• Outputs

• Mass Balance

• Energy Balance

• Cash Flow

• Charts and Graphs

• Sensitivity Analysis

The model accepted the total daily wet feedstock throughput as an input and used MC, VS content as a fraction of the TS content (VS:TS), biodegradability, and biogas yield potential to calculate the mass balance of residual solids and biogas produced. Description of the APS Digester System Configuration Scenarios Modeled

The purpose of this study was to maintain a consistent size of digester and quality of feedstock while comparing a variety of additional unit processes, such as a boiler for heat recovery, PSA unit for upgrading biogas to CNG, electrical generator with heat recovery system for combined heat and power production, and a combination of these (Table 13). The first configuration modeled assumed that all the biogas produced would be burned in a reciprocating engine generator with heat recovery. In the second scenario modeled, the engine generator was substituted with a PSA unit to make a CNG substitute, and since no heat was captured from the PSA unit a boiler was included to provide the electricity and heat required for operating the digester system and energy conversion units and a PSA unit for upgrading the remaining biogas to CNG.

All the scenarios included the capital cost of a boiler, since it would be needed for startup of the digester system and also for back‐up purposes. It was assumed for this study that unused recovered heat had no value. However, the model was designed with the ability to include heat sales.

124

Table 13. Scenarios Evaluated and Unit Processes Compared Using the Financial and Mass/Energy Balance Model

Engine Generator with Heat Recovery

Pressure Swing Adsorption and Compressor

Boiler with Heat Exchanger

CG x x

CNG x x

CG + CNG x x x

The digester was sized for the working volume and head space required. The working volume was determined for waste with a constant TS and VS throughput rate loaded at the maximum sustainable organic loading rate as follows:

OLRm

V VSW

&=

(1)

where VW = working volume (m3), �VS = VS mass input rate (t d‐1), and OLR = VS organic loading rate (t m‐3 d‐1).

The total reactor volume required was calculated from the working volume adjusted by a headspace factor. The biogasification‐to‐hydrolysis volume ratio was used to determine how much of the total reactor volume would be divided between the total number of biogasification and hydrolysis tanks, and the reactor volume required for each biogasification and hydrolysis reactor was calculated:

HS

WR F

VV =

(2)

where VR = total reactor volume (m3), VW = total working volume (m3), and FHS = headspace factor ( percent of working volume).

( )BRBHHR

RHR NFN

VV+

= (3)

HRBHBR VFV = (4)

where VHR = individual hydrolysis reactor volume (m3), VR = total reactor volume (m3), NHR = number of hydrolysis reactors, FBH = biogasification‐to‐hydrolysis volume ratio, NBR = number of biogasification reactors, and VBR = individual biogasification reactor volume (m3).

The feedstock was diluted with enough process water to reach the required initial MC. The MC was defined on a wet mass basis as follows:

125

pwfeed

pwfw

mmmm

MC&&

&&

+

+=0 (5)

where MC0 = initial MC of diluted feedstock ( percent wet basis), �fw = mass flow rate of water in the undiluted feedstock (t d‐1), �pw = mass flow rate of recycled process water for feedstock dilution (t d‐1), and �feed = wet feedstock input rate (t d‐1).

The mass flow rate of water in the feedstock was expressed in terms of the MC of the feedstock:

feed

fwf m

mMC

&

&= (6)

where MCf = feedstock MC ( percent wet basis), �fw = mass flow rate of water in the undiluted feedstock (t d‐1), �feed = wet feedstock input rate (t d‐1).

Combining equations (5) and (6) allowed for the calculation of the mass of process water required to achieve the desired diluted feedstock MC:

0

0

1 MCMCMC

mm ffeedpw −

−= &&

(7)

where �pw = mass flow rate of process water added (t d‐1), MC0 = initial MC of diluted feedstock ( percent wet basis), and MCf = feedstock MC ( percent wet basis).

The volume of process water added was calculated as the product of equation (7) and the specific volume of water. Note that the volume of water added per unit of feedstock mass treated would thus be calculated as:

( )0TS

MCvv w

pwΔ

= (8)

where vpw = volume of (process) water added per unit mass of feedstock treated (m3 t‐1), vw = specific volume of water (m3 t‐1), ΔMC = difference in MC between diluted and undiluted feedstock ( percent wet basis), TS0 = TS content of the diluted feedstock ( percent wet basis).

This was added to the volumetric throughput of the feedstock which was calculated based on its true specific volume to find the overall flow rate as follows:

feedfeedwpw vmvmV &&& += (9)

where V& = volumetric flow rate through the digester (m3 d‐1), �pw = mass flow rate of process water added (t d‐1), vw = specific volume of water (m3 t‐1), �feed = wet feedstock input rate (t d‐1), and vfeed = specific volume of the feedstock (m3 t‐1).

The waste was assumed to be completely saturated and degassed within the reactor such that no pore spaces would persist. Therefore, the specific volume was used. If the

126

waste were found to contain pore spaces, which would be more likely for drier feedstocks such as green waste, the bulk density would need to be used instead of the specific volume.

The overall hydraulic retention time (HRT) was calculated as the quotient of the working volume calculated in equation (1) and the overall flow rate calculated using equation (9):

VV

HRT W&

= (10)

where HRT = hydraulic retention time (d), V& = volumetric flow rate through the digester (m3 d‐1), and Vw = working volume (m3).

Calculation of the Mass Balance of the APS Digester System

The mass flow shown in Figure 69 was used for calculating the mass balance for solids, liquids, and gases in through the APS Digester system. The mass of water entering the system was calculated based on the feedstock characteristics. It was assumed that no additional water was added and no process water was retained. Water leaving the system via the biogas was calculated based on the assumption that biogas was saturated with water vapor. The water balance was assumed to leave the digester system as liquid. The mass of solids entering and leaving the digester system were calculated based on the assumed feedstock characteristics and the biodegradability. The remaining mass balance was assumed to leave the digester as biogas. However, the biogas yield was reported on a volumetric basis. The mass of biogas would thus depend on the full spectrum of gaseous components, which was unknown. The primary components of biogas were assumed to include methane, carbon dioxide, water vapor, and hydrogen sulfide. These were expected to compose over 99 percent of the biogas on a molar basis; however on a mass basis, relatively heavy trace contaminants could make up a larger fraction of the biogas. Therefore, the ideal gas equation of state was used to calculate the masses of the primary biogas components. These were then summed and subtracted from the mass balance. The difference was assigned to the unknown trace biogas contaminants, such as volatile organic compounds, ammonia, and siloxanes.

127

Figure 69. Mass Flow Diagram for Calculating the Mass Balance of the Anaerobic Digester, Including All Possible Biogas Uses. Dashed Lines Indicate Biogas Flow

Mass and Composition of the Feedstock

The daily mass of feedstock added to the digester was specified by the user along with the feedstock MC, VS content, biodegradability (% percent VS destroyed), biogas yield (per unit mass of VS destroyed), methane content of biogas, and reactor temperature. A capacity factor was defined as the fraction of a year during which the digester was operated. The daily feedstock mass input rate was multiplied by the number of operational days in a year to calculate an average annual feedstock mass throughput rate. Using the assumed feedstock characteristics, the average annual throughput rates for water and solids were calculated. Therefore, the assumed feedstock characteristics were also aggregate annual values.

The feedstock was assumed to consist of water, VS, and FS. The VS content was specified per unit mass of TS, which was calculated based on the MC:

feedfeed MCTS −= 1 (11)

where TSfeed = TS content of the feedstock ( percent wet basis) and MCfeed = MC of the feedstock ( percent wet basis).

The annual masses of water, VS, and FS in the feedstock were calculated based on the user specified MC and VS content as follows:

feedfeedWf MCmm = (12)

where mWf = mass of water in the feedstock (t), mfeed = wet mass of feedstock (t), and MCfeed = MC of the feedstock ( percent wet basis)

128

feedfeedfeedVSf VSTSmm = (13)

where mVSf = mass of VS in the feedstock (t), mfeed = wet mass of feedstock (t), TSfeed = TS content of the feedstock ( percent wet basis), and VSfeed = VS content of the feedstock ( percent of TS).

( )feedfeedfeedFS VSTSmm −= 1 (14)

where mFS = mass of FS (t), mfeed = wet mass of feedstock (t), TSfeed = TS content of the feedstock ( percent wet basis), and VSfeed = VS content of the feedstock ( percent of TS).

Since the annual mass of feedstock added to the digester was a function of the capacity factor, the annual masses of water, VS, and FS were also affected by changes in the assumed capacity factor, as were any calculations based on these values.

Mass Changes Resulting from Versus Reduction Via Biodegradation After anaerobic digestion, the residual material was assumed to consist of dry solids and water. The dry mass of residual solids were calculated based on the assumption that FS were not involved in the anaerobic digestion process and that leaching of FS was negligible; therefore, the mass of FS in the residuals was assumed to equal the mass of FS in the feedstock. The dry mass of undigested VS was calculated based on the user specified biodegradability as follows:

( )Bmm VSfVSR −= 1 (15)

where mVSR = dry mass residual VS (t) and B = biodegradability ( percent VS reduction).

The sum of the mass of FS and undigested VS was assumed to equal the dry mass of residual TS. The degradation of VS was also used to calculate the volume of biogas produced.

The product of the biodegradability and biogas yield potential was used to calculate the volume of biogas produced:

pVSfTstd BYmV = (16)

where VTstd = volume of biogas produced (m3), Yp = biogas yield potential (m3 t‐1 VS destroyed), B = biodegradability ( percent of VS destroyed), mVSf = mass of VS in the feedstock (t).

In practice, biogas volumes are measured at ambient or above ambient temperatures. Biogas yields are then adjusted to the standard temperature and pressure (STP) of 273 K and 101.325kPa, but water vapor is not typically accounted for in the biogas yield measurement. Therefore, for this model the volume of biogas at STP was adjusted to the volume occupied at the reactor temperature using the ideal gas equation of state as follows:

129

⎟⎟⎠

⎞⎜⎜⎝

⎛=

std

rTstdTr T

TVV (17)

where VTr = volume of biogas (m3) produced at the reactor temperature (assuming no other state changes), VTstd = STP volume of biogas produced (m3), Tr = absolute reactor temperature (328 K), Tstd = absolute standard temperature (273 K),

The amount of biogas produced was computed using the ideal gas equation of state:

( )

r

TrstdTBG RT

VPn r = (18)

where ( )rTBGn = amount of biogas (kmol) leaving the reactor, Pstd = standard pressure

(101.325 kPa), VTr = volume of biogas (m3) produced at the reactor temperature, Tr = absolute reactor temperature (328 K), and R = gas constant (J mol‐1 K‐1).

The molar composition of the biogas was then determined. The overall molar biogas composition was adjusted for the presence of hydrogen sulfide and water vapor while maintaining the stoichiometric ratio of methane to carbon dioxide measured in the lab. Anaerobic digesters typically produce biogas with 500 – 2,000 ppm H2S (wet molar basis), depending on the feedstock. For this study a value was assumed (Table 18), but in the future the H2S content of the biogas produced will be measured. The molar concentration of water vapor was calculated based on the partial pressure of water vapor, assuming the total pressure of biogas was at standard ambient air pressure. The following equations were used to determine the amount of known gases present in the biogas:

( )OHSHCOCH

T nnnnn r

BG 2224+++= (19)

2

4

CO

CHlab n

nR =

(20)

( )r

BG

TSHSH nCn

22= (21)

( ) ( )

( )( )r

BG

r

gOHr

OH

T

std

TT n

P

Pn 2

2=

(22)

where ( )r

BG

Tn = amount of biogas produced (kmol) at the reactor temperature, nCH4 =

amount of methane produced (kmol), nCO2 = amount of carbon dioxide produced (kmol), nH2S = amount of hydrogen sulfide produced (kmol), nH2O = amount of water vapor exiting the reactor (kmol), Rlab = ratio of methane to carbon dioxide in lab‐based measurements, CH2S = hydrogen sulfide content of the biogas (mol:mol), ( )r

OH

Tn2 = amount

130

of water vapor (kmol) in biogas exiting the digester, ( )

( )r

gOH

TP2

= saturation pressure (kPa) of

water vapor at the reactor temperature, and Pstd = standard pressure (101.325 kPa).

Combining equations (19) ‐ (22), the actual amounts of methane and carbon dioxide produced were:

( ) ( )

( )

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛−−

+=

stdSH

lab

labTCH P

PC

RR

nnTr

gOHr

BG

2

24 11

(23)

( ) ( )

( )

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛−−

+=

stdSH

lab

TCO P

PC

Rnn

Tr

gOHr

BG

2

22 11

1

(24)

where nCH4 = amount of methane produced (kmol), nCO2 = amount of carbon dioxide produced (kmol), ( )r

BG

Tn = amount of biogas produced (kmol) at the reactor temperature,

Rlab = ratio of methane to carbon dioxide in lab‐based measurements, CH2S = hydrogen sulfide content of the biogas (mol:mol),

( )

( )r

gOH

TP2

= saturation pressure (kPa) of water vapor

at the reactor temperature, and Pstd = standard pressure (101.325 kPa).

The molar composition of the biogas was thus a function of temperature as determined by the saturation pressure of water vapor. This assumed that the biogas was saturated with water vapor, which was confirmed by data from the literature [70].

The mass of water vapor in the biogas was computed using the molar mass of water:

( ) ( ) 310222

−= rr TOHOH

TOH nMm (25)

where ( )rTOHm

2 = mass of water in the biogas (t) at the reactor temperature, OHM

2 = molar

mass of water (kg kmol‐1), and ( )r

OH

Tn2 = amount of water vapor (kmol) in biogas exiting the

digester.

This was used to calculate the mass of water in the residual stream based on the balance of water in the digester under the assumption that water consumed during hydrolysis was offset by other processes such that there was no significant net consumption or generation of water within the digester.

Water Mass Balance Calculations The mass of liquid residuals was calculated based on the water balance, assuming no water was generated or consumed, as follows:

( )r

OH

TWfWR mmm

2−= (26)

131

where mWR = mass of residual water (t), mWf = mass of water in the feedstock (t) as calculated in equation (12), and ( )rT

OHm2 = mass of water in the biogas (t) at the reactor

temperature, as calculated in equation (15).

The wet mass of total residuals was calculated as the sum of the mass of residual water as calculated in equation (26) and the mass of residual TS, which was the sum of FS and residual VS calculated in equations (14) and (15), respectively. This was subtracted from the wet mass of the feedstock, and the mass balance was ascribed to the biogas, such that:

WRVSRFSfeedBG mmmmm −−−= (27)

where mBG = mass of biogas produced (t), mfeed = mass of feedstock added (t), mFS = mass of FS in the feedstock (t), mVSR = mass residual VS after digestion (t), and mWR = mass of residual water (t).

The molar composition of the biogas as it left the digester at the reactor temperature was calculated using equations (21) to (24). From the volumetric biogas yield, the masses and volumes of the known biogas components were calculated and subtracted from the biogas mass balance. The difference was attributed to trace contaminants in the biogas.

Post-Digestion Mass Flow Calculations

The mass of residuals was divided into a wastewater and a recoverable/compostable fraction. The dewatering efficiency of the solid/liquid separation unit was used to determine the mass of dry TS in the compostable fraction as follows:

( )VSRFSDWTSc mmm +=υ (28)

where mTSc = dry mass of the recovered TS (t), νDW = dewatering efficiency ( percent of dry TS recovered), mFS = dry mass of FS (t), and mVSR = dry mass of residual VS (t).

This resulted in a press cake which was collected and a wastewater stream which was treated on‐site. The MC of the press cake was used to determine the wet‐weight of the compostable fraction as follows:

c

TScc MC

mm−

=1 (29)

where mc = wet mass of the compostable fraction (t), mTSc = dry mass of the recovered TS (t), and MCc = MC ( percent wet basis) of the dewatered solids.

The remaining wet mass of residuals was assumed to exit via the wastewater stream. Since the majority of the solids were recovered in the compostable fraction, the density of the wastewater stream was assumed to equal that of water (1.0 t m‐3). This was used to calculate the volume (m3) of wastewater produced.

132

The volume of biogas calculated in equation (17) was assumed to leave the digester in equilibrium with the state of the biogas in the digester headspace. As biogas cooled from the reactor temperature to ambient temperature, it was assumed that water vapor would condense, leading to a state change that included a temperature change as well as a change in amount of biogas present (Error! Reference source not found.).

Figure 70. State Change as Biogas Cools from the Reactor Temperature (Tr) to Ambient Temperature (Tamb) and Water Condenses (ncondense = Moles of

Water Condensed, nBG = Moles of Biogas)

Assuming no chemical reactions, the molar balance was used to calculate the amount of water vapor condensed as follows:

( ) ( )ambr TOH

TOHcondense nnn

22−= (30)

where ncondense = amount of water vapor condensed (kmol), ( )rTOHn

2 = amount of water vapor

(kmol) in the biogas headspace at the reactor temperature, and ( )ambT

OHn2 = amount of water

vapor remaining in the biogas (kmol) at ambient temperature.

The saturation pressure of water vapor was substituted into equation (30) as follows:

( )( )

( ) ( )( )

( )( )condenseT

BGstd

TgOHT

BGstd

TgOH

condense nnP

Pn

PP

n r

amb

r

r

−−= 22

(31)

where ( )( )rT

gOHP2

and ( )( )ambT

gOHP2

= pressure of saturated water vapor (kPa) at the reactor and

ambient temperatures, respectively, Pstd = standard pressure (101.325 kPa), ( )rTBGn =

amount of wet biogas exiting the reactor (kmol), and ncondense = amount of water vapor condensed (kmol).

This was rearranged to solve for the amount of water condensed as follows:

( )( )

( )( )

( )( )

( )r

amb

ambrT

BGTgOHstd

TgOH

TgOH

condense nPP

PPn ⎟

⎟⎠

⎞⎜⎜⎝

⎛

−

−=

2

22

(32)

State 1 Tr n(Tr)

BG Pstd

ncondense

State 2 Tamb

n(Tamb)BG = n(Tr)

BG – ncondense Pstd

133

where ncondense = amount of water vapor condensed (kmol), ( )( )rT

gOHP2

and ( )( )ambT

gOHP2

= pressure

of saturated water vapor (kPa) at the reactor and ambient temperatures, respectively, Pstd = standard atmospheric pressure (101.325 kPa), and

( )rTBGn = amount of wet biogas exiting

the reactor (kmol).

From this the amount of biogas available at ambient temperature after the water condensed was calculated as follows:

( ) ( )condense

TBG

TBG nnn ramb −= (33)

where ( )ambTBGn = amount of wet biogas (kmol) remaining at ambient temperature,

( )rTBGn =

amount of wet biogas exiting the APS Digester system (kmol), and ncondense = amount of water vapor condensed (kmol).

The mass of water condensed was calculated using the molar mass of water, and the volume of the remaining biogas at ambient temperature was computed using the ideal gas state equation:

( )

std

ambT

BGTamb P

RTnVamb

= (34)

where VTamb = volume of biogas (m3) at ambient temperature (after accounting for condensation losses), ( )ambT

BGn = amount of wet biogas (kmol) remaining at ambient temperature, R = gas constant (8.314 J mol‐1 K‐1), Tamb = ambient temperature (273 K), and Pstd = standard atmospheric pressure (101.325 kPa).

As the biogas passed though the hydrogen sulfide scrubber, the amount of hydrogen sulfide absorbed by the scrubber media was calculated based on the hydrogen sulfide removal efficiency (mol:mol basis) reported by the manufacturer.

( )SHSH

AbsorbedSH rnn

222= (35)

where ( )AbsorbedSHn

2 = amount of hydrogen sulfide absorbed by the scrubbing unit (kmol),

nH2S = amount of hydrogen sulfide produced (kmol), rH2S = hydrogen sulfide removal efficiency (mol:mol).

This was subtracted from the amount of biogas entering the scrubber to determine the amount of biogas available for energy conversion. The methane content of the biogas leaving the scrubber was calculated based on the assumption that no methane was absorbed by the scrubber, hence:

( ) ( )AbsorbedSH

TBG

CH

nnn

CambCH

2

4

4 −=

(36)

134

where CCH4 = methane content ( percent mol:mol) of biogas exiting the hydrogen sulfide scrubber, and nCH4 = amount of methane produced (kmol), ( )Absorbed

SHn2

= amount of

hydrogen sulfide absorbed by the scrubbing unit (kmol), and ( )ambTBGn = amount of wet

biogas (kmol) remaining at ambient temperature.

Similar calculations were made for each of the known gas species. Biogas was then sent to either the generator, boiler, or biogas upgrade unit depending on the scenario and the energy requirements.

As biogas passed through the engine generator, boiler, and PSA unit, the gaseous species were either physically removed or oxidized during combustion. The mass of gas leaving the engine generator and boiler unit was higher than the mass entering the units due to the addition of oxygen during combustion. Neither nitrogen nor carbon dioxide in the air used for combustion was included in the mass balance calculations, but carbon dioxide produced during combustion was included. The chemical reactions used for calculating mass balance on combustion reactions were:

OHCOOCH 2224 22 +→+ (37)

OHSOOSH 2222 2232 +→+ (38)

The reduced forms of carbon and sulfur (CH4 and H2S, respectively) were assumed to be completely oxidized during combustion following reactions equations (37) and (38). Based on this assumption the mass of the oxidized form was calculated as follows:

reduced

oxidizedreducedoxidized M

MSmm =

(39)

where moxidized = mass of the oxidized form (t), mreduced = mass of the reduced form (t), S = stoichiometric ratio of the oxidized form to the reduced form (mol:mol), Moxidized = molar mass of the oxidized form (g mol‐1), and Mreduced = molar mass of the reduced form (g mol‐1).

For the biogas upgrade unit, physical gas removal was calculated based on the removal efficiencies (molar basis) reported by the manufacturer of the PSA unit being used.

Calculation of the Energy Balance of the APS Digester System

The system’s energy requirements were calculated based on the heating and electrical power requirements of each unit process. The amount of heat and electricity produced by the system were calculated based on the amount of methane produced (from the mass balance calculations), the higher heating value of methane, and the assumed conversion and heat‐capture efficiencies of the boiler, generator, and compressors.

Heat Requirement and Production Calculations The model calculated the heat balance for the APS Digester system by assuming that heat recovered from the boiler and engine generator was used to provide the heat

135

needed to bring feed from ambient temperature to the working temperature (feed heat) and replace heat lost via conduction and convection (maintenance heat). Heat losses were calculated using an idealized lumped model which assumed that reactor contents were homogeneous, isotropic, and well mixed, the effects of radiation and biochemical heat generation were negligible, and there was no contact resistance between surfaces. Heat loss through the pipes was not included, but heat loss through the roof was included. Heat loss through the heat exchanger was included in the heat recovery factor of the boiler.

The amount of heat required to bring the feedstock from ambient temperature to the desired reactor temperature was calculated using the following equation:

( )MJ

kWhTTtCmQ ambrpf 6.31

−Δ= & (40)

where Qf = amount of heat (kWh) required to heat feed material over a given time period, Δt (d); m& = feed throughput rate (t d‐1); CP = feed heat capacity (MJ t‐1 K‐1); Tr = reactor temperature (K); Tamb = average ambient temperature (K) over the period of interest.

The amount of heat lost from the tanks due to conduction was calculated from the insulation’s thermal resistance (commonly reported by insulation manufacturers as the R‐value). When the thickness of the insulation is small relative to the diameter of the reactor, the thermal resistance can be modeled as:

κdR = (41)

where R = thermal resistance of the insulation (m2 K W‐1), d = insulation thickness (m), and κ = thermal conductivity of the insulation material (W m‐1 K‐1)

Conductive heat transfer can be modeled as:

( )wallrcond TTARQ −= −1& (42)

where condQ& = heat flow rate (W); R = thermal resistance of the insulation (m); A = reactor surface area (m2); Tr = internal reactor temperature (K); and Twall = outer reactor wall surface temperature (K).

Convective heat transfer can be modeled with a heat transfer coefficient such that:

( )wallambconv TThAQ −−=& (43)

where condQ& = heat flow rate (W); h = heat transfer coefficient for air (W m‐2 K‐1); A = reactor surface area (m2); Twall = outer reactor wall surface temperature (K); and Tamb = ambient air temperature (K).

136

Conduction through the insulation equals convection in the air around the reactor, such that:

( ) ( )ambwallwallr TThTTR −=−−1 (44)

where R = insulation thermal resistance (W m‐2 K‐1), Tr = reactor temperature (K), Twall = reactor wall’s outer surface temperature (K), h = heat transfer coefficient of air (W m‐2 K‐1), and Tamb = ambient air temperature (K).

Solving for the outer wall surface temperature:

( )hRhTTR

T ambrwall +

+=

−

−

1

1

(45)

Substituting equation (45) into equation (42), the heat loss rate was then calculated to be:

( ) ( )ambr TTAhR

hRQ −+

=−

−

1

1&

(46)

where Q& = heat loss rate (W); R = insulation thermal resistance (m2 K W‐1); h = heat transfer coefficient (W m‐2 K‐1); A = reactor surface area (m2); Tr = internal reactor temperature (K); and Tamb = ambient air temperature (K).

It was assumed that the capacity factor applied only to loading, and that the reactor temperature would be maintained even if the capacity factor were less than100 percent. Therefore, the heat loss rate was multiplied by a factor of 8.766 kWh W‐1 to compute the annual maintenance heat required for the APS Digester (kWh y‐1).

The amounts of heat recovered from direct conversion in the boiler and from the engine generator were calculated as:

DCDCDC HnQ γ= (47)

( ) CGCGCG HnQ γη−= 1 (48)

where Q = heat recovered (kWh), nDC = amount of methane diverted to the boiler (kmol), nCG = amount of methane diverted to the electricity generator (kmol), H = higher heating value of methane (kWh kmol‐1), η = electrical conversion efficiency ( percent of biogas energy content converted to electricity), γDC =boiler heat recovery factor ( percent of biogas energy content recovered as heat), and γCG = cogen heat recovery factor ( percent of waste heat recovered).

From equations (47) and (48), molar heat availability factors were defined for the boiler and generator to facilitate calculations on a mole basis as:

DCDC HF γ= (49)

137

( ) CGCG HF γη−= 1 (50)

where FDC = boiler heat availability factor (kWh kmol‐1 CH4), H = higher heating value of methane (kWh kmol‐1), γDC =boiler heat recovery factor ( percent of biogas energy content recovered as heat), FCG = cogen heat availability factor (kWh kmol‐1 CH4), η = electrical conversion efficiency ( percent of biogas energy content converted to electricity, based on the HHV of CH4 ), and γCG = cogen heat recovery factor ( percent of waste heat recovered).

Electricity Production and Requirement Calculations

The amount of electricity recovered from the engine generator was calculated as

ηHnE CGCG = (51)

where ECG = electricity generated (kWh), nCG = amount of methane diverted to the generator (kmol), H = higher heating value of methane (kWh kmol‐1), and η = electrical conversion efficiency ( percent of biogas energy content converted to electricity, based on the HHV of CH4) .

The molar electrical generation factor was defined as:

ηα H= (52)

where α = generator electricity availability factor (kWh kmol-1 CH4), H = higher heating value of methane (kWh kmol‐1), and η = electrical conversion efficiency ( percent of biogas energy content converted to electricity).

The electrical consumption of the feedstock handling, circulation, and heating systems was based on the pumps’ power ratings and durations of operation. Thus the digester’s electrical requirement was a function of the capacity factor. The control system power rating was added to this to calculate the overall digester electrical requirement (Ereq), which was assumed to be independent of the choice of biogas usage.

The electrical requirement for the biogas upgrade unit depended on the choice of biogas usage. However, an electrical usage factor per mole of methane treated was calculated for biogas with the determined methane content, based on the electrical requirement of the compressors in the PSA unit. Three compressors were required: two for purifying the biogas and one for storing it. Compressor power ratings as a function of volume flow rate were available for the inlet and outlet pressures prescribed by the PSA unit. To calculate the amount of electricity required to purify and compress a mole of methane, biogas was assumed to pass through the compressors without altering the biogas temperature or composition (although in practice this would have to be achieved with a cooling system which would increase the electrical demand). Thus, the amount of biogas (kmol) containing a kmol of methane was calculated using the molar fraction of methane in post‐scrubber biogas as calculated previously in equation (47):

138

4

1

CHBGM C

n = (53)

where nBGM = amount of biogas (kmol) containing 1 kmol methane, and CCH4 = molar fraction of methane (mol:mol) in post‐scrubber biogas.

The volume of post‐scrubber biogas containing one mole of methane was then calculated using the ideal gas equation of state, as follows:

std

ambBGMBGM P

RTnV = (54)

where VBGM = volume of post‐scrubber biogas (m3) containing 1 kmol methane, nBGM = amount of biogas (kmol) containing 1 kmol methane, R = gas constant (8.314 J mol‐1 K‐1), Tamb = ambient temperature (K), and Pstd = standard pressure (101.325 kPa).

This was used to calculate the molar electrical demand for the PSA unit in purifying one mole of methane, as follows:

BGMPSAPSA VUE = (55)

where EPSA = electrical demand from the PSA unit (kWh kmol‐1 CH4) for purifying 1 kmol of methane from biogas, UPSA = volumetric electrical demand factor for the PSA unit (kWh m‐3 biogas compressed), and VBGM = volume of biogas (m3) calculated in equation (54).

The PSA unit removed hydrogen sulfide, water, and carbon dioxide based on the removal efficiencies indicated by the manufacturer. The manufacturer’s methane recovery efficiency was converted to removal efficiency so that the total amount of gas removed could be calculated as follows:

( )∑= iBGMiremoved Cnrn (56)

where ri = removal efficiency ( percent) for biogas component i, nBGM = amount of biogas (kmol) containing 1 kmol methane, and Ci = fraction of biogas component i in post‐scrubber biogas (mol:mol).

The amount of biogas remaining to be compressed after the PSA unit had purified the methane was calculated by subtracting the amount of gas removed from the amount of biogas input to the PSA unit. The net result was the following equation:

( )∑−= iiBGMcompress Crnn 1 (57)

where ncompress= amount of cleaned biogas (kmol) compressed to CNG, nBGM = amount of biogas (kmol) containing 1 kmol methane, ri = removal efficiency ( percent) for biogas component i, and Ci = molar concentration of biogas component i in post‐scrubber biogas (mol:mol).

139

This was converted to a volume of biogas at ambient temperature using the ideal gas equation of state for calculation of the compressor’s electrical demand, such that:

std

ambcompresscompress P

RTnV =

(58)

where Vcompress = volume of purified biogas (m3) per kmol of CH4 compressed to its final storage pressure, R = gas constant (8.314 J mol‐1 K‐1), Tamb = ambient temperature (K), and Pstd = standard pressure (101.325 kPa).

Similar to equation (55), the equation used for calculating the electrical requirement for the storage compressor was:

compresscompresscompress VUE = (59)

where Ecompress = electrical requirement (kWh) for the storage compressor, Ucompress = volumetric electrical power rating (kWh m‐3 biogas) for the storage compressor, Vcompress = volume of biogas compressed (m3).

The net result of the above calculations was the following expression for the electrical requirement for upgrading one mole of methane to CNG:

( )[ ]stdCH

ambiicompressPSACNG PC

RTCrUUu4

1 ∑−+= (60)

where uCNG = specific biogas upgrade electrical demand (kWh kmol‐1 CH4), UPSA = volumetric electrical power rating (kWh m‐3) for the PSA unit, Ucompress = volumetric electrical power rating (kWh m‐3) for the storage compressor, ri = removal efficiency ( percent) for biogas component i, and Ci = molar concentration of biogas component i in post‐scrubber biogas (mol:mol), CCH4 = molar concentration of methane in the post‐scrubber biogas (mol:mol), R = gas constant (8.314 J mol‐1 K‐1), Tamb = ambient temperature (K), and Pstd = standard pressure (101.325 kPa).

Calculation of the Diversion of Biogas to the Energy Conversion Units

The model was designed to calculate how much biogas should be distributed to the various processing units to produce the most energy for sale while providing the parasitic heat and electricity demands (the latter was purchased from the local utility in the CNG scenario). For each scenario it was assumed that the heat required by the digester would be provided by burning biogas if necessary. If the scenario included an engine generator with heat recovery, the amount of heat provided by the generator was subtracted from the digester requirement before diverting biogas to the boiler. For the CG and CNG scenarios, it was assumed that biogas not used directly for heat would be diverted to either the generator in the first case or the PSA unit in the latter case for upgrading to CNG. For the CG+CNG scenario, it was assumed that enough biogas would be burned in the generator to provide the system’s electrical requirements prior to diverting the remaining biogas to the boiler (if necessary) and the PSA unit.

140

The sum of the amounts of methane sent to all different units had to equal the total amount of methane produced:

CNGDCCGM nnnn ++= (61)

where nM = total amount of methane produced (kmol), nCG = amount of methane sent to the engine generator (kmol), nDC = amount of methane sent to the boiler for heating the digester (kmol), and nCNG = amount of methane sent to the PSA unit (kmol).

For the CNG scenario, the amount of methane needed to heat the digester was sent to the boiler, and the remaining biogas was sent to the biogas upgrade unit. In this case the amount of biogas diverted to the boiler for heat and to the biogas upgrade unit for CNG was determined as follows:

DC

reqDC F

Qn =

(62)

DCMCNG nnn −= (63)

where nDC = amount of methane diverted to the boiler (kmol), Qreq = digester heat requirement (kWh), FDC = boiler heat recovery factor (kWh kmol‐1), nCNG = amount of methane diverted to the PSA unit (kmol), and nM = amount of total methane available (kmol).

For the remaining scenarios, a series of conditional statements were used to determine the appropriate biogas diversion. The CG scenario required a set of decisions that did not include the biogas upgrade unit (Figure 70). The CNG+CG scenario required a different set of decisions based on the assumption that the system defaulted to using the available biogas for self‐heating first, then for self‐powering and self‐heating, and finally for power, heat, and CNG if enough biogas were available.

141

Figure 71. Flow Chart for Determining the Distribution of Biogas in the CG Scenario

For the CG scenario, if converting all the biogas was found to generate enough heat to maintain the reactor temperature, nCG = 100 percent and no further calculation was necessary. Similarly, if converting all the biogas to heat provided less than the system’s total heat requirement, nDC = 100 percent, and the additional heat and electricity required by the system would be purchased from the utility company. In the special case that the generator could not produce enough heat to maintain the reactor temperature but excess biogas would be available if it were all converted to heat, nDC and nCG had to be solved simultaneously. In this case, the two constraints used were:

The heat recovered from the generator and the boiler had to equal the heat required for the digester.

The methane used for electricity generation and the methane used for heat generation had to equal the total amount of methane produced by the digester.

reqCGCGDCDC QnFnF =+ (64)

MCGDC nnn =+ (65)

where FDC and FCG = molar heat recovery factors (kWh kmol‐1 CH4) for of the boiler and electricity generator, respectively, nDC and nCG = amount of methane diverted (kmol) to the boiler and electricity generator, Qreq = digester heat requirement (kWh), nM = total amount of methane produced (kmol).

142

Equations (64) and (65) were solved for the unknown amounts of methane diverted to the boiler and cogen as follows:

DCCG

reqMCGDC FF

QnFn

−

−=

(66)

DCCG

MDCreqCG FF

nFQn

−

−=

(67)

where FDC and FCG = molar heat recovery factors (kWh kmol-1 CH4) for of the boiler and electricity generator, respectively, nDC and nCG = amount of methane diverted (kmol) to the boiler and electricity generator, Qreq = digester heat requirement (kWh), nM = total amount of methane produced (kmol).

The CNG+CG case required more complicated evaluation (Error! Reference source not found.). First, if the biogas could not provide enough heat for the digester, all of the biogas would be burned for heat generation. Next, if the biogas could not provide enough heat and electricity simultaneously, the diversion had to be solved using equations (66) ‐ (67). In order to determine whether the biogas could provide enough heat and electricity simultaneously, the following conditions had to be met:

reqCG En =α (68)

reqDCDCCGCG QnFnF ≥+ (69)

MDCCG nnn =+ (70)

where α = electricity generation factor (kWh kmol‐1 CH4), nCG = amount of methane diverted to the generator (kmol), Ereq = total electrical demand of the digester and PSA unit (kWh), FDC and FCG = heat recovery factors (kWh kmol‐1 CH4) for of the boiler and electricity generator, respectively, nDC = amount of methane diverted to the boiler (kmol), Qreq = digester heat requirement (kWh), nM = total amount of methane produced (kmol).

These were combined to give the following condition:

reqreq

MDCreq

CG QE

nFE

F ≥⎟⎟⎠

⎞⎜⎜⎝

⎛−+

αα (71)

where α = electricity generation factor (kWh kmol‐1 CH4), Ereq = total electrical demand of the digester and PSA unit (kWh), FDC and FCG = heat recovery factors (kWh kmol‐1 CH4) for of the boiler and electricity generator, respectively, Qreq = digester heat requirement (kWh), nM = total amount of methane produced (kmol).

Equation (71) was provided as a conditional statement in Excel®. If this condition were met, the model checked whether the heat captured from the generator provided the

143

system’s heat demand when just enough electricity was produced to provide the system’s electricity demands. In this case, nDC = 0 and constraints solved were:

reqCNGCNGCG EnUn +=α (72)

MCNGCG nnn =+ (73)

where nCG and nCNG = amount of methane diverted (kmol) to the electricity generator and PSA unit, respectively, α = electricity generation factor (kWh kmol‐1 CH4), UCNG = biogas upgrade electrical requirement factor (kWh mol‐1 CH4), Ereq = total electrical demand of the digester and PSA unit (kWh), and nM = total amount of methane produced (kmol).

The solution to this system of equations was:

CNG

reqMCNGCG U

EnUn

+

+=

α (74)

CNG

reqMCNG U

Enn

+

−=αα

(75)

where nCG and nCNG = amount of methane diverted (kmol) to the electricity generator and PSA unit, respectively, α = electricity generation factor (kWh kmol‐1 CH4), UCNG = biogas upgrade electrical requirement factor (kWh mol‐1 CH4), Ereq = total electrical demand of the digester and PSA unit (kWh), and nM = total amount of methane produced (kmol).

Note that nCNG + nCG = nM as required, and that nCNG > 0 since the total amount of recoverable electricity (αVM) was already determined to be greater than the digester’s electrical requirements.

If it were found that the result of equation (74) did not divert enough biogas to the generator to provide the heat requirement, biogas would have to be burned in the boiler to provide the heat requirement. In this case nDC, nCG, and nCNG would have to be solved simultaneously. The following constraint would be added:

reqDCDCCGCG QnFnF =+ (76)

where FDC and FCG = heat recovery factors (kWh kmol‐1 CH4) for of the boiler and electricity generator, respectively, nDC and nCG = amount of methane diverted (kmol) to the boiler and electricity generator, respectively, and Qreq = digester heat requirement (kWh).

The system of equations (74), (75) and (76) could then be solved simultaneously, resulting in the following equations:

( )( ) DCCNGCGDC

reqDCreqMDCCNGCG FUFF

EFQnFUn

α+−

+−= (77)

144

( ) ( )( ) DCCNGCGDC

reqreqCGMCGreqCNGDC FUFF

QEFnFQUn

αα

+−

−−−= (78)

( ) ( )( ) DCCNGCGDC

reqreqCGreqMDCCNG FUFF

QEFEnFn

ααα

+−

−+−= (79)

where FDC and FCG = heat recovery factors (kWh kmol‐1 CH4) for of the boiler and electricity generator, respectively, and nCG, nDC, and nCNG = amount of methane diverted (kmol) to the electricity generator, boiler, and PSA unit, respectively, Qreq = digester heat requirement (kWh), α = electricity generation factor (kWh kmol‐1 CH4), UCNG = biogas upgrade electrical requirement factor (kWh mol‐1 CH4), Ereq = total electrical demand of the digester and PSA unit (kWh), and nM = total amount of methane produced (kmol).

Note that equations (77), (78), and (79) sum to nM as required.

145

Figure 72. Flow Chart for Determining the Distribution of Biogas When Converting to Heat, Electricity and CNG (CG+CNG Scenario)

Stop

Emax > EreqQDC + QCG > Qreq

(Can the digester heat and power itself?)

QDCmax > Qreq?(Can the digester

heat itself?)

Let PCNG = 0 and solve for PCG and PDC while maximizing PCG

Let PDC = 0 and solve for PCG and PCNGsimultaneously such that ECG = Ereq

PCG = 0%PDC = 100%PCNG = 0%

No

Yes No

Yes

QCG ≥ Qreq?(Does cogen heat cover the digester

requirements)Yes

Solve for PDC, PCNG and PCG simultaneously

No

QDCmax maximum heat recoverable from the boiler using all the biogasQreq digester heat requirementEmax maximum possible electricity productionEreq digester electrical requirementECG electricity produced by the generatorQCG heat recovered from the generatorQDC heat recovered from the boilerPCG percent of biogas sent to the generatorPDC percent of biogas sent to boilerPCNG percent of biogas upgraded to CNG

Calculation of the financial performance of the APS Digester system

The results of the mass and energy balance were fed into the cost and revenue calculators. Onsite Power Systems, Inc. provided itemized cost estimates based on their experience building the pilot APS Digester system. The costs were divided into capital and ordinary expenses as shown in Table 14.

146

Table 14. Cost Categories Considered for the APS Digester System Cost Calculator Capital Costs Ordinary Expenses

Pre-construction Digester feed

Digester Labor

Biogas cleaning system Maintenance

Electricity generator with heat recovery Utilities

Direct combustion heat recovery system Administration

Post-treatment unit Property costs

Contingency Contingency

Pre‐construction costs were estimated for a stand‐alone digester project based on the footprint needed for the computed digester size (Table 15). The capital cost of the digester included the cost of the reactors, material handling system, the circulation, heating and biogas collection system, and the computer control and data acquisition system. The biogas cleaning system included the cost of a hydrogen sulfide scrubbing unit, a moisture removal unit, and a PSA unit (estimate provided by QuestAir, Inc., 2008) which included the compressor for bringing the biogas to its storage pressure. The PSA capital cost was only included in the appropriate scenarios. The cost of the reactors was calculated based on the reactor volume. The rest of the digester capital costs were computed specifically for a digester sized to treat 115 t d‐1 of mixed food and green waste. The electricity generation system was sized based on the amount of biogas converted to electricity.

able 15. List of Pre-Construction Cost Items Used for Estimating the Pre-Construction Cost of a New Stand-Alone APS Digester System Project

Site Preparation Costs Permitting

Clear and grub Environmental Impact Report

Rough grading Conditional Use Permit

Soils testing and surveys Miscellaneous (water, construction, etc.)

Excavation and compacting State permits (local enforcement agency and waste management)

Utilities infrastructure

Off-site infrastructure

Clarifier

147

The operating and maintenance costs for industrial equipment typically lump labor and equipment maintenance costs, but the labor requirements for the APS Digester system were unknown. Therefore, the digester operating company provided estimates of the number of laborers and expected wages, separately from estimates on the maintenance costs. The maintenance cost for the hydrogen sulfide removal unit was based on the cost of replacing the hydrogen sulfide absorptive media, which was in turn based on the amount of biogas purified. The maintenance costs for the digester heating system and electrical generator were calculated based on the amount of energy produced by those units, but the remaining equipment maintenance costs were assumed at a rate based on the equipment’s capital cost. Utility costs included charges for electricity consumed (in the scenario that did not include a generator), a meter charge for all three scenarios, and demand charges where appropriate.

Revenues for the APS Digester system were assumed to come from a combination of the following sources: energy products (electricity and natural gas), residual solids, tax credits, and tipping fees. The model used the mass balance calculations to determine the amount of residual solids available for sale as compost. Similarly, the energy balance was used to calculate the amount of excess electricity and natural gas available for sale. Although the model was designed with the capability of including revenue from heat, water, and debt reserve, these revenues were set to zero for the current analysis.

A 20‐year annual cash flow was calculated for the resulting digester model. The annual cash flows were discounted at the MARR and summed to compute the NPW of each scenario. The IRR was calculated using the Excel® IRR function which uses an iterative trial‐and‐error method of finding the discount rate for which the NPW becomes zero [69]. This method of calculating IRR can have multiple solutions and does not distinguish between net positive and net negative cash flows. Therefore, the cash flow pattern was included for interpreting the IRR. Furthermore, the IRR does not depend on the scale of the initial investment, thus the magnitude of the IRR cannot be used to compare mutually exclusive projects such as the three scenarios analyzed here [69]. However, because it does not depend on the MARR assumption it is useful when the precise MARR is unknown. Therefore the IRR was calculated for each scenario, and the incremental IRR was calculated for the pair‐wise comparison between scenarios. The revenue requirements method was used to determine the levelized COE required to generate the required MARR [71]. The COE was, therefore, equal to the price of energy (either electricity or natural gas depending on the scenario) that set the NPW equal to zero. The COT was also calculated using the revenue requirements method; however, the required revenue was from tipping‐fees instead of energy. Since the amount of waste treated was the same for all three scenarios, the COT provided a consistent value for comparing the scenarios. For testing changes in the values assumed for computing the three scenarios, only the NPW was used to simplify the analysis.

The after‐tax cash flow for each year of the project was calculated as follows:

148

EIbpOCF CTDDERA −−+−−= (80)

where Acf = after‐tax cash flow ($ y‐1), R = revenues ($ y‐1), EO = ordinary expenses ($ y‐1), Dp = debt payments ($ y‐1), Db = debt borrowing ($ y‐1), TI = income taxes ($ y‐1), and CE = capital expenses ($ y‐1).

Capital expenses and debt borrowing occurred in year zero. Revenues and expenses were escalated annually beginning in year one at a user‐designated constant rate as follows:

( )nen rVV += 10 (81)

where Vn = value ($) in year n, V0 = current value ($), re = escalation rate ( percent y‐1), and n = time elapsed (y).

Income taxes were calculated as in equation (82) on the taxable income based on the user‐designated aggregate federal and state tax rates, adjusted to avoid double taxing as in equation (83). Taxable income was calculated as the difference between revenues, ordinary expenses, debt interest, and depreciation as in equation (84).

TAI ITT = (82)

SFFSA TTTTT −+= (83)

MDERI IOT −−−= (84)

where TA = adjusted tax rate ( percent), TI = income tax ($ y‐1), IT = taxable income ($ y‐1), TF = aggregate federal income tax rate ( percent), TS = aggregate state income tax rate ( percent), R = revenues ($ y‐1), EO = ordinary expenses ($ y‐1), DI = debt interest ($ y‐1), and M = depreciation ($ y‐1).

The NPW of the project was calculated by discounting the after tax cash flow at the user‐supplied minimum attractive rate of return for each year and summing the net cash flow over the life of the project, as follows:

( )∑=

−+=L

n

nmCFN rAW

01 (85)

where WN = net present worth ($), n = time elapsed (y), L = project lifetime (y), ACF = after‐tax cash flow ($ y‐1), and rm = discount rate ( percent y‐1) which was set equal to the minimum attractive rate of return.

4.3.2 Assumptions Made When Modeling The APS Digester System The APS Digester system was designed to accept 115 t d‐1 of a mixture of food and green waste, every day year‐round. The digester size calculation was based on the design assumptions in Table 16. Each hydrolysis reactor was assumed to be of the same volume and dimensions, and every reactor was assigned the same headspace allowance. The

149

reactor surface areas were calculated based on the assumption that the reactors were cylindrical, and the floor and ceiling areas were included even when using the surface area for calculating heat losses.

Table 16. Design Parameters Used to Size The APS Digester System and Compute the Amount and Volume of Waste Treated

Parameter Value

Daily feedstock treatment capacity (t d-1)

115

Organic loading rate (kg m-3 d-1) 8.0

Specific volume of feedstock (m3 t-1) 1.00

Headspace allowance (% of working volume)

15

Number of hydrolysis reactors 4

Number of gasification reactors 1

Biogasification-to-hydrolysis volume ratio

1.00

Hydrolysis aspect ratio (height/diameter)

0.5

Gasification aspect ratio (height/diameter)

0.5

Capacity factor (%) 100

Samples of food and green waste from municipal sources in the Bay Area have been characterized and tested separately and together in bench‐scale anaerobic digestion trials [72‐74]. Combined food and green waste yielded slightly more biogas than the theoretical average of the separate feeds (Table 17), which may indicate an interaction effect that enhanced the biogas yield. However, difficulties during the study reduced the reliability of the results; therefore, the theoretical average feedstock characteristics and biogas yield of a 1:1 mixture (VS basis) of food and green waste was used as a 20‐year aggregated average for the current model, and no interaction effects were assumed.

150

Table 17. Characteristics and Anaerobic Degradability of Municipal Food Waste (FW) and Green Waste (GW) Collected from the Bay Area in California

Reference Source [75] [73] [74]

Experimental Design

30 day, 1 L batch

Theoretical calculation

Lab-scale APS

Theoretical calculation

Lab-scale APS

Substrate FW GW Mix (1:1 VS basis)

FW GW Mix (1:1 VS basis)

Mix (1:1 VS basis)

MC (%) 69 75 72 69 73 71 76

TS (%) 31 25 28 31 27 29 24

VS (% of TS) 85 88 86 87 78 82 85

VS (% of wet weight)

26 22 24 26 21 24 20

Biogas Yield Potential* (m3 t-1 VS destroyed)

745 504 792 705 562 642.3 718

Methane Content (%)

73 55 65 64 58 61 75-80

Biodegradability (% VS reduction)

81 79 80 85 78 82 75

Methane Yield* (m3 t-1 VS)

435 226 279 379 252 271 430

* Gas volumes were adjusted to standard temperature and pressure (273 K, 101.325 kPa).

Theoretical Biogas and Methane Yield Calculation and Comparison With Experimentally Determined Yields Assuming that all of the VS consumed were converted to biogas (neglecting cell growth, for the moment), mass balance would require that 1 kg of VS would yield 1 kg of biogas, or 1 kg kg‐1 VS consumed. Assuming that biogas contains only CH4 and CO2, the methane yield can be calculated from the biogas yield. Since some of the VS consumed should be converted to cellular biomass instead of biogas, the yield calculated in this manner would be a maximum theoretical yield.

151

To calculate the theoretical biogas yield per unit mass of VS consumed (and biogas produced), the following relation must hold, based on the above assumptions:

2244 COCOCHCHbg MnMnm += (86)

where bgm = 1kg biogas, nCH4 = amount of methane (kmol) in 1 kg biogas, nCO2 = amount

of carbon dioxide (kmol) in 1 kg biogas, MCH4 = molar mass of methane (kg kmol‐1), and MCO2 = molar mass of carbon dioxide (kg kmol‐1).

The biogas methane content can be expressed on a molar basis as follows:

42

4

4CHCO

CHCH nn

nP

+=

(87)

where PCH4 = biogas methane content (mol:mol), nCH4 = amount of methane (kmol), and nCO2 = amount of carbon dioxide (kmol).

The volumes of CH4 and biogas occupied can be calculated based on the ideal gas equation of state for CH4 and the observation that the gas is composed of only two species.

44 CHstd

stdCH n

PRT

V = (88)

4

4

CH

CHBG P

VV =

(89)

where PCH4 = biogas methane content (mol:mol), nCH4 = amount of methane (kmol) in 1 kg biogas, nCO2 = amount of carbon dioxide (kmol) in 1 kg biogas, VCH4 = volume of methane (m3) at STP, R = gas constant (8.314 J mol‐1 K‐1); Tstd = standard temperature (273 K); Pstd = standard pressure (101.325 kPa), VBG = volume of biogas (m3) at STP.

This yield could be calculated as a function of the methane content by substituting the equations (88) and (89) relations into equation (86), which results in the following equation for biogas yield:

( ) ⎥⎥⎦

⎤

⎢⎢⎣

⎡

+−=

2244

1

COCOCHCHstd

stdBG MMMPP

RTY

(90)

where YBG = biogas yield (m3 kg‐1 VS) at STP, MCH4 = molar mass of methane (kg kmol‐1), MCO2 = molar mass of carbon dioxide (kg kmol‐1), PCH4 = biogas methane content (mol:mol), R = gas constant (8.314 J mol‐1 K‐1), Tstd = standard temperature (273 K), Pstd = standard pressure (101.325 kPa).

Assuming MCH4 = 16 kg kmol‐1 and MCO2 = 44 kg kmol‐1 and multiplying by 1000 kg t‐1, equation (90) becomes:

152

457.1

873

CHBG P

Y−

= (91)

where YBG = biogas yield (m3 t‐1 VS) and PCH4 = biogas methane content (mol:mol).

From this the theoretical maximum methane yield was calculated for methane contents from 40 percent to 70 percent (Figure 73). For glucose, the stoichiometric ratio of methane to carbon dioxide can be determined from the following balanced reaction:

246126 33 COCHOHC +→ (92)

In this case the methane content in the biogas (PCH4) would be 50 percent, which leads to a theoretical biogas yield of 816 m3 t‐1 VS at STP and a theoretical methane yield of 408 m3 t‐1 VS at STP. Different biochemical reactions would yield different stoichiometric ratios.

For the assumed methane content of 61 percent used in the model, the theoretical maximum biogas yield potential would be 909 m3 t‐1 VS destroyed. The value assumed in the model (642 m3 t‐1 VS destroyed) was 71 percent of the maximum. Therefore, the assumed value was within realistic expectations for biogas yield and may be lower than the optimal yield.

Figure 73. Theoretical Maximum Biogas Yield Potential at 298 K,101.325 Kpa as a Function of Methane Content.

40 45 50 55 60 65 70700

750

800

850

900

950

1000

1050

Methane Content (%)

Biog

as Y

ield

Pot

entia

l (m

3 t-1 V

S de

stroy

ed)

Assumptions Made For Calculating the Mass Balance for Gases Produced During Anaerobic Digestion and Subsequent Energy Conversion

153

Although biogas consisted primarily of methane and carbon dioxide, water vapor, hydrogen sulfide, and trace contaminants were also present (Table 18). The model assumed that biogas exited the digester at the maintenance temperature of 55°C carrying water vapor which condensed as the biogas cooled to ambient temperature. The mean annual ambient temperature assumed was 16°C, which was the 2007 mean annual temperature for California, according to the National Climatic Data Center4.

The saturation pressure of water vapor at a given temperature was calculated by fitting a fourth order polynomial to data taken from the steam tables of Keenan et al [76] (Figure 74). This resulted in the following equation:

( )47

36232

10772.6

10485.610851.110137.4615.02

T

TTTP gOH

−

−−−

×+

×+×+×+= (93)

where PH2O(g) = saturation pressure of water vapor (kPa) and T = temperature (°C).

Although the form of the curve‐fit was chosen arbitrarily, the polynomial equation was found to predict water vapor density and saturation pressure within 0.4 percent of the published steam table data and biogas vapor content within 1.6 percent of experimental data. Biogas vapor content was calculated by dividing the saturation pressure of water vapor by the standard pressure for air (101.325 kPa), which assumed that biogas was at equilibrium with atmospheric pressure. The predicted biogas vapor content was found to closely match that found in empirical studies [70].

4 Data accessed at http://www.ncdc.noaa.gov/oa/climate/research/cag3/regional.html on March 20, 2008.

154

Figure 74. Water Vapor Density and Biogas Vapor Content as a Function of Temperature. Curves Were Fit to Published Data [70, 76].

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0%

5%

10%

15%

20%

25%

0 20 40 60D

ensit

y of

Sat

urat

ed W

ater

Vap

or (k

g m

-3)

Bio

gas V

apor

Con

tent

(%)

Temperature (°C)

Biogas Vapor Content from Saturation Tables (%)

Biogas Vapor Content Data from Dupont, 2006

Density of Saturated Water Vapor (kg/m3)

Biogas Vapor Content Curve-Fit

Vapor Density Curve-Fit

The hydrogen sulfide scrubber used at the pilot plant incorporated an iron‐impregnated solid granular medium (SulfaTreat, Chesterfield, MO). The model used the company’s reported hydrogen sulfide removal efficiency to calculate the mass balance (Table 18). Water removal by condensation was calculated for cooling of saturated biogas from the reactor temperature to ambient temperature as described earlier. Additional drying equipment required to remove the remaining moisture was included in the capital cost, but neglected for the mass balance calculation. The pressure swing adsorption unit proposed for use at the pilot site (QuestAir Technologies, Inc., Vancouver, Canada) was assumed to reduce the amount of remaining hydrogen sulfide, ammonia, and moisture in addition to carbon dioxide with removal efficiencies assumed in Table 18. Some methane was assumed to be lost during purification, therefore the methane recovery efficiency was less than 100 percent.

155

Table 18. Base Assumptions Used for Calculating Mass Balance on Non-Combustion Biogas Processing Equipment

Parameter Value

Hydrogen Sulfide Scrubbing Unit

H2S Removal Efficiency (%) 95

Hydrogen Sulfide Content of Biogas (ppm) 500

Pressure Swing Adsorption Unit

Hydrogen sulfide removal efficiency (%) 95

Water vapor removal efficiency (%) 99.5

Carbon dioxide removal efficiency (%) 93.5

Methane recovery efficiency (%) 93.2

The US EPA performed a life‐cycle assessment for MSW disposal of many items, including food scraps, yard trimmings, leaves, and grass clippings, in landfills with and without biogas recovery systems [78]. For this study, the green waste was assumed to be highly biodegradable, therefore the carbon dioxide emissions factor for grass clippings was used to represent the green waste, and the factor for food scraps was used to represent the food waste. The landfill configuration was assumed to match the average landfill configuration for the US. The wet‐weight fractions of food and green waste in the mixed stream modeled were 44 and 56 percent, respectively. This resulted in a composite carbon dioxide emissions factor of 342 kg CO2 t‐1 waste.

Assumptions made for calculating the mass balance of solids and liquids exiting the digester For the current model, it was assumed that no additional fresh water was added and no process water was retained. The model could be used to predict the mass balance resulting from fresh water addition by revising the MC of the feedstock to include the fresh water added. If process water were retained, it would contain some TS and the TS content would change as a function of the volume of process water retained. Therefore, the mass balance would have to be recalculated to account for these dynamics. The calculation of the mass balance on solids and liquids was performed as described earlier. The solid/liquid separation unit was assumed to extract water from the residual slurry with a solids recovery efficiency of 80 percent and a press cake MC of 65 percent, which was typical of screw presses treating agricultural waste [79]. The amount of waste‐water to be treated was calculated through the mass balance, but it was assumed that effluent would be treated on‐site.

Assumptions Made For Calculating the Energy Balance of the APS Digester System

156

The energy balance of the digester was calculated by estimating the heat and electricity required to operate the system and subtracting the result from the energy content of the methane gas produced by the digester. The higher heating value of methane was used throughout (Table 19). It was assumed that heat was only consumed in heating the reactors. Heat was assumed to be recovered from the boiler and engine generator only. Heat from the compressors used by the PSA unit was not included in the model. Electricity was assumed to be consumed by the digester control and material handling systems and the PSA unit.

Assumptions made for calculating the heat consumption and production for the APS Digester system For calculating the heat requirements, the heat capacity of the reactor contents, ambient and reactor temperatures, insulation heat‐transfer coefficient, and reactor surface area were required. The heat capacity of the feedstock was not computed directly, but previous research on organic matter yielded the following empirical equation for heat capacity [80]:

wafpcp XXXXXc 187.4837.0675.1549.142.1 ++++= (94)

where cp = heat capacity, Xi = mass fraction of component i, w = water, c = carbohydrate, p = protein, f = fat, and a = ash.

For the waste stream modeled here, the percentage of carbohydrate, protein, and fat was not measured. However, in a study of foodwaste from several different sources in Sacramento, CA the organic matter consisted on average of 70 percent carbohydrate, 20 percent protein, and 10 percent fat [81]. Assuming the VS was composed primarily of organic matter, equation (94) can be re‐written as:

( ) ( )VSVSMCMCC p 50.1)1(837.0187.4 +−−+= (95)

where MC = moisture content (wet basis), and VS = volatile solids content (wet basis).

In the absence of precise composition data, equation (95) was used to approximate the heat capacity of organic waste streams with known MC and VS values. For the food and green waste modeled here, the estimated heat capacity of the feedstock was 3.41 kJ kg‐1 °C‐1.

The R‐value and thickness of the insulation was reported by Onsite Power Systems from their pilot digester trials (Table 19). The heat transfer coefficient will depend on the conditions at the digester site. For a single cylindrical tank with forced‐air convection in cross‐flow, the heat transfer coefficient would depend on the properties of the fluids and surfaces involved [82]. Typically, at windy sites the heat transfer coefficient can range from 20 – 250 W m‐2 °C‐1 [83]. Here, the heat transfer coefficient was set to 50 W m‐2 °C‐1. When using the model for a specific site, the precise heat transfer coefficient can be determined.

157

It was assumed that all excess heat would be dissipated to the environment, although in some settings an industrial neighbor may be able to use some of the excess heat.

Table 19. Base Values Assumed for Modeling the Heat Production and Consumption of the APS Digester System

Parameter Value

Heat Production

Higher heating value of methane (kJ mol-1) 890

Generator heat recovery factor (%) 60

Direct combustion boiler heat recovery factor (%) 60

Generator efficiency -- methane to electricity (%) 30

Heat Consumption

Ambient Temperature (°C) 16

Reactor Temperature (°C) 55

Feed Solids Heat Capacity (kJ kg-1 °C-1) 3.41

Heat Transfer Coefficient (W m-2 °C-1) 50

Insulation thermal conductivity (W m-1 °C-1) 0.0146

Insulation thickness, d (cm) 2.54

Insulation R-value (m² °C W-1) 1.74

Assumptions Made for Calculating the Electricity Consumption and Production

Onsite Power Systems reported that the digester required a 300 HP pump operating at 33% capacity for mixing and material handling. The feeding system required a 21 HP chopper pump operating at 33 percent capacity. The heating system required the equivalent of a 5 HP pump operating at 100 percent capacity for circulating digester liquid through the heat exchanger. The control system was assumed to draw 1 kW of electricity at full time. Altogether, this gave the digester a gross power rating of 84.52 kW (assuming a 100 percent capacity factor). The PSA unit included two compressors, each of which compressed the biogas from 14 kPa to 690 kPa gauge pressure with an electrical requirement of 0.116 kWh m‐3 gas compressed. The natural gas storage compressor was assumed to compress the natural gas exiting the PSA unit at 655 kPa to 24,800 kPa gauge pressure with an electrical requirement of 0.212 kWh m‐3 gas compressed.

158

For the first scenario, it was assumed that all of the biogas produced would be used to generate electricity at 30 percent efficiency based on the higher heating value of methane (Table 20). The generator was sized based on the predicted electrical production and a 90 percent capacity factor. In the CNG scenario the entire electrical demand was assumed to be supplied by the utility grid.

Table 20. Base Values Assumed for Modeling the Electricity Production and Consumption for the Digester and Biogas Upgrade Unit

Parameter Value

Electricity Production

Generator efficiency -- methane to electricity (%) 30

Generator capacity factor (%) 90

Electricity Consumption

Recirculation system power rating (kW) 74.6

Feed system power rating (kW) 5.2

Heating system power rating (kW) 3.7

Control system power rating (kW) 1.0

PSA unit power rating, including two compressors (kWh m-3) 0.232

Storage compressor power rating (kWh m-3) 0.212

Assumptions Made For Calculating the Financial Performance of the APS Digester System Under Various Process Configurations

From an energy standpoint, the APS Digester system was treated as a distributed power generator. For the electricity producing scenarios the digester could either be considered a retail generator or a self‐generation system, depending on whether the digester was connected to the grid as a stand‐alone unit or behind the meter of an electricity customer with demands at least as large as the production capacity of the digester. A stand‐alone digester would not need to consume any grid power, therefore it would not be assessed any of the demand charges levied by the utility company for providing the capacity to meet a customer’s maximum demand throughout the year. A self‐generation digester would incur a demand charge based on the total installed capacity of the system if it should be shut down at any point while the customer demand remained unchanged. However, the stand‐alone digester would only be able to sell the electricity at the wholesale rate (unless renewable‐electricity feed‐in tariffs were considered), whereas the self‐generation digester would offset the consumer’s electricity which was valued at the retail rate. Both cases were evaluated but the stand‐alone digester assumptions were

159

used for comparison with the non‐electricity producing scenarios since co‐location would also alter the capital costs and operating expenses which was not considered in this analysis.

Assumptions Made for Calculating Capital Costs for the APS Digester System

For the current model, land value was omitted from the capital cost calculations because the effect of land on the NPW would be the same for all three scenarios. The preconstruction costs were itemized (Table 15) and estimated by Waste Systems Inc. who had partnered with Onsite Power Systems in evaluating the potential of digesting MSW for the City of Industry in Southern California. The digester costs, biogas cleaning equipment costs, and boiler cost were estimated by Onsite Power Systems (Table 21). The electricity generator was sized differently in the CG and CG+CNG scenarios; therefore an economy of scale was used to calculate the capital cost of the generator. All of the capital was assumed to have no salvage value after 20 years.

Table 21. Assumed Costs Used to Calculate the Capital Costs for the APS Digester System Parameter Value

Preconstruction Costs

Site Prep ($) 130,000

Predevelopment Costs ($) 275,000

Digester Costs

Volumetric Tankage Cost ($ m-3) 660

Circulation, Heating, and Biogas Collection System ($) 2,000,000

Feedstock Handling System ($) 750,000

Water Treatment System, i.e. SBR ($) 30,000

Biogas Cleaning Equipment Costs

Pressure Swing Adsorption Unit ($) 1,000,000

Hydrogen Sulfide Removal Unit ($) 50,000

Moisture Removal Unit ($) 10,000

Energy Conversion Equipment Costs

Cogen System Capital Cost Scaling Factor -0.102

Cogen System Capital Cost ($ kWh-1)* 1,106

Boiler Heating System Specific Capital Cost ($ kW-1) 30.00

* Cost of generator sized to convert 100 percent of biogas produced.

160

The engine generator electrical capacities ranged from 200 kW to 1.3 MW. Therefore, the capital and maintenance costs were calculated based on the published cost data for distributed combined heat and power engine generators (Error! Reference source not found.). A power function was fit to the cost data and the resulting equations were used by the model to predict the capital and maintenance cost of the generator as follows:

102.02267 −= gencap PC (96)191.0

int 043.0 −= genma PC (97)

where Pgen = generator electrical capacity (kW), capC = specific capital cost ($ kW‐1

electrical capacity), and intmaC = maintenance cost ($ kWh‐1 electricity produced).

Figure 75. Capital and Maintenance Cost Data and Curve Fitting for Combined Heat and Power Engine Generators [62]

y = 2267.1x-0.102

R² = 0.9657

y = 0.043x-0.191

R² = 0.9599

0.00

0.01

0.01

0.02

0.02

0.03

0200400600800

10001200140016001800

0 1000 2000 3000 4000 5000 6000

Mai

nten

ance

Cos

t ($

kWh-

1 ) 2

007

Cap

ital C

ost (

$ kW

-1)

2007

Maximum Capacity (kW)

Capital Cost Maintenance Cost

The capital cost of a boiler sized to provide the entire heat demand of the digester was included in every scenario, since it would be required both for startup and as a backup. The maintenance cost of the boiler, however, was determined based on the amount of heat produced. No economy of scale was calculated, but the maintenance cost for water heating systems was assumed to be relatively low ($0.001 kWh‐1), therefore scaling would have little influence on overall operating costs.

Assumptions Made for Estimating the Ordinary Expenses

Maintenance costs and labor were estimated separately. The annual maintenance cost assumed for the digester and wastewater treatment was 3 percent of the capital cost (Table 22). The rate assumed for the PSA unit was higher (5 percent) due to the novelty and uniqueness of the technology. According to Onsite Power Systems, the newly

161

designed computer control system reduced the labor costs from the preliminary estimates. Two full‐time laborers, one plant manager and a technician at half‐time were envisioned to operate and maintain the APS Digester system, paid at the prevailing wage rates (Table 22). The labor requirement was assumed to be equal for all of the scenarios. The cost of one contracted administrator working 10 hours per week was included in the administrative costs along with management and licensing fees. The hydrogen sulfide unit’s maintenance cost was based on the amount of hydrogen sulfide that the absorption medium could absorb before it had to be replaced and the cost of the medium. Feedstock was assumed to have no price in the current model since tipping fees were charged.

The tariff schedules from Pacific Gas and Electric (PG&E) were used to determine the utilities costs (Table 22). Since the entire electrical demand for the system was less than 500 kW, the A‐10 rate schedule (Revised Cal. P.U.C. Sheet No. 24876‐E, effective May 1, 2006) was consulted for the retail electricity charge. It was assumed that the electrical demand would be constant throughout the year (heat was provided by burning biogas); therefore the summer and winter non‐FTA primary voltage charges were averaged. Demand charges were calculated based on the peak electrical capacity required over the course of a year. An industry that offset its electricity use with power from a digester would normally incur a demand charge on the electrical capacity not provided by the digester. However, if the digester were unable to provide electricity at any point during the year, the demand charge would include the electrical capacity normally offset by the digester. Therefore, the demand charge for covering the additional electrical capacity was included in the digester’s ordinary expenses at the schedule E‐20 summer peak demand charge rate (Revised Cal. P.U.C. Sheet No. 26954‐E, effective Jan. 1, 2008). It was assumed that a stand‐alone digester that provided its own electrical demands would shut down its electrical systems when the generator was not running, therefore no demand charges would be incurred. For the scenario in which the digester electrical demand was purchased, the A‐10 rate schedule summer peak demand charge was used.

In July 2007, the California Public Utilities Commission (CPUC) required electric utility companies to implement feed‐in tariffs for renewable energy projects producing less than 1.5 MW under Decision (D.) 07‐07‐027. CPUC Resolution E – 4137, released on February 14, 2008, set the Market Price Referent for a 20 year project beginning in 2008 at $0.09572 kWh‐1. Because this is a new program, the feed‐in tariff price was not used for the base‐case analysis. However, the new Market Price Referent was used for comparing independent and co‐located systems.

162

Table 22. Assumed Costs Used to Calculate the Running Ordinary Expenses for the APS Digester System

Parameter Value

Maintenance Costs

Digester ($ $-1 cap y-1) 0.03

Boiler ($ kWh-1) 0.001

Pressure Swing Adsorption Unit ($ $-1 cap y-1) 0.05

Hydrogen Sulfide Removal Unit ($ kg-1 H2S removed) 7.56

Wastewater Treatment Unit ($ $-1 cap y-1) 0.03

Labor Costs

Operation Manager Salary ($ y-1) 70,000

Number of Managers 1

Hourly Laborer Wage ($ h-1) 12

Number of Laborers 2

Annual Technician Wage ($ y-1) 40,000

Number of Technicians 0.5

Labor Burden Rate (% of wages) 15

Administrative Costs

Administrator Contract Labor Wage ($ yr-1) 50,000

Number of Administrators 0.25

Management and Licensing Fees ($ y-1) 50,000

Utilities Costs

Retail Electricity Charge ($ kWh-1) 0.1033

Meter Charge ($ d-1) 3.94

E-20 Demand Charge ($ kW-1) 10.66

A-10 Demand Charge ($ kW-1) 8.11

Additional Costs

Insurance ($ y-1) 15,000

Property Tax Rate (% on equipment) 1.25

163

In order to account for uncertainties in the estimates, the gross costs were augmented by a contingency factor of 10 percent on capital and 5 percent on ordinary expenses.

Assumptions Made for Calculating Revenues for the Various Scenarios Modeled

Stand‐alone digesters were assumed to be able to sell all of their excess electricity at the going wholesale rate. The price used was the average 2007 wholesale price for a California trading hub, published by the Energy Information Administration (EIA)5. Digesters co‐located with a large electricity consumer were credited with the full value of the electricity offset. The price used was the non‐weighted average of the electricity rates published in the E‐20 customer rate schedule. This assumed that the digester produced electricity consistently and that the consumer used all of the electricity produced. In fact, some form of energy storage (for example as biogas or electricity) would allow the digester to produce electricity during peak demand periods; however, the generator would then need to be sized for peak production. Whether the increased offset rate would compensate for the added costs would depend on the size of the storage unit and the difference in electricity prices. For simplicity, an even electricity production and an aggregate average offset price were assumed.

The price used for natural gas revenues was the November 2007 City Gate6 price published by the EIA, converted from the reported price per thousand cubic feet to the price per kWh using the HHV for methane (Table 23). The price for the residual solids was set to half of the price of mature compost in Northern California. It was assumed that stabilization of the residual solids via aeration would be accomplished off‐site. The tipping fee used in this study (Table 23) was the price negotiated by Onsite Power Systems, Inc. for a potential project in Southern California. The negotiated price was for separated and ground organic waste. In California, landfills received an average of $34 t‐1 in 2000, which was actually slightly lower than the average 1995 tipping fee, according to the California Integrated Waste Management Board. Within California tipping fees ranged from $2.75 – $94 t‐1, with a trend toward higher fees in Southern California. Revenue from tipping fees was assumed to be the same for all the scenarios tested.

5 http://www.eia.doe.gov/cneaf/electricity/wholesale/wholesale.html

6 “A point or measuring station at which a distributing gas utility receives gas from a natural gas pipeline company or transmission system.” – EIA, 2007.

164

Table 23. Prices (in 2007 Dollars) Used for Calculating Revenues for the APS Digester System

Revenue Source Price

Wholesale electricity price ($ kWh-1) 0.0670

Retail electricity offset value($ kWh-1) 0.0872

Natural gas price/offset value ($ kWh-1) 0.0247

Residual solids price ($ t-1) 11.02

Net Tipping Fee ($ t-1) 24.25

Assumptions Made to Calculate Cash Flows and Overall Financial Performance

The cash flow was calculated beginning with capital costs and debt borrowing beginning in year zero. For subsequent years, revenues and expenses were escalated at the non‐seasonally adjusted ten‐year average Consumer Price Index inflation rate (1997 – 2007 CPI‐U, US City Average, All Items) escalation rate (Table 24). The model was designed with the ability to apply different escalation rates for heat, electricity, compost, and natural gas, but the same escalation rate was used for all revenues and expenses in the current analysis. It should be noted, however, that electricity prices have been more volatile. The average escalation rate for electricity increased over the past ten years and the true escalation of energy prices was somewhat uncertain.7

For each subsequent year, the net after‐tax cash flow was computed and discounted at the MARR to a present value as described earlier. The model used an aggregate tax rate. However, this was a simplification that may not apply if revenues were small enough to justify using the tax tables or if the project were to post losses which would need to be accounted for using a tax loss carry forward computation. The model allowed tax losses, assuming that either such losses would be carried forward or counted against additional tax liability from other revenue generating activities. A modified accelerated cost recovery system (MACRS) of the general recovery system type was used for calculating depreciation of the assets. As a waste recovery and resource reduction plant, the APS Digester system fell under asset class 49.5 of the IRS depreciation guide (2006 Publication 946, Table B‐2) which mandated the use of a 7‐year MACRS depreciation schedule.

7 Consumer Price Index and electricity price data was obtained from the U.S. Department of Labor, Bureau of Labor Statistics, http://www.bls.gov/cpi/.

165

Figure 76. Ten-Year Trend in the Annual Escalation Rate for Electricity in the U.S.

Tax credits were applied as a taxable revenue stream in the current analysis (Table 24). The Energy Policy Act of 1992 (Section 1212.e.2) established a $0.015 kWh‐1 federal tax credit for electricity produced from renewable resources. According to IRS Federal Tax Form 8835, the 2006 inflation adjusted credit came to $0.019 kWh‐1, but electricity produced from biomass not specifically grown as an energy crop was only allowed to take half of the credit, or $0.0095 kWh‐1. Furthermore, the eligibility period for landfill gas and agricultural waste projects established after 2005 was 10 years, but projects using cellulosic waste were only eligible to take the tax credit for 5 years. It was not clear which rule would apply to the APS Digester system treating organic waste. For the current analysis, the more conservative 5‐year eligibility period was used. It was assumed that the tax credit would be available over the full first five years of operation. However, the production tax credit has expired and been extended several times since its inception, and it was not clear how often it would be extended in the future.

For the scenarios in which natural gas was produced, two possible tax credits could be claimed. The Nonconventional‐Source Fuel Credit could be claimed using IRS Federal Tax Form 8907 for renewably produced natural gas. At the time of publication the April 2007 bulletin reported the credit at $4.72 per barrel‐oil‐equivalent. The tax form explicitly set the conversion rate for barrels‐oil‐equivalent at 5.8 million BTU per barrel. If the natural gas were used as a vehicle fuel, however, under Internal Revenue Code §§ 34, 6426(d), 6426(e), and 6427(e) the Alternative Vehicle Fuel Credit could be claimed using IRS Tax Form 4136. The credit amount would be $0.50 per gallon gasoline equivalent with a conversion rate of 121 cubic feet of CNG per gallon‐gasoline‐equivalent. Section 11113(b)(2) of the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA‐LU) defined a gasoline gallon equivalent as the amount of such fuel having a BTU content of 124,800 (higher heating

166

value). The tax code did not specify the temperature, pressure, or heating value of CNG to use. Therefore, the SAFETEA‐LU definition was used for calculating the value of the tax credit per kWh of CNG produced (Table 24). These credits were escalated during the eligibility period at the general inflation rate. No eligibility period for the Alternative Vehicle Fuel tax credit was currently delimited, but as a conservative measure it was assumed that a five‐year eligibility period would be imposed.

Table 24. Financial Assumptions Used for Calculating Cash Flows Parameter Value

Taxes

Federal tax rate (%) 35.0

State tax rate (%) 9.3

Producer tax credit ($ kWh-1) 0.0095

Nonconventional fuel tax credit ($ kWh-1) 0.0028

Alternative vehicle fuel tax credit ($ kWh-1) 0.0137

Depreciation schedule used MACRS 7 year

Financing

Escalation rate (% y-1) 2.57

Debt ratio (%) 0

Interest rate on debt (% y-1) 7.50

Debt Term (y) 20

Economic life (y) 20

Minimum attractive rate of return (% y-1) 4.42

Debt reserve ($) 0

The debt ratio was set to zero (all of the initial capital expenses were assumed to be paid for with equity) to remove confusion about the effect of debt interest on the project. If a project were found to be merited under conditions of zero debt it would be merited for any debt ratio, therefore this was a conservative measure. The MARR represented the expected return on the equity invested. Firms typically choose different MARR depending on the rate they expect their money to earn in alternative investments. However, if the MARR were set to the after‐tax interest rate on debt, varying the debt

167

ratio would not affect the NPW. Therefore for this analysis, the MARR was set equal to the after‐tax interest rate as follows:

( )AdAT Trr −= 1 (98)

where rAT = after‐tax interest rate, rd = interest rate on debt, and TA = adjusted tax rate.

The interest rate on debt (rd) used for the model was the prime loan interest rate reported by the Federal Reserve Bank as of November, 2007. At the time of this publication, the prime rate was decreasing. Using the after-tax interest rate on debt could also be justified by arguing that a project which earned this minimum return on investment would allow a firm to pay interest on debt borrowing for other projects. Furthermore, the real return on investment (the IRR) was reported when possible so that any MARR could be compared against the reported IRR.

Figure 77. Effect of Setting the Minimum Attractive Rate of return (MARR) Equal to the After-Tax Interest Rate on Debt. Data Generated By the Present Model

0 20 40 60 80 100

Net

Pre

sent

Wor

th

Debt Ratio (%)

MARR < After-Tax Interest RateMARR = After-Tax Interest RateMARR > After-Tax Interest Rate

4.3.3 Validation of the Present Model by Comparing Results with Previously Published Studies The model described in this study required testing and validation before using it to make predictions. In the absence of available real data, an effort was made to populate the model with the same assumptions used in a previously published economic analysis of the same system and compare the results. As opposed to the Cash Flow model used here, the previously published study used a Revenue Requirements model for calculating the COE required to generate enough revenues to provide a specified rate of return [52, 71]. The Revenue Requirements model assumed that revenues could vary from year to year to offset the annual costs, including taxes which were computed using the following equation for income tax [71]:

168

( )ondepreciatiROIprincipalA

AI PPP

TTT −+−

=1 (99)

where TA = adjusted tax rate, Pprincipal = debt principal payment (based on the debt interest rate and amount of unpaid principal), PROI = annual payment on the equity investment (based on the expected rate of return on investment), and Pdepreciation = annual depreciation of capital.

The net revenues were discounted and the resulting NPW of revenues was adjusted by the following inflation adjustment factor to generate the constant dollar levelized COE:

( )( )

( )( )Lmm

L

e

m

Lm

L

e

m

e

m

rrrr

rrr

rr

K

+⎟⎟

⎠

⎞

⎜⎜

⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛++

−+⎟⎟⎠

⎞⎜⎜⎝

⎛++

⎟⎟⎠

⎞⎜⎜⎝

⎛−

++

=

1111

1111

111

(100)

where K = inflation adjustment factor, rm = minimum attractive rate of return, re = general escalation rate, L=project lifetime.

The model created for this research incorporated a complete mass balance calculation to predict solid residues. It was found that the previous model significantly over‐estimated the residual solids production (Table 25). Assuming 29 percent TS in the feed, 82 percent of which were volatile, 82 percent biodegradability and 80 percent recovery of undigested solids at 65 percent MC, 9,115 t of residual solids were available for sale. This was less than half of the previously reported amount [52]. Therefore, to compare the results of the two models, the price of residual solids was increased so that the revenues from residual solids would be the same in both models.

The discount rate used in the previous study was the average of the MARR on equity and the debt interest rate, weighted by the debt ratio. This was most likely selected because it was the rate commonly used by the utilities industry when using revenue requirements to compute the levelized COE. However, choosing a discount rate other than the MARR on equity would not result in a COE that caused the NPW of the Cash Flow model to equal zero. Therefore, the Revenue Requirements model used in the previous study was obtained from the author and the discount rate was set equal to the MARR on equity. The constant dollar levelized COE dropped from the published value of $0.080 kWh‐1 to $0.063 kWh‐1, which was equal to the value determined by the current model (Table 25).

169

Table 25. Comparison of the Current Model with Previously Published Results of the APS Digester System

Previous Study Current Model

Assumptions

Feedstock Throughput (t yr-1) 69,160 69,160

Digester Volume (m3) NC 10,987

Methane Production (m3 y-1) 3,990,532 3,990,532

Compost Production (t y-1) 29,473 14,005

Net Electricity Production (kWh y-1) 10,961,583 10,961,583

Debt Ratio (%) 85 85

Escalation Rate (% yr-1) 2.1 2.1

MARR (% yr-1) 15 15

Debt Interest Rate (% yr-1) 6 6

Costs

Total Capital Cost ($) 16,908,278 16,908,278

Total O&M Expenses ($ yr-1) 1,307,776 1,307,776

Revenues

Electricity ($ yr-1) revenue required

661,898

Compost ($ yr-1) 330,274 330,274

Tipping Fees ($ yr-1) 1,372,169 1,372,134

Results

Constant Dollar Levelized Annual COE ($ yr-

1) 0.080 (published [52])

0.063

Revised COE ($ kWh-1) with Discount Rate = MARR

0.063 0.063

170

4.4 Results and Conclusions of the Economic and Financial Analysis

4.4.1 Results of the Mass and Energy Balance Calculations

All three scenarios used the same assumptions on the OLR and throughput of mixed food and green waste with the same physical characteristics. These assumptions resulted in a digester with an overall working volume of 3,418 m3 and a total digester volume of 3,931 m3 split evenly between the five reactors, giving each reactor a volume of 786 m3 (Table 26). Assuming cylindrical reactors with an aspect ratio (height : diameter) of 0.5, the exposed surface area of each reactor (including roof and floor) was 499 m2. The buffer tank was set to 10 percent of the volume of the biogasification reactor. The STP‐adjusted biogas yield and production rate were calculated to be 527 m3 t‐1 VS and 4.21 m3 m‐3 d‐1, respectively. The STP‐adjusted methane yield and production rate were calculated to be 271 m3 t‐1 VS and 2.17 m3 m‐3 d‐1, respectively. Adjusted for the VS content of the feedstock, the biogas and methane yields became 125 and 64 m3 t‐1 wet waste. The gross energy content was 29,868 MWh y‐1 and the energy yield was 712 kWh t‐1 wet‐waste treated, which represented a pre‐conversion power potential of 3.4 MW.

Table 26. Overall Digester Performance Results

Parameter Value

Digester Size

Wet Feedstock Throughput (t y-1) 41,975

Total Working Volume (m3) 3,418

Total Digester Volume (m3) 3,931

Hydrolysis Reactor Volume (m3) 786

Biogasification Reactor Volume (m3) 786

Buffer Tank Volume (m3) 79

Biogas Production

Biogas Produced (kmol y-1) 234,692

Methane Produced (kmol y-1) 120,815

Carbon Dioxide Produced (kmol y-1) 77,243

Hydrogen Sulfide Produced (kmol y-1) 3,990

Biogas Yield at STP (m3 t-1 VS) 633

Methane Yield at STP (m3 t-1 VS) 271

Residuals

171

Parameter Value

Dry Solids Recovered (t y-1) 3,190

Wet Solids Recovered (t y-1) 9,145

Wastewater Produced (m3 y-1) b 24,015

Water Vapor Produced (m3 y-1) 658

Energy

Energy Content of Biogas (kWh y-1) 29,868,276

Biogas Energy Yield (kWh t-1) 712

Feed Heat Required (kWh y-1) c 1,553,540

Heat Losses (kWh y-1) 453,802

Digester Electrical Requirement (kWh y-1) 740,889

a Biogas mass was estimated from the mass balance of solids and liquids.

b Effluent had a solids content of 3 percent, but the assumed specific volume was that of liquid water (1 m3 t‐1).

c Feed heat was the amount of heat required to bring the feed from the ambient temperature (16°C) to the maintenance temperature (55°C).

Results of the Mass Balance Calculations

The mass balance for the APS Digester system was the same for all three scenarios analyzed when the boundary did not include the biogas energy conversion unit. The assumed throughput of 115 t d‐1 and 100 percent capacity factor resulted in an annual wet‐waste throughput of 41,975 t y‐1. Assuming the density of the waste was approximately that of water, the volumetric waste throughput rate was 41,975 m3 y‐1 (115 m3 d‐1). At the assumed feedstock MC of 71 percent, this represented 82 m3 d‐1 of water added which resulted in an overall hydraulic retention time of 30 d. Reducing the HRT would require diluting the feed stream with water. At the assumed biodegradability 8,185 t y‐1 of VS would be converted to biogas, leaving 1,797 t y‐1 of undigested VS for the residual stream. This along with 2,191 t y‐1 of recalcitrant food solids (FS) resulted in a TS reduction of 67 percent. Based on the biodegradability the residual stream was expected to contain 3,988 t y‐1 of dry solids, of which 3,190 t y‐1 would be recovered for compost production and the rest would enter the wastewater stream. At the assumed biogas yield potential 4.8 million m3 y‐1 (545 m3 h‐1) of biogas

172

would be available at ambient temperature (after accounting for volume reduction due to water vapor condensation). The biogas in the reactor headspace was predicted to consist of methane (51 percent), carbon dioxide (33 percent), water vapor (16 percent), and hydrogen sulfide (500 ppm) (Table 27). Based on the water content of the biogas, it was estimated that 660 m3 y‐1 (1.8 m3 d‐1) of liquid water equivalent would exit the digester via the biogas stream, leaving 29,142 m3 y‐1 of liquid water for the residual stream. Based on the MC of the recovered press cake, 23,217 m3 y‐1 of liquid water would be present in the wastewater stream along with 798 t y‐1 of dry TS, resulting in 24,015 m3 y‐1 (66 m3 d‐1) of wastewater with a TS content of 3 percent. Given that the sum of the maintenance cost and the amortized capital cost of the wastewater treatment system (amortized over the project lifetime at the after‐tax loan interest rate of 4 percent) was $3,191 y‐1, the cost of treating the wastewater was $0.13 m‐3.

The balance between the mass of feedstock and residuals was attributed to biogas produced, resulting in the production of 8,845 t y‐1 of biogas wet mass, which was 21 percent of the mass of the wet feedstock. As the biogas temperature equilibrated with the ambient air temperature, the water vapor was calculated to condense to 595 m3 y‐1 (1.6 m3 d‐1) of liquid water. The biogas produced was calculated to consist of 1,933 t y‐1 (5.3 t d‐1) of methane and 3,399 t y‐1 (9.3 t d‐1) of carbon dioxide. In addition, 3.99 t y‐1 (0.011 t d‐1) of hydrogen sulfide and 2,852 t y‐1 (7.8 t d‐1) of trace contaminants would be present in the biogas.

Figure 78. Overall Mass and Water Balance for the APS Digester System. Percentages Given as a Fraction of the Mass of Feed Input to the Digester May Not Sum to 100 Percent

Due to Rounding

APS Digester system

Feed

41,975 t y-1

(71.0% H2O)

Biogas

21.1%

(4.6% CH4)

(8.1% CO2)

(1.6% H2O)

Recoverable Solids

21.7%

(14.1% H2O)

Wastewater

57.2%

(55.3% H2O)

173

This suggested that the mass of biogas calculated via the mass balance did not match the mass calculated via the volumetric production rate. Because the volumetric biogas yield and the mass balance yield are related via the biodegradability, it is possible to find a biodegradability for which the mass balance and volumetric yield result in the same mass of biogas. In laboratory‐scale studies, mass of residual solids can be difficult to measure accurately, especially for continuously fed digesters. Therefore, the biogas yield is often more accurate than the yield potential which depends on the VS reduction. Thus, the biogas yield was fixed at the value used in the current analysis by making the biogas yield potential a function of biodegradability as follows:

BYYp = (101)

where Yp = biogas yield potential (yield per unit VS destroyed), Ῡ = fixed biogas yield, B = biodegradability (VS destroyed per unit VS fed).

Both YP and the mass of trace gases were plotted as a function of B and it was found that biodegradability would have to approach 50 percent with a yield potential of 1,000 m3 t‐1 VS destroyed for the mass of trace gases to approach zero. However, reducing the biodegradability from the assumed 82 percent to 70 percent would only increase the yield potential by 17 percent from the assumed 642 m3 t‐1 VS while reducing the mass attributed to trace gases by 42 percent. The larger quantities of materials processed at the pilot‐scale APS Digester facility should improve the accuracy of the yield potential and biodegradability measurements. In addition, trace contaminants in the biogas should be carefully measured to improve the accuracy of the volumetric biogas mass calculation.

Figure 79. Mass of Trace Gases Calculated Based on the Mass Balance over a Range of Biodegradabilities, with the Biogas Yield Fixed at 526.7 m3 t-1 VS Fed

0

500

1,000

1,500

2,000

2,500

3,000

-4,000

-3,000

-2,000

-1,000

0

1,000

2,000

3,000

4,000

0 20 40 60 80 100

Yie

ld P

oten

tial (

m3

t-1V

S de

stro

yed)

Mas

s of T

race

Gas

ses (

t y-1

)

Biodegradability (% VS Reduction)

Mass of Trace Gasses Biogas Yield Potential

174

As the biogas left the digester and passed through the hydrogen sulfide scrubber, the composition was assumed to change as water vapor and hydrogen sulfide were removed (Table 27). Most of the hydrogen sulfide gas (3.8 t y‐1) was assumed to be captured by the hydrogen sulfide removal unit. The ultimate fates of methane and carbon dioxide were different depending on the scenario (Table 28).

Table 27. Compositional Breakdown of Biogas a Several Points in the Digester Process In The

Headspace After Cooling

After H2S Scrubbing

Methane (%) 51 60 60

Carbon Dioxide (%) 33 38 38

Water Vapor (%) 16 2 2

Hydrogen Sulfide (ppm)

500 582 29

Burning all of the biogas in the CG scenario resulted in 8,715 t y‐1 of carbon dioxide emissions from the electrical generator. This included 3,399 t y‐1 (39 percent) of carbon dioxide that evolved during anaerobic digestion and 5,316 t y‐1 (61 percent) from combustion of methane. Based on the energy balance calculation, this resulted in a carbon dioxide emission factor of 292 kg CO2 MWh‐1 of biogas and 1,060 kg CO2 MWh‐1 electricity (which did not include parasitic electricity demand). All of the CO2 in this scenario was derived from renewable resources. For the CNG producing scenarios, the carbon dioxide emissions were lower because the majority of the methane in the biogas was captured (Table 28). However, 117 and 101 t y‐1 of methane was included in the PSA emissions for the CNG and CG+CNG scenarios, respectively, due to inefficiencies in the biogas cleaning process. Because methane is 21 times more potent than carbon dioxide as a greenhouse gas, this quantity of methane was calculated to be equivalent to 2,451 and 2,129 t y‐1 of anthropogenic carbon dioxide emissions. The carbon dioxide offset due to the diversion of waste from landfill was the same for the three scenarios, but the carbon dioxide offset due to displaced energy was higher for the CNG producing scenarios (Table 28).

175

Table 28. Fate of Biogas, Methane, and Carbon Dioxide Generated by the APS Digester System, and the Overall Greenhouse Gas Balance for the System

CG CNG CG+CNG

Biogas Diversion

Electrical Fraction (%) 100 0 23

Heat Fraction (%) 0 11 0

CNG Fraction (%) 0 89 77

Ultimate Fate of Methane and Carbon Dioxide

Mass of CO2 Emitted (t y-1) 8,715 3,798 4,444

By the Generator 8,715 0 1,994

By the Boiler 0 976 0

By the PSA Unit 0 2,822 2,451

Mass of CH4 Emitted (t y-1) 0 117 101

Anthropogenic GHG Emissions Balance

GHG Emissions (t y-1 CO2 equiv.) 0 2,694 2,129

Due to Methane Emissions 0 2,451 2,129

Due to Electricity Purchased 0 243 0

GHG Offset (t y-1 CO2 equiv.) 15,243 18,835 18,246

Electricity Displacement 888 0 0

Natural Gas Displacement 0 4,479 3,890

Waste Disposal Reduction 14,355 14,355 14,355

Net GHG Emissions (t y-1 CO2 equiv.) -15,243 -16,384 -16,117

Results of the Energy Balance Calculations and Differences Between Scenarios

The biogas produced in a year would contain 29,868 MWh of energy, which would equate to a biogas energy yield of 712 kWh t‐1 of wet food and green waste. This varied depending on the feedstock characteristics and digestibility. For example, the green waste characteristics would result in a specific energy yield of 646 kWh t‐1, while those of food waste would result in a specific energy yield of 963 kWh t‐1. The digester required 2,007 MWh of heat annually, 77 percent due to feed heating and 23 percent due to maintenance heat. The effect of the mean annual ambient temperature on the total heat requirement is shown in Error! Reference source not found..

176

Figure 80. Effect of Mean Annual Ambient Temperature on the Amount of Heat Required Annually By the APS Digester System

0

500

1,000

1,500

2,000

2,500

3,000

0 5 10 15 20 25 30 35 40

Hea

t Req

uire

men

t (M

Wh

y-1)

Mean Annual Ambient Temperature (°C)

Feed HeatMaintenance HeatTotal Digester Heat Req

For the CG scenario, with all of the biogas converted to electricity, 72 percent of the available energy would be captured: 30 percent as electricity and 40 percent as heat. The remainder would be lost as waste heat. However, only 16 percent of the usable heat would be required by the digester. The remaining 10,537 MWh y‐1 of usable heat could either be used for other processes or vented. Of the 8,960 MWh y‐1 of electricity produced, 8 percent would be used parasitically to operate the digester. The remaining 8,220 MWh y‐1 could be sold. This would be equivalent to a net electricity yield of 196 kWh t‐1 of waste treated.

Table 29. Energy Balance for Each Scenario (Percentages Calculated Per Total Energy Available in the Biogas)

CG CNG CG+CNG

Energy Content of Biogas (kWh y-1) 29,868,276

29,868,276 29,868,276

Electricity Produced (%) 30.0 0.0 6.9

Heat Produced (%) 70.0 11.2 16.0

Usable (%) 42.0 6.7 9.6

Waste (%) 28.0 4.5 6.4

Methane Lost During Biogas Upgrade (%)

0.0 6.0 5.2

177

CG CNG CG+CNG

Natural Gas Produced (%) 0.0 82.8 71.9

Volume of Natural Gas (m3 y-1) at STP

0.0 2,339,989

2,032,286

Total Energy Balance (%) 100.0 100.0 100.0

Electricity Demand (%) 2.5 7.5* 6.9

Digester (%) 2.5 2.5 2.5

PSA (%) 0.0 5.0 4.4

Heat Demand(%) 6.7 6.7 6.7

* Electricity provided by the grid, not by converting biogas.

The CNG scenario would consume 7 percent of the energy content of the biogas to provide the digester’s heat requirement. No excess usable heat would be produced. An additional 740,889 kWh y‐1 would have to be purchased from the utility to provide the electrical demand for the digester and PSA unit. Inefficiencies in the PSA unit would result in the loss of 6 percent of the energy content, resulting in the capture of 83 percent of the energy content as CNG. The PSA unit was calculated to require twice as much electricity as the digester in this scenario. If the boiler were able to capture 80 percent of the energy content of burned biogas instead of the 60 percent assumed here, the amount of CNG produced would increase by 3 percent (Figure 81).

Figure 81. Effect on the Amount of CNG Produced of Changing the Boiler Recovery Efficiency (for the CNG Scenario)

23,000

23,500

24,000

24,500

25,000

25,500

26,000

30 40 50 60 70 80 90

CN

G P

rodu

ced

(MW

h y-1

)

Boiler Heat Recovery Factor (%)

178

The results of the CG+CNG scenario demonstrated that adding a generator sized to provide the parasitic electrical requirement would reduce the amount of CNG produced by 13 percent while producing 863 MWh y‐1 of heat in excess of the digester’s requirement. The generator would have to have a capacity of 260 kW if it were operated at 90 percent capacity. If the fraction of heat recovered from the generator remained above 45 percent, the amount of CNG produced would be maximized for the scenario. At lower heat recovery factors, some of the biogas would have to be burned in the boiler to provide the required heat, reducing the amount of biogas available for upgrading to CNG.

Figure 82. Effect of the Fraction of Heat Captured from the Generator on the Amount of CNG and Heat Produced (for the CG+CNG Scenario)

19,80020,00020,20020,40020,60020,80021,00021,20021,40021,600

0

200

400

600

800

1,000

1,200

20 25 30 35 40 45 50 55

CN

G P

rodu

ced

(MW

h y-1

)

Boi

ler H

eat P

rodu

ced

(MW

h y-1

)

Generator Heat Recovery Factor (%)

Boiler Heat Captured CNG Produced

4.4.2 Results of the Financial Analysis Overall, all three projects would be financially feasible (NPW > 0) given the base assumptions (Table 30). The CG and CG+CNG scenarios exhibited similar financial feasibility. The lower performance of the CNG scenario compared with the CG+CNG scenario suggested that adding a generator to provide the system’s electrical demand would increase the overall financial performance of the APS Digester configured to produce natural gas, which was confirmed by the incremental IRR analysis. The NPW for the APS Digester configured to producing natural gas was also shown to be optimized when the generator provided just enough electricity to provide the system’s demand.

Table 30. Comparison of Overall Financial Performance of the Three Scenarios Analyzed (for a Discount Rate of 4.42 Percent Y-1)

179

Scenario CG CNG CG+CNG

Net Preset Worth ($) 5,381,642 4,236,645 5,264,947

Discounted Payback Period (y) 9 10 9

Levelized Cost of Waste Treatment ($ t-1) 11.20 13.98 11.48

* The cost of treatment calculation assumed energy and compost were sold at their market values ($0.067 kWh‐1 for electricity, $0.0247 kWh‐1 for natural gas, and $11 t‐1 for compost).

The estimated capital costs for the three scenarios were similar, with the cost of the APS Digester (including the reactors and material handling systems) making up 70 percent of the total capital cost (

). The cost of a generator sized for converting the total biogas output was slightly higher than the cost of the biogas upgrade unit ($1.3 vs. $1.0 million). This caused the capital cost of the CNG scenario to be 4 percent lower than that of the CG scenario. The addition of a 250 kW generator brought the total capital cost of the CNG+CG scenario to within 1 percent of the CG scenario. Preconstruction costs made up 5 percent of the total capital cost for all three scenarios.

Figure 83. Capital Cost Breakdown for the Three Scenarios

0

1

2

3

4

5

6

7

8

9

CG CNG CG + CNG

Cap

ital C

ost (

106

$)

10% Contingency

Boiler

Cogeneration System

Biogas Cleaning System

Digester

Preconstruction Costs

Unlike the capital costs, the ordinary expenses for the CNG scenario were 30 percent higher than the other two scenarios largely due to the cost of purchasing electricity ($235,656 y‐1) which made up 29 percent of the ordinary expenses (Figure 84). The

180

equipment maintenance costs were 5‐15 percent lower for the CNG producing scenarios because the generator was responsible for $100,446 y‐1 in the CG scenario while the PSA unit incurred $50,000 y‐1 in maintenance costs. Overall, the equipment maintenance costs made up 47 percent, 30 percent, and 45 percent of the ordinary expenses for the CG, CNG and CG+CNG scenarios, respectively. The digester was a portion of the equipment maintenance cost and incurred $161,945 y‐1 which made up 26, 20, and 27 percent of the ordinary expenses. Labor incurred an additional $158,700 y‐1 which made up 25, 19, and 26 percent of the ordinary expenses.

Figure 84. Ordinary Expenses Breakdown for the Three Scenarios

0

100

200

300

400

500

600

700

800

900

CG CNG CG + CNG

Ord

inar

y Ex

pens

es (1

03$

y-1)

5% Contingency

Property Maintenance

Administration

Utilities

Equipment Maintenance

Labor

The revenues for the three streams were similar with the CNG scenario earning 4‐5 percent more than the other two scenarios (Figure 85). Energy revenue for the CG scenario came only from electricity, which resulted in the production of 33‐38 percent as much energy as the CNG producing scenarios, but the assumed wholesale price (per kWh) of CNG was 37 percent lower than the price of electricity. The choice of energy type had no effect on the revenues from tipping‐fees and residual‐solids sales. The annual net revenue (annual revenue minus annual expenses) for the CG, CG+CNG, and CNG scenarios was $1,044,926, $912,141, and $1,043,631, respectively.

181

Figure 85. Revenue breakdown for the Three Scenarios (Not Including Tax Credits)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

CG CNG CG + CNG

Rev

enue

s (10

6$

y-1)

Tipping Fee Income

Compost Income

Energy Income

In all three scenarios the cash flow pattern represented a simple investment cash flow, with only one sign change occurring in year 1 (Figure 86). Thus, the IRR calculation would result in only one solution which would represent the true return on investment. The year zero cash flow was equal to the capital cost because the assumed debt ratio was zero. The remaining cash flows reflected the difference between the revenues and expenses for each scenario. The 7‐year MACRS depreciation schedule caused the cash flows to be higher and more variable in years 1 – 8, followed by steadily increasing cash flows after year 8 due to escalation.

Figure 86. 20-Year Cash Flows for the Three Scenarios Analyzed

(9)

(8)

(7)

(6)

(5)

(4)

(3)

(2)

(1)

0

1

2

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Aft

er-T

ax C

ash

Flow

(106

$)

Time (y)

CG CNG CG + CNG

182

Results of the Revenue Requirements (COE and COT) Analyses

AD is of interest in both the energy sector and the waste treatment sector. The COE analysis treated the project as an energy production project credited with revenue earned from non‐energy sources (tipping fees and compost), while the COT analysis treated the project as a waste treatment project credited with revenue from the products. The revenue requirements method allows decision makers to compare the current‐dollar expected price of a good with the levelized cost of providing that good. For AD, the primary goods of interest are energy and waste treatment. When the good provided is energy, the form energy must be considered and standardized (for example to gallons gasoline equivalent) if different forms of energy are compared. Since each scenario only produced one form of energy (electricity for the CG scenario, natural gas for the CNG and CG+CNG scenarios), no standardization was used. This allowed for comparison with prices of electricity and natural gas, but precluded comparison of the scenarios. Conversely, the COT could be used to compare the scenarios.

The levelized COE for the CG scenario was close to zero ($0.00035 kWh-1) because revenue provided by the $24.25 t-1 tipping fee provided enough revenue to cover the costs and provide the 4.42 percent MARR. However, most analysts would expect a higher MARR. With the MARR set to 15 percent the COE was $0.10 kWh-1 (in 2007 dollars), which was higher than the average 2007 wholesale price of electricity in California ($0.067 kWh-1) but only slightly higher than the 20-yr 1998 Market Price Referent ($0.09572 kWh-1) for the new CPUC feed-in tariffs for renewable-electricity producers. Furthermore, if 25 percent of the capital cost was paid for with a 20 yr loan (7.5 percent market interest rate), the COE fell to the assumed wholesale electricity price. If 75 percent of capital cost were paid with the same loan terms, the COE became negative, indicating that under these conditions electricity revenues were not necessary. Reducing the loan term to 10 years insulated the COE from the effect of changing the debt ratio. With a 10yr loan paying for 75 percent of the capital, the COE (15 percent MARR) was less than the current price of electricity.

The levelized COE of the CNG and CG+CNG scenarios was $0.0072 and ‐$0.0003 kWh‐1, respectively. A negative COE for the CG+CNG scenario indicates that no revenue was required from the natural gas sold to provide the 4.42 percent MARR. With the MARR set to 15 percent, the COE was $0.0386 and $0.0377 kWh‐1, which was higher than the 2007 average City Gate price of natural gas of $0.0247 kWh‐1. Changing the debt ratio and the loan term had an effect on the COE of the natural gas scenarios that was similar to the effect on the CG scenario (Error! Reference source not found.).

183

Figure 87. Effect of Financing on Levelized Cost of Energy ($ kWh-1) for the CG Scenario, Where Energy = Electricity (HHV, 30 Percent Efficiency), and the CNG and CG+CNG

Scenarios, Where Energy = Natural Gas (HHV)

0.000

0.020

0.040

0.060

0.080

0.100

0.120

CG CNG CG+CNG

Cos

t of

Ene

rgy

($ k

Wh-

1 )

Assumes MARR = 15%

No Loan - Zero Debt20 yr Loan - 25% Debt20 yr Loan - 75% Debt10 yr Loan - 25% Debt10 yr Loan - 75% Debt

The level annual tipping fee revenue requirement for the CG scenario was $470,105 y‐1 assuming that electricity and residual solids were sold at the market rates. This revenue requirement was then standardized per unit wet mass of OFMSW treated annually to calculate the levelized COT. For the CNG and CG+CNG scenarios, the level annual tipping fee revenue requirement was $586,660 and $ 481,984 y‐1, respectively, assuming natural gas and residual solids were sold at the market rates. The resulting COT was similar for the CG and CG+CNG scenarios regardless of the financing and MARR used. The COT ranged from $10 ‐ $30 t‐1. Although the tipping fee assumed for this study was $24 t‐1, the 2004 average landfill tipping fee was $38 t‐1 [53]. In this study, the assumed tipping fee was lower than landfill tipping fees because the OFMSW was assumed to have been sorted prior to delivery at the APS Digester facility. The true cost of sorting the OFMSW is unknown, however, and merits further study.

184

Figure 88. Effect of Financing a 15 Percent MARR on the Levelized Cost of Treatment ($ t-1 Wet Waste), with the Base Case (Zero Debt, 4.42 Percent MARR) Included for Comparison

0

5

10

15

20

25

30

35

CG CNG CG+CNG

Cos

t of T

reat

men

t ($

t-1)

Zero Debt - No Loan - 4.42% MARRZero Debt - No Loan - 15% MARR25% Debt - 10 yr Loan - 15% MARR25% Debt - 20 yr Loan - 15% MARR75% Debt - 10 yr Loan - 15% MARR75% Debt - 20 yr Loan - 15% MARR

Results of the Incremental IRR Analysis

The lowest costing scenario (CNG) had an IRR of 10.9 percent (Table 31). The incremental IRRs between the CNG and the next most costly scenarios (CG and CG+CNG, in ascending order) were 37.5 percent and 27.7 percent y‐1, respectively. CG scenario had an IRR of 12.1 percent y‐1. The increase in expenditure between the CG and CG+CNG scenario did not generate enough additional revenue to justify the expense. Therefore, the incremental analysis reveals that adding a PSA unit to produce natural gas is not fiscally advisable, given the assumed values for capital cost, expenses, and revenues. However, this result is specific to the assumptions used in this analysis. Similarly, the MARR must be less than 12.1 percent y‐1 assuming none of the other assumptions (for example debt ratio) change.

Table 31. Results of the Incremental IRR Analysis ( Percent y-1) Capital Cost (106 $) CNG CG CG+CNG

CNG 7.59 10.9*

CG 7.88 37.5 12.1*

CG+CNG 7.96 27.7 NC** 11.9*

* Values on the diagonal are the IRRs calculated for the indicated scenario. Off‐diagonal values are incremental IRRs.

** NC = not calculable.

185

Results of the NPW Analysis

The NPW of the CG scenario was 2 percent higher than that of the CG+CNG scenario and 27 percent higher than that of the CNG scenario (Table 26). As in the incremental IRR analysis, the CG scenario would be the preferred project, given the current assumptions. However, the assumptions contain uncertainties that could easily result in a different financial decision because the difference between the NPWs of the CG and CG+CNG scenarios is small. Therefore, the current study suggested that the financial viability of a project that converts biogas to CNG is similar to that of one which converts biogas to electricity. However, if the price of electricity were to increase, the CG scenario would prove to be the more financially sound choice. For example, using the CPUC feed‐in tariff Market Price Referent ($0.09572) as the price of electricity, the NPW increased by 43 percent to $7,700,682.

The difference between the NPWs of the CNG and CG+CNG scenarios was due primarily offsetting the system’s electricity usage. Therefore, the benefit of adding a generator to the CNG producing system would be expected to decline as the retail price of electricity fell. The NPW of the CNG scenario was found to equal that of the CG+CNG scenario when the retail price of electricity was reduced from the market rate of $0.10 kWh‐1 to $0.059 kWh‐1 (assuming all other things remained equal). However, if the retail price of electricity were to decline, the price of natural gas would also be likely to decline which could preserve the relative NPW of the two scenarios.

The base case revenues included the Renewable Electricity Producer (REP) tax credit for the CG scenario and the Non‐Conventional Fuel (NCF) tax credit for the CNG and CG+CNG scenarios. The REP credit added $78,086 y‐1 to the revenue stream of the CG scenario. Removing this credit would only reduce the NPW by 4 percent (Table 32). The NCF credit added $68,719 y‐1 and $59,683 y‐1 to the CNG and CG+CNG scenarios, respectively. Removing this credit would only reduce the NPWs by 5 percent and 3 percent. Thus, the tax credits were net essential for the viability of the projects. Furthermore, if the CNG were used as a vehicle fuel, the Alternative Vehicle Fuel (AVF) tax credit could be claimed which would add $337,921 y‐1 and $293,485 y‐1 to the CNG and CG+CNG scenarios, respectively. This would increase the NPWs of the natural gas producing scenarios by 15 percent and 11 percent, making natural gas financially preferable to electricity when using the natural gas as a transportation fuel. This assumes, of course, that no additional expenses would be incurred for injecting natural gas into CNG‐fueled vehicles.

186

Table 32. Effect of the Renewable Electricity Producer (REP) Credit, the Nonconventional Fuel (NCF) Credit, and the Alternative Vehicle Fuel (AVF) Credit on the Net PresentWorth

(NPW) and Cost of Waste Treatment (COT) Scenario Tax Credit ($ kWh-1) NPW ($) Change from

Assumed Value (%)

COT ($ t-1)

CG None 5,163,423 -4.2 11.73

REP (0.010) 5,381,642 0.0 11.20

CNG None 4,044,602 -4.7 14.44

NCF (0.003) 4,236,645 0.0 13.98

AVF (0.014) 4,988,956 15.1 12.15

CG+CNG None 5,098,157 -3.3 11.89

NCF (0.003) 5,264,947 0.0 11.48

AVF (0.014) 5,918,330 11.0 9.90

Maximization of NPW for the CG+CNG Scenario

The biogas distribution was optimized to maximize NPW for the APS Digester configured with an engine generator and PSA unit operating in different combinations by varying the fraction of biogas diverted away from the PSA unit for heat and electricity production by 0‐20 percent and 0‐30 percent, respectively. The resulting surface plot revealed that NPW was maximized when the biogas was distributed as assumed for the CG+CNG scenario (Figure 89). Diverting biogas to the boiler reduced the NPW because excess heat had no value. Converting less than 23 percent of the biogas to electricity required purchasing electricity from the utility company which reduced the NPW, even though additional biogas was available for sale as natural gas. Converting more than 23 percent of the biogas to electricity added revenue from electricity sales, but the generator size and cost increased. The PSA capital and maintenance cost did not scale with the amount of biogas treated. If it had, the maximization point could have been different, and future modeling efforts could include a cost function for the PSA unit to test the maximization point. However, PSA units are currently sold in fixed sizes and true cost functions are unavailable and unrealistic.

187

Figure 89. Maximization of NPW for the CNG Producing Case over a Range of Biogas Diversion Schemes. Biogas Not Diverted to the Generator o Boiler Was Converted to CNG.

0 10 20

4.0

4.2

4.4

4.6

4.8

5.0

5.2

0

10

20

30

Boiler (%)

Net

Pre

sent

Wor

th (1

06$)

Generator (%)

Converting the biogas to natural gas for storage and then converting it to electricity during peak hours may actually improve the NPW, depending on the peak price of electricity. In order to model this, the capital and operating cost of the PSA unit was included and the capacity factor of the generator was reduced by 66 percent. This brought the total capital cost to $11,299,042 and the ordinary expenses to $697,120. The peak price of electricity would have to be $0.1043 (54 percent higher than the base assumption) for the NPW to remain the same. Peak prices can be 50‐100 percent higher than daily the mean price. In addition, low and medium pressure storage options for biogas could be less expensive than PSA and should be investigated.

Effect of Co-Locating the APS Digester System with Other Industries

The financial performance of a digester that offset the electrical usage of another industry was compared to that of an independent energy producing digester that sold its electricity. The co‐located digester received a higher price for the electricity produced but incurred demand charges on the electrical capacity of the digester. No reduction of capital costs or operating expenses was considered. The NPW of the co‐located digester was 28 percent higher than the independent digester. Cost sharing would further improve the financial performance of a co‐located digester. However, for the comparison with CNG‐producing systems the digester was considered to be independent of any other power users.

188

Figure 90. Effect of C-Location (Demand Charge) Vs. Independent Location (No Demand Charge) of APS Digester on NPW (Assuming Equal Capital Costs). Retail = $0.0857 kWh-1,

Wholesale = $0.067 kWh-1, Feed-In = $0.09572 kWh-1

6.90

5.38

7.70

0

1

2

3

4

5

6

7

8

9

Co-Located (retail electricity price)

Independent (wholesale electricity price)

Independent (feed-in tariff electricity price)

Net

Pre

sent

Wor

th (1

06 $)

Figure 91. Effect of Co-Location (Demand Charge) Vs. Independent Location (No Demand Charge) of APS Digester on Revenues and Expenses (Assuming Equal Capital Costs).

Retail = $0.0857 kWh-1, Wholesale = $0.067 kWh-1, Feed-In = $0.09572 kWh-1

1.841.67

1.91

0.64 0.62 0.62

0.0

0.5

1.0

1.5

2.0

2.5

Co-Located (retail electricity price)

Independent (wholesale electricity price)

Independent (feed-in tariff electricity price)

Cas

h Fl

ow (1

06$

y-1)

Revenues Expenses

189

Sensitivity of the Model to Changes in the Assumed Values

The slope of the sensitivity curve indicated the magnitude of the sensitivity; therefore the slopes of the linear parameters were compared. The capacity factor used in the analysis was 100 percent, therefore only lower factors were tested.

Figure 92. Sensitivity of the CG Scenario’s Financial Performance to Changes in the Assumptions (Top) and a Comparison of the Slopes of the Linear Sensitivity Curves

(Bottom)

0

1

2

3

4

5

6

7

8

9

10

-50 -40 -30 -20 -10 0 10 20 30 40 50

Net

Pre

sent

Wor

th (1

06$)

Change from Base Value (% )

Tipping FeePrice of ElectricityBiogas Yield PotentialLab Methane ContentCapacity FactorLoading RateBiodegradability/VS ReductionEscalation of ElectricityPrice of Residual SolidsDigester Power RequirementFederal Tax RateMARROrdinary ExpensesCapital Cost

-100 -50 0 50 100 150 200

Capacity FactorTipping Fee

Price of ElectricityLab Methane ContentBiogas Yield Potential

Biodegradability/VS ReductionEscalation of ElectricityPrice of Residual Solids

Digester Power RequirementFederal Tax Rate

MARROrdinary Expenses

Capital Cost

Slope of Sensitivity Curve (103 $ %-1)

190

Figure 93. Sensitivity of the CNG Scenario’s Financial Performance to Changes in the Assumptions (Top) and a Comparison of the Slopes of the Linear Sensitivity Curves

(Bottom)

0123456789

10

-50 -40 -30 -20 -10 0 10 20 30 40 50

Net

Pre

sent

Wor

th (1

06$)

Change from Base Value (%)

Tipping Fee

Natural Gas Price

Biogas Yield

Loading Rate

Ambient Temp

PSA Maintenance Cost

PSA Power Rating

PSA Unit Capital Cost

Capital Cost

Ordinary Expenses

Capacity Factor

PSA Methane Recovery Factor

-100 -50 0 50 100 150 200


Natural Gas PricePSA Methane Recovery Factor

Biogas YieldAmbient Temp

PSA Maintenance CostPSA Power Rating

PSA Unit Capital CostCapital Cost

Ordinary Expenses


191

Figure 94. Sensitivity of the CG+CNG Scenario’s Financial Performance to Changes in the Assumptions (Top) and a Comparison of the Slopes of the Linear Sensitivity Curves

(Bottom)

0

1

2

3

4

5

6

7

8

9

10

-50 -40 -30 -20 -10 0 10 20 30 40 50

Net

Pre

sent

Wor

th (1

06$)

Change from Base Value (%)

Tipping FeeBiogas YieldNatural Gas PricePSA Methane Recovery FactorCNG EscalationAmbient TempCogen Capital CostCogen Maintenance CostPSA Maintenance CostPSA Power RatingPSA Unit Capital CostOrdinary ExpensesCapital CostCapacity Factor

-100 -50 0 50 100 150 200


Biogas YieldNatural Gas Price

PSA Methane Recovery FactorCNG EscalationAmbient Temp

Cogen Capital CostCogen Maintenance Cost

PSA Maintenance CostPSA Power Rating

PSA Unit Capital CostOrdinary Expenses

Capital Cost


192

Often, additional capital expenditure results in cost savings for the system. Since the capital cost is not discounted, NPW varies inversely proportionately with the capital cost. In other words, each dollar spent on capital reduces the NPW by a dollar. For the CG scenario, the slope of the NPW of as a function of the annual ordinary expenses was ‐$10 $‐1 y‐1 (Figure 95). Therefore, when facing a choice between two pieces of equipment, each extra dollar spent on the more expensive choice would have to result in at least ten cents per year worth of savings on annual expenses to be justifiable financially.

Figure 95. NPW of the CG Scenario as a Function of Annual Ordinary Expenses

y = -9.82x + 11,513,486.15

0

1

2

3

4

5

6

7

8

9

200 400 600 800 1000

Net

Pre

sent

Wor

th (1

06$)

Annual Ordinary Expenses (103 $ y-1)

In addition, the sensitivity of the CG scenario to the feedstock MC was investigated. The effect of changing the feedstock’s initial MC was unpredictable because it simultaneously affected the digester size, biogas production, compost production, wastewater treatment, and heat requirement (via the heat capacity). It should be noted that the model assumed a constant cost for the wastewater treatment equipment, and the only capital cost item to vary with reactor volume was the tankage cost. At lower MC, with more VS to treat, the capital cost was higher because the digester was larger, but the revenues were also higher due to the increased biogas and compost production. At higher MC, the capital cost was reduced but the revenues were also lower. The offset resulted in a linear reduction in NPW across the range of MC analyzed (

193

96). If the model calculated capital costs and wastewater treatment costs using a cost function, the results could be different. However, based on this analysis, if two feedstocks are available for the same tipping fee (feedstock transportation costs were assumed to be embedded in the tipping fee) the drier feedstock will be financially preferable. The financial effect of drying the feedstock would have to include the cost of the drying process.

Figure 96. NPW and IRR of the CG Scenario as a Function of the Feed MC (All Other Feed Characteristics Held Constant)

0

1

2

3

4

5

6

7

8

9

30 40 50 60 70 80 90

Net

Pre

sent

Wor

th (1

06$)

Feed Moisture Content (%)

Effect of Heat Sales and Heating Efficiency on Financial Performance

Adding heat sales to the CG and CG+CNG scenario increased the NPW by 58 percent and 5 percent, assuming 100 percent of the available useable heat were sold at a price of 0.0273 $ kWh‐1 (Error! Reference source not found.). The CNG scenario produced no excess usable heat. The effect of adding heat sales to the CG+CNG scenario was small because only the generator produced useable heat, 70 percent of which was used for heating the digester.

194

Figure 97. Effect on NPW of the CG and CG+CNG Scenario When Including Heat Sales at a Rate of $0.0273 kWh-1.

45566778899

0 20 40 60 80 100

Net

Pre

sent

Wor

th (1

06 $)

Fraction of Recoverable Heat Sold (%)

CG

CG+CNG

The mean annual ambient temperature had a negligible influence on the NPW of any of the scenarios analyzed, even when heat sales were included in the revenue stream (Figure 98). When the biogas was converted to electricity, the CG and CG+CNG scenarios produced 10.4 and 2.4 times more heat than the digester required, respectively, when the average annual ambient temperature was 16°C. The CNG scenario diverted just enough biogas to the boiler to satisfy the digester heat requirement, thus the amount of biogas converted to CNG varied linearly with the ambient temperature. The CG+CNG scenario varied inversely with the ambient temperature because the PSA unit required less electricity to compress the cooler biogas. However, even when the highest heat producing scenario was credited with income from heat sales, the NPW changed by less that 2 percent per 10 °C increase in mean annual ambient temperature.

195

Figure 98. Effect of Mean Annual Ambient Temperature on the NPW of All Three Scenarios, Including the CG Scenario with All Usable Heat Sold for $0.0273 kWh-1

-6

-4

-2

0

2

4

6

0 10 20 30 40 50

Rel

ativ

e Cha

nge i

n N

PW (

%)

Average Annual Ambient Temperature (°C)

CG

CNG

CG+CNG

CG w/ Heat Sales

Similarly the R‐value of the insulation (inverse of the thermal conduction) had little effect on the NPW of the digester, as long as the R‐value was above 0.35 m2 K W‐1 (99). Commercially available insulation R‐values typically range from 0.3 to 1.8 m2 K W‐1. The amount of heat required annually by the digester decreased by 71 percent when the R‐value changed from 0.176 to 3.52 m2 K W‐1, but the NPW of the CG, CNG, and CG+CNG scenarios increased by 0.2 percent, 43 percent, and 28 percent, respectively. The NPW of the CG scenario did not change as a function of the R‐value because extra heat had no value. However, when heat sales at $0.0273 kWh‐1 were included in the revenue stream, the NPW of the CG scenario increased by 18 percent.

196

Figure 99. Effect of the R-Value of the Insulation on the Digester Heat Requirement and NPW of All Three Scenarios and the CG Scenario When Heat Sales Were Included

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2

3

4

5

6

7

8

9

0 1 2 3 4

Ann

ual H

eat R

equi

rem

ent (

MW

h y-1

)

Net

Pre

sent

Wor

th (1

06 $ )

Insulation R-Value (m2·K·W-1)

CG

CG w/ Heat Sales

CNG

CG+CNG

Total Heat Requirement

4.4.3 Conclusions from the Model of the Financial and Engineering Performance of the APS Digester System

The calculated reactor volume, biogas and methane yield, and energy production were similar to those seen for anaerobic digestion systems treating equal quantities of OFMSW. For example, methane yields for full‐scale MSW digesters have been reported to range from 226 – 400 m3 t‐1 VS for source‐separated organics from MSW [11, 37, 38]. Therefore, the predicted yield for the APS Digester system (271 m3 t‐1 VS) could be seen to be relatively conservative. Similarly, the APS Digester system’s predicted biogas production rate (4.21 m3 m‐3 d‐1) was within the 3 – 5 m3 m‐3 d‐1 range reported for thermophilic high‐solids anaerobic digesters [43]. The overall mass balance predicted for the APS Digester system was also similar to the measured mass balances of several full‐scale digesters (Table 9). The carbon dioxide balance calculations revealed that the anaerobic digestion of organic wastes would reduce net GHG emissions as compared with landfilling the waste. The reduction of GHG could be enhanced by oxidizing the methane emitted from the PSA unit and venting or sequestering the CO2. However, this study did not take into account any emissions that would occur after digestate was removed, for example during post‐digestion compost aeration. LCA analyses of anaerobic digesters treating MSW have demonstrated that these emissions can contain large quantities of methane, therefore they should be included in a complete GHG emissions inventory [49]. This study also did not include GHG emissions due to plant construction or transportation of material to and from the facility, whereas the emissions calculation for landfilling food and green waste did. In the future a complete carbon balance should be calculated for the APS Digester system via LCA.

197

The carbon dioxide emission factor for biogas not derived from landfill gas or sludge digestion has been reported to range from 166 – 238 kg MWh‐1 [84]. However, the carbon dioxide emission factor for the APS Digester system was higher (292 kg MWh‐1) because biogenic carbon dioxide was included. For performing carbon dioxide inventories, the GHG emissions factors for digesters should either be revised to include biogenic carbon dioxide, or if the inventory includes only anthropogenic carbon dioxide the factor should be negative to reflect carbon dioxide offsets.

The model revealed that, under the current assumptions, the financial performance of an APS Digester system that converted biogas to natural gas would approximately equal that of a system that converted the biogas to electricity. The natural gas producing system could be combined with electricity generation to further enhance the NPW of the project. In the current analysis, the NPW was maximized when the generator produced just enough electricity to cover the system’s own demand. Changes in the prices of electricity and natural gas could change the maximization point. One way of effectively increasing the price of electricity would be to include biogas storage, but this would incur additional costs. The cost of adding a PSA unit to store electricity and increasing the size of the generator to produce electricity only during off‐peak hours was tested, and the price of electricity was found to increase 54 percent to keep the same NPW. Since peak electricity prices can be 50‐100 percent greater than the 24hr average price, this may be a viable alternative. It would also provide the option of selling natural gas if needed. However, if natural gas were not needed, the biogas could be stored at lower pressures which may be less expensive. The cost of adding medium and low‐pressure biogas storage should be investigated. Also, a higher NPW could be obtained by selling electricity at off‐peak and shoulder rates as well since the generator and biogas storage unit size could be reduced. In the future, a model could be used to optimize NPW or incremental IRR based on the above considerations.

Co‐locating the digester with another industry could further improve the financial performance. When the retail value of electricity was $0.1033 kWh‐1, the demand charges incurred in the co‐located situation did not offset price difference from the wholesale value of $0.067 kWh‐1. Therefore, the costs saved by not paying the utility’s demand charge favored independent location of the APS Digester system. However, when the wholesale electricity value was replaced with the market price referent for the CPUC feed‐in tariff of $0.0957 kWh‐1 (which is slightly lower than the time‐of‐delivery adjusted rate) the demand charge did offset the price difference. The improvement in financial performance due to cost sharing and offset of heat use was not investigated. Including these items could favor co‐location even if the higher tariffs were available. This implies that the feed‐in tariffs may not play a role in the development of MSW digestion if the digesters tend to be co‐located. However, negotiating co‐location deals could be difficult when the waste stream does not come from a single source. Furthermore, the feed‐in tariffs would not be available to systems much larger than the one modeled here, in which case co‐location would be favorable.

198

The scenarios analyzed assumed that excess available heat had no value, but many situations exist in which the heat could have high value. The CG scenario resulted in the largest amount of unused heat available, and selling 100 percent of the available heat increased the NPW by 60 percent which would have a significant impact on the feasibility of the project. The NPW of the CG+CNG scenario only increased by 4 percent with the addition of heat sales, since a smaller amount of usable heat was available for sale, but including such sales would provide an additional advantage to including a generator with the system. It would also provide an incentive for capturing heat from the PSA unit’s compressors. Without heat sales, the NPW of the CG+CNG scenario approached that of the CG scenario when the insulation had an R‐value >0.88. At lower R‐values, the NPW of the CG+CNG scenario fell while the CG scenario remained constant because the generator produced enough heat to keep the reactors at 55°C even though the heat requirement increased 300 percent. Even with uninsulated steel tanks (2cm thick, k = 46 W m‐2 °C‐1) the NPW of the CG scenario remained within 1 percent of the base‐case as long as the heat transfer coefficient were kept to less than 13 W m‐2 °C‐1, such as by using a wind break. Even when heat sales were included in the CG scenario, insulation’s thermal resistance had a larger relative effect on the NPW of the natural gas producing scenarios than the CG scenario. Assuming the reactors were well insulated, the average annual ambient temperature had a negligible effect on the NPW for all the scenarios, including the CG scenario when heat sales were included. Thus, the APS digester could be expected to be a financially viable option even in cold climates, as long as the capital cost accurately reflects the cost of the insulation. In the future, the model should be designed to calculate the cost of the insulation needed as a function of the assumed average annual ambient temperature.

The model could be used to test the effect of the heating efficiency on the value of any combination of base assumptions. In the current analysis, improving the boiler efficiency had a relatively small influence on the amount of CNG produced. The cost of improving the boiler efficiency was not investigated, but it can be inferred that under the current assumptions such investment may not be justified. In general, it was determined that each extra dollar spent on capital upgrades would have to result in >$0.10 y‐1 in cost savings to improve the NPW.

In general, the sensitivity analyses indicated that financial parameters such as capital cost, operating cost, tipping fees, electricity and natural gas prices, and MARR had a greater effect on the NPW than physical parameters such as biogas yield, biodegradability, and methane content. Even when the feed characteristics of food waste were compared with those of green waste, the NPW changed by half as much as the energy yield. This may in part be due to the fact that compost production decreased as energy yield increased, but it was mostly due to the large influence of tipping fees on the present worth. Tipping fees were assumed to be independent of the waste quality. Some of the financial parameters did not have a significant influence on NPW. Notably, the escalation rate of electricity and CNG price did not have a large effect on NPW, although the assumed 2007 base price of electricity and CNG did have a relatively large effect. In

199

addition, the PSA unit’s capital and operating cost did not have a large effect on NPW and there was uncertainty in these values due to the novelty of using PSA to purify biogas. However, the methane recovery efficiency of the PSA unity had a relatively large effect on NPW. Since the assumed methane recovery efficiency was over 90 percent, there would be little room for improvement and a large potential for the PSA unit to perform worse than expected. Also, when choosing biogas cleaning technologies, the performance of the technology may be more important than its cost.

The magnitude of the sensitivity to the MARR assumption was moderate, but the MARR used in the analysis was low. Therefore, the range of MARRs that could be applied to the project would be expected to be much larger than the uncertainty in prices and biological performance. Therefore, even though it appears to be a minor factor, it could have a much larger influence on the financial feasibility of an APS Digester project than the other factors tested in the sensitivity analysis. The IRR analysis revealed that, assuming all other factors remained the same, none of the scenarios would be financially feasible for MARR >12.1 percent. Therefore, this project may be more suitable for risk tolerant investors, unless the bottom line could be improved. For all three scenarios, reducing the capital cost by 20 percent or increasing the tipping fee by 30 percent caused the IRR to increase to >15 percent y‐1.

The capacity factor for the whole digester had the largest effect on NPW. Decreasing the capacity factor did not affect the digester capital or maintenance cost because the system was designed based on the assumed daily throughput rate. The capacity factor was then used to calculate the annual throughput which was used to calculate the biogas production, and by extension the energy production. Therefore, the energy conversion system costs as well as energy revenues decreased along with the capacity factor, as did revenues from compost and tipping fees. The APS Digester may be able to tolerate fluctuations in OLR, which would allow system managers to stockpile feed while waiting to make repairs. This could be an area for future research. The multiple reactors in the APS Digester system could also offer redundancy that would allow for repairs without shutting down the operation. However, if the loading equipment were to fail, operation would have to cease. The APS Digester should be managed well to reduce the time required to fix common problems, and regular maintenance should be scheduled. In reality, this may not be possible, especially in light of the lack of experience with the technology. Therefore, future modeling efforts should incorporate a lower assumed capacity factor with the understanding that the financial performance of full‐scale APS Digester systems will improve as the technology matures.

This analysis assumed that the CNG was injected into the natural gas grid, in which case the Nonconventional Fuel Tax Credit would be available. If the CNG were used as a transportation fuel, the Alternative Vehicle Fuel Tax Credit would increase the present worth of the CNG producing system beyond that of the electricity producing system, barring any additional expenses incurred. Tax credits did not add considerably to the NPW of the APS Digester, which suggests that the technology could be financially viable without governmental support. However, the MARR assumed was relatively low, and

200

the financial feasibility may be marginal if higher MARRs are required. The size of the tipping fee had a much greater influence on the financial feasibility of the system. AD systems in general should be expected to receive more revenue from tipping fees than energy sales. Therefore, if policy makers wish to support AD (especially in light of its high carbon emissions reduction value) tipping‐fee enhancements or price‐supports would be more effective than tax credits. Furthermore, tipping fees for waste treatment may be more volatile than energy prices, especially since the OFMSW treatment sector has not been fully developed. Understanding the economic behavior of tipping fees will be crucial for the successful development of AD of MSW.

4.4.4 Recommendations for Continued Development of the APS Digester System Model

In this study, a large number of the cost items were estimated for the system as sized here. In the future, the cost calculators should be replaced with scaling factors for capital and operating cost based on the design capacity. Other MSW digesters were found to exhibit cost curves that fit the form Y = αXβ, where α and β are constants with economies of scale occurring when β < 1 and X is the design capacity in tons per year [46]. Ideally, the capital and operating costs of APS Digester systems with several different design capacities should be estimated as was done for this study, and the resulting curves should be fit with a power equation. This would allow the performance of the digester to be calculated for any size system. In addition, cost functions for wastewater treatment (as a function of the wastewater throughput rate and VS content) and for a PSA unit would allow the model to be used to optimize energy production more accurately. The biogas production should be a function of the retention time which would be a function of the waste density and volume of process water retained. With these modifications the model could be used to maximize the NPW (or any other financial performance indicator) for different HRT.

Future generations of the model could incorporate more sophisticated heat transfer modeling that includes the effects of radiation and pipe losses and more accurately models thermal properties such as the thermal resistivity of cylindrical reactors and heat transfer coefficients around cylindrical reactors. Other geometries and special orientations could also be included. However, based on the low impact of heat on NPW found in this model, these efforts may be more useful for financial forecasting if heat is given a monetary value.

Finally, the model should be validated with data from the pilot APS Digester system and compared against the financial performance of the first full‐scale versions of the system as they are developed. In particular, the heat and electricity requirements of the pilot‐scale digester should be compared with the model predictions, and if necessary a more detailed heat transfer model should be developed. When the APS Digester system is built at full scale, attention should be paid to the real capital and operating costs since these had the largest influence on NPW.

201

CHAPTER 5: Waste-to-Energy Project for a Food Processing Plant 5.1 Introduction There are several food processing plants in Sacramento, California and this Preliminary Design Report (PDR) is for a food processing plant (FPP) that processes meat, seafood, and vegetables to produce, soups, sauces, and dips. Wastewater generated from equipment and floor cleaning operations is screened prior to discharging to the city sewer, producing 9 tons per day (TPD) of wastewater solids. Trial run, damaged, or out of date products in metal cans and glass or plastic bottles constitute nearly 16 TPD of additional waste organic material. Both the wastewater solids and waste cans and bottles are currently land filled at an estimated annual cost of nearly $450,000 as means of disposal.

The University of California at Davis (UCD) contracted RvL Associates (RvL) to prepare a PDR that uses the UCD patented anaerobic phased solids (APS) digestion system to process wastewater solids and the organic material from discarded cans and bottles. Biogas from the proposed facility will be conditioned and used by Sacramento Municipal Utilities District (SMUD) co‐located on the FPP site. Digested solids will be dewatered and processed for removal and use off‐site whereas liquid effluent will be treated and then discharged to the city sewer with the FPP other wastewater.

The following chapters describe the project setting, project description, and cost information. The project setting chapter identifies the location, environmental conditions, applicable regulations, and waste characteristics. The project description provides information about how the materials are processed, how biogas is produced, conditioned and sold, and how the digestate is processed into dewatered solids and treated liquid for sewer disposal. The last chapter presents capital and operating cost opinions as well as projected savings and revenues of the proposed project. Figures are also presented in support of the information presented in the report, including a preliminary process flow diagram and site layout, mass and energy balance, and detailed information on capital and operating cost opinions.

5.2 Project Setting

5.2.1 Location and Environmental Setting The proposed container processing and digestion facility will be located at the FPP facility west of the main plant, north of the SMUD steam power plant and adjacent to the railroad tracks. The proposed site has good access and egress for dump trucks and pallet trucks to deliver feed stocks and remove containers and solids from the APS facility for recycling and disposal. Its proximity to the SMUD facility provides a short

202

conveyance distance to deliver the biogas and access to steam from SMUD for heating water for process use.

The site is at the approximate elevation of 20 feet above sea level and is generally flat. The monthly average temperature is 60.8° Fahrenheit (F) with average daily highs of 73.5° F and average daily lows of 48.1° F; maximum: 115 °F, minimum of 18°F. Day time temperatures are typically above freezing in the winter. However, it is not unusual for daytime temperatures to be at or above 105° F in the summer

Yearly average rainfall is 17.2 inches with monthly and daily maximums of 12.64 and 5.59 inches, respectively. Records also show a maximum of 2.0 inches of snow (water equivalent). Wind is generally from the southwest and averages 7.8 miles per hour (mph) with historic maximum of 74 mph. There is seismic activity in the area and structures must be reinforced appropriately.

5.2.2 Regulatory Requirements FPP and SMUD are in an area zoned as industrial. SMUD produces 160 megawatts of power for FPP and distribution to their other customers. SMUD also produces and sells steam to FPP. In addition to production processes, FPP treats wastewater, and recycles cardboard and other recyclables. FPP treats wastewater by screening the plant discharge before it enters the local sewer. Under an industrial wastewater discharge permit, FPP discharges wastewater into the Sacramento Regional County Sanitation District. Pallets and cardboard are recycled at a covered facility, but no special permit is required for this activity.

As with any major construction activity, building and construction permits would be required and the project ushered through the City Community Development Department. It is likely that FPP would need to modify the industrial wastewater discharge permit to include the discharge from the Waste‐to‐Energy project. Because it is a change in the wastewater solids treatment process, we believe a permit modification is needed. However, the wastewater volume would not significantly increase and the wastewater characteristics are also expected to remain the same. We believe that the proposed recycling activities for metal, plastic, and glass will not require any new special permits.

The Air Pollution Control District will probably require the project to have a permit to construct, permit to operate, and will likely have requirements to comply with or at least address AB32 Greenhouse Gas concerns. The emergency flare to combust biogas will need to be permitted. The only other air pollution sources that will have to be addressed will be odor from the preprocessing area and the digestate treatment area. No fired heaters are planned. Although the project will use electricity to power equipment, because the project will reduce methane emissions from the landfill, we believe there will actually be a reduction in greenhouse gases.

203

5.2.3 Waste Characteristics Solids screened from wastewater are currently loaded into a hopper and then into dump trucks for delivery to and disposal at the local landfill. The material consists mainly of chopped pieces of vegetables, meats, poultry, and seafood. Solids are contaminated with a very small percentage of metal (nuts, bolts, maintenance tools, aggregate, and other inorganic materials).

Metal and plastic cans are stored on pallets, usually in shrink wrap and/or boxes. Metal, cardboard, plastic containers and shrink wrap, and pallets can be recycled if segregated. Contents of the plastic, metal, and glass containers can be emptied to release their contents of organic material for conversion to biogas in the APS system. The proposed system will unfill the metal, plastic, and glass containers, wash out the inside, crushing the plastic and metal containers and breaking the glass containers for recycling. Plastic lids from glass containers are a small percentage of the total weight and will be put in the trash.

For purposes of this study the wastewater solids are considered to have slightly different characteristics than the containerized waste. Average container weight is based on data collected from selected cans, bottles, and glass containers made by a FPP with these types of products. Table 33 shows the quantity and characteristics of the waste. Characteristics of the containerized waste are after removal from the containers.

Table 33. Waste Quantities and Characteristics Parameter Wastewater Solids Containerized Wastes Quantitya 9 tons per day (TPD) 16 TPD Moisture 80% 85% Total Solids (TS) 20% 15% Volatile Solids (VS) 96% of TS 96% of TS Feed rate of VSb 1.728 TPD 2.304TPD Carbon 50% of dry mass 50% of dry mass Nitrogen 3.3% of dry mass 3.3% of dry mass Proteinc 20.4% of dry mass 20.4% of VS dry mass Fat 15% of dry mass 15% of dry mass Metal cansc 65% of total (10.4 TPD)a

Plastic cans/bottlesd

10% of total (1.6 TPD)a

Glass bottlese 25% of total (4 TPD)a

COD/BOD5 1.4:1 1.4:1 a Does not include weight of container; b Calculated from the above information c Containers represent an estimated 9.1 percent of the total weight processed or1.02 TPD d Containers represent an estimated 4.1 percent of the total weight processed or 0.07TPD e Containers represent an estimated 34.2 percent of the total weight processed or 2.12 TPD

204

5.2.4 General Process Description Wastewater solids and pre‐processed container wastes from FPP are chopped and then conveyed as slurry to an equalization tank to homogenize the waste and provide storage for continuous feeding over the weekend and holidays. The equalization tank can also store waste from containers that must be destroyed in a short period of time when there is a safety or corporate requirement to do so. Waste is fed to one of two, first‐stage digesters operating at thermophilic temperatures, approximately 131° F. Each digester is fed for 10 days before beginning to feed the second first‐stage digester. The first‐stage digesters are periodically mixed and heated using a central pumping station and heat exchanger. Using a patented liquid/solids separator, a small amount of liquid from the first‐stage digesters will be transferred to a blend tank. An equal volume of liquid from the second‐stage digester is transferred to the first‐stage digesters. Blend tank liquid is then transferred to the second‐stage digester. The second‐stage digester is mixed and heated to thermophilic temperatures and allowed to settle before effluent is conveyed to the wastewater treatment system and/or recycled back to the first‐stage digester to recycle nutrients and maintain chemical balance in the system.

Solids are periodically removed from the first‐stage digesters and dewatered. Solids are disposed of or used for beneficial purposes off‐site. Liquid from solids dewatering is returned to the first‐stage digester.

Biogas from the first‐ and second‐stage digesters, blend tank and batch treatment tanks are conveyed through a knock‐out drum to remove free moisture before biogas is metered and transferred to SMUD. In case of an emergency, the biogas is combusted in a flare. A small amount of air is conveyed into the top of the digesters to oxidize hydrogen sulfide in the biogas.

5.2.5 Project Assumptions The following are the project assumptions used in developing the PDR:

1. Solids screened from wastewater treatment will be delivered to the site in dump trucks during an 8‐hour period to a covered area at the proposed facility for processing. The front‐end handling and processing equipment is designed to process at least 2 tons per hour of wastewater solids.

2. Waste metal cans, plastic bottles, and glass jars will be delivered on pallets to Unfill line. Cans, plastic bottles, and glass jars will all be of similar size (~10‐24 ounces) and can be automatically loaded into the pre‐processing system to remove the lids of glass jars, tilted to remove the contents, washed and air dried, and conveyed to an intermediate storage bin. Metal cans and plastic bottles are similarly conveyed to a slicer to allow release of their contents, washed and then conveyed to intermediate storage and separation of excess water.

3. Each pallet only contains one type and size of container and containers are in corrugated boxes.

205

4. Automated crushing will be designed for 3 tons per hour. 5. A limited number of pallets may be placed in the area due to space limitations. 6. Power, water, steam, and sewer will be available within 10 feet of the site. 7. While the energy in the biogas produced by the proposed facility is significant,

the volume is a very small volume compared to the total volume of natural gas used by SMUD at the steam power plant. Because of the large difference in relative volume, biogas impurities will not negatively impact the methane fed to the power plant.

8. SMUD sulfur concentration limits for biogas are assumed to be 100 parts per million by volume.

9. Existing pipe support for steam and condensate return between the FPP and SMUD can be used for biogas conveyance.

10. Only a drip trap is required at SMUD to remove moisture from the gas and SMUD will accept the moisture to their local drain. Only a knock‐out drum is required at the proposed facility to remove water from the biogas. SMUD will pay $5.00 per therm for biogas, about 20 percent above current market price for natural gas.

11. Based on a recent conversation with a Waste Management Company representative, washed crushed cans have a recycle value of $225/ton and are recycled in 20‐ton loads; washed crushed plastic bottles have a recycle value of $85/ton and are recycled in 5‐ton loads; cardboard has a recycle value of $155/ton and are recycled in 15 ton loads; and recycled glass costs $15/ton for recycling (versus $45/ton for disposal) and are recycled in 10 ton loads.

12. Hauling charges for recycling are $178.71 per load. 13. Wastewater from the AD system will be treated to match or be cleaner than the

existing wastewater discharge. 14. Ground water is at least 10 feet below the surface. 15. Site only requires minimal grading. 16. Steam is available from the SMUD steam line through a pressure reducing valve

and flow meter for heating water. 17. Dewatered digestate will have a no or minimal cost for disposal and, because of

its organic nature and potential use as a soil amendment or compost, will be hauled away without cost.

5.3 Project Description The following is a more detailed description of the process proposed for the FPP Waste‐to‐Energy facility. A process flow diagram is provided in the appendix for reference. The

206

process calculations used to size the equipment can also be found with the mass and energy balance.

5.3.1 Waste Preprocessing and Equalization

Wastewater Solids

Solids screened from the wastewater are currently discharged into a hopper and then into a dump truck. For this project, the dump truck will deliver the wastewater solids to a covered area at the project site and discharge them into a depressed area equipped with a drain. A small front‐end loader will scoop up the solids and lift them into a hopper equipped with a Solids Grinder and Pump, G‐101. Fresh and/or recycled water will be added to the hopper as the chopper pump located at the discharge of the hopper is operated. The pump associated with G‐101 will be equipped with a rock trap to capture metal or rocks that may damage the pump. Solids will be chopped and recirculated back to the hopper for a pre‐determined amount of time and then transferred to the AD equalization tank (T‐101). Flow to T‐101 will be monitored and totalized.

Containerized Waste—Glass Bottles

Metal cans, glass jars, and plastic bottles will be brought from the plant to the facility on pallets. Shrink‐wrap and non‐recyclable packaging materials securing the containers will be removed by hand and placed in the trash. As noted above, we assumed that the pallet contains only one size and type of container for processing. The operator will selected one type of containers and set‐up the Unfill Line for that container type. Available adjustments are provided to handle containers of a specific size and type (glass containers versus cans or plastic bottles). The pallets will be placed on a pallet lifter and the height of the pallet adjusted so that the operator does not have to bend over to pick up the next box of containers.

The corrugated box holding the case of containers will be scored and cut in half around the perimeter of the case. The top half of the corrugated box will be put into a bin for recycling. The bottom half of the box with the containers will be placed on the Case Inverter/Indexer Conveyor (refer to drawing P‐6 in Appendix A). The Case Inverter/Indexer Conveyor will turn the bottom half of the box over, discarding the corrugated into a recycled corrugated bin and then convey the containers to the next conveyor for processing.

Cans, bottles, jars are automatically transferred to a Singulation conveyor that orients the containers into a single row of containers. Depending on the container type being processed at that time, a diverter will direct the containers to one or the other conveyor, Glass Bottle Conveyor or Metal and Plastic Can Orientor Conveyor.

If the line has been set up for a glass container, the container’s lid will be removed automatically and placed into a trash bin. The glass container will be automatically loaded into the Glass Tip and Cleanout Station. This device tips the container 135° from

207

vertical releasing most of the contents into the hopper below. The hopper has a pyramidal bottom to concentrate the solids and separate the liquid for reuse. Liquid and water from washing the containers is conveyed with the Container Solids Recirculation Pump, P‐101 and used to provide an initial rinse of the container. Hot fresh water will be used as a final rinse and a burst of dry air used to remove moisture before the clean container is transferred to the Glass Bottle Collection Conveyor which elevates the container before depositing them into a bin. The bin will be emptied into a 10 cubic yard roll‐off bin that is loaded to up to 10 tons before hauling it to the recycling center.

Containerized Waste-Metal Cans or Plastic Bottles

If the Unfill Line is processing metal cans or plastic bottles, non‐recyclable packaging will be removed and the same process will be used to cut open the box and place it on the Case Inverter/Indexer Conveyor. The cans or plastic bottles will then be processed as described above using the Singulation Conveyor. The diverter will transfer the cans or plastic bottles to the Orientor Conveyor and then to the Slicer and Cleanout Station where the cans or bottles are sliced with two diamond‐edged cutters. The contents will fall into the same hopper as described above. A burst of recycled water from P‐101 and a hot fresh‐water rinse will be used to clean the can or bottle. The can or bottle is then discharged to a Can Crusher and then to the Crushed Can Elevating Conveyor before being deposited into a bin.

Recycling Packaging and Containers

The bin of corrugated is periodically taken over to the existing compactor where the corrugated is loaded, automatically crushed, compacted, and then stored in a 40 cubic yard (cy) container. When full, the container is taken to a recycler, weighed, emptied and the container returned. Lids for the glass containers are considered trash and hauled off site for disposal. Bins of glass, plastic, and steel are emptied into separate roll‐off bins and when full, transported to the recycler and returned. Glass is restricted to 10 tons per load. Plastic and steel are put into separate 40‐cy containers. Pallets are also recycled and picked up on site.

Equalization

Container Solids Pump, P‐102 will convey solids from the bottom of the Wash Hopper to the same transfer line used to convey wastewater solids and will combine prior to the flow going through the flow meter to the Equalization Tank, T‐101. T‐101 is needed to accommodate will be periodically mixed to homogenize the contents including just before solids is fed to one of the first‐stage digesters. Mixing homogenizes the feed and also provides additional size reduction of the solids for better hydrolysis. The pump is also equipped with a rock trap should any metal pieces get through the hopper and positive displacement pump.

208

5.3.2 Anaerobic Phased Solids (APS) Digestion Waste in the Equalization Tank will be periodically mixed and then fed to the two‐stage thermophilic digester system. Pump P‐103 will feed one of two first‐stage digesters, T‐201 and T‐202. Feeding the waste to the first‐stage digesters will alternate between the two first‐stage digesters every 10 days. After loading, the first‐stage digester will be periodically mixed and heated using a central pump station and heat exchanger.

The first‐stage digesters also will be allowed to settle after mixing. Each first‐stage digesters is equipped with four patented liquid/solids separation augers. There are three transfers involved in the APS system. The first is that the augers allow wastewater to flow by gravity to a small Blend Tank, T‐203.

The second‐stage digester is also operating at thermophilic temperatures and is where methanogensis occurs. It contains neutrally buoyant media to support and retain biological growth. The second‐stage digester is periodically mixed and heated using the central pump station and heat exchanger and allowed to settle. After settling, flow is conveyed by through the central pump station back to the working first‐stage digester. This is the second transfer step in the APS system. Wastewater will be conveyed from the Blend Tank, T‐203 through the central pump station to the second‐stage digester, T‐204. This completes the third transfer.

If needed, caustic from the Caustic Tote, T‐502 and Chemical Feed Pump CP‐502 is conveyed through Central Pump Station piping to the second‐stage digester to adjust the pH. Micro and macro‐nutrients will be pumped into the first‐stage digester to maintain optimum conditions for gas production and digestion.

5.3.3 Digestate and AD Effluent Processing After settling, solids from the first‐stage digesters are periodically removed and conveyed to one of two conical bottomed, decant tanks that operate in a batch‐process mode. Once filled, the variable speed mixer, MX‐601 or MX‐602 is turned on, sodium hydroxide is added to adjust the pH if needed for coagulation and flocculation and to reduce odors. Alum will be added as a coagulant and polymer added as flocculent. The mixer speed is reduced to flocculate the solids which are then conveyed to an elevated Filter Press, FP‐601. Solids are dewatered to an estimated 30 percent solids and discharged to a bin below the press for off‐site disposal, beneficial use as a soil amendment, or in compost. Filter press filtrate is discharged back to the first‐stage digester through the central pump station to complete the digestion process.

After settling, the second‐stage digester decant will be transferred to the second decant tank system in an amount equivalent to the influent flow. Once the decant tank is filled, the variable speed mixer is turned on, sodium hydroxide is added to adjust the pH if needed for coagulation and flocculation. Alum will be added as a coagulant, and polymer added as flocculent. The mixer speed is reduced to flocculate the solids and allowed to settle. Clear liquid is decanted to the FPP sewer by gravity through a flow meter. Solids can be returned to the first or second‐stage digester to help increase the

209

system sludge retention time or conveyed to the other decant tank for dewatering in the filter press.

5.3.4 Biogas Biogas will be generated from the equalization tank, first‐stage and second‐stage digesters, and blend tank and collected in a common line. The vast majority of the hydrogen sulfide and other mercaptans will be generated in the first and second‐stage digesters. A small volume of pressurized air will be fed into the top of the digesters to nurture sulfide‐consuming bacteria that oxidize the sulfide to elemental sulfur which will precipitate and be removed as a solid. Free moisture will be removed in a knockout drum and drained to sewer. The biogas normally flows through a mass flow meter to SMUD for consumption. In case of an emergency, biogas is burned in an enclosed flare. Natural gas is used for fueling the pilot light and or when there is low BTU gas.

5.3.5 Mass and Energy Balance

Table 34. Digester Heat Transfer Calculations

Warm-up value unit Water Xfer 8000 gal Volume of Water to be Recirculated 66400 lbs Heat Reqd 1900000 Btu Total heat reqd to heat 10 tons/vessel 28.6 Btu/lb[ H2O] Tf 135 oF Desired final operating temp Ti 163.6 oF Required water temp for heat xfr Holding Time 16 hrs water to be heated overnight Heat Rate 118750 Btu/hr heat required for water Heat Loss 15000 Btu/hr storage tank loss Total Heat 133750 Btu/hr heat exchanger requirement U 140 Btu/hr/ft2.oF assumed heat transfer rate of exchanger A 83 ft2 heat exchanger surface area 11620 Btu/hr-.oF dT 11.5 oF temperature difference across exchanger Thot max 177 oF maximum heating water temperature

210

Table 35. Overall Thermal Energy Requirements Maintenance q=UAdT Warm-Up * q=mC(Tf-Ti) Available Heat (Peak Night Load w/100% convection loss) (Material at average daily temperature) Direct Combustion of Biogas Tank Dia 10.7 Weight In 10 tons/load 62% Total Gas 66 therms/day Tank HT 14.6 m 20000 lbs/load Moisture 275000 Btu/hr Volume 9722 gal C 1 Btu/lb. o F Boiler Efficiency 80% HHV A_ wall 488.7 ft2 Tf 135 o F Gas Heat 220000 Btu/hr A_ top+ bot 178.7 ft2 Ti (min) 40 o F A_total 667.4 ft2 dT(max) 95 o F Generator Waste Heat q (max) 1900000 Btu/load Generator 22 kWe Insulation Heating Time 16.0 hrs Heat Rate 12000 Btu/kWh R 6 Spray-on Max Warm-Up 118750 Btu/hr Gen Efficiency 28% HHV U 0.17 Btu/hr.ft2. o F *Only needed Mon+Thur nights Gross Heat 188914 Btu/hr Tj Jacket Temp 140 o F Recovery 60% Ts Surface Temp 32 o F Waste Heat 113348 Btu/hr dT 108 o F A 667.4 ft2 q per reactor 12013 Btu/hr no. reactors 5 Max Maint 60067 Btu/hr

211

Table 36. Maintenance Heat Exchanger Requirements

Maintenance Heat (jacket water) Recirc Rate 5.0 Recirculation Rate, gpm 2490 Mass flow rate, lb/hr Heat Load 12013 Maximum heat load (winter nights), Btu/hr 4.8 Btu/lb [H2O] dT 4.8 Temp Rise reqd for heat xfer, oF Tin 135 Desired Reactor Operating Temp, oF Tout 140 Required Heat Exchanger Outlet Temp, oF Seasonal Variations Heat Load Winter Spring/Fall Summer Description Maintenance 12013 10456 8899 Heat Loss per tank (Btu/hr) Flow Rate 5.0 4.0 3.4 Jacket Water flow rate, gpm (per tank) Total Circ 25.0 20.0 17.0 Total jacket water flow, gpm Total variation of flow (per tank): Max Flow 5 gpm coldest winter night Min Flow 2 gpm hot summer day (neglecting radiant gains)

Table 37. Warm-Up Heat Exchange Requirements

Warm-up value unit Water Xfer 8000 gal Volume of Water to be Recirculated 66400 lbs Heat Reqd 1900000 Btu Total heat reqd to heat 10 tons/vessel 28.6 Btu/lb[ H2O] Tf 135 oF Desired final operating temp Ti 163.6 oF Required water temp for heat xfr Holding Time 16 hrs water to be heated overnight Heat Rate 118750 Btu/hr heat required for water Heat Loss 15000 Btu/hr storage tank loss Total Heat 133750 Btu/hr heat exchanger requirement U 140 Btu/hr/ft2.oF assumed heat transfer rate of exchanger A 83 ft2 heat exchanger surface area 11620 Btu/hr-.oF dT 11.5 oF temperature difference across exchanger Thot max 177 oF maximum heating water temperature

212

Table 38. Mass Balance Calculations Parameter Value Unit Mass In 10 tons/batch MC 70 % Water In 7.0 tons water 1686.7 gallons water in TS 3.0 tons [TS]/batch VS/TS 80 % VS 2.4 tons [VS]/batch Destruction 70 % 1.7 tons VS destroyed Remaining 0.7 tons VS remain 1.3 tons TS remain MC residual 60 % Mass Residuals 3.3 tons residual 2.0 tons water residual 477 gallons water residual 1210 gallons water out 5.0 tons water out 8.3 total tons out Biogas Density CH4 0.042 lb/ft3 CO2 0.115 lb/ft3 Methane Content 65 % CH4 0.027 lb/ft3 [biogas] CO2 0.040 lb/ft3 [biogas Biogas 0.068 lb/ft3 [biogas Methane yield 6 ft3/lb [VS] Biogas yield 9.2 ft3/lb [VS] 0.24 lb [VS]/lb[wet] 2.2 ft3 [biogas]/lb [wet] 4431 ft3[ biogas]/ton [wet] 44308 ft3[ biogas]/batch 2993 lb biogas Moisture 0.15 lb/lb [biogas] 447 lb vapor 3440 lb wet biogas 1.7 tons wet biogas

213

Table 39. Energy Balance Calculations

Energy In In Feedstock In Biogas Mass In 2.4 Tons[VS]/batch Biogas Out 44308 ft3/batch Energy Content 18.75 MJ/kg[VS] Energy Content 650 BTU/ft3 17773 Btu/kg[VS] Energy Out 288.0 therms/batch 8078 Btu/lb[VS] Conversion Eff1 74% therms out/therms in Energy In 388 therms/batch Energy Out Electricity Prod Co-generation Efficiency 28% Waste Heat 207.4 therms/batch Electricity Out 80.6 therms/batch Heat Recovery 60% 3413 Btu/kWh Recoverable 124.4 therms/batch 2363 kWh/batch Conversion Eff2 21% therms elec/therms in Backloads Winter Summer Electrical 350 kWh/batch Thermal, Maint 38 28 therms/batch 15% of generated Thermal, Warm 19 15 therms/batch Net Elec Out 2013 kWh/batch 57 43 therms/batch 68.7 therms/batch 46% 35% of recoverable Net Energy Overall Net Conversion Efficiency Winter Feed Biogas Gross* Thermal 67.6 therms/batch 17% 23% 54% Electricity 68.7 therms/batch 18% 24% 85% 136.3 therms/batch 35% 47% 66% Summer Thermal 81.4 therms/batch 21% 28% 65% Electricity 68.7 therms/batch 18% 24% 85% 150.1 therms/batch 39% 52% 73%

* Note: Gross thermal is net fraction of total recoverable heat and gross electricity is net fraction of total electricity produced

5.4 Cost Options Currently waste solids and containerized waste is sent to the landfill for disposal at a rate of approximately $45 ton. Total annual cost is estimated to be $445,000. This project is an alternative to the current method. In simple terms, the net operating cost (operations and

214

maintenance costs plus depreciation minus revenue from gas sales and recycling) should be less than the current method.

5.4.1 Capital Cost Opinion Table 2 provides a summary of our capital cost opinion for the system described above. The cost opinion uses vendor equipment pricing information for the sizes shown in the drawings or prorated from recent other projects that use the same equipment, but having a different capacity. For mechanical equipment, we used the following formula to develop the prorated capital cost:

$larger unit = $smaller unit * (Capacitylarger unit /Capacitysmaller unit)0.6

Appendix B provides a detailed equipment list and cost opinion as well as the system layout (Drawing C‐1) which includes the equipment described above, a roof (only) over the Unfill Line, Filter Press, and the Chemical Feed Systems. A building for the laboratory, office, and storage area building is also shown. Roll‐off bins will be on the same concrete pad as the other equipment. The capital cost also includes a black‐top area for turning around the trucks. Labor will be required to load containers and remove and empty the bins. The APS system will require an operator to monitor and record process variables, schedule and add bulk chemicals, and empty the filter press. The APS digester system will be fully instrumented and automatically controlled so that no routine labor will be required on weekends. The capital cost opinion includes a 25 percent contingency, which is standard for a cost estimate based on the available information and level of detail used to develop the cost opinion.

215

Table 40. FPP Waste-to-Energy Capital Cost Opinion Item Amount EQUIPMENT

Pre-processing equipment $141,840 Anaerobic Digester (APS) $639,892 Digestate and Effluent Treatment $124,895 Biogas Treatment and Flare $112,000 Hot Water, Chemical Feed, Compressor $30,922 Freight and Taxes $157,432 SUBTOTAL $1,206,982

INSTALLATION Contractor mob/demob, testing, office trailer $150,000 Engineering/permitting $350,000 Site work, pad, utilities, building $464,693 Of equipment plus Piping, Electrical,

Instrumentation $801,436 SUBTOTAL $1,766,129 SUBTOTAL EQUIPMENT & INSTALLATION $2,973,111 SUBTOTAL ROUNDED $2,970,000

Contingency @ 25% $742,500 SUBTOTAL $3,712,500

Bonds/Insurance $445,500 Contractor O/H & Profit $556,875

SUBTOTAL $1,002,375 TOTAL $4,710,000

5.4.2 Operating Cost Opinion Operating costs consist of depreciation, labor, chemicals, utilities, replacement parts, repairs, and consumables. These are offset by revenue from the biogas and recycling of cans, plastic, and glass. We assumed a straight‐line 20‐year depreciation schedule of the capital costs. We assumed two full‐time operators will be needed, splitting their time between the Unfill Line, process testing, and monitoring digestate treatment system, and other process equipment. The operators will need to be crossed‐trained and will work 40‐hour work weeks. It also includes quarter‐time each for a mechanic, electrician, and instrumentation and control technician; and a quarter‐time supervisor.

Revenue will come from the sale of the biogas to SMUD and rebates on the recycled steel, plastic, and cardboard materials. While there is no rebate on glass, there is a reduction from $45 to $15/ton for recycling. Table 3 summarizes the operating cost opinion and offsetting revenue.

216

Table 41. FPP Waste-to-Energy Annual Operating Cost Opinion Item Amount Costs

Labor $113,000 Chemicals & Steam $30,000 Electricity $1,000 Maintenance at 2% of Capital $94,200 Trash Disposal $2,255 Depreciation-20 yrs $235,500

SUBTOTAL $475,955 Revenue

Biomethane $96,250 Cardboard $34,057 Steel $80,307 Glass $(11,116)Pallets $5,600 Cost to Haul Recycle Materials $(19,902)

SUBTOTAL $205,099 NET OPERATING COSTS $271,000

217

CHAPTER 6: Characterization of Food and Green Wastes as Feedstock for Anaerobic Digesters 6.1 Introduction Many technologies are currently applied for treatment of food and green wastes. To choose a certain technology for waste treatment, the characteristics of the feedstock should be considered. Anaerobic digestion is widely applied for treatment of organic wastes that are easily biodegradable and have relatively high moisture contents. Recently, it is also applied to materials of low moisture contents (Ten Braummeler, 1993).

From a pollution control standpoint, green (GW) and food (FW) wastes are not hazardous wastes but the large volume produced takes up valuable landfill space (Thassitou and Arvanitoyannis, 2001). The physical and chemical characteristics of the food and green wastes are important information for designing and operating anaerobic digesters for biogas generation. According to Kroyer (1995), there are some common characteristics of food wastes: (1) large amounts of organic materials such as proteins, carbohydrate and lipids, (2) high concentrations of chemical oxygen demand (COD) and nitrogen, and (3) high variations in pH. The feedstock composition is one of major factors that affect biogas production and process stability during digestion. Biodegradability is an important design parameter for anaerobic digesters. It determines the maximum amount of methane that can be produced from a feedstock after subjecting it to anaerobic digestion for a sufficient digestion time. It can also be a good indication of the degradation rate of particulate organic substrates. The rate and extent of digestion dramatically depends on the composition of the feedstock. According to Veeken and Hamelers (1999), wastes collected indoors are largely comprised of non‐lignocellulosic food remainders while those collected outdoors are mainly composed of lignocellulosic plant materials. The former generally degrade easily and quickly.

218

Table 42. Characteristics of Green and Food Wastes Reported in the Literature Waste Source Components Characteristics Country Reference

MC (%)

VS/TS C/N

Food wastes A dining hall Grains, vegetables and meat

79.5 0.95 14.7 Korea Han and Shin (2004)

Food wastes University’s cafeteria

Mainly cooked meal residues

80.03 0.94 Korea Kwon and Lee (2004)

Food wastes A dining hall 93.2 0.94 18.3 Korea Shin et al. (2004)

Food wastes A dining hall 84.1 0.96 Korea Kim et al. (2004)

Food and beverage industry

Food processing Sludge, processing wastes (e.g. sugar, tobacco, beer production, meat and fish)

10 0.85 Netherlands Faaij et al. (1997)

Swill Food remains released at restaurants, hospitals, etc.


Food wastes Emanating from fruit and vegetable markets, household and juices centers

85 0.89 36.4 India Rao and Singh (2004)

Verge grass Mowed grass released during maintenance of road sides


Household organic wastes

Remains of vegetables, fruit, plants and garden waste


Straw Residues from cereal production


Green wastes

Gardens and parks 52.1 0.97 United Kingdom

Manios and Stentiford (2004)

Chao et al (1995) assessed the biochemical methane potentials of different food wastes at 37oC and 28 days of digestion time. The ultimate methane yields were 482, 294, 277, and 472 L/ kg VS for cooked meat, boiled rice, fresh cabbage and mixed food wastes (MFW) respectively. Based on elemental composition of these food wastes, the values of anaerobic biodegradability were, respectively, 82, 72, 73 and 86 percent of the stoichiometric methane yield. Gunaseelan (2004) determined the biochemical methane potential of 54 fruits and vegetable wastes samples and eight standard biomass samples at 35oC and 100 days of digestion time. His data revealed that little, if any, methane was collected during the period from day 30 to day 100. The ultimate methane yields ranged from 180 to 732 and 190 to 410 L/kgVS for fruit and vegetable wastes, respectively. For sorghum, the methane yield ranged from 228 to 538 L/kgVS for roots and

219

inflorescence with grains respectively while for Napier grass, the methane yield ranged from 342 to 419 L/kgVS for sheath and microcrystaline cellulose, respectively.

To study the performance of an anaerobic phased digester system (APS‐Digester) and support the UC Davis pilot digester project funded by California Energy Commission, short and long‐term characterization of the feedstock is required. The main objective of this study was to characterize the proposed feedstock, preprocessed green and food wastes, and to determine the overall variability and consistency of these materials over time. According to the requirement of APS‐Digester system design, thermophilic temperature was used for anaerobic digestion tests. The extensive literature search showed that there is a scarcity of information concerning the biodegradability of food and green waste under thermophilic conditions.

6.2 Materials and Methods

6.2.1 Characterization of Fresh and Green Wastes The food and green wastes studied in this project were provided by Norcal Waste Systems, Inc. who operates a composting facility in Dixon, CA and processes over 200 tons of food and green wastes daily. The food waste was collected from the City of San Francisco and was separated and processed at a transfer station. The green waste was collected from the City of Vacaville and surrounding areas and processed at the composting facility. The food and green waste streams were sampled and analyzed with the procedures described as follows.

Food Waste

The food waste material is currently collected and processed in San Francisco before being sent to Norcal’s Hay Road facility for composting or the East Bay Municipal Utility District (EBMUD) digesters for treatment and disposal. Collection and handling operations include the normal daily pickup of source‐separated material from commercial routes within the city. Currently five routes each day have been included in this study. Each route was initially selected based on estimates of organic content by operations personnel. The material was dropped off at a transfer station in San Francisco where it was first pre‐screened to remove the coarse contaminants (wood, metal, cardboard, glass, plastics, and so forth). The larger material (overs) passed over the screen and was conveyed to a truck for transport to a landfill facility. The material that passed through the screen (unders) was sent to a grinder for final size reduction before loaded onto a truck and transported to the anaerobic digesters of EBMUD.

Initial comparison testing was conducted and resulted in nearly identical findings between the two sources, ground and un‐ground. The un‐ground material tested was collected from the under‐drum conveyor as it fell onto the discharge belt. A scoop shovel was used to collect the material before it left the screen, and several small samples, collected periodically, were mixed in a 5‐gallon bucket to form a single composite sample. The ground samples were pulled at periodic intervals (15 minutes) from the grinder discharge belt and placed in separate zip‐lock bags, put on ice, and delivered to the Bioenvironmental Engineering Research Lab at UC Davis for testing.

220

In order to assess the ability of the APS‐Digester to produce a relatively constant gas production rates, an understanding of the feedstock variability is required. Daily and weekly variations were studied. Samples were initially collected at 15 minute intervals, from each day of the week (Monday through Friday) and then one day (Tuesday) was selected as a representative day of the week, and samples were taken on that same day of the week for a period of 8 weeks.

Green Waste

Samples of green waste were obtained from the composting facility of Norcal Waste Systems, Inc., in Dixon, CA. The green waste was collected from urban areas in and near Vacaville. The collected green waste was first processed through a screen of 4 in openings to separate grasses and similar materials from the large materials. Samples were taken from the materials that passed the screen and transported to the laboratory at UC Davis for analysis. Bi‐weekly samples were taken over a four‐month period from July to October, 2004 with an attempt to determine the variation, as well as average values, in the characteristics of sieved green waste.

Analysis

The Total Solids (TS) and Volatile Solids (VS) contents were measured according to the standard methods. The nutrients contents were measured according the standard methods in DANR Analytical Laboratory (http://danranlab.ucanr.org/). Figure (1) shows the main components of a specific waste, moisture and total solids. The TS contains VS and ash (fixed solids, FS)

Figure 100. Main Components of a Specific Waste: Not In a Scale

VS

Ash

Water

TSMC

6.2.2 Anaerobic Digestion Tests Equipment and Method

Batch digesters are the common systems used to assess the anaerobic biodegradability of different materials. These systems are simple both in design and operation and therefore very

221

attractive (Lettinga, 2001). These systems start with a certain amount of inoculum (for example seed bacterial culture), which should be taken directly from a running reactor and then the substrate to be digested is added with the inoculum at the beginning of digestion period. Afterwards, anaerobic conditions are maintained until the depletion of the available substrate (for example end of the digestion).

Batch anaerobic digestion tests were performed in duplicate using 1‐L anaerobic reactors at thermophilic temperature (50°C ±2). The effective volume of each digester was 500 ml. In the beginning of the tests, 150 ml of inoculum were mixed with a certain amount of food or green waste. Two different initial loadings of food waste were studied: 6.8 and 10.5 gVS/L. For green waste, two different experimental setups were applied. In the first, the green waste was added loosely in the liquid while in the second the green wastes was added in nylon bags. The average initial loading rate in both setups was 13.35 g VS/L.

After the inoculum and intended waste were added, each digester was filled up to 500 ml with tap water. The digesters were tightly closed with a rubber septa and a screw cap. To assure the anaerobic conditions, the head space was purged with helium gas. Two blank digesters that contained inoculum only were also incubated at the same temperature to correct for the biogas produced from the inoculum. Each blank digester contained the same amount of inoculum and was filled with tap water until 500 ml. All the digesters were incubated for a time period until little if any biogas was produced (Figures 5 and 7). Mixing was provided once a day.

Inoculum

The thermophilic inoculum used in the digestion tests was anaerobic sludge collected from a thermophilic anaerobic digester at East Bay MUD in Oakland, California. The inoculum used in the food waste digestion experiments had average TS and VS/TS of 2.56 percent and 50.84 percent, respectively, while the inoculum used in green waste digestion had average TS and VS/TS of 3.40 percent and 61.0 percent, respectively. It should be mentioned that the lower TS and VS/TS for the first inoculum might be due to the storage of the sludge without feeding for about two weeks at 50 °C.

Gas Measurements

Daily methane production from each anaerobic digester was measured and biogas composition (CH4 and CO2 contents) was measured periodically (see Figures 9 and 10) using gas chromatography (GC). The measurements were duplicated for each digester. Biogas production was adjusted to standard temperature (0°C) and Pressure (STP) (1 atm).

6.3 Results and Discussion

6.3.1 Characteristics of Food Wastes

The average MC, VS, and VS/TS of daily samples of food waste are shown in Figure 2 together with standard deviations as indicated by the Y error bars. The MC of food waste ranged from 66 percent to 73 percent. Taking into consideration the standard deviation, it can be deduced that the MC is almost constant over five days of the week, with an average value of 70 percent.

222

Similarly, VS remained fairly constant at an average of 25 percent. The VS/TS was relatively high, which favors anaerobic conversion. For all the samples VS/TS was rather constant, 84‐87 percent, except for Monday’s sample where an average ratio of 70 percent was measured.

Figure 101. Daily Average MC, VS (Weight Basis) and VS/TS of Food Wastes with the Standard Deviations as Indicated By Y Error Bars

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Avg MC Avg VS Avg VS/TS

Mon Tue Wed Thu Fri Average

Figure 104 shows the weekly variations in the average values of MC, VS and VS/TS of the samples tested. The standard deviations are indicated by the Y error bars. The MC, VS and VS/TS were almost constant (74 percent, 23 percent and 87 percent on average, respectively) for all studied weeks. The measured values of VS/TS are similar to the values reported for other countries Appendix. However, the MC of food waste samples collected in this study was higher than those reported. This could be due to many factors, including weather conditions during sample collections and preparation of samples. In the present study, the pre‐screening might take out some of the dry‐ highly carbonaceous material. This also may explain the lower C/N ratio of both substrates (Table 43).

223

Figure 102. Daily Average MC, VS (Weight Basis) and VS/TS of Food Wastes with the Standard Deviations as Indicated By Y Error Bars

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Avg MC Avg VS Avg VS/TS

13-Jul 20-Jul 3-Aug 10-Aug 17-Aug 14-Sep Average

Table 43 shows the nutrient elements of both food and green wastes, revealing that well balanced macro and micro nutrients are available for anaerobic microorganisms. Since the total concentration of each of these nutrients will not change significantly during the digestion, the digester effluents should provide the essential elements for plant growth if they are used as organic fertilizers.

Table 43. Elemental Composition of Food and Green Wastes Components Unit Food wastes Green wastes

C (Total) %* 46.78 41.75

N (Total) % 3.16 3.14

P (Total) % 0.52 0.53

K % 0.90 2.73

Ca (Total) % 2.16 1.43

Mg (Total) % 0.14 0.47

S (Total) ppm** 2507.50 4215.00

NH4-N ppm 972.50 600.00

NO3-N ppm 117.50 35.00

224

Components Unit Food wastes Green wastes

Al ppm 1202.00 1055.50

Fe (Total) ppm 765.75 1500.00

B (Total) ppm 11.25 49.50

Zn (Total) ppm 76.00 125.00

Mn (Total) ppm 59.75 107.00

Cu (Total) ppm 31.13 22.50

Cd ppm <1 <1

Cr ppm 2.50 2.00

Pb ppm 3.50 2.00

Ni ppm 2.00 7.50

* percent of dry matter

** wet material

6.3.2 Characteristics of Green Wastes

Figure 103 shows the characteristics of green wastes from the last week of July until the last week of October 2004. As can be seen, more or less constant average MC was obtained during the study period except for the samples collected on 9/14/2004, which had relatively low moisture content. This may be due to the fact that the temperature was high causing much evaporation. For the first two samples (23rd of July and 14th of September) almost the same VS/TS was determined. Likewise, the last three samples had almost the same VS/TS. Over the whole period, the average values for MC, VS, and VS/TS were 64.9 percent, 26.8 percent, and 77.8 percent, respectively. As for food waste, these results are not completely in agreement with comparable studies from other countries (Appendix B). This may be due to variations in the components of this type of waste.

225

Figure 103. Weekly Average MC, VS (Wet Weight Basis) and VS/TS of Green Wastes. Y Error Bars Show the Standard Deviations

0%10%20%30%40%50%60%70%80%90%

100%

Avg MC Avg VS Avg VS/TS23-Jul-04 14-Sep-04 14-Sep-04 21-Sep-04 22-Oct-04

Similar to the food waste, green waste also had the essential nutrients needed by microorganisms (Table 43). Comparing the composition of food waste with that of green waste, it is clear that both materials had very similar C/N ratios: 14.8 and 13.3 for food and green wastes, respectively. The green waste contained higher sulfur (S) content (4215 ppm) compared to food waste (2507 ppm). Calcium contents were 2.16 percent and 1.43 percent for food and green wastes, respectively. Magnesium concentration in the food waste was about one third of that of the green waste. The ammonia concentration of food wastes was about 1.5 times of that of green waste. In addition, both materials contained metals and other micro elements.

6.3.3 Anaerobic Digestion Tests Table 44 shows the characteristics of food and green waste used in the digestion experiments. It should be mentioned that these data are the average of duplicates and triplicates for food wastes and green waste, respectively. Standard deviations are shown in parentheses.

Table 44. Characteristics of the Feedstock Used in the Experiments with Standard Deviation Between Brackets

Parameter Food waste Green waste

Total Solids (TS), % 30.90 (0.07) 24.70 (2.39)

Volatile Solids (VS), % 26.35 (0.14) 21.64 (2.21)

Fixed Solids (FS), % 4.54 (0.21) 3.06 (0.19)

VS/TS, % 85.30 (0.65) 87.57 (0.51)

226

Methane Production

The methane yield (mL/gVS) and methane production rate (mL/L per day) during the digestion of food wastes are shown in Figures 5 and 6, respectively. The average methane yield from the two digesters that had 6.8 gVS/L initial loading was about 425 mL/g VS. While for the two digesters that had 10.5 gVS/L initial loading, the methane yield was 445 mL/gVS. Essentially, there is no significant difference between the two different initial loadings. So, food waste yielded 435mL of methane per g VS on average after 28 days of digestion. About 80 percent of the methane yield was obtained after the first 10 days of digestion. As can be seen in Figure 104, methane production increased until day 16, and then remained almost constant at a low level until the end of experiments.

The methane yield obtained in this study is comparable to the values reported in the literature. A methane yield of 472 mL/g VS was reported by Cho et al (1995) during the digestion of municipal food waste at 37oC and 25 days. It should be mentioned that the VS/TS of their feedstock was 95 percent. They calculated that the stoichiometric methane yield was 546 mL/g VS, thus the measured biodegradability was 86 percent.

Figure 104. Methane yield of Food Waste During Anaerobic Digestion at 50°C

0

50

100

150

200

250

300

350

400

450

500

0 5 10 15 20 25 30Digestion time (days)

Met

hane

yie

ld (m

L/g

VS)

initial loading = 6.8 gVS/L initial loading = 10.5 gVS/L

Figure 105 shows that the methane production rate was relatively low during the first five days of digestion, increased to reach a peak at the sixth day of digestion, and then declined again. Maximum methane production rates of about 602 and 762 mL/L per day could be achieved for the digesters started at 6.8 gVS/L and 10.5 g VS/L initial loadings, respectively. However, during the increasing period (ca day 2 to day 11), the average methane production rate per g of VS was almost the same for both starting conditions. This was deduced from plotting the data of methane yield (Y) against digestion time (X) and estimating the slopes of the straight lines which represent the average methane production rates for each g of VS during that period. The

227

equations obtained were: Y = 31.032 X with R2 = 0.8858 at 6.8 gVS/L initial loading and Y = 31.309X with R2 = 0.9322 at 10.5 gVS/L initial loading. Thus the calculated average methane rates during that period were 31 and 31 mL/g VS per day at 6.8 and 10.5 g VS/L initial loadings, respectively. This may suggest that the anaerobic sludge used in the experiments had high methanogenic activity. Therefore, it can be concluded that under similar operational conditions, it is possible to operate a single batch reactor, treating food wastes, with an initial loading of up to 10.5 g VS/L without jeopardizing process stability and reduction of methane yield.

Figure 105. Daily Methane Production During Digestion of Food Wastes

0

100

200

300

400

500

600

700

800

900

0 5 10 15 20 25 30

Digestion time (days)

Met

hane

pro

duct

ion

rate

(mL/

L.d)

initial loading = 6.8 gVS/L initial loading = 10.5 gVS/L

Figure 106 shows the methane yield obtained from the experiments conducted with placing the green waste loosely in the liquid. As can be seen, the methane yield (mL/g VS) of green waste during 28 days of digestion. There was a short lag phase for about one day before biogas production commenced. A methane yield of about 226 ml/gVS was obtained after 28 days, out of which 90 percent was produced during the first 10 days. As can also be seen that, after day 20 methane production was negligible. A similar methane yield was obtained (215 mL/g VS) from the experiments conducted with adding the green waste in nylon bags after 20 days digestion time. But lower methane production rate was evident (Fig.107).

228

Figure 106. Methane Yield During the Digestion of Green Waste Placed Loosely in the Reactor

0

50

100

150

200

250

0 5 10 15 20 25 30


Met

hane

yie

ld (m

l[CH

4]/g

[VS

adde

d])

Figure 107. Methane Yield During the Digestion of Green Waste Placed in Nylon Bags Submerged in the Reactor.

0

50

100

150

200

250

0 2 4 6 8 10 12 14 16 18 20


Met

hane

Yie

ld (m

L / g

VS

229

Figure 108 shows the methane production rate (mL/L per day) from the digestion of green waste placed loosely in the reactors. Like food waste digestion, a peak methane production rate of 1567 mL/L per day was achieved after about two days of digestion. This may be due to the higher initial amount of volatile solids, or it may suggest that the green waste used in the experiments had elements (for example volatile fatty acids) that were very accessible to the methanogenic bacteria. According to Chanakya et al. (1999), significant fractions of most biomass feedstocks are readily biodegradable. For the green waste placed in nylon bags, the maximum methane production rate was obtained after 4 days at a level of 667 mL/L per day. Moreover, there was another production peak on day 7 at a level of 616 mL/L per day. Though the methane yields from both experimental setups were almost the same, the maximum methane production rate was lower using the nylon bags. From these results, it can be concluded that adding the waste in nylon bags might have a limitation effect of the contact between the substrate and the microorganisms.

Figure 108. Daily Methane Production During Digestion of Green Waste Placed Loosely in the Reactor

0

200

400

600

800

1000

1200

1400

1600

1800

0 5 10 15 20 25 30


Met

hane

pro

duct

ion

rate

(ml[C

H4]

/l[ef

fect

ive

volu

me]

.day

)

230

Figure 109. Daily Methane Production During Digestion of Green Waste Placed in Nylon Bags Submerged in the Reactor

0

100

200

300

400

500

600

700

0 2 4 6 8 10 12 14 16 18 20Digestion time (days)

Met

hane

pro

duct

ion

rate

(mL

/ L.d

)

Biogas Composition

The biogas composition during digestion of food wastes at both studied initial loadings is shown in Figure 110. Almost constant methane content was obtained under each studied initial loading. Moreover, the biogas produced from the digesters of lower loading had higher methane content compared with the digesters that had higher initial loading. The average CH4

and CO2 contents during the whole experiment were 73.14 percent, and 26.86 percent respectively. Thus, an average energy content of 731.4 Btu/ft3 could be estimated for the biogas produced from food waste. The average values of the biogas composition with standard deviation in parentheses are shown in Table 38.

The biogas composition obtained in this study is comparable to those obtained by other researchers. Wang et al. (2005) studied the anaerobic batch digestion at 35oC of food waste using lab and pilot‐scale hybrid solid‐liquid anaerobic digesters. The results showed that the methane contents of the biogases produced were 71 percent and 72 percent, respectively. The total VS destruction for food waste after digestion was 77 percent and 78 percent after 10 and 25 days, respectively.

231

Figure 110. Biogas Composition during Food Wastes Digestion at Two Different Initial Loadings (6.8 and 10.5 gVS/L)

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20

Digestion time (dyas)

Biog

as c

ompo

sitio

n (%

)

CH4, 6.8 gVS/L CO2, 6.8 gVS/L

CH4,10.5 gVS/L CO2,10.5 gVS/L

The biogas composition of green waste is shown in Figure 111. The methane content increased slightly during the first week of digestion and then remained relatively constant. An average methane content of 54.7 percent was determined over the whole digestion time. Consequently, an average energy content of 547 Btu/ft3 could be estimated for the biogas produced from green waste.

Figure 111. Biogas Composition During Green Waste Digestion

0

10

20

30

40

50

60

70

0 2 4 6 8 10 12


Bio

gas c

ompo

sitio

n (%

)

CH4 CO2

232

Based on these results, the biogas and methane yields and biogas composition are summarized in Table 45. The biogas yields from food and green wastes were 5077.8 and 2862.5 ft3 [biogas]/ton [w.w.], respectively. The data in Table 45 also show that, at the end of the digestion experiments, the VS destructions were 80.6 percent and 79.3 percent and the pH were 7.57 and 7.51 for food and green wastes, respectively.

Table 45. Average Biogas and Methane Yields and Biogas Composition for Food and Green Waste. The Numbers Between Brackets are Standard Deviations

w.w. wet weight. d.m. dry matter.

6.4 Conclusions The overall agreement between the daily and weekly samples, as shown by the narrow ranges of MC, VS, and VS/TS, indicates a reasonably consistent feedstock for both food and green waste. For food waste, the daily average values of MC and VS/TS from five samples taken over a one‐week period were 70 percent and 83 percent, respectively, and the weekly average values from an eight‐week sampling period were 74 percent and 87 percent, respectively. For green waste, the average values from six samples taken over a four‐month period were 65 percent and 78 percent, respectively. The nutrient contents of food and green wastes assure the availability

Parameter Unit Food wastes Green waste

Average methane content % 73.14 (3.64) 54.73 (2.13)

Average carbon dioxide content % 26.86 (3.64) 45.33 (2.12)

VS destruction % 80.57 (3.1) 79.30 (1.7)

Average pH at the end 7.57 (0.13) 7.51 (0.0)

Methane yield L/g VS 0.44 0.23

ft3/Ib VS 7.0 3.6

ft3/ton[d.w.] 12019.2 6342.3

ft3/ton[w.w.] 3713.9 1566.6

Biogas yield L/g VS 0.6 0.4

ft3/Ib VS 9.6 6.6

ft3/ton[d.w.] 16433.1 11588.3

ft3/ton[w.w.] 5077.8 2862.5

233

of nutrients required for anaerobic microorganisms. On average, food and green wastes had C/N of 14.8 and 13.3, respectively. The results of the anaerobic digestion tests showed that:

• Food and green wastes had average methane yields of 435 and 226 mL/gVS, respectively, after 28 days of digestion.

• More than 80 percent of the methane produced from both materials was produced in the first 10 days of digestion.

• The biogas produced from food and green wastes had methane content of 73 percent and 55 percent, respectively.

Based on the sources of food and green wastes, it is expected that the biodegradability of food waste will be relatively consistent throughout the year but the biodegradability of green waste could have seasonable variation. The results of this study on the green waste would well represent the characteristics of green waste in the summer and early fall. Further study is needed for characterizing the green waste generated in other seasons.

234

CHAPTER 7: Thermophilic Digestion of Green and Food Wastes with an Anaerobic Phased Solids Digester System 7.1 Introduction

This section describes lab‐scale research performed prior to the design and construction of the APS Digester system pilot plant described previously. Successful operation of the APS Digester system in the lab was critical for the initial design and subsequent construction of the pilot plant facility.

7.2 Background and Research Objectives

The combination of rising energy prices and needs for reducing the consumption of fossil fuels and greenhouse gas emissions drives development of anaerobic digestion technology as an alternative energy generation technology from renewable natural resources, such as biomass materials. The anaerobic digestion processes use mixed microorganisms to convert organic matter into a methane‐rich biogas, which can be used as a fuel for energy generation. The breakdown of organic matter is carried out in four major steps: hydrolysis, acidification, acetogenesis and methanogenesis (Pavlostathis and Giraldo‐Gomez, 1991).

The anaerobic digestion systems can be designed to be a single stage or two‐phase systems, depending on the characteristics of the substrate and economic and technical aspects. In two‐phase systems, most of hydrolytic ad acidogenic microorganisms are maintained in the first‐phase reactor(s) and most of methanogenic microorganisms are maintained in the second‐phase reactor(s). Feed material is loaded into the hydrolysis reactors therein disintegrated and converted to simple organic acids. These acids are collected and transferred to the biogasification reactor, wherein they are reduced further into biogas. However, in some cases there is no distinctive separation between different phases, depending on the feedstock being digested and pH level maintained in the hydrolysis reactors. There are currently many types of anaerobic digester technologies on the market for wastewater treatment, however there are only a few technologies that are suitable for the treatment of solids wastes. For the digestion of organic solid wastes, two‐phase systems may offer several advantages over the single‐stage systems, such as better process stability and ability of handling higher solids loading rates, (Cho et al., 1995; Xu et al., 2002). For easily biodegradable substrates, such as food waste, performance of two‐phase systems has been proven to be better than single‐stage system (Ghosh, 1977; Ghosh, 1987; Sarada and Joseph, 1995).

This study focused on the continuing development of Anaerobic Phased Solids Digester technology (APS‐Digester), which was recently developed for digestion of solid waste as well as liquid waste (US Patent No. 6342378). The APS‐Digester was developed based on the previous research carried out by many researchers on two‐phase or two‐stage digestion (such as Verrier et al., 1987). Each hydrolysis reactor is operated as a batch reactor while the biogasification

235

reactor is operated as a continuous reactor. The whole system is operated as a continuous system by staggering the feeding schedule of four batch hydrolysis reactors in sequential fashion. The feedstock material is retained in each hydrolysis reactor for a predetermined digestion period, and then the residual solids are removed from the reactor and processed through a solid‐liquid separation device. The recovered liquid is put back into the hydrolysis reactor as the new feedstock is loaded. Depending on the moisture content of the feedstock, all or part of the recovered liquid may be returned to the system. Several previous studies have been published evaluating the performance of the APS‐Digester system in digesting rice straw (Zhang and Zhang, 1999), food garlic wastes (Zhang and Zhang, 2002) and mixed municipal and agricultural wastes (Hartman, 2003). However, the previous studies only tested part of the APS‐Digester system, with one hydrolysis reactor coupled with one biogasfication reactor. This study was initiated to experimentally evaluate an APS‐sigester system consisting of four hydrolysis reactors coupled with one biogasification reactor. These experiments were designed to use food waste and green waste individually or as a mixture of both as feedstock in preparation for the development and optimization of a pilot‐scale APS system being installed at the University of California at Davis (UC Davis).

The overall objective of this research was to determine the efficacy of using the APS‐Digester system to treat food and green wastes. The specific objectives were to: (1) evaluate the performance of the APS‐Digester system in terms of biogas and methane production rates and yields and solids reduction during the digestion of food waste, green waste and a mixture of the two at a thermophilic temperature of 55oC, and (2) determine the effects of different volume ratios of biogasification reactor to hydrolysis reactors on the performance of APS‐Digester system.


7.3.1 Experimental Design

The first part of the experiment was dedicated to the start‐up of the APS‐Digester system with green waste and allowing the system to stabilize for about 72 days. After the system had stabilized, co‐digestion experiments were conducted in which the performance of the APS‐Digester system fed with a mixture of food and green wastes was tested at two BR/HR ratios. The BR/HR was calculated as the volume of biogasification reactor divided by the total volume of four hydrolysis reactors. The volume ratios of 1.0 and 0.5 were selected for evaluation; however the operating conditions gave slightly different liquid volumes in the biogasification reactor, resulting in these ratios to be 0.93 and 0.55, respectively. At the higher ratio, the volume of biogasification reactor and the total volume of four hydrolysis reactors were 3.7 and 4 L, respectively, while at the lower ratio, they were 2.2 and 4 L, respectively. After finishing these two experimental runs (first and second), three more experimental runs were conducted: including digestion of food waste at a BR/HR of 0.55 (third run) and digestion of green waste at two BR/HR ratios, 0.55 and 0.25 (fourth and fifth runs). Table 46 shows the experimental design together with the characteristics of the substrates used in each experiment. It should be

236

mentioned that the BR/HR affect the hydraulic retention time (HRT) of the biogasification reactor when other system parameters (system cycle time, loading rates, feed volumes, and so forth) were the same. Therefore, the calculated HRT of the biogasification reactor was 2, 1.2 and 0.54 days at a BR/HR ratio of 0.93, 0.55 and 0.25 respectively.

7.3.2 Feedstock Collection and Preparation

The green waste used in the startup period of the APS‐Digester was collected from residential homes near Vacaville, CA. After the collection, the green waste was passed through a screen with 10‐cm openings to remove the impurities such as glass, wood and metals. The green waste used in the fourth and fifth experimental runs was lawn clippings collected on the campus of the University of California at Davis. Food waste was provided by a waste management company in Sacramento, California. It was collected from restaurants in the city of San Francisco and prepared by screening and grinding. Details about the sources, processing procedures and characteristics of the food waste are described by Zhang et al. (2007). After collection, the wastes were sampled and taken to the UC Davis Bioenvironmental Engineering Laboratory where they were analyzed and then stored at 4oC until used for feeding the reactors. The characteristics of the green and food wastes used in the experiments are shown in Table 46. The total solids (TS), moisture content (MC) and volatile solids (VS) data are the average of at least three samples. Intention was made to use the same green and food wastes for all the digestion experiments but due to the spoilage of food waste in storage and availability of food waste, food waste collected at different times was used, which resulted in the slightly different characteristics of food waste as shown in Table 46. When the food and green wastes were digested together, enough feedstock for 24 days of digester operation was made from equal amounts of food and green wastes based on VS.

237

Table 46. Experimental Design and Substrate Characteristics Used (GW = Green Waste, FW = Food Waste)

Run

no.

Subs

trate

s

BR

/HR

TS

(%

, w.b

.)

MC

(%

, w.b

.)

VS

(%, w

.b.)

VS/

TS

(%)

C

(g/k

g)

N

(g/k

g)

C/N

NH

4-N

(p

pm)

Start up

GW 0.93 21.6 78.4 19.0 87.7 --- -- -- --

1 FW+GW 0.93 25.7 74.3 21.4 83.2 -- -- -- --

2 FW+GW 0.55 34.0 66.0 29.9 88.0 -- -- --- --

3 FW 0.55 30.9 69.1 26.0 87.0 46.8 3.2 14.8 972.5

4 GW 0.55 27.4 72.6 21 78 41.8 3.1 13.3 117.5

5 GW 0.25 27.4 72.6 21 78 41.8 3.1 13.3 117.5

7.3.3 APS-Digester System Description

The APS‐Digester system was operated at 55 ± 2oC. The temperature was controlled by housing all the reactors in a heated chamber. The experimental set‐up of the APS‐Digester system is shown in Figs.1 and 2 with the liquid and gas flow paths indicated. The system consisted of four identical hydrolysis reactors (denoted as HR1, HR2, HR3 and HR3), a liquid collection tank and a biogasification reactor (BR).

The biogasification and hydrolysis reactors were constructed from acrylic tubing and capped with Poly Vinyl Chloride (PVC) fittings. Each hydrolysis reactor had a total volume of 2.4 L. Packed feed solids were kept submerged in the liquid at the 1‐L level by a screen. Several ports were provided on the top and bottom of each reactor for biogas collection and liquid draining and return. The hydrolysis reactors received the liquid from the biogasficiation reactor on the top and drained from the bottom for liquid collection prior to its transfer to the biogasification reactor. Each hydrolysis reactor was operated as a batch reactor with liquid being transferred in and out during the 12‐day solids digestion time. Loadings of the reactors with fresh feedstock were staggered so that one reactor was loaded every three days.

The liquid collection tank, with a total volume of 2 L, collected the liquid drained from the four hydrolysis reactors, and held it until the next feed cycle of biogasification reactor. The headspace of the collection tank was connected to the headspace of the biogasification reactor, but isolated from the headspaces of the hydrolysis reactors.

238

The biogasification reactor had a maximum total volume of 4.2 L. Different working volumes of biogasification were adjusted depending on different testing BR/HR ratios shown in Table 46. The reactor was operated as an Anaerobic Mixed Biofilm Reactor (AMBR). Approximately 600 mL ring‐ type polyethylene biomedia pellets, acquired from a company in Norway, were suspended in the liquid near the top to provide surface area for microorganisms to attach to. The biomedia pellets had dimensions of approximately 1 cm tall by 1cm in diameter and a density of 0.95 g/cm3. The reactor was decanted at the middle depth of the liquid. The whole digester system was controlled using an automated timer/controller (Model XL, Chrontrol, San Diego, CA).

Figure 112. Schematic of Laboratory Set-Up of APS Digester System

HR1 HR2 HR3 HR4

Liquid recirculation Biogas Biogas Biogas Biogas Biogas

BR

Liquid collection Tank

Pump Pump

Pump

239

Figure 113. Photo of APS-Digester System in the Laboratory

7.3.4 Analysis and Measurements

Samples were taken from the feedstocks (green and food waste) and digested solids from hydrolysis reactors and analyzed for TS and VS contents according to the standard methods (APHA, 1998). The seed sludge used for reactor startup was analyzed for mixed liquor volatile suspended solids (MLVSS), mixed liquor suspended solids (MLSS), TS, and VS according to standard methods (APHA, 1998). The characteristics of seed sludge are shown in Table 2. The nutrients contents were measured by the DANR Analytical Laboratory (http://danranlab.ucanr.org as described by Zhang et al. (2007). Daily biogas production from each reactor was measured using a wet tip gas meters (http://wettipgasmeter.com/), which were held in a controlled environment at 35oC. The measured daily biogas volume was adjusted to the volume at standard temperature (0°C) and pressure (1 atm). Collected biogas was analyzed periodically for methane and carbon dioxide contents using a gas chromatograph (GC) (Model HP5890A, Hewlett Packard, Avondale, PA) equipped with a thermal conductivity detector. The pH of liquid samples from the biogasification effluent and the liquid collection tank were measured prior to loading of each hydrolysis reactor using a pH meter (Accumet AR50, Fisher Scientific, Pittsburg, PA). For the experiment conducted with food waste (3rd experimental run), the biogas was analyzed for H2, CH4 and CO2. Each GC analysis was run in duplicates.

Biogasification

Reactors

Hydrolysis reactors

240

7.3.5 APS- Digester System Startup and Operational Procedures

APS-Digester System Startup

The biogasification reactor was initially seeded with the sludge taken from a thermophilic anaerobic digester at East Bay Municipal Utility District’s (EBMUD) wastewater treatment facility in Oakland, CA. The characteristics of the thermophilic seed sludge used for biogasification reactor are shown in Table 2. After inoculating the biogasification reactor, the first hydrolysis reactor was loaded with 100 gVS of green waste while the other three were loaded with water. The green waste loading was determined by the maximum amount of green waste that could be manually packed into a hydrolysis reactor. The remaining three hydrolysis reactors were brought online over a period of nine days by loading one reactor every three days.

Table 47. Characteristics of the Thermophilic Seed Sludge for Biogasification Reactor Parameter Value

TS (g/L) 20.27

VS (g/L) 11.67

VS(%TS) 57.6

MLSS (g/L) 16.73

MLVSS (g/L) 10.45

APS-Digester System Operation

Following the digester system’s startup with green waste, the feedstock was changed to a mixture of food and green wastes. The mixture was composed of 50 percent food waste and 50 percent green waste (based on VS contents). Then the other experiments were conducted sequentially as shown in Table 46.The data reported here are from a 12‐days digestion periods conducted after the digester system had stabilized under the reported operating conditions. Each hydrolysis reactor was unloaded and reloaded every three days. Unloading was accomplished by draining all of the liquid from the reactor and entirely removing the reactor from the digester system. The drained liquid was saved for the reloading phase. Solids were then manually pressed through a screen with 841‐μm openings to separate the liquid which was then added with the new feed. For each feeding day, the pressed solids were measured for total weight and three samples were taken and analyzed for TS and VS. After being cleaned, the hydrolysis reactor was reloaded with fresh feedstock that contained 100 g VS. The recovered liquid was put back into the reactor with any required make‐up tap

241

water to reach a total volume of 1 L. The reloaded reactor was then returned to the APS‐Digester system.

7.3.6 Data Analysis

Daily biogas yields were calculated for each day in the digestion trial using biogas using OLR values and rate data. One way ANOVA analysis was performed to determine if there was any statistical difference between biogas and methane production at different BR/HR ratios using the twelve data points from each experiment. These tests were performed using Microsoft Excel’s data analysis package at a significance level of ��= 0.05. Because the system is operating under steady‐state conditions each day was considered to be a repeat trial. The ANOVA analysis used each data point as a replication and compared daily production values from the first digestion experiment to those in the second experiment.


7.4.1 Anaerobic Digester System Startup

The performance of APS‐Digester system was characterized by two parameters: (a) ability to produce methane‐rich biogas, which was quantified by daily biogas production volume and methane content of biogas, and (b) effective treatment of the solids waste, which was quantified by TS and VS reductions in the feedstock after digestion. System stability was determined by monitoring pH and daily biogas production of all reactors.

The biogas production of the system during the first 60 days was variable and the biogas production data collected were not accurate due to the limitations of gas meters used (data not shown). The meters were replaced and the system stabilized before the data were collected for reporting. The pH in the system over the startup period was monitored to determine the stability of all reactors. Over a 72‐day period, pH in all reactors increased to and stabilized near 8.25 (Fig. 3). The pH in each hydrolysis reactor cycled between a low value following each batch loading of the feedstock and higher values near the end of the batch digestion period. The low pH value in the hydrolysis reactors at the initial period was mainly due to the rapid production of organic acids from the degradation of readily hydrolyzed substances in the feedstock and then the pH increased as methanogenic bacteria established themselves within the reactor and began to consume the organic acids. Such a result concurs with the studies of anaerobic batch reactors (El‐Shimi et al., 1994; Zhang and Zhang, 1999). Methane contents in the hydrolysis reactors were found to vary with respect to the digestion time. A low methane content of about 45 percent was found shortly after loading, whereas a high methane content of about 75 percent was reached towards the end of the digestion cycle. The methane content in the biogas produced in the biogasification reactor was consistently higher than the methane content in the biogas produced in the hydrolysis reactors at any point in their batch cycle (data not shown). This may indicate that the methanogenic and hydrolytic bacteria had been separated to some extent into their respective reactors. However, increasing methane content of the biogas

242

produced in the hydrolysis reactors indicated that methanogenic bacteria were establishing themselves in the hydrolysis reactors over the 12‐day batch digestion period.

Figure 114. pH of the Five Reactors in the APS-Digester System During the Digestion Of Green Waste in the APS-Digester System During the 72-Day Startup Period

6

6.5

7

7.5

8

8.5

0 10 20 30 40 50 60 70 80

Operating time (days)

pH

BR HR1 HR2 HR3 HR4

7.4.2 Digestion of Green and Food Waste Mixture

The cumulative biogas and methane production at both studied BR/HR ratios during digestion of food and green waste mixture is shown in Figure 115. The cumulative biogas and methane production was slightly higher at BR/HR of 0.93 than at BR/HR of 0.55.

Figure 115. Cumulative Biogas and Methane Production in the APS-Digester System at BR/HRs of 0.93 and 0.55

0

20

40

60

80

100

120

140

160

180

200

0 2 4 6 8 10 12 14Digestion time (days)

Cum

ulat

ive

gas p

rodu

ced

(L)

biogas (BR/HR 0.93) methane (BR/HR 0.93) biogas (BR/HR 0.55) methane (BR/HR 0.55)

243

The biogas and methane yields and TS and VS reduction are shown in Table 48. Biogas and methane values are the average of 12 measurements. TS and VS values are reported as the average of three measurements. The biogas yields from the digestion of the mixture were 537 and 461 mL/gVS, respectively for BR/HR of 0.93 and 0.55 and the methane yields were 430 and 319 mL/g VS, respectively. The solids reductions at BR/HR 0.55 were slightly higher than the reductions at BR/HR of 0.93. The increased VS and TS reductions suggest that the system would have produced greater amounts of biogas and methane, however the yields were slightly lower (Table 48). Statistical analysis of the data for digestion of food and green waste mixture showed no significant difference in biogas and methane yields between the two studied BR/HR ratios. This indicates that operating the APS‐Digester system at the BR/HR of 0.55 would be both economically and functionally superior to the BR/HR of 0.93.

7.4.3 Digestion of Green Waste

The daily biogas production rates from the five reactors in the APS‐Digester system during the digestion of green waste at the BR/HR of 0.55 and 0.25 are shown in Figures 116 and 117, respectively. Daily biogas production rates from each of the hydrolysis reactor were varied over the batch digestion time (12 days) with a higher rate shown after the loading of each hydrolysis reactor. On the other hand, the biogas production rate from the biogasification reactor was fairly constant over the digestion period. The total biogas production rate from the system varied between a maximum of 0.96 L/L.day and a minimum of 0.82 L/L.day for the BR/HR ratio of 0.55 and a maximum of 1.02 L/L.day and a minimum of 0.93 L/L.day for the BR/HR ratio of 0.25. An average biogas production rate of 0.86 and 0.98 L/L.day could be determined, respectively for the APS system having BR/HR ratio of 0.55 and 0.25. The biogas production from hydrolysis reactors in the APS‐ Digester system having BR/HR ratio of 0.55 is higher than that of 0.25. This may be attributed to the increase of the HRT from 2.4 to14.3 days in the hydrolysis reactors.

244

Figure 116. Daily Biogas Production in Five Reactors of the APS-Digester System During Digestion of Green Waste at a BR/HR of 0.55

0

1

2

3

4

5

6

7

0 2 4 6 8 10 12 14


Bio

gas

prod

uctio

n (L

/day

)HR1 HR2 HR3 HR4 BR Total

Figure 117. Daily Biogas Production of Five Reactors of the APS-Digester System During Digestion of Green Waste at a BR/HR of 0.25

0

1

2

3

4

5

6

7

0 2 4 6 8 10 12 14


Bio

gas p

rodu

ctio

n (L

/day

)

HR1 HR2 HR3 HR4 BR Total

The average methane contents of the biogas produced from the APS‐Digester systems are shown in Figure 118. The methane content was almost constant for both BR/HR ratios tested. On average, a methane content of 55 percent was obtained in the hydrolysis reactors for both systems (data not shown). For both systems, the biogas produced in the biogasification reactor had a higher methane content than the biogas produced from the hydrolysis reactors (data not shown). The methane content of biogas produced in this study was lower than that (for

245

example71 percent) reported by Yu (2002) during the digestion of grass waste using a mesophilic two phase anaerobic digestion system.

Figure 118. Methane and Carbon Dioxide Contents of Biogas Produced in the APS-Digester System during Digestion of Green Waste at Two Different BR/HR Ratios

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14


Biog

as c

ompo

sitio

n (%

)Methane BR/HR=0.55 Carbon dioxide BR/HR=0.55Methane BR/HR=0.25 Carbon dioxide at BR/HR=0.25

The cumulative biogas and methane production from the APS systems are shown in Figure 119. The biogas and methane production increased linearly over time. The average biogas and methane yields of green waste were determined to be, respectively, 438 and 252 mL/ for BR/HR of 0.55, and 318 and 175 mL/gVS for BR/HR of 0.25. The average TS and VS reductions in the feedstock after digestion were 57.8 percent and 78.0 percent, respectively, for the system of 0.55 BR/HR and 59.1 percent and 75.8 percent, respectively, for the system of 0.25 BR/HR. The average pH value of the liquid of the BR and the HR were 7.8 and 8.2, respectively with the BR/HR of 0.55 and 7.6 and 8.1 with the BR/HR of 0.25.

246

Figure 119. Cumulative Biogas and Methane Production in the APS-Digester System During Digestion of Green Waste at BR/HR of 0.55 and 0.25.

0

10

20

30

40

50

60

70


Cum

ulat

ive

gas (

L/da

y)

Biogas at BR/HR=0.25 Methane at BR/HR=0.25Methane at BR/HR=0.25 Biogas at BR/HR=0.55

7.4.4 Digestion of Food Waste

Biogas production rates during the digestion of food wastes using the APS‐Digester at BR/HR of 0.55 are shown in Figure 120. The biogas production rate increased directly after loading a hydrolysis reactor and then declined until loading a next hydrolysis reactor. The total biogas production from the system varied between a maximum of 3.8 L/L.day to a minimum of 3.0 L/L.day with an average biogas production rate of 3.2 L/L.day. The composition of biogas produced in one of the hydrolysis reactors over the 12‐day digestion period is shown in Figure121. It is interesting to note that the biogas produced during the first day contained up to 30 percent hydrogen. Then hydrogen content gradually decreased and then vanished after the third day. The biogas, methane and hydrogen yields from the digestion of food waste are shown in Figure 122. A linear increase of both biogas and methane production can be seen. The calculated average biogas and methane yields were 596 and 379 mL/gVS, respectively. An accumulative hydrogen production of 29 mL/gVS could be calculated after 12 days of digestion. Hydrogen production yield represented 4.9 percent of the biogas production yield in the system. More research is underway in our laboratory to improve the hydrogen production in the APS‐Digester system from different types of organic wastes.

The TS and VS reductions in the food waste after 12‐day digestion were measured to be 72.4 and 84.5 percent, respectively (Table 3). These values are higher than those of green waste. The VS reduction values are in line with those reported by Zhang et al. (2007). The measured pH values were 6.7 and 7.8 for hydrolysis and biogasification reactor, respectively.

247

Figure 120. Daily Biogas Production of Five Reactors in the APS-Digester System During Digestion of Food Waste at a BR/HR of 0.55

0.00

5.00

10.00

15.00

20.00

25.00

30.00


Biog

as p

rodu

ctio

n ra

te (L

/day

)

HR1 HR2 HR3 HR4 BR Total

Figure 121. Composition of Biogas Produced From One of the Hydrolysis Reactors During Digestion of Food Waste in APS-Digester System at a BR/HR of 0.55

0

10

20

30

40

50

60

70

80

90

0 2 4 6 8 10 12 14


Biog

as c

ompo

sitio

n (%

)

Hydrogen Methane Carbon dioxide

248

Figure 122. Cumulative Production of Biogas, Methane and Hydrogen During Digestion of Food Waste in APS-Digester System at a BR/HR of 0.55 (3rd Experimental Run)

F

y = 19.436xR2 = 0.9986

y = 1.0561xR2 = 0.9278

y = 10.988xR2 = 0.9511

0

50

100

150

200

250


Cum

ulat

ive

gas (

L/da

y)

Biogas Methane Hydrogen

Table 48. Experimental Design and Performance of the APS-Digester Under Different Conditions (GW = Green Waste, FW = Food Waste)

Run No. Substrates BR/HR

Biogas yield (mL/gVS)

Methane yield (mL/gVS)

TS destruction (%)

VS destruction (%)

1 FW+GW 0.93 537 430 72.4 74.8 2 FW+GW 0.55 461 319 77.3 81.8 3 FW 0.55 596 379 72.4 84.5 4 GW 0.55 438 252 57.8 78.0 5 GW 0.25 438 247 59.1 75.8

7.5 Conclusions

Application of APS‐Digester system for digestion of food waste and green wastes as well as their mixtures was evaluated in the laboratory under thermophilic conditions. Based on the research results, BR/HR of 0.55 is recommended for the digestion of food waste and mixture of food waste and green waste and BR/HR of 0.25 is recommended for the digestion of green wastes. Under these operation conditions, the average biogas and methane yields were, respectively, 596 and 379 ml/gVS for food waste, 461 and 319 ml/gVS for mixture of food and green wastes, and 438 and 247 ml/gVS for green waste. Hydrogen production was measured when the food waste was digested. It was found that hydrogen yield was 4.9 percent of the biogas yield. Research is underway in the laboratory to measure and improve the hydrogen production from different types of organic wastes. The TS and VS reductions were, respectively,

249

72.4 percent and 84.5 percent for food waste, 77.3 percent and 81.8 percent for mixture of food and green wastes, and 59.1 percent and 75.8 percent for green waste.

250

CHAPTER 8: Biodegradability and Soil Amendment Potential of Anaerobically Digested Residues 8.1 Introduction The following section describes research on digestate generated by a lab‐scale APS Digester treating mixed food and green waste. The research was performed in the lab prior to construction of the APS Digester pilot plant. The ability to utilize digester residuals in agricultural and horticultural industries can be essential for the financial feasibility of commercial anaerobic digester facilities.

8.2 Background and Research Objectives The U.S. generated 251.3 million tons of municipal solid waste; 162.9 million tons of this waste was organic matter in 2006 (EPA, 2006). About 66.1 million tons of the organic waste was recovered, either through composting, recycling or combustion for energy generation. Given current tipping fees of $25 per ton, the 96.8 million tons of land‐filled organic matter is equivalent to 2.4 billion dollars of wasted revenue (Zhang, 2007). In landfills, organic matter takes up space and releases methane; when added as a soil amendment, it can help reduce erosion, reduce fertilizer usage, and increase crop yields (Atiyeh et al., 2000; Golueke, 1972; Hoitink, 1997; Martin, 1992).

Anaerobic digestion is an alternative method of organic matter decomposition to aerobic composting. Anaerobic digestion is a biochemical process to convert the organic substrates into methane and hydrogen. The benefits of anaerobic digestion are production of biogas, a renewable energy source, reduced greenhouse gas emissions, and the possible byproduct of a soil amendment. Anaerobic digestion typically converts about 50‐75 percent of the initial total solids to gaseous form, leaving 25‐50 percent as anaerobically digested solid residues (Zhang, 2006). Anaerobically digested residues have been reported to have excellent fertilizing properties, increase soil organic matter content, and decrease plant pathogen populations (Bath and Ramert, 1999; Dahiya and Vasudevan, 1986; Garg et al., 2005; Mehta and Daftardar, 1984; Salminen et al., 2001; Shen, 1997; Shenlin, 1992; Veeken, 2005; Vermeulen et al., 1993).

Historically, anaerobic composts were among the first mechanized composting processes; the Beccari process combines an anaerobic stage in an enclosed reactor with gradual venting to the final aerobic stage, and was first patented in 1920 (Golueke, 1972). Crops grown with anaerobically digested residues have been shown to have higher yields than with aerobic compost, mainly due to its higher mineral nitrogen content (Bath and Ramert, 1999). Both aerobic composts and anaerobically digested residues have been reported to prevent disease in horticultural crops (Hoitink, 1997; Jingru, 1992). In one study, biogas residues resulted in higher yields and grain quality for wheat and barley than aerobic composts for the same amount of nitrogen applied (Svensson et al., 2004).

251

Anaerobically digested composts may have different qualities than their aerobic counterparts, which may make them more or less suitable for potting media. Some research has been done on processing anaerobically digested paper and kitchen wastes into plant growth substrates, but a full comparison of the aerobic and anaerobic pathways has not been completed (Vermeulen et al., 1993). A comprehensive nutrient and heavy metal content analysis of anaerobically digested residues is essential to widespread agricultural adoption of these residues. Carbon to nitrogen ratios should ideally be 25:1 to allow nitrogen mineralization after soil incorporation (Martin, 1992). Aerobic composting can have large amounts of ammonia volatilization, which reduces the nitrogen fertilizer in the final product (Kirchmann and Witter, 1989). Anaerobic digestion has the potential to increase the amount of mineral nitrogen in the residue, as compared to aerobic composting. Excess mineral nutrients, such as calcium, potassium, or magnesium, can present a salinity problem for both aerobic and anaerobic composts. However, excess salts in compost used for horticultural purposes can be alleviated with leaching (Chong, 2005). Nevertheless, saline organic wastes used to amend mineral soils can salinize both soil and groundwater, if appropriate rates are exceeded (Hao, 2003). At present in the U.S., heavy metal content of composts must meet the regulatory limits for bio solids. Compared to European standards, the US heavy metals limit is approximately twenty times as high (Brinton, 2000).

Stability is an important criterion in the evaluation of biogas residues as potential soil amendments. Stability is the degree to which the biodegradable organic matter will resist further decomposition, and can be determined by a variety of measurements, including CO2 production, O2 consumption, Dewar self‐heating tests, or ammonia to nitrate ratios (Brinton, 2000; Gomez et al., 2006). Dynamic respirometry is the technique of measuring the changes in CO2 production and O2 consumption over a specified time interval, ranging from 3 days to 2 weeks. By comparing the maximum rates with published standards, a compost or soil amendment can be classified as stable or unstable (Brinton, 2000). Maturity is the level of completeness of the composting process. Seed germination, growth tests, and tests for the absence of phytotoxic compounds can all be used to determine the maturity level (Brinton, 2000). Stability is correlated with maturity, and many composting facilities use it as a proxy measure of maturity. By measuring the stability of the anaerobically digested residues, it will be possible to predict whether an aerobic post‐treatment process will be necessary to achieve a high‐quality soil amendment.

Most of the published data, on using the digested materials as soil amendments, are from digesters treating liquid substrates. However, digestion of solid and semi‐solid substrates has a great importance from economical and environmental points of view. The Anaerobic Phased Digester (APS‐Digester) system is a two stage system capable of digesting solid feed‐stocks, and is fully described by (Zhang, 2006) and Zhang and Zhang (1999). The digested materials could be applied as soil amendments. Key areas that must be investigated for agriculture to begin using biogas residues as fertilizer are: uniformity and reproducibility of nutrient content, physical properties analysis, stability, demonstration of lack of phytotoxicity with plant growth trials, and analysis for heavy metals and other toxins.

252

The first objective of this study was to evaluate the nutrients and heavy metals contents of the digested food and green wastes. The second objective was to measure the biochemical stability of the compost produced from anaerobically digested food and green wastes.


8.3.1 Food and Green Wastes

Fresh food waste, food compost and green‐waste compost were supplied by NorCal Waste Systems (Dixon, CA). The food waste used for digestion and composting was similar and collected from San Francisco by NorCal Waste Systems. The initial total solid of the food waste was 30.9 percent, on a wet basis (w.b.), and the volatile solids to total solids ratio (VS/TS) was 87 percent.

NorCal Waste Systems prepared the aerobic composts by first grinding up food and green waste with an industrial‐sized grinder. The ground food waste was mixed with ground pallets for composting. The ground green‐waste was not mixed with other wastes before composting. The ground material was placed in plastic bags with forced aeration (Ag‐Bag Composting Technology, Ag Bags International Ltd., St. Nazianz, WI). After sixty days, the compost was removed and was allowed to cure for thirty days in windrows before sampling occurred.

The green waste samples used for anaerobic digestion were taken from UC Davis lawn clippings. The initial total solid (TS) of the grass waste was 27.4 percent (w.b.) and the VS/TS ratio was 78 percent. The grass waste is representative of the process of screening of fresh green waste for use in the digester, since the large size of un‐composted wood chips would be unacceptable for the digester’s pumping system. Although the digested and composted feed‐stocks were not identical, the finished composts and digested residues were representative of the final product farmers or horticulturalists would receive.

8.3.2 Digestion System

A laboratory‐scale Anaerobic Phased Solids (APS) digester system was used to process food and green waste, as described by Zhang (2006). The digester system consisted of four hydrolysis reactors and one biogasification reactor, with working volumes of 1 L and 3.7 L, respectively. During digestion, liquid was circulated once every two hours to transport the fermented feedstock to the biogasification reactor. Food and green wastes were digested for 12 days each, at 55oC. The solid food and grass residues were removed manually from the digester and new batches of food and grass wastes were loaded. The solid residues were stored at 4oC for 1‐3 months until the respirometry experiment.

8.3.3 Particle Size Distribution Due to their larger particle size, the digested residuals were briefly ground in a blender to achieve similar particle sizes as the corresponding aerobic compost (Figure 123). The particle size distribution of the digested residuals was measured using nested sieves at an approximate moisture content of 10 percent. The digested grass waste has a smaller particle size than the

253

composted green waste, especially in the <0.05 mm and >2 mm range. The pH of the digested and composted wastes was measured using 1:5 slurry as described in standard test methods (Thompson, 2002).

Figure 123. Particle Size Distribution for Different Materials.

Composted food waste (CFW), digested food waste (DFW), digested grass waste (DGW) and composted green waste (CGW).

8.3.4 Carbon Dioxide and Oxygen Consumption Measurements

The carbon dioxide evolution and oxygen consumption of composted and digested samples of both food and green wastes was measured by dynamic respirometry. A potting soil mix (1:1 peat and vermiculite on a volume basis) was used as a control. All digested residuals, composts, and potting soils were stored at 4oC until the testing began.

Each waste was mixed with the potting soil in a 1:1 ratio, on a dry weight basis (d.b.). The moisture content of the mixes was determined by drying the samples at 70oC for 24 hours. The mixtures were adjusted to approximately 60 percent moisture content (d.b.). The mixtures were allowed to equilibrate at 4oC overnight. The moisture contents at the beginning and end of the respirometry experiment are shown in Table 49.

254

Table 49. Initial and Final Moisture Content of the 1:1 Mix, Final pHof the 1:1 Mix, the Final Bulk Density of the Mix, and the Initial VS/TS Ratio of Each Type of Waste With Potting Soil

Organic Media Initial Moisture Content (%)

Final Moisture Content (%)

Final Bulk Density (g/cm3)

Final pH VS/TS Ratio

Potting Soil 59.6 57.9 0.089 6.9 0.34

Digested Food 57.1 60.3 0.116 7.2 0.55

Food Compost 60.0 59.8 0.121 6.8 0.47

Digested Grass 56.2 58.8 0.084 7.1 0.54

Green Compost 57.1 58.4 0.121 6.8 0.41

For anaerobic digestion, three replicate 25 g (dry weight) samples of the potting mix control or 1:1 mixtures of potting mix and either digested food, digested grass, composted food, composted green waste were loaded into 250 mL, 7 cm diameter reactors and placed in an incubator (Istotemp Incubator, Fisher Scientific) at 35°C for two weeks. An unfilled reactor was included to measure background levels of oxygen and carbon dioxide. A schematic of the respirometry experiment system is shown in Figure 124 A flowmeter (RateMaster Flowmeter, Dwyer Instruments) maintained a forced air‐flow of 20 mL/min through the sixteen reactors. The flow rates were measured every 2 to 3 days throughout the experiment using a mass flow meter (Humonics Veri‐flow 500, J & W Instruments). A chamber identical to the reaction vessels was filled with water to create a bubbler, which humidified the air before it reached the flowmeter. After passing through each sample, the air was dehumidified with a molecular sieve. A 16‐position multiposition valve (VICI Valco, Houston, TX) was used to pass exhaust air from each reactor through CO2 and O2 sensors. The valve was switched every 20 minutes, so that each reactor was sampled once during 320‐minute cycles. Oxygen concentrations were measured using a Zirconia oxide oxygen sensor (Neuwghent Technologies, LaGrangeville, NY), and carbon dioxide concentrations were measured using an infrared CO2 sensor (Vaisala, Suffolk, UK).

The flowmeter maintained the flow of air to the 16 sample reactors at approximately 20 mL/min. The multi‐position valve allowed the data acquisition system to measure the CO2 and O2 concentrations for each reactor for twenty minutes. Each reactor was measured once during a 5 hour, 20 minute cycle. The data acquisition system (LabTechpro Build Time, Laboratory Techologies Corp., Inc.) recorded the average molar concentration of CO2 and O2 over the twenty‐minute intervals for each reactor.

255

Figure 124. Schematic of the Set-Up for the Respirometry Experiment.

The carbon dioxide evolution and oxygen consumption were calculated from the difference in carbon dioxide and oxygen concentration, respectively, between the reactors with samples and the blank reactor. The carbon dioxide evolution rates (CER) and oxygen uptake rates (OUR) were calculated by using the following equations:

CER = F (CO2out −CO2in ) /gV .S .

..22 /)( SVinout gOOFOUR −=

where F is the air flow rate (mg air/day), CO2,out is the concentration in the outgoing air (mg CO2

/mg air), CO2,in is the concentration of carbon dioxide in the incoming air from the blank reactor (McEachin and VanderGheynst, 2006), O2,in and O2,out are the respective concentrations of oxygen in the incoming and outgoing air (mg O2 /mg air). The dynamic respiratory index (DRI) was calculated by averaging the oxygen consumption rate over the 24‐hour period with the greatest biological activity (Scaglia et al., 2000). The maximum carbon dioxide evolution index (MCI) was calculated similarly, by averaging the CO2 evolution rate over the 24‐hour period with the highest rates, although the rates are typically compared on a per day basis (Brinton, 2000).

8.3.5 Data Analysis

The DRI and MCI were analyzed by a one‐factor ANOVA method using Excel software 2003. The cumulative CO2 evolution and O2 production were also analyzed using a one‐factor

256

ANOVA. The differences in each of the treatments were compared using Tukey’s test at α = 0.05.


8.4.1 Element Contents

Many of the elements in the food and grass wastes increased in concentration (based on the dry solids) during digestion, most likely because all of the elements are nonorganic and the volatile solids are undergone degradation and converted into biogas and therefore they decreased during digestion. But the nutrient content within the system did not decrease (Table 50). During solid extraction from the digester effluent, a large proportion of insoluble nutrients will stay with the solid residue. The nitrogen content decreased in the digested food waste, and the potassium decreased in both the digested solid fraction of food and green wastes. The decrease of nitrogen and potassium could have been caused by these more soluble ions dissolving in the liquid medium that remained in the APS Digester. The large increase of aluminum and iron during digestion may have resulted from interaction of the bacteria and Archaea with metal components of the APS Digester. Sulfate‐reducing bacteria have been shown to corrode iron and aluminum in anaerobic environments; sulfate reduction could have occurred at the beginning of digestion (Hamilton, 1985).

Table 50. Elemental Content (Percent of Dry Solids) of the Fresh and Digested Food and Green Waste Is Shown

Element Food Waste Green Waste Percent Increase

(%) Fresh Digested Fresh Digested Food Green

Nitrogen 3.16 1.94 2.56 2.88 -39 13

Phosphorus 0.52 0.58 0.52 0.63 12 21

Potassium 0.90 0.52 2.22 0.89 -42 -60

Calcium 2.16 2.83 1.06 1.71 31 61

Magnesium 0.14 0.19 0.45 0.55 41 22

Sulphur 0.25 0.30 0.37 0.45 19 23

Aluminum 0.12 0.63 0.03 0.11 425 317

Iron 0.08 0.71 0.04 0.15 829 327

257

The heavy metal content of both green and food wastes increased during digestion (Table 51). The US Biosolids limits from EPA rule 503 (also applied to land application of processed sewage sludge) are shown in Table 3 as well as the recommended maximum for intensive compost use with vegetables (Brinton, 2000).

The final heavy metal content is still significantly below the US Biosolids limit, set in EPA rule 503, which is the only legal limit on heavy metals in compost in the U.S. (Brinton, 2000). When judged against the German Agriculture and Horticultural Association standards for heavy metal in intensive vegetable production (Table 51), the heavy metal content was low enough to allow intensive use in vegetable production, except for the nickel and zinc content in the digested food waste, and the copper content in both types of waste.

Table 51. Heavy Metal Content (mg/L) of the Fresh and Digested Food and Green Waste

Heavy Metal

Food Waste

(mg/L)

Green Waste

(mg/L)

Bio-solids Limit

(mg/L)

Intensive Use

(mg/L)

Percent Increase

Fresh Digest Fresh Digest Food Green

B 11 18 24 131 NA NA 60 446

Cd <1 <1 <1 <1 39 0.75 NA NA

Cr 2.5 13.2 2.0 9.7 1200 75 428 397

Cu 31 107 14 107 1500 50 245 644

Pb 4 27 6 10 300 75 657 67

Mn 60 84 54 102 NA NA 41 91

Ni 2 51 7 16 420 30 2425 129

Zn 76 266 38 154 2800 200 249 305

8.4.2 Compost Suitability The compost suitability parameters for the fresh and digested food and green waste are shown in Table 52. The total concentration of available nitrogen in both digested food waste and digested green waste exceeded the 300 mg/l level recommended by Wood’s End Laboratory (Brinton, 2000). The pH is 8.6 for the digested food waste and 7.6 for the digested grass waste,

258

although the pH decreased to 7.2 and 7.1, respectively, after two weeks of aerobic composting with peat moss, as indicated in Table 50. The pH is recommended for potting mixes to be between 6 and 7, and both residues are higher than this recommendation (Brinton, 2000). The total carbon concentration was 47.2 percent for the digested food waste and 42.7 percent for the digested grass waste. The amount of organic matter is recommended to be greater than 30 percent; both residues satisfy the organic matter requirement (Brinton, 2000). The electrical conductivity was not measured for the fresh food waste, but the salt content decreased during digestion for the green waste. The EC of the digested food waste was 2.44 dS/m, while the digested green waste was 1.88 dS/m. The electrical conductivity of potting soils is recommended to be below 2.0 dS/m (in a 1:1.5 dilution) to avoid salinity damage in general nursery conditions (Handreck and Black, 2002). However, excessive salts in potting media can be leached by standard irrigation procedures or by mixing with other, non‐saline potting media (Chong, 2005).

The fresh and digested food and green wastes are classified as immature, using both the ammonia/nitrate parameter and the C/N parameter (Table 52). The ratio of ammonia to nitrate is a maturity parameter; the lower the ratio, the more nitrification has occurred in the compost. The digested food waste had a ratio of 52.5 and the digested grass waste had a ratio of 65.0. A ratio less than 0.5 is very mature, from 0.5‐3.0 is mature, and above 3 is considered immature (Brinton, 2000). The C/N of the digested residues are both less than 25. The ratio of C/N is an indicator of maturity, and using soil amendments with a C/N greater than 25 will generally cause nitrogen draw‐down and poor plant growth (Brinton, 2000).

The nutrients content of the digested wastes has the potential to replace slow‐release fertilizers in a potting mix. The ammonium content in the digested food and grass wastes (525 and 650 ppm, respectively) is much higher than the recommended limit of 25 ppm in greenhouse soils (Newman, 2008). However, the nitrate content for the two wastes is <10 ppm and 10 ppm, respectively, which is lower than the recommended minimum of 25 ppm (Newman, 2008). Aerobic processing should help correct this imbalance, since ammonia will both volatilize and be oxidized to the nitrate form. However, the negative effect of ammonia emissions during composting process should be controlled. The recommended levels of phosphorus, potassium, magnesium, calcium, and sulfur for greenhouse soils (50, 175, 4000, 500 and 50 ppm, respectively) were all exceeded by the levels in the digested food and grass waste (Newman, 2008). This problem could be corrected either by leaching, or by mixing the digested wastes with media lacking nutrients, like peat moss.

259

Table 52. Compost Suitability Parameters For The Digested And Fresh Food And Green Waste Compost Suitability Parameters

Food Waste Green Waste Percent Increase

Fresh Digested Fresh Digested Food Green

NH4-N (ppm) 973 525 630 650 -46 3

NO3-N (ppm) 118 <10 25 10 NA -60

NH4/NO3 8.3 52.5 25.2 65.0 534 -52

Total Carbon (%) 46.8 47.2 41.3 42.7 1 4

C/N 14.8 24.3 16.1 14.8 64 -8

pH (H2O 1:5) 5.78 8.65 8.05 7.62 50 -7

EC (H2O 1:5) (dS/m) NA 2.44 3.94 1.88 NA -52

8.4.3 Respirometry Experiments The anaerobic residues produced more CO2 than their aerobic counterparts (Figure 125). Each curve is the average for the three reactors for each treatment and the error bars show one standard deviation. The baseline measurements are shown for the day before the experiment. The experiment begins at day 0. Each data point represents a ten hour period.

The composted food waste, though, produced more CO2 than the digested green waste. The total cumulative CO2 evolution for each reactor was analyzed using a one‐factor ANOVA test and Tukey’s test, at α = 0.05. All treatments had significantly different cumulative evolutions, except between the composted green waste and the potting soil.

The digested food waste, composted food waste and digested green waste all reached their maximum rate of CO2 evolution within the first two days of the experiment (Figure 126). Each curve is the average for the three reactors for each treatment and the error bars show one standard deviation. The large error in the first day is due to slightly different lag times for each reactor.

The lag times, or the time before the maximum rate is achieved, varied between the three treatments. The digested grass waste had the shortest lag time, followed by the digested food waste and the composted food waste. This indicates that readily degradable compounds were present in both of the digested wastes. The rates of carbon dioxide evolution peaked twice for the digested and composted food wastes, and to a lesser degree for the digested grass waste. Microbial succession, or a change in the microbial populations, could possibly explain this result.

260

The cumulative oxygen consumption and rate of oxygen consumption are seen in Figures 127 and 128, respectively. The O2 production profiles look almost identical to the CO2 evolution, indicating that aerobic processes dominated during the respirometry experiment. The respiratory quotient (RQ), or ratio of CO2 to O2, was approximately 0.6 for the digested grass and food wastes and the composted food waste. The RQ of the composted green waste was 0.39 and the potting soil was 0.12. The RQ can vary depending on available nutrients, oxygen and the carbon substrate (Dilly, 2003; Gea et al., 2004). The RQ appears to be roughly correlated to the stability of the wastes but more research is necessary to verify this relationship.

Figure 125. Cumulative CO2 Produced For Different Materials

Figure 126. Average Rate of Carbon Dioxide Evolution (Mg CO2 C/Hr/G V.S.) Over the Two Week Experiment

261

Figure 127. Cumulative Oxygen Consumption (mg O2/g V.S) Over the Two Week Experiment for Different Materials

Figure 128. Average Rate of Oxygen Consumption of the Three Reactors for Different Materials

The maximum carbon dioxide evolution index (MCI) is shown in Figure 129, and the Dynamic Respiration Index (DRI) is shown in Figure 8.The measured MCI values were 32.8, 12.7, 7.7, 1.5 and 0.2 mg CO2 C/day/gVS for digested food waste, composted food waste, digested grass waste, composted green waste and potting soil, respectively. The measured DRI values were

262

5.7, 1.9, 1.0, 0.4 and 0.1 mg O2/hr/g VS for digested food waste, composted food waste, digested grass waste, composted green waste and potting soil, respectively

The conditions of the experiment (temperature, pH, pre‐incubation time, air flow rate, and moisture content) affect both indices and makes comparison with existing standards difficult. Wood’s End considers an MCI greater than 14 mg CO2/day/g V.S. to be unstable, and below 2 mg CO2/day/g V.S. to be very stable (Brinton, 2000). Since temperature has the largest effect on the index, and the Wood’s End tests are carried out at 34oC, these standards may be the most applicable (ADAS, 2005). Using those standards, the digested food waste is classified as unstable, while the composted food waste and digested grass waste are stable, and the composted green waste and potting soil are both very stable.

Wood’s End considers a DRI for oxygen above 2 mg O/hr/g VS to be unstable, and below 0.5 to be very stable (Brinton, 2000). By the DRI, the composted food waste is also considered unstable, while the other classifications remained the same. Using different index standards for the MCI (of >8 mg CO2/day/g V.S. is unstable and < 2 mg CO2/day/g V.S is stable), the classifications for both DRI and MCI would have been equivalent (Brinton, 2000).

The digested food waste is unstable, while the composted food waste and digested grass waste are stable, and the composted green waste and potting soil are very stable.

Figure 129. Maximum Carbon Dioxide Evolution Index (MCI) for Different Materials

Digested food waste (DFW), composted food waste (CFW), digested grass waste (DGW), composted green waste (CGW) and potting soil (PS).

263

Figure 130. Dynamic Respiration Index (DRI) for Different Materials.

Digested food waste (DFW), composted food waste (CFW), digested grass waste (DGW), composted green waste (CGW) and potting soil (PS).

8.5 Conclusions The digested food waste was found to be unstable and immature. The digested grass waste would be considered stable using respirometry techniques, but the ratio of ammonia to nitrate indicates that the digested waste was still not mature. Both of the anaerobically digested residues would need aerobic post‐processing. The high salinity is a cause for concern, but not an insurmountable problem. The heavy metal content is low enough to allow legal use in the U.S.; however the levels of nickel, zinc and copper may restrict use of the digested residues in intensive, vegetable production. The respirometry experiment provided valuable data on the stability of the different wastes and their composting potential, but evaluation of the experimental set‐up with other stability tests, like the Solvita test and the Dewar Self‐Heating test, is needed. Reproducibility, uniformity and physical properties of the digested residues are all areas worth investigation. Finally, plant growth trials are needed to fully evaluate the potential of residues as soil amendments.

264

GLOSSARY

AD Anaerobic Digestion APS Anaerobic Phased Solids AVF Alternative Vehicle Fuel BOD Biological Oxygen Demand BR Biogasification Reactor CAAA Clean Air Act Amendments of 1990 CAMD Clean Air Markets Division CARB California Air Resources Board CC Combined Cycle CDS Commercial Demonstration System CEMS Continuous Emissions Monitoring System CER Carbon Dioxide Evolution Rates CFW Composted Food Waste CGW Composted Green Waste CNG Compressed Natural Gas COD Chemical Oxygen Demand COE Cost of Energy COP California Oregon Border COT Cost of Waste Treatment CPI Consumer Price Index CPU Central Processing Unit CPUC California Public Utilities Commission CY Cubic Yard DFW Digested Food Waste DGW Digested Grass Waste EBMUD East Bay Municipal Utility District EE Energy Efficiency EGRID Emissions and Generation Resource Integrated Database EGU Electricity Generating Units EIA Energy Information Administration Energy Commission California Energy Commission EPA U.S. Environmental Protection Agency FDG Flue Gas Desulfurization FERC Federal Energy Regulatory Commission FPP Food Processing Plant FS Fixed Solids

265

FS* Food Solids FW Food Waste GC Gas Chromatography GC* Gas Chromatograph GT Gasification Tank GW Green Waste HHV Higher Heating Value HRT Hydraulic Retention Time HT Hydrolysis Tanks IEPR Integrated Energy Policy Report IPM Integrated Planning Model IPP Intermountain Power Project IRR Internal Rate of Return ISO Independent System Operator MACRS Modified Accelerated Cost Recovery System MARR Minimum Attractive Rate of Return MC Moisture Content MSW Municipal Solid Waste NCF Non Conventional Fuel NPW Net Present Worth OLR Organic Loading Rate OM Operations and Maintenance OUR Oxygen Uptake Rates PBM Pressure Balanced Manifold PDR Preliminary Design Report PM Project Manager PS Potting Soil PSA Pressure Swing Adsorption PVC Poly Vinyl Chloride RWT Reclaimed Water Tank or Effluent Water Tank SCADA Supervisory Control and Data Acquisition Sebac Sequential Batch Anaerobic Composting SQL Software (System) Query Language STP Standard Temperature and Pressure SUBBOR Super Blue Box Recycling TDP Tons Per Day TEAM Onsite Power Systems Design Team TS Total Solids

266

UC Davis University of California, Davis UCD‐RT UC Davis Research Team VS Volatile Solids

267

REFERENCES

Abad, M., Noguera, P., and S. Bures, 2001. National inventory of organic wastes for use as growing media for ornamental potted plant production: case study in Spain. Bioresource Technology 77: 197-200.

ADAS, 2005. Assessment of Options and Requirements for Stability and Maturity Testing of Composts. The Waste and Resources Action Programme, Banbury, England.

ANR, 2006. Methods of Analyses: Plant Analysis. UC Davis Agriculture and Natural Resources Analytical Lab. http://groups.ucanr.org/danranlab/Plant%5FAnalysis

APHA (1998). American Public Health Association, Standard Methods: Examination of Water and Wastewater, 20th Ed, United Book Association, Washington D.C.

Atiyeh, R.M., S. Subler, C.A. Edwards, G. Bachman, J.D. Metzger, and W. Shuster, 2000. Effects of vermicomposts and composts on plant growth in horticultural container media and soil. Pedobiologia 44: 579-590.

Bath, B. and B. Ramert, 1999. Organic household wastes as a nitrogen source in leek production. Acta Agriculturae Scandinavica Section B-Soil and Plant Science 49: 201-208.

Beck, R.W., Final report: Anaerobic digestion feasibility study for the Bluestem Solid Waste Agency and Iowa Department of Natural Resources., in Final Report: Anaerobic Digestion Feasibility Study for the Bluestem Solid Waste Agency and Iowa Department of Natural Resources. 2004, Bluestem Solid Waste Agency.

Beggs, R.D., R. Konwinski, R.H. Zhang, and P. Shaffer. Anaerobic phased solids digestion of mixed wastes. in Proceedings of Water Environment Federation’s Annual Technical Exhibition and Conference. 2007.

Blischke, J., Combining anaerobic digestion with enclosed tunnel composting. BioCycle, 2004. 45(4): p. 49-68.

Bolzonella, D., P. Battistoni, C. Susini, and F. Cecchi, Anaerobic codigestion of waste activated sludge and OFMSW: The experiences of Viareggio and Treviso plants (Italy). Water Science and Technology, 2006. 53(8): p. 203-211.

Bolzonella, D., P. Pavan, S. Mace, and F. Cecchi, Dry anaerobic digestion of differently sorted organic municipal solid waste: A full-scale experience. Water Science and Technology, 2006. 53(8): p. 23-32.

Brinton, W.F., 2000. Compost Quality Standards and Guidelines. Woods End Research Laboratory. compost.css.cornell.edu/Brinton.pdf.

BTA, BTA-technology for the utilization of biodegradable waste, in Short Information. 2007, BTA.

Cavenati, S., C.A. Grande, and A.E. Rodrigues, Upgrade of methane from landfill gas by pressure swing adsorption. Energy & Fuels, 2005. 19(6): p. 2545-2555.

268

Cecchi, F., P. Traverso, P. Pavan, D. Bolzonella, and L. Innocenti, Characteristics of the OFMSW and behaviour of the anaerobic digestion process., in Biomethanization of the organic fraction of municipal solid waste, J. Mata-Alvarez, Editor. 2002, IWA: Cornwall, U.K. p. 141-180.

Cengel, Y.A., Heat transfer: A practical approach. International ed. 1998, Boston: McGraw-Hill.

Chanakya, H.N.; Srikumar, K.G.; Anand, V.; Modak, J. and Jagadish, K.S. (1999). Fermentation properties of agro-residues, leaf biomass and urban market garbage in a solid phase biogas fermenter. Biomass and Bioenergy 16:417-429

Chavez-Vazquez, M. and D. Bagley. Evaluation of the performance of different anaerobic digestion technologies for solid waste treatment. in Proceedings of the CSCE/EWRI of ASCE Enviro. Eng. Conf. 2002. Niagara.

Cho, J.K. and Park, S.C. (1995). Biochemical methane potential and solid state anaerobic digestion of Korean food wastes. Bioresource Technology 52(3):245-253

Cho, J.K., Park, S.C., 1995. Biochemical methane potential and solid state anaerobic digestion of Korean food wastes. Bioresource Technology 52(3), 245-253.

Chong, C., 2005. Experiences with wastes and composts in nursery substrates. HortTechnology 15: 739-747.

Chynoweth, D.P., G. Bosch, J.F.K. Earle, R. Legrand, and K.X. Liu, A novel process for anaerobic composting of municipal solid-waste. Applied Biochemistry and Biotechnology, 1991. 28-9: p. 421-432.

Chynoweth, D.P., J. Owens, D. Okeefe, J.F.K. Earle, G. Bosch, and R. Legrand, Sequential batch anaerobic composting of the organic fraction of municipal solid-waste. Water Science and Technology, 1992. 25(7): p. 327-339.

Clarke, W.P., Cost-benefit analysis of introducing technology to rapidly degrade municipal solid waste. Waste Management & Research, 2000. 18(6): p. 510-524.

Dahiya, A.K. and P. Vasudevan, 1986. Biogas Plant Slurry as an Alternative to Chemical Fertilizers. Biomass 9: 67-74.

Daigle, J., and H. Gautreau-Daigle, 2001. Canadian Peat Harvesting and the Environment. Sustaining Wetlands Issues. North American Wetlands Conservation Council Committee, Ottawa, Ontario.

De Baere, L., Anaerobic digestion of solid waste: State-of-the-art. Water Science and Technology, 2000. 41(3): p. 283-290.

De Baere, L., Will anaerobic digestion of solid waste survive in the future? Water Science and Technology, 2006. 53(8): p. 187-194

DeLaclos, H.F., S. Desbois, and C. Saint-Joly, Anaerobic disgestion of municipal solid organic waste: Valorga full-scale plant in Tilburg, the Netherlands. Water Science & Technology, 1997. 36(6-7): p. 457-462.

269

Dilly, O., 2003. Regulation of the respiratory quotient of soil microbiota by availability of nutrients. FEMS Microbiology Ecology 43: 375-381.

Dupont, L. and A. Accorsi, Explosion characteristics of synthesised biogas at various temperatures. Journal of Hazardous Materials, 2006. 136(3): p. 520-525.

Edelmann, W., K. Schleiss, and A. Joss, Ecological, energetic and economic comparison of anaerobic digestion with different competing technologies to treat biogenic wastes. Water Science and Technology, 2000. 41(3): p. 263-273.

Edelmann, W., Products, impacts and economy of anaerobic digestion of OFMSW, in Biomethanization of the organic fraction of municipal solid wastes. Iwa publishing, uk, J. Mata-Alvarez, Editor. 2003, IWA Publishing: Barcelona. p. 265-301.

Eggleston, S., L. Buendia, K. Miwa, T. Ngara, and K. Tanabe, 2006 IPCC guidelines for national greenhouse gas inventories, in IPCC national Greenhouse Gas Inventories Programme, T.F.o.N.G.G. Inventories, Editor. 2006, Intergovernmental Panel on Climate Change: Hayama, Japan.

El-Shimi, S.A., M. El-Housseini, B.E. Ali, M.M. El-Shinnawi, 1992, Biogas generation from food-processing waste, Resources, Conservation and Recycling 6(1992) 315-327

EPA, 2006. Municipal Solid Waste Generation, Recycling and Disposal in the United States. EPA, Washington, DC, pp. 1-11.

EPA, Municipal solid waste in the United States: 2005 facts and figures. 2005, US Environmental Protection Agency, Office of Solid Waste: Washington, DC.

EPA, Solid waste management and greenhouse gases: A life-cycle assessment of emissions and sinks. 2006, United States Environmental Protection Agency.

Faaij,A.; van Doorn, J.; Curvers, T.; Waldheim, L.; Olsson, E.; van Wijk, A. and Daey-Ouwens, C. (1997). Characterization and availability of biomass waste and residues in the Netherlands for gasification. Biomass and Bioenergy 12(4):225-240

Firestone, R., Distributed energy resources customer adoption model technology data, E.E.T. Division, Editor. 2004, Ernest Orlando Lawrence Berkeley National Laboratory: Berkeley.

Forster-Carneiro, T., L.A. Fernandez, M. Perez, L.I. Romero, and C.J. Alvarez, Optimization of SEBAC start up phase of municipal solid waste anaerobic digestion. Chemical and Biochemical Engineering Quarterly, 2004. 18(4): p. 429-439.

Gallert, C., A. Henning, and J. Winter, Scale-up of anaerobic digestion of the biowaste fraction from domestic wastes. Water Research, 2003. 37(6): p. 1433-1441.

Garg, R.N., H. Pathak, D.K. Das, and R.K. Tomar, 2005. Use of Flyash and Biogas Slurry for Improving Wheat Yield and Physical Properties of Soil. Environmental Monitoring and Assessment 107: 1-9.

Gea, T., R. Barrena, and A. Artola, 2004. Monitoring the Biological Activity of the Composting Process: Oxygen Uptake Rate (OUR), Repirometrix Index (RI) and Respiratory Quotient (RQ). Biotechnology and Bioengineering 88: 520-527.

270

Ghosh S., 1987. Improved sludge gasification by two-phase anaerobic digestion. J Environ Engin.113, 1265–84.

Ghosh, 1977, Two phase anaerobic digestion, US patent number 4022665

Golueke, C.G., 1972. Composting: A Study of the Process and its Principles. Rodale Press, Inc., Emmaus, PA.

Gomez, R.B., F.V. Lima, and A.S. Ferrer, 2006. The use of respiration indices in the composting process: a review. Waste Management & Research 24: 37-47.

Gunaseelan, V. N. (2004). Biochemical methane potential of fruits and vegetable solid waste feedstocks. Biomass and Bioenergy 26: 389 – 399

Hamilton, W., 1985. Sulfate-Reducing Bacteria and Anaerobic Corrosion. Annual Review of Microbiology 39:195-217.

Han, S.K. and Shin, H.S. (2004). Biohydrogen production by anaerobic fermentation of food waste. International Journal of Hydrogen energy 29:569-577

Handreck, K. and N. Black, 2002. Growing Media for Ornamental Plants and Turf. University of New South Wales Press Ltd., Sydney, Australia.

Hao, X., 2003. Does long-term heavy cattle manure application increase salinity of clay loam soil in semi-arid southern Alberta? Agriculture Ecosystems & Environment 94: 89-103.

Hartman, K., An APS-digester for the treatment of organic solid waste and power generation: UC Davis digester design, Master of Engineering at University of California, Davis, Biological and Agricultural Engineering Thesis, 2004, pp. 54.

Hartman, K.M. 2004. An APS-Digester for the treatment of organic solid waste and energy generation: UCD Digester Design. M.S. Thesis, University of California, Davis.

Hartmann, H. and B.K. Ahring, Strategies for the anaerobic digestion of the organic fraction of municipal solid waste: An overview. Water Science and Technology, 2006. 53(8): p. 7-22.

Hoitink, H.A.J., 1997. Suppression of Plant Diseases by Composts. HortScience 32: 184-187.

Jenkins, B.M., Background for the energy cost calculator, California Biomass Collaborative: http://faculty.engineering.ucdavis.edu/jenkins/CBC/Calculator/index.html.

Jingru, Z., 1992. Studies on Increasing Crop Yield and Controlling Diseases By Digested Slurry and Sludge. Biogas and Sustainable Agriculture. Department of Environmental Protection and Energy, People's Republic of China, Bremen Overseas Research and Development Association, Yichang City, Hubei Province, pp. 74-82.

Johansen, J.E. and R. Bakke, Enhancing hydrolysis with microaeration. Water Science and Technology, 2006. 53(8): p. 43-50.

271

Juanga, J.P., C. Visvanathan, and J. Trankler, Optimization of anaerobic digestion of municipal solid waste in combined process and sequential staging. Waste Management & Research, 2007. 25(1): p. 30-38.

Keenan, J.H., Steam tables: Thermodynamic properties of water including vapor, liquid, and solid phases. 1969, New York: Wiley. pp. 162.

Kim, S.H.; Han, S.K. and Shin, H.S. (2004). Feasibility of biohydrogen production by anaerobic co-digestion of food waste and sewage sludge. International Journal of Hydrogen energy 29:1607-1616

Kirchmann, H. and E. Witter, 1989. Ammonia Volatilization During Aerobic and Anaerobic Manure Decomposition. Plant and Soil 115: 35-41.

Krich, K., D. Augenstein, J.P. Batmale, J. Benemann, B. Rutledge, and D. Salour, Biomethane from dairy waste: A sourcebook for the production and use of renewable natural gas in California. 2005: Western United Dairymen.

Kroyer, G.Th. (1995). Impact of food processing on the environment- an overview. LWT, 28(6):547-552

Kubler, H., K. Hoppenheidt, P. Hirsch, A. Kottmair, R. Nimmrichter, H. Nordsieck, W. Mucke, and M. Swerev, Full scale co-digestion of organic waste. Water Sci Technol, 2000. 41(3): p. 195-202.

Kwon, S.H. and Lee, D.H. (2004). Evaluation of Korean food waste composting with fed-batch operations I: using water extractable total organic carbon content (TOCw). Process Biochemistry 39:1183-1194.

Lettinga, G. (2001). Digestion and degradation, air for life. Wat. Sci. Technol. 44(8):157-176

Lissens, G., P. Vandevivere, L. De Baere, E.M. Biey, and W. Verstraete, Solid waste digesters: Process performance and practice for municipal solid waste digestion. Water Science and Technology, 2001. 44(8): p. 91-102.

Liu, G.Q., R. Zhang, and H. El-Mashad, Biogasification of green and food wastes using the anaerobic phased solids digester system. In Press (UC Davis, BAE Department), 2008.

Liu, H.W., H.K. Walter, G.M. Vogt, H.S. Vogt, and B.E. Holbein, Steam pressure disruption of municipal solid waste enhances anaerobic digestion kinetics and biogas yield. Biotechnology and Bioengineering, 2002. 77(2): p. 121-130.

Liu, T.C. and S. Ghosh, Phase separation during anaerobic fermentation of solid substrates in an innovative plug-flow reactor. Water Science and Technology, 1997. 36(6-7): p. 303-310.

Lusk, P.D., Comparative economic-analysis - anaerobic digester case-study. Bioresource Technology, 1991. 36(3): p. 223-228.

Lyum, M., Life and times of an organics recycling company. BioCycle, 1999. 40(11): p. 34.

Manios, T. and Stentiford, E.I. (2004). Sanitary aspect of using partially treated landfill leachate as a water source in green waste composting. Waste Management 24:107-110

272

Marnay, C., D. Fisher, S. Murtishaw, A. Phadke, L. Price, and J. Sathaye, Estimating carbon dioxide emission factors for the California electric power sector. 2002, Lawrence Berkeley National Laboratory: Berkeley, CA.

Martin, D., and G. Gershuny, 1992. The Rodale Book of Composting. Rodale Press, Emmaus, PA.

Mata-Alvarez, J., ed. Anaerobic digestion of the organic fraction of municipal solid waste: A perspective. Biomethanization of the organic fraction of municipal solid waste ed. J. Mata-Alvarez. Vol. 17. 2003, IWA Publishing: London.

Mata-Alvarez, J., F. Cecchi, P. Pavan, and P. Llabres, The performances of digesters treating the organic fraction of municipal solid wastes differently sorted. Biological Wastes, 1990. 33(3): p. 181-199.

McEachin, D.N. and J.S. VanderGheynst, 2006. Development of probabilistic models to predict phytotoxicity of compost amended soil from compost stability measurements; In submission.

Mehta, S.A. and S.A. Daftardar, 1984. Effects of Anaerobically Prepared Wheat Straw Composts and City Garbage Composts on Yield and N and P Uptake by Wheat. Agricultural Wastes 10: 37-46.

Menear, J.R. and L.W. Smith, Dairy-cattle manure liquid-solid separation with a screw press. Journal of Animal Science, 1973. 36(4): p. 788-791.

Moran, M.J. and H.N. Shapiro, Fundamentals of engineering thermodynamics. 5th ed. 2004, Hoboken, New Jersey: Wiley.

National Institute of Standards and, T., NIST chemistry webbook [electronic resource]. 1997, National Institute of Standards and Technology.

Newman, J., 2008. Management Practices to Protect Water Quality: A Manual for Greenhouses and Nurseries. ANR Cooperative Extension Report, Ventura County; In submission.

Nichols, C.E., Overview of anaerobic digestion technologies in Europe. BioCycle, 2004. 45(1): p. 47-53.

Okeefe, D.M., D.P. Chynoweth, A.W. Barkdoll, R.A. Nordstedt, J.M. Owens, and J. Sifontes, Sequential batch anaerobic composting of municipal solid-waste (MSW) and yard waste. Water Science and Technology, 1993. 27(2): p. 77-86.

Onovwiona, H.I. and V.I. Ugursal, Residential cogeneration systems: Review of the current technology. Renewable and Sustainable Energy Reviews, 2006. 10(5): p. 389-431.

Oppenheimer, J., 1997. Measurements of Air-filled Porosity in Unsaturated Organic Matrices Using a Pycnometer. Bioresource Technology 59: 241-247.

Pall, R., and N. Mohsenin, 1980. A Soil Air Pycnometer for Determination of Porosity and Particle Density. Transactions of the ASAE: 735-745.

Park, C.S., Contemporary engineering economics. Third ed. 2002, Upper Saddle River, New Jersey: Prentice Hall.

273

Pavlostathis, S.G. and Giraldo-Gomez, E.(1991). Kinetics of anaerobic treatment: A critical review. Crit. Rev.Environ. 21 (5/6): 411-490

Persson, M., O. Jonsson, and A. Wellinger, Biogas upgrading to vehicle fuel standards and grid injection, in Task 37 - Energy from Biogas and Landfill Gas. 2006, IEA Bioenergy: European Union.

Pirt, S.J. and Y.K. Lee, Enhancement of methanogenesis by traces of oxygen in bacterial digestion of biomass. FEMS Microbiology Letters, 1983. 18(1-2): p. 61-63.

Rao, M.S. and Singh, S.P. (2004). Bioenergy conversion studies of organic fraction of MSW: kinetic studies and gas yield-organic loading relationships for process optimization. Bioresource technology 95:173-185

Raviv, M., 2005. Production of high-quality composts for horticultural purposes: A mini-review. HortTechnology 151: 52-57.

Repa, E.W., Nswma’s 2005 tip fee survey, in NSWMA Research Bulletin. 2005, National Solid Wastes Management Association. p. 05-3.

Richard, T.L., 2004. Air-filled Porosity and Permeability Relationships during Solid-State Fermentation. Biotechnology Progress 20: 1372-1381.

RIS, Feasibility of generating green power through anaerobic digestion of garden refuse from Sacramento area, in Report to SMUD Advanced Renewable and Distributed Generation Program, R.I. Ltd., Editor. 2005, Sacramento Municipal Utilities District: Toronto.

Ross, C.C., T.J. Drake, and J.L. Walsh, The handbook of biogas utilization, ed. P. Southeastern Regional Biomass Energy. 1996, Atlanta, Ga.: Environmental Treatment Systems Inc.

Saint-Joly, C., S. Desbois, and J.P. Lotti, Determinant impact of waste collection and composition on anaerobic digestion performance: Industrial results. Water Science and Technology, 2000. 41(3): p. 291-297.

Salminen, E., J. Rintala, J. Harkonen, M. Kuitunen, H. Hogmander, and A. Oikari, 2001. Anaerobically digested poultry slaughterhouse wastes as fertiliser in agriculture. Bioresource Technology 78: 81-88.

Sarada R. and Joseph R., 1995. A comparative study if single and two stage processes for methane production from tomato processing waste. Process Biochemistry 31(4), 337-340.

Scaglia, B., F. Tambone, P.L. Genevini, and F. Adani, 2000. Respiration index determination: Dynamic and static approaches. Compost Science & Utilization 8: 90-98.

Schilstra, A.J., 2001. How sustainable is the use of peat for commercial energy production? Ecological Economics 39: 285-293.

Shen, D., 1997. Microbial diversity and application of microbial products for agricultural purposes in China. Agriculture, Ecosystems and Environment 62: 237-245.

274

Shenlin, J., 1992. Studies on the Performance of Biogas Manure for Preserving Fertilizer in the Soil and its Effect on Raising Crop Yields. Biogas and Sustainable Agriculture. Bremen Overseas Research and Development Association, Yichang City, China.

Shin, H.S.; Youn, J.H. and Kim S.H. (2004).Hydrogen production from food waste in anaerobic mesophilic and thermophilic acidogenesis. International Journal of Hydrogen energy 29: 1355-1363

Simmons, P., N. Goldstein, S.M. Kaufman, N.J. Themelis, and J. Thompson, The state of garbage in America - 15th nationwide survey of MSW management in the United States. Biocycle, 2006. 47(4): p. 26-43.

Simons, G. California's dairy power production program. AgSTAR National Conference 2004 [cited 2007 April 6, 2007].

Singh, R.P. and D.R. Heldman, Introduction to food engineering. 1984, Orlando: Academic Press. pp. xii, 306.

Six, W. and L. DeBaere, Dry anaerobic conversion of municipal solid-waste by means of the dranco process. Water Science and Technology, 1992. 25(7): p. 295-300.

Sprague, J.C. and J.D. Whittaker, Economic analysis for engineers and managers. 1986: Prentice-Hall.

Svensson, K., M. Odlare, and M. Pell, 2004. The fertilizing effect of compost and biogas residues from source separated household waste. Journal of Agricultural Science 142: 461-467.

Svensson, L.M., K. Christensson, and L. Bjornsson, Biogas production from crop residues on a farm-scale level: Is it economically feasible under conditions in Sweden? Bioprocess and Biosystems Engineering, 2005. 28(3): p. 139-148.

Ten Braummeler, E. (1993). Dry anerobic digestion of the organic fraction of municipal solid wastes. Ph.D. thesis, Wageningen University, The Netherlands

ten Brummeler, E., Full scale experience with the BIOCEL process. Water Science and Technology, 2000. 41(3): p. 299-304.

ten Brummeler, E., H. Horbach, and I.W. Koster, Dry anaerobic batch digestion of the organic fraction of municipal solid-waste. Journal of Chemical Technology and Biotechnology, 1991. 50(2): p. 191-209.

ten Brummeler, E., M.M.J. Aarnink, and I.W. Koster, Dry anaerobic-digestion of solid organic waste in a BIOCEL reactor at pilot-plant scale. Water Science and Technology, 1992. 25(7): p. 301-310.

Thassitou, P.K. and Arvanitoyannis, I.S. (2001). Bioremediation:a novel approach to food waste management. Trends in Food Science& Technology 12:185-196.

Thompson, W.H., 2002. Test Methods for the Examination of Composting and Compost. US Composting Council, USDA.

Thorneloe, S.A., K. Weitz, and J. Jambeck, Application of the US decision support tool for materials and waste management. Waste Management, 2007. 27(8): p. 1006-1020.

275

Tsilemou, K. and D. Panagiotakopoulos, Approximate cost functions for solid waste treatment facilities. Waste Management & Research, 2006. 24(4): p. 310-322.

US Patent 6,342,378. Biogasification of Solid Wastes by Anaerobic Phased Solids Digester System. Issued in January, 2002. Inventors: Zhang, R.H. and Z. Zhang

Vandevivere, P., L. De Baere, and W. Verstraete, Types of anaerobic digesters for solid wastes, in Biomethanization of the organic fraction of municipal solid wastes, J. Mata-Alvarez, Editor. 2002, IWA Publishing: Barcelona. p. 111-140.

Veeken, A. and Hamelers, B. (1999). Effect of temperature on hydrolysis rates of selected biowaste components. Bioresource Technology 69: 249-254

Veeken, A.H.M., 2005. Improving quality of composted biowaste to enhance disease suppressiveness of compost-amended, peat-based potting mixes. Soil Biology and Biochemistry 37: 2131-2140.

Verma, S., Anaerobic digestion of biodegradable organics in municipal solid wastes, Master of Science at Columbia University. Thesis, 2002.

Vermeulen, J., A. Huysmans, M. Crespo, and A. Van Lierde, 1993. Processing of Biowaste by Anaerobic Composting to Plant Growth Substrates. Water Science Technology 27: 109-119.

Verrier D , Roy. F, Albagnac G, 1987. Two-stage methanization of solid vegetable wastes. Biological Waste 22, 163-177.

Vogt, G.M., H.W. Liu, K.J. Kennedy, H.S. Vogt, and B.E. Holbein, Super blue box recycling (SUBBOR) enhanced two-stage anaerobic digestion process for recycling municipal solid waste: Laboratory pilot studies. Bioresource Technology, 2002. 85(3): p. 291-299.

Wang, J.Y.; Zhang, H.; Stabnikova, O. and Tay, J.H. (2005). Comparison of lab-scale and pilot-scale hybrid anaerobic solid–liquid systems operated in batch and semi-continuous modes. Process Biochemistry xxx, xxx–xxx

Weiland, P., Anaerobic waste digestion in Germany–status and recent developments. Biodegradation, 2000. 11(6): p. 415-421.

Wellinger, A. and A. Lindberg, Biogas upgrading and utilisation, in Task 24: Energy from biological conversion of organic waste, I. Bioenergy, Editor. 2004, International Energy Association.

Whyte, R. and G. Perry, A rough guide to anaerobic digestion costs and MSW diversion. Biocycle, 2001. 42(10): p. 30-33.

Williams, R.B., B.M. Jenkins, and D. Nguyen, Solid waste conversion: A review and database of current and emerging technologies. 2003, California Integrated Waste Management Board.

Williams, R.B., UC Davis technology assessment for advanced biomass power generation, in PIER Consultation Report. 2005, California Energy Commission: Sacramento, CA.

Withrow, W.B., Thermophilic digestion of municipal organic solid waste in an anaerobic phased solids digester system, Masters at UC Davis, Biological and Agricultural Engineering. Thesis, 2006.

276

Xu HW, Wang JY, Tay JH, 2002. A hybrid anaerobic solid-liquid bioreactor for food waste digestion. Biotechnol Lett. 24, 757-761.

Yu H.W., Samani Z., Hanson A., Smith G., 2002. Energy recovery from grass using two-phase anaerobic digestion. Waste Management, 22, 1-5.

Zhang R., El-Mashad H.M., Hartman K., Wang F., Liu G., 2007. Characterization of Food Waste as Feedstock for Anaerobic Digestion. Bioresource Technology, 98(4):929-935

Zhang R., Zhang Z., 1999. Biogasification of rice straw with an anaerobic-phased solids digester system. Bioresource Technology 68, 235-245.

Zhang, R. and Z. Zhang, Biogasification of rice straw with an anaerobic-phased solids digester system. Bioresource Technology, 1999. 68(3): p. 235-245.

Zhang, R. and Z. Zhang, Biogasification of solid waste with an anaerobic-phased solids-digester system, USPTO, Patent No. 6342378, 2002, USA.

Zhang, R., Anaerobic phased solids digester for biogas production from organic solid wastes, WIPO, Patent No. WO/2007/075,762, 2007, International.

Zhang, R.H. and Z.Q. Zhang, 1999. Biogasification of rice straw with an anaerobic-phased solids digester system. Bioresource Technology 68: 235-245.

Zhang, R.H. and Z.Q. Zhang, Biogasification of rice straw with an anaerobic-phased solids digester system. Bioresource Technology, 1999. 68(3): p. 235-245.

Zhang, R.H., 2006. Thermophilic Digestion of Green and Food Wastes with an Anaerobic Phased Solids Digester System. California Energy Commission Report.

Zhang, R.H., D. Konwinski, K. Hartman, S. Archibald, H. El-Mashed, G. Matteson. 2005. Anaerobic Phased Solids Digester Demonstration Project – Pilot Digester Design and Test Plan. Report submitted to Public Interest Energy Research (PIER) program of California Energy Commission. March 2005.

Zhang, R.H., D. Konwinski, K.M. Hartman, S. Archibald, H. El-Mashad, and G. Matteson, Pilot digester design and test plan, in Anaerobic Phased Solids Digester Pilot Demonstration Project. 2005, California Energy Commission: Sacramento. p. 43.

Zhang, R.H., El-Mashad, H., and E. Torbert, 2007. Environmental Impact Analysis for Co-digestion of Food Waste and Dairy Manure. Sacramento Municipal Utility District, Sacramento, CA.

Zhang, R.H., H. El-Mashad, K. Hartman, F. Wang, J.L. Rapport, C. Choate, and P. Gamble, Characterization of food and green wastes as feedstock for anaerobic digesters, in Anaerobic Phased Solids Digester Pilot Demonstration Project. 2005, California Energy Commission.

Zhang, R.H., H.M. El-Mashad, K. Hartman, F.Y. Wang, G.Q. Liu, C. Choate, and P. Gamble, Characterization of food waste as feedstock for anaerobic digestion. Bioresource Technology, 2007. 98(4): p. 929-935.

277

Zhang, R.H., R.T. Romano, and H. El-Mashad, Anaerobic digestion of five food waste streams (batch digestion study). 2006, Sacramento Municipal Utility District.

Zhang, R.H., Z. Zhang, 2002, Anaerobic digestion of vegetable waste with an anaerobic phased solids digester system, Transactions of CSAE 18(5), 134‐139 Zitomer, D.H. and J.D. Shrout, Feasibility and benefits of methanogenesis under oxygen-limited conditions. Waste Management, 1998. 18(2): p. 107-116.

A‐1

APPENDIX A: List of Patents, Publications, Journal Articles and Presentations

Research Contract Reports Submitted to Energy Commission:

von Langen, R. and R. Zhang, 2011. Anaerobic Phased Solids Digester Pilot Demonstration Project –Design of A Commercial Digester Plant for A Food Processing Plant. Research report submitted to Public Energy Research Program of California Energy Commission. July 2011.

Rapport, J., R.H. Zhang, D. Konwinski. 2010. Anaerobic Phased Solids Digester Pilot Demonstration Project –Engineering, Economic and Environmental Analysis of APS Digester System. Research report submitted to Public Energy Research Program of California Energy Commission. March 10.

Zhang, R., E. Torbert, R. Evans, J. Rapport, H. El‐Mashed. 2010. Biodegradability and soil amendment potential of anaerobically digested residues. Research report submitted to Public Energy Research Program of California Energy Commission. April 10.

Zhang, R., J. Rapport, D. Konwinski, I. Clark, X. Chen and S. Jiang. 2010. Anaerobic Phased Solids Digester Pilot Demonstration Project –Operation and testing of pilot digester plant with solid food waste. Research report submitted to Public Energy Research Program of California Energy Commission. April 10.

Konwinski, D., R. H. Zhang, S. Archibald, H. El‐Mashad, J. Rapport, I. Clark. 2008. Anaerobic Phased Solids Digester Pilot Demonstration Project – Pilot Digester Construction Report. Research report submitted to Public Energy Research Program of California Energy Commission. August, 2008.

Zhang, R.H., D. Konwinski, K. Hartman, S. Archibald, H. El‐Mashed, G. Matteson. 2005. Phased Solids Digester Pilot Demonstration Project – Pilot Digester Design and Test Plan. Research report submitted to Public Energy Research Program of California Energy Commission. March, 2005.

Zhang, R.H., H. El‐Mashed, W. Withrow, G. Liu. 2006. Anaerobic Phased Solids Digester Pilot Demonstration Project –Thermophilic Digestion of Food and Green Wastes with Anaerobic Phased Solids Digester System. Research report submitted to Public Energy Research Program of California Energy Commission. October, 2006.

Zhang, R. H., H. El‐Mashed, K. Hartman, F. Wang, J. Rapport, G. Liu, C. Choate, and P. Gamble. 2005. Anaerobic phased solids digester pilot demonstration project –characterization of food and green wastes as feedstock for anaerobic digesters. Research report submitted to Public Energy Research Program of California Energy Commission. September, 2005.

Patents:

A‐2

Konwinski, D. and R.H. Zhang. Fermentation System Having Strainer Assembly and Method of Continuously Producing Biogas. Provision application filed on August 1, 2006, Non‐provisional US and PCT application filed on August 1, 2007. UC Case No. 2006‐278‐2.

Zhang, R.H. Anaerobic phased solids digester for biogas production from organic solid wastes. US Patent 7,556,737. July 7, 2009.

Journal Articles:

Liu, G. R.H. Zhang, H.M. El‐Mashed and R. Dong. 2009. Effect of Feed to Inoculum Ratios on Biogas Yields of Food and Green Wastes. Bioresource Technology. 100(2009)5103‐5108.

Liu, G., R. Zhang and H. El‐Mashad. 2011. Biogas production from green and food wastes using anaerobic phased solids digester system. Biochemistry and Biotechnology. In Press.

Pan, J. R. Zhang, H.M. El‐Mashad, H. Sun, Y. Ying. 2008. Effect of food to microorganism ratio on biohydrogen production from food waste via anaerobic fermentation, International Journal of Hydrogen Energy. 33(2008)6968‐6975.

Rapport, J. , R. Zhang, B. M. Jenkins, B. R. Hartsough and T. P. Tomich. 2011. Modeling Paper. Modeling the Performance of the Anaerobic Phased Solids Digester System for Biogas Energy Production. Biomass & Bioenergy. 35(2011)1263 ‐ 1272.

Rapport, J. L., R.H. Zhang, R.Williams and B.M. Jenkins. 2011. Anaerobic Digestion Technologies for the Treatment of Municipal Solid Waste. International Journal of Environment and Waste Management. In Press.

Zhang, R. H., H. M. El‐Mashed, K. Hartman, F. Wang, G. Liu, C. Choate, and P. Gamble. 2006. Characterization of food waste as feedstock for anaerobic digestion. Bioresource Technology 98(2007)929‐935.

Zhang, R.H. 2009. Anaerobic Phased Solids Digester – Innovation for Waste to Energy Conversion. International Journal of Agricultural and Biological Engineering. 2(3).

Conference Papers and Presentations:

Rapport, J. L, E. Torbent, J. VanderGhynst, R. Evans, and R. H. Zhang. 2006. Composting and land application of anaerobically digested solids. Poster presented at ASABE (American Society of Agricultural and Biological Engineers) Annual International Meeting, Portland, OR. July.

Rapport, J., R. Zhang, B.M. Jenkins, B. R. Hartsough and T. P. Tomich. 2009. Modeling the performance of anaerobic phased solids digester system for biogas energy production. Paper presented at 2009 ASABE Annual International Meeting, Reno, Nevada, June 21‐24. Paper Number 097328. ASABE, 2950 Niles Road, St. Joseph, MI. 49085‐9659.

Rapport, J., R. Zhang, B.M. Jenkins, B. R. Hartsough and T. P. Tomich. 2009. Modeling the design and financial feasibility of an anaerobic digester producing energy from organic solid waste. Presentation at US Composting Councilʹs 17th Annual Conference and Exhibition, January 26‐29, 2009, Houston, TX.

A‐3

Zhang, R. 2010. Anaerobic Digesters for Solid Feedstock Treatment. Presentation at Anaerobic Digestion Workshop, ASABE Annual Meeting, June 19, 2010, Pittsburg, PA.

Zhang, R. 2010. UC Davis Biogas Energy Project. Invited Presentation at Green Energy‐Canada‐California Consortium Symposium, May 12, 2010. UC Davis.

Zhang, R. H. 2005. Advanced anaerobic digestion technologies for bioenergy and bioproduct production. Seminar Presented at Novozymes, Inc., Davis. June.

Zhang, R. H. 2005. Advanced anaerobic digestion technologies for bioenergy and bioproduct production. Invited Seminar, UCD IAD Program. October.

Zhang, R. H. 2006. Anaerobic digestion ‐ organics to energy conversion technology. Presentation at Workshop on Converting Organic Wastes to Energy, Sonoma County Waste Management Agency. January.

Zhang, R. H. 2006. Bioenergy production from agricultural and food wastes. Presentation at Global Environmental Change, UC Davis CAES Dean’s Advisory Council Meeting. October.

Zhang, R. H. 2006. Leftovers to lights: bioenergy technology update. Presented at Davis Explorit Science Center 2006 Lecture Series: The Cutting Edge of Science, UCD, Power Point Presentation. September.

Zhang, R. H. 2006. Organic waste to energy: research to commercial application. Presented at University of California President’s Advisory Council Meeting, Oakland, CA, Power Point Presentation. October.

Zhang, R. H., and Z. Pan. 2006. Present and future development of renewable energy in the United States. Seminar at Department of Agriculture, Shanxi Province, China, July 29, and at Chinese Academy of Agricultural Mechanization and Sciences, August 2.

Zhang, R. H., H. El‐Mashed, Z. Pan, B. M. Jenkins, and R. Evans. 2006. Biogas energy demonstration project ‐ partnership among university, industry and state government. Poster presentation at International Conference on The Future of Agriculture: Science, Stewardship, and Sustainabilit, Sacramento, CA. August.

Zhang, R. Transforming Food Processing Waste by Anaerobic Digestion into Biogas Energy. Invited Presentation at CIFAR Conference XVIII, May 13, 2010, UC Davis.

Zhang, R.H, J. Rapport, I. Clark, X. Chen, S. Jiang and D. Konwinski. 2009. Anaerobic Digestion of Food Residuals for Biogas Energy Production. Invited Presentation at ASABE Bioenergy Engineering Conference, Bellevue, WA, 10/12‐14, 2009

Zhang, R.H. 2008. Advanced Anaerobic Digestion Technologies for Organic Waste Conversion. Invited Presentation at China‐Canada‐California (CCC) FORUM on Energy and the Environment: Climate Change, Agriculture, Biorefineries/Biofuels, November 1‐3, 2008, Wuxi, China.

A‐4

Zhang, R.H. 2010. Biogas Energy Production and Utilization Technologies‐Current Status and Future Development, Keynote Presentation at International Conference On Biomass Energy Technologies, August 20 – 23, 2010, Beijing Conference Center, Beijing, China.

B‐1

Appendix B: Digester System Illustrations

B‐2

Figure 131. Pilot Digester System Site Plan

B‐3

Figure 132. Hydrolysis Reactor Vessel Detail

B‐4

Figure 133. Material Handling System for Pilot Digester System – Plan View

B‐5

Figure 134. Material Handling System for Pilot Digester System – Elevation View A

B‐6

Figure 135. Material Handling System for Pilot Digester System – Elevation View B

B‐7

P LA NT L AYOUT

2 16 F Stree t, No . 3 Dav is, CA 956 16 Te l: 55 9-270 5760 Fa x: 53 0-758 -2909

D ra wn By: S cott Arc hibald

U C DA VIS BIOG AS E NE RGY PROJECT

1 0001 00NoneS ca le: 1 o f 1

0DD av e Ko nwinskiC hec ked By:

R elea se d Da te:

D at e: 01/12/08

01/16/0801/16/08D at e:

E ng in ee r: S cott Arc hibald

D ra w in g Title:

P ro je ct:

Figure 136. Layout Diagram of UC Davis Biogas Energy Demonstration Plant

anaerobic phased solids digester pilot demonstration project · 2013-09-09 · anaerobic phased...

Documents