<%BANNER%>

Integrating Technologies in the Biodiesel Process Coupling Ultrasonication Solar Thermal Energy and Anaerobic Digestion ...

Permanent Link: http://ufdc.ufl.edu/UFE0041354/00001

Material Information

Title: Integrating Technologies in the Biodiesel Process Coupling Ultrasonication Solar Thermal Energy and Anaerobic Digestion of Coproducts
Physical Description: 1 online resource (88 p.)
Language: english
Creator: Renk, Douglas
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anaerobic, biodiesel, campus, energy, solar, ultrasound
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre: Agricultural and Biological Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Concern over high transportation fuel costs, trade deficits, depleting resources, energy security, and mounting evidence of global climate change has led to re-investigation of fossil fuel alternatives. The U.S. Department of Energy promotes bio-energy as a means to transform our nation s abundant renewable biomass resources into cost competitive, high power biofuels and products. Biodiesel offers one renewable fuel option that can be produced from waste cooking oil sources however, biofuels including biodiesel, are often criticized for low energy yield per unit of energy input and inability to reduce greenhouse gas emissions. Those claims challenge agricultural researchers to find solutions to improve yields and reduce environmental impacts. Can the coupling of biological, solar and ultrasonic technologies enhance the overall energy balance of biodiesel production? This study models and develops a system that builds on information and technology developed through prior research at UF-ABE. This project demonstrates the feasibility of producing renewable fuel from a campus waste stream. It starts with establishing methods in lab trials later tested in field applications. Construction and management of a research-scale production facility was a foray into the logistics of developing an inter-agency unit. Objectives include developing a lab for fuel and feedstock quality testing and for monitoring process. Integration of several components to possibly improve the energy balance and to reduce environmental externalities was investigated. Glycerin co-product bio-gasification, ultrasonic resonance, ethanol inputs and modeling of solar energy inputs were all studied. Improvements in mixing were shown using ultrasonication in liquid mediums. An increase in the rate of the biodiesel reaction occurred using ultrasound. Use of the glycerin co-product from biogas production can provide enough energy for all of the heat required in the process.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Douglas Renk.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Pullammanappallil, Pratap C.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041354:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041354/00001

Material Information

Title: Integrating Technologies in the Biodiesel Process Coupling Ultrasonication Solar Thermal Energy and Anaerobic Digestion of Coproducts
Physical Description: 1 online resource (88 p.)
Language: english
Creator: Renk, Douglas
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anaerobic, biodiesel, campus, energy, solar, ultrasound
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre: Agricultural and Biological Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Concern over high transportation fuel costs, trade deficits, depleting resources, energy security, and mounting evidence of global climate change has led to re-investigation of fossil fuel alternatives. The U.S. Department of Energy promotes bio-energy as a means to transform our nation s abundant renewable biomass resources into cost competitive, high power biofuels and products. Biodiesel offers one renewable fuel option that can be produced from waste cooking oil sources however, biofuels including biodiesel, are often criticized for low energy yield per unit of energy input and inability to reduce greenhouse gas emissions. Those claims challenge agricultural researchers to find solutions to improve yields and reduce environmental impacts. Can the coupling of biological, solar and ultrasonic technologies enhance the overall energy balance of biodiesel production? This study models and develops a system that builds on information and technology developed through prior research at UF-ABE. This project demonstrates the feasibility of producing renewable fuel from a campus waste stream. It starts with establishing methods in lab trials later tested in field applications. Construction and management of a research-scale production facility was a foray into the logistics of developing an inter-agency unit. Objectives include developing a lab for fuel and feedstock quality testing and for monitoring process. Integration of several components to possibly improve the energy balance and to reduce environmental externalities was investigated. Glycerin co-product bio-gasification, ultrasonic resonance, ethanol inputs and modeling of solar energy inputs were all studied. Improvements in mixing were shown using ultrasonication in liquid mediums. An increase in the rate of the biodiesel reaction occurred using ultrasound. Use of the glycerin co-product from biogas production can provide enough energy for all of the heat required in the process.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Douglas Renk.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Pullammanappallil, Pratap C.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041354:00001


This item has the following downloads:


Full Text

PAGE 1

1 INTEGRATING TECHNOLOGIES IN THE BIODIESEL PROCESS; COUPLING ULTRASONICATION, SOLAR THERMAL EN ERGY AND ANAEROBI C DIGESTION OF CO-PRODUCTS By DOUGLAS FRANK RENK A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Douglas F. Renk

PAGE 3

3 This work is dedicated to my darling daughter Jennifer, my faith and hope for the future endure in you

PAGE 4

4 ACKNOWLEDGMENTS I would like to recognize those who have helped through the twists and turns of my journey at University of Florida. I have trusted on the words of encourag ement, advice, technical support and strategies that come from everyones experiences. Thanks first and foremost to Dr. Pratap Pullammanappallil for his patience, financ ial support and academic direction. Thanks for giving me the opportunity to show my potential. When the time for focus, direction and academic legitimacy was needed, you were ther e. Without your brilliance and support, the project would not have happened. Much thanks goes to ABE department chair Dr. Dorota Haman for her adoption of this technology and new direction in our depart ment. Thanks to my chief Dr. Wendell Porter for all the hours and hours of counse ling in the office and br inging the element of reality to what sometimes seemed overly esot eric and obscure. Your sensible wealth of knowledge in energy systems is invaluable. Ther e were times where I might not have pulled through without your wisdom and encouragement. I remain most humble and appreciative to my mentor, Dr. David Chynoweth, who speaks truth to power. Your leadership in the renewable energy field le aves me awestruck. We are all grateful for the groundbreaking work that your career has left for us I remain deeply honored to have been affiliated with you. Hats off to my sponsors Jim Jaworski at Telsonic Incorporated and Ray Inman at DSI Fabrication. Your involvement ga ve the biodiesel projec t practical application that bridged academia and industry. Much thanks goes to ABE graduate advisor Dr. Ray Bucklin for direction, wisdom, advice and sometimes pulling a few strings. Thanks to committee member Dr. Ben Koopman for rounding out my educational experi ence and connecting agriculture to the planets greater environmental concerns. Thanks to Dedee Delongre-Johnston, UF Direct or of Sustainability and Suzanne Lewis, Sustainability Coordinator for Aramark Cor poration for your endorsement of this waste

PAGE 5

5 reduction effort. You are both national leaders on the path to sustainabi lity in higher education. Thanks for opening the doors with campus bureaucr ats and internal agencies and featuring this flagship sustainability effort, which gave it lo cal and national attention. Thanks to Dr. Anne Wilkie for putting everything in pe rspective and setting me straig ht when I needed it. You are my idol of strength, perseveran ce and dedication to sustainabl e bio-energy on campus. Thanks especially to Jon Priest, coor dinator of PPD Motor Pool, who has been a believer and supporter throughout this project. Much appreciation to Eric Davidson and Kathleen Tillett at UF Tax Services, without whose help we would not have established what seemed impossible at the start. Also, much thanks for all the wisdom and tech nical support from Ralph Hoffman, Larry Miller, Paul Lane and Steve Feagle. Thanks also to Jo e Thompson at University of Idaho for clutch coaching and analysis. I am grateful for the teamwork of future profes sionals that I have been so fortunate to be part of in the ABE bioprocess engineering lab. Y ou are friends for life! Thanks to Yanni, Mandu, Abhay, Sachin, Patrick,Troy, Cesar, Flora, Aaron, JD and Jaimeit takes a village! Our village would not function without the staff that pushes all the right buttons. Much thanks to Jane Elholm, Max Williams, Mary Hall, Dawn Mendoza and Jeanette Wilson for pushing my buttons! Above all I acknowledge the love of my life, Julie Garrett, a mender of the Earth, upon whose faith, nurturing and trust I relied on da ily. I am going to make you proud, sweetheart!

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................8LIST OF FIGURES .........................................................................................................................9LIST OF ABBREVIATIONS ........................................................................................................ 11ABSTRACT ...................................................................................................................... .............12 CHAP TER 1 INTRODUCTION .................................................................................................................. 14General Concerns ....................................................................................................................14Specific Problem .....................................................................................................................16Chapter Summaries .................................................................................................................172 BIODIESEL PRODUCTION FROM FA ST-F OOD OILS AND ON-CAMPUS FUEL UTILIZATION ................................................................................................................... ....19Introduction .................................................................................................................. ...........19Materials and Methods ...........................................................................................................25Materials and Equipment ................................................................................................. 25Materials and System Design ..........................................................................................26Oil collection equipment .......................................................................................... 27Pretreatment equipment ............................................................................................ 27Alcohol/ catalyst mix tank ........................................................................................ 28Main reactor .............................................................................................................28Post reaction settling tank ........................................................................................ 29Water wash tank ....................................................................................................... 29Drying tank ...............................................................................................................29Fuel storage and delivery equipment ....................................................................... 30Side stream product storage ..................................................................................... 30Methods ...........................................................................................................................30Pretreatment of the feedstock ...................................................................................31Main biodiesel reaction ............................................................................................33Post treatment of the esters ....................................................................................... 34Management of glycerin co-product ........................................................................363 APPLICATION OF ULTRASONIC RESONANCE ON OIL MIXES .................................42Introduction .................................................................................................................. ...........42Materials and Equipment ....................................................................................................... .44

PAGE 7

7 Lab Procedures .......................................................................................................................45Results and Discussion ........................................................................................................ ...46Experiment One: Tracer Dye in Water ............................................................................ 46Experiment Two: Oil Mixing .......................................................................................... 474 ULTRASOUND AND THE BI ODIESEL RE ACTION ........................................................57Introduction .................................................................................................................. ...........57Materials and Methods ...........................................................................................................59Equipment and Materials ........................................................................................................60Experimental Design ....................................................................................................... 61Sampling ...................................................................................................................... ....61Results and Conclusions ....................................................................................................... ..625 INTEGRATING ULTRASONICATION, ANAE ROBIC DIGESTION OF GLYCERIN CO-PRODUCT FROM BIODIESEL AND SOLAR PROCESS HEAT; AN ENERGY ANALYSIS ...................................................................................................................... .......67Introduction .................................................................................................................. ...........67Power and Energy Inputs ........................................................................................................69Energy Analysis of Ultrasonic Inputs .....................................................................................70Anaerobic Digestion of Glycerin ............................................................................................ 70Solar Thermal and Photovoltaic Inputs for Process Heat and Power ..................................... 73Conclusions .............................................................................................................................746 FINAL COMMENTS AND FUTURE WORK ..................................................................... 79APPENDIX: TABLES FROM MIXING EXPERIMENTS .......................................................... 80LIST OF REFERENCES ...............................................................................................................86BIOGRAPHICAL SKETCH .........................................................................................................88

PAGE 8

8 LIST OF TABLES Table page 4-1 Chart shows the amount of unreacted mate rial (m ono, di, tri) and free glycerin as grams per 100 grams of sample. ........................................................................................ 66A-1 These data are the resu lts from Ultraviolet photo-sp ectroscopic measurement of absorbance of known concentrations of dye ...................................................................... 80A-2 Data below were used for calibration curve shown in Figure two from prepared concentrations analyzed by UV photo-spectro scopy. These concentrations were used for analyzing absorbance on field samples. ....................................................................... 80A-3 Measurements of concentration of dye in samples taken over 30 minutes of time during Experiment 1.1 using pum p recirculation mixing only .......................................... 81A-4 Experiment 1.2 Pump mixing w ith ultrasonic assistance ..................................................82A-5 These data show concentrations of methylene blue dye over time from experiment 1.3 which uses ultrasonication only. .................................................................................. 83A-6 Experiment 2 Trial 2.1us es pump recirculation only. ..................................................... 84A-7 Experiment 2 Trial 2.2uses pump r ecirculation with ul trasonic assist. ........................... 84A-8 Experiment 2 Trial 2.3 uses ultrasonic r ods only, no pump .............................................. 85

PAGE 9

9 LIST OF FIGURES Figure page 2-1 Schematic of biodiesel pilot project at UF Agricultural and Biological Engineering field station.........................................................................................................................382-2 Transesterification is the biodiesel reaction. The R1, R2 and R3 represent the hydrocarbon chain of the fatty acy l groups of the triglyceride. ......................................... 392-3 Soap formation reaction; R denotes hydrocarbon fatty group ........................................... 392-4 Main reactor vessel ............................................................................................................402-5 Glycerin digesters in trailer ............................................................................................. ...413-1 The chart above plots the linear relationship of absorban ce to dye concentration. ........... 503-2 Calibration graph shows concentration of biodiesel as a tracer dye in pure soybean oil. .......................................................................................................................... ............503-3 Diagram of reactor ve ssel shows sample ports .................................................................. 513-4 Graph of absorbance versus time Experiment 1.1 sample port Atop of reactor .............. 513-5 Graph shows absorbance versus time Expe riment 1.1 from sample port Bbottom of cone. ......................................................................................................................... ..........523-6 Graph of Table 3-4 shows absorban ce versus time in experiment 1.2A. .......................... 523-7 Graph of Table 3.4 shows experiment 1.2B Bottom sample port concentrations over time .......................................................................................................................... ..........533-8 Experiment 1.3A data from table 4. Shows concentrations of dye over time ................... 533-9 Experiment 1.3B shows concentrations of blue dye present, measured over time. .......... 543-10 Experiment 2.1 uses pump mixing only. ........................................................................... 543-11 This graph plots Experiment 2.2, top and bottom ports for side-by-side comparison. ...... 553-12 This graph plots the concentration of biodiesel versus time mixed in Trial 2.3. ............... 553-13 Representation of cylindrical sound wa ves emitted between the nodes of resonating rods .......................................................................................................................... ...........564-1 Steel reactor vessel, St Johns County Florida .................................................................... 644-2 Tube resonator assemblies for reaction kinetic study ........................................................ 65

PAGE 10

10 5-1 Stoichiometric balance s hows theoretical yield of m eth ane, carbon dioxide and water products from glycerol and the product s from the combustion of methane. ..................... 745-2 This schematic shows the process flow of heat and power from biogas derived from anaerobic digestion of glycerin product. ............................................................................ 755-3 Integrated concept of biodiesel research facility with feedstock development and complete solar application. ................................................................................................ 765-4 Solar assisted methanol recovery system ........................................................................... 77

PAGE 11

11 LIST OF ABBREVIATIONS ABE Agricultural and Biological Engineering AD Anaerobic digestion BMP Biochemical methane pote ntial, determined assay Btu British thermal unit FFAs Free fatty acids GBP Glycerin-by-product HDPE High-density polyethylene KOH Potassium hydroxide LCA Life-cycle assessment or analysis NBB National Biodiesel Board NOx Oxides of nitrogen, one of three em issions related smog precursors, higher in biodiesel, mostly due to the oxygen content of the fuel PPD University of Florida Physical Plant Division STP Standard temperature and atmospheri c pressure, 0 degrees Celsius and one atmosphere at sea-level TAG Triacylglycerides or triglycerides, technical for fat and oil UVO Used vegetable oil

PAGE 12

12 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INTEGRATING TECHNOLOGIES IN THE BIODIESEL PROCESS; COUPLING ULTRASONICATION, SOLAR THERMAL EN ERGY AND ANAEROBI C DIGESTION OF CO-PRODUCTS By Douglas Frank Renk December 2009 Chair: Pratap C Pullammanappallil Major: Agricultural and Biological Engineering Concern over high transportation fuel costs, trade deficits, depleting resources, energy security, and mounting evidence of global climate change has led to re-investigation of fossil fuel alternatives. The U.S. Department of Ener gy promotes bio-energy as a means to transform our nations abundant renewable bi omass resources into cost competitive, high power biofuels and products. Biodiesel offers one renewable fuel option that can be produced from waste cooking oil sources however, biofuels including biodiesel, are often cri ticized for low energy yield per unit of energy input and inability to reduce greenhouse gas emissions. Those claims challenge agricultural researchers to find soluti ons to improve yields and reduce environmental impacts. Can the coupling of biological, solar and ultrasonic technologie s enhance the overall energy balance of biodiesel product ion? This study models and develops a system that builds on information and technology developed through prior research at UF-ABE. This project demonstrates the feasibility of producing renewable fuel from a campus waste stream. It starts with establishing methods in lab trials later tested in field applications. Construction and management of a research-sca le production facility was a foray into the logistics of developing an inte r-agency unit. Objectives include developing a lab for fuel and feedstock quality testing and fo r monitoring process. Integrat ion of several components to

PAGE 13

13 possibly improve the energy balanc e and to reduce environmental ex ternalities was investigated. Glycerin co-product bio-gasificati on, ultrasonic resonance, ethanol inputs and modeling of solar energy inputs were all studied. Improvements in mixing were shown using ultrasonication in liquid mediums. An increase in the rate of th e biodiesel reaction occu rred using ultrasound. Use of the glycerin co-product from biogas production can provide enough energy for all of the heat required in the process.

PAGE 14

14 CHAPTER 1 INTRODUCTION The Agricultural and Biological Engineering Departm ent of UF has been involved in waste-to-energy research such as cellulosic ethanol and biomethane technology for several decades and much of this research has been rev italized in the latest wave of projects underway. UF Biodiesel was conceived in the July 2006 and was ultimately sized to produce 500 gallons per month of biodiesel from used cooking oil from the dining halls, food courts and sport arenas for over 60,000 students, faculty and staff (Figure 2-2). The UF Motor Pool successfully uses the fuel produced in vehicles and equipment. The pr oject was designed primarily as a demonstration and research program to investigate the chemical, biological and indus trial process, train researchers on use of analytical equipment for fu el quality, and develop safety procedures for small-scale producers of biofuels. General Concerns Geo-political issues are at the heart of m ovements toward homegrown fuels like biodiesel. The history of federal policy directed towards energy independence extends several decades. The first Energy Security Act was signed into law in June 1980 shortly afte r market upsets during conflicts in the middle-eastern oil producing nati ons. This act paved the way for domestic coal gasification, geothermal and solar energy technologies. Howe ver, the projects were soon thereafter discontinued after world oil prices fe ll. With the signing of the first Energy Policy Act (EPAct) in 1992 came provisions for both supply and demand-side management for many energy sectors. The EPAct of 1992 offici ally recognized biodiesel as a transportation fuel and legally defined its chemical composition. The mandate of th is act was to displace 30% of the petroleum used in the U.S. by 2010. The EPAct was amende d in 1998 to grant cred it to those using B20

PAGE 15

15 (20% biodiesel blend) in vehi cles as meeting the requireme nt equivalent of purchasing alternative fuel vehicles. This action set th e course for the expansion of the industry. President Bushs February 1, 2006 State of th e Union Address declared we are a nation addicted to oil and mandated increased ener gy independence. Concerns surfaced over supply capacity of domestically produced fossil fuels and led to congressional enactment of the Energy Independence and Security Act of 2007. This legislation includes provision for alternative fuels, fuel economy and diversification of transportation fuels. The biodiesel goal was to provide a 5% blend in all diesel fuel produced. First generation biofuels, which use readily available feedstoc k, are produced through conventional means, such as biomass, biogasification of waste products, and conversion of food crops such as corn, sugar cane and soybeans. The aim is to quickly provide alternative transportation fuels that utilize existing infrastructu re such as pipelines, terminals, vehicles and equipment with little to no m odification. Another advantage is the availability of hundreds of years of agricultural t echnology and yield optimization for ro w crop production. The objective is to wean our economy off of traditional fossil-base d fuels and pave the course for more advanced fuels that show more promise for meeting greater needs. Funding was included in the Energy Independe nce and Security Act of 2007 (EISA) for EPA studies to assess impacts of bio-based fuels. Many of those reports are being presented at the time of this publication. The greater con cern comes after criticis m of energy benefits (Pimentel et al., 2005, 2009)) and revi ew of the viability of biofuels. Here lie considerable policy considerations whether or not to move in th is direction. Policymakers might question the prudence to continue investment incentives for food-based biofuels. Land-use changes associated with large-scale monocultures negate the goal of reducing greenhouse gas emissions (EPA,

PAGE 16

16 2009). Burning and clearing, and habitat destruc tion are most notable in Brazil and Indonesia. However in less developed regions where agricultural mechanization and local petroleum supply is unstable, use of locally grown biofuels will increase yields far beyond the land cost for this fuel production (World Watch Institute, 2006). The inte ntion of this research is to investigate and quantify the application of vari ous techniques to help maximize energy available from vegetable oil as a fuel source by conservi ng energy within the process. Specific Problem Typical vegetable oil-to-biodi esel conversion plants are energy intensive operations. The m ain energy requirements are process heat and electricity to power pumps, centrifuges and controls. On site raw materi al production and downstream processing of waste streams is sometimes called closing the loop, where manage d systems are vertically integrated through the supply chain to meet the needs of production. The National Renewable Energy Laboratory defines a biorefinery as a facility that integrates biomass convers ion processes to produce fuels, power and chemicals. The power production redu ces costs and avoids greenhouse gas emissions related to the refini ng process itself. Three subsystems were analyzed for integr ation into a proposed biorefinery prototype: solar photovoltaic and thermal pow er, anaerobic digestion of biodi esel glycerin co-product and ultrasonic wave resonance for possible enhanced chemical reaction. Florida, with its warm weather, high rainfall and long growing season is poised to become the nations leading biomass producer. Climatic advantages give Florida an edge for solar energy production and anaerobic digestion, an engineered biologi cal process for production of biogas, a natural gas substitute. Ultrasonic resonance, common in many chemical dispersion applic ations, is new to biodiesel production. Ultrasonic sound waves, generated at a certain range of frequency, create high shear forces in fluid mixtures. This is a potential clean energy input.

PAGE 17

17 Biodiesel production is primarily an engineered chemical pro cess, yet this study presents rationale for integrating mechanical and biological inputs. As far as literature reveals, this system is unique. Can the coupling of biological, solar and ultrasonic technologi es enhance the overall return of energy on the energy invested in the production of biodiesel? Can these technologies reduce need for traditional fossil fuel inputs? This study tests and evaluates the appropriate integration of these technologies in smallscale production applications. Substa ntial effort was put in to th e instigation of new biodiesel research at Agricultural and Bi ological Engineering (ABE). Assembly of a research scale unit south of campus provides a field location to demonstrate biodiese l production and allows us to collect and bio-gasify residual glycerin produced in the reaction. Lab pr ocedures for quality analysis were established based on previous literature and studies. Feedstock and fuel refinement techniques were also simulated in the lab setting and later fiel d-tested. Ultrasonic tube-shaped resonating rods were installed in the reactor. Mixing and reaction ra tes were observed and evaluated. Chapter Summaries The second chapter reports on th e course of action followed for biodiesel production from used fast-food oils and its use on campus. Information collected here makes recommendations whether or not a sustained effort in collection and processing of this resource is feasible. Included are design strategies a nd rationale for environmentally friendly integrations. Chapter 3 explains the experimental design of ultras ound in tandem with common mixing practices in small batch reactors. Chapter 4 explores the applicability of sonochemistry in biodiesel production for purpose of affecting reaction rates of conversion. This could prove relevant to industry if throughput is increased significantly by ultrasound. The final chapter strives to show any value to the entire study in terms of energy savings by co upling ultrasound technology with

PAGE 18

18 delivery of solar process heat and energy recovere d from glycerin, a co-pro duct of biodiesel that is becoming a disposal issue. Research associates at UF/ABE have performed solar and glycerin studies, in conjunction with this project. This study would be beneficial to any group contemplating biodiesel research in and operation of a small-scale facility.

PAGE 19

19 CHAPTER 2 BIODIESEL PRODUCTION FROM FA ST-F OOD OILS AND ON-CAMPUS FUEL UTILIZATION Introduction The path to the creation of th e reacto r design was cleared from a collection of sources. The authors initial interest was pi qued by a wide variety of sources both academic and otherwise. A fringe interest in domestic energy self reliance and decentralization in th e United States has led to open-source support networks within a commun ity of small-scale bi odiesel enthusiasts. Although some ideas were initiated through biod iesel community networks, academic studies were necessary for scientific verification. This study focuses on peer-reviewed journal sources. The exploration of concepts re quired investigation into stud ies performed at key research institutions in the United States such as Na tional Renewable Energy La b (NREL), United States Department of Agriculture (USDA) and studies at Iowa State, North Carolina State, and University of Idaho to name a few. These artic les were reviewed to consider advantages and disadvantages of various process designs and to build on an established base of industrial knowledge. Chapter 2 of this thesis stre tches the boundary of pursuit of research knowledge to execute direct action towards sustainable resource res ponsibility on campus. William Kemp in his book, Biodiesel Basics and Beyond, covers the spectr um on energy issues, global and local, which provides rationale for pursuing biodiesel production for farm and fleet use (Kemp, 2006). The book discusses the environmental and economic cons equences and advantages that need to be clearly thought through as this st udy unfolds. Ethical concerns are the compass on this path. The journey begins with a schol arly look into the past and how the industry has come to this crossroad. One review disclosed an understandi ng into the history of fuel from oils and its evolution into modern biodiesel (Knothe et al., 1997). This work gave direction as to the most

PAGE 20

20 customary and reliable means of production, given th e typical raw material resources available in the U.S. The understanding of the physical and ch emical properties of oils, such as cold-flow, shelf life and importance of complete conversion presented serve as an important reference for managing feedstock and system design for Florida conditions. All aspects of establishing a functioning program were consid ered. One thesis looked into the feasibility of an Alachua County operated biod iesel plant (Grant 2003). Much like all fuel production, biodiesel is resource limited. There was much applicable information for the availability of nearby resources for continued success. A noteworthy discovery was that local officials and mechanics alike would embrace cha nge in operations for the goals of reduced emissions and local fuel security. Inquiries at the university Office of Sustai nability yielded support and direction for establishing biodiesel use on campus. The path le d to directors at food services, the business college and the campus motor pool fleet director The idea was welcomed and a strategy was put in place to proceed. After a planning team was established, information was found which serves as a general overview for poten tial producers of biodiesel who are contemplating entering the biodiesel business (Van Gerpen, 2004). A sound gr asp of the growth potential and the limits to growth was necessary. Issues such as material acquisition and proper storage and handling, workplace safety, are addressed. The book wa s useful for learning tax and regulatory requirements and operational concerns. The book e xplained the limits to production. Currently, supply constraints limit biodiesel to a niche market. Essentiall y, there are not enough oils and fats produced to make a dent in the demand for di esel fuel in the U.S. Sixty billion gallons of diesel are used annually, forty billion on th e highway, while only 4.6 bi llion gallons of fats,

PAGE 21

21 grease, and oils produced domestic ally. However, this amount coul d supplant all of the diesel fuel needed for domestic agricultural mach inery and railroad applications (EIA). Understanding that used cooki ng oils on campus was the give n feedstock, this publication first created the awarene ss of contending with inadvertent soap formation as a side reaction in the process (Van Gerpen, 2004). Prolonged use of the campus cooking oil and the thermal breakdown of fats to free-fatty acids (FFAs) nece ssitated a search for academic material that spelled out procedures for remediating broken do wn fats. Acid catalyzed methods discussed in literature were used in experi ments and early field trails (Canakci and Van Gerpen, 2001) Methods learned enabled us to recover broken down fats and convert to us eful biodiesel product. As lab trials were conducted, more detailed pr ocedures for removal of FFAs through alkali refining were taken from Baileys Industrial Oil and Fat Products (Bailey et al., 1979). A prototype reactor was constructed that allows flexibility to integrate attachments or expand capacity of the system if needed. Va n Gerpen (2005) and NREL Biodiesel Business (2004) presented typical schematics. The system was modified as lab trials demonstrated safer and more effective components. Concepts for batch systems and continuous operations were reviewed. Batched systems operate in a closed sy stem where all of the ingredients are added in the beginning and nothing is removed until comp lete. Continuous processes allow material to enter and leave the system, ideally at steady rates. Van Gerpen (2005) describes how small producers prefer batch reactors du e to relative low startup cost while Peterson (2002) provides consideration of a continuous flow system. Lessons learned through these previous trials led to design concepts used in a third incarnation of the UF prototype unit. This unit functioned as a semi-continuously operating system, where some product was removed during batch operation

PAGE 22

22 while other material was simultaneously m oved downstream through multiple refinement processes. A means to convey liquids through the system was needed. Work disclosing particular reactor design was pivotal in choices for react or design (Canakci and Van Gerpen, 2001). Pumps served not only as mode of transfer, but also mimics the mixing method embraced by Iowa State University in their batch 190-liter pilot plant. This paper also provide d a remedy for re-reacting sub-spec fuel to meet quality standards for total glycerin. Another element of standard procedures gathered from these trials was the two-stage base transester ification reaction, which will be explained in more detail in system design and methods section. In the first stage, 75% of the reactants are delivere d, partially reacted and allowed to se ttle. A fraction of glycerin product is decanted and then the remaining 25% of the meth anol reactant/ catalyst mixture is added in the second stage. The reaction equili brium shifts further towards th e product side, in turn producing quality fuel low in unreacted glycerides. Earlier accounts describe the need for pur ifying the crude biodies el to remove any contaminants and debunk any myths to the cont rary (Kemp, 2006). American Society of Testing Materials (ASTM) defines criteria for quality fu el that meets the definition of biodiesel under federal law (Van Gerpen, 2004). After realizing the need for post treatment of biodiesel, the issue of remediating production by-products becomes evident. Some previously established methods for water-mist extraction and the need for water to be pure and warm to enhance solubility of soap in the water (Canakci, 2001). Feasibility st udies show cost of production can be competitive with petroleum distillate fuel only if the value of methanol recovery and purification of glycerin co-product is calculated into the business plan (Grant, 2003) Planning beyond the monetary bottom line includes capturing energy va lue from residues in the process.

PAGE 23

23 Quality assurance was paramount to convince th e research team and fleet operators that locally produced biodiesel was a suitable alternative fuel in campus equipment. Van Gerpen (2004) briefly laid out analytical methods that helped set up lab procedures for testing for soap content, Water present limits the reaction rate in the proce ss (Komers, 2002; Canakci, 2001), Tracking soaps indicates relative water formation. The need for precise quantification of water content led to purchasing a Karl Fisher titrator unit testing water content in the finished product as well as in alcohol and oil r eactants. To determine the reacti on completion, reference is given to the American Oil Chemists Society (AOCS) method Ca 14-56 Free, Combined and Total Glycerol; Iodometric and Periodi c Acid Method from the Official Methods and Practices of the AOCS 1998. Safe storage and transport of bi odiesel has been promoted by the National Biodiesel Board (NBB), which overs ees industry standards (NBB, 2009) With map and compass in hand, the journey begins. On Campus Sustainability Day October 2005, University of Florida President Dr. Bernie Machen announced a new challenge for faculty, staff and students. He envisioned a waste-free campus in ten years; an even broader pursuit was launched to reduce wast e on campus. One of the lower-hanging fruit in this effort is conversion of used cooking oil on campus to biodiesel. Biodiesel is an agricultural-based, renewable fuel for use in all compression ignition engines and home heating oil applications. It can be produced from preand post-consumer fa t, oil and grease. It holds great promise as a ready substitute for petroleum di esel and can be mixed at any pe rcentage with petroleum diesel to form a biodiesel blend. The University of Florida Agricultural a nd Biological Engineering Department (ABE), in cooperation with multiple departments, groups and clubs on campus, has produced nearly 2000 gallons of biodiesel fuel fo r use in the Physical Plant Divisions (PPD) vehicles and equipment.

PAGE 24

24 The literature reviewed and the lab experiments conducted set the course for simple process design in the UF pilot project (Figure 2-1). The common biodies el reaction or basecatalyzed transesterification, described in simplest terms, reacts any fat, oil or grease, commonly referred to as triglyceride (TG) with an alcohol, such as methanol in the presence of an alkaline catalyst such as lye with added heat to pr oduce both biodiesel, known as fatty esters and glycerol. The balanced stoichiometric reaction is shown in Figure 2-2. The development of a functioning system re quires disciplinary know ledge in chemistry, fluid dynamics, and thermodynamics, together with engineering, process controls, management and construction skills to coordi nate. The engineering challenge in any new technology is taking an idea from the bench-scale to ve rify lab results in p ilot setting. In an effort to attempt scale-up of lab experiments, an objective for phase one of the biodiesel project wa s to build a prototype unit. This also allowed the waste oil on campus to be managed in-house while producing fuel for the UF PPD Motor Pool. Moneta ry compensation for the fuel helped provide funding for continued research. PPD agreed to accept up to 1600 gallons/ month making a 20% blend (B20) with petroleum diesel. This re search program enabled experime ntation with components that could be installed and tested for process improve ments or energy efficiency such as solar and biogas additions. With considerations of limited start-up cap ital, the project was in itiated as a volunteer effort prior to authors academic pursuit. The intent was to institute a reasonably functional unit that would meet immediate demonstration n eeds while allowing the project to commence. During which time new funding opportunities were explored. As a demo nstration unit, the project successfully informed students of all ages, clubs and bu sinesses of the possibilities for alternative transportation fuels. Th e project also helped bring nati onal attention to sustainability

PAGE 25

25 in higher education. The University of Florida s efforts were featured in Sierra Magazines cover story Ten Coolest Schools on College Green Guide (Hartog and Fox, 2008) and The Chronicles of Higher Educations lead story, In Search of the Sustainable Campus (Carlson, 2006). Materials and Methods The Material and Methods portion of this chapter is divided into three sections: A materials and equipment list, brok en down by component or ingredient A description of the design of each com ponent and how the materials were used Methods attempted and the path to final choices of process Materials and Equipment Feedstock c ollection One-ton 5x9 trailer with 10-HP diesel centrifugal pump, two-inch pipes, ball valves, 5-inch diameter strainer basket on flex ible PVC hose, excludes particles greater than 1.5 cm diameter, 275-gallon capacity tote Sixteen 55-gallon colle ction drums with lids at dining f acilities, Blue with identifying labels for used cooking oil Settling and storage Two 275-gallon settling totes, 6000-gallon main storage tank, 500-gallon day tank storage for preheated, pretreated oil, -HP 110 VAC Chicago transfer pump Pretreatment Filtration: 40-mesh intake screen on settling ta nk transfer, 2 FSI sock style filters, 2-inch inlet/outlet 700-micron and 100-micron particles excluded Dewateringheat evaporated 1500Watt pipe-plug immersion heater Caustic strip vessel: 115-gallon carbon st eel cylindrical vess el, 1500-Watt electric heating tape Alcohol/ catalyst mix tank 40-gallon DSI custom fabricated steel mix tank and stand with stainless steel basket insert, -HP magnetic-drive recirculation pump, -inch steel pipes, valves and fittings Main reaction vessel 115-gallon capacity HDPE conical bottom reactor, 1/2-HP magnetic drive Little Giant 35 gallon/ minute pump Post reaction separator Elevated 275-gallon settling tote with bo ttom drain, floating outlet to wash process

PAGE 26

26 Wash tank Insulated 55-gallon drum, heating tape stri ps four-nozzle mist apparatus, FisherScientific air stone, aquarium pump with 3/16 tubing, -HP Ch icago centrifugal transfer pump Drying tank Insulated 55-gallon drum, Ogden 1500-Watt thermostatically c ontrolled pipe-plug immersion heater, -inch steel piping and valves Residue storage Two 275-gallon bulk container totes Glycerin digesters Insulated 190-liter food grade wide mouth barrels, Materials-reactants Methanol, Brenntag MidSouth (99.8%) Potassium Hydroxide flakes, O xyChem 90% minimum alkalinity Sodium hydroxide flakes Chem One Ltd 98.0% min Sulfuric acid, technical grade 93% Water, de-ionized, filtered Used vegetable oil, mixed Materials and System Design A com plete system schematic is shown at the end of this chapter (F igure 2-1). Materials used for the unit were often recycled, donated or re-used items to demonstrate the manner in which the community scale biodie sel processor can be assembled from salvaged or otherwise readily available material. Another goal of th is unit was to minimize the use of grid-supplied electric. Use of gravity for material flow and se ttling was incorporated wherever possible into the feedstock pretreatment, and intermediate stages of refinement. This sustainable mission guided material and method selections toward those th at tend to reduce energy requirements in the process, re-use salvaged equipment and recycle all side streams and co-products. The glycerol product, referred to in this work as glycerinby product (GBP), can contain residual soaps, excess alcohol reactant, catalyst, water and trace organic compounds. To a lesser degree, these substances can exist in the ester layer and are treated as contaminants that must be removed downstream.

PAGE 27

27 Oil collection equipment The first used vegetable oil (UVO) was co llected fro m campus in September of 2006. Coordinating with GDS and restaurant managers 16 55-gallon drums were painted, labeled and placed at 7 dining facilities across campus. A portable self-contained collection unit was designed that allowed the collect ion of up to 275 gallons of oil (Figure 2-3). A 10-HP diesel powered centrifugal trash pump, operated on 100% biodiesel and was fixed on a one-ton utility trailer, plumbed to a 275-gallon HDPE bulk-shipping container with necessary hoses and valves. A 5-inch cylindrical basket diffuser was attached at the end of the main 2-inch suction hose to vacuum oil from the drums. Three-way valves, enabled use of the same pump for emptying of the collection tank. Pretreatment equipment Two 275-gallon settling tanks were set up for the first step of oil clean-up. This was sized to allow time for water and food par ticles to settle out of the used oil. A scum layer formed at the top of the oil. A special floating intake from these settling tanks was devised to permit only the cleaner oil, located just below the surface, to be strained with a 40 mesh screen. The oil then moved through a check-valve by wa y of HP electric transfer pu mp and -inch piping sending the strained oil to the top of the main 6000-gallon holding tank Two large capacity FSI sock filters cartridges were purchased to use for filtering of f ood particles. To prevent frequent fouling of filters, recirculation of the UVO through the f ilters was done during warm weather only when the fats are less viscous. The filters were to be arranged in series, 700-micron followed by a 100micron unit. This is adequate for pretreatment, as most of the finer particles are removed in ester post-treatment.

PAGE 28

28 The field station made use of a 110-gallon steel open top tank, as a cauldron for mixing caustic material, such as sodium hydroxide, to bi nd FFAs as soap salts of FFAs (Figure 2-3). The industrial term for this is alkali refining (Bailey et al., 1979) or caustic strip (Van Gerpen, 2004). A -inch PVC pipeline routs the filtered UVO to the soap reaction tank. The need for preheating is dependent on seasonal temperatures. Heating tape was wrapped around the tank to bring temperature to about 22 0C for winter processing. Alcohol/ catalyst mix tank Mixing of dry flaked catalyst with methanol oc curred in a custom-built 40-gallon steel mix tank with stand designed and assembled by DSI Fa brication. The unit featured a stainless steel basket for holding the flaked potassium hydroxide catalyst, while the magnetic drive chemical pump delivers the alcohol from the storage drum to the top of the mix tank to dissolve the flakes. The volatile gases formed were vented and extreme caution was practiced when handling these chemicals. The tank was fashioned with a calibrated sight tube fo r measuring alcohol. Main reactor The m ain types of reactors in the industry ar e batch reactors and plugflow reactors that function similarly to mini-batch style reactors (Turner, 2005). Continuously fed stirred tank reactors (CSTRs) are effective only in two-stage systems, whic h remove product between stages (Van Gerpen, 2005). We utilized tanks and drums that served as batch reactors, most commonly used in pilot or research scale settings. Through most of the course of the project, the main biodiesel reaction occurred in a 115-gallon insulated HDPE tank (Figure 2-4), with hp magnetic drive pump for both material transfer and vigorous mixing. The conical bottom allowed for easier draining and separation of the denser glycerin product. Sight tubes allowed the operator to observe color changes between phases. They were placed in the discharge lines when

PAGE 29

29 gravity drained to the storage totes downhill. Th e esters could be further reacted in a second stage to assure a complete reaction. Post reaction settling tank The ester layer was pumped to the center of the adjacen t settling tank to allow time for gravitational force to remove soaps, glycerin, wate r and sediment. The purpose of this stage is to minimize the amount of time and water needed in the subsequent refinement processes. After settling, a thin wax layer often formed on the top, so the design of a floating outlet was devised to exclude this layer and also prevented the disr uption of the settled glyc erin co-product residue on the bottom. From this tank, the esters are sent to the wash process. A manifold with multiple -inch ball valves was designed to allow the flex ible use of a single -HP transfer pump for washing, drying and storage. Care was required to avoid cros s-contamination. Water wash tank The rem oval of any remaining trace contaminan ts takes place in the 55-gallon heated and insulated wash drum. This step is critical for meeting fuel quality specifications. A mist apparatus was built from -inch pipe with fogge r mist nozzles deliver pur e water to the top of the esters, The unit was shaped in a cross pattern to lay flat across the top of the tank. Following mist application, a solvent resistant air stone was used for continuous overnight air bubbling, which leaves the fuel polished and almost clear. Drying tank A final drum tank was designed for drying the es ters. The e sters we re heated through a pipe-plug immersion heater while recirculated within the drying drum. The heated esters were sprayed tangentially to the upper inner surface of th e open drums. This spraying of the hot esters allowed maximum surface area for water evaporation. The drying process required

PAGE 30

30 approximately one hour from the tim e the oil is heated to about 70 0C depending on humidity and ambient temperature. Fuel storage and delivery equipment The dry fuel was filtered through a 10-micr on Goldenrod cartridge filter and a second Goldenrod water separator unit on the discharg e hose. The finished material was metered through a fuel-rated, mechanical flow meter and stored in a 175-gallon hi gh-density polyethylene (HDPE) fuel tank until full and sent across campus to the UF PPD Motor Pool. Side stream product storage Glycerin-by-product was stored in 275-gallon HDPE totes a nd delivered by peristaltic pum p to tank bioreactors. Methane recovered is a valuable natural gas substitute. Floating drum collectors were used for gas collect ion and U-tube type meters were used for calculating rates of production. Methods All standard m ethods were performed first in co ntrolled lab environment and verified to be repeatable in field trials. The base-catalyzed trans-esterificatio n reaction is the most common and predictable method. The first batches were made from lightly used restaurant oil donated by local affiliates. A 6:1 molar ratio of methanol to triglycerides, twice in excess, was found in literature to be most effec tive for complete conversion (F reedman, 1984; Van Gerpen, 2005). This amount is about 20% of total mixture we ight. An amount of potassium hydroxide (KOH) catalyst about 1% of total weight is dissolved in the methanol reactant and delivered to the heated oil. Additional KOH was added to FFAs. If left un-neutralized, the FFA s will bind the free catalyst. Formation of soap and water hinders the reaction until the KOH is no longer available to function as a catalys t (Fig 2-3). An acid/base titration method was used to determine the acid value, which is the amount of titrant solution necessary to neutralize FFAs. The method,

PAGE 31

31 a modified ASTM version, was adopted from the laboratory experiments in Biodiesel Analytical Methods (Van Gerpen, 2004). The concentration of titrant used most frequently was 0.017 normal. The formula used for determining acid value with units in ml KOH/gram sample is: [(A-B)*N*56.1]/W, where: A=number of ml of titrant to neutralize sample B=amount needed for the blank N=normality of the titrant solution W=weight of the sample in grams If KOH was the base titrant, then the percent FFA is approximately equal to half this value. Pretreatment of the feedstock During spring and fall class sessions, bi-week ly pick-up was required to allow adequate storage capacity at the dining halls. E arly in th e project, cold-flow/ visc osity problems greatly hindered pumping. Additional challenges arose when oils sampled from the local dining halls showed high oxidation and food contamination. The high volume of activity at these restaurants showed the used cooking oil to contain no le ss than 5% FFAs and often nearly 4% water. Committed partners to the campus waste reduction initiative learned to assume the role of used cooking oil collector and renderer. The material was offered to the project at no cost, but the pretreatment of the oil required coarse straining, filtering, dewatering and removal or repair of broken down fats. Discussion was opened to work with the Gator Dining Services to manage oil filtration at the dining facility and type of oil used and frequency of oil change in fryers. Preheating, straining and filtering, described ea rlier in Materials a nd Equipment, were practiced in the lab using MilliP ore filtration to remove sediment and suspended food particles larger than 10 microns that might interfere in the reaction. Dissolved food colloids showed no adverse affect on the biodiesel reaction. In addition, when fuel is combusted in vehicles, an aromatic quality is produced ch aracteristic only of esters from used vegetable oil (UVO).

PAGE 32

32 The first pretreatment reaction attemp ted in production was an acid-catalyzed esterification first stage followed by a base-cataly zed second stage. The goal is to increase the yield while simultaneously lowering the acid value of the UVO. Directions for this pretreatment process result from literature (Canakci, 1999). Methyl ester product forms from FFAs with the addition of methanol at a 20:1 molar ratio metha nol to FFAs in the presence of a sulfuric acid catalyst. This method successfully produced the fi rst biodiesel with UVO from the dining halls. Prior to this many emulsificati on problems were experienced due to the water formed in the neutralization of FFAs (Fig 2-4). Acid esterification reduced the FFA content from as high as 7% down to less than 1% in most cases. The amount of catalyst was adjusted between 0.05 grams to 0.1 grams of sulfuric acid per gram of FFA. Catalyst concentration of 0.1 grams per gram produced the best reduction of FFAs as tested with standard titr ation. This method was used for production from March through June of 2007. Although exceptional results were achieved, the process had th ree major drawbacks. First of which is logistical concern; the reactor vessel was occupied for twice as long as the single stage reaction and the heat requ ired for the reaction doubled th e electrical energy usage. Secondly, the reaction produced particles that were prone to restricting th e pipeline and scaling temperature probes and heating element. The th ird and most significant drawback was higher material cost for the excess amount of methanol required for the reaction when compared to the value of the net increase in yield. Commercial applications practici ng this method always retrieve excess methanol via reflux distillation to increase cost effectiveness (Zhang et al., 2003). An alternative pretreatment reaction implem ented was a saponification or deliberate soap making of the FFAs, sometimes referred to as alkali-strip method. In this process, a neutralizing solution of potassium hydroxide (KOH ) or sodium hydroxide (NaOH) in de-ionized

PAGE 33

33 water is gently folded into the high FFA oil, w ith the hope of forming salts of long-chain fatty acids (soap) with the most charged particles in the UVO, namely the FFA molecular chains. This reaction usually requires no added heat and can be performed in a simple cauldron type vessel with little additional plumbing, thus freeing up the reactor vessel for only the main reaction. There is also no need for dewatering of the oil pr ior to this reaction, as a lye/ water solution is added anyway. The co-product, soapstock, from this process, is a fine semi -liquid soap that was investigated for possible uses. A bout 18% of feedstock is bound to the soapstock in this process and is inseparable by settling of the soap prec ipitate in the soap reaction vessel. About 48-50 gallons of UVO are needed to yield 38-40 gallons of biodiesel. The soap salts can be broken using hydrochloric or sulfuric ac id solution to recover the FFAs and neutral oil (Bailey et al., 1979). Recovered FFAs can be late r converted to esters. Other benefits of this method of pretreatment include cost savings due to lo wer electrical consum ption and no methanol requirement. Main biodiesel reaction The pretreated oil was transfe rred to the m ain reactor vessel and preheated by recirculation of the oil past a thermostatically contro lled pipe-plug immersion heater up to 70 0C. This continued for several hours to allow any entrained water to ev aporate to atmosphere. Water levels were reduced to less than 500 parts per million or 0.05% by volume. The potassium hydroxide (KOH) catalyst is soluble only in the alcohol and dissociates throughout the mixture. The KOH is a homogeneous catalyst and is intended for a single use. Literature recommends about 6 grams of KOH catalyst pe r liter of oil reacted or appr oximately 0.5% by weight for reactions performed at 60 0C (Van Gerpen, 2005). The biodiesel reaction is a heterogeneous reac tion affected by mass tr ansfer properties and concentration gradients characteristic of the reactants used. The reac tion starts with two

PAGE 34

34 immiscible liquids, the alcoholic catalyst solution and oil. As the reaction proceeds from triglycerides (TG) to esters, the two distinct phases become a single phase momentarily and then two phases develop once again as the polarity changes. At exact stoichiometric ratios of reactants, 3:1 molar ratio of al cohol to TG, the reaction proceeds until a dynamic equilibrium occurs. At 6:1 molar ratio, twice the theoretical amount of alcohol needed, the reaction proceeds nearly to completion (94%) in a single stage at 60 0C, yet still short of meeting fuel standards. The main reaction of the oil was determined to be most complete when conducted in two stages where product is removed after a period of reaction (Canakci, 2001). In the first stage, the reaction is conducted with 80% of the alcohol solution, af ter draining some of the co-product, the remaining 20% of the reactants ar e added, this time the KOH catalyst is lowered to 0.1% of total concentration to minimize inadvertent hydrolyzing of the esters to soap due to traces of water usually present. Due to the low solubility of the glycerin in esters, the two phases generally separate quickly (Van Gerpen, 2005). The denser glycerin can be decanted to storage by gravitational force while observing color change through si ght tubes when ester layer is reached. Post treatment of the esters Al most nothing could more quickly end a progr am designed to merge research with fleet operations than equipment failure from a shipment of poorly processed fuel. To prevent engine damage, the downstream processing is essential fo r producing quality fuel. Water is the universal solvent. The contaminants remaining in the fuel are more soluble in water than the biodiesel. A cross-shaped misting apparatus placed on a wide -top tank allowed gentle application of a warm water mist to the crude esters. The water woul d settle to the bottom and decanted to a holding tank that held the used water for remediation. Af ter three applications of mist water, a final continuous wash, called bubble-wa sh, was applied by sparging air from the bottom through an

PAGE 35

35 added layer of warm, clean, water. The esters would need to be kept at minimum of 400 C to enhance solubility. The water-coated air bubble s rise to the surface and collapse. Soaps, methanol, catalyst or free glycerin dissolved in the water settle back to the bottom. Thorough mist washes are important for removing larg e amounts of soaps prior to this bubble wash; otherwise the bubble agitat ion here could create an emulsi on difficult to break. Bubble washing was normally done twice, first for about two hour s, then drained, and followed by an overnight polishing; this produced very clean fuel, almo st clear. Washing is the most time and power consuming, yet critical part of the process as heat is applied throughout process residence time. Industry leaders question airbubbling practice due to oxidativ e breakdown concerns (Van Gerpen 2007). During oxidative breakdown, the long fa tty chains of the esters are reduced to shorter chains, usually where double carbon bonds occur. Oxidative stab ility, tested by the Rancimat Oxidative Stability method, is of great concern for long-term storage. If fuel on campus were to be stored for many months, an anti-oxidant additive would be recommended. Typically any biodiesel derived from used cooki ng oils (Pahgova et al., 2008) requires the addition of an oxidative stabilizer. Additives are sold under multiple brand names. Many biodiesel producers are adop ting dry wash for the refine ment of the esters instead of water wash and ester drying method. Dry washi ng uses an adsorbant or ionic exchange resin beads to draw out contaminants. Dry washing also eliminates the need for water removal from the esters, an energy intensive step. Water cons umption is always a concern in fuel production (Gerbens-Leens, 2009); more of a concern is pro duction and remediation of contaminated water. Optimization of contact for contaminant abso rption, which could minimize water consumption, was another related study in this project (see appendix). Using so ap neutralization techniques and ionic exchange methods, the water can be purif ied and reused. Further study is needed to

PAGE 36

36 determine the energy costs of water purification as compared to mining or synthesizing of magnesium silicate commonly used as an adsorban t in the dry wash. Life-cycle assessments (LCA) for Mg SiO remain to be seen, but uncertainty as to the mining impact and disposal issues for a project touted as a wast e-reduction initiative directed th e project toward water wash. After the water wash, the fuel is very sa turated with almost 2000 parts per million of water, which must be removed almost entirely to avoid fuel de gradation in storage and engine fuel filter problems. The esters are heated to 65 0C and sprayed to allow the water to evaporate until levels of less than 500 part s per million are recorded. Heat inputs necessary in this evaporation process contributed to the energy requirements. Management of glycerin co-product For each gallon of biodiesel produced r oughly two pounds of glycerin are produced. Mismanagement of glycerin co-product might end a project quicker th an poor quality fuel delivery, this time with possible fines or prison time. Glycerin is an environmentally benign substance, but methanol, in any concentration, in the glycerin renders the entire solution an Flisted EPA hazardous waste (Kemp, 2006). The cr ude glycerin by-product is only about 50% glycerol with the remaining fractions consisting of water, soaps and residual methanol and food particles. Awareness of the abundan ce of glycerin in the production of biodiesel and Btu value of the glycerin was the original impetus for this project. The intent was to utilize methanol contaminated glycerin through an aerobic digestion (AD), an e ngineered biological process. Methanol and glycerin are 100% biodegradabl e to carbon dioxide, methane and water through AD. Complex substrates, such as soaps, FFA s and oils are also biodegradable in AD (Angelidaki, 1998). Anaerobic digestion produces biogas, typically 60% methane, and functions as a natural gas substitute. The glycerin by-pr oduct requires no pretreatment to serve as feedstock in AD. With additional nutritional suppor t, the glycerin is an excellent carbon source

PAGE 37

37 for the anaerobic micro-fauna. The glycerin dige sters (Figure 2-5) were located in a portable demonstration trailer where simple biogas techno logy could be showcased at alternative energy expositions. Experiments were performed in the lab to show the use of biogas to neutralize the pH of the wash water produced in the extraction process. Biogas, which can contain almost 40% carbon dioxide, can bond with potassium or sodium from the soap salts in the wash water forming potassium or sodium carbonates. These trials de termined that enough biogas can be produced to neutralize all of the wash water formed, but th ere was not enough wash water produced to scrub all of the carbon dioxide from the biogas. Scr ubbing or removing the carbon dioxide from biogas increased the Btu value by raising the pe rcentage of methane in the biogas. Experiments were conducted to show the us e of glycerin as a solvent to extract contaminants from the biodiesel without the co ncerns of water use pr eviously mentioned. Glycerin is readily availabl e and can be recovered and re used, therefore minimizing or eliminating the need for further post treatment. Soaps, methanol and water are always found in higher concentrations in the glyc erin layer. Benefits include no need for separate methanol recovery system for the esters and minimal need for adsorbants such as silicates, or ionic exchange resins. Extraction expe riments were conducted in small batches, but ideally, a countercurrent packed column liquid-liquid extraction could provide continued contact between the esters and purified glycerin. Although the energy required for methanol and water removal from the glycerin solvent may compare to energy needed for direct removal of excess methanol from the esters, the need for additional evaporator equipment is avoided.

PAGE 38

38 The project functioned for near ly three years, produced ove r 2000 gallons of high quality fuel, and set the course for future work. A visi on of a fully integrated process utilizing all resources on site was developed during this time. Figures Figure 2-1. Schematic of biodiesel pilot project at UF Agricult ural and Biological Engineering field station

PAGE 39

39 Figure 2-2. Transesterification is the biodiesel reaction. The R1, R2 and R3 represent the hydrocarbon chain of the fatty acyl groups of the triglyceride. In their free form, fatty acids have the configuration shown belo w where R is a hydrocarbon chain greater than or equal to 10 carbon atoms Figure 2-3. Soap formation reaction; R denotes hydrocarbon fatty group

PAGE 40

40 Figure 2-4. Main reactor vessel

PAGE 41

41 Figure 2-5. Glycerin di gesters in trailer

PAGE 42

42 CHAPTER 3 APPLICATION OF ULTRASONI C RESONANCE ON OIL MIXES Introduction Ultrasound has been used for a variety of pur poses that includes areas as diverse as communication with anim als (dog whistles), the detection of flaws in concrete buildings, to the synthesis of fine chemicals and the diagnosis and treatment of di sease. The work presented here explains time-honored methods of enhanced mixing or improved chemical reactivity through ultrasonic technology and test ing new applications to the biodiesel industry. The ability of substances to mix, dissolve or disperse in one a nother is dependent on physical properties and intermolecular forces between solids, liquids or gasses. At room temperature, liquids with similar structures, de nsity or polarity can be mixed in all proportions within a contained space. In contrast, the biodie sel reaction starts with two immiscible liquids: alcoholic homogeneous catalyst solution and feed stock oil. As the reac tion proceeds step-wise from triglycerides to esters, th e two distinct phases of alcohol and oil will behave as a single emulsified phase (Boocock et al., 1998) for some time and then two phases develop once again as the polarity changes during the formation of final ester and glycerin phases. The biodiesel reaction is a heterogeneous r eaction affected by ma ss transfer propertie s and concentration gradients characteristic of th e reactants used. In any liquidliquid reaction, the size of the interfacial area is critical in determining the rate of mass transfer. Agitation must be adequate to ensure the methanol/methoxide droplets are small, therefore increasing surface area and improving contact with the unreacted oil. The reacti on between these species occurs only in this interfacial region between liquids (Collucci et al., 2005). Thes e chapter three tests are simplified as the effect of mass transfer properties is di sregarded for the sake of only investigating the macroscopic effect of ultrasonic acous tic mixing in a recirculated tank.

PAGE 43

43 Mixing techniques used by some small-sc ale producers are often accomplished by recirculating with a centrifugal pump where high shear is produced by the pump impellor. Shear in fluid dynamics is produced when two liquid st reams flow perpendicula rly to one another. Centrifugal pump impeller blades produce turbulent kinetic energy and recirculation within the cells of the impeller (Anagnostopoulos, 2006). Biodies el mixing has been cleverly described as being shaken, not stirred as the material from the bottom is brought to the top during recirculation as if it were turned repeatedly up side down. This is important to prevent the two phases, with markedly different densities from settling or separating in any way during the reaction. Laminar flow occurring through simple cylindrical tanks is referred to as channeling and is to be minimized in a mixed or stirred r eactor to prevent areas of inactivity in a vessel known as dead zones. In chemical engineering, pract ical use of the term mixing can be thought of in more than one context. Mixing throughout the vessel, distri bution versus dispersion, shear mixing occurring at a molecular level. Homogenization of liquids in a vessel by ultrasound is called acoustic stream mixing in sonochemistry (Colucci et al ., 2005) while molecular shear-mixing has been explained by the phenomenon known as cavitati on in the ultrasound industry, which provides a localized microscopic effect. Sonochemistry is a branch of chemical study and application pertaining to the effect of u ltrasonic sound waves in chemical reactivity. No information was found in literature searches for the applicati on of ultrasonic resonance for mixing of oils throughout a reactor vessel for batch production of biodiesel. Tracer experiments for detecting changing concentrations of substances over time and at different locations in the vessel were set up to simulate the introduction of chemical reactants needed for the biodiesel reaction. The series of tracer tests was designed to observe ultrasounds

PAGE 44

44 affect on the presence of miscible substances or the ability to reduce the time needed to achieve homogeneity throughout the vessel. The mixing experiments were c onducted first with aqueous so lutions and secondly with oil solutions. The work utilized methylene blue dye and biodiesel (esters) as a dye with its characteristic auburn color, measured at various concentrations in water and virgin soybean oil respectively through use of u ltraviolet photo-spectr oscopy. In the second set of mixing experiments, predetermined quantit ies of biodiesel (esters) were used for testing oil mixing. Both sets of experiments were divided into three tr ials. Each experiment te sted traditional pump mixing alone in the first trial, traditional pum p mixing with ultrasonic assistance in the second trial and ultrasonic resonance alone in the third trial. The ultraviolet photo-spectrometer allowed us to compare concentration of material at absorbed at signature wavelengths. Absorbance of light ( ) varies inversely and exponentially to transmittance ( T ) of light in a homogenous substance. (II / I0) represents the comparison between the initial measurement in the cuvette (IO) and the concentration (c) after some time (II ) (Braslavsky and Houk, 1988). Methylene blue has a strong absorption band cen tered at 660 nanometers, in the red region of the visible spectrum, and transmits wavele ngths below 600 nanometers, bestowing a blue color to the dye (Green, 1990). Materials and Equipment Mixing and reaction vessel is an insulate d 115-gallon HDPE cone-bottom vessel equipped with a HP-12 gallon per m inute Chicago Pump, a 1500-Watt Ogden pipe-plug immersion heater, two 25 kilo-Hertz Telsoni c ultrasonic resonating rods with generators. Rods are usually

PAGE 45

45 sized proportionally to re actor as prescribed for ultrasonic cleaning vessels, typically about 10.5 Watts/liter. The initial trials in this work tested effectivenes s at 7.1 Watts/liter. Manufacturer representatives recommended that the returning liquids from the pump exhaust just below the surface to avoid splashing. Splashing could dissolve gasses from the headspace into the liquid below. Dissolved gasses inte rrupt the flow of sound wave s through the liquid, therefore decreasing the effectiveness of the ultrasound. Samp le ports at two regions of the vessel were chosen; top center (A) and bottom return port (B) (Figure 3-3). An alytical equipment used was a Milton/Roy 21D ultraviolet photo-spectrometer. Lab Procedures Lab procedures established standards and cal ibration curves f or quantifying dye and ester concentrations in future field samples. The plot of prepared concentrations of methylene blue dye (Figure 3-1) shows a linear relationship of abso rbance at their respectiv e wavelengths to known concentrations of dye up to the point of saturation. The concen tration of methylene blue at saturation is where the photo-spectrometer is unable to detect higher concentration. Concentrations above this sa turation limit are no l onger linear in relation to absorbance. Biodiesel produced from campus used cooking oil possesses a characteristic amber hue where absorbance corresponds most closely to the 450 nanometer range of the visible light spectrum The calibration curve plotting concentration of biodiesel in pure soy oil versus absorbance also shows a straight-line rela tionship. Homogenous mixing, given measured quantities of the substances, gives concentrations at known absorbance levels along the line (Figure 3-2). The slope/ intercept formula allows us to solve for th e concentration of biodiesel in our field samples when absorbance is measured. No saturation limit was found with high concentrations of biodiesel. Absorbance increas ed up to concentration of pure biodiesel.

PAGE 46

46 Results and Discussion Experiment One: Tracer Dye in Water The first experim ent in this work, labeled Experiment 1.1, used conventional mixing technique of centrifugal pump agitation. 1710 m illigrams dry weight, of the concentrated methylene blue dye dissolved in water was introduced at the t op of the reactor (t0) into 90 gallons (342 liters) of water. The expected final c oncentration, if homogenous, would be 5 mg/liter for all three water/ dye trials. Data was collected at pr ecise time intervals in all trials. Power settings on the generator were relative to amplitude; full power was applied. Experiment 1 showed that soon after the dye was introduced, a highe r concentration was found at the bottom of the vessel and never homogenized despite a half hour of mixing with the pump (Figure 3-4). One might have expected that the absorbance measured in top and bottom would equilibrate after some time and approach the concentration of 5 mg/ liter, which is found at absorbance close to 0.48. Ta bles with values for graphs can be found in the appendix. If absorbance increases with concentration, then one can dedu ce from the results that the concentration is higher at the bo ttom of the tank. Therefore, insuffi cient mixing occurs at the top of the reactor. An unexplained sudden drop in co ncentration occurs at both the top and bottom after 14 minutes. The chart shows little variation in concentra tion before and after this; no improved uniformity is apparent. The dye material seems to immediately dissipate to lower part of tank and not adequately retained at the top. The second set of water/ dye trials, labeled Experiment 1.2, used typical pump mixing, but this time with ultrasound assistance. This set of mixtures allowed comparison of data from conventional over-the-top pump mixing alone in Experiment 1.1 with the combined pump and ultrasonic assist.

PAGE 47

47 Figure 3-6 shows a high concentration of dye occuring at the bottom after 15 seconds following the introduction of the dye materi al, indicating channeli ng through the vessel. Continued recirculation dissipa tes the tracer. Figures 3-6 an d 3-7, with ultrasound, show a 14 percent improvement in the final concentration of dye in the upper layer when compared to that same region in Figure three, pump only. Although uniformity throughout vessel is not achieved, there is a 5.2% higher concentr ation in the upper vessel region on average in Trial 1.2 with ultrasound and pump agitation as compared to Trial 1.1 with pump mixing alone. This shows a degree of imrovement with ultrasound. The final run of the water/ dye trials made use of the ultr asound equipment onlyno pump was run during this third and final test. The only forces at work here were gravity, as the dye settles and applied ultrasonic ener gy. These data show the most c onsistent blend over time. After the first 5 minutes of fluctuati on, a nearly consistent concentration develops with no fluctuation thereafter. Information gathered from these graphic repres entations is limited to only two regions of the vessel, but gives a clear idea of material transport through the ve ssel and shows adequate side-by-side comparisons. This experiment f unctioned as a good exercise in detection of dissolved solid material movement in the vessel, but may have limited application for comparison of oil mixing as it relates to the biodiesel reaction, a liquid/ liquid mixture. Experiment Two: Oil Mixing Adequate mixing of reactants is crucial for effective chem ical processes. To determine degree of mixing at various zones in the vessel, we needed to be able to detect substances that would simulate the introduction and infusion of reactants into the oil feedstock. Since dyes mixed do not bind and create new compounds as in the biodiesel reaction, compatible materials were chosen, capable of forming a uniform homogenous blend, thereby simulating the single-

PAGE 48

48 phase emulsion that occurs in a thoroughly blended oil/alcohol mixture. This uniform mix occurs readily with biodiesel and vegetable oil. Experiment 2 was also divided into three mixing trials. Trial 2.1 used pump mixing only, Trial 2.2 used pump and ultrasonication, and Tr ial 2.3 used ultrasonic ation only. The 115-gallon reactor was filled with 60 gallons of purified biodiesel. Each tr ial added 15 gallons of pure food grade soybean oil, creating a weak er dilution of biodiesel for each subsequent trial. The expected final concentrations of biodiesel and oil, if fully homogenized, should correspond to the level of light absorbance at that con centration when measured by UV photo-spectroscopy. The expected final concentrations for Experiment 2 Trials 2.1, 2.2, and 2.3, are 80%, 67% and 57.2% respectively. Concentrations of biodiesel in oil were measured from samples taken over three hours of mixing. There was a considerable difference in resu lts depending on the method or combination of mixing methods used. The biodiesel remained more concentrated at the bottom in Trial 2.1 with the pump mixing. An initial 100% co ncentration at the bo ttom of the vessel (series 2) lowered to 80% initially, but increased over time beyond the expected concentration of homogeneity. Even though the biodiesel layer was retu rned to the top by the pump, the biodiesel concentration at the top of the vessel did not improve. Experiment 2, Trial 2.2, shows a more uniform blend at both the top and bottom sample ports. An interesting note is that the biodiesel remained more concen trated in the upper section of the vessel than when compared to the trial wi th the pump only. The biodiesel was at the bottom of the vessel prior to the introduction of more soybean oil and was intr oduced to the top by the recirculation pump. The chart shows more co-ming ling of the two liquids from the onset of the mixing.

PAGE 49

49 Experiment 2, Trial 2.3, showed no vertical mi xing between the laye rs after the initial addition of the soybean oil. Measurement of the concentration of bi odiesel shows how the soybean oil appears to have immediately settled to the bottom of the reactor vessel. The top layer showed a higher concentra tion of biodiesel throughout the three-hour trial. To conclude, observations showed significant channeling occurred in the vessel. None of the trials achieved the expect ed concentrations that woul d indicate thorough homogenous mixing. However, Trial 2.2, where pump mixing was coupled with ultrasound, showed the most consistency throughout the tank. Th e resonating rods are designed to send sound waves laterally, which reverberate off the sides of the tank, perpendicularly to the rod and the laminar flow of the liquid through the vessel. There are five nodes spaced along the length of the rod that emit circular waves of sound (Figure 3-13). Trial 2.2 s howed how this lateral movement affected the transport of material by ultras ound. The lateral movement appeared to break the channeling of fluid stream and prevented the se ttling of higher concentrations of dye toward the bottom region of vessel as observed in Trial 2.1. This effect best exemplified the acoustic stream mixing described earlier, where material is divert ed perpendicularly to the flow of liquid. Many lessons were learned in these mixing experiments. Shortcomings of our process design were identified and corrected in the proc ess. A need was recognized for most vigorous agitation or mixing to reach a homogenous state quicker. A new pump was installed for future studies; specific volume flow rate increased from 12 to 37 gallons/ minute. The return line on the pump was modified to create a more even distri bution of reactants. A revised manifold delivery recirculates the liquid to mu ltiple points in the tank. A 2.5 0C increase in temperature per hour was noted throughout the experiments. Since this work began operati ng at room temperature, this

PAGE 50

50 effect was not of concern, but allowances were made for heat additions in Chapter 4 kinetics studies with ultrasound. Figures: Figure 3-1. The chart above plots the linear relationship of absorb ance to dye concentration. The Y-axis represents the absorbance of the methylene blue in absorbance units. The concentration of methylene blue in mill igrams/ liter is shown on the X-axis. Figure 3-2. Calibration graph shows concentration of biodiesel as a tracer dye in pure soybean oil.

PAGE 51

51 Figure 3-3. Diagram of reactor ve ssel shows sample ports A, top (Series one on oil mix graphs) and B, bottom (Series 2 on oil mix graphs), ultrasound rods, and pump placement for recirculation. Figure 3-4. Graph of absorbance versus time E xperiment 1.1 sample port Atop of reactor

PAGE 52

52 Figure 3-5. Graph shows absorbance versus time Experiment 1.1 from sample port Bbottom of cone. Absorbance is a function of the con centration of dye at the bottom of vessel Figure 3-6. Graph of Table 3-4 shows absorban ce versus time in experiment 1.2A. Samples taken from top sample portshows concentr ation in upper portion of reactor vessel

PAGE 53

53 Figure 3-7. Graph of Table 3.4 shows experiment 1.2B Bottom sample port concentrations over time Figure 3-8. Experiment 1.3A data from table 4. S hows concentrations of dye over time, samples taken from top of reactor

PAGE 54

54 Figure 3-9. Experiment 1.3B shows concentrations of blue dye present, measured over time. Samples taken from bottom of reactor Figure 3-10. Experiment 2.1 uses pump mixi ng only. Series one graphs the biodiesel concentration from samples taken from top sample port. Series 2 represents samples from the bottom port. The plot of these da ta from Table 3-6 represents time versus changes in concentration of biodiesel in oil

PAGE 55

55 Figure 3-11. This graph plots Experiment 2.2, t op and bottom ports for side-by-side comparison. Experiment 2.2 used mixing pump with th e ultrasound assistance. Final expected concentration is 67%. Figure 3-12. This graph plots the concentration of biodiesel versus time mixed in Trial 2.3. Only ultrasonication was used to mix. Expected final concentration 57.2%

PAGE 56

56 Figure 3-13. Representation of cylindrical s ound waves emitted between the nodes of resonating rods (courtesy of Telsoni c promotional literature)

PAGE 57

57 CHAPTER 4 ULTRASOUND AND THE BI ODIESEL RE ACTION Introduction In sim plest terms, the most typical industr ial practice for produc ing biodiesel reacts triglycerides (fats, oils, grease) and alcohol such as methanol in the presence of an alkaline catalyst at 60 0C to produce mono-alkyl fatty esters and glycerin. (Figure 22) Can the addition of ultrasonic resonance have a significant effect on the convers ion of oil to esters? Typical advantages to ultrasound include : no moving parts to wear out or maintain, it provides a small amount of heat that can make up for any heat loss during the reaction and the unit has a long lifespan. The equipment is design ed to operate for minimum of 6000 hours and can be operated continuously. Rods are usually designed with th icker metal for continuous reaction. The lifespan is expected to be longer in oil mixes than with wa ter. Solids or particles in the feedstock should be well filtered to prevent wear on metals. Industrial ultrasonic app lications exist for producing emulsions from normally immiscible substances such as mixing tolu ene with water (Mason and Cintas, 2002). These applications pertain to similar transport tribul ations with vegetable oil and alcohol in the biodiesel process. Power ultrasound, the trade term for sound waves occurring in the range of 20-25 kilohertz, produces its affects through the phenomenon of cavitation. Like any sound wave, ultrasound is transmitted via waves which a lternately compress and stretch the molecular structure of the medium through which is passe s. During each stretching phase (rarefaction), provided that the negative pressure is strong enough to overcome intermolecular binding forces, a fluid medium can be literally torn apart pr oducing tiny cavities or mi crobubbles. These bubbles are formed from dissolved gasses in the liquid me dium, such as methanol or water vapor or atmospheric gasses. In succeeding compression cycl es, these cavities can collapse violently with

PAGE 58

58 the release of large amounts of energy in the immediate vicinity of the microbubbles. It has been estimated that temperatures of up to several t housand degrees Kelvin and pressures of several hundred atmospheres are produced during this collapse. The shock wave produced on bubble collapse can disrupt solvent structure, such as th e methanol in the biodie sel reaction and this can influence reactivity by alte ring solvation of the tr iglycerides. (Mason, 1991) The mechanical and chemical effects of the collapsing bubble will be felt in two distinct regions, first within the bubble itself which can be thought of as a high pressure micro-reactor, and secondly in the immediate vicinity of the bubble where the shockwave produced on collapse will create enormous shear forces (Mason, 1991). The cavitation bubble can be considered a nano-reactor. This work in this paper was designed to test the hypothesis of Colucci et al. (2005) at a production scale. Colucci et al. (2005) conclude d in bench-scale experi ments that ultrasonic resonance reduced energy of activation needed in the biodiesel reaction. Th ey report reaction rate constants three to five times higher than that reported in literature for mechanical agitation. The paper however relies entirely on these comparisons of typical react ion rates cited in literature, rather than comparing the effect of ultras ound in their specialized 250-milliliter reaction chamber. Data were also sent from sponsor Telsonic laboratories depicting biodiesel trials with ultrasound. These previous lab-s cale ultrasound experiments reac ted small batches, 600 ml and 1000 ml, using various feedstocks. Methods used for determining effect of ultrasound on the reaction included measuring and comparing viscosit y, percent yield of biod iesel and glycerin-byproduct by weight and volume. Although viscosity can be an indicator of reaction completion if using identical feedstock, it does no t give a complete picture of th e degree of completion of the reaction (Knothe et al., 1997). Also the figures for the yields of esters and glycerin do not reflect

PAGE 59

59 other substances contained in the respective layers. The differe nt phase layers can contain unquantified amounts of residual metha nol, water, soaps and catalyst. The extent of the literature seems to stop at the bench scale and complications occur in trying to reproduce work at the production level due to the diffe rent shapes of vessels and positioning of the transducers with respect to the base (Mason and Cintas, 2002). Repeatability is essential for rigorous hypothesis te sting and validation of trials. Materials and Methods The experim ent was designed to allow quantific ation of the completion of the reaction over a period of time. Comparison of samples analyzed from oils reacted with or without ultrasound will show the effect of the ultrasonication on the reaction kinetics. Bound glycerides are noted as fuel contaminants and indicate quantities of unreacted oils. An alytical methods tried include AOCS Method Ca 14-56 test for free and to tal glycerin, NIR spectroscopy and gas chromatography method ASTM-D6584. Parameters to be measured: Changes of rate of conversion at diffe rent amplitudes of sound waves provided by controller unit. Any measurable temperature change from th e application of the ultra-sonic mixers Settling time Testing for any difference in soap formation, an undesirable side reaction Energy consumption analysis will be prepar ed for comparison with conventional mixing Can this experiment, conducted at the small production level, produce results comparable to bench scale work performed in 2005 (Colucci et al., 2005), which showed a decrease in energy of activation?

PAGE 60

60 We are assuming the feedstock contains 100% triglycerides. It is possible for oils to inherently contain small fractions (less than 0.5%) of mono and di-gly cerides. The biodiesel reaction occurs step-wise as the triglycerides r eact with the methanol to form methyl esters, while the glycerides are reduced to pure glycer ol. The reaction kinetics are complicated by the reversible nature of each step. For the sake of simplicity, our analysis wi ll consider the rates of change from triglycerides to esters. The overall biodiesel reaction most closely follows pseudosecond-order rates of reacti on (Colucci et al., 2005): Where k is the overall rate constant, Ea is the Arrhenius energy of activation required, R is the ideal gas constant, and T is abso lute temperature. A is the pr e-factor or frequency factor. This model shows the reactions temperature dependence and how a catalyst, which lowers the activation energy can a ffect rate of reaction. Ultrasound, if found to have a significant effect on increasing the reaction rate would therefore increase throughput and he lp minimize reactor size and capital costs. Equipment and Materials Ultrasound equipm ent, provided by Telsonic Co rporation includes two 1500-watt model TI 25 48X tube resonators, maximum frequency 25 kilo-Hertz (kHz) each with generator model ECO20/25/30/40XX. Reactions in these trials were conducted using 120-gallon carbon steel tanks at St Johns County Florida Public Works biofuel production (Figure 4-1). Steel tanks are more successful for reflecting sound waves, wh ere plastic tanks can absorb sound energy. For use with flammable or explosive liquids such as methanol, the manuf acturer recommends the transducer be fully immersed and the double-insulated high volta ge power cable sleeved in a steel casing (Figure 4-2) The re flectance of sound propagates greater cavitation effect (Jaworski).

PAGE 61

61 The manufacturers suggested sizing for the ro ds to reactor working volume is 40 watts per gallon of working volume, which in our case is approximately 110 gallons. Given a total 3000 Watts of total power demand from the equipment, the ratio is about 27 Watts per gallon. A -HP magnetic drive centrifugal pump provided ma terial delivery and recirculation. Oil used in the process was from recycled cook ing oil pretreated via sulfuric acid-catalyzed esterification to nearly 1% free-fatty acid. Th e acid value, titrated to first permanent phenolphthalein indication, was 2.2 mg of sodium hydroxide per gram of sample. Additional catalyst was added in the first stage of the basecatalyzed transesterification to neutralize these free fatty acids to form soap salts of long-ch ain fatty acids and water (Figure 2-3). Univar provided technical grade methanol greater than 99.8% pure. Sodi um hydroxide was chosen as catalyst, supplied at 0.4% of oil weight. Experimental Design The trials w ere designed to compare reacti ons with and without the assistance of ultrasound mixing. Samples were ta ken precisely at ten-minute intervals in a two-stage base catalyzed transesterification r eaction. A central processing unit and chemical metering pump monitored and controlled identi cal side-by-side reactors. Vo lume and type of feedstock, temperature, pump agitation and reaction times we re consistent to meas ure only the effect of ultrasound. Sampling It is n ecessary to halt the reaction at the time of sampling to obtain a true snapshot of the extent of completion of the reacti on at preselected time intervals. There are four possible ways to prevent the reaction from proceeding: freeze be low favorable reaction temperature, boil off methanol reactant, add water to bind catalyst, and neutralize base catalyst with acidic solution. In the field conditions, we chose the water method of stopping the reaction. It serves as the quickest

PAGE 62

62 way to halt the reaction and still permit to analyze soap formation. Use of acids to neutralize the catalyst could inadvertently break the soaps, a si de reaction we are tryi ng to measure. However the addition of water to halt th e reaction will inadvertently hydrolyze esters to soaps due to presence of free catalyst, so the samples taken for soap analysis were not halted. Samples for triglyceride analysis were taken at 10-minute intervals with initial time at first contact of the methoxide solution to the preheated oil. Eighty-one gallons of oil and nearly 15 gallons of methanol were reacted in each batch. The reaction occurred in two stages; first stage (A) used 90% of the catal yst and 80% of the methanol to make a 6:1 molar ratio methanol to oil needed for the reaction. The first stage pro ceeded for 45 minutes, pump and ultrasound stopped for 1 hour. About half of the glycerin product la yer was removed prior to the second stage to help favor reaction completion. The second stage delivered the remaining 20% of methanol and the last 10% of the catalyst needed to comple te the reaction. Mixing wa s allowed for another 45 minutes. Results and Conclusions The sam ples were observed for visible sepa ration differences and tested for soap formation. The samples were each placed in a separatory funnel and glycerin-by-product decanted. The samples were thoroughly water wash ed and dried for gas chromatography analysis to determine free and total glycerin by AS TM method D6584. The rate of conversion is calculated by plotting the change in con centration of bound glycerides versus time. The results showed an improvement in the c onversion of triglycerides to esters at all intervals (Table 4-1). Most importantly are the amount of bound glycerides (i.e. mono, di and triglycerides), which are fats not fully converted to esters, and total glycerin, which is the total amount of the glycerin portion of the bound glyc erides and free unbound glycerin. Free glycerin is water soluble and normally removed by post-reaction refining. The total conversion to 0.4%

PAGE 63

63 free and total glycerin falls just short of the ASTM specification of 0.24% permitted. Note, after 90 minutes of mixing, the final amo unt of unreacted triglycerides in the batch using ultrasound is half that of the batch with conventional mixing alone. Analysis of the soap reaction is determined by titrating the mixed phase samples against a solution of hydrochloric acid to the bromophenol blue end point. These tests showed this complete within ten minutes for the batch with ultrasound, and 20 minutes without ultrasound. No change was detected after these times. Settling of the partially reacted material is an indicator of the degree of conversion. The high amount unreacted methanol forms an inhibitiv e complex with the low amounts of glycerin product formed and slows the phase separati on. Samples were taken and observed for any improvement in settling. A negligible improvement may have been detected with the ultrasonic enhanced reaction, however too subj ective for any conclusive data. Resulting from the energy provided by power ultrasound, sonochemistry is finding a niche in clean technology for the future (Mason and Cintas, 2002). These resu lts show promise for improved throughput in reaction, which could be optimized by greater power input or smaller reactor sizing. The manufactur er recommended optimal operating power at about 40 Watts per gallon. This is based on typical applications for ultrasonic cleaning systems. While this work operated at about 27 Watts per gallon. Colucci conducted his trials in a 250 ml glass reactor, applying 14.49 Watts, the equivalent of nearly 220 Watts per ga llon, 9 times the power provided in the UF pilot system. The cost would be fina ncially impractical for batch scale reactors to apply this much power needed to simulate these results. Both studies show potential for applications of high power ultras ound in a smaller reaction chambe r with much shorter residence times.

PAGE 64

64 Figure 4-1 Steel reactor vesse l, St Johns County Florida

PAGE 65

65 Figure 4-2 Tube resonator assemblies for reaction kinetic study

PAGE 66

66 Table 4-1. Chart shows the amount of unreacted ma terial (mono, di, tri) and free glycerin as grams per 100 grams of sample. Sample seri es 3 is from the reaction with ultrasound. Samples labeled A are from first stage of reaction, those marked B are from the second stage, F is final after second stage. Bound represents the glycerin fraction of bound glycerides. Total includes the free glycer in. Courtesy of University of Idaho Sample Free Mono Di Tri Bound Total Time 3.1A 0.001 0.677 5.323 23.129 3.382 3.383 10 minutes 3.2A 0.032 0.476 3.686 18.477 2.601 2.633 20 minutes 3.3A 0.007 0.378 2.827 14.149 1.996 2.003 30 minutes 3.4A 0.001 0.388 2.388 14.026 1.920 1.921 40 minutes 3.1B 0.019 0.345 1.412 7.075 1.038 1.057 10 minutes 3.2B 0.002 0.325 0.960 4.520 0.699 0.701 20 minutes 3F 0.000 0.551 0.477 1.783 0.400 0.400 45 minutes 4.1A 0.024 0.848 6.502 32.319 4.561 4.585 10 minutes 4.2A 0.001 0.444 3.939 19.738 2.762 2.763 20 minutes 4.3A 0.000 0.408 3.187 16.934 2.348 2.348 30 minutes 4.4A 0.043 0.440 2.585 15.745 2.142 2.185 40 minutes 4.1B 0.012 0.326 1.456 8.702 1.210 1.222 10 minutes 4.2B 0.003 0.335 1.150 6.913 0.980 0.982 20 minutes 4F 0.000 0.597 0.799 3.875 0.678 0.678 45 minutes

PAGE 67

67 CHAPTER 5 INTEGRATING ULTRASONICATION, ANAERO BIC DIGESTION OF GLYCERIN COPRODUCT FROM BIODIESEL AND SOLAR PROCESS HEAT; AN ENERGY ANALYSIS Introduction W hen the internal combustion engine was first introduced, it was touted as a healthy alternative to the animal draw n wagons, where manure and disease vector flies were problematic. Oil spills, lead toxic ity, MTBE in our water supply, sm og, greenhouse gas emissions, are all negative externalities somewhat addressed by the petroleum industr y. Alternative fuels, such as biodiesel are hailed as an answer due to its closed-carbon loop. Questions of mass deforestation and habitat destruction in many pa rts of the world for large-scale oil crops have ta inted biofuels reputation. Propositions have been placed in areas such as Seattle, Washington to ban land-based biofuels, such as ethanol and biodi esel. If biofuels are to become a viable alternative to fossil fuels, then their production needs to show bene fits when replacing any portion of fossil fuel consumption. These concerns greatly challenge us as researchers and natural resource managers. Prior to the coal, oil and natural gas explora tion, (fossilized biofuels) human society relied on biomass for heat and steam. Direct combustion of forest products such as wood and peat was the norm. Common biofuels like bio-methane, et hanol and biodiesel could be considered second generation biofuels and more advanced fuels like cellulosic ethanol, bio-butanol and algalbiodiesel as third generation. Florida is an energy dependent state, relying on coal, natura l gas, fuel oil and nuclear power, all sourced outside the stat e. Florida Department of Agri culture and Consumer Services Farm-to Fuel 20/20 initiative shoots for producin g 20% of Floridas energy needs from farmland without compromising food supply by the year 2020. Objectives include ge nerating state revenue and providing greater income to farmers ofte n tempted to sell arable farmland to housing developers.

PAGE 68

68 The analysis described in this chapter collects data from collaborative efforts throughout the UF Agricultural and Biologi cal Engineering Bio-energy Res earch Laboratory. Many research associates conducted experiments, simulations and field trials from 2005-2009. Teams formed to study efficient reactor design, solar thermal heat exchange and photovoltaic inputs, as well as solar powered methanol recovery. Typically a life-cycle assessment (LCA) is s ponsored or overseen by the US Department of Energy for any fuel entering into the market. An LCA accounts for complete birth-to-grave tally of inputs to the process, including energy for manufacturing, feedstock production, transportation of materials to and from markets, and even ener gy embedded in structures needed to produce or distribute the fuel. The National Renewable Energy Lab has conducted a full LCA, which showed 3.2 units of energy produced for every unit of energy put in to the system (Sheehan et al., 1998). This energy analysis pays no regard to the source of th e energy inputs, whether or not it is from renewable sources. One could argue whet her or not a fuel is truly renewable if it is merely consuming fossil fuels at a slower rate; the result remains the sa me, depletion of vast deposits that occur perhaps on ce in the life of a planet. Rather than a complete life-cycle inventor y, which could develop into its own research study, consideration was given onl y to energy conserved or produced compared to process energy invested. This balance is similar to an input output (I/O) modeling for economic analysis. The boundary, across which this energy flows for our system, is shown on Figure 2-1. For the sake of comparison, Btu per gall on, a standard unit of measurement was selected for reference across all materials (Table 5-1) The gross energy in Btu pe r gallon of biodiesel is 118,000. This simplified analysis ignores the monetary value of certain products such as recovered methanol, acidulated soapstocks and glycerin, which could be valued more as feedstock inputs or

PAGE 69

69 commodity than their Btu conten t alone. We are merely measuri ng and comparing Btus in (I) versus Btus out (O). Various combinations of in tegrated systems will be presented to show tradeoffs in costs and benefits. Assumptions include: No inputs or energy costs for operator energy expended Biodiesel is valued for transportation by PPD not the used cooking oil, as modification of all engines for straight vegetable oi l use is not acceptable or practical Life of used vegetable oil begins at the waste receptacle, pro duction energy inputs for feedstock were for previous cooking use Only energy balance is considered, not economic costs or value All heat and power can be produced on site with solar engineering All values reported are standardized for unit comparison, averaged from1600 gallon per month target production rate. Process methods are similar to those described in Chapter 2 Power and Energy Inputs Energy inputs include major categories: transportation (I1), Btu value of methanol (I2), process heat and power (I3), energy in catalyst production (I4), waste disposal transportation energy for wash water and glycerin (I5), and heat needed for methanol recovery (I6). Energy outputs are limited to the Btu value of the biodiesel fuel (O1) and the energy from biogasification of co-products (O2) (Table 5-1). Thermodynamics of every component in our fi eld process were evaluated including: heat of reaction, heat needed for li quid/ liquid extraction, evaporat ion, and power requirements for water pressure, oil pumps, centr al processors and valves.

PAGE 70

70 Energy Analysis of Ultrasonic Inputs Ultrasonic resonance showed im provements in the rate of reaction, which in turn lowers the Arrhenius energy of activation needed. Activati on energy can be expre ssed in units of energy per unit of product (Colucci et al ., 2005). The goal of increasing rate of conversion is to increase throughput affecting capital investme nt cost more than energy cost savings. The energy cost of running the ultrasonic generators is 3.0 kWh/ hour of reaction. Operating the rods provides enough heat to compensate for loss of heat thr ough pipes and reactors surface area during the reaction. Two to four degrees Celsius temperatur e rise was noted during ultrasonic resonance. These benefits show increased settling time, which would require less volume for the settling tank in-between stages or batches. All of these be nefits as concluded in Chapter 4 have led to the adoption of ultrasonic addition to the process for improved ra tes of conversion and throughput. The power and heat consumption in the energy balance reflect the trade-offs, increase power consumption for ultrasonic generators, but less heat and pump energy needed. Anaerobic Digestion of Glycerin For every gallon of biodiesel produced, about two pounds of glycerin co-product form UF Professor Emeritus Dr David Chynoweth inspired the UF Biodiesel project in January of 2006 after attending a BEST Society bi odiesel seminar hosted by Dr A nne Wilkie of IFAS Soil and Water Sciences. At this seminar, Lyle Estill, founder of Piedmont Biofue ls Industrial expressed the need for the industrys improved management of the glycerin co-product. Increased biodiesel production has caused pric e drops in the international gl ycerin market. Many new startups had optimistic financial projections that depended on revenues from refined glycerin, which at one time was valued higher than the bi odiesel fuel itself (K nothe and Dunn, 1997). The tremendous growth in biodiesel production ha s driven prices to between $0.025 / pound to $0.05 / pound (Patzer, 2007). With the assistance of Dr C hynoweth, a specialist in anaerobic digestion,

PAGE 71

71 the biodiesel project began with the intent of deriving all proce ss heat and power from the biogas produced from biogasification of glycerin-by-product. Anaerobic digestion (AD), a process by which organically degradable substances are conv erted to minerals including methane gas, is often engineered for managing agricultural residues. The gross heat of combustion of glycerin is approximately the same whether burned as a fuel directly or converted to intermediary methane through anaerobic digestion and then combusted (Thompson and He, 2006). Anaerobi c digestion enables processing of GBP containing water, soaps, alcohol and food particles. Also ther e are no chlorine emissions or concern of acrolein toxicity. Ac rolein, an aldehyde formed from the decomposition of glycerin at or above 280 0C, is a pulmonary irritant and causes serious skin damage. It is also linked to lung cancer (Feng, et al., 2006). Use of this as an energy source on site retains energy in the system. Us e of glycerin in its crude form can eliminate the need for storage, transportation and disposal costs. In anaerobic digestion, glycerin is fully degr adable to water, carbon dioxide and methane (Figure 5-1), and the methane, in the form of biogas, can be directly combusted in a furnace or modified natural gas water heater. If potassium hydroxide is used as catalyst and neutralizer, then nutritive compounds formed in AD effluent can be land applied. Bio-chemical methane potential (BMP) of th e glycerin by-product (GBP) was determined by standardized methods. Effective conversion of GBP showed the potenti al to be 450 ml of methane at STP, slightly higher than theoretical stoichiometric yield of 418 ml methane/ ml of pure glycerin. Still, the effectiv e yield of methane from BMP anal ysis was less than theoretical however, at 340 ml/ml of pure glycerin. The GBP analyzed was about 36% glycerin. The remaining 64% consisted of 37% residual methanol and 27% soaps and water product from side

PAGE 72

72 reactions of free fatty acid ( FFA) neutralization. The excess me thanol needed to drive the reaction more nearly to completion remains distri buted in both the este r and glycerin phases, proportionately 2:3. This soap amount is variable depending on quality of feedstock. Typically crude glycerin is refined to 85% purity for market sale ; 12% is typical wate r content after soap acidulation. The methanol and some water are re moved through flash evaporation and methanol is distilled for reuse. The soaps would be rec overed as FFAs through soap acidulation, usually by addition of hydrochloric acid (HCl ) in water solution. The HCl added leaves dissolved sodium or potassium salt, which pose chlorine emission problems when combusted in boilers (Patzer, 2007). The soaps, if not acidulated and recovered as FFAs, and tr ace organic colloids could be degraded in AD and contribute to the BMP. Consider this economic scenario based on th e expected monthly production of glycerin-byproduct (GBP). The natural gas or me thane produced is as follows: 1000 Btu/ cubic ft methane @ STP 320 gallons (1216 liters) waste GBP/month Theoretical yield-418 liters methane/liter of GBP 418 liters *1216 liters= 508,288 liters methane 508,288* 1 cubic ft/28 liters=18,153 cu ft 18153 cu ft* 1000 BTUs/cu ft= 18,153,000 Btu 18.15 MM Btu/mo @ $40.00/ MM Btu (retail) =$726/mo Compare with value of crude gl ycerin @$0.05/lb (Patzer, 2007) 300 gallons X 7.5 lbs/gallon=2250 lbs. 2250 lbs X $0.05/lb (wholesale)=$112.00 A comparative analysis in standard units (Btu) of energy consumed per ga llon of biodiesel shows the energy value, used saved and pro duced in the process (Table 5-1) Heating Value of Methane: 50,016 Kg KJ Heat Produced from combustion: 2.36e7 KJ or 6550 kWh or about 4 kWh/ gallon of fuel.

PAGE 73

73 Calculations show the goal of an energy self -sufficient biodiesel reactor would be met using the closed-loop design system that utilizes methane gas created from GBP of the biodiesel reaction. The production sufficiently yields enough biogas to heat water for all processes and the anaerobic digester itself. The sy stem design is economically feasib le and can be duplicated, as all materials were simple industrial and commercial elements. Solar Thermal and Photovoltaic Inputs for Process Heat and Power The power availab le with solar heat exchange could eliminate the need for the natural gas hot water heater for the biodiesel process heat. The addition of so lar thermal heat exchange also offers options for energy expended by the glycerin digestion to heat the anaerobic digesters. Providing process heat from solar sources, allo ws the biogas produced from the glycerin digestion to instead be used in our 13.5 kW modi fied dual-fuel diesel ge nerator. The waste heat from the liquid cooled generator can provide eno ugh heat to replace the electric hot water heater used for maintaining the AD operating temperat ure (Figure 5-2). The energy produced by the generator can be net-metered and help offs et power consumption by the university. The solar thermal system was designed to ope rate along side a sola r photovoltaic system that would function as both a backup with batter ies and inverter and to also provide additional heat needed in reactor and eva porator. The upper limit of the solar thermal system used in field trials was 55 0C. Six meter squared of panel surface pow ers 600 watts per m^2 at 53% efficiency. This field system provided 11.4 kWh/day averaging 6 good hours of Florida sunlight daily. To provide all thermal process heat, based on 75% e fficient conversion, requires nine meter squared of thermal collection surface. (see appendix). This proposed system would provide 24.3 kWh/day or about 729 kWh/ month for the 1600 ga llons of fuel produced. This is the Btu equivalent of 2.49 mega-Btu. The heat requir ement for operating the reactor 20 days/month, which will produce 1600 gallons of biodiesel is but 504 kWh or 1.72 MBTU. Solar inputs can

PAGE 74

74 always be properly sized to accommodate heat and power needs. The steady energy demanded to operate the equipment was 15.5 kW. Conclusions The vision f or the integration of these co mponents was shared and developed by many students acknowledged in the forwar d to this thesis. Research associates aptly depict the conceived model for a sustainable bioenergy cente r, where all of the f eedstock is produced and all of the waste products become building blocks for the new products (Figure 5-3) Economics influences choices in process desi gn. Conventional methods prevail in industry because of familiarity and reliab ility of energy systems. But the calculations of energy required versus return outputs favor anaerobic dige stion, methanol recovery and solar. Figures: 4C3O3H8 7CH4 + 5CO2 + 2H2O glycerol methane carbon dioxide water CH4 + 2O2 CO2 + 2H2O + heat Methane oxygen carbon dioxide water + heat Figure 5-1. Stoichiometric balan ce shows theoretical yield of methane, carbon dioxide and water products from glycerol and the products from the combustion of methane. Heat can be used in the process eith er by flue gas or heat tr ansferred from liquid cooled generator.

PAGE 75

75 Figure 5-2. This schematic shows the process flow of heat and power from biogas derived from anaerobic digestion of glycerin product.

PAGE 76

76 Figure 5-3 Integrated concept of biodiesel res earch facility with f eedstock development and complete solar application.

PAGE 77

77 Figure 5-4. Solar assisted me thanol recovery system

PAGE 78

78 Table 5-1. Energy balance table calculates all energy requirements.

PAGE 79

79 CHAPTER SIX FINAL COMMENTS AND FUTURE WORK When this project was conceived as a resear ch proposal and academ ic plan, a tentative proposal was introduced to gradua te coordinator Dr Ray Buckli n. He thought enough work was described for multiple PhD candidates, not one Masters of Science. The scope needed to be narrowed to a more specific focus. The addition of student teams enabled the project to expand and integrate more functions. The thesis question, Can the coupling of bi ological, solar and u ltrasonic technologies enhance the overall return of energy on the energy invested in the producti on of biodiesel? is a bit of a foregone conclusion in effect. The use of the sun to supplant the energy requirement alone favors a more positive energy balance. Solar can be unreliable and the need for electrical back-up is still necessary. There is much experimental work needed to determine applica tion of carbon dioxide scrubbing of biogas from glycerin dige ster using alkaline wash water. The use of refined GBP as a solvent to extract soluble fuel contaminants, such as soaps, water and free methanol shows promise for more integration of unit operations. There is a need for a more thorough life-cycle assessment for the entire integrated process showing feedstock development from residues produced on site. Residues, such as carbon dioxide from the generator, soaps and waste heat can be used fo r photo-bioreactors or co vered aquatic systems. Where ultrasonic studies show increased rates of reaction, further studies are needed to quantify economic advantages, if any. More expe riments should be conducted using increased power or development of a small flow-through chamber reactor which could function continuously. The unit shows possi bilities for more kinetic st udies, which could reduce the amount of methanol reactant.

PAGE 80

80 APPENDIX TABLES FROM MIXING EXPERIMENTS Table A-1. These data are th e results from Ultr aviolet photo-spectroscopic measurement of absorbance of known concentrations of dye Concentrationof Dye in mg/l Absorbance Fit 0 0 0 0.2 0.028 0.020342294 0.5 0.053 0.050855735 1 0.101 0.10171147 3.33 0.292 0.338699195 5 0.482 0.50855735 10 1.105 1.0171147 11.11 1.13 1.130014432 16.67 1.66 1.695530205 Table A-2. Data below were used for calibration curve shown in Figure two from prepared concentrations analyzed by UV photo-spectro scopy. These concentrations were used for analyzing absorbance on field samples. % Biodiesel in oil Amount of biodiesel Amount of soy oil Absorbance % Soy oil 1 .1 ml 9.9 ml 0.09 99% soy oil 2 .2 ml 9.8 ml 0.106 98% soy oil 20 2 ml 8 ml 0.194 80% soy oil 40 4 ml 6 ml 0.326 60% soy oil 60 6 ml 4 ml 0.426 40% soy oil 80 8 ml 2 ml 0.499 20% soy oil

PAGE 81

81 Table A-3. Measurements of concentration of dye in samples taken over 30 minutes of time during Experiment 1.1 using pum p recirculation mixing only TIME/MIN ABSORBANCE UNIT EXPERIMENT TIME/MIN ABSORBANCE UNIT EXPERIMENT 0.25 0.035 1.1A 0.25 0.728 1.1B 0.5 0.814 1.1A 0.5 0.128 1.1B 0.75 0.083 1.1A 0.75 0.513 1.1B 1 0.57 1.1A 1 0.29 1.1B 1.25 0.213 1.1A 1.25 0.523 1.1B 1.5 0.452 1.1A 1.5 0.474 1.1B 1.75 0.4 1.1A 1.75 0.371 1.1B 2 0.261 1.1A 2 0.284 1.1B 2.25 0.159 1.1A 2.25 0.342 1.1B 2.5 0.029 1.1A 2.5 0.472 1.1B 2.75 0.348 1.1A 2.75 0.483 1.1B 3 0.251 1.1A 3 0.46 1.1B 3.5 0.384 1.1A 3.5 0.53 1.1B 4 0.339 1.1A 4 0.495 1.1B 4.5 0.379 1.1A 4.5 0.51 1.1B 5 0.341 1.1A 5 0.447 1.1B 5.5 0.208 1.1A 5.5 0.371 1.1B 6 0.297 11.A 6 0.472 1.1B 6.5 0.342 11.A 6.5 0.526 1.1B 7 0.378 11A 7 0.519 1.1B 8 0.381 1.1A 8 0.52 1.1B 9 0.376 1.1A 9 0.519 1.1B 10 0.341 1.1A 10 0.485 1.1B 12 0.301 1.1A 12 0.408 1.1B 14 0.182 1.1A 14 0.342 1.1B 16 0.227 1.1A 16 0.465 1.1B 18 0.256 1.1A 18 0.509 1.1B 20 0.307 1.1A 20 0.498 1.1B 25 0.283 11.A 25 0.513 1.1B 30 0.304 1.1A 30 0.524 1.1B

PAGE 82

82 Table A-4. Experiment 1.2 Pump mixi ng with ultrasonic assistance TIME/MIN ABSORBANCE UNIT EXPERIMENT TIME/MIN ABSORBANCE UNIT EXPERIMENT 0.25 -0.01 1.2A 0.25 0.598 1.2B 0.5 0.617 1.2A 0.5 0.346 1.2B 0.75 0.064 1.2A 0.75 0.794 1.2B 1 0.438 1.2A 1 0.38 1.2B 1.25 0.3 1.2A 1.25 0.489 1.2B 1.5 0.315 1.2A 1.5 0.491 1.2B 1.75 0.314 1.2A 1.75 0.353 1.2B 2 0.298 1.2A 2 0.509 1.2B 2.25 0.168 1.2A 2.25 0.408 1.2B 2.5 0.292 1.2A 2.5 0.464 1.2B 2.75 0.365 1.2A 2.75 0.478 1.2B 3 0.389 1.2A 3 0.526 1.2B 3.5 0.385 1.2A 3.5 0.471 1.2B 4 0.403 1.2A 4 0.492 1.2B 4.5 0.374 1.2A 4.5 0.463 1.2B 5 0.351 1.2A 5 0.443 1.2B 5.5 0.209 1.2A 5.5 0.45 1.2B 6 0.293 1.2A 6 0.49 1.2B 6.5 0.367 1.2A 6.5 0.52 1.2B 7 0.373 1.2A 7 0.504 1.2B 8 0.369 1.2A 8 0.5 1.2B 9 0.376 1.2A 9 0.484 1.2B 10 0.384 1.2A 10 0.514 1.2B 12 0.352 1.2A 12 0.412 1.2B 14 0.231 1.2A 14 0.406 1.2B 16 0.313 1.2A 16 0.409 1.2B 18 0.332 1.2A 18 0.513 1.2B 20 0.364 1.2A 20 0.478 1.2B 25 0.366 1.2A 25 0.502 1.2B 30 0.346 1.2A 30 0.445 1.2B

PAGE 83

83 Table A-5. These data show concentrations of methylene blue dye over time from experiment 1.3 which uses ultrasonication only. TIME/MIN ABSORBANCE UNIT EXPERIMENT TIME/MIN ABSORBANCE UNIT EXPERIMENT 0.25 -0.006 1.3A 0.25 0 1.3B 0.5 0.003 1.3A 0.5 0.003 1.3B 0.75 0.96 1.3A 0.75 0.002 1.3B 1 0.327 1.3A 1 0.005 1.3B 1.25 0.468 1.3A 1.25 0.153 1.3B 1.5 0.549 1.3A 1.5 0.207 1.3B 1.75 0.493 1.3A 1.75 0.353 1.3B 2 0.407 1.3A 2 0.461 1.3B 2.25 0.435 1.3A 2.25 0.574 1.3B 2.5 0.492 1.3A 2.5 0.609 1.3B 2.75 0.485 1.3A 2.75 0.622 1.3B 3 0.532 1.3A 3 0.517 1.3B 3.5 0.564 1.3A 3.5 0.631 1.3B 4 0.59 1.3A 4 0.582 1.3B 4.5 0.585 1.3A 4.5 0.583 1.3B 5 0.583 1.3A 5 0.552 1.3B 5.5 0.558 1.3A 5.5 0.59 1.3B 6 0.57 1.3A 6 0.517 1.3B 6.5 0.584 1.3A 6.5 0.522 1.3B 7 0.576 1.3A 7 0.511 1.3B 8 0.595 1.3A 8 0.518 1.3B 9 0.593 1.3A 9 0.54 1.3B 10 0.588 1.3A 10 0.577 1.3B 12 0.591 1.3A 12 0.538 1.3B 14 0.577 1.3A 14 0.553 1.3B 16 0.575 1.3A 16 0.526 1.3B 18 0.569 1.3A 18 0.499 1.3B 20 0.597 1.3A 20 0.516 1.3B 25 0.528 1.3A 25 0.521 1.3B 30 0.545 1.3A 30 0.562 1.3B

PAGE 84

84 Table A-6. Experiment 2 Trial 2.1 uses pump recirculation only. 15 gallons of Soybean oil introduced to 60 gallons of esters Dilution 15/75 or .2 soy oil concentration OR 0.8 (80%) BD in oil Expected final absorbance at A and B when well mixed = 0.503 Calibration: Abs = Int + (Slope*Percent) OR Percent = (Abs-Int)/Slope; Int = 0.09962; Slope = 0.005045 Time/ min Absorb ance Exp/Port BD % Time (Min) Absorbance Exp/Port BD % 5 0.446 2.1A 68.6 5 0.499 2.1B 79.2 10 0.491 2.1A 77.5 10 0.491 2.1B 77.6 15 0.457 2.1A 70.8 15 0.497 2.1B 78.8 20 0.449 2.1A 69.2 20 0.498 2.1B 78.9 25 0.453 2.1A 70.0 25 0.501 2.1B 79.5 30 0.448 2.1A 69.0 30 0.517 2.1B 82.7 40 0.449 2.1A 69.2 40 0.501 2.1B 79.5 50 0.444 2.1A 68.2 50 0.496 2.1B 78.8 60 0.45 2.1A 69.4 60 0.484 2.1B 76.2 75 0.453 2.1A 70.0 75 0.48 2.1B 75.4 90 0.45 2.1A 69.4 90 0.489 2.1B 77.2 105 0.45 2.1A 69.4 105 0.499 2.1B 79.2 123 0.445 2.1A 68.4 120 0.505 2.1B 80.3 150 0.453 2.1A 70.0 150 0.526 2.1B 84.5 180 0.463 2.1A 72.0 180 0.53 2.1B 85.3 Table A-7. Experiment 2 Trial 2.2uses pum p recirculation with ultrasonic assist. 30 gallons of soybean oil into 60 gallons of biodiesel Dilution 30/90 or .33 soy oil concentration OR 67% BD Expected final absorbance at Ports A and B when well mixed = 0.438 Calibration: Abs = Int + (Slope*Percent) OR Percen t = (Abs-Int)/Slope; Int = 0.09962; Slope = 0.005045 Time/Min Absorb ance Exp/port % BD Time/m in Absorbance Exp/Port % BD 5 0.409 2.2A 61.3 5 0.397 2.2B 58.9 10 0.402 2.2A 59.9 10 0.395 2.2B 58.5 15 0.411 2.2A 61.7 15 0.39 2.2B 57.5 20 0.424 2.2A 64.3 20 0.392 2.2B 57.9 25 0.411 2.2A 61.7 25 0.392 2.2B 57.9 30 0.404 2.2A 60.3 30 0.408 2.2B 61.1 40 0.436 2.2A 66.6 40 0.392 2.2B 57.9 50 0.416 2.2A 62.7 50 0.394 2.2B 58.3 60 0.41 2.2A 61.5 60 0.402 2.2B 59.9 75 0.408 2.2A 61.1 75 0.386 2.2B 56.7 90 0.412 2.2A 61.9 90 0.388 2.2B 57.2 105 0.414 2.2A 62.3 105 0.388 2.2B 57.2 120 0.407 2.2A 60.9 120 0.392 2.2B 57.9 150 0.397 2.2A 58.9 150 0.39 2.2B 57.5 180 0.405 2.2A 60.5 180 0.398 2.2B 59.1

PAGE 85

85 Table A-8. Experiment 2 Trial 2.3 uses ultrasonic rods only, no pump 45 gallons of soybean oil into 60 gallons of biodiesel Dilution 45/105 or .428 soy oil concentration OR 57.2% BD Expected final absorbance at A and B when well mixed = 0.388 Calibration: Abs = Int + (Slope*Percent) OR Percen t = (Abs-Int)/Slope; Int = 0.09962; Slope = 0.005045 Time/ min Absorbance Exp/port %BD Time/min Absorbance Exp/port %BD 5 0.397 2.3A 58.9 5 0.335 2.3B 46.6 10 0.404 2.3A 60.3 10 0.324 2.3B 44.5 15 0.385 2.3A 56.5 15 0.316 2.3B 42.9 20 0.391 2.3A 57.7 20 0.33 2.3B 45.6 25 0.387 2.3A 56.9 25 0.323 2.3B 44.3 30 0.389 2.3A 57.3 30 0.321 2.3B 43.9 40 0.376 2.3A 54.7 40 0.323 2.3B 44.3 50 0.372 2.3A 53.9 50 0.323 2.3B 44.3 60 0.378 2.3A 55.2 60 0.326 2.3B 44.9 75 0.373 2.3A 54.2 75 0.323 2.3B 44.3 90 0.368 2.3A 53.2 90 0.323 2.3B 44.3 105 0.363 2.3A 52.2 105 0.319 2.3B 43.5 120 0.368 2.3A 53.2 120 0.337 2.3B 47.0 150 0.364 2.3A 52.4 150 0.317 2.3B 43.1 180 0.363 2.3A 52.2 180 0.323 2.3B 44.3

PAGE 86

86 LIST OF REFERENCES Am erican Oil Chemists Society, 1998. Official methods and recommended practices of the AOCS, fifth ed. American Oil Chemis ts Society, Champaign, Illinois. Anagnostopoulos, J.S., 2006. CFD analysis and de sign effects in a radial pump impeller. WSEAS Transactions on Flui d Mechanisms. 7 (1), 763-770. Bailey, A.E., Swern, D., Formo, M.W., Applewhite T.H., 1979. Baileys industrial oil and fat products, fourth ed. Wiley, New York. Boocock, D.G.B., Konar, S.K., Mao, V., Lee, C., Buligan, S., 1998. Fast formation of highpurity methyl esters from vegetable oils. Jour nal of the American Oil Chemists Society. 75 (9), 1167-1172. Braslavsky, S. E. & Houk, K. N. (Eds.), 1988. Glo ssary of terms used in photochemistry, in: International Union of Pure and Applied Chemistry (IUPAC). Canakci, M., 2001. Production of biodiesel from feedstock with high free fatty acids and its effect on diesel engine performance and emissions. Iowa State University, Ames. Canakci, M., Van Gerpen, J., 2003. A pilot plant to produce biodiesel from high free-fatty acid feedstock. American Society of Agricultural Engineers, St. Joseph, Michigan. Carlson, S., 2006. In search of the sustainable cam pus: With eyes on the future, universities try to clean up their acts. The Ch ronicle of Higher Education. Colucci, J.A., Borrero, E.E., Alape, F., 2005. Biodi esel from an alkalin e transesterification reaction of soybean oil using ultrasonic mixi ng. J. of the American Oil Chemists Society. 82 (7), 525-530. EPA, 2009. EPA Lifecycle analys is of greenhouse gas emissions from renewable fuels. http://www.epa.gov/oms/renewablefuels/420F09024.pdf accesed 12/03 /2009 Freedman, B., Pryde, E.H., Mounts, T.L., 1984. Variable s affecting the yields of fatty esters from transesterified vegetable oils. Journal of th e American Oil Chemists Society (JAOCS). 61 (10), 1638-1643. Gerbens-Leenes, W., Hoekstra, A.Y., van de r Meer, T.H., 2009. The water footprint of bioenergy. PNAS. 106 (25), 10219-10223. Grant, A.T.J., 2003. A model for potential coope rative biodiesel production in Gainesville, Alachua County, Florida. University of Florida, Gainesville. Green, F.J., 1990. The Sigma Aldrich Handbook of Stains, Dyes and Indicators.

PAGE 87

87 Hansen, K.H., Angelidaki, I., Ahring, B.K., 1998. Anaerobic digestion of swine manure: Inhibition by ammonia. Wa ter Research. 32 (1), 5. Hartog, L., Fox, M., 2008. Cool schools: Ten that get it. Sierra. Kemp, W.H., 2006. Biodiesel basics and beyond: A comprehensiv e guide to production and use for the home and farm. Aztext Press, Tamworth, Ontario. Knothe, G., Dunn, R.O., Bagby, M.O., 1997. Biodiesel: The use of vegetable oils and their derivatives as alternative diesel fuels. Fuels and Chemicals from Biomass. 666, 172-208. Komers, K., Skopal, F., Stloukal, R., 2002. Kine stics and mechanism of the KOH catalyzed methanolysis of rapeseed oil for biodiesel production. European Journal of Lipid Science and Technology. 104 (11), 728-737. Mason, T.J., 1991. The use of sonochemistry in or ganic reactions: The effect of ultrasonic waves on chemical reactivity. Canadian Chemical News. 43, 25-26. Mason, T.J., Cintas, P., 2002. Sonochemistry, in: Clark, J., Macquarrie, D. (Eds.), Handbook of Green Chemistry and Technology. Blackwe ll Science, Malden, MA, pp. 372-396. Pahgova, J., Jorikova, L., Cvengros, J., 2008. Study of FAME stability. Energy and Fuels, 22, 1991-1996 Peterson, C.L., Cook, J.L., Thompson, J.C., Tabe rski, J.S., 2002. Continuous flow biodiesel production. Applied Engineering in Agriculture. 18 (1), 5-11. Turner, T. L., 2005. Modeling and simulation of r eaction kinetics for biodiesel production, North Carolina State University, Raleigh, NC Van Gerpen, J., 2005. Biodiesel processing and production. Fuel Processing Technology. 86 (10), 1097-1107. Van Gerpen, J.H., 2004. Business management fo r biodiesel producers: August 2002 January 2004. National Renewable Energy La boratory, Golden, Colorado. Zhang, Y., Dube`. M.A., McLean, D.D., Kate s, M., 2003. Biodiesel production from waste cooking oil: 1. Process design and technological assessment. Bioresource Technology 89 (2003) 1-16

PAGE 88

88 BIOGRAPHICAL SKETCH Douglas Renk has retu rned to academics to apply management skills acquired in the construction industry to the area of environmental resource management. His past interest in crop production as an ornamental nu rseryman has followed him to agricultural operations. Mr. Renk holds a Bachelor of Science degree in environm ental resource management from Rollins College Winter Park, Florida. Concerns over depletion of resources, anthropogenic causes of climate change, and increasing unrest in politically charged oil-producing regi ons have driven his interest for contributing to renewable fuel programs at the University of Florida. The intention of this research is to investigate and quantify th e application of various techniques to increase energy available from vegetable oil as a fuel s ource. Renk is an Agricultural Operations major and has been working on developing a program to convert the campus waste oil to biodiesel, a flagship project for campus waste reduction and sust ainability initiatives. He received a research assistantship to help solve some of the concer ns for biofuel use globally while establishing and managing a small production facility at a nearby field station.