Green Grass: Developing Grass for Sustainable Gaseous Biofuel.pdf

Dr. Abdul-Sattar Nizami

Download Free PDF

Green Grass: Developing Grass for Sustainable Gaseous Biofuel.pdf

Dr. Abdul-Sattar Nizami

Grass is ubiquitous in Ireland and temperature northern Europe. It is a low input perennial crop; farmers are well versed in its production and storage (ensiling). Anaerobic digestion is a well understood technology. However the level of comfort with the technology can mask the difficulties associated with digestion of high solid content feedstocks especially grass silage. It is not simply a matter of using a digester designed for slurry or for Maize to produce biogas from grass silage. Grass is a lignocellulosic feedstock which is fibrous; it can readily cause difficulties with moving parts (wrapping around mixers); it also has a tendency to float. This thesis has an ambition of establishing the ideal digester configuration for production of biogas from grass. Extensive analysis of the literature is undertaken on the optimal production of grass silage and the associated biodigester configurations. As a result of this analysis two different digester systems were designed, fabricated, commissioned and operated for over a year. The first system was a two stage wet continuous system commonly referred to as a Continuously Stirred Tank Reactor (SCTR). The second was a two stage, two phase system employing Sequentially Fed Leach Beds complete with an Upflow Anaerobic Sludge Blanket (SLBR-UASB). These were operated on the same grass silage cut from the same field at the same time. Small biomethane potential (BMP) assays were also evaluated for the same grass silage. The results indicated that the CSTR system produced 451 L CH4 kg-1 VS added at a retention time of 50 days while effecting a 90% destruction in volatile dry solids. The SLBR-UASB produced 341 L CH4 kg-1 VS added effecting a 75% reduction in volatile solids at a retention time of 30 days. The BMP assays generated results in the range 350 to 493 L CH4 kg-1 VS added. This thesis concludes that a disparity exists in the BMP tests used in the industry. It is suggested that the larger BMP (2L with a 1.5 L working volume) gives a good upper limit on methane production. The micro BMP (100 ml) gave a relatively low result. The CSTR when designed specifically for grass silage is shown to be extremely effective in methane production. The SLBR-UASB has significant potential to allow for lower retention times with good levels of methane production. This technology has more potential for research and improvement especially in enzymatic hydrolysis and for use of digestate in added value products.

Department of Civil & Environmental Engineering And Environmental Research Institute University College Cork Green Grass: Developing Grass for Sustainable Gaseous Biofuel Abdul-Sattar Nizami BSc (Hons) MScEng Thesis submitted for the degree of Doctor of Philosophy to the National University of Ireland, Cork Supervisor: Dr. Jerry Murphy Head of Department: Dr. Michael Creed March 2011 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Dedication I dedicate this thesis to the ladies of my life: Humera Aslam my younger sister for her friendship, affection and trust; Shaheena Nizami my wife, my life and my love for her evergreen kindness, humbleness and love for me. Abdul-Sattar Nizami ii Green Grass: Developing Grass for Sustainable Gaseous Biofuel Acknowledgements This thesis could not be produced without the help and support of the following people and institutions: Dr. Jerry D Murphy for giving me the opportunity to carry out this research project. His motivated and patient attitude and belief led me to work in a total freedom of mind and heart to the full extent of my abilities. He guided me throughout my research and supported me with his valued advice to rectify any troubles I encountered. Department of Agriculture, Fisheries and Food (DAFF) for funding the project through the research Stimulus Fund Project “GreenGrass”; Dr. Padraig O‟Kiely and Dr. Joseph McEniry from Teagasc, Eddie Appelbe from Erneside Engineering, and Richard Kearney from Cork Institute of Technology who kindly provided technical and laboratory support to the research; Thanasit Thamsiriroj, Beatrice M Smyth, James Browne, Nicholas E Korres, and Anoop Singh of the Biofuels Research Group, for all the discussions and constructive criticism; My friends who gave me cheer and laughter with their jokes and support. They never made me feel alone… Abdul-Sattar Nizami iii Green Grass: Developing Grass for Sustainable Gaseous Biofuel Declaration I hereby declare that this thesis is my own work and that it has not been submitted for another degree, either at University College Cork or elsewhere. Where other sources of information have been used, they have been acknowledged. Signature: ………………………………………. Date: ……………………………………………. Abdul-Sattar Nizami iv Green Grass: Developing Grass for Sustainable Gaseous Biofuel Executive Summary Grass is ubiquitous in Ireland and temperature northern Europe. It is a low input perennial crop; farmers are well versed in its production and storage (ensiling). Anaerobic digestion is a well understood technology. However the level of comfort with the technology can mask the difficulties associated with digestion of high solid content feedstocks especially grass silage. It is not simply a matter of using a digester designed for slurry or for Maize to produce biogas from grass silage. Grass is a lignocellulosic feedstock which is fibrous; it can readily cause difficulties with moving parts (wrapping around mixers); it also has a tendency to float. This thesis has an ambition of establishing the ideal digester configuration for production of biogas from grass. Extensive analysis of the literature is undertaken on the optimal production of grass silage and the associated biodigester configurations. As a result of this analysis two different digester systems were designed, fabricated, commissioned and operated for over a year. The first system was a two stage wet continuous system commonly referred to as a Continuously Stirred Tank Reactor (SCTR). The second was a two stage, two phase system employing Sequentially Fed Leach Beds complete with an Upflow Anaerobic Sludge Blanket (SLBR-UASB). These were operated on the same grass silage cut from the same field at the same time. Small biomethane potential (BMP) assays were also evaluated for the same grass silage. The results indicated that the CSTR system produced 451 L CH4 kg-1 VS added at a retention time of 50 days while effecting a 90% destruction in volatile dry solids. The SLBR-UASB produced 341 L CH4 kg-1 VS added effecting a 75% reduction in volatile solids at a retention time of 30 days. The BMP assays generated results in the range 350 to 493 L CH4 kg-1 VS added. This thesis concludes that a disparity exists in the BMP tests used in the industry. It is suggested that the larger BMP (2L with a 1.5 L working volume) gives a good upper limit on methane production. The micro BMP (100 ml) gave a relatively low result. The CSTR when designed specifically for grass silage is shown to be extremely effective in methane production. The SLBR-UASB has significant potential to allow for lower retention times with good levels of methane production. This technology has more potential for research and improvement especially in enzymatic hydrolysis and for use of digestate in added value products. Abdul-Sattar Nizami v Green Grass: Developing Grass for Sustainable Gaseous Biofuel Thesis outputs Peer Reviewed Journal Papers  AS Nizami, JD Murphy. 2010. What type of digester configurations should be employed to produce biomethane from grass silage? Renewable and Sustainable Energy Reviews, 14:1558–1568.  AS Nizami, NE Korres, JD Murphy. 2009. A review of the integrated process for the production of grass biomethane. Environmental Science and Technology, 43(22), 8496–8508.  AS Nizami, T Thamsiriroj, A Singh, JD Murphy. 2010. The role of leaching and hydrolysis in a two phase grass digestion system. Energy and Fuels, 24 (8), 4549–4559.  AS Nizami, A Singh, JD Murphy. 2010. Design, commissioning, and start-up of a sequentially fed leach bed reactor complete with an upflow anaerobic sludge blanket digesting grass silage. Energy and Fuels, 25 (2), 823–834.  AS Nizami, JD Murphy. An optimized two phase digestion system for the production of gaseous biofuel from grass silage, a high solids content feedstock. Energy & Environmental Science. March, 2011.  AS Nizami, A Orozcob, E Groom, JD Murphy. How can we optimise production of biomethane from grass silage? Energy & Environmental Science, March, 2011.  A Singh, AS Nizami, NE Korres, JD Murphy. 2010. The effect of reactor design on the sustainability of grass biomethane. Renewable and Sustainable Energy Reviews, 15(3), 1567-1574.  Nicholas E. Korres, Anoop Singh, AS Nizami, JD Murphy. 2010. Is grass biomethane a sustainable transport biofuel? Biofuels, Bioproducts, Biorefinery, 4:310–325. Peer Reviewed Conference Papers  AS Nizami, A Orozcob, E Groom, JD Murphy. Comparison of Wet and Dry Grass Digestion Systems. International IWA-Symposium on Anaerobic Digestion of Solid Waste and Energy Crops. Vienna, Austria, August 28th – September 1st, 2011.  A Orozcob, AS Nizami, JD Murphy, E Groom. Evaluation of a pretreatment process for improved methane production from grass silage. Progress in Biogas II. Biogas production from agricultural biomass and organic waste. International Congress, March 30-April 1, 2011. University of Hohenheim, Stuttgart, Germany.  JD Murphy, B Smyth, AS Nizami, T Thamsiriroj, A Singh, N Korres. The potential for biomethane as a transport fuel in Ireland. In, Biofuels Directive to bio based Transport Systems in 2020. IEA Bioenergy Task 39 Subtask Policy and Implementation Workshop, Dresden, Germany, June 2-5, 2009. Abdul-Sattar Nizami vi Green Grass: Developing Grass for Sustainable Gaseous Biofuel  A Singh, NE Korres, AS Nizami, JD Murphy. Biomethane from agricultural waste: A clean vehicular biofuel. International Conference on Environment Energy and Development from Johensberg to Kopenhagen ICEED-2010, Sambalpur University, Orissa, India. December 10-12, 2010.  NE Korres, CO‟Brien, B Smyth, AS Nizami, T Thamisirioj, R Schulte, JD Murphy. A preliminary analysis of energy balance and greenhouse gas (GHG) emissions of biomethane production as a transport fuel from grass/silage. A case study for Ireland. Society of Environmental Toxicology and Chemistry (SETAC) Conference, Goteborg, Sweden, 31 May – 4 June, 2009. Miscellaneous  JD Murphy, NE Korres, A Singh, B Smyth, AS Nizami, T Thamsiriroj. 2010. Grass Biomethane. Environmental Protection Agency (EPA), Ireland. EPA Climate Change Research Programme 20072013. (2007-CCRP-1.7), Johnstown Castle, Co. Wexford, Ireland.  NE Korres, T Thamsiriroj, BM Smyth, AS Nizami, A Singh, JD Murphy. 2010. Grass biomethane for agriculture and energy. Chapter in, „Sustainable Agriculture Reviews‟. In press, Publisher: SpringerVerlag London Ltd. Invited Lectures  AS Nizami. 2010. Potential digester configuration for Grass Biomethane. Environmental Protection Agency (EPA) conference on „Grass as a source of renewable gaseous fuel‟. EPA Grass Biomethane, University College Cork (UCC), Cork, Ireland. 15 April. 2010.  A Singh, NE Korres, AS Nizami, JD Murphy. 2010. Biomethane from agricultural waste: A clean vehicular biofuel. International Conference on Environment Energy and Development from Johensberg to Kopenhagen ICEED-2010, Sambalpur University, Orissa, India. December 10-12, 2010. Poster Presentations  AS Nizami, JD Murphy. Anaerobic digestion of grass using a continuously stirred tank reactor (CSTR). International Energy Agency Biofuels Symposium, University College Cork (UCC), Cork, Ireland. September 14, 2008.  AS Nizami, JD Murphy. Leach Beds coupled with UASB digester for optimal biogas production from Grass. In: Biofuels Directive to bio based Transport Systems in 2020, IEA Bioenergy Task 39, Subtask Policy and Implementation Workshop, Dresden, Germany, June 2-5, 2009. Abdul-Sattar Nizami vii Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table of Content Dedication……………………………………………………………………………………………………..¡¡ Acknowledgements…………………………………………………………………………………………....¡¡¡ Executive Summary…………………………………………………………………………………………...v Thesis output…………………………………………………………………………………………………v¡ Chapter 1: Introduction ............................................................................................................. 1 1.1 Introduction........................................................................................................................................................................ 2 1.1.1 Background ............................................................................................................................2 1.1.2 Thesis aims.............................................................................................................................2 1.1.3 Thesis objectives ....................................................................................................................2 1.2 Thesis in brief ..................................................................................................................................................................... 3 Chapter 2: What Type of Digester Configurations should be employed to Produce Biomethane from Grass Silage? ................................................................................................. 5 Abstract ...................................................................................................................................................................................... 6 2.1 Introduction........................................................................................................................................................................ 7 2.1.1 Focus of the paper .................................................................................................................7 2.1.2 Anaerobic digestion................................................................................................................7 2.1.3 Grass; a new way towards renewable energy ...........................................................................7 2.1.4 Biomethanation of grass silage ...............................................................................................8 2.1.5 Design of anaerobic digester ..................................................................................................8 2.1.6 Properties of grass silage ........................................................................................................9 2.2 Potential digester configurations .................................................................................................................................... 9 2.2.1 One-stage versus two-stage digesters ......................................................................................9 2.2.2 Dry versus wet digesters ...................................................................................................... 14 2.2.3 Batch versus continuous digesters ....................................................................................... 16 2.2.4 High-rate digesters .............................................................................................................. 16 2.3 Digester configurations suitable for grass silage ........................................................................................................ 16 2.3.1 Wet continuous digester ...................................................................................................... 16 2.3.2 Leach bed system connected with high-rate digesters .......................................................... 17 2.3.3 Dry continuous digester ...................................................................................................... 18 2.3.4 Batch digester...................................................................................................................... 18 Abdul-Sattar Nizami viii Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.4 Arguments for and against different digester configurations for grass silage digestion ..................................... 19 2.4.1 Biogas production per unit of grass silage based on volatile solids destruction ..................... 19 2.4.2 The wet continuous two- stage process ............................................................................... 20 2.4.3 The dry batch process ......................................................................................................... 20 2.4.4 The dry continuous process................................................................................................. 20 2.4.5 The leach bed system combined with UASB ....................................................................... 20 2.5 Discussion ......................................................................................................................................................................... 21 2.5.1 Potential improvements in anaerobic digesters .................................................................... 21 2.5.2 Research required ................................................................................................................ 21 2.6 Conclusion ........................................................................................................................................................................ 22 Acknowledgements ................................................................................................................................................................ 22 References ............................................................................................................................................................................... 22 Chapter 3: A review of the Integrated Process for the Production of Grass Biomethane ....... 27 Abstract .................................................................................................................................................................................... 28 3.1 Introduction...................................................................................................................................................................... 29 3.1.1 Characteristics of grass silage ............................................................................................... 29 3.1.2 Anaerobic Digesters ............................................................................................................ 30 3.1.3 Anaerobic biodegradation of grass silage ............................................................................. 30 3.1.3 Treatment; pre/during/post ................................................................................................ 31 3.2. Anaerobic reactors ......................................................................................................................................................... 34 3.2.1. Single Phase Digestion ....................................................................................................... 34 3.2.2 Two Phase Digestion ......................................................................................................... 35 3.3. Agronomic factors .......................................................................................................................................................... 36 3.3.1 Potential grass types ............................................................................................................ 36 3.3.2 Harvest date and frequency ................................................................................................. 39 3.3.3 Ensiling of grass .................................................................................................................. 40 3.4. Operational procedures ................................................................................................................................................. 42 3.4.1 Temperature........................................................................................................................ 42 3.4.2 pH range ............................................................................................................................. 42 3.4.3 Mixing ................................................................................................................................. 42 3.4.4 Particle size ......................................................................................................................... 43 3.4.5 Retention time..................................................................................................................... 43 3.4.6 Co-digestion ........................................................................................................................ 43 Abdul-Sattar Nizami ix Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.4.7 Nutrients, inoculum and inhibition ...................................................................................... 44 3.4.8 Recirculation of leachate/water ........................................................................................... 44 3.5. Pretreatments options .................................................................................................................................................... 44 3.5.1 Pre-treatment ...................................................................................................................... 44 3.5.2 Physical pre-treatment ......................................................................................................... 45 3.5.3 Chemical pre-treatment ....................................................................................................... 45 3.5.4 Thermal pre-treatment ........................................................................................................ 46 3.5.5 Biological pre-treatment ...................................................................................................... 47 3.5.6 Combining pre-treatments................................................................................................... 47 3.6. Research required to improve grass digestion ........................................................................................................... 47 3.6.1 Process control system ........................................................................................................ 47 3.6.2 Rumen fluid/saliva .............................................................................................................. 48 3.6.3 The use of fungi .................................................................................................................. 48 3.6.4 Laboratory scale up-scaling to commercial/industrial scale .................................................. 48 3.7. Conclusion ....................................................................................................................................................................... 49 Acknowledgements ................................................................................................................................................................ 49 References ............................................................................................................................................................................... 49 Chapter 4: Role of Leaching and Hydrolysis in A Two Phase Grass Digestion System ......... 58 Abstract .................................................................................................................................................................................... 59 4.1 Introduction...................................................................................................................................................................... 60 4.1.1 The benefits of grass biomethane ........................................................................................ 60 4.1.2 Technologies for grass mono-digestion ............................................................................... 60 4.1.3 Relevance and objective of research .................................................................................... 61 4.2 Hydrolysis and leaching .................................................................................................................................................. 61 4.2.1 Soluble and insoluble substrates .......................................................................................... 61 4.2.2 Leaching: sprinkling versus flooding .................................................................................... 62 4.2.3 Hydrolysis of grass .............................................................................................................. 62 4.2.4 Increased hydrolysis rate by pretreatments .......................................................................... 63 4.2.5 Focus of paper ................................................................................................................... 63 4.3 Materials and methods .................................................................................................................................................... 64 4.3.1 Feedstock ............................................................................................................................ 64 4.3.2 Experimental setup ............................................................................................................ 65 Abdul-Sattar Nizami x Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.3.3 Experimental procedure ...................................................................................................... 65 4.3.4 Analytical methods .............................................................................................................. 67 4.4 Results................................................................................................................................................................................ 67 4.4.1 Variation in VS and COD ................................................................................................... 67 4.4.2 Changes of grass structure because of leaching .................................................................... 68 4.5 Kinetic modelling ............................................................................................................................................................ 70 4.5.1 Characteristics of the model ................................................................................................ 70 4.5.2 Fitting parameters to the model ........................................................................................... 70 4.6 Application of experimental output and model ......................................................................................................... 72 4.6.1 One- stage and two- stage digestions.................................................................................. 72 4.6.2 Application of SLBR-UASB to grass digestion .................................................................... 72 4.6.3 Achievable methane yields from SLBR-UASB..................................................................... 74 4.6.4 Suggested operation of the SLBR-UASB ............................................................................. 74 4.7 Conclusions ...................................................................................................................................................................... 76 Acknowledgements ................................................................................................................................................................ 77 References ............................................................................................................................................................................... 77 Chapter 5: Design, Commissioning, and Start-Up of a Sequentially Fed Leach Bed Reactor Complete with an Upflow Anaerobic Sludge Blanket Digesting Grass Silage ....................... 79 Abstract .................................................................................................................................................................................... 80 5.1 Introduction...................................................................................................................................................................... 81 5.1.1 Significance of digester design ............................................................................................. 81 5.1.2 Monodigestion of grass silage .............................................................................................. 81 5.1.3 The dry batch process ......................................................................................................... 82 5.1.4 Sequentially fed leach bed reactor complete with upflow anaerobic sludge blanket .............. 82 5.1.5 Need for process control system ......................................................................................... 82 5.1.6 Focus of the paper .............................................................................................................. 82 5.2 Design of the digester system........................................................................................................................................ 83 5.2.1 Characteristics of the grass silage ......................................................................................... 83 5.2.2 Design of UASB reactor ...................................................................................................... 83 5.2.3 Digester scheme .................................................................................................................. 85 5.2.4 Digester components .......................................................................................................... 85 5.2.5 System Operation ................................................................................................................ 87 5.2.6 Process Control................................................................................................................... 89 Abdul-Sattar Nizami xi Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.3 Methodology and initial start-up ................................................................................................................................... 90 5.3.1 Analytical Methods .............................................................................................................. 90 5.3.2 Set-up of small batch experiments ....................................................................................... 90 5.3.3 Seeding the UASB reactor ................................................................................................... 91 5.3.4 Experiment layout for SLBR-UASB .................................................................................... 91 5.4 Problems and modifications .......................................................................................................................................... 91 5.4.1 Foam formulation ............................................................................................................... 91 5.4.2 U-tube connection .............................................................................................................. 92 5.4.3 Clogging .............................................................................................................................. 93 5.4.4 Start-up of UASB ................................................................................................................ 93 5.5 Results................................................................................................................................................................................ 94 5.5.1 Digester operation............................................................................................................... 94 5.5.2 Methane production ............................................................................................................ 94 5.5.3 UASB efficiency .................................................................................................................. 95 5.5.4 Dry solid removal efficiency ................................................................................................ 95 5.5.5 Volatile solid removal efficiency .......................................................................................... 97 5.5.6 pH ...................................................................................................................................... 98 5.6 Discussion ......................................................................................................................................................................... 98 5.6.1 Biological process ................................................................................................................ 98 5.6.2 Methane production ............................................................................................................ 98 5.6.3. Methane content of biogas ................................................................................................. 99 5.6.4 Retention times ................................................................................................................... 99 5.6.5 UASB efficiency ................................................................................................................ 100 5.6.6 Improving the efficiency of the system .............................................................................. 100 5.7 Conclusion ......................................................................................................................................................................100 Acknowledgements ..............................................................................................................................................................101 References .............................................................................................................................................................................101 Chapter 6: An Optimized Two Phase Digestion System for the Production of Gaseous Biofuel from Grass Silage, A High Solids Content Feedstock ...............................................105 Abstract ..................................................................................................................................................................................106 6.1 Introduction ...................................................................................................................................................................107 6.1.1 Anaerobic reactor design for high solid content feedstock ................................................. 107 6.1.2 Methane yields from grass biomethane digesters ............................................................... 107 Abdul-Sattar Nizami xii Green Grass: Developing Grass for Sustainable Gaseous Biofuel 6.1.3 The requirement for a stable UASB system ....................................................................... 107 6.1.4 Digestate: a value-added product ....................................................................................... 108 6.1.5 Previous work on two stage grass digestion ....................................................................... 108 6.1.6 Focus of paper .................................................................................................................. 108 6.2 Methodology ..................................................................................................................................................................109 6.2.1 Characteristics of grass silage ............................................................................................. 109 6.2.2 Scheme of SLBR-UASB system ........................................................................................ 110 6.2.3 Experimental set-up .......................................................................................................... 110 6.2.4 Analytical Methods ............................................................................................................ 111 6.3 Results..............................................................................................................................................................................112 6.3.1 Methane production .......................................................................................................... 112 6.3.2 UASB efficiency ................................................................................................................ 112 6.3.3 Dry solid removal efficiency .............................................................................................. 114 6.3.4 Volatile solid removal efficiency ........................................................................................ 114 6.3.5 Alkalinity and total volatile fatty acids ................................................................................ 114 6.3.6 pH .................................................................................................................................... 114 6.3.7 Morphology and structure of granule ................................................................................ 114 6.3.8 Digestate value .................................................................................................................. 116 6.4 Discussion of results .....................................................................................................................................................116 6.4.1 Stability of biological process ............................................................................................ 116 6.4.2 Methane production and retention time ............................................................................ 117 6.4.3 Methane content of biogas ................................................................................................ 117 6.4.4 UASB efficiency ................................................................................................................ 118 6.4.5 Alkalinity and TVFA ......................................................................................................... 118 6.4.6 UASB granule structure and morphology .......................................................................... 118 6.4.7 Digestate: a valuable product ............................................................................................. 119 6.5 Conclusions ....................................................................................................................................................................119 Acknowledgements ..............................................................................................................................................................120 References .............................................................................................................................................................................120 Chapter 7: How Can We Optimise Production of Biomethane from Grass Silage? ..............123 Abstract ..................................................................................................................................................................................124 7.1 Introduction....................................................................................................................................................................125 Abdul-Sattar Nizami xiii Green Grass: Developing Grass for Sustainable Gaseous Biofuel 7.1.1 Difficulties in comparing digester configurations from scientific literature ......................... 125 7.1.2 Comparability of different biomethane potential assays ..................................................... 125 7.1.3 Wet and dry digestion technologies ................................................................................... 125 7.1.4 Documented grass biomethane potential ........................................................................... 126 7.1.5 Focus of the paper ............................................................................................................ 126 7.2 Methodology ..................................................................................................................................................................128 7.2.1 Preparation of grass silage ................................................................................................. 128 7.2.2 Experimental set-up .......................................................................................................... 128 7.2.3 Analytical analysis .............................................................................................................. 130 7.3 Results..............................................................................................................................................................................130 7.3.1 Methane production .......................................................................................................... 130 7.3.2 Alkalinity and total volatile fatty acids ................................................................................ 131 7.3.3 Dry and volatile solid contents .......................................................................................... 131 7.3.4 Ammonia .......................................................................................................................... 131 7.3.5 pH .................................................................................................................................... 131 7.4 Discussion .......................................................................................................................................................................134 7.4.1 Biological process.............................................................................................................. 134 7.4.2 Organic loading rate .......................................................................................................... 134 7.4.3 Methane production .......................................................................................................... 134 7.4.4 Retention times ................................................................................................................. 135 7.4.5 Methane content ............................................................................................................... 135 7.4.6 Process inhibition .............................................................................................................. 135 7.4.7 Digestate and value added products................................................................................... 136 7.4.8 Post methanation potential ................................................................................................ 136 7.5 Conclusion ......................................................................................................................................................................136 Acknowledgements ..............................................................................................................................................................137 References .............................................................................................................................................................................137 Chapter 8: Recommendations ................................................................................................139 Recommendations ...............................................................................................................................................................140 Abdul-Sattar Nizami xiv Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 1: Introduction Abdul-Sattar Nizami 1 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 1.1 Introduction 1.1.1 Background Grass biomethane (renewable natural gas) has been shown to be a sustainable gaseous biofuel with an energy balance superior to alternative first generation indigenous liquid biofuels in Ireland. Grass biomethane has a minimum impact on sensitive environments; the crop (grass) and its storage mechanism (ensiling) is well understood by farmers. Grass is a perennial crop which sequesters carbon. It has been shown to allow economic viability both to the producer and the consumer. The gaseous biofuel industry in Ireland is in its infancy. Many facilities are at planning stage. Numerous farmers are interested in the industry due to the low farm family income associated in particular with beef farming. Production of grass biomethane requires a holistic approach to agriculture and bioenergy production; the variable nature of grass must be considered when choosing a digester configuration. A primary driver of this thesis is to assess the preferred digestion technology through comparison of different digester configurations using the same grass silage, cut from the same field, at the same time, ensiled in the same manner. Two digestion systems are designed and fabricated at small pilot scale. One is a Sequential fed Leach Bed Reactor complete with an Upflow Anaerobic Sludge Blanket (SLBR-UASB); the second is a 2-stage Continuously Stirred Tank Reactor (CSTR). A series of biomethane potential (BMP) assays were conducted at small and large scale to determine the upper limit for methane production. The same feedstock (grass silage) was used for all processes. 1.1.2 Thesis aims The aims of the thesis are to:  Undertake a literature review of all aspects of a grass biomethane system;  Identify appropriate anaerobic digestion technologies suitable for grass silage digestion. 1.1.3 Thesis objectives The objectives of this thesis are to:  Design, commission and operate two small pilot scale systems; a two-stage CSTR and a SLBR-UASB system;  Estimate the methane yields, the optimum organic loading rate and the residence time of grass silage in both systems;  Determine the upper limit for grass biomethane from various types of BMP assays; Abdul-Sattar Nizami 2 Green Grass: Developing Grass for Sustainable Gaseous Biofuel  Determine the optimum organic loading rates and residence time of the grass in different digester designs;  Evaluate the nutrient characteristics of the digestate for value added products. 1.2 Thesis in brief The thesis is composed of 8 chapters; chapters 2 to 7 are the original manuscripts of papers published/under review in international peer review scientific press. Each chapter is summarized below. Chapter 2. What Type of Digester Configurations should be employed to Produce Biomethane from Grass Silage? This chapter aims to explain and evaluate various technologies/configurations of anaerobic digestion (AD) and to suggest optimal configurations for producing biomethane from grass. AD technologies/configurations vary from one-stage batch dry systems to two-stage wet continuous systems; from one-stage continuous wet systems to two-stage two phase systems (e.g. a batch dry reactor coupled with a second stage high-rate reactor). The chapter reviews the scientific literature and AD practice at commercial scale. Chapter 3. A review of the Integrated Process for the Production of Grass Biomethane Production of biogas from grass requires a holistic approach to agriculture and bioenergy production. The structural and chemical composition of grass changes with different locations, soil type, time of harvesting (both date and time of day) and grass species. These factors along with different reactor configurations, operating conditions and pre-treatment affect methane production. The chapter reviews in detail the various processes and techniques involved in the production of grass silage for digestion, the potential for pretreatment, and digestion of grass silage. Chapter 4. Role of Leaching and Hydrolysis in A Two Phase Grass Digestion System This chapter focuses on the hydrolysis of grass in a batch leach bed reactor under aerobic conditions. Sprinkling and flooding conditions are assessed for bale and pit silages. A mathematical simulation of experimental analysis is developed to assess the optimal hydrolysis parameters (sprinkling or flooding and bale or pit silage). The model also allows the prediction of the methane production and retention time of the SLBRUASB system. Chapter 5. Design, Commissioning, and Start-Up of a Sequentially Fed Leach Bed Reactor Complete with an Upflow Anaerobic Sludge Blanket Digesting Grass Silage Abdul-Sattar Nizami 3 Green Grass: Developing Grass for Sustainable Gaseous Biofuel This chapter deals with design and commissioning of a small pilot scale Sequential fed Leach Bed Reactor coupled with an Upflow Anaerobic Sludge Blanket (SLBR-UASB). The chapter outlines in detail the numerous modifications that were required to bring the system to operational stability. Chapter 6. An Optimized Two Phase Digestion System for the Production of Gaseous Biofuel from Grass Silage, A High Solids Content Feedstock This chapter deals with the optimization of the SLBR-UASB system. The hydraulic retention time of the digester is reduced to assess the effect on methane production. The effect of adding a second pump to the system for higher hydrolysis rates is also investigated. The nutrient characteristics of the digestate is analysed for value added products. Chapter 7. Conclusions: How Can We Optimise Production of Biomethane from Grass Silage? This chapter is the conclusion of the thesis. The results of the two systems are compared to a series of biomethane potential (BMP) assays. The results indicate the range of potential values for methane production per kg volatile solid that may be obtained from the same sample of grass silage. Abdul-Sattar Nizami 4 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 2: What Type of Digester Configurations should be employed to Produce Biomethane from Grass Silage? Abdul-Sattar Nizami 5 Green Grass: Developing Grass for Sustainable Gaseous Biofuel What type of digester configurations should be employed to produce biomethane from grass silage? Abdul-Sattar Nizamia,b, Jerry D Murphya,b,1 aDepartment of Civil and Environmental Engineering, University College Cork, Cork, Ireland. bEnvironmental Research Institute, University College Cork, Ireland. Abstract Grass is an excellent energy crop; it may be classified as a high yielding, low energy input, perennial crop. Over 90% of Irish agricultural land is under grass; thus farmers are familiar with, and comfortable with, this crop as opposed to a “new energy crop” such as Miscanthus. Of issue therefore is not the crop, but the methodology of generating energy from the crop. Numerous farmers across Europe (in particular Germany and Austria) use grass silage as a feed-stock for biogas production; in a number of cases the produced biogas is scrubbed to biomethane and used as a transport fuel or injected into the natural gas grid. Many Irish farmers are considering converting from conventional farming such as beef production to grass biomethane production. Numerous technologies and combinations of such technologies are available; from one-stage batch dry systems to twostage wet continuous systems; from one-stage continuous wet systems to two-stage systems incorporating a batch dry reactor coupled with a second stage high rate reactor. This paper reviews work carried out both in the scientific literature and in practice at commercial scale. Keywords: grass silage; biogas; biomethane; anaerobic digester 1 Corresponding author. Tel +353 21 4902286 Fax +353 21 4276648 E-mail address: jerry.murphy@ucc.ie Abdul-Sattar Nizami 6 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.1 Introduction 2.1.1 Focus of the paper The number of on-farm digesters is increasing significantly across Europe. There is tremendous potential for these on-farm digesters using grass silage as a bioenergy feedstock, especially in Ireland where 91% of agricultural land is under grass [1]. The biogas generated may be used on site as a source of combined heat and power. Alternatively it may be upgraded to biomethane and distributed via the natural gas grid where better efficiencies and markets may be available. For example grass biomethane offers great potential as a transport fuel [1]; it may also be used as a source of renewable heat to existing housing stock connected to the gas grid. The industry in Ireland is in its infancy, however numerous technology providers are selling their wares; many facilities are at planning stage and numerous farmers are interested in the industry due to the low farm family income associated in particular with beef farming. The big issue is what type of digester should be utilized. This paper reviews the present state-of-the-art anaerobic digester configurations for high solid content feedstocks, and their application to grass silage. This paper has an ambition of explaining and evaluating various anaerobic digester configurations. 2.1.2 Anaerobic digestion The process of anaerobic digestion has now become a more attractive source of renewable energy due to reduced technological cost and increased process efficiency [2]. A plethora of substrates such as wastewaters [3], animal wastes [4] and sewage sludge [5] are extensively used for anaerobic digestion [6]. Additionally, during the last few years, the use of lignocellulosic substrates [7] and feedstocks with a high solids content such as the OFMSW (Organic Fraction of Municipal Solid Waste) [8], crops [9], crop residues [10] and grass silage [11, 12] have received considerable attention particularly in Europe [13]. Various researchers have reviewed and compared various digester types suitable for digesting solid wastes [14, 15]. Digesters which are optimized for OFMSW may not be ideal for grass silage because the volatile solid content of grass silage is of the order of 92% where as the volatile solid content of OFMSW may be as low as 60% [1] 2.1.3 Grass; a new way towards renewable energy In Ireland [16] and generally in temperate regions, grassland is the most predominant form of land use providing most of the feed requirements for ruminants [17] either through grazing or after conservation as hay or, more recently, silage [18]. Despite its ubiquity in European lands, there is still a considerable risk of its conversion into surplus land [19] if the land is not used productively. Nevertheless usage of grassland as a renewable source of energy through biogas production will contribute significantly to the protection of the environment, due to the ability of grass to sequester carbon into the soil matrix [20]. Additionally, many socioeconomic benefits [21] can be achieved without harming the food industry [22]; this is particularly true for Abdul-Sattar Nizami 7 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Ireland, where land area available to grow grass is 10 times more than for arable land. Furthermore, to affect a 10% reduction in emissions from the agriculture sector, the National Climate Change Strategy for Ireland [23] recommended reductions in the national herd. This eventually, in combination with the preservation of grassland, will necessitate grass growth as a new source of renewable energy in Ireland [1]. 2.1.4 Biomethanation of grass silage Optimal digestion of grass silage is an area still under active research. Most of the work on optimising grass silage digestion is conducted at laboratory and pilot trials. The interest in using grass silage as a feedstock for bioenergy and biorefinery systems is due to its high yield potential in terms of methane production per hectare [1]; however its lignin and cellulose content [24] makes it suitable as a multiple source of energy and products [25]. Grass silage is the most important substrate after maize silage for biogas plants in Germany [26] and one of the most used co-substrates in agricultural biogas plants between 2002 and 2004 in Germany [27]. Still the use of biomethane from substrates like grass silage in Europe is modest compared to other raw materials [28]. The high potential of methane production from grass silage has been shown in the studies of Amon et al. [29], Mähnert et al. [12] and Lehtomäki et al. [30]. 2.1.5 Design of anaerobic digester The optimization of digester design i.e. higher OLR (Organic Loading Rate), reduced HRT (Hydraulic Retention Time) and higher methane yields is of great importance [31]. Operational parameters such as HRT, mixing, number of tanks and temperature [14] along with the properties of the feedstock [32] form the basis of digester design. Moreover, in digesting lignocellulosic substrates such as grass silage, the dry matter content, the solubility and hydrolysis rates, play a critical role [33]. A detailed description of operational aspects of various digester types was undertaken by Hobson and Wheatley in 1993 [34]. There is a need to examine the digester configuration as applied to grass silage in 2009. Various digester configurations are employed which use different approaches such as one-stage or two-stage digesters [35], wet or dry/semi-dry digesters [15], batch or continuous digesters [36], attached or non-attached biomass digesters [37], high-rate digesters [38] and digesters with combination of different approaches (Figure 2.1). Abdul-Sattar Nizami 8 Green Grass: Developing Grass for Sustainable Gaseous Biofuel One stage digester Two-stage digester Dry digester Wet digester Continuous digester Batch digester One stage digester Two-stage digester Dry digester Wet digester Continuous Batch digester digester Figure 2.1. Possible combination of various digester types 2.1.6 Properties of grass silage Grass silage is wet (less than 20% dry solids content) or dry (20 - 40% dry solids) depending on whether it is wilted, weather conditions at time of harvesting and storage conditions (baled or pit) [1]. In Ireland the dry solids content of grass silage is of the order of 20% for pit silage and 30% for baled silage. The D-value reflects the digestibility and may be defined as the organic matter digested by the cow divided by the dry matter digested. The ME (Metabolizable Energy) value is defined as the energy available for the cow. In Scotland values for grass silage cut in early June were found to have a D-value of 67% and an ME Value of 10.7 MJ kg-1; for late cut silage (cut in late June) values of 65% and 10.4 MJ kg -1 were recorded [39]. Thus early cut grass silage is more digestible than late cut silage. In Southern Ireland D-values and ME values would be considerably higher [39]. 2.2 Potential digester configurations 2.2.1 One-stage versus two-stage digesters In one-stage digestion all the microbiological phases of anaerobic digestion occur in one tank [37]. In two-stage digestion different microbial phases can be separated [14]. Conversely two stage digestion may allow both stages to be complete microbial processes with the second stage incorporating storage of digestate and remedial gas collection [35]. When the microbial phases are separated the hydrolytic and acidification phases may occur in the first reactor and acetogenesis and methanogenesis occur in the second reactor [9]. The concept of twostage digestion is driven by optimization of the digestion process [40], resulting in potentially higher yields of biogas in smaller digesters (Table 2.1). Parawira et al. [43] scrutinized pilot and lab scale two-stage systems for Abdul-Sattar Nizami 9 Green Grass: Developing Grass for Sustainable Gaseous Biofuel anaerobic digestion of MSW, agricultural residues and market waste. The one-stage system is still popular at industrial scale because of the simplicity in operation, reduced costs and lesser technical problems (Table 2.2). Process reviews were undertaken by Weiland et al. [47] on one-stage digesters and by Demired and Yenigun, [48] on two-stage digesters. However the scientific literature is relatively sparse on one-stage digestion of grass silage [49], which is the normal application for commercial scale. Most of the studies conducted at laboratory and pilot scale use two-stage digesters (Table 2.3), which are not available at commercial scale. In the one-stage process, either dry batch systems or wet continuous systems are used [15], whereas in the two-stage process, continuous and wet processes are preferred (Figures 2.2-2.4). Table 2.1. Comparison of different digester configurations for high solid content feedstocks [14, 34, 41, 42] Criteria One-stage versus twostage digesters Dry versus wet digesters Batch versus continuous digesters One-stage Two-stage Dry Wet Batch Continuous Biogas production Irregular and discontinuous Higher and stable Higher Less and irregular Irregular and discontinuous Continuous Continuous and higher Solid content 10-40% 2-40% 20-50% 2-12% 25-40% 2-15% <4-15% Cost Less More Less More Less More More Volatile solids destruction HRT (days) Low to high High 40-70% 40-75% 40-70% 40-75% 75-98% 10-60 10-15 14-60 25 – 60 30-60 30-60 0.5-12 OLR (kg VS m-3 d-1) 0.7-15 10-15 for second stage 12 – 15 <5 12-15 0.7-1.4 10 -15 Abdul-Sattar Nizami 10 High-rate bioreactors Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 2.2. Comparison of process weaknesses and benefits of various digester types [14, 44-46] System One-stage versus two-stage digesters One-stage Two-stage Strengths  Simpler design  Less technical failure  Low cost Weaknesses  Higher retention time  Foam and scum formation    Dry versus wet digesters Dry Wet Batch versus continuous digesters Batch           Continuous High rate bioreactors Abdul-Sattar Nizami        Efficient substrate degradation owing to recirculation of digestate Constant feeding rate to second stage More robust process Less susceptible to failure Higher biomass retention Controlled feeding Simpler pre-treatment Lower parasitic energy demands Good operating history Degree of process control is higher No mixing, stirring or pumping Low input process and mechanical needs Cost-effective Simplicity in design and operation Low capital costs Higher biomass retention Controlled feeding Lower investment cost No support material is needed 11                Complex and expensive to build and maintain Solid particles need to be removed from second stage Complex handling of feedstock Mostly structured substrates are used Material handling and mixing is difficult Scum formation High consumption of water and energy Short-circuiting Sensitive to shock loads Channelling and clogging Larger volume Lower biogas yield Rapid acidification Larger VFA (Volatile Fatty Acid) production Larger start-up times Channelling at low feeding rates Table 2.3. Performance data of different anaerobic digesters applied for silage/grass digestion Abdul-Sattar Nizami Digester characteristics Operating temperature (oC) Mono/co-digestion Retention time (days) Characteristics of substrate Biogas yield (m3 /kg VS added) Methane yield (m3 /kg VS added) Prochnow et al. [11] Continuous digester, Laboratory scale Co-digestion 18-36 Extensive grassland cut, silage and hay 0.5-0.55 Not reported 20 Extensive grassland cut, silage 0.5-0.55 Not reported 25 Intensive grassland cut, first cut in June, fresh and ensiled 0.7-0.72 Not reported Extensive grassland cut, first cut in August, fresh, silage and hay Extensive grassland cut, silage and hay 0.54-0.58 Not reported 0.5-0.6 Not reported 35 Continuous digester, Farm scale Baserga and Egger, [50] 2 Batch digester, Laboratory scale 35 Co-digestion 12 Baserga, 1998 2 Continuous digester, Farm scale 35 Co-digestion 20 Extensive grassland cut, silage 0.5-0.55 Not reported Mähnert et al., 2002; Mähnert, 2002 2 Batch digester, Laboratory scale 35 Mono-digestion 28 0.65-0.86 0.31-0.36 35 Mono-digestion 28 Three grass species, first cut in mid-May, fresh ad ensiled Three grass species, second cut, ensiled 0.56-0.61 0.3-0.32 Semi-continuous digester, Laboratory scale Green Grass: Developing Grass for Sustainable Gaseous Biofuel Studies Table 2.3. Continued, Abdul-Sattar Nizami Studies Digester characteristics Operating temperature (oC) Amon et al., 2003 Batch digester, 2 Laboratory scale 37-39 Characteristics of substrate Mono-digestion Intensive forage mixture of grassland and clover, ensiled, 0.53 Mid-May (before anthesis i.e. when the flower is ready for pollination) 59 Biogas yield (m3 /kg VS added) mid-May (before End of May (anthesis 0.47 anthesis) ) Mid-June (after 0.42 ) anthesis) 13 Lemmer and Semi-continuous 37 Oechsner, 2002 2 digester, Laboratory scale and farm scale Lehtomäki et al. Batch leach bed-USB 37 [30] reactors Co-digestion Mono-digestion 25-60 55 Methane yield (m3 /kg VS added) 0.37 0.32 0.29 Grass from intensively 0.39 used sites, 4 cuts per year ensiled Not-reported Grass from extensively 0.22 used sites, 2 cuts per year, ensiled Not reported Grass from landscape 0.08 management Not-reported Grass silage (twostage leach bed process without pH) 0.27-0.39 0.197 Grass silage (two stage leach bed process with pH) 0.16 0.1 One stage leach bed process 0.2 0.06 These investigations investigations on on biogas biogas production production from from grassland grassland vegetation vegetation are are tabulated tabulated by by Prochnow Prochnow etet al. al. [11] [11] 22These Green Grass: Developing Grass for Sustainable Gaseous Biofuel Mono-codigestion Retention time (days) Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 2.4. Five year development in different digesters types [15] Period 1991-1995 1996-2000 2001-2005 2006-2010 (estimated) One-stage versus two-stage digesters Wet versus dry digesters One-stage 85% 91% 92% 98% Wet 37% 38% 59% 29% Two-stage 15% 9% 8% 2% Dry 63% 62% 41% 71% 2.2.2 Dry versus wet digesters Vandevivere and co-workers [14] classify dry and wet systems as follows. Digesters, in which the feedstock used consists of 20-40% dry matter, are known as dry anaerobic digesters; those with less than 20% dry matter are classified as wet digesters. Therefore, pre-treatment (i.e. pulping and slurrying) is required for grass silage in wet digesters. Currently, one-stage dry continuous (Figure 2.3) and dry batch digesters (Figure 2.4) are relatively new and innovative digesters used for MSW, biowaste and grass silage; their use is expected to continue in the coming years (Table 2.4). One-stage dry batch systems typically employ a system whereby high solids content feed-stock is entered into a vessel without initial dilution. Recirculation of water/leachate is employed. Feeding is by front actor, no mixing takes place, and as such parasitic energy demands are very low [14]. Vertical CSTR (Continuously Stirred Tank Reactor) configuration (Figure 2.2) is the most commonly used configuration in 90% of the newly erected wet digesters [27]. Parasitic energy demand for wet digesters is higher than for dry digesters due to the requirement to dilute grass silage, pump slurries and mix reactors for the total retention time (Table 2.2). Biogas Biogas Recycle Recycle Substrate Biogas Effluent Effluent Substrate One stage;wet/continuous/batch/dry/ CSTR Two- stage; wet/ Continuous /batch / CSTR Figure 2.2. Design variations in one-stage and two-stage digesters Abdul-Sattar Nizami 14 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Biogas Biogas Biogas Effluent Effluent Mixer Substrate Kompogas Effluent Substrate Substrate Biogas recirculation Mixer Valorga DRANCO Figure 2.3. Various types of one-stage dry continuous digesters [14] a Biogas b Biogas Biogas Biogas Biogas Biogas Substrate UASB Percolation liquid storage tank New Pump Mature Old Two-stage / Sequential-batch Hybrid batch-UASB Biogas c Leach beds Substrate UASB Leachate tank Figure 2.4. (a) One-stage dry batch digester [14] (b) Two-stage dry batch digesters [14] (c) Sequencing fed leach bed digesters coupled with UASB Abdul-Sattar Nizami 15 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.2.3 Batch versus continuous digesters In batch digesters (Figure 2.4), the reactor vessel is loaded once with raw feedstock for a certain period of time (and inoculated with digestate from another reactor). It is then sealed and left until complete degradation has occurred [51]. On the contrary, in continuous digesters (Figure 2.2), the substrate is regularly and continuously fed either mechanically or by force of the newly entered substrate [52]. In continuous digesters, plug flow, CSTR, anaerobic filters and UASB (Upflow Anaerobic Sludge Blanket) systems are used, while in batch digesters, one-stage, sequential batch and hybrid batch digesters are used. According to Bouallagui et al. [53], about 90% of the industrial scale plants currently operating in Europe are different continuous type digesters in configurations such as a continuous onestage digester [54] used for anaerobic digestion of OFMSW, solid waste and biowaste. However, batch digesters maybe more suitable for grass silage digestion due to the dry solid contents (bailed silage has a solids content of about 32%) and fibrous characteristics of grass silage and the reduced parasitic energy demands (Table 2.2). This is particularly advantageous when using more than one batch digester with different start up times to guarantee a continuous yield of biogas [55]. 2.2.4 High-rate digesters In these digesters high solid retention time is achieved through attachment of biomass to high density carriers and formation of highly settleable granules [56]. Upflow anaerobic filters, UASB, anaerobic packed-bed and fluidized bed reactors are utilised as high-rate digesters both at lab and industrial scale. The use of UASB among high rate digesters has increased and widened in recent years [57] by taking feed with solid contents less than 4% or up to 15% [41, 42] at retention times of 0.5 to 12 days [14]. Marchaim [41] suggests solids content of less than 4% in UASB, while Barnett et al. [42] allow for solids content of up to 15% in UASB (Table 2.1). Moreover, the UASB reactor is suggested by various authors [58, 59] to offer benefits over other high-rate digesters when applied to high organic loading rates. For digestion of grass silage, high-rate digesters are applied in connection with leach beds [30, 35], or may be used with CSTR in two or multi-stage fashion. 2.3 Digester configurations suitable for grass silage 2.3.1 Wet continuous digester The popularity of the one-stage and two-stage CSTR systems in wet continuous digesters is due to the simplicity of the system in design and operation and the low capital costs (Table 2.2). The digestion of grass silage in the CSTR is facilitated by the use of a separate preprocessing tank with chopper pump, screw-feeder and flushing system (Figure 2.5). In addition, solid contents are reduced by recirculation and mixing of leachate with fresh matter [55]. In a laboratory scale one-stage CSTR, the biogas yield of three fresh grass species as mono-substrate was 0.61 and 0.56 m3 kg-1 VS added at an OLR of 0.7 and 1.4 kg-1 VS added at 35oC [12]. While, at commercial scale a two-stage CSTR in Eugendorf, Austria, the methane yield from grass silage as mono-substrate is 0.3 m3 kg-1 VS at an OLR of 1.4 kg VS m-3 day-1 [60]. In Eugendorf the biogas was 55% methane, thus the biogas yield was 0.55 m 3 kg-1 VS at 1.4 kg VS m-3 day-1; almost the same result as Mähnert and co-workers [12]. Abdul-Sattar Nizami 16 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.3.2 Leach bed system connected with high-rate digesters In this system one or several solid-bed reactors (leach beds) [61] are connected with high-rate digesters such as a UASB or an anaerobic filter [35]. These leach beds are sequentially batch loaded; leachate is recirculated [62] to facilitate a near continuous system in terms of biogas production. The system requires a high conversion of volatile solids to COD (Chemical Oxygen Demand) in the leach beds followed by a high rate of conversion of COD to methane in the UASB to effect stable and relatively high biogas production (Table 2.5). Liu et al. [64] reported that green waste can be digested within 12 days yielding steady and high production of biogas. In a study by Lehtomäki et al. [30] using a two-stage process, the total methane yield originating from the UASB was 76-98%; the remaining biogas was produced in the leach beds using grass silage as digester substrate. In another study by Lehtomäki and Björsson [65] at pilot scale using batch leach bed digesters coupled with an anaerobic filter, methane yields of 0.39 m3CH4 kg-1 VS added were obtained at 59% VS removal after 50 days digestion of grass silage. With the same digester scheme using grass silage at pilot scale, Yu et al. [35] and Cirne et al. [66] obtained 0.165 m3 CH4 kg-1 VS added and 0.27 m3 CH4 kg-1 VS added at 67% VS and 60% VS removal respectively. Lehtomäki et al. [30] explained the lower CH4 production at higher volatile solids destruction in the following manner. The grass mixtures used as substrates in these studies had different composition of lignin and nitrogen. Higher contents of lignin mean lesser contents of volatile solids, which results in less overall degradation and less methane. Lower contents of nitrogen in grass silage result in a less than optimal C:N ratio, which effects microbial degradation, which ultimately affects the methane concentration of biogas. Table 2.5. Comparison of the optimal anaerobic digesters for grass silage [14, 41, 42, 60, 63] System Example Pretreatment Process Quality of digestate HRT (days) Solid contents (%) Operating temperature (oC) Cost Destruction of volatile solids (%) OLR (kg VS m-3 d-1) Wet continuous one/two-stage digester CSTR Pulping, chopping, slurry, hydrolyzed Twostage (can be onestage) Juice rich in protein and nutrients, soil conditioner >60 2-14 35-40 Med ium 40-70 0.7-1.4 Two-stage sequential batch digester connected with high-rate bioreactor Leach bed with UASB Chopping, pulping Two or multistage Soil conditioner, fertilizer, fibrous materials 12 20-40 35 Hig h 75-98 from UASB 10 -15 One-stage dry continuous digester DRANCO Shredding, Chopping One stage Dewatered, good, fibrous materials 15-30 20-50 50-58 Low 40-70 12 One or multistage dry batch digester BEKON Chopping Onestage Dewatered, good, fibrous materials 40-70 30-40 35 Low 40-70 12-15 Abdul-Sattar Nizami 17 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.3.3 Dry continuous digester DRANCO [67, 68], Kompogas and Volarga systems are dry continuous systems [14]. In DRANCO, the digestate/leachate is recirculated back vertically, while the Kompogas system works horizontal (Figure 2.3). Both systems may operate in the thermophilic temperature range. Slow moving impellers are used in Kompogas to homogenize and re-suspend denser particles [69]. The Valorga system operates in the mesophilic temperature range; this system employs recirculation of biogas at the bottom of the reactors through injection ports to effect mixing [70]. One technical drawback with the Valorga digester is clogging of the gas injection ports; maintenance of these systems is difficult [14]. A high level of OLR is achievable in the DRANCO process (Table 2.5). The DRANCO plant at Nustedt, Germany treats 12,500 t a-1 of agricultural crops. The feedstock comprises maize (6200 t a -1), sunflowers (2400 t a-1), rye (2000 t a-1) and grass (600 t a-1). The grass adds biogas at a rate of 90-120 Nm3 t-1. The total biogas production is 145 Nm3 t-1 [71]. Figure 2.5. Preprocessing systems for high solid substrates [57] 2.3.4 Batch digester Dry batch digesters, such as the BEKON processes (Figure 2.4a), are used widely in Europe for dry solids content up to 50% [63]. In this digester type, the leachate is recirculated/sprayed back on to the feedstock. After completion of digestion, the digester is reopened, half unloaded and half of the feedstock is left as inoculum; it is refilled with fresh feedstock and the cycle continues [51]. In addition to BEKON, garage type, bag type, immersion liquid storage vat type and wet-dry combination digesters are in the first phase of commercialization [72]. In a lab-scale batch digester, the biogas and methane yield of fresh and ensiled grass species were examined and reported by Mähnert et al. [12]. The observed biogas and methane yield were in the range of 0.65-0.86 and 0.31-0.36 m3 kg-1 VS respectively. With the same digester scheme at laboratory scale, Baserga and Egger [50] and KTBL, [73] reported the values of fresh cut grass in the range of 0.5-0.6 m3 biogas kg-1 VS added. The values for biogas production from grass silage in m3 kg-1 VS added range from 0.54 by Linke et al. [74], 0.58 by Niebaum and Döhler [75], 0.63 by KTBL [73] to 0.81 by Jäkel [76]. Among different grass species both fresh and ensiled, Mähnert and coworkers [12] found the highest biogas yields (0.83 and 0.86 m 3 kg-1 VS added) were achieved for perennial ryegrass and the lowest (0.72 and 0.65 m3kg-1 VS added) for fresh cocksfoot and silage. Abdul-Sattar Nizami 18 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.4 Arguments for and against different digester configurations for grass silage digestion 2.4.1 Biogas production per unit of grass silage based on volatile solids destruction Various grass silage samples from a farm in Cork, Ireland were analysed by the authors. The dry solids content of grass silage varied from 20-40%. The volatile dry solid contents for various tests averaged 92% when expressed as a percentage of dry solids. An ultimate analysis yielded the following stoichiometric equation for dry grass: C28.4H44.5O17.7N. The ratio of carbon to nitrogen (C:N) in grass silage was found to be 25:1 which is very suitable for digestion [39]. The biogas production per unit of grass silage is an essential descriptor of a system. Box 2.1 highlights the biogas production associated with 60% destruction of volatile solids as follows:  509 L CH4 kg-1 VS destroyed or 305 L CH4 kg-1 VS added;  945 L biogas kg-1 VS destroyed or 567 L biogas kg-1 VS added; It is shown that the CH4 content is of the order of 54% by volume. Improvements upon these figures necessitate a destruction of volatile solids in excess of 60%. These figures are in line with the preceding literature; including such data as Mähnert et al. [12] who generated levels of 560 L biogas kg -1 VS added at 1.4 kg VS m-3 day-1. Box 2.1. Biogas production per unit of grass from first principles Stoichiometry: C28.4H44.5O17.7N + 8.425H2O → 15.335 CH4 + 13.065 CO2 668.5 + 151.6 → 245 + 575 820 → 820 300kg solids + 68 kg water → 110kg CH4 + 258kg CO2 (30% dry solids) 276kg VS + 62.5 kg water → 101kg CH4 + 237kg CO2 (92% volatiles) 165kg VSdest + 37.5 kg water → 60.6kg CH4 + 142kg CO2 (60% destruction) Density of CH4 = 16/22.412 m3 kg-1 = 0.714kg m-3, Density of CO2 = 44/22.412 m3 kg-1 = 1.96kg m-3 Thus the proportion of gas by volume → 84m3CH4 + 72m3CO2 = 156m3 biogas Thus biogas contains approximately → 53.8% CH4 + 46.2% CO2 by volume Energy Balance: 1m3 CH4 ≈ 37.78MJ 1m3 biogas @ 53.8% CH4 = 20.3MJ m-3 1 t VS = 18.77 GJ [77] 84 m3 CH4 = 3.17 GJ; 165 kg VS dest = 3.10 GJ Biogas production per unit: 84 m3 CH4/ 165 kg VS dest = 509 L CH4 /kg VS dest = 305 L CH4 kg-1 VS added 156 m3 biogas/ 165 kg VS dest = 945 L biogas kg-1 VS dest = 567 L biogas kg-1 VS added Abdul-Sattar Nizami 19 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 2.4.2 The wet continuous two- stage process Two tanks are employed in series with recirculation of leachate to dilute the grass feedstock [60]. Maceration is required to reduce the potential for the grass silage to get caught in moving parts [39]. This type of system is in place in Eugendorf, Austria. The retention time for grass silage is relatively high (60 days) and the OLR is correspondingly low (ca. 1.4 kg VS m-3d-1). Gas production of 300 L CH4 kg-1 VS added was recorded in operating plants at a volatile solids destruction of 60% [60]. Of concern is the tendency of grass silage to float on the liquid surface of the digester; this may be overcome through good mixing design such as paddle system that breaks the liquid surface [39]. 2.4.3 The dry batch process The dry batch has a significant advantage; simplicity [63]. There are few moving parts; little pretreatment is required as the grass silage does not come into contact with moving parts; the feedstock is not diluted; as a result energy input is low (Table 2.2). The time between loading and unloading is greater than 30 days, but as half the substrate is left in place as an inoculum, the actual retention times is of the order of 45 days. Gas production starts from zero, increases, peaks and decreases; thus a series of batch digesters are required which are fed sequentially to generate a gas curve with a relatively constant output [51]. A disadvantage of the process is the lack of facilities actually operating on grass silage. We simply do not know if grass silage is suitable for vertical garage door batch digester systems. These systems were designed originally for treatment of OFMSW (in lieu of composting). OFMSW has a lower volatile solids content [1] and thus a higher quantity of solids in the digestate. There is a fear that grass digestate, which has a lower solids content than OFMSW digestate, will flow out the vertical door of the digester on emptying. It is perceived that the batch process will not effect the volatile solids reduction of a continually mixed wet process and will not therefore generate the same level of gas production. 2.4.4 The dry continuous process Again the dry continuous process was designed originally for biowaste and OFMSW. The low solids content of the digestate from grass silage may be problematic, especially for pumping. There is little recorded evidence of grass silage digestion in these systems. Disadvantages will potentially include: requirement for size pretreatment; requirement to pump the digestate up the digester a number of times; significant energy input (ca. 80kWeh t-1 feedstock is documented for OFMSW) [68]. This is extremely high when compared to the dry batch process that simply involves using a front actor to push substrate into a chamber. 2.4.5 The leach bed system combined with UASB This system as outlined in Figure 2.4c was shown by Lehtomäki et al. [30] to pull the gas production curve to the left as compared to the dry batch process; in effect this significantly shortens the required retention time. This system is different to the others in that the final reactor (the UASB) receives a liquid waste high in COD. The benefit of the UASB is that it can be loaded to 20 kg COD m-3day-1 while effecting a 90% destruction of COD [77]. Nizami and co-workers [77] showed that 1 kg of VS destroyed generates 1.4 kg of COD and 1 kg of COD destroyed produces Abdul-Sattar Nizami 20 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 350 L CH4. Thus if the UASB effects a 90% destruction in volatile solids then each kg of VS destroyed can generate 441 L of CH4. Therefore an increase in volatile solids destruction is required to obtain the same gas production. In section 2.4.1 it is shown that 509 L CH4 is generated per kg volatile solids destroyed in a CSTR system; to effect the same methane production 69% destruction of volatiles needs to occur in the leach bed. The leach bed may be optimized through thermal or enzymatic treatments [77]. Lehtomäki and Björnsson [65] used batch leach bed digesters coupled with an anaerobic filter to obtain methane yields of 390 L CH4 kg-1 VS added after 50 days digestion of grass silage. This may be compared with 305 L CH4 kg-1 VS added for 60% destruction of volatiles in section 2.4.1. 2.5 Discussion 2.5.1 Potential improvements in anaerobic digesters So far limited and inconclusive work has been undertaken to improve the design of reactors for enhanced biogas production from grass silage. The ubiquitous wet continuous process (CSTR) has been shown to be an effective process for grass silage digestion at loading rates of approximately 1.4 kg VS m -3d-1[60]. The UASB reactor offers great potential as it can withstand loading as high as 20 kg COD m -3 d-1; thus by converting the feedstock from volatile solids to soluble COD, loading rates may be increased, retention times may be shortened, and optimistically biogas production may be increased [77]. The leach bed may be optimized for hydrolysis through thermal and enzymatic pretreatment of the grass silage while the UASB may be optimized for COD removal. 2.5.2 Research required Data from the literature on the best digester configuration using grass silage as a feedstock is inconclusive. There is a tendency to utilize data on digester configurations utilizing different feedstocks (OFMSW, green waste) and to apply the outcome to grass silage. The feedstock is an important criterion. High solid content substrates may require different reactor configurations; food waste and the OFMSW though both having similar dry solids contents to grass, are very different to grass silage. Grass silage has a far higher volatile solids content than OFMSW (ca. 92% versus 60%) [1] and thus a more complete dry matter removal takes place in the reactor. The digestate from grass silage thus has a lower solids content than the digestate from OFMSW; it is more liquid in nature. This has led to suggestions that dry batch reactors fed through a vertical door are more suitable for OFMSW than for grass silage as the grass silage digestate may require additional handling due to the vertical containment of a digestate low in solids content. There is another argument that bailed silage (32% dry solids content) is more suited to a dry batch digestion process while clamp or pit silage (22% dry solids content) is more suited to a wet continuous process. These arguments need to be interrogated. Even digester configurations assessed using grass silage is problematic. Of significant concern is the comparison of data from different countries, different institutes, digesting different species of grass, at different dry solids content, cut at different times of year and times of day [77]. Nizami and co-workers [77] outlined the significant difference in digestibility of grass; for example the water soluble carbohydrates are higher in the afternoon than the morning resulting in higher biogas production for grass cut in the afternoon. It is suggested that a number of reactor Abdul-Sattar Nizami 21 Green Grass: Developing Grass for Sustainable Gaseous Biofuel configurations should be compared in real time treating similar quantities of grass silage under similar loading rates to evaluate the optimal configuration. Indeed grass cut at different times of the year and from different locations can have different D-values, lignin contents and N (Nitrogen) values [77]. Thus as shown by Lehtomäki et al. [30] in section 2.3.2, incorrect conclusions may be drawn from direct comparisons of reactors treating grass silage in different plants (or labs) from different countries, cut at different times of the year and hence with different compositions. 2.6 Conclusion Grass digestion offers opportunities to farmers. Instead of managing herds of beef cattle, an industry which offers a low Farm Family Income, the farmer can continue to draw down single farm payments while cutting silage two or three times a year as a feedstock for an anaerobic digester. Biogas/biomethane is now the end product rather than beef; the work load is considerably less; greenhouse gas production is eased significantly [1]. This is in line with the objective of the National Climate Change Strategy for Ireland to reduce the herd and to effect an overall reduction in greenhouse gas emissions from the agriculture sector by 10% [23]. Many digester systems are available for the farmer to choose. Technology providers are selling digesters that were not initially designed for grass silage. The authors believe that much work needs to be undertaken to ascertain optimal digester configurations for production of grass biomethane. The CSTR system appears to be a safe technology if the mixing system is adapted to deal with the tendency for grass silage to float. However there is a significant potential to assess the benefits of leach beds followed by a high rate digester (such as a UASB). The leach beds will permit pretreatment technologies (thermal and enzymatic) to optimize hydrolysis while the UASB may be optimized for methanogensis. Optimal configurations can only be established by operating the different configurations in parallel, in real time, digesting the same grass silage feed-stock. Acknowledgements Nicholas E Korres, Anoop Singh, Beatrice M Smyth, and Thanasit Thamsiriroj for advice, brainstorming sessions, conversations, and critiques. Funding sources: Department of Agriculture and Food (DAFF) Research Stimulus: “GreenGrass.”; Environmental Protection Agency (EPA) Strive Programme: “Compressed biomethane generated from grass used as a transport fuel.” References [1] Murphy JD, Power NM. An argument for using biomethane generated from grass as a biofuel in Ireland. Biomass and Bioenergy. Volume 33 Issue 3 (March 2009) 504-512. [2] EEA (European Environment Agency), briefing, 02. How much biomass can Europe use without harming the environment? 2005. ISSN: 1830-2246. [3] Ramakrishna C, Desai JD. High rate anaerobic digestion of a petrochemical wastewater using biomass support particles. World J. Microbiol. Biotechnol. 13 (1997) 329-334. Abdul-Sattar Nizami 22 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [4] Moller HB, Sommer SG, Ahring B. Methane productivity of manure, straw and solid fractions of manure. Biomass and bioenergy 26 (2004) 485-495. [5] Edelmann W, Engeli H, Gradenecker, M. Co-digestion of organic solid waste and sludge from sewage treatment. Water science and technology 41 (2000) 213-221. [6] Verstraete W, Vandevivere P. New and broader applications of anaerobic digestion. Crit. Rev. Environ. Sci. Technol. 29 (1999) 151-173. [7] Petersson A, Thomsen MH, Hauggaard-Nielsen H, Thomson, AB. Potential bioethanol and biogas production using lignocellulosic biomass from winter rye, oilseed rape and faba bean. Biomass and bioenergy 31 (2007) 812819. [8] Macias-Corral M, Samani Z, Hanson A, Smith G, Funk P, Yu H, Longworth J. Anaerobic digestion of municipal solid waste and agricultural waste and the effect of co-digestion with dairy cow manure. Bioresource technology. 2008; 99(17): 8288-93. [9] Gunaseelan VN. Anaerobic digestion of biomass for methane production: a review. Biomass and bioenergy. 1997; 13(1/2): 83-114. [10] Bohn I, Bjornsson L, Mattiasson B. The energy balance in farm scale anaerobic digestion of crop residues at 1137 oC. Process biochemistry 42 (2007) 57-64. [11] Prochnow A, Heiermann M, Drenckhan A, Schelle H. Seasonal Pattern of Biomethanisation of Grass from Landscape Management. Agricultural Engineering International: the CIGR Ejournal. Manuscript EE 05 011. Vol. VII, 2005. [12] Mähnert P, Heiermann M, Linke B. Batch- and semi- continuous production from different grass species. Agricultural Engineering International: The CIGRE E Journal. Manuscript EE 05 010. Vol. V11, 2005. [13] Braun R, Steffen R. Anaerobic digestion of agroindustrial byproducts and wastes. In: Anaerobic conversions for environmental protection, sanitation and re-use of residues. REUR Technological series 51, 1997 (ISSN 10242368), pp. 27-41. Rome, FAO. [14] Vandevivere P, De Baere L, Verstraete W. Types of anaerobic digester for solid wastes. In: Mata-Alvarez J, editor. Biomethanization of the organic fraction of municipal solid wastes. London: IWA Press. p 112-140, 2003. [15] De Baere LD, Mattheeuws B. State-of-the-art 2008 - Anaerobic digestion of solid waste. Waste management world 2008, Volume 9, Issue 5. [16] Feehan J. Farming in Ireland. History, Heritage and Environment. 2003. University College Dublin, Faculty of Agriculture. [17] Hopkins A, Wilkins RJ. Temperate grassland: key developments in the last century and future perspectives. Journal of agricultural sciences 144 (2006) 503-523. [18] Brockman JS, Wilkins RJ. Grassland. In: Primrose McConnell's The Agricultural Notebook, 20th edition. 2003. Ed. Soffe R. J. Blackwell Publishing. [19] Rounsevell MDA, Ewert F, Reginster I, Leemans R, Carter TR. Future scenarios of European agricultural land use: II. Projecting changes in cropland and grassland. Agriculture, ecosystems and environment. 2005; 107(23):101–116. [20] Tilman D, Hill J, Lehman C. Carbon-negative biofuels from low-input high-diversity grassland biomass. Science 314 (2006) 1598-1600. Abdul-Sattar Nizami 23 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [21] Baier U, Grass S. Bioraffination of grass, Anaerobic digestion 2001–9th world congress for anaerobic conversion for sustainability, Antwerpen 2001. [22] Rösch C, Raab K, Stelzer V. Surplus grassland-a new source of bio-energy? 14th European biomass conference, 17-21 October, 2005, Paris, France. [23] National Climate Change Strategy Ireland October 2000. Available from: http://www.environ.ie/DOEI/doeipub.nsf/0/7d411c497cb4fbdb80256f88003b0961/$FILE/pccexsuminside %5B1%5D.pdf. [24] Lewandowski I, Scurlock JMO, Lindvall E, Christou M. The development and current status of perennial rhizomatous grasses as energy crops in the U.S. and Europe. Biomass and bioenergy 25(2003) 335–361. [25] Kromus S, Wachter B, Koschuh W, Mandle M, Krotscheck C, Narodoslawsky M. The green biorefinery austriadevelopment of an integrated system for green biomass utilization. Chemical and biochemical engineering quarterly. 2004; 18 (1): 7-12. [26] Rösch C, Raab K, Sharka J, Stelzer V. Sustainability of bioenergy production from grassland concept, indicators and results. 15th European biomass conference and exhibition, 7-11 May 2007, Berlin, Germany. [27] Weiland P. Biomass digestion in agriculture: a successful pathway for the energy production and waste treatment in Germany. Eng. Life Sci. 6, 2006. No.3 [28] Abraham ER, Ramachandran S, Ramalingam V. Biogas: can it be an important source of energy? Env Sci Pollut Res. 2007; 14 (1): 67-71. [29] Amon T, Kryvoruchko V, Amon B, Zollitsch W, Pötsch E. Biogas production from maize and clover grass estimated with the methane energy value system. In: EurAgEng: Eng2004 Engineering the Future, 12 – 16, Leuven, Belgium. [30] Lehtomäki A, Huttunen S, Lehtinen TM, Rintala JA. Anaerobic digestion of grass silage in batch leach bed processes for methane production. Bioresource technology 99 (2008) 3267-3278. [31] Ward AJ, Hobbs PJ, Holliman PJ, Jones DL. Optimization of the anaerobic digestion of agricultural resources, review. Bioresource technology 99 (2008) 7928-7940. [32] Igoni AH, Ayotamuno MJ, Eze CL, Ogaji SOT, Probert SD. Design of anaerobic digesters for producing biogas from municipal solid-waste. Applied energy 85 (2008) 430-438. [33] Qi BC, Aldrich C, Lorenzen L, Wolfaardt GW. Acidogenic fermentation of lignocellulosic substrate with activated sludge. Chemical engineering communications 192 (2005) 1221-1242. [34] Hobson PN, Wheatley AD. Anaerobic digestion: modern theory and practice. Essex, UK: Elsevier science Publishers. 1993. [35] Yu HW, Samani Z, Hanson A, Smith G. Energy recovery from grass using two-phase anaerobic digestion. Waste management 22 (2002) 1-5. [36] Chowdhury RBS, Fulford DJ. Batch and semi continuous anaerobic digestion system. Renew Energy Int J. 1992. 24/5, pp. 391–400. [37] Callander IJ, Barford JP. Recent advances in anaerobic digestion technology. Process biochemistry. August 1983: 24-30. [38] Pol LH, Lettinga G. New technologies for anaerobic wastewater treatment. Wat. Sci. Tech. 1986; 18(12): 41-53. [39] Dieterich, B. Energy crops for anaerobic digestion (AD) in Westray. Report written for Heat and Power Ltd., Westray, Orkney, UK. Available from: colin@crisenergy.co.uk Abdul-Sattar Nizami 24 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [40] Liu GT, Peng XY, Long TR. Advance in high-solid anaerobic digestion of organic fraction of municipal solid waste. Journal of central south university of technology 13 (2006) 151–157. [41] Marchaim U. Biogas processes for sustainable development. 1992. FAO Agricultural services bulletin. Rome. [42] Barnett A, Pyle L, Subramanian SK. Biogas technology in the third world: a multi-disciplinary review. 1978. Ottawa: IDRC Publications. [43] Parawira W, Read JS, Mattiasson B, Bjornsson L. Energy production from agricultural residues: high methane yields in pilot-scale two-stage anaerobic digestion. Biomass and Bioenergy 32 (2008) 44-50. [44] Liu G, Zhang R, Li X, Dong R. Research progress in anaerobic digestion of high moisture organic solid waste. Agricultural Engineering International: the CIGR Ejournal. Invited overview No.13. Vol. IX. 2007. [45] Deublein D, Steinhauser A. Biogas from waste and renewable resources: an introduction. Publisher, 2008. Wiley-VCH. [46] Rajeshwari KV, Balakrishnan M, Kansale A, Lata K, Kishore VVN. State-of-the-art of anaerobic digestion technology for industrial wastewater treatment. Renewable and sustainable energy reviews 4 (2000) 135-156. [47] Weiland P, Melcher F, Rieger Ch, Ehrmann Th, Hel.rich D, Kissel R. Biogas-Messprogramm – Bundesweite Bewertung von Biogasanlagen aus technologischer Sicht. Report, published by FAL, 2004. Bundesforschungsanstalt fur Landwirtschaft. [48] Demired B, Yenigun O. Two-phase anaerobic digestion process: a review. J Chem. Technol. Biotechnol 77 (2002) 743-755. [49] Sachs JV, Meyer U, Rys P, Feikenhauer H. New approach to control the methanogenic reactor of two-phase anaerobic digestion system. Water Res. 37 (2003) 973-982. [50] Baserga U, Egger K. Vergärung von Energiegras zur Biogasgewinnung. Bundesamt für Energiewirtschaft, Forschungsprogramm Biomasse, 1997. Tänikon, 41 p. [51] Parawira W. Anaerobic treatment of agricultural residues and wastewater, Application of high-rate reactors. PhD thesis. 2004. Department of Biotechnology, Lund University, Sweden. [52] Demirbas A, Ozturk T. Anaerobic digestion of agricultural solid residues. Int J Gren Energy 4 (2005) 483–494. [53] Bouallagui H, Touhami Y, Cheikh RB, Hamdi M. Bioreactor performance in anaerobic digestion of fruit and vegetable wastes. Process Biochemistry 40 (2005) 989-995. [54] Lissens G, Vandevivere P, De Baere L, Biey EM, Verstrae W. Solid waste digesters: process performance and practice for municipal solid waste digestion. Water Sci. Technol. 44 (2001) 91–102. [55] Weiland P. Production and energetic use of biogas from energy crops and wastes in Germany. Applied biochemistry and biotechnology. 2003. Vol.109. [56] Lettinga G. Anaerobic digestion and wastewater treatment systems. Antonie van Leeuwenhoek 67 (1995): 3–28. [57] Weiland P, Rozzi A. The start up, operation and monitoring of high rate anaerobic treatment systems: discussers report. Water Sci. Technol. 24 (1991) 257–77. [58] Bal AS, Dhagat NN. Upflow anaerobic sludge blanket-a review. Indian journal of environmental health 2001; 43 (2) 1-82. [59] Paula JDR, Foresti E. Kinetic studies on a UASB reactor subjected to increasing COD concentration. Water Science and technology 1992; 25 (7) 103-111. [60] Energiewerkstatt. Projekt: Graskraftwerk Reitbach Biogas aus Wiesengras – Energie ohne Ende, Technisches Büro und Verein zur Förderung erneuerbarer Energie, Energiewerkstatt, 7 November 2007. Abdul-Sattar Nizami 25 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [61] Lai TE, Nopharatana A, Pullammanappallil PC, Clarke WP. Cellulolytic activity in leachate during leach-bed anaerobic digestion of municipal solid waste. Bioresource technology 80 (2001) 205-210. [62] Chynoweth DP, Owens J, O‟Keefe D, Earle JFK, Bodch G, Legrand R. Sequential batch anaerobic composting of the organic fraction of municipal solid waste. Water Sci. Technol., 1992; 25(7) 327-329. [63] BEKON. New BEKON Biogas Technology Batch Process Dry Fermentation (Secured by Various Patents). BEKON energy technologies GmbH and Co. KG. 2008. [64] Liu G, Zhang R, Hamed M, El-Mashad, Withrow W, Dong R. Biogasification from kitchen and grass wastes using batch and two-phased digestion. Journal of China agricultural University 2006; 11(6) 111-115. [65] Lehtomäki A, Björnsson L. Two-stage anaerobic digestion of energy crops: Methane production, nitrogen mineralization and heavy metal mobilisation. Environ. Technol. 27 (2006) 209–218. [66] Cirne DG, Lehtomäki A, Björnsson L, Blackall LL. Hydrolysis and microbial community analyses in two-stage anaerobic digestion of energy crops. Applied microbiology 103 (2007) 516–527. [67] Six W, De Baere L. Dry anaerobic conversion of municipal solid waste by means of the DRANCO process. Water. Sci. Technol. 1992; 25(7): 295-300. [68] Murphy JD, Power N. A technical, economic, and environmental analysis of energy production from newspaper in Ireland. Waste Management 27 (2007) 177–92. [69] Thurm F, Schmid W. Renewable energy by fermentation of organic waste with the Kompogas process. In II Int. Symp. Anaerobic Dig. Solid Waste, held in Barcelona, June 15-17, 1999. (eds. J. Mata-Alvarez, A. Tilche and F. Cecchi), vol. 2, pp. 342-3345, Int. Assoc. Wat. Qual. [70] De Laclos FH, Desbois S, Saint-Joly C. Anaerobic digestion of municipal solid organic waste: Valorga full-scale plant in Tilburg, The Netherlands. In Proc. 8th Int. Conf. on Anaerobic Dig., held in Sendai, May 25-29, 1997, vol. 2, pp. 232-238, Int. Assoc. Wat. Qual. [71] De Baere L. Dry continuous anaerobic digestion of energy crops. Publications. 2007. (www.ows.se). [72] Köttner M. Biogas and fertilizer production from solid waste and biomass through dry fermentation in batch method. In: Wilderer, P. & Moletta, R. (eds), Anaerobic Digestion of Solid Wastes III. 2002. IWA Publishing, London. [73] KTBL. Gaserträge. Gasausbeuten in landwirtschaftlichen Biogasanlagen. Association for Technology and Structures in Agriculture, KTBL (ed.), 2005. Darmstadt, 10-16. [74] Linke B, Heiermann M, Grundmann P, Hertwig, F. Grundlagen, Verfahren und Potenzial der Biogasgewinnung im Land Brandenburg. Biogas in der Landwirtschaft, eds. 2003. Ministry of Agriculture, Environmental Protection and Regional Planning, 2nd edition, 10-23, Potsdam. [75] Niebaum A, Döhler H. Modellanlagen. Handreichung Biogasgewinnung und –nutzung. eds. 2004. Agency of renewable resources (FNR), Leipzig, 117-136. [76] Jäkel K. Grundlagen der Biogasproduktion. Bauen für die Landwirtschaft 2000; 3(37) 3-7. [77] Nizami AS, Korres NE, Murphy JD. A review of the integrated process for the production of grass biomethane. Environmental Science and Technology, October 12, 2009, doi:10.1021/es901533j Abdul-Sattar Nizami 26 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 3: A review of the Integrated Process for the Production of Grass Biomethane Abdul-Sattar Nizami 27 Green Grass: Developing Grass for Sustainable Gaseous Biofuel A review of the integrated process for the production of grass biomethane Abdul-Sattar Nizamia,b, Nicholas E Korresa,b , Jerry D Murphya,b,2 aDepartment of Civil and Environmental Engineering, University College Cork, Cork, Ireland. bEnvironmental Research Institute, University College Cork, Ireland. Abstract Production of grass biomethane is an integrated process which involves numerous stages with numerous permutations. The grass grown can be of numerous species, and it can involve numerous cuts. The lignocellulosic content of grass increases with maturity of grass; the first cut offers more methane potential than the later cuts. Water-soluble carbohydrates (WSC) are higher (and as such methane potential is higher) for grass cut in the afternoon as opposed to that cut in the morning. The method of ensiling has a significant effect on the dry solids content of the grass silage. Pit or clamp silage in southern Germany and Austria has a solids content of about 40%; warm dry summers allow wilting of the grass before ensiling. In temperate oceanic climates like Ireland, pit silage has a solids content of about 21% while bale silage has a solids content of 32%. Biogas production is related to mass of volatile solids rather than mass of silage; typically one ton of volatile solid produces 300 m 3 of methane. The dry solids content of the silage has a significant impact on the biodigester configuration. Silage with a high solids content would lend itself to a two-stage process; a leach bed where volatile solids are converted to a leachate high in chemical oxygen demand (COD), followed by an upflow anaerobic sludge blanket where the COD can be converted efficiently to CH4. Alternative configurations include wet continuous processes such as the ubiquitous continuously stirred tank reactor; this necessitates significant dilution of the feedstock to effect a solids content of 12%. Various pretreatment methods may be employed especially if the hydrolytic step is separated from the methanogenic step. Size reduction, thermal, and enzymatic methodologies are used. Good digester design is to seek to emulate the cow, thus rumen fluid offers great potential for hydrolysis. Keywords: grass silage, biomethane, hydrolysis 2 Corresponding author. Tel +353 21 4902286 Fax +353 21 4276648 E-mail address: jerry.murphy@ucc.ie Abdul-Sattar Nizami 28 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.1 Introduction Digestion of grass silage [1, 2] and resultant production of grass biomethane for use as a transport fuel in Ireland [3, 4] has received substantial interest in recent years. The practice of producing and using grass biomethane as a transport fuel or as a displacement/replacement for natural gas in the gas grid is at an early stage in Germany and Austria [4, 5]. In Germany, codigestion of grass silage with maize silage [6] is common in agricultural biogas plants [7]. The substantial increase in the use of grass as a cosubstrate is because of its high biodegradability and associated biogas production rates when compared to organic fraction of municipal solid waste (OFMSW) [8]. In Ireland, farmers are considering the production of grass biomethane as an occupation in lieu of beef production, due to the low farm family income associated with animal husbandry [4]. Furthermore, considering that 91% of Ireland‟s agricultural land is under grass, cross compliance rules [9] would decree that grass is the most ubiquitous bioenergy feedstock in an Irish context [3]. Ireland‟s 9% arable land is fully utilized [10], which negates the potential for a bioenergy strategy based on energy crops. This paper reviews in detail the various processes and techniques involved in the production of grass silage for digestion, the potential for pretreatment, and digestion of grass silage. 3.1.1 Characteristics of grass silage Grass silage can be utilized as a beneficial feedstock for the production of biomethane, especially in Ireland, due to its high yield (11-15 t dry solid (DS) ha-1 a-1), its perennial nature (a low energy input crop), the high volatile solid (VS) content (ca. 92%), and the associated relatively high biomethane yield [3, 4]. The chemical composition of grass may be noted in Table 3.1. An ultimate analysis of grass silage was carried out by the authors (Table 3.2). The analysis yielded the following stoichiometric equation for dry grass: C28.4H44.5O17.7N. This yields a carbon to nitrogen ratio (C:N) of 24:1. The energy content of grass silage based on the modified Dulong formula [14] is found (Box 3.1) to be 18.77 MJ kg-1 VS. Box 3.1. Energy content of grass silage Formula for silage on dry matter basis = C28.4H44.5O17.7N From modified Dulong Formula [14]; Energy Content of fuel (kJ kg-1) = 337C + 1419(H -1/8O) + 93S + 23.26 N = 337(49.93) +1419(6.52-41.49/8) +23.26(2.05) = 18,770 kJ kg-1 Energy content of silage on dry & ash free basis (VS) = 18.77 MJ kg -1 Cellulose, hemicellulose, and lignin are the three main types of polymers that constitute lignocellulosic materials like grass. The strong interlinkage of these polymers with noncovalent forces and covalent cross-linkage gives a stable shape and structure to the plant [15]. Cellulose and hemicellulose are macromolecules consisting of the same or different carbohydrate units, while lignin is an aromatic polymer made by phenyl propanoid precursors [16]. Secondary components are protein, lipids extracts, pectin, and nonstructural carbohydrates [17]. The chemical composition of grass changes as the plant matures [18]. For example, the lignin content starts increasing with plant maturity and growing season, i.e., after anthesis (Figure 3.1). Abdul-Sattar Nizami 29 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.1.2 Anaerobic Digesters The use of a two-stage digester configuration for high solid content feedstocks (such as grass silage) supports high growth rates of hydrolytic and methanogenic bacteria [20-22]. In particular the incorporation of a high rate reactor, such as an upflow anaerobic sludge blanket (UASB) downstream of the initial digester, has been studied by workers including Lehtomäki and Björnsson [23] and Yu et al. [24]. A viewpoint of Nizami and Murphy [5] is that sequencing fed leach bed reactors (SLBR) coupled with a UASB, and continuously stirred tank reactors again coupled with a UASB, offer great potential for grass biomethane production (Figure 3.2). These advantages include potentially higher organic loading rates (OLR) and methane yields at reduced digester volumes [5]. Table 3.1. Chemical composition of grass [11-13] Cellulose (%) Grass 25-40 Hemicellulose (%) Lignin (%) Ash (%) C (%) N (%) K (%) Ca (%) Mg (%) B (mg kg-1) Cu (mg kg-1) Fe (mg kg-1) Zn (mg kg-1) Na (mg kg-1) 15-50 10-30 1.5 49.93 2.05 1.51 1.91 0.26 23 13 0.29 72 915 3.1.3 Anaerobic biodegradation of grass silage The crystalline structure of cellulose is a barrier in the penetration of microbes and enzymes into the cellulosic parts [25]. Therefore, cellulose is resistant to hydrolysis while hemicellulose is weak to hydrolysis due to its amorphous structure. It is easily hydrolyzed by dilute acid or base as well as many hemicellulase enzymes [26]. Lignin is the most recalcitrant part of the structural carbohydrates because of its nonwater-soluble nature; lignin is resistant to microbial action and oxidative forces [27,28]. Therefore, the biodegradability of grass silage in a digester is limited by the crystallinity of cellulose and lignin which is enhanced by the high contents of fiber (ca. 30%) in grass silage [23]. Inefficient biodegradation results in reduced solubilization of grass silage, which limits the conversion of volatile solids to chemical oxygen demand (COD). In a system incorporating two stages with the first serving as a vessel for hydrolysis and acidogenesis and the second as a methanogenic stage, the production of COD in the first vessel is essential [5]. The protein, lipid, and extracted fractions of carbohydrates are the soluble parts of grass silage, while the fibrous components represent the structural carbohydrates of the plant, which are difficult to solubilize. Solubility of hemicellulose triggers cellulose accessibility [28] to enzymes and microbes. Considering lignin, another important Abdul-Sattar Nizami 30 Green Grass: Developing Grass for Sustainable Gaseous Biofuel constituent of structural carbohydrates, Lehtomäki et al. [1] stated that it is removed more efficiently due to solubilization than to degradation. This result is in line with the observation made by Kivaisi et al. [29]. Thus, the biodegradation and solubilisation of lignin and hemicellulose could lead to more efficient cellulose hydrolysis of grass silage. In hydrolysis, cellulose is converted into the simple sugar glucose [30]. Hydrolysis determines the range of loading and operational measures of the digester, as it holds a critical and rate limiting role in the anaerobic digestion of lignocellulosic substrate [31]. A complex set of enzymes generally known as cellulases is associated with hydrolysis of grass silage, which further enables a range species of bacteria to work on cellulosic surfaces. Therefore, hydrolysis is considered first as an enzymatic hydrolysis and then microbial hydrolysis [32]. The hydrolysis rates are determined by the attachment of hydrolytic enzymes to the surface of biodegradable material [33]. These rates are accelerated by treating the substrates chemically, physically, and/or biologically prior to feeding to the digester [34]. As a result, the structural and compositional barriers of grass silage to digestion such as the lignin seal, cellulose crystallinity, and degree of polymerization, are altered or removed. This enhances the efficiency of methane production [35]. Table 3.2. Chemical composition of grass as assessed by authors C H O N Total Atom wt 12 1 16 14 No. of atom 28.4 44.5 17.7 1 Contribution 340.8 44.5 283.2 14 682.5 % contribution 49.93 6.52 41.49 2.05 100 3.1.3 Treatment; pre/during/post Different treatment methods are used to overcome the limitation of hydrolysis of grass silage [36] based on biodegradation and solubilization of lignin and hemicellulose, with the aim of facilitating optimal methane production [37]. The system of anaerobic digestion from the production of raw material to the end-use product is represented using a modiﬁcation of circular diagrams [38] (Figure 3.3). In this methodical illustration the inter relationship between the components of inputs (agronomic factors, operational measures, and pretreatments) and of outputs (biomethane and digestate) from the digester are highlighted. Abdul-Sattar Nizami 31 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 7 Protein 3 33 Cell Contents 65 % 25 Lipid 10 Sugars 5 Cell Contents 40 % ____ 10 Minerals ____ Cell Walls 12 23 Hemicellulose 14 Cell walls 30 Cellulose 18 35 % Lignin 3 Stage of maturity Figure 3.1. Chemical composition of grass with advancing maturity [19] Abdul-Sattar Nizami 32 7 60 % Green Grass: Developing Grass for Sustainable Gaseous Biofuel (a) (b) (c) Figure 3.2. (a) Two-stage Continuously Stirred Tank Reactor; (b) Two stage two phase digester; (c) Sequencing fed Leach Bed Reactors complete with UASB (SLBR-UASB) [5] Abdul-Sattar Nizami 33 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Retention time Recir culat ion ix in g pH le Partic size ure rat M nti a typ l gra ss e e mp Te Operational measures Ino c nu ulu m ad trien / dit t ion Agricultural procedures Ensiling Output st rve Ha me ti est Harv cy en u q fre Anaerobic Digester li rti Fe In su bo lat ar ion d Post-treatment r ze Physical Chemical Biological Thermal Pre-treatment Co tion diges Po te Input Biom s Gras ane eth Pro t e rich ju in ice Figure 3.3. A circular diagram of the main factors determining grass biomethane yield of an anaerobic digester 3.2. Anaerobic reactors 3.2.1. Single Phase Digestion The anaerobic process including the conﬁguration of the digesters is a crucial element of the process [5]. Initially digesters tended to be simple, one-stage processes in which all stages of the anaerobic microbiological process were encouraged to coexist. Figure 3.2a indicates a single-phase digestion process; the process is divided into two chambers but both chambers have a complete array of anaerobic bacteria (hydrolytic, fermentative, obligate proton reducing acetogenic, acetoclastic, and hydrogenophilic methanogenic bacteria). This is based upon a facility visited by the authors in Eugendorf, Austria and documented by Smyth et al. [4]. The facility digests silage at 42% DS. The silage is diluted to about 12% DS in the ﬁrst chamber by return of leachate from the ﬁnal chamber. A loading rate of -3 -1 3 -1 1.4 kg VS m day was utilized [4]. Box 3.2 outlines the biogas production as 206 m biogas t silage. The methane 3 -1 production equates to 0.295 m CH4 kg VS added. This value is as would be expected from the scientiﬁc literature [3-5]. In an Irish context where pit silage is typically at 21% DS, the biogas production would be of the order of 100 3 -1 3 -1 m t (56 m CH4 t ) [4]. Abdul-Sattar Nizami 34 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Box 3.2. Biogas production in single phase digestion of grass (adapted from Smyth et al. [4]) 1 tonne of silage @ 42% dry solids = 420kg VS t-1 @ 92% VS = 386 kg VS t-1 60% destruction of volatiles = 232 kg VSdest t-1 grass 1 kg VS = 18.77MJ; 1 m3CH4 = 37.78MJ; Biogas @ 55.5% CH4 = 21MJ; Thus destruction of 1 kg VS = 0.89m3 biogas @ 55% CH4 Thus 60% destruction of volatiles in silage @ 42% dry solids yields: 206 m3 biogas t-1 silage 114 m3 CH4 t-1 silage CH4 production per kg VS added yields: 0.295 m3 CH4 kg-1 VS added Thus if silage from a pit at 21% Dry Solids is digested the expected yields of gas are as follows: 103 m3 biogas t-1 silage 57 m3 CH4 t-1 silage 3.2.2 Two Phase Digestion The two-stage process gained popularity, initially in the laboratory, as each bacteria stage enjoys different environmental conditions [21,23,24]. Furthermore it was realized that certain stages may be limiting for certain feedstocks. In particular digestion of feedstock with high dry solids content and more particularly high lignocellulosic content is rate limited by hydrolysis [39]. Figure 3.2b highlights a two-phase system whereby the acidogenesis stages (hydrolytic and fermentative) are separated from the methanogenic stages. Figure 3.2c highlights a system that incorporates leaching of high solids content feedstock (such as grass silage) to a leachate/wastewater high in COD. The leachate is treated in an upﬂow anaerobic -3 -1 sludge blanket (UASB) which may be subjected to very high organic loading (up to 20 kg COD m d ) while effecting a 90% destruction in COD [40]. The digested leachate is returned to the leach beds to generate more COD. This system may be termed a sequencing batch leach bed reactor complete with UASB (SLBR-UASB). In Box 3.3 the potential for COD production is estimated and the associated CH4 production is calculated. The same silage (42% DS and 92% VS) [4] is utilized in the calculation. It is shown in Box 3.2 and Box 3.3 that if 67% of volatiles are 3 -1 destroyed by the digestion process the same gas production (114 m CH4 t silage) may be achieved as by the single stage process. This is to be expected as 60% of volatiles are destroyed in the single-phase system; in the SLBR-UASB 90% of COD is converted, thus 67% of volatiles must be destroyed (60/0.9). The potential for this system is the -3 -1 -3 higher loading rate achievable in the UASB (20 kg COD m day ) versus the single-phase system (1.4 kg VS m day 1 -1 - -1 ) [4]. Allowing for 1.4 kg COD kg VS the UASB may undergo OLR 10 times higher than the single-phase process. Thus for optimal methane production within a reactor, the critical step is the hydrolysis and fermentation of the grass silage in the leach beds. This is dependent on the grass production process, the operation and environmental conditions within the digester, and pretreatments employed as discussed in the following sections. Abdul-Sattar Nizami 35 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Box 3.3. Biogas production in two phase digestion Relationship between VS and COD Based on energy basis Energy content of methane = 37.78 MJ m-3 Energy content of silage = 18.77 MJ kg-1VS From Sperling et al. [41], 1 kg COD produces = 0.35 m3 CH4 Energy content = 0.35*37.78 = 13.22 MJ kg-1 COD Relationship of VS and COD = 18.77 MJ kg-1 VS / 13.22 MJ kg-1 COD = 1.42 kg COD kg-1 VS Based on chemical reaction The basis for the COD test is that nearly all organic compounds can be fully oxidized to carbon dioxide with a strong oxidizing agent under acidic conditions. The amount of oxygen required to oxidize an organic compound to carbon dioxide, ammonia, and water is given by: CnHaObNc + (n+a/4-b/2-3c/4) O2  nCO2 + (a/2-3c/2) H2O + cNH3 For silage @ 42%DS, 92%VS: C28.4H44.5O17.7N + 29.925O2  28.4 CO2 + 20.75H2O + NH3 For 1 kg silage: 0.3864 kg VS + 0.542 kg O2  … Therefore, the relationship is 1.40 kg COD kg-1 VS Biogas potential @ 42% DS & 92% VS 1 t silage @42%DS, 92%VDS = 386.4 kgVS For 67% VSdest: 386.4*0.67 = 258 kg VSdest = 362 kg COD For 90% removal of COD: 362*0.9 = 325 kg CODremoved As 1 kg COD produces 0.35 m3 CH4; 1 t silage produces = 325*0.35 = 114 m3CH4  0.295 m3CH4 kg-1 VSadded  0.44 m3CH4 kg-1 VSdest 3.3. Agronomic factors 3.3.1 Potential grass types Depending on usage of grass either as a feed for livestock or as a feedstock for an anaerobic digestion process, the agricultural production of grass differs. This difference is due to the diversity of environmental and operational measures in growing and harvesting grass, and the microbiology of the anaerobic digester as opposed to rumen [42]. For example, the level of cellulose degradation is up to 80% in biogas plants [43] with retention times of 30-80 days, while according to Gray [44] it is 40-60% in rumen with retention of about 2 days. To make grass silage more suitable for a digester, a different harvesting frequency and time may be required. In the temperate grassland region, particularly in Ireland, the perennial ryegrass (Lolium perenne) is preferred for anaerobic digestion because of high digestibility values (D- values) [45], water-soluble carbohydrates (WSC) levels [46], and reduced quantities of crude ﬁber (Table 3.3). D-Value may be equated to the potential digestibility of the silage in cattle paunch [45,47]. In work by Mähnert et al. [2] perennial ryegrass gave the highest biogas yield (0.83- Abdul-Sattar Nizami 36 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3 -1 0.86 m kg VS added), compared to other grass species both fresh and ensiled. For example Dactylis glomerata 3 -1 (cocksfoot) gave a biogas yield of 0.65-0.72 m kg VS added. Other grasses used as a substrate for a digestion process include Phleum pratense (Timothy) and Lolium multiﬂorum (Italian ryegrass). Tetraploid ryegrass varieties are recommended due to high sugars levels [47]. In recent times diploid varieties have tended to dominate mixtures in Ireland, but tetraploid varieties remain an important component of grass seed mixtures because of their higher WSC content, their increased palatability that determines higher intake by livestock, and their tolerance to drought. However, they tend to have lower tiller densities resulting in more open swards and lower dry matter compared with diploids. Seeding rates for tetraploid grasses will need to be higher because of their larger seed size [48]. Mixtures of grasses and grass silage increase methane yield as when compared to a single grass type such as in Cynodon spp. (Bermuda grass) [20]. Additionally, Plochl and Heiermann [49] reported methane production from 3 -1 forage and paddock mixtures of 297-370 and 246 m t organic dry matter (ODM), respectively. Prochnow et al. [50] -1 reported methane yields of 370 L kg VS for ensiled mixture of grass and clover, cut before anthesis (mid-May) compared to after anthesis stage (mid-June). According to the same authors, feedstock from late cut yielded less methane when compared to the ﬁrst cut because of the increased crude ﬁber content in comparison with the ﬁrst cut. The efficiency of anaerobic digestion can be considerably improved in mixed feedstock such as that of grass with legumes because the neutral detergent ﬁber (NDF) concentration of grasses is usually greater than that of legumes, which is caused mostly by differences in NDF concentration of grass and legumes leaves [51]. Hence, increasing the proportion of legumes and consequently the leaf to stem ratio of forage results in lower cell wall concentration, in other words reduced indigestible material and increases in digestibility of feedstock. Additionally, the beneﬁts of grass-clover swards mixtures as stated by Stinner et al. [52] can be extended when the digestate from the anaerobic digestion process is relocated back to the ﬁeld resulting in higher nitrogen (N) availability to the plants due to the transformation of digestate organic N to readily available ammonia-N. Abdul-Sattar Nizami 37 Table 3.3. Comparison of fresh and ensiled grass characteristics in batch and CSTR digesters [2] Abdul-Sattar Nizami Batch digester CSTR Fresh grasses Grass silage Fresh grasses Cocksfoot Meadow foxtail Perennial ryegrass Cocksfoot Perennial ryegrass Cocksfoot Meadow foxtail Mixture 17.6 18.6 15.8 18.7 27.3 25.6 22.9 24.2 24.2 VDS (% TS) 90.1 89.1 91.1 88.5 88.8 90.6 88.8 90.6 90.0 VFA (g kg-1 FM) 0.5 0.5 0.3 6.9 14.3 0.7 0.5 0.6 0.6 pH 6.5 6.7 6.6 4.6 6.1 6.5 7.1 7.1 6.9 C:N 16.4 13.7 - 15.5 14.3 19.8 12.0 13.5 15.1 XP (% TS) 14.7 18.5 - 17.0 18.4 11.8 21.4 18.8 17.4 XF (% TS) 24.8 24.8 25.3 31.3 30.1 29.1 28.0 31.5 29.5 Saccharides (% TS) 10.8 9.8 3.3 3.4 3.1 19.3 9.8 9.1 12.7 XL (% TS) 2.1 2.3 2.2 4.9 4.6 2.4 2.6 2.1 2.4 TS (% FM) 38 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Perennial ryegrass Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.3.2 Harvest date and frequency Grass for silage is usually harvested at a less mature stage of growth (leafy and nonligniﬁed) (Table 3.4) since the aim is to obtain a crop with a relatively high content of fermentable substrate and a low content of ﬁber. The crop at this stage usually has a high leaf-stem ratio [54]; similar ﬁndings were made by Gunaseelan and Nallathambi [20] using leaves of Pennisetum purpureum (Napier grass). In a study by Amon et al. [42] on a multifaceted crop rotation to increase the yield of methane per hectare, the ﬁrst cut at vegetation stage is selected as an optimum option for harvesting. Furthermore, De Boever et al. [55] found signiﬁcant increases in structural carbohydrates (i.e., NDF and acid detergent ﬁber (ADF)) and lignin between early and late ﬁrst cut in a permanent pasture consisting of 50:50 ratio between diploid and tetraploid varieties of perennial ryegrass. The same was observed in timothy regarding NDF at two different locations within three years of experimentation [56]. Table 3.4. Analysis of grass silage characteristics [1, 23, 53] Parameters Dry matter digestibility (DMD) Measuring unit g kg-1 Acid detergent fiber (ADF) Neutral detergent fiber (NDF) Ash Crude protein g kg-1 DM g kg-1 DM g kg-1 DM g kg-1 DM pH - Lactic acid g kg-1 DM Volatile fatty acids (VFA) g kg-1 DM Ammonia-N (g/kg of total N) g kg-1 N Carbohydrates Extractives Proteins SCOD N tot Crude fiber NDF ADF Lignin (% TS) (% TS) (% TS) (mg g-1 TS) (mg g-1 TS) (% TS) (% TS) (% TS) (% TS) Range 780 (excellent leafy stage) 500 (poor late cut silage) 220-450 400-750 < 100 > 160 (leafy young tissue) 100-160 (grass for silage) < 100 (very stemmy herbage) 3.8-4.2 (wet silage) 5.0 (wilted silage) 6.0-6.3 (fresh herbage) 80-120 (well preserved) < 50 (poorly preserved) Acetic acid: 20-50 (well preserved) Butyric acid: < 10 (good quality) Propionic acid: < 10 < 100 (well fermented) 150-200 (poorly fermented) 45.0 8.4 10.4 228 16.9 20.5 28.7 28.8 5.4 Production of methane per VS depends on harvesting time and species (Trifolium spp., Lolium perenne, Phleum pratense and Trifolium spp. in mixed swards) since the results produced by various experiments were not consistent. As such Kaparaju et al. [57] found that Trifolium spp. (clover) produced 50% more methane per VS at vegetative stage than at ﬂowering stage. The results obtained were different when the same experiments were performed by Pouech et al. [58], where 32% lower methane yield per VS was recorded at vegetative stage than at ﬂowering stage. Dieterich [47] imposes signiﬁcant differences on the digestibility of grass silage which was reduced between ﬁrst cut and second Abdul-Sattar Nizami 39 Green Grass: Developing Grass for Sustainable Gaseous Biofuel cut. Additionally, in Orkney (island north of Scotland) the D-value and metabolizable energy (ME) value for early cut -1 silage (early June) were 67% and 10.7 MJ kg , while the values for late cut silage (late June) were 65% and 10.4 MJ -1 kg , respectively [47]. Prochnow et al. [50] described more biogas yield in second cut silage than ﬁrst cut. But, in spite of high DS and VS contents present in late cut grass, a lower methane yield was established [50,59]. The total solid (TS) and VS content of grass depend on several factors such as location and origin, seasonal variations, cultivation practices, type of soil, pretreatment of the biomass, and nutrient composition of the grass [60]; all of which affect the production and yield of grass biomethane [5]. The methane content can also be increased if grass is cut in the afternoon as it increases the concentration of WSC. It has been stated by White [61] that the concentration of total nonstructural carbohydrates in several grass species when investigated was lowest at 6 a.m. and increased linearly to a high at 6 p.m. Another important factor affecting the qualitative characteristics of grass silage related to harvesting management is harvest frequency. Forage quality among others is a function of high nutrient value and digestibility and is usually determined by animal performance when forages are fed to livestock [62,51]. Even more ideal pasture management would maximize both the quantity and quality of forage available to livestock [63]. The authors cannot ﬁnd any reason why the same indices cannot be used for silage evaluation as a feedstock for biomethane production particularly when experimental data regarding this important husbandry factor in relation to biomethane production are lacking. Geber [64] showed a decrease in dry matter yields and the dry organic matter with more than two harvests per year in case of Phalaris arundinacea (reed canary grass). According to Holliday [65], the 2 or 4 week cycle for cutting of grass is considered optimum for anaerobic digestion in terms of their C:N ratios; this cycle will also lead to a reduction in ﬂoat or scum formation. However, Murdoch [66] has reported that an early ﬁrst cut followed by cutting intervals of about six weeks will produce grass silage of high digestibility, whereas Motazedian and Sharrow [63] reported reductions in digestibility and crude protein in mixed pastures comprising of ryegrass and Trifolium subterraneum (subclover) as the defoliation interval increased from 7 to 49 days. Finally, intensity of harvest in case of mixed pastures [67], old or hill pastures [63], or even abandoned land can affect botanical composition, a perspective that merits further investigation since it could be used for efficient utilization of this land type considering the energy potential of the existed species for biomethane production along with biodiversity issues. In support of this argument Keating and O‟Kiely [68] reported lower WSC in old pastures in comparison with pastures dominated with Lolium multiﬂorum and L. perenne. 3.3.3 Ensiling of grass Grass and in particular grass silage form the basal diet for the vast majority of ruminants in many parts of the world during the winter feeding period [69,70]. Lee et al. [71] stated that good quality silages will support high levels of performance in animals without the need of additional concentrate supplementation. Taking under consideration the potential of grass and grass silage as a feedstock to biomethane production as a biofuel [3] and the need to increase biofuel penetration in line with the European Directive for the use of biofuels (EC/30/2003), then this necessitates the rapprochement of grass and grass silage production and their characteristics that make them suitable for both feed and biofuel. To keep a constant quality and supply of substrate to an anaerobic digester facility, ensiling of grass as silage is preferable to utilization of fresh grass. In addition, grass silage produced higher methane per tonne of ODM than Abdul-Sattar Nizami 40 Green Grass: Developing Grass for Sustainable Gaseous Biofuel fresh grass [65]. There may be potential to batch digest fresh grass in the summer months and to utilize other energy crops or biomass in the winter months. Ensiled grass in comparison to dried and stored grass ensures lower organic matter losses and independency of weather conditions that might cause damages to the dried feedstock [72]. During ensiling the resistive polysaccharides are degraded and intermediates such as volatile fatty acid (VFA) for methanogens are produced, which increase methane yield in the digester [73]. Mshandete et al. [74] reported that initial high concentrations of VFA (especially propionic acid) negatively inﬂuences the microbial population of methanogens, while at the same time VFA can also be toxic to the anaerobic process. Based on evidence from animal production research, VFA can account for up to two-thirds of the ruminant energy requirement [55], although according to Bannink [75] a signiﬁcant proportion of VFA is absorbed by cells of epithelial tissues in the rumen wall and is not used as energy source, resulting in less efficient use of the feedstock by cattle. However, this is speciﬁc to the rumen and may not be signiﬁcant to biomethane production; hence several issues still need to be addressed:  the mechanism of microbial fermentation, with conversion of feed matter into large amounts of VFA is critical for a correct description of such systems;  conservation of WSC as a higher resource of energy as compared to VFA [70] allowing increased potential for higher biomethane production;  ensiling that results in less VFA production through methods such as wilting [51,69,76] or the use of additives such as formic acid [76] merits more attention. Nevertheless, the use of additives in silage preparation did not increase methane as recorded by Neureiter et al. [77], Madhukara et al. [78], and Rani and Nand [79]. Conversely, according to Lehtomäki [25] formic acid addition resulted in higher methane production. During ensiling, nutrient losses occur due to plant respiration, fermentation, and storage. Aerobic respiration involves the action of enzymes on the carbohydrate fraction of the herbage and consequently produces water, carbon dioxide, and heat. If sufficient oxygen is present two major effects can occur. The ﬁrst is the reduction in carbohydrate supply, which restricts the quantity of lactic acid formed in the silage. Second, the heat produced raises the temperature of the silage and, if this rises above 40 °C, the digestibility of crude protein can be reduced markedly [66]. Moreover, there is a need to restrict clostridia growth that consumes lactic acid [66] and causes deterioration in silage quality [54]. Ammonia may act as a simple index of silage fermentation quality [70] since it is predominantly a product of clostridia fermentation of amino acids and its excess indicates a low quality product. Keady et al. [80] reported that improvements in silage fermentation could be indicated by decreases in pH and ammonia-N due to various factors such as use of additives, e.g., formic acid. Additionally, the formation of monosaccharides starts the hydrolysis during ensiling. Therefore, acidic conditions are suggested during the whole process of ensiling [81] to produce efficient silage for the digester. Conservation of grass can be in the form of silage in bales, pit, and/or clamp and hay. In an Irish context bale silage tends to have a solids content of 32% while pit silage has a solids content of 21%. A bale of silage has a mass of 660 kg and contains the same dry matter as one tonne of pit silage (211 kg DS) [4]. Abdul-Sattar Nizami 41 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.4. Operational procedures 3.4.1 Temperature Generally, increases in temperature result in higher levels of solubilization [13] since xylan, the major component of hemicellulose in grass, is not stable at high temperatures [28]. Therefore, short-chain fragments are formed due to increased temperature, resulting in higher biological suitability of substrate to microorganisms [8]. The literature contains contradictory evidence and variations on the beneﬁts of mesophilic or thermophilic temperature ranges. One explanation of these variations is when temperature shifts from mesophilic to thermophilic, a time period is required to ensure that a sufficient microbial population has grown, otherwise it will result in a decrease in production of biogas and methane [34]. In a study by Bouallagui et al. [82] more than 95% VS is removed to produce -1 methane at rate of 420 L kg VS added, when the ﬁrst and second stage of the digester were operated at thermophilic and mesophilic temperatures, respectively. The problem in using thermophilic temperature ranges in the digester is the high parasitic energy demand [34]. Therefore, if two-stage and two-phase digestion is used (Figure 3.2b and c) the ﬁrst stage should be operated at thermophilic temperatures and the second stage should be operated at mesophilic temperatures to accelerate the grass hydrolysis and ultimately the methane yield. 3.4.2 pH range According to Ward et al. [34], the most suitable pH for anaerobic digestion is 6.8-7.2. Below a pH level of 6.6, the methanogenic population decreases. Excessive alkalinity also results in the failure of the process by disintegration of microbial granules [83]. pH values for the ﬁrst and second stage of two-stage two-phase digesters vary according to Yu et al. [24] and Kim et al. [84]. According to Yu et al. [24], in the ﬁrst stage of a grass digester the pH should be between 4.0 and 6.5. However, Dinamarca et al. [85] and Babel et al. [86], both working on solid waste, both stated that a pH less than 7 in the hydrolysis tanks did not enhance the hydrolysis rate. In the second stage of a two-phase digester the pH should be around 7.0 [84]. Increases of pH in the leachate tank of the SLBR (Figure 3.2c) to 7.2 indicate the beginning of methanogenesis and methane production [87]. The pH alters when the total VFA concen-1 tration exceeds 4 g L and glucose is inhibited for fermentation [88]. A constant pH in two-stage digestion can also be maintained by modifying the inoculum to feed ratio [89]. 3.4.3 Mixing Gentle mixing increases digester stability against shock loading; it also increases availability of substrate to bacteria and improves methane yield [90]. Excessive mixing reduces the oxidation of fatty acids due to disruption of granulation [91], which is formed from protein and carbohydrate extracellular polymers [92]. If the substrate is mixed with support media such as carbon ﬁlter, rock wool, loofah sponge [93], pored glass material [94], or clay minerals such as sepiolite and stevensite [95] higher levels of SCOD are achieved [96]. Support media are often used in highrate digesters for wastewater/leachate treatment, such as anaerobic ﬁlters; these are often used as the second stage of a two-phase process in lieu of a UASB (Figure 3.2c). Mixing the digestate with fresh substrate improves bacterial performance [97]. There is a need to use a paddle mixer in continuously stirred tank reactors (CSTR) to overcome Abdul-Sattar Nizami 42 Green Grass: Developing Grass for Sustainable Gaseous Biofuel ﬂoating or scum formation when the dry matter of the feedstock is below 5-6% [47]. Gas mixing is one efficient option especially for reactors treating plant biomass. Gas mixing is also used as a mixing option in dry continuous digesters such as “Volarga” systems [5]. 3.4.4 Particle size Particle size can affect methane yield signiﬁcantly because of the increased availability of surface area for ﬁber degradation through hydrolyzing enzymes and bacteria [34]. According to Mshandete et al. [98], methane yield increased when the size of the particles reduced from 100 to 2 mm whereas the threshold limit of particle size, particularly for grasses, was set at 0.40 mm by Sharma et al. [99], at 1 mm by Chynoweth et al. [100], at 3 mm by Braun [101] and at 10 mm by Kaparaju et al. [57]. In case of Cynodon dactylon (Bermuda grass) the least effect was found at a size of 0.40 mm; below there was no effect on biogas production until a size of 0.088 mm was reached. Parasitic energy demand dictates against reduction in size of grass silage below 1 mm in commercial-scale plants [20]. 3.4.5 Retention time In a study of agricultural solid residues by Demirbas and Ozturk [97], 80-85% of biogas was produced in the ﬁrst 18 days of a total of 30 days digestion period. Qi et al. [13] also proposed a period of 2-3 weeks as an optimum HRT for lignocellulosic substrates. Silvey et al. [102] suggested that the batch leach bed digester should be loaded for 18-30 days, rather than 60-90 days when digesting unsorted municipal solid waste. Demirbas and Ozturk [97], state that the composition of VFA in shorter or longer digestion periods does not change signiﬁcantly. This agreement on an approximate 30-day HRT is also supported by Lehtomäki and Björnsson [23], when they obtained 85% of total methane production from grass within 30 days, using a two-stage batch digester at pilot scale. However, Chugh et al. [103] obtained 95% total methane within 45 days digesting unsorted solid waste. This is further testiﬁed in a study by Lehtomäki et al. [1], using a leach bed digester with UASB, operating at 55 days retention time using grass as a feedstock. They recorded 66% of methane potential, while 39% of carbohydrates were removed after 49 days retention time. The relationship between OLR and HRT is important. Silage may be at 40% dry solids content in Austria or Germany when wilted or 22% dry solids content in Ireland when pit silage is produced. The retention time must be decided by the mass of volatiles added rather than the mass of silage added. An operating facility visited in Austria used a loading rate of 1.4kg VS m-3 reactor day-1[4]. 3.4.6 Co-digestion Methane production is improved by codigesting grass silage with manure [104] even though the C:N ratio of grass makes it a suitable substrate as a mono-substrate [5]. Hashimoto [105] found increased microbiological stability and buffering capacity in combination with reduced nutrient deﬁciency when converting straw to biomethane with manure as a cosubstrate. Macias et al. [106] outlined the positive role of cellulase enzymes and methane bacteria present in manure on the digestion process. Additional nitrogen can be supplied by codigestion with manure [107] or urea or food wastes [97]. Codigestion of grass silage with slurries improves the digestion process through a reduction of ammonia and H2S in the digester; both of which are inhibitors to the digestion process in elevated quantities Abdul-Sattar Nizami 43 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [108]. In a study by Lehtomäki et al. [109], 53% VS removal was achieved at methane yield of 0.268 m 3 CH4 kg-1 VS added in codigestion of grass with cow manure in the ratio 2:5. Enhanced hydrolysis rates were observed by Qi et al. [13] in a study codigesting turf grass with activated sludge in batch and 2-stage semicontinuous digester conﬁguration at laboratory scale. 3.4.7 Nutrients, inoculum and inhibition Gunaseelan and Nallathambi [20] found a 40% higher methane production and reduced VFA concentration with the addition of nickel (Ni), cobalt (Co), molybdenum (Mo), selenium (Se), and sulfate. Nutrients may be added if the methanogens need some supplementary trace elements such as Ni, Co, Mo, and Se [110]. The ideal nutrient ratio for hydrolysis and acidogenesis is C:N:P:S of 500:15:5:3 and for methanogenesis is 600:15:5:3 [47]. Different additives can also increase the production rate of methane. However, they need to be evaluated based on ﬁnancial terms [34]. By supplementing grass silage where nutrient deﬁciency is evident, an increase in methane production may be observed. For example, digesting Bermuda grass increased methane production by 96% with the addition of NH4Cl [20]. In the presence of large quantities of inoculum in the second stage of a two-phase process such as a SLBR-UASB system (Figure 3.2c), the process is very stable, high methane yields are achieved, and pH adjustment is not required [20]. This statement is in line with the observation made by Lehtomäki et al. [1] digesting grass silage in a two-phase process. Ammonia, sulphide, light metal ions, and heavy metals are some of the different inhibitory substances for anaerobic digesters [111]. During digestion of energy crops, the nitrogen discharged from plant protein in the form of free ammonia and ammonium is an inhibitory substance [59]. When short-chain fatty acids are in higher concentration, they inhibit methanogens. An increase in VFA is an indicator of overloading rate [34]. 3.4.8 Recirculation of leachate/water Hydrolysis of organic matter was found to be accelerated by the recycling of leachate from the methanogenic process in an UASB to the ﬁrst reactor in a two-stage/two-phase system [112]. Initially, an increase in VFA and COD levels accompanied by a reduction in pH levels was observed. According to Sponza and Ağdağ [113], and Lai et al. [114] the optimum period of leachate recirculation for efﬁcient digestion is a recirculation frequency of 4 times per week; when the leachate volume is equal to or greater than 10% of the bed volume the frequency is 7-12 days. As Demirbas and Ozturk [97] stated the supplemental nitrogen is converted to water-soluble ammonium after its release from the digested matter, thus by recycling of leachate the need for more nitrogen is also reduced. 3.5. Pretreatments options 3.5.1 Pre-treatment According to Qi et al. [13] a combination of different pretreatment approaches with different operational procedures is required for optimum cost-effective pretreatment manipulation accompanied with minimum parasitic energy Abdul-Sattar Nizami 44 Green Grass: Developing Grass for Sustainable Gaseous Biofuel demand. Besides the pretreatment application attention should be focused to avoid any excessive formation of inhibitory byproducts such as furfural, hydroxymethyl furfural, and levulinic acid [115]. For lignocellulosic substrates, economically viable and operationally efficient pretreatment options include steam pretreatment, lime pretreatment, liquid hot water (LHW), and ammonia-based pretreatments (Table 3.5). Less optimal options include concentrated acid, wet oxidation, solvents, and metal complexes [35]. The established methane potential of grass silage without pretreatment in many studies was recorded at 0.3 m 3 kg-1 VS added [20]. This is in line with the results in Box 3.2 and Box3.3. 3.5.2 Physical pre-treatment Physical/mechanical pretreatments increase pore-size of grass silage by releasing intercellular components [116]. Particle size and available surface area in combination with pore size of the substrate in comparison to the size of the enzymes is also a limiting factor in hydrolysis [117]. With increasing pore size, the hydrolysis of hemicellulose increases, which further accelerates cellulose hydrolysis [118] and lignin degradation [119]. Drying is not favorable after pretreatment because this causes the collapse of pore structure, which reduces hydrolyzable substrates [120] in the digester. The milling effect can increase methane content from 5 to 25% [28] but higher parasitic demands make milling treatment less attractive [121]. 3.5.3 Chemical pre-treatment Chemical pretreatment increases surface accessibility for enzymes and bacteria by decreasing cellulose crystallinity [122]. In chemical pretreatment, application of NaOH, NH4OH, or a combination of both to grass ﬁber increases the potential of methane yield [8]. Acid pretreatment might be a preferential choice for grass silage because of the enhanced degradation of xylan (the major component of hemicellulose) in acidic environments [28]. Formic acid can be used as ensiling method, but it also serves as chemical pretreatment. The reactor conﬁguration is of essence in this pretreatment option. Acid pretreatment is a preferred choice only in the ﬁrst stage/phase of a two-phase process such as the leach bed in a SLBR-UASB (Figure 3.2c), due to methanogenesis in the UASB unit which can regulate any possible incoming inhibitory compounds, i.e., furfural and HMF [123]. However, lower methane yield is observed when sulphuric or nitric acids are used due to formation of H2S and N2 [28]. An increase of 100% in the yield of methane was observed due to alkaline pretreatment when wheat straw was used as a digester substrate [39]. Although, according to Pettersen [124], this could lead to the formation of denser and more thermodynamically ﬁrm cellulose than innate cellulose. Therefore, the use of alkaline pretreatment in a CSTR digester can cause formation of toxic compounds which degrade acetate and glucose 15% and 50%, respectively, through the saponiﬁcation reaction [125]. Abdul-Sattar Nizami 45 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 3.5. Effects of various pre-treatments on lignocellulosic substrates [28, 35] Accessibility of surface areas + + Cellulose crystallinity Mechanical + Steam pretreatment/steam explosion LHW (batch) + LHW (flow + through) Acid + Alkaline + Oxidative + Thermal +acid + Thermal +alkaline + (lime) Thermal+ + oxidative Thermal + +oxidative +alkaline Ammonia + + (AFEX) CO2 explosion + (+) Major effect (-) Minor effect (Blank) Not determined/yet unknown Hemicellulose solubilization Lignin solubilization Lignin structure alteration Formation of furfural/HMF + - + + + + +- - - + + - ++++- + + + + + + + - - +- + - - +- + - - + + - + 3.5.4 Thermal pre-treatment Thermal pretreatment affects the degradation of lignin [126] and hemicellulose which further increases hemicellulose hydrolysis by the acids formed from the thermal treatment [28]. Inhibitory or toxic effects may be caused to bacteria, yeast, and methanogens due to the phenolic compounds produced from the solubilization of hemicellulose and lignin at 160 °C [127]. Recondensation and precipitation on feedstock may result if soluble lignin compounds are not removed quickly [128]. The composition of the lignocellulosic substrates determines the thermal reactivity [129]. Thermal pretreatment may be divided into two categories: liquid hot water and steam pretreatment. 3.5.4.1 Liquid hot water (LHW) LHW solubilizes hemicellulose, thus enhancing the accessibility of cellulose [28]. One advantage of using LHW in a two-phase process (such as a SLBR-UASB digester) is to achieve higher concentrations of soluble carbohydrates such as xylan [128]; degradation of xylan is more effective by the use of LHW as compared to steam pretreatment [130]. The increase in enzymatic hydrolysis of the lignocellulosic material is increased 6-fold after LHW treatment. Due to higher water input in LHW, the concentration of soluble hemicellulose and lignin are reduced (due to dilution) in comparison to steam pretreatment, hence the risk of condensation and precipitation of lignin is reduced [28]. Abdul-Sattar Nizami 46 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.5.4.2 Steam pre-treatment Steam pretreatment [120] loosens cellulose ﬁber by removing large fractions of hemicellulose. This results in increased accessibility of the enzymes to cellulose [131], although as Hendriks and Zeeman [28] mentioned this method could yield lower methane volume due to condensation and precipitation of lignin over lignocellulosic substrate. 3.5.5 Biological pre-treatment Biological pretreatment offers a cost-effective solution in comparison to other pretreatments but it requires a speciﬁc environment to work efficiently [34]. As an example, treatment with cellulase enzymes [26] during silage preparation caused an increased degradation of plant cell wall constituents that were more susceptible to bacterial decomposition. Additionally, Ridla and Vehida, Sajko et al. (cited in Clavero and Razz, [26]) reported that the addition of cellulase enzymes facilitated the breakdown of a component of structural carbohydrates during ensiling which resulted in improved degradation during silage fermentation. Considering the use of inoculants it has been proposed that heterofermentative bacteria (as compared to homofermentative bacteria) could be more beneﬁcial for efficient anaerobic digestion, since they facilitate the production of intermediates for methanogens [54,132]. Additional beneﬁts include the reduction in quantity of digestate and reduced parasitic energy demands [133]. The use of the ﬁlamentous fungi, especially white-rot fungi, as a biological pretreatment has been studied recently due to its potential to degrade lignin [16]. 3.5.6 Combining pre-treatments Combined pretreatment at thermophilic temperature ranges (55 °C) [13] in the ﬁrst phase of a two-stage/phase system could prove an efficient solution in hydrolysis of grass silage feedstock. Physiochemical pretreatments can increase solubility of substrates [8]. Size reduction and preincubation with hot water have a positive effect in accelerating hydrolysis [20]. 3.6. Research required to improve grass digestion There is a wealth of research waiting to be undertaken in the area of agronomy, particularly in optimization of ensiling, grass types/mixtures, and harvest time. The list below is a few areas we feel are relatively important but not often discussed in the literature of anaerobic digestion. 3.6.1 Process control system Commercial scale anaerobic digesters are often run outside their optimum loading rates due to less effective monitoring and control mechanisms [34]. This can trigger process instability in digesters by forming inhibitory substances, such as elevated fatty acid levels which effect methanogens [134]. Therefore, online closed-loop monitoring through sensory devices, data-reading software, and automated control are beneﬁcial. Research in the Abdul-Sattar Nizami 47 Green Grass: Developing Grass for Sustainable Gaseous Biofuel application of nanotechnology in monitoring and controlling process parameters (such as pH, temperature, VFA, COD, etc.) through chips and sensors at laboratory/pilot-scale digesters will allow for new ways to overcome inhibitory effects in digesters operating at industrial/commercial scale. 3.6.2 Rumen fluid/saliva It should be borne in mind that cattle can breakdown the volatile solids in grass silage to a level of over 50% in 2 days. This may be compared with 60% destruction of volatile solids with a retention time of over 60 days in a CSTR digester in Austria [4]. Rumen ﬂuid increases the formation of fatty acids [135] which stimulates grass silage biodegradation in leach bed/hydrolyzing tanks. Rumen microorganisms are a promising source of inoculum in methanogenic stages of two-phase processes. Rumen microbes convert carbohydrates into energy-producing substances such as acetic, propionic, and butyric acids [136]. Therefore, the use of rumen ﬂuid/saliva in accelerating hydrolysis rates in leach beds and CSTR, and cow dung as inoculum in the second phase of two-phase digestion, has a signiﬁcant potential to enhance methane yield. 3.6.3 The use of fungi The dynamics of hydrolysis of grass silage in terms of microbial population and process performance is yet not fully established and understood because of the complexity of different microbes associated with grass silage and with the digestion process [87]. However, the efficient system of enzymes makes fungi a very suitable tool to achieve optimized hydrolysis of lignocellulosic substrates. Among different fungi groups, the ﬁlamentous fungi (white-rot fungi) have the ability to degrade lignin; although they are a smaller group in comparison to many other microbial groups. The oxidative nature and low substrate speciﬁcity to lignolytic enzymes make white-rot fungi efficient in biodegrading lignocellulosic materials. At commercial scale, the use of fungi can result in more cost efficient digester systems [16]. 3.6.4 Laboratory scale up-scaling to commercial/industrial scale Procedures adopted at the laboratory/pilot scale are different from procedures adopted in industrial-scale digesters in terms of both operational and process parameters. The major reason is economics [8]; potential for up-scaling is often ignored at laboratory/pilot scale research. Technologies employed at the laboratory scale must incorporate the ability to up-scale to industrial/commercial scale. Two-phase processes are quite common in the scientiﬁc literature but are less ubiquitous at commercial scale. Technical limitations exist such as: leaks in the solid phase of the twophase system due to the lack of a water seal; and leakage due to operational requirements involved in opening and closing during loading [106]. Innovation must be applied to up-scaling and commercialization of lab-scale systems. In emptying and reloading solid-phase batch chambers the chamber should be emptied of biogas by recirculation of CO2 from the gas combustion system, followed by insertion into the chamber of oxygen. This has a 3-fold effect: (1) the operatives are not overcome by noxious fumes when the chamber is opened; (2) the risk of explosion is eliminated; and (3) biogas losses are minimized. The range of operational and process measures adopted at laboratory/pilot scale should be in accordance with the range applicable and comparable to commercial-scale plants. Abdul-Sattar Nizami 48 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 3.7. Conclusion Production of biogas from grass must be seen as a holistic approach to agriculture and bioenergy production. Initially some differentiation must be observed between silage production for livestock and silage production for biodigestion. It is still important to generate a high D-value. However perennial rye grasses when cut young, before ﬂowering stage, offer optimal methane potential. The ﬁrst cut offers more methane than later cuts. For grass biomethane production an extra cut of silage is beneﬁcial as compared to silage for livestock. Laboratory work would suggest that two-phase digestion offers advantages over one-phase digestion and furthermore would suggest that hydrolysis of the grass is the rate limiting step. The importance of the integration of agronomy and biotechnology may be highlighted by noting that young grass has less lignocellulosic material than more mature grass. Biogas production of about 0.3 m3 CH4 kg-1 VS added is expected for standard digestion of grass silage. This can be improved by pretreatment. Pretreatment techniques suitable for grass silage include size reduction and thermal treatment such as liquid hot water. Codigestion with slurries can offer more stable processes, more beneﬁcial than monodigestion of either grass or slurry. Methods of improving biogas production from grass include for more sophisticated process control and utilization of rumen ﬂuids to increase hydrolysis. Good grass digester design is an attempt to emulate the paunch of cattle. Upgrading biogas to a renewable gas with 97% plus methane (termed biomethane) offers great potential as a renewable natural gas which can be distributed via the existing natural gas grid and allow countries, industries, and institutions to meet renewable energy targets particularly for transport and heat [137]. Acknowledgements Anoop Singh, Beatrice M Smyth, and Thanasit Thamsiriroj for advice, brainstorming sessions, conversations, and critiques. Funding sources: Department of Agriculture and Food (DAFF) Research Stimulus: “GreenGrass.”; Environmental Protection Agency (EPA) Strive Programme: “Compressed biomethane generated from grass used as a transport fuel.” References [1] Lehtomäki, A.; Huttunen, S.; Lehtinen, T. M.; Rintala, J. A. Anaerobic digestion of grass silage in batch leach bed processes for methane production. Bioresource technology. 2008, 99 (8), 3267-3278. [2] Mähnert, P.; Heiermann, M.; Linke, B. Batch- and semi- continuous production from different grass species. Agricultural Engineering International: The CIGRE E Journal. 2005, Manuscript EE 05 010. Vol. V11. [3] Murphy, J. D.; Power, N. M. An argument for using biomethane generated from grass as a biofuel in Ireland. Biomass and bioenergy. 2009, 33 (3), 504–512. [4] Smyth, B.; Murphy, J. D.; O‟Brien, C. What is the energy balance of grass biomethane in Ireland and other temperate northern European climates? Renew Sustain Energy Rev (2009), doi:10.1016/j.rser.2009.04.003 [5] Nizami, A. S.; Murphy, J. D. What is the Optimal Digester Configuration for Producing Grass Biomethane? Renewable Energy. 2008, Submitted December. Abdul-Sattar Nizami 49 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [6] Rösch, C.; Raab, K.; Sharka, J.; Stelzer, V. Sustainability of bioenergy production from grassland concept, indicators and results. 15th European biomass conference and exhibition, 7-11 May 2007, Berlin, Germany. [7] Weiland, P. Biomass digestion in agriculture: a successful pathway for the energy production and waste treatment in Germany. Eng. Life Sci. 2006, 6 (3). [8] Mata-Alvarez, J.; Mace, S.; Llabres, P. Anaerobic digestion of organic solid wastes. An overview of research achievements and perspectives. Bioresource Technology. 2000, 74 (1), 3–16. [9] Council regulation EC/1782/2003. Establishing common rules for direct support schemes under the common agricultural policy and establishing certain support schemes for farmers. Official Journal of European Union. 2003, I.270/1, 21.10. [10] O'Mara, F. Country Pasture/Forage Resource Profile. FAO. 2008. Agriculture Dept. Crop and Grassland Service (http://www.fao.org/ag/AGP/AGPC/doc/Counprof/Ireland/Ireland.htm). [11] Howard, R. L.; Abotsi, E.; Jansen van Rensburg, E. L.; Howard, S. Lignocellulose biotechnology: issues of bioconversion and enzyme production. Afr J Biotechnol. 2003, 2 (12), 602–619. [12] Orr R. M.; Kirk J. A. Animal physiology and nutrition. In Primrose McConnell’s The Agricultural Notebook; Soffe, R. J., 20th Eds.; Blackwell Publishing, 2003. [13] Qi, B. C.; Aldrich, C.; Lorenzen, L.; Wolfaardt, G. W. Acidogenic fermentation of lignocellulosic substrate with activated sludge. Chemical engineering communications. 2005, 192 (7-9), 1221-1242. [14] Pichtel, J. Waste management practices: Municipal, hazardous and industrial. Boca Raton-Singapore: Taylor & Francis.; 2005. [15] Pérez, J.; Muñoz-Dorado, J.; De-la-Rubia, T.; Martínez, J. Biodegradation and biological treatments of cellulose, hemicellulose and lignin: an overview. Int Microbiol . 2002, 5 (2), 53–63. [16] Sánchez, C. Lignocellulosic residues; biodegradation and bioconversion by Fungi. Biotechnology advances. 2009, 27(2), 185-194. [17] McDonald, P.; Henderson, N.; Heron, S. The Biochemistry of Silage. 2nd Eds. Chalcombe Publications: Marlow U.K., 1991. [18] Carpita, N. C. Structure and biogenesis of the cell walls of grasses. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1996, 47 (1), 445-76. [19] Holmes, W. Grass: its production and utilisation. Blackwell Scientific Publications: Oxford, U.K., 1980. [20] Gunaseelan, V.; Nallathambi, S. Anaerobic digestion of biomass for methane production: a review. Biomass and bioenergy. 1997, 13(1-2), 83-114. [21] Verrier, D.; Roy, F.; Albagnac, G. Two-phase methanization of solid vegetable waste. Biological wastes. 1987, 22 (3) 163-177. [22] Mata-Alvarez, J. A dynamic simulation of a two-phase anaerobic digestion system for solid wastes. Biotechnol. & Bioeng. 1987, 30 (7), 844-851. [23] Lehtomäki, A.; Björnsson, L. Two-stage anaerobic digestion of energy crops: Methane production, nitrogen mineralization and heavy metal mobilisation. Environmental technology. 2006, 27 (2), 209–218. [24] Yu, H. W.; Samani, Z.; Hanson, A.; Smith, G. Energy recovery from grass using two-phase anaerobic digestion. Waste management. 2002, 22 (1), 1-5. [25] Lehtomäki, A. Biogas production from energy crops and crop residues. PhD dissertation, Jyväskylä studies in biological and environmental science. University of Jyväskylä, Finland, 2006. Abdul-Sattar Nizami 50 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [26] Clavero, T.; Razz, R. Effects of biological additives on silage composition of mott dwarf elephantgrass and animal performance. Revista Cientifica. 2002, 12(4), 313-316. [27] Lewis, N. G.; Davin, L.B. The biochemical control of monolignol coupling and structure during lignan and lignin biosynthesis. In Lignin and Lignan Biosynthesis; Lewis, N. G.; Sarkanen, S. Eds.; Chem. Soc: Washington DC 1998; pp 334-361. [28] Hendriks, A. T. W. M.; Zeeman, G. Pretreatments to enhance the digestibility of lignocellulosic biomass, review. Bioresource technology. 2009, 100 (1), 10-18. [29] Kivaisi, A. K.; Op den Camp, H. J. M.; Lubberding, H. J.; Boon, J. J.; Vogels, G. D. Generation of soluble lignin derived compounds during the degradation of barley straw in an artificial rumen reactor. Appl. Microbiol. Biotechnol. 1990, 33 (1), 93–98. [30] Murphy, J. D.; McCarthy, K. Ethanol production from energy crops and wastes for use as a transport fuel in Ireland. Applied energy. 2005, 82 (2), 148-166. [31] Lynd LR, Weimer PJ, van Zyl WH, Pretorius IS. Microbial cellulose utilization: fundamentals and biotechnology. Microbiol. Mol. Biol. Rev. 2002, 66 (4), 506–577. [32] Haruta, S.; Cui, Z.; Huang, Z.; Li, M.; Ishii, M.; Igarashi, Y. Construction of a stable microbial community with high cellulose-degradation ability. Appl Microbiol Biotechnol. 2002, 59 (4-5), 529–534. [33] Veeken, A. H. M.; Hamelers, B. V. M. Effect of temperature on hydrolysis rates of selected biowaste components. Bioresource technology. 1999, 69 (3), 249±254. [34] Ward, A. J.; Hobbs, P. J.; Holliman, P. J.; Jones, D. L. Optimization of the anaerobic digestion of agricultural resources, review. Bioresource technology. 2008, 99 (17), 7928-7940. [35] Mosier, N.; Wyman, C.; Dale, B.; Elander, R.; Lee, Y. Y.; Holtzapple, M.; Ladisch, M. Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresource technology. 2005, 96 (6), 673–686. [36] Cecchi, F.; Pavan, P.; Musacco, A.; Mata-Alvarez, J.; Vallini, G. Digesting the organic fraction of municipal solid waste: Moving from mesophilic (37OC) to thermophilic (55 OC) conditions, Waste Manage. Res. 1993, 11 (5), 403– 414. [37] Petersson, A.; Thomsen, M. H.; Hauggaard-Nielsen, H.; Thomsen, A. B. Potential bioethanol and biogas production using lignocellulosic biomass from winter rye, oilseed rape and faba bean. Biomass and Bioenergy. 2007, 31 (11-12), 812–819. [38] Spedding, C. R. W. An introduction to agricultural systems; Applied Science Publishers Ltd, London., 1979. [39] Pavlostathis, S. G.; Gossett, J. M. Alkaline treatment of wheat straw for increasing anaerobic biodegradability. Biotechnol. Bioeng. 1985, 27 (3), 334–344. [40] Lepisto, S. S.; Rintala, J. A. Start-up and operation of laboratory-scale thermophilic upflow anaerobic sludge blanket reactors treating vegetable processing wastewaters. Journal of Chemical Technology and Biotechnology. 1997, 68 (3), 331-339. [41] Sperling, M.; Chernicharo, C. A. L. Biological Wastewater Treatment in Warm Climate Regions; IWA Publishing: London 2005. pp 1452. [42] Amon, T.; Amon, B.; Kryvoruchko, V.; Machmuller, A.; Hopfner-Sixt, K.; Bodiroza, V.; Hrbek, R.; Friedel, J.; Potsch, E.; Wagentristl, H.; Schreiner, M.; Zollitsch, W. Methane production through anaerobic digestion of various energy crops grown in sustainable crop rotations. Bioresource Technol. 2007, 98 (17), 3204–3212. Abdul-Sattar Nizami 51 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [43] Ress, B. B.; Calvert, P. P.; Pettigrew, C. A.; Barlaz, M. A. Testing anaerobic biodegradability of polymers in a laboratory-scale simulated landfill. Environ. Sci. Technol. 1998, 32 (6), 821–827. [44] Gray, F. V. The digestion of cellulose by sheep. J. Exp. Biol. 1947, 24 (1-2), 15–19. [45] Robson, M. J.; Parsons, A. J.; Williams, T. E. Herbage production: grasses and legumes. In Grass. Its production and utilization; Holmes, W., 2nd Eds.; Oxford: Blackwell 1989, pp 7-88. [46] Smith, K. F.; Culvenor, R. A.; Humphreys, M. O.; Simpson, R. J. Growth and carbon partitioning in perennial ryegrass (Lolium perenne) cultivars selected for high water-soluble carbohydrate concentrations. Journal of Agricultural Science. 2002, 138 (4), 375-385. [47] Dieterich, B. Energy crops for anaerobic digestion (AD) in Westray. Report written for Heat and Power Ltd., Westray, Orkney, UK. Available from: http://cngireland.com/bio- methane%20info/Burkart%20Grass%20Report%20-%20Final170308.pdf 2008 [48] Anonymous. Grass and clover. Recommended List Varieties for Ireland, Dept. of Agriculture, Fisheries and Food, 2008. [49] Plochl, M.; Heiermann, M. Biogas farming in Central and Northern Europe: A strategy for developing countries? Invited Overview Agricultural Engineering International: the CIGR Ejournal. 2006, 3 (8). [50] Prochnow, A.; Heiermann, M.; Drenckhan, A.; Schelle, H. Seasonal Pattern of Biomethanisation of Grass from Landscape Management. Agricultural Engineering International: the CIGR Ejournal. 2005, Manuscript EE 05 011. Vol. VII. [51] Buxton, D. R. Quality-related characteristics of forages as influenced by plant environment and agronomic factors. Anim. Feed Sci. Techol. 1996, 59 (1-3), 37-49. [52] Stinner, W.; Moller, K.; Leithold, G. Effects of biogas digestion of clover/grass-leys, cover crops and crop residues on nitrogen cycle and crop yield in organic stockless farming systems. European Journal of Agronomy. 2008, 29 (2-3), 125-134. [53] McEniry, J. Routine analysis of herbage samples. Grange laboratories, Teagasc, Ireland. January 2009. Available from: joseph.mceniry@teagasc.ie [54] Woolford, M. K. The silage fermentation; Marcel Dekker Inc. New York, Basel., 1984. [55] De Boever, J. L.; De Smet, A.; De Brabander, D. L.; Boucque, C. V. Evaluation of physical structure. 1. Grass silage. Journal of Dairy Science. 1993, 76 (1), 140-153. [56] Nordheim-Viken, H.; Volden, H. Effects of maturity stage, nitrogen fertilization and seasonal variation on ruminal degradation characteristics of neutral detergent fibre in timothy (Phleum pratense). Animal Feed Science & Technology. 2009, 149 (1-2), 30-59. [57] Kaparaju, P.; Luostarinen, S.; Kalmari, E.; Kalmari, J.; Rintala, J. Co-digestion of energy crops and industrial confectionery by-products with cow manure: Batch scale and farm-scale evaluation. Wat. Sci. Technol. 2002, 45 (10) 275–280. [58] Pouech, P.; Fruteau, H.; Bewa, H. Agricultural crops for biogas production on anaerobic digestion plants. In: Proceedings of the 10th European conference on biomass for energy and industry, 8.-11.6. Wurzburg, Germany; 1998. p. 163–165. [59] Lemmer, A. Kofermentation von Grüngut in landwirtschaftlichen Biogasanlagen. PhD Dissertation, University of Hohenheim, Germany. 2005. Abdul-Sattar Nizami 52 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [60] Bauer, A.; Hrbek, R.; Amon, B.; Kryvoruchko, V.; Machmüller, A.; Hopfner-Sixt, K.; Bodiroza, V.; Wagentristl, H.; Pötsch, E.; Zollitsch, W.; Amon, T. Potential of biogas production in sustainable biorefinery concepts. Proceedings of the 5th research and development conference of central- and eastern European institutes of agricultural engineering-part 2, 2, 20-31; ISBN: 966-8302-16-08. 2007. [61] White, L. M. Carbohydrate Reserves of grasses. A review. Journal of range management. 1973, 26 (1), 13-18. [62] Wheeler, J. L.; Corbett, J. L. Criteria for breeding forages of improved feeding value: Results of a Delphi survey. Grass, Forage Science. 1989, 44 (1), 77-83. [63] Motazedian, I.; Sharrow, S. H. Defoliation frequency and intensity effects on pasture for age quality. Journal of Range Management. 1990, 43 (3), 198-201. [64] Geber, U. Cutting Frequency and Stubble Height of Reed Canary Grass (Phalaris arundinacea L.): Influence on Quality and Quantity of Biomass for Biogas Production. Grass and Forage Science. 2002, 57 (4) 389-394. [65] Holliday, L. Rye-grass as an energy crop using biogas technology. DTI report no. B/CR/00801/00/00. 2005. [66] Murdoch, J. C. The conservation of grass. In Grass: its production and utilisation; Holmes, W., Eds.; Publ. for the British Grassland Society by Blackwell Scientific Publications, 1980. [67] Barthram, G. T.; Bolton, R. G.; Elston, D. A. The effects of cutting intensity and neighbour species on plants of Lolium perenne, Poa annua, Poa trivialis and Trifolium repens. Agronomie. 1999, 19 (6), 445-456. [68] Keating, T.; O‟Kiely, P. Comparison of old permanent grassland, Lolium perenne and Lolium multiflorum swards grown for silage. Irish Journal of Agricultural and Food Research. 2000, 39 (1), 25-32. [69] Bohstedt, G. The preparation and nutritive value of grass silage. Journal of Animal Science. 1940, pp 55-63. [70] Charmley, E. Towards improved silage quality-a review. Canadian Journal of Animal Science. 2001, 81, pp 157-168. [71] Lee, M. R. F.; Evans, P. R.; Nute, G. R.; Richardson, R. I.; Scollan N. D. A comparison between red clover silage and grass silage feeding on fatty acid composition, meat quality and sensory quality of the M. Longissimus muscle of dairy cull cows. Meat Science. 2009, 81 (4), 738-744. [72] Egg, R.; Coble, C.; Engler, C.; Lewis, D. Feedstock storage, handling and processing. Biomass Bioenergy. 1993, 5 (1), 71–94. [73] Madhukara, K.; Nand, K.; Raju, N. R.; Srilahta, H. R. Ensilage of mango peel for methane generation. Process biochemistry. 1993, 28 (2), 119–123. [74] Mshandete, A. M.; Björnsson, L.; Kivaisi, A. K.; Rubindamayugi, M. S. T.; Mattiasson, B. Effect of aerobic pretreatment on production of hydrolases and volatile fatty acids during anaerobic digestion of solid sisal leaf decortications residues. African Journal of Biochemistry Research. 2008, 2 (5), 111-119. [75] Bannink, A. Modelling volatile fatty acid dynamics and rumen function in lactating cows. Ph.D. Dissertation, Wageningen University, The Netherlands. 2007. [76] Dawson, L.E. R.; Ferris, C. P.; Steen, R. W. J.; Gordon, F. J.; Kilpatrick, D. J. The effects of wilting grass before ensiling on silage intake. Grass and Forage Science. 1999, 54 (3), 237-247. [77] Neureiter, M.; dos Santos, J. T. P.; Lopez, C. P.; Pichler, H.; Kirchmayr, R.; Braun, R. Effect of silage preparation on methane yields from whole crop maize silages. In: Ahring, B. K. & Hartmann, H. (eds), Proc. 4th Int. Symp. on Anaerobic Digestion of Solid Waste, Vol. 1: 109–115. BioCentrum-DTU, Copenhagen. 2005. [78] Madhukara, K.; Srilatha, H. R.; Srinath, K.; Bharathi, K.; Nand, K. Production of methane from green pea shells in floating dome digesters. Process biochemistry. 1997, 32 (6), 509–513. Abdul-Sattar Nizami 53 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [79] Rani, D. S.; Nand, K. Ensilage of pineapple processing waste for methane generation. Waste Manage. 2004, 24 (5) 523–528. [80] Keady, T. W. J.; Mayne, C. S.; Fitzpatrick, D. A. Prediction of silage feeding value from the analysis of the herbage at ensiling and effects of nitrogen fertilizer, date of harvest and additive treatment on grass silage composition. Journal of Agricultural Science. 2000, 134 (4), 353-368. [81] Mosier, N.; Hendrickson, R.; Ho, N.; Sedlak, M.; Ladisch, M. R. Optimization of pH controlled liquid hot water pretreatment of corn stover. Bioresource technology. 2005, 96 (18), 1986–1993. [82] Bouallagui, H.; Touhami, Y.; Cheikh, R. B.; Hamdi, M. Bioreactor performance in anaerobic digestion of fruit and vegetable wastes. Process Biochemistry. 2005, 40 (3-4), 989-995. [83] Sandberg, M.; Ahring, B. K. Anaerobic treatment of fish-meal process wastewater in a UASB reactor at high pH. Applied Microbiology and Biotechnology. 1992, 36 (6), 800–804. [84] Kim, J.; Park, C.; Kim, T. H.; Lee, M.; Kim, S.; Kim, S.W.; Lee, J. Effects of various pretreatments for enhanced anaerobic digestion with waste activated sludge. Journal of Bioscience and Bioengineering. 2003, 95 (3), 271–275. [85] Dinamarca, S.; Aroca, G.; Chamy, R.; Guerrero, L. The influence of pH in the hydrolytic stage of anaerobic digestion of the organic fraction of urban solid waste. Water Sci. Technol. 2003, 48 (6), 249–254. [86] Babel, S.; Fukushi, K.; Sitanrassamee, B. Effect of acid speciation on solid waste liquefaction in an anaerobic digester. Water Res. 2004, 38 (9), 2417–2423. [87] Cirne, D. G.; Lehtomäki, A.; Björnsson, L.; Blackall, L. L. Hydrolysis and microbial community analyses in twostage anaerobic digestion of energy crops. Applied microbiology. 2007, 103 (3), 516–527. [88] Siegert, I.; Banks, C. The effect of volatile fatty acid additions on the anaerobic digestion of cellulose and glucose in batch reactors. Process biochemistry. 2005, 40 (11), 3412–3418. [89] Gunaseelan, V. N. Effect of inoculum substrate ratio and pretreatments on methane yield from Parthenium. Biomass and bioenergy. 1995, 8 (1), 39–44. [90] Gomez, X.; Cuetos, M. J.; Cara, J.; Moran, A.; Garcia, A. I. Anaerobic co-digestion of primary sludge and the fruit and vegetable fraction of the municipal solid wastes – conditions for mixing and evaluation of the organic loading rate. Renewable Energy. 2006, 31 (12), 2017–2024. [91] McMahon, K. D.; Stroot, P. G.; Mackie, R. I.; Raskin, L. Anaerobic codigestion of municipal solid waste and biosolids under various mixing conditions–II: Microbial population dynamics. Water Research. 2001, 35 (7), 1817– 1827. [92] Liu, J.; Olsson, G.; Mattiasson, B. Monitoring and control of an anaerobic upflow fixed-bed reactor for highloading-rate operation and rejection of disturbances. Biotechnology and Bioengineering. 2004, 87 (1), 43–53. [93] Yang, Y.; Tada, C.; Miah, M. S.; Tsukahara, K.; Yagishita, T.; Sawayama, S. Influence of bed materials on methanogenic characteristics and immobilized microbes in anaerobic digester. Mater. Sci. Eng. 2004, 24 (3), 413– 419. [94] Show, K. Y.; Tay, J. H. Influence of support media on biomass growth and retention in anaerobic filters. Water Research. 1999, 33 (6), 1471–1481. [95] Maqueda, C.; Perez-Rodriguez, J. L.; Lebrato, J. An evaluation of clay minerals as support materials in anaerobic digesters. Environmental Technology. 1998, 19 (8), 811–819. Abdul-Sattar Nizami 54 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [96] Pinho, S. C.; Ratusznei, S. M.; Rodrigues, J. A. D.; Foresti, E.; Zaiat, M. Influence of the agitation rate on the treatment of partially soluble wastewater in anaerobic sequencing batch biofilm reactor. Water Research. 2004, 38 (19), 4117–4124. [97] Demirbas, A.; Ozturk, T. Anaerobic digestion of agricultural solid residues. Int J Gren Energy. 2004, 1(4), 483– 494. [98] Mshandete, A.; Bjornsson, L.; Kivaisi, A. K.; Rubindamayugi, M. S. T.; Mattiasson, B. Effect of particle size on biogas yield from sisal fibre waste. Renewable Energy. 2006, 31 (14), 2385–2392. [99] Sharma, S. K.; Mishra, I. M.; Sharma, M. P.; Saini, J. S. Effect of particle size on biogas generation from biomass residues. Biomass. 1988, 17 (4), 251-263. [100] Chynoweth, D. P.; Turick, C. E.; Owens, J. M.; Jerger, D. E.; Peck, M. W. Biochemical methane potential of biomass and waste feedstocks. Biomass Bioenergy. 1993, 5 (1), 95–111. [101] Braun, R. Anaerobic digestion: a multi-faceted process for energy, environmental management and rural development. In Improvement of crop plants for industrial end-uses; Ranalli, P., Eds.; Springer: Berlin 2007; pp 335-416. [102] Silvey, P.; Blackall, L.; Pullammanappallil, P. Microbial ecology of the leach-bed anaerobic digestion of unsorted municipal solid waste. In Proceedings of the Second International Symposium on Anaerobic Digestion of Solid Wastes; Mata-Alvarez, J.; Tilche, A.; Cecchi, F., Eds.; Barcelona, vol. 1. June 1999. pp 17-24. [103] Chugh, S.; Chynoweth, D. P.; Clarke, W. P.; Pullammanappallil, P.; Rudolph, V. Degradation of unsorted solid waste by a leach-bed process. Bioresource Technology. 1999, 69 (2), 103-115. [104] Romano, R. T.; Zhang, R. H. Co-digestion of onion juice and wastewater sludge using an anaerobic mixed biofilm reactor. Bioresource technology. 2008, 99 (3), 631–637. [105] Hashimoto, A. G. Conversion of straw-manure mixtures to methane at mesophilic and thermophilic temperatures. Biotechnol. Bioeng. 1983, 25 (1), 185–200. [106] Macias-Corral, M.; Samani, Z.; Hanson, A.; Smith, G.; Funk, P.; Yu, H.; Longworth, J. Anaerobic digestion of municipal solid waste and agricultural waste and the effect of co-digestion with dairy cow manure. Bioresou Technol. 2008 Nov; 99(17):8288-93. [107] Shyam, M.; Sharma, P. K. Solid-state anaerobic digestion of cattle dung and agro-residues in small capacity field digesters. Bioresource technology. 1994, 48 (3), 203-207. [108] Sterling, Jr. MC.; Lacey, R. E.; Engler, C. R.; Ricke, S. C. Effects of ammonia nitrogen on H 2 and CH4 production during anaerobic digestion of cattle manure. Bioresource technology. 2001, 77 (1), 9-18. [109] Lehtomäki, A.; Huttunen, S.; Rintala, J. Laboratory investigations on co-digestion of energy crops and crop residues with cow manure for methane production: Effect of crop to manure ratio. Resources, Conservation and Recycling. 2007, 51 (3), 591–609. [110] Weiland, P. Grundlagen der Methangärung – Biologie und Substrate. VDI-Berichte 2001, Nr.1620, pp. 19-32. [111] Chen, Y.; Cheng, J. J.; Creamer, K. S. Inhibition of anaerobic digestion process: a review. Bioresource technology. 2008, 99 (10), 4044-4064. [112] Jiang, W. Z.; Kitamura, Y.; Li, B. Improving acidogenic performance in anaerobic degradation of solid organic waste using a rotational drum fermentation system. Bioresource technology. 2005, 96 (14), 1537-1543. [113] Sponza, D. T.; Ağdağ, O. N. Impact of leachate recirculation and recirculation volume on stabilization of municipal solid wastes in simulated anaerobic bioreactors. Process biochemistry. 2004, 39 (12), 2157-2165. Abdul-Sattar Nizami 55 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [114] Lai, T. E.; Nopharatana, A.; Pullammanappallil, P. C.; Clarke, W. P. Cellulolytic activity in leachate during leach-bed anaerobic digestion of municipal solid waste. Bioresource Technology. 2001, 80 (3), 205-210. [115] Speece, R. E. In Anaerobic Digestion of Biomass; Chynoweth, D. P.; Isaacson, R., Eds.; Elsevier Applied Science 1987. pp 129–140. [116] Palmowski, L.; Muller, J. Influence of the size reduction of organic waste on their anaerobic digestion. In: II International Symposium on Anaerobic Digestion of Solid Waste. Barcelona 15–17 June, 1999. pp 137–144. [117] Thompson, D. N.; Chen, H-C.; Grethlein, H. E. Comparison of pre-treatment methods on the basis of available surface area. Bioresource technology. 1992, 39 (2), 155–163. [118] Palonen, H.; Thomsen, A. B.; Tenkanen, M.; Schmidt, A. S.; Viikari, L. Evaluation of wet oxidation pretreatment for enzymatic hydrolysis of softwood. Appl. Biochem. Biotechnol. 2004, 117 (1), 1–17. [119] Gong, C. S.; Cao, N. J.; Du, J.; Tsao, G. T. Ethanol production for renewable resources. Adv Biochem Eng Biotechnol. 1999, 65 (2), 207–41. [120] Grous, W. R.; Converse, A. O.; Grethlein, H. E. Effect of steam explosion pretreatment on pore size and enzymatic hydrolysis of poplar. Enzyme Microbiol. Technol. 1986, 8 (5), 274–280. [121] Ramos, L. P. The chemistry involved in the steam treatment of lignocellulosic materials. Quim. Nova. 2003, 26 (6), 863–871. [122] McMillan, J. D. Pretreatment of lignocellulosic biomass. In Enzymatic conversion of biomass for fuels; Himmel, M. E.; Baker, J. O.; Overend, R. P., editors.; American Chemical Society: Washington DC 1994; pp 292–324. [123] Xiao, W.; Clarkson, W. W. Acid solubilization of lignin and bioconversion of treated newsprint to methane. Biodegradation. 1997, 8 (1), 61–66. [124] Pettersen, R. C. The chemical composition of wood. In The chemistry of solid wood, Advances in Chemistry Series, vol. 207; Rowell, R. M.; Eds.; American Chemical Society: Washington DC, 1984; p 984. [125] Mouneimne, A. H.; Carrere, H.; Bernet, N.; Delgenes, J. P. Effect of saponification on the anaerobic digestion of solid fatty residues. Bioresource technology. 2003, 90 (1), 89–94. [126] Tuomela, M.; Vikman, M.; Hatakka, A.; Itävaara, M. Biodegradation of lignin in a compost environment: a review. Biores. Technol. 2000, 72 (2), 169–183. [127] Gossett, J. M.; Stuckey, D. C.; Owen, W. F.; McCarty, P. L. Heat treatment and anaerobic digestion of refuse. J. Environ. Eng. Div. 1982, 108 (3), 437–454. [128] Liu, C.; Wyman, C. E. The effect of flow rate of compressed hot water on xylan, lignin and total mass removal from corn stover. Ind. Eng. Chem. Res. 2003, 42 (21), 5409–5416. [129] Hon, D. N. S.; Shirashi, N. Wood and Cellulosic Chemistry. Marcel Dekker: New York. 1991. [130] Bobleter, O. Hydrothermal degradation of polymers derived from plants. Prog. Polym. Sci. 1994, 19 (5), 797– 841. [131] Laser, M.; Schulman, D.; Allen, S. G.; Lichwa, J.; Antal, Jr. M. J.; Lynd, L. R. A comparison of liquid hot water and steam pretreatments of sugar cane bagasse for bioconversion to ethanol. Bioresource technology. 2002, 81 (1), 33–44. [132] Idler, C.; Heckel, M.; Herrmann, C.; Heiermann, M. Influence of biological additives in grass silages on the biogas yield. Research papers of IAg Eng & LU of Ag. 2007, 39 (4), 69-82. [133] Zimbardi, F.; Viggiano, D.; Nanna, F.; Demichele, M.; Cuna, D.; Cardinale, G. Steam explosion of straw in batch and continuous systems. Appl Biochem Biotechnol. 1999, Vol. 77-79, pp 117–25. Abdul-Sattar Nizami 56 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [134] Pullammanappallil, P. C.; Svoronos, S. A.; Chynoweth, D. P.; Lyberatos, G. Expert system for control of anaerobic digesters. Biotechnology and Bioengineering. 1998, 58 (1), 13–22. [135] Hu, Z. H.; Yu, H. Q. Anaerobic digestion of cattail by rumen cultures. Waste Management. 2006, 26 (11), 1222–1228. [136] Albores, S.; Pianzzola, M. J.; Soubes, M.; Cerdeiras, M. P. Biodegradation of agroindustrial wastes by Pleurotus spp for its use as ruminant feed. Electr J Biotechnol. 2006, 9 (3), 215–20. [137] Singh, A.; Smyth, B. M.; Murphy, J. D. “A biofuel strategy for Ireland with an emphasis on production of biomethane and minimization of land take,” Renewable and Sustainable Energy Reviews, (2009), doi:10.1016/j.rser.2009.07.004 Abdul-Sattar Nizami 57 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 4: Role of Leaching and Hydrolysis in A Two Phase Grass Digestion System Abdul-Sattar Nizami 58 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Role of leaching and hydrolysis in a two phase grass digestion system A.S. Nizami1.2, T. Thamsiriroj1, 2, A Singh1,2, J.D. Murphy1, 2 1Department of Civil and Environmental Engineering, University College Cork, Ireland 2Environmental Research Institute, University College Cork, Ireland Abstract Grassland is ubiquitous in Ireland, covering over 91% of agricultural land. Grass biomethane has shown to be a sustainable biofuel with a very strong energy balance. Anaerobic digestion is a mature technology, particularly wet continuous digestion. However, the retention periods for grass digestion are relatively long, typically over 60 days. Recently, dry batch digestion has become quiet prevalent; retention times are lower, at about 30 days, but because half of the feedstock is left in the digester for a second cycle as an innoculum, the actual retention time is of the order of 45 days. A methodology that is at the development stage is a two-stage system. The first stage is a dry batch leaching stage (hydrolysis and acidogenesis). The leachate produced is treated in an upflow anaerobic sludge blanket (UASB), where methanogenisis occurs. This should allow for the shorter retention times of the dry batch process because there is no need to leave half of the feedstock in the digester as an innoculum for a second cycle. This paper concerns itself with the leaching process. How should it be carried out? What recirculation rate should be used? Should the grass silage be from a pit (ca. 20% dry solids) or from a bale (ca. 30% dry solids)? Should the grass silage be flooded or sprinkled? An experimental process was set up that allowed for four scenarios. These scenarios included sprinkling and flooding of pit silage and bale silage. The results of the analysis were used to generate a model that predicted the application of the leach beds with a UASB. The results suggested that sprinkling of bale 3 silage was the preferable option. It suggested that, with a 40 day retention time, gas production of 0.4 m of CH4/kg of volatile solids added could be achieved. This would be a similar value to a wet continuous system operating at a 60 day retention time and more efficient than a one-stage dry batch process. Keywords: grass silage, leaching, biomethane, kinetic modelling; UASB A.S. Nizami and T. Thamsiriroj are joint first authors of this paper. * Corresponding author, Ireland: Tel.: + 353 21 490 2286 E-mail: jerry.murphy@ucc.ie Abdul-Sattar Nizami 59 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.1 Introduction 4.1.1 The benefits of grass biomethane -1 -1 In the Irish context, grass is a high-yielding crop producing 11-15 tons of dry solids (DS) ha annum ; this may be -1 -1 compared to 8-10 tons of DS ha annum in central Europe [1,2]. Grassland does not require annual ploughing of soil and is a carbon sink. It covers about 91% of total agricultural land in the country. Farmers in Ireland are wellversed in all aspects of grass husbandry. Grass biomethane (renewable natural gas) has been suggested as the optimal non-residue biofuel in Ireland [3]. Singh and co-workers [4] suggested that biomethane could substitute between 7 and 33% (practical scenario to technical scenario) of natural gas in Ireland using indigenous feedstocks. The practical scenario (7% substitution) would involve the construction of 4 slaughter digesters, 4 municipal digesters, and 183 -1 rural/centralized anaerobic digesters based on slurry and grass all at a scale of about 50 kilotons annum of feedstock [4]. Grass biomethane has a minimum impact on sensitive environments in Ireland. The size of the cattle herd in Ireland is in decline, leading to an estimated present excess of 100 kha of grassland; this excess may not be converted to arable land because of cross-compliance [4]. As a result, grass biomethane will not result in reduced food production [3]. Smyth et al. [1] reported that the energy balance of grass biomethane based on a two-stage continuously stirred tank reactor (CSTR) facility is far superior to indigenous European biofuel systems, such as rapeseed biodiesel and wheat -1 -1 ethanol. The gross energy of grass biomethane (122 GJ ha annum ) is comparable to tropical biofuels, such as palm -1 -1 oil biodiesel from southeast Asia (120 GJ ha annum ) and sugar cane ethanol from Latin America (135 GJ ha -1 annum-1 [1,5,6]. The greenhouse gas savings of grass biomethane including carbon sequestration benefit is well above the 60% reduction level required to satisfy the European Union (EU) sustainability criteria for biofuels for facilities built after 2017 [7]. 4.1.2 Technologies for grass mono-digestion In Ireland, the technologies proposed for grass monodigestion are either dry batch digesters or CSTR digesters. In the dry batch digester, grass silage is introduced into the batch digester and initially water is sprinkled over the feedstock, generating leachate, which is recirculated and resprinkled over the feedstock. Gas production starts, increases, peaks, decreases, and ceases. The digester is reopened; half of the feedstock is unloaded; and half is left as an inoculum for the next batch [8]. The length of each cycle (time between loading and unloading) is about 30 days, but because half of the substrate is left as an inoculum, the actual retention time is of the order of 45 days [9]. CSTR digesters may be of a single or multi-stage type. The two-stage system involves the recycle of liquid digestate back to the first digester. The hydraulic retention time (HRT) of these systems is normally beyond 60 days, with an organic -3 -1 loading rate (OLR) of less than 3 kg of VS m day . A full-scale grass digester facility (two-stage CSTR) in Austria -3 -1 operated at a HRT of 70-80 days, with an OLR of 1.4kg of VS m day [10]. A suggested improvement on both of these systems is the combination of a dry batch digester (leaching) system with a high rate reactor, such as an upflow anaerobic sludge blanket (UASB) reactor. The benefit of the UASB is that it Abdul-Sattar Nizami 60 Green Grass: Developing Grass for Sustainable Gaseous Biofuel -3 -1 -3 -1 can be loaded to 20 kg of chemical oxygen demand (COD) m day (equivalent to 14 kg of VS m day in the case of grass silage feedstock) while effecting a 90% destruction of COD [9,11]. This is an OLR of potentially 10 times higher than for the CSTR system. The leach beds will not require half of the feedstock to be left in the batch as inoculum as required in the dry batch system. Therefore, the actual retention time can be reduced below 45 days. Using a number of batch digesters fed in sequence, a consistent level of COD can be produced and supplied to the UASB, allowing for a consistent level of biogas production. The suggested scheme for grass digestion is known as sequencing leach bed reactor complete with UASB (SLBR-UASB; Figure 4.1). Second-phase (UASB) First-phase (Sequencing Batch Leach bed) Figure 4.1. Sequencing Leach Bed Reactors reactor complete with UASB (SLBR-UASB) 4.1.3 Relevance and objective of research Because the efficiency of the UASB has been proven through many applications in full-scale anaerobic digestion facilities, the focus of this paper is on the performance of the leaching process. The result from the study is used to model the performance, including the time-related volatile solids (VS) destruction, the level and consistency of COD produced, and prediction of the achievable methane through the SLBR-UASB process. It is an objective of this work to assess the potential methane production and the retention time of the SLBR-UASB. 4.2 Hydrolysis and leaching 4.2.1 Soluble and insoluble substrates When we consider hydrolysis, the substrate may be divided into two types: soluble and insoluble. Soluble substrate includes monomers (monosaccharides, amino acids, and long-chain fatty acids) that are readily solubilized, absorbed into the cells of microorganisms, and metabolized. Insoluble substrate includes macromolecules (disaccharides, oligosaccharides, proteins, and lipids) that require enzymatic hydrolysis to break down to their constituent monomers [12]. The major substrate in lignocellulosic material is insoluble substrate in the form of celluloses and hemicelluloses. Abdul-Sattar Nizami 61 Green Grass: Developing Grass for Sustainable Gaseous Biofuel During the leaching process, soluble substrates that are hydrolyzed from insoluble substrate accumulate in the liquor. The concentration of this hydrolyzed substrate is represented by the mass of COD in the liquid. As soluble substrate increases, insoluble substrate decreases. 4.2.2 Leaching: sprinkling versus flooding Leaching involves passing of liquid (initially water) through the feedstock. This may be achieved by either sprinkling onto the feedstock or flooding. When the flooding methodology is used, the feedstock is submersed at all time and a liquid is recirculated through the submersed feedstock. In the sprinkling system, the feedstock is soaked by the liquid falling from a height onto the feedstock in a closed loop. The benefit of the flooding method is that all of the feedstock is submersed, in theory allowing for the soluble substrate to be solubilized more rapidly. It also allows the bacterial biomass and enzymes to have constant access to the feedstock and, thus, should result in a high hydrolysis rate of insoluble substrate. On the other hand, according to the first-order kinetic reaction, the hydrolysis rate depends upon neither the bacterial biomass concentration nor the enzyme concentration and only depends upon a parameter called the hydrolysis rate constant [13]. If the first-order kinetic equation is true, the hydrolysis rate constant could be dependent upon physical conditions, such as the sprinkling rate, bulk density of feedstock, distribution of liquid over feedstock, and force of liquid, which is dependent upon the vertical distance between the sprinkling head and feedstock. Potentially, the sprinkling method could be more beneficial because the liquid is sprinkled from a height, increasing the hydrolysis rate constant and, consequently, the hydrolysis rate of insoluble substrate. 4.2.3 Hydrolysis of grass As a lignocellulosic material, grass is composed of three main components: cellulose, hemicellulose, and lignin (Figure 4.2). Lignin represents the most recalcitrant part of the plant structure because of its non-water-soluble nature [14]. Cellulose is resistant to hydrolysis because of its crystallinity and the strong protection offered by hemicellulose and the lignin seal. Hemicellulose shows less resistance against hydrolysis and can be easily solubilized because of its random amorphous structure. The elements of grass that are difficult to solubilize include the fibrous components of the structural carbohydrates [11,15]. Hydrolysis is considered a rate-limiting step. A fraction of the VS in grass is not amenable to hydrolysis. According to Nizami et al. [9] for the grass silage used in this experiment, 1 kg of VS when destroyed produces 1.4 kg of COD. However, also, a fraction of the COD converted from the hydrolyzed substrate may not be amenable to degradation. Typically, an UASB can generate 90% destruction of 3 COD and produce 0.35 m of CH4 for each kg of COD destroyed [11]. Abdul-Sattar Nizami 62 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Lignin cellulose hemicellulose Treatment Cells Transverse section Plant Figure 4.2. Lignocelluloses before and after pretreatment 4.2.4 Increased hydrolysis rate by pretreatments To reduce the effect of the rate-limiting hydrolytic stage associated with inefficient solubilization of various components because of the lignin seal, cellulose crystallinity, and degree of polymerization, different pretreatment methods may be used [16,17]. The desirable effect of pretreatment on grass is indicated in Figure 4.2. Promising pretreatments for agricultural residues and herbaceous crops include steam pretreatment, lime pretreatment, liquid hot water, and ammonia-based pretreatment [15,16]. Alkali, ammonia fiber/ freeze explosion (AFEX), and liquid hot water are also reported as promising pretreatments [18-21]. Among these, the use of liquid hot water is simple. It enhances the accessibility of cellulosic surfaces to various enzymes and microbes and the solubilization of the hemicellulose. Additionally, the solubilization of soluble carbohydrates, such as xylan, is significantly increased, resulting in a high COD strength in the liquid leachate. It was observed that the use of liquid hot water could increase the hydrolysis rate of lignocellulosic material by a factor of 6 [11]. Moreover, the higher water input in the digester can reduce the risk of condensation and precipitation of lignin and hemicelluloses over the cellulosic surfaces [15]. 4.2.5 Focus of paper This paper focuses on the hydrolysis of grass in a batch leach bed reactor under aerobic conditions. Sprinkling and flooding conditions are assessed for bale and pit silages. Hot water (ca. 40 oC) is the leaching medium and may also be considered as a simple pretreatment. A kinetic model proposed by Pelillo et al. [22] is used to assess the optimal hydrolysis parameters (sprinkling or flooding and bale or pit silage). The model will also allow for the prediction of the methane production and retention time of the SLBR-UASB digester system. Abdul-Sattar Nizami 63 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.3 Materials and methods 4.3.1 Feedstock Bale and pit silages were used in the analysis (Table 4.1). Bale silage is differentiated from pit silage predominately by the solids concentration; bale silage is of the order of 30% solids, while pit silage is of the order of 20% solids content. Table 4.1. Characteristics of pit and bale silages pH Ammonia (% of total N) Protein (% DS) ME (MJ kg-1 DS) DMD or D-value (% DS or D-value) Silage intake or Palatability (g-1 kg W0.75) Lactic acid (% DS) Lactic acid (% total acids) VFA (% DS) PAL (meq kg-1 DS) NDF (% DS) Soluble sugars (% DS) FME (MJ kg-1 DS) FME/ME ratio Oil (%DS) C (% DS) H (% DS) N (% DS) Dry solids (% total) Volatile solids (% total) Pit silage 4.4 12 12.3 8.8 60 63 2.1 65 1.1 757 66 0.1 5.8 0.71 2.9 46.225 5.85 1.93 19 90 Bale silage 4.3 9 9.5 10 64 89 4.3 7.3 1.6 821 59 5 8.2 0.81 3.3 43.035 5.82 1.61 30.66 92.46 Bale silage was obtained from the Irish Agricultural Institute (Teagasc). The herbage was harvested on June 2 (first cut, early mature) from a homogeneous perennial ryegrass (Lolium perenne) dominant plot. The herbage was fieldwilted for 24 h before being baled. According to Nizami et al. [11], ensiling the grass silage through wilting results in higher conservation of water-soluble carbohydrates (WSCs) and reduced volatile fatty acids (VFA) in the silage. Higher contents of WSCs result in higher COD in the reactor; this accelerates the hydrolysis process and increases the biomethane production. The bales of herbage were wrapped in 6 layers of polythene stretch film and stored for around 5 weeks to allow for ensilage to take place. They were then repackaged as small square bales stored at the normal room temperature ready for experimental use. Pit silage in this work was taken from the pit of a farm in Cork, Ireland. It originated from a pasture dominated by ryegrass (L.perenne), which was conserved unwilted and without any additive. It had been left covered with polyethylene plastic for 3 months to ferment. The pit silage collected for the experiment was stored in a freezer at a temperature below 0 oC. It was allowed to defrost naturally before use. -1 Silage was chopped in a mobile macerator (approximated capacity of 10 kg of silage h ) to a particle size in the range of 20 mm. Abdul-Sattar Nizami 64 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.3.2 Experimental setup The experiment was carried out in a 50 L stainless-steel leach bed reactor (Figure 4.3). The silage was supported on a fine stainless-steel mesh screen to prevent feedstock throughput. A sprinkle head is installed on the bottom of the lid to distribute liquid leachate across the grass feedstock. A holding cup holds the leachate until a certain level is reached, and an automatic valve opens when the cup is full, allowing the leachate to flow to the sprinkle head. After passing through the leach bed, the leachate is collected in a 36 L stainless-steel leachate tank (Figure 4.3C). The leachate may be sampled from this tank, in particular, for COD assessment. A variable-speed pump connects the leachate tank to the holding cup. A heating panel is installed below the leachate tank and is used to control the temperature of both the leachate bed and leachate tank. 4.3.3 Experimental procedure In the study, four sub-experiments were performed as per Table 4.2: sprinkling-pit silage (S-P), sprinkling-bale silage (S-B), flooding-pit silage (F-P), and flooding-bale silage (F-B). The experiment was started by adding water to fill up the leachate tank in the case of sprinkling experiments. For the flooding experiments, additional water was added to flood the leach beds. A total of 40 L of water is added to the system for the sprinkling experiments, and a total of 75 L of water is added to the system for the flooding experiments. The added water was heated to 40 oC before experiments began. This water was entirely removed from the system upon completion of a sub-experiment, and new water was then added for the next sub-experiment. In all cases, 5 kg of macerated grass silage was placed in the leach bed. For flooding experiments, grass in the leach bed was entirely submersed in water, while in the case of sprinkling, the water level was set under the mesh screen and, therefore, well below the grass feedstock. The pump -1 started to circulate the water and distribute through the grass feedstock at a rate of 100 L day in the case of -1 sprinkling and 50 L day in the case of flooding. Liquid leachate was sampled from the leachate tank to measure the COD concentration. Each sub-experiment was continued for a 30 day period. At the end of each sub-experiment, post-leaching grass was weighted and tested for DS and VS contents. Because it is not possible to take a VS sample midway through a hydrolysis test, each sub-experiment was repeated over a shorter time frame to generate VS on intermediate days (for example, day 10 or 15). This is required for curve fitting in the kinetic modeling process discussed later (refer to Figure 4.6). Abdul-Sattar Nizami 65 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Leach Bed Temperature probe pH probe Sampling point Temperature probe Pump Sampling point Leachate tank Figure 4.3a. Scheme of the experimented reactor Figure 4.3b. Batch leach bed (left) and sprinkling head inside the batch (right) Figure 4.3c. Leachate tank (left) and leachate holding cup (right) Abdul-Sattar Nizami 66 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 4.2. Summary of experimental cases Run Leaching method Silage type Grass loaded (kg) DS Loaded (kg) VS loaded (kg) Water added (L) Theoretical COD loaded1/ (kg) Recirculat ion rate (L d-1) 1.23 Maximum theoretical COD 2/ concentration (g L-1) 28.0 S-P Sprinkling Pit 5.0 0.99 0.88 40 S-B Sprinkling Bale 5.0 1.5 1.38 40 1.93 44.41 100.0 F-P Flooding Pit 5.0 1.15 1.06 75 1.48 18.79 50.0 F-B Flooding Bale 5.0 1.5 1.38 75 1.93 24.61 50.0 100.0 Note 1/ Based on 1 kg VS = 1.4 kg COD, e.g. Run S-P: 0.88 kg VS *1.4 = 1.23 kg COD 2/ Based on full destruction of volatiles. Takes into account water added and moisture in silage, e.g. Run S-P: moisture = 4.01 kg (4.01 L); COD concentration = 1.23 kg COD *1000/(4.01+40) = 28.0 g L-1 4.3.4 Analytical methods DS and VS contents were measured using methods detailed by the American Public Health Association (APHA) [23]. The COD concentration was measured by a COD analyzer set, model HACH DRB200 and DR/2800. A complete grass silage analysis report based on their feeding value for dairy cattle was conducted by the Agri-Food and Biosciences Institute (AFBI), Belfast, U.K. C, H, and N contents in grass silage were analyzed by the Department of Chemistry, University College Cork, Cork, Ireland, using the ultimate analysis method. Scanning electron microscopy (SEM) was carried out on a FEI Inspect F system operating at 10 kV to examine the changes within the grass silage structure under flooding conditions after a 30 day leaching period. Samples were placed on a conductive carbon tape prior to imaging. 4.4 Results 4.4.1 Variation in VS and COD Figure 4.4 indicates the variation of COD and pH over time for the four experimental runs. In most cases, COD increases up to day 30, except the run F-B, where the COD peaks at day 3 and subsequently declines. Because experiment runs were not performed under anaerobic conditions, COD would be partially degraded by the bacterial biomass naturally occurring in the grass feedstock. The decrease of COD indicates the higher rate of COD degradation than the hydrolysis rate. The VS removal after 30 days for each run is shown in Table 4.3. The highest VS removal is shown in the run S-B (70.6%). It may be concluded that the sprinkling of bale silage use is the best condition for hydrolysis. The hydrolysis rate for the runs S-P and F-P are similar (60.6% VS removed as compared to 63.8%). Therefore, the sprinkling method may be more favorable than the flooding method, which requires more liquid and, hence, extra thermal energy to heat. The low percentage of VS destruction in the run F-B may be caused by the pH inhibition. It is suggested by Veeken et al. [24] that, in the pH range of 5.0-7.0, decreased pH inhibits the hydrolysis rate as a linear function. The run F-B shows an obvious difference in pH to the other runs; the pH remains below 6.0 to day 15. Abdul-Sattar Nizami 67 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 14 8 12 7.5 7 10 pH COD (g L-1) 6.5 8 Sprinkling-Pit Sprinkling-Bale Flooding-Pit Flooding-Bale 6 4 5.5 Sprinkling-Pit Sprinkling-Bale Flooding-Pit Flooding-Bale 5 2 0 6 4.5 0 5 10 15 20 25 4 30 0 5 Days 10 15 Days 20 25 30 Figure 4.4. Measured COD and pH for different experimental runs Table 4.3. Percent of dry solids and volatile solids destruction after 30 days of leaching Experimental Run %DS destroyed %VS destroyed Measured COD at day 30 (g L-1) S-P S-B F-P F-B 61.4 70.6 64.2 35.3 60.6 70.6 63.8 31.9 9.8 10.98 12.6 1.37 COD as percentage of max theoretical 35.0% 24.7% 67.1% 5.6% 4.4.2 Changes of grass structure because of leaching Patterns of the grass structure before and after leaching, as observed through the eye of SEM, are shown in Figure 4.5. The structure of grass feedstock preleaching (Figure 4.5a) shows the covering of hemicellulose and lignin seal over the crystalline structure of cellulose. This covering prevents the cellulose from exposure to external environments. The purpose of the leaching process (as per Figure 4.2) is to break through the structure by disordering the lignin and hemicellulose covering, allowing for hydrolysis to take place within the internal cellulose core. Figure 4.5b shows the surface of grass silage after leaching under flooding conditions. The hemicellulose, which ties up cellulose strands, is loosened and ruptured. The lignin and cellulose structures, which originally stretch in straight lines, are now significantly distorted. Abdul-Sattar Nizami 68 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Lignin Cellulose 500 m 300m Lignin Cellulose 50 m 100 m Hemicellulose Figure 4.5a. Grass silage structure before leaching Cellulose Lignin 100m 50m 50m 50m Figure 4.5b. Grass silage structure after hydrolysis under flooding condition Abdul-Sattar Nizami 69 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.5 Kinetic modelling 4.5.1 Characteristics of the model To generate a relationship between VS removal and observed COD, a simple kinetic model is built. The model was proposed by Pelillo et al. [22] who studied the aerobic degradation of olive mill effluents. The model comprises three different equations, which describe the rate of change of three state variables, including insoluble substrate (SI), soluble substrate (SS), and bacterial biomass (X). Insoluble substrate requires hydrolysis to convert to soluble substrate. The soluble substrate thus increases as the insoluble substrate decreases during the hydrolysis process. Hydrolysis takes place in the enzymatic reaction. Enzymes are produced by bacterial biomass, which metabolizes soluble products. Equation 1 de-of change of three state variables, including insoluble substrate scribes the reduction of insoluble substrate during the hydrolysis process. Equation 2 describes the increase of soluble substrate from hydrolysis of insoluble substrate minus the total consumption of soluble substrate by the bacterial biomass. Equation 3 describes the bacterial biomass growth; this equates to the yield of new bacterial biomass minus the death of existing biomass [22]. All parameters in these equations are defined in Table 4.4. dSI/dt = –kh (SI – SIo) X (1) dSS/dt = kh (SI – SIo) X – kc (SS – SSo) X (2) dX/dt = Y kc (SS – SSo) X – kd X (3) 4.5.2 Fitting parameters to the model The differential equation set of this model was solved using the ODE45 solver, a built-in function of MATLAB. Runge-Kutta integration techniques are applied in the solver. Parameters to fit the observed data are shown in Table 4.5, and the fitting curves for each run are shown in Figure 4.6. 4.5.2.1 Initial Conditions The initial condition in the model was input to parameters SIini, SSini,and Xini, which represent initial insoluble substrate, soluble substrate, and bacterial biomass, respectively. SSini and Xini are based on a curve fitting. SIini was estimated from the VS content in grass feedstock input, taking into account the COD equivalent of VS at -1 1.4 g of COD g of VS (previously calculated in Table 4.2). SSini is dependent upon not only the type of silage but also the leaching condition. The soluble substrate COD curve -1 of the run F-Bindicates a SSini value of 3.5 g of CODS L , while for the runs S-P, S-B, and F-P, the value is 0. This indicates the better initially solubilizing capability of bale silage under flooding than sprinkling conditions. With reference to silage properties in Table 4.1, it may be noted that bale silage has a higher percentage of soluble sugar (5%), as compared to only 0.1% in pit silage. This soluble sugar can be quickly solubilized in the flooding liquid, producing an initial COD concentration at an early stage of the leaching process. Xini is unlikely to be determined by measurement unless a known amount of bacterial biomass is initially added to the leach bed. However, this experiment aims to examine the system without additive, which can be considered as a base Abdul-Sattar Nizami 70 Green Grass: Developing Grass for Sustainable Gaseous Biofuel case for an actual application. 4.5.2.2 Fitting Parameters Parameters kh, kc, kd, Y, SIo, and SSo are estimated by the curve fitting. They show some variations in the value among different runs. Flooding cases have a greater kh than sprinkling cases. This indicates a rapid hydrolysis response under flooding conditions. Parameters kc, kd, and Y are more related to conditions of bacterial biomass during the leaching process; thus, they vary. The parameter SIo is significantly high in the run F-B. In effect, this indicates a limited conversion from insoluble to soluble substrate for bale silage under flooding conditions, although the conversion through hydrolysis occurs at a high rate. In contrast, bale silage under sprinkling conditions (run S-B) shows a value for SIo of 0. This indicates that all of the VS content can be hydrolyzed if a sufficient retention time of silage is allowed in the reactor. In the case of pit silage (runs S-P and F-P), the parameter SIo is relatively close to 0, indicating that more than 90% of the insoluble substrate can be degraded. For parameter SSo, the value is low in all cases, indicating that most of the insoluble substrate converted to soluble substrate can be degraded and, therefore, may be digested in an anaerobic reactor. Table 4.4. Parameters used in the leaching model [22] State variables SI SS X Insoluble substrate concentration (g CODS L-1) Soluble substrate concentration (g CODS L-1) Bacterial biomass concentration (g CODX L-1). kh kc kd Y SIo SSo Fitting parameters Hydrolysis rate constant (L g-1 CODX d-1) Consumption rate of soluble substrate by bacterial biomass (L g -1 CODX d-1) Bacterial biomass death rate (d-1) Bacterial biomass yield on soluble substrate (g CODX g-1 CODS) Non-biodegradable fraction in the insoluble substrate (g CODS L -1) Non-biodegradable fraction in the soluble substrate (g CODS L -1). SIini SSini Xini Initial conditions Initial insoluble substrate concentration before the leaching (g CODS L-1) Initial soluble substrate concentration before the leaching (g CODS L -1) Initial bacterial biomass concentration before the leaching (g CODX L -1) CODX CODS Note COD contributed by bacterial biomass Substrate COD (COD equivalent in case of insoluble substrate). Table 4.5. Parameters in the kinetic model S-P SIini (g CODS L-1) 28.0 SSini (g CODS L-1) 0.0 Xini (g CODX L-1) 0.50 kh (L g-1 CODX d-1) 0.085 kc (L g-1 CODX d-1) 0.09 kd (d-1) 0.05 Y (g CODX g-1 CODS) 0.10 SIo (g CODS L-1) 2.0 SSo (g CODS L-1) 1.20 Abdul-Sattar Nizami S-B 44.41 0.0 0.33 0.045 0.09 0.02 0.14 0.0 0.0 F-P 18.79 0.0 0.20 0.36 0.01 0.045 0.06 1.10 0.60 71 F-B 21.11 3.5 0.55 0.44 0.17 0.08 0.12 14.5 0.20 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.6 Application of experimental output and model 4.6.1 One- stage and two- stage digestions Hydrolysis and acidification of substrate take place in the leaching process in the first phase of a two-phase digester system. Methanogenesis occurs in the second-phase reactor. Hofenk et al. [25] studied two-phase anaerobic digestion of organic fraction of municipal solid waste (OFMSW) and concluded that one-and two-phase digesters could produce a similar biogas yield unless the hydrogen produced in the acidification process (in the first-phase reactor) can be captured and converted to methane. However, Gunaseelan [26] suggested that the two-phase digester allows for optimal growth conditions in separate chambers for hydrolytic and methanogenic bacteria, and thus, a high digester performance resulting in a high methane yield can be expected. Lin and Ouyang [27] compared a singlephase UASB and a CSTR acid-phase digester connected with an upflow methane-phase digester in sludge digestion. They reported higher VS destruction in the two-phase system than in the single-phase system. 4.6.2 Application of SLBR-UASB to grass digestion Lehtomäki al. [28] investigated the performance of the single-stage dry batch leach bed anaerobic digester in comparison to the same leach bed followed by an UASB, using grass silage as a feedstock. They found 83% of the extracted COD converted to methane in the single-stage leach bed, as opposed to 92-95% in the leach bed-UASB system. Moreover, they reported a high UASB efficiency (above 90%) corresponding to a high influent COD -1 strength; this efficiency decreased to 45-55% when the influent dropped to ca. 1 g L [28]. A similarity is found in the study by Shin et al.29 in a SLBR-UASB system for food waste digestion. They reported COD removal efficiency over 96% at a high influent COD (between 6.6 and 8.6 g L-1). The influent COD in their experiment was maintained consistently high by using a series of batch leach beds with a time lag of 2 days between feeding [29]. There is significant potential for this system (SLBR-UASB) applied to grass digestion in Ireland. A study by Smyth et al. [1] suggested a size of about 140 ha under grass to be amenable for a farm-based grass-biomethane facility [1]. Thamsiriroj and Murphy [30] using a two-stage CSTR grass digestion facility at small pilot scale found a methane 3 -1 yield from bale silage in Ireland of 455 m kg of VS added. This is equivalent to over 90% destruction of VS. -3 However, the experiment was performed under a long HRT and low OLR (HRT, 221 days; OLR, 0.5kgofVSm day - 1 ). The results of this study suggest 70.6% destruction of VS from bale silage under sprinkling conditions (S-B) within 30 days. There is significant potential for the batch leach bed to achieve a higher percent destruction of volatiles with additional retention time and, thus, to compete with the CSTR system. Abdul-Sattar Nizami 72 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Experimental Run S-P Experimental Run S-P 12 Soluble Substrate (g CODS L-1) Insoluble Substrate (g CODS L-1) 30 25 20 15 10 5 0 0 10 20 Days 30 10 8 6 4 2 0 0 40 Experimental Run S-B Soluble Substrate (g CODS L-1) Insoluble Substrate (g CODS L-1) 30 20 10 10 20 Days 30 8 6 4 2 0 0 Experimental Run F-P 20 Days 30 40 Experimental Run F-P Soluble Substrate (g CODS L-1) Insoluble Substrate (g CODS L-1) 10 14 15 10 5 10 20 Days 30 12 10 8 6 4 2 0 0 40 Experimental Run F-B 10 20 Days 30 40 Experimental Run F-B 6 Soluble Substrate (g CODS L-1) 25 Insoluble Substrate (g CODS L-1) 40 10 40 20 20 15 10 5 0 0 30 12 40 0 0 20 Days Experimental Run S-B 50 0 0 10 10 20 Days 30 40 5 4 3 2 1 0 0 10 20 Days 30 40 Figure 4.6. Variation of insoluble and soluble substrates: o Experimental data; – Simulating model Note: The data for insoluble substrate (volatile solids) is by necessity taken from more than one run of the experiment. We can not take a VS sample midway through a hydrolysis test. The data from day 0 and day 30 is from one run but the intermediate data is by necessity from other runs. The soluble substrate (COD) is from only one run of the experiment and is associated with the run from which the first and last (day 0 and day 30) insoluble substrate is taken. Abdul-Sattar Nizami 73 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 4.6.3 Achievable methane yields from SLBR-UASB The kinetic model (on the basis of fitting parameters from Table 4.5) can be further applied to examine extended retention times. Figure 4.7 shows cumulative COD in the leachate simulated in the kinetic model over a 60 day leaching period without aerobic degradation. The percentage of VS destruction is also indicated. The COD extracted would be used to produce biogas in the UASB reactor (retention time of 1 day). It is assumed that 85% of the COD obtained from the leaching step is converted to biogas in the UASB reactor. The 15% remaining includes consumption by aerobic (in leach bed) and anaerobic (in UASB) bacterial biomass and the non-biodegradable fraction of COD. It is also assumed that the hydrolysis characteristics of the leach bed in connection to UASB do not change as compared to the single leach bed as in our experiment. This assumption is on the conservative side, because normally, the effluent from the UASB reactor, which is recirculated to the leach bed, will contain a significantly lower level of COD content; this would facilitate a high leaching performance (increase in the kh value), as compared to the liquid leachate, which carries a high COD strength. From Figure 4.7, the percentage of VS destruction for sprinkling cases is higher than flooding cases. It may therefore be concluded that the sprinkling method is more efficient for grass silage hydrolysis than flooding. Sprinkling would result in a shorter retention time and a relatively higher methane yield than flooding. 4.6.4 Suggested operation of the SLBR-UASB A proposed operation of a SLBR-UASB system would involve insertion of feedstock to the leach beds on a weekly basis. For example, if a system is composed of 6 batch leach beds, the retention time of grass feedstock in each batch will be 42 days. From Figure 4.7, the COD removed (converted to CH4) at day 42, is 0.934, 1.143, 0.855, and 0.489 g -1 of COD g of VS added for the runs S-P, S-B, F-P, and F-B, respectively. Theoretically, 1 kg of COD produces 0.35 3 3 -1 m of CH4. Methane yields estimated from this basis are equal to 0.33, 0.40, 0.30, and 0.17 m kg of VS added for the runs S-P, S-B, F-P, and F-B, respectively. Bale silage under the sprinkling conditions could produce the highest 3 methane yield, followed by pit silage under the sprinkling conditions. This is comparable to the 0.455 m of CH4 kg -1 of VS added in a two-stage CSTR, with a HRT of 221 days [30]. Figure 4.8 shows the influent COD input to the UASB reactor simulated for the run S-B with 6 leach beds, 7 days between sequential feeding, and a retention time of 42 days. For example, the first batch is filled on day 1, batch 2 on day 8, etc. On day 42, digestate from the first batch is removed and batch 1 is refilled with new grass feedstock. The top curve in Figure 4.8 shows the cumulative COD as a sum of the daily COD produced from the 6 batches. Thus, when the leach beds are operated in series, a consistent COD influent can be achieved. For this particular run (S-B), -1 the average COD can be maintained at about 6 g of COD L to feed the UASB reactor. Shin et al. [29] reported -1 COD removal efficiency over 96% at a COD between 6.6 and 8.6 g L . Abdul-Sattar Nizami 74 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Run S-P Run S-P 1.4 100 90 Volatile Solids Destruction (%) Yield of Soluble Substrate (g CODS g-1 VSadded) 1.2 1 0.8 0.6 0.4 Total COD potential COD converted to CH4 0.2 0 0 10 20 30 Days 40 50 80 70 60 50 40 30 20 10 0 60 0 10 20 Run S-B Volatile Solids Destruction (%) Yield of Soluble Substrate (g CODS g-1 VSadded) 0.8 0.6 0.4 Total COD potential COD converted to CH4 0.2 0 10 20 30 Days 40 50 50 60 40 50 60 40 50 60 70 60 50 40 30 20 0 60 0 10 20 Run F-P 30 Days Run F-P 100 90 Volatile Solids Destruction (%) 1.2 Yield of Soluble Substrate (g CODS g-1 VSadded) 40 80 10 1.4 1 0.8 0.6 0.4 Total COD potential COD converted to CH 4 0.2 80 70 60 50 40 30 20 10 0 10 20 30 Days 40 50 0 60 0 10 20 Run F-B 30 Days Run F-B 1.4 100 90 Volatile Solids Destruction (%) 1.2 Yield of Soluble Substrate (g CODS g-1 VSadded) 60 90 1 Total COD potential COD converted to CH4 1 0.8 0.6 0.4 0.2 0 50 100 1.2 0 40 Run S-B 1.4 0 30 Days 80 70 60 50 40 30 20 10 0 10 20 30 Days 40 50 0 60 0 10 20 30 Days Figure 4.7. Cumulative yield of soluble substrate and corresponding %VS destruction Abdul-Sattar Nizami 75 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Run S-B 7 -1 Soluble Substrate (g CODS L ) 6 5 Daily COD produced/6 batches Daily COD produced/batch 4 3 Batch: 2 1 2 3 4 5 6 1 2 3 4 1 0 0 20 40 60 80 100 Days Figure 4.8. Influent COD to UASB reactor simulated for sprinkling bale silage (S-B) in a sequentially fed six leach bed reactor system with feeding every 7 days The result in Figure 4.8 is slightly below this range. If a higher COD level is required to feed the UASB and achieve higher COD efficiency, either the initial quantity of water should be reduced (increasing concentration) or the retention time should be reduced somewhat (reducing overall reduction in VS). Alternatively, the sprinkling rate could be increased to increase the hydrolysis rate. Caution must be exercised in combining the two systems. The UASB is typically designed on two aspects: OLR and upflow velocity. The upflow velocity should be less than 0.1 m h-1. This may clash with the sprinkling rate required for the leaching process. Two pumps may be added, but this may add to the complexity and cost of an up-scaled system. Alternatively, once a new batch of grass is filled, a high sprinkling rate may be performed for a short period (e.g., 24 h), while the UASB is disconnected to obtain an initial high COD strength. Subsequently, the UASB is connected, and the leachate is fed at a lower rate to the UASB. This will thus reduce the electrical energy demand of the pump, and a consistently high COD level can be achieved. 4.7 Conclusions In a two-phase anaerobic digestion system, the first stage must concern itself with conversion of VS to COD. The second phase is concerned with the conversion of the COD to CH4. This process examined the hydrolysis of grass silage in two forms: bale silage (at 30% DS content) and pit silage (at 19% DS content). It also compared sprinkling as a method of conversion of solids to COD to flooding. The results of an experimental process suggest the best case is sprinkling of bale silage. A kinetic model of the process was generated that allowed for simulation of a SLBRUASB reactor. The model suggested that, if 6 leach beds are fed sequentially every 7 days (cycle time of 42 days) and a sprinkling hydrolysis is effected on bale silage, then there is potential to achieve a consistent leachate with a COD -1 3 of 6 g L and generate 0.4 m of CH4 kg-1 of VS added. This offers great advantages over existing anaerobic digestion systems proposed for grass silage in terms of both retention time and methane production. The methane production 3 -1 potential is in the high range suggested for grass digestion, which is typically of the order of 0.3-0.4 m of CH4 kg of VS added [11]. Dropping the retention time from over 60 days for a wet process digesting grass to 42 days for the two-stage process described here is very significant. The 30% reduction in size must lead to a marked decrease in capital costs, leading to lower production costs and more profitable renewable energy production. Abdul-Sattar Nizami 76 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Acknowledgements Research funding was obtained from the Department of Agriculture, Fisheries, and Food (DAFF) Research Stimulas Fund Project “GreenGrass” and the Higher Education Authority Programme for Research in Third Level Institutes Cycle 4 (HEA PRTLI Cycle 4). We thank Eddie Appelbe and Erneside Engineering for the fabrication, modification, and input into reactors and Padraig O‟Kiely and Joe McInerney from Teagasc, Grange, and Richard Kearney from the Cork Institute of Technology (CIT) for the supply of grass silage. Beatrice M Smyth for advice, brainstorming sessions, conversations, and critiques. References [1] Smyth, B. M.; Murphy, J. D.; O‟Brien, C. M. What is the energy balance of grass biomethane in Ireland and other temperate northern European climates? Renewable and Sustainable Energy Reviews. 2009, 13 (9) 2349-2360. [2] Garin, P. A.; Vliegen, F.; Jossart, J. M. Energy and CO2 balance of maize and grass as energy crops for anaerobic digestion. Bioresource Technology. 2008, 99 (7) 2620-2627. [3] Murphy, J. D.; Power, N. M. An argument for using biomethane generated from grass as a biofuel in Ireland. Biomass and Bioenergy. 2009, 33 (3) 504-512. [4] Singh, A.; Smyth, B. M.; Murphy, J. D. A biofuel strategy for Ireland with an emphasis on production of biomethane and minimization of land-take. Renewable and Sustainable Energy reviews. 2010, 14 (1) 277-288. [5] Thamsiriroj, T.; Murphy, J. D. Is it better to import palm oil from Thailand to produce biodiesel in Ireland than to produce biodiesel from indigenous Irish rape seed? Applied Energy. 2009, 86 (5) 595-604. [6] Bourne, J. Green Dreams. Biofuels. National Geographic; October 2007. http://ngm.nationalgeographic.com/2007/10/biofuels/biofuels-text/ (accessed April 2010). [7] Korres, N. E.; Singh, A.; Nizami, A. S.; Murphy, J. D. Is grass biomethane a sustainable transport biofuel? Biofuels, Bioproducts and Biorefining. 2010, 4 (3) 310-325. [8] Parawira, W. Anaerobic treatment of agricultural residues and wastewater, Application of high-rate reactors. PhD thesis. Department of Biotechnology, Lund University, Sweden; 2004. [9] Nizami, A. S.; Murphy, J. D. What type of digester configurations should be employed to produce biomethane from grass silage? Renewable and Sustainable Energy Reviews. 2010, 14 (6) 1558-1568. [10] Gollackner, M. Projekt: Graskraftwerk Reitbach. Biogas aus Wiesengras – Energie ohne Ende. GRASKRAFT Reitbach reg.Gen.m.b.H. http://www.klimaaktiv.at/filemanager/download/19237/ (accessed January 2010). [11] Nizami, A. S.; Korres, N. E,; Murphy, J. D. Review of the integrated process for the production of grass biomethane. Environmental Science & Technology. 2009, 43 (22) 8496-8508. [12] Lai, T. E.; Koppar, A. K.; Pullammanappallil, P. C.; Clarke, W. P. Mathematical modeling of batch, single stage, leach bed anaerobic digestion of organic fraction of municipal solid waste. In: Optimization in the Energy Industry. Springer Berlin Heidelberg. 2009. [13] Vavilin, V. A.; Fernandez, B.; Palatsi, J.; Flotats, X. Hydrolysis kinetics in anaerobic degradation of particulate organic material: An overview. Waste Management. 2008, 28 (6) 939-951. Abdul-Sattar Nizami 77 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [14] Lewis, N. G.; Davin, L. B. The biochemical control of monolignol coupling and structure during lignan and lignin biosynthesis. In Lignin and Lignan Biosynthesis; Lewis, N. G.; Sarkanen, S. Eds.; American Chemical Society: Washington, DC, 1998; pp 334-361. [15] Hendriks, A. T. W. M.; Zeeman, G. Pretreatments to enhance the digestibility of lignocellulosic biomass, review. Bioresource Technology. 2009, 100 (1) 10-18. [16] Mosier, N.; Wyman, C.; Dale, B.; Elander, R.; Lee, Y. Y.; Holtzapple, M.; Ladisch, M. Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresource. Technology. 2005, 96 (6) 673-686. [17] Petersson, A.; Thomsen, M. H.; Hauggaard-Nielsen, H.; Thomsen, A. B. Potential bioethanol and biogas production using lignocellulosic biomass from winter rye, oilseed rape and faba bean. Biomass and Bioenergy. 2007, 31 (11-12) 812-819. [18] Kim, J. S.; Lee, Y. Y.; Park, S. C. Pretreatment of wastepaper and pulp mill sludge by aqueous ammonia and hydrogen peroxide. Applied Biochemistry and Biotechnology. 2000, 84-86 (1-9) 129-139. [19] Verga, E.; Szengyel, Z.; Réczey, K. Chemical pretreatments of corn stover for enhancing enzymatic digestibility. Applied Biochemistry and Biotechnology. 2002, 98-100 (1-9) 73-87. [20] Van Walsum, G. P.; Laser, M. S.; Lynd, L. R. Conversion of lignocellulosics pretreated with liquid hot water to ethanol. Applied Biochemistry and Biotechnology. 1996, 57-58 (1) 157-170. [21] Holtzapple, M. T.; Jun, J. H.; Ashok, G.; Patibandla, S. L.; Dale, B. E. The ammonia freeze explosion (AFEX) process: a practical lignocellulose pretreatment. Applied Biochemistry and Biotechnology. 1991, 28-29 (1) 59-74. [22] Pelillo, M.; Rincon, B.; Raposo, F.; Martin, A.; Borja, R. Mathematical modelling of the aerobic degradation of two-phase olive mill effluents in a batch reactor. Biochemical Engineering Journal. 2006, 30 (3) 308-315. [23] APHA. Standard methods for the examination of water and wastewater. 20 th ed. American Public Health Association, Washington DC. 1998. [24] Veeken, A.; Kalyuzhnyi, S.; Scharff, H.; Hamelers, B. Effect of pH and VFA on hydrolysis of organic solid waste. Journal of Environmental Engineering. 2000, 126 (12) 1076-1081. [25] Hofenk, G.; Lips, S.; Rijkens, B. A.; Voetberg, J. W. Two-phase anaerobic digestion of solid organic wastes yielding biogas and compost, EC Contract Final Report ESE-E-R-040-NL, pp.57 (1984). [26] Gunaseelan, N. V. Anaerobic digestion of biomass for methane production: A review. Biomass and Bioenergy. 1997, 13 (1-2) 83-114. [27] Lin, H. Y.; Ouyang, C. F. Upflow anaerobic sludge digestion in a phase separation system. Water Science and Technology. 1993, 28 (7) 133-138. [28] Lehtomäki, A.; Huttunen, S.; Lehtinen, T. M.; Rintala, J. A. Anaerobic digestion of grass silage in batch leach bed processes for methane production. Bioresource Technology. 2008, 99 (8) 3267-3278. [29] Shin, H. S.; Han, S. K.; Song, Y. C.; Lee, C. Y. Performance of UASB reactor treating leachate from acidogenic fermeter in the two-phase anaerobic digestion of food waste. Water Research. 2001, 35 (14) 34413447. [30] Thamsiriroj, T.; Murphy, J. D. The difficulties associated with mono-digestion of grass as exemplified by commissioning a pilot scale digester. Manuscript submitted to Energy & Fuels, February 2010. [31] Von Sperling, M.; de Lemos Chernicharo, C. A. Biological wastewater treatment in warm climate regions. IWA Publishing, London, 2005. Abdul-Sattar Nizami 78 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 5: Design, Commissioning, and Start-Up of a Sequentially Fed Leach Bed Reactor Complete with an Upflow Anaerobic Sludge Blanket Digesting Grass Silage Abdul-Sattar Nizami 79 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Design, commissioning, and start-up of a sequentially fed leach bed reactor complete with an upflow anaerobic sludge blanket digesting grass silage Abdul-Sattar Nizamia,b, Anoop Singha,b, Jerry D. Murphy,a,b* a Department of Civil and Environmental Engineering, University College Cork, Cork, Ireland b Biofuels Research Group, Environmental Research Institute, University College Cork, Cork, Ireland Abstract In a wet digestion process, it is necessary to dilute high solid content feedstocks, such as grass silage. However, grass silage tends to be a problematic feedstock for wet digestion due to its tendency to float, to wrap around moving parts, and to cause inhibition to the microbial process due to production of ammonia. Grass silage may be better suited to batch digestion. However, in a batch process, half the feedstock is left behind after each cycle to provide innoculum for the next batch of feedstock. This reduces the effective reactor volume and increases capital costs. A solution is to combine the leach beds with a high-rate reactor. The system employed in this paper is termed a SLBRUASB and is a two-phase process. The leach beds are the conduit for hydrolysis, and the methane production takes place in the UASB. The leach beds may be emptied at each cycle, reducing the size requirement of the leach beds to 67% of a pure batch system. This paper documents the problems in designing and commissioning a small pilot-scale SLBR-UASB system. The SLBR-UASB showed itself to be a reliable system when the commissioning was -1 completed. A batch test suggested the upper limit for methane production of 350 L CH 4 kg VS added. The recorded gas production when the system was operated as designed was 305 L CH 4 kg -1 VS added (87% of gas production from batch test) at a retention time of 42 days, effecting a volatile solid reduction of 68%. The first 5 days of the 7-day cycle resulted in 86% of CH4 production. Keywords: biogas; UASB; grass silage, sequencing leach bed reactor, UASB * Corresponding author. Tel +353 21 4902286 Fax +353 21 4276648 E-mail address: jerry.murphy@ucc.ie Abdul-Sattar Nizami 80 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.1 Introduction 5.1.1 Significance of digester design Design criteria for an anerobic digester includes for biological, technical, economic, and environmental criteria. Different disciplines emphasize different aspects of the design. Scientists (biologists and biochemists) are concerned with the rate, stability, and completion of the biochemical reactions. An engineer, on the other hand, is preoccupied with the retention time, the path taken by the feedstock, the separation of liquid and solid streams, and process concerns (such as wear and tear of machine parts and the maintenance of electromechanical devices). The developer is concerned about finance; he is willing to take risk with money but is risk averse to the technology. He expects stable operation with maximum energy production. The environmentalist places the emphasis on energy and greenhouse gas balance [1]. The substrate and its properties are an essential design input to the digester design. What one may consider to be the same substrate can vary greatly from country to country, from season to season, from farm to farm, and from site to site. For example, grass silage can vary in dry solids content from 20% to over 40%. The water-soluble content is maximized in an afternoon cut, the lignin content increases with the season, and the potential to float in a wet digester increases with increased particle size [2]. The digester design must also be robust and allow for expansion and flexibility. The potential to maximize profit and provide sustainable employment benefits is based upon the ability to seize upon opportunities. Can a digester designed for monodigestion of grass silage allow for codigestion with other feedstocks that bring a gate fee and a high specific methanogenic capacity? A good reactor seeks to emulate the digestive system of the bovine: efficient, stable digestion with high loading rates and short retention times. The components of a good digester could mimic the bovine and encompass an acid-phase digester (stomach), a methane-phase digester (intestine), and a process control system (brain). 5.1.2 Monodigestion of grass silage Grass, due to its specific gravity, tends to float on the digester fluid surface [3], forming sods of undigested grass that do not digest and trap any produced gas within the liquor. Grass, due to its fibrous nature, tends to wrap around moving devices [4]. Thus, grass silage as a monosubstrate causes abrasion [5] and leads to the failure of the biological process particularly in wet digestion systems. A well-designed stirring system is essential for wet monodigestion of grass silage. Prochnow et al. [5] stated that the technical and biological problems associated with using grass silage for anaerobic digestion processes are not addressed in the scientific literature. Recently, articles have appeared on this topic; Thamsiriroj and Murphy [6] addressed the difficulties in commissioning a wet two-stage digester operating with grass silage as the sole feedstock. Abdul-Sattar Nizami 81 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.1.3 The dry batch process The dry batch process is particularly suitable for grass digestion because of its ability to take a high solid content (up to 50%). However, there are not many facilities operating on grass silage [2]. The dry batch process overcomes the problems of grass flotation and the specific need for an effective mixing mechanism. Moreover, ammonia inhibition, which may be problematic in mono-digestion of grass silage in a wet process [7], is not reported in the dry batch process. Dry digestion offers simplicity in operation with little pretreatment, minimal dilution, and reduced energy inputs [2]. Feeding is also relatively simple. Half of the substrate is left behind in the digester on first unloading as an innoculum. Thus, for example, a cycle time between opening and closing the batch digester of 30 days corresponds to a retention time of 45 days. With a single batch, the gas production starts, increases, peaks, decreases, and ceases. To overcome this normal curve production of gas, a number of digesters are required, which are fed sequentially to get a relatively constant supply of biogas production [2]. 5.1.4 Sequentially fed leach bed reactor complete with upflow anaerobic sludge blanket In the history of AD technologies, one of the most notable developments is the upflow anaerobic sludge blanket (UASB) reactor [8]. There are now more than 1000 UASB units in operation all over the world [9]. The distinguishing phenomenon of the UASB over other high-rate reactors is the formation of granular sludge [10,11]. The potential -3 -1 loading rate of the UASB is of the order of 20 kg soluble chemical oxygen demand (COD) m reactor d . This is of the order of 10 times higher than a wet system [2]. By coupling the UASB to a dry batch system, the batch system is converted to a leaching/hydrolysis system providing COD for the UASB, which converts the COD to methane [12]. Thus, in operation of the batch system, it is possible to empty the whole batch reactor and thus reduce the retention time by 33% (from 45 to 30 days in the previous example). All biogas is generated in the UASB. 5.1.5 Need for process control system Commercial-scale anaerobic digesters may not achieve their optimum loading rates if the frequency of monitoring of the digestion activity is limited [13,14]. Esteves et al. [15,16] stressed the importance of measuring parameters, such as temperature, pH, liquid flow rates, and gas production, in a combined control system. The control system can allow a digester system to be loaded to the maximum of its ability, optimizing reactor size and energy production. The control system may consist of online closed-loop monitoring facilitated by sensory devices, data-reading software, and automated control, all of which is recorded, logged, and stored [2]. 5.1.6 Focus of the paper The paper deals with designing, commissioning, and the start-up of a small pilot-scale sequentially fed leach bed reactor complete with an upflow anaerobic sludge blanket (SLBR-UASB). The emphasis of this paper is on the design and commissioning of the reactor system rather than the optimal biological performance over time. This paper is in the domain of the engineering designer. The paper outlines in detail the numerous issues that led to problems in the effective operation of the designed system. The numerous modifications that were required to bring Abdul-Sattar Nizami 82 Green Grass: Developing Grass for Sustainable Gaseous Biofuel the system to operational stability are described in detail. The paper is concerned with scale-up and with engineering design; however, the loading regime required to bring to steady biological stability is also described. 5.2 Design of the digester system 5.2.1 Characteristics of the grass silage Baled grass silage prepared by the Irish Agricultural Institute “Teagasc” was used in this study (Table 5.1). The silage consists of homogeneous perennial ryegrass. The silage was the first cut of the year at an early mature stage of the grass. The herbage was field wilted for 24 h prior to baling. Polythene stretch-film was used to wrap the bales, which were initially stored for five weeks. Subsequently, small square bales were prepared at a suitable scale for experimental use. Table 5.1. Characteristics of Grass Silage in the study Lactic acid (g kg-1DS) Ethanol (g kg-1 DS) Acetic acid (g kg-1 DS) Propionic acid (g kg-1 DS) Butyric acid (g kg-1 DS) VFA (g kg-1 DS) Ammonia (g kg-1N) WSC (g kg-1 DS) pH Protein (% DS) ME (MJ kg-1 DS) DMD or D-value (% DS or D-value) Silage intake or Palatability (g kg-1 W0.75) PAL (meq kg-1 DS) NDF (% DS) FME (MJ kg-1 DS) FME/ME ratio Oil (% DS) C (% DS) H (% DS) N (% DS) DS (%) VS (%) 26.95 11.54 3.93 0.25 1.43 5.61 46.18 49.83 4.3 9.5 10 64 89 821 59 8.2 0.81 3.3 43.03 5.82 1.61 30.66 92.46 5.2.2 Design of UASB reactor The system (Figure 5.1) is similar in concept to a system tested by Lehtomäki et al. [17], who combined leach beds with a high-rate anaerobic filter at lab scale. This system differs as it is at a small pilot scale, which introduces engineering complexity. It also differs in that it employs a UASB. A UASB was chosen because it can tolerate loading -3 up to 20 kg COD m while effecting a 90% destruction of COD. Nizami et al. [14] showed that 1 kg of volatile Abdul-Sattar Nizami 83 Green Grass: Developing Grass for Sustainable Gaseous Biofuel solids (VS) destroyed generates 1.4 kg of COD and 1 kg of COD destroyed produces 350 L of CH 4. Therefore, each kilogram of VS destroyed can generate 441 L of CH 4 if the UASB effects 90% destruction of COD. The relationship between COD, biogas, and methane for this experimental process are outlined in Table 5.2. An upflow velocity of -1 less than 0.1 m h is a crucial operating parameter in UASB design [18,19]. If the upflow velocity criterion is not met, operational problems will occur. These include sludge bed flotation, foaming in the gas-liquid interface, and undergraded ingredients, which reduce the efficiencies of the UASB [20]. The other design criteria include hydraulic load, superficial biogas velocity, sludge retention time (SRT), temperature of the reactor [21], and influent COD concentration [22,23]. Table 5.3 outlines the capacity of the UASB reactor. 5a 6c 6b First-phase (Sequencing Batch Leach bed) 2c Second-phase (UASB) 3c 5d 4g 5e 5f 5b 5c 6a 2d 3a 2a 4f 4d 3d 4c 4a 3b 2b 3e 4b 4e Figure 5.1. Schematic of Sequencing Leach Bed Reactors reactor complete with UASB (SLBR-UASB) indicating views of photographs indicated in figures. (for example 2a refers to the photo in figure 2a) Abdul-Sattar Nizami 84 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 5.2. Theoretical relationship between COD, biogas and methane production (adopted from Nizami and Murphy [2], Nizami et al. [12] and Nizami et al. [14]) Grass silage DS content @ 30% VS content @92 % of DS VS conversion @ 70% Retention time Feedstock = 21 kg = 6.3 kg DS = 5.8 kg VS = 4.06 kg VS = 42 days COD production = 1.4 kg COD kg-1 VS = 4.06kg VS *1.4 kg COD kg-1 VS = 5.68 kg COD = 5.68 kg COD / 42 days = 0.135 kg COD d-1 Methane production Methane produced @ 90% COD removal efficiency of UASB = 0.35 m3CH4 kg-1 COD = 0.35 m3CH4 kg-1 COD *5.68 kg COD *0.9 = 1.79 m3 CH4 = 0.44 m3 CH4 kg-1 VS destructed = 0.31 m3 CH4 kg-1 VS added Biogas and Methane production Methane produced = 1.79 m3 CH4 / 42 days = 0.04 m3 CH4 d-1 Biogas produced @55% methane = 3.25 m3 biogas / 42 days = 77.48 L d-1 COD produced 5.2.3 Digester scheme Grass silage is loaded on a substrate holding cage within a cylindrical chamber (leach bed). Water/leachate is sprayed onto the grass silage beds and leaches organic material from the grass silage to the liquid. The leachate contains high amounts of organic matter (COD) and volatile fatty acids (VFAs). The leachate collects in a leachate tank from where it will be pumped either to a UASB reactor or back to the leach beds to increase the COD content. The COD in the leachate is converted to CH4 in the UASB reactor. Biogas is collected over the UASB; gas flow is measured by a gas flow meter. The collected biogas is further tested for CH 4 concentration. The treated leachate is recirculated back over the leach beds. Leach bed reactors are arranged in parallel so that the leachate obtained is uniform in COD concentration. 5.2.4 Digester components 5.2.4.1 Leach beds There are six leach beds (Figure 5.2a) of similar diameter (400 mm) and height (400 mm). Each leach bed consists of a substrate holding cage that is positioned on a sieve plate. The capacity of the substrate holding cage (depth of 150 -3 mm) is 16 L. On average, the cage accepts 3.5 kg of silage (220 kg m ). The total volume of each leach bed is 50 L. Sixteen liters is left as a head space to provide for effective sprinkling over the substrate; a further 16 L is also Abdul-Sattar Nizami 85 Green Grass: Developing Grass for Sustainable Gaseous Biofuel provided under each leach bed for the collection of the leachate. Table 5.3. COD capacity of UASB reactor DS content of inoculum VS content of inoculum Reactor size Designed VS concentration Quantity of Inoculum added Silage weight Water in the system DS: liquid ratio Innoculum = 6.05 % total = 80.14 %DS = 48.5 gVS L-1 innoculum = 31.4 L = 10 gVS L-1 reactor = (10 gVS L-1 reactor / 48.5 gVS L-1) * 31.4 L = 6.5 L = 315 g VS Dry solid to liquid ratio = 21 kg @ 30% dry solids = 6.3 kg DS = 40 L = 1: 6.3 Upflow velocity of the leachate and maximum pump rate = 31.4 litre = 0.75 m = 0.1 m hr-1 = 0.75m /7.5 hr Retention time of leachate = 7.5 hr Designed leachate flow rate = 4.18 L hr-1 = 100.4 L d-1 Pump speed = 5.9 L d-1 rpm-1 Pump speed required to reach 100.4 l/d = 17 rpm Volume of the UASB Height of the UASB Upflow velocity (m/hr) COD capacity of innoculum Digestion activity of the inoculum = 0.6 gCOD g-1VS d-1 COD conversion efficiency of the inoculum = 90% COD capacity for 6.5 litre innoculum = 315 g VS * 0.6 gCOD g-1 VS d-1 /0.9 = 210 gCOD d-1 COD capacity for 15 litre innoculum = (15/6.5) 210 g COD d-1 = 481.5 g COD d-1 Loading rate of COD for UASB start-up Theoretical COD produced from silage = 5.68 kg COD or 135 g COD/d (Box 1) Maximum concentration = 5.68 kg COD / 40 L = 142 g COD L-1 Potential increase in concentration per day = 135 g COD d-1 / 40L = 3.37 g COD L-1d-1 Practical COD in 1st 24 hours (operation) = 13.5 g COD L-1 = 540 g COD d-1 UASB capacity for COD (15 L innoculum) = 481.5 g COD d-1 Pump rate at low flow = 5.9 L d-1 rpm-1 Loading rate = 481.5 g COD d-1 / 13.5 g COD L-1 = 35.7 L d-1 Pump on/off every = 2 hours COD feed @ 12 rpm = 71.3L d-1 5.2.4.2 Leachate tank A common horizontal leachate tank (Figure 5.2b) with a volume of 40 L serves as the collection chamber for leachate generated in the leach beds. This leachate may either be recirculated back to the leachate holding cups or be pumped up into the UASB. A sampling point and connection valve for the leachate tank, located before the pump, facilitates analytical analysis and flow control. Abdul-Sattar Nizami 86 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.2.4.3 Water/Leachate Holding Cups A suitable starting point in this recirculation system is the highest point. The leachate is held in six water/leachate holding cups with a holding capacity of 1.5 L per cup (Figure 5.2c). These cups fill to a similar level before a valve allows the simultaneous release of the leachate over the leach beds below. 5.2.4.4 UASB Reactor The UASB reactor (Figure 5.2d) is 250 mm in diameter and 750 mm in height. The total volume of the UASB reactor is 37 L with a working volume of 31.4 L. The sampling points and connecting valves at both ends of the reactor allow analysis of experimental data and control of the flow of leachate. A wet gas flow meter is connected to the top of the UASB to measure the flow of biogas. b a d c Figure 5.2. Digester components (a) Leach beds, (b) Leachate tank, (c) Water/leachate holding cups (d) UASB reactor 5.2.5 System Operation 5.2.5.1 Feeding Substrate The substrate/feedstock is placed in the holding cage fitted inside each leach bed (Figure 5.3a). The leach bed is closed with a sealed lid. Water is sprinkled from the leachate holding cups through the spray nozzle onto the substrate in the leach bed. Abdul-Sattar Nizami 87 Green Grass: Developing Grass for Sustainable Gaseous Biofuel a b c d e Figure 5.3. Operation of system (a) cage of the leach bed (b) Low rate pump (c) Leachate discharge nozzle (d) Sieve plate (e) Filters for leachate 5.2.5.2 Variable-Speed Pump A pump with variable speed is connected to recirculate the leachate around the system, from the leachate tank to the UASB, or to the leachate holding cups and back to the leach bed and leachate tank (Figure 5.3b). High sprinkling rates ensure good hydrolysis and efficient conversion of volatile solids to COD. 5.2.5.3. Sprinkling System The sprinkling system consists of the aforementioned pump (Figure 5.3b), six leachate holding cups (Figure 5.2c), pipes, and leachate discharge nozzles (Figure 5.3c). Each leachate holding cup is connected at the bottom level to the next leachate holding cup to ensure the filling regime results in a similar level of leachate in each cup. The valve that releases the leachate from the cups is set to function when a sufficient quantity of leachate is stored. Small holes in the leachate discharge nozzle together with the quick discharge of the leachate from the full holding cups ensure a maximum spread of leachate over the substrate. The six cups are connected with a pipe at the top of the cups to discharge the leachate in case of overflow. This pipe also serves to prevent excessive suction pressures and allows the leachate to be sprinkled over the substrate (Figure 5.2c). 5.2.5.4. Percolation System A sieving system is in place within the bioreactor to ensure that particulate matter does not clog moving parts. Different filters and sieve plates are positioned for this purpose. Within the leach beds, coarse particles are filtered by 4 mm diameter holes at the center of the holding cage (Figure 5.3a). The sieve plate (Figure 5.3d) beneath the leach bed has 1.5 mm diameter holes. The leachate is further filtered for fine particles through a 1 mm diameter size filter, when the leachate is conveyed from the leachate tank to the UASB or directly to the leach beds (Figure 5.3e). Abdul-Sattar Nizami 88 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.2.6 Process Control 5.2.6.1 Digital Control System A digital control system consisting of a programmable logic controller (PLC) and a supervisory control and data acquisition system (SCADA) is designed to monitor and control the various process parameters of the bioreactor, including temperature, pH, flow of circulating leachate, pump rates, leachate holding and discharge frequency, and the flow and quantity of biogas. 5.2.6.2 Temperature A temperature control and command system is provided to each main component of the reactor, such as the leach bed, the leachate tank, and the UASB reactor. Each leach bed is connected with a temperature sensor probe (Figure 5.4a). To attain the required temperature in the leachate tank, two folding coils are employed: one placed beneath the leachate tank and the other wrapped around the tank (Figure 5.4b). The UASB reactor is connected with embedded heating elements (PT100 Elements Omega) (Figure 5.4c) and temperature sensor probes (PT100 Probes Omega) at two different locations (Figure 5.4d). 5.2.6.3 pH pH is a crucial process parameter in anaerobic digestion; it indicates the level of acidity or alkalinity. pH sensor probes (Signet 2754-2757 pH DryLoc pH/ORP electrodes) are positioned in the leachate tank (Figure 5.4e) and in the UASB (Figure 5.4f). 5.2.6.4 Biogas Biogas is the ultimate outcome of the anaerobic digestion process. To monitor and measure the production of biogas, a gas flow meter (FMA-1600A Omega) was attached on the head of the UASB reactor (Figure 5.4g). To measure the composition of the gas (mainly CH4 and CO2), a separate sample point was provided in the outlet pipe for the biogas. Abdul-Sattar Nizami 89 Green Grass: Developing Grass for Sustainable Gaseous Biofuel a b d e c f g Figure 5.4. Process control mechanism (a) Temperature sensor for leach bed (b) Heating coils at leachate tank (c) Heating coils for UASB (d) Temperature sensor for UASB reactor (e) pH sensor of leachate tank (f) pH sensor of UASB (g) Wet gas flow meter 5.3 Methodology and initial start-up 5.3.1 Analytical Methods DS and VS contents were measured using methods detailed in APHA [24]. For every new batch of silage inserted into the leach bed, three samples were taken. These samples were measured for DS and VS; the average value was used for analysis. Similarly, when the batch was removed after 42 days, three samples were taken, the DS and VS were measured, and the average value was used for analysis. The COD concentration was measured by a COD analyzer set, model HACH DRB200 and DR/2800, USA. A complete grass silage analysis report based on their feeding value for dairy cattle was conducted by Agri-Food and Biosciences Institute (AFBI), Belfast, U.K. and the Irish Agricultural Institute “Teagasc”. C, H, and N contents in the grass silage were analyzed by the Department of Chemistry, University College Cork, Cork, Ireland, using the ultimate analysis method. A portable biogas analyzer PGD3-IR (Scientific Controls Ltd.) was used to measure the composition of the biogas. The Hohenheim biogas yield test (HBT) [25] is used to determine the potential methane yield of the grass silage. 5.3.2 Set-up of small batch experiments As a countermeasure and check of the process, an expected upper limit on methane production is required. The achievable methane yield of grass silage was determined in discontinuous digesters using the Hohenheim biogas yield test (HBT) [25]. In this technique, a 100 mL glass syringe is used as a flask sampler; 500 mg of grass silage is used as a test substrate. The test samples are dried at 60 0C over 48 h and ground to a size of less than 1 mm. The test substrates are mixed with 30 mL of inoculum and digested at 37 0C in three replicates. In total, six samples are prepared and tested to measure the methane potential. Biogas produced was recorded periodically for its volume and Abdul-Sattar Nizami 90 Green Grass: Developing Grass for Sustainable Gaseous Biofuel methane content. A separate inoculum was also digested with three replicates to get zero variant. 5.3.3 Seeding the UASB reactor Fresh inoculum was collected from an operating UASB anaerobic digestion plant (Carbery Milk products). Initially, -1 the VS content in the inoculum was adjusted to 10 g VS L [26] by blending with tap water. At this VS concentration, the inoculum required was 6.5 L (Table 5.3), which, on initial experiments, proved to be too low in comparison to the size of the UASB reactor and thus resulted in high dilution of the feed and inoculum. The process of biogas production ceased very quickly after its initiation. In the second trial, half of the UASB was filled with inoculum from the reactor (15 L), by adjusting the VS content in the inoculum to 23 g VS, and half with leachate -1 from the leachate tank. The COD capability of the UASB was calculated to be 210 g COD d for 6.5 L of inoculum -1 and 482 g COD d for 15 L of inoculum (Table 5.3). 5.3.4 Experiment layout for SLBR-UASB A 3.5 kg portion of grass silage was loaded to each leach bed in a 7 day sequence. With six leach beds, this leads to an overall cycle time of 42 days. The leachate tank (capacity of 40 L) was filled with tap water. In the first 24 h, the UASB reactor was disconnected from the cycle to allow build up of the COD concentration. A pump flow rate of -1 -1 1180 L d was chosen, dispersing 200 L d over each batch. The COD concentration observed in the first 24 h was -1 13.5 g COD L . This equates to 540 g of COD, which is above the COD capacity of the UASB reactor (481.5 g COD) (Table 5.3). The feeding rate and the corresponding pump speed were calculated accordingly (Table 5.3). It -1 was noted that the pump speed cannot go above 17 rpm (100 L d ) when feeding the UASB as the upflow velocity -1 will exceed 10 cm hr and cause washout of the innoculum from the UASB reactor (Table 5.3). 5.4 Problems and modifications 5.4.1 Foam formulation Foam forms in the UASB when the reactor is fed at loading rates beyond the capacity of the bacterial population [27,28]. Accumulation of long-chain fatty acids within the reactor causes foaming [29]. If insoluble components with floating properties, such as fats and lipids, are not degraded, this can also result in foaming [30]. Foaming occurred during the start-up period of the UASB. The reactor was fed at high loading rates as it was believed that biogas was not being produced. In actuality, the biogas was being produced but was not recorded due to leakage of biogas through the U-tube connection connecting the UASB to the leachate holding cups (Figure 5.5c). Foam rose to the head of the UASB, condensed, and solidified around the walls of the pipes leading to the gas flow meter, exasperating the problem. To overcome this foaming problem, the following steps were taken: a) The loading rates were reduced to suit the capacity of the innoculum. b) A water holding cup with a capacity of 886 mL was installed at the top of the UASB to backwash the gas pipes and reduce buildup of foam at the top of the UASB (Figure 5.5a). Abdul-Sattar Nizami 91 Green Grass: Developing Grass for Sustainable Gaseous Biofuel c) A water filter with a volume of 2 L was positioned between the UASB and the gas flow meter to absorb any foam (Figure 5.5b). All gas passed through this water filter. e a b d c f Figure 5.5. Modifications for foaming and U-tube (a) water holding cup (b) water filter (c) small U-tube (d) deeper U-tube (e) plastic pipe (f) cylindrical unit 5.4.2 U-tube connection The leachate leaving the UASB en route to the leachate holding cups traveled via a U-tube connection. The gas flow meter was manufactured to have a restricted flow of biogas in order to give an accurate measurement. The diameter of the pipe leading to the gas flow meter was thus small. Because of the high pressure of the biogas at the top of the UASB and the restricted flow capacity of the gas pipes, the biogas traveled through the U-tube (Figure 5.5c) to the leachate holding cups, instead of passing through the narrow pipe to the flow meter. To overcome this problem, a deeper U-tube was installed (Figure 5.5d). However, after a few weeks of operation, this, too, began to allow the biogas to pass through, which pushed the leachate from the U-tube. It was noted that sometimes the biogas escaped from the U-tube in bubbles due to foaming. The ultimate solution involved the removal of the U-tube and the installation of two small cylindrical vessels (1.8 L volume) (Figure 5.5f). The leachate leaves the UASB through a straight plastic pipe and is discharged to the first cylinder. From this, the leachate passes to the second cylinder and finally to the leachate holding cups. The leachate has the same level at the top of both cylinders. The transparent plastic pipe allows observation of any potential risk of foaming. An additional gas collection point was installed in the first cylinder to capture any escaped biogas from the UASB (Figure 5.5e). Abdul-Sattar Nizami 92 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.4.3 Clogging The leachate was observed to have a considerable quantity of coarse particles; this resulted in clogging of various joints and components. This was initially noted at the nozzles of the sprinkling heads (Figure 5.6b). It resulted in irregular and weak sprinkling, rather than the intended smooth, homogeneous leachate shower over the grass silage in the leach bed. The particles also resulted in the failure of the sprinkling control mechanism. This mechanism consists of solenoid valves, which are designed to open and close at a certain intervals. The coarse particles coagulated at the surface of solenoid valves (Figure 5.6a), giving false signals to the process control system and resulting in valves remaining open continuously. Thus, the leachate holding cups never filled, and the water merely trickled out of the sprinkling heads. This situation would not lead to hydrolysis of the grass silage. The solution was to insert a removable sieve with pore sizes of 1.2 mm in diameter at the first leachate holding cup (Figure 5.6c). a b c Figure 5.6. Modifications in sprinkling mechanism (a) solenoid valves protecting valve (b) extended nozzles (c) removable sieve cup 5.4.4 Start-up of UASB These problems were detrimental to the start-up process. In particular, for a UASB system, granulation must be enhanced; the hydraulic and organic loading rate should increase the size and density of granules to achieve the optimum efficiency from the UASB. Changing the feed from cheese whey to grass silage liquor is already a challenge to the bacterial population besides the stop/go, overfeed/underfeed regime of this start-up process. These disturbances affect the UASB; a significant period of time is required to recover due to the slow growth rate of methanogenic bacteria [31]. As a result of overloading, there is an accumulation of VFAs. This induces the production of carbon dioxide and hydrogen gas in the biogas. Once the partial pressure of the hydrogen gas Abdul-Sattar Nizami 93 Green Grass: Developing Grass for Sustainable Gaseous Biofuel -4 increases to values of ca. 10 atm, disturbances occur in the metabolic pathways. It is beyond the limit of slowgrowing methanogenic bacteria to eliminate this level of H 2. As a result, the degradation rates of propionate, butyrate, and lactate are inhibited [32]. The combined effect is a significant change in the production of biogas and its composition [33]. 5.5 Results 5.5.1 Digester operation These problems dominated the first 84 days of operation, and as a result, a very low volume of biogas was produced (Figure 5.7a). On solving the miscellaneous previously described problems, the biological process of the UASB began to stabilize. The loading rates were increased with recirculation (Figure 5.7b) rates at an interval of 21 days (Figure 5.7c). The methane content of the biogas increased, and the volume of biogas produced increased with increasing loading rate (Figure 5.7a). 5.5.2 Methane production Using the Hohenheim methodology, an upper limit on methane production on this particular grass silage was -1 suggested at 350 L CH4 kg VS (Figure 5.8). It could be stated that this is the upper limit on the potential methane production. The methane yield from the SLBR-UASB is divided into different stages. All of these stages include for six cycles of 7 days between feeding, yielding an overall retention time of 42 days. For stages 2-5, the first day of feeding (every 7th day), the UASB was disconnected and a high recirculation rate was used to generate a high -1 -1 strength leachate. The pump was set to 1180 L d , which is roughly equivalent to 200 L d over each leach bed. This value was chosen based on a previous work on leaching and hydrolysis by Nizami and co-workers [12]. For days 2-7, the UASB is connected. Thus, it should be noted for stages 2-5 that the gas production is produced primarily in 6 days as the UASB is disconnected on 1 day in the 7-day cycle and produces little gas on that day. Stage 1 (day 1 to day 84): This stage accounts for the first 84 days after seeding the UASB and leaving all the gas off -1 from the inoculum. The methane production was 52 L CH 4 kg VS added. The low methane production was due to the aforementioned problems of foaming, gas leaking, clogging, and resultant poor sprinkling. -1 Stage 2 (85 to day 105): The feeding to the UASB was 3 rpm or 17.7 L d , which corresponds to an upflow velocity -1 of 0.017 m hr , which is 5.7 times less than the design upflow at steady state. The methane production in this 21-day -1 period was 139 L CH4 kg VS (more than twice the methane produced from the first 84 days). -1 Stage 3 (day 106 to day 126): The pump was set to 6 rpm or 35.4L d , which corresponds to an upflow velocity of -1 -1 0.034 m hr . The methane production increased to 175 L CH4 kg VS. -1 Stage 4 (day 127 to day 147): The pump was set to 12 rpm or 70.8 L d , which corresponds to an upflow velocity -1 -1 of 0.068 m hr . A methane production of 243.84 L CH4 kg VS was recorded. -1 Stage 5 (day 148 to day 189): The pump was set to 17 rpm (100.4 L d ). The upflow velocity is now at a maximum Abdul-Sattar Nizami 94 Green Grass: Developing Grass for Sustainable Gaseous Biofuel -1 -1 for the UASB (0.1 m hr ). A methane production of 305 L CH4 kg VS was recorded. This is 87% of the gas production in the batch test. 5.5.3 UASB efficiency The UASB efficiency is calculated by the expression in Equation 1 UASB efficiency (%) = 100 × COD in –COD out /COD in. Equation 1 where CODin is the COD flowing into the UASB and CODout is the COD leaving the UASB. The conversion rates of COD to methane in the UASB increased with increasing organic loading rates. The concentration of COD increased when the recirculation rates were increased. A UASB efficiency of 80% was achieved at -1 the highest recirculation rates employed: 100 L d from day 148 to day 189 (Figure 5.7b). 5.5.4 Dry solid removal efficiency Days 1-84: With a retention time of 42 days, the dry solids removal efficiency varied from 49% to 62%, averaging 53.5%. Days 85-189: After improving the hydrolysis mechanism and increasing the recirculation rates, the DS removal efficiency increased from 43.2% to 68.9%. -1 Days 148-189: During the last 42 days operating at the highest recirculation rates (100 L d ), the average DS removal was 66.6%. Abdul-Sattar Nizami 95 Green Grass: Developing Grass for Sustainable Gaseous Biofuel a b c Abdul-Sattar Nizami 96 Green Grass: Developing Grass for Sustainable Gaseous Biofuel d e Figure 5.7. Results of the experimental SLBR-UASB operation 5.5.5 Volatile solid removal efficiency Days 1-84: The volatile solids removal efficiency varied from 23.7% to 62.2%. Days 85-189: The VS removal efficiency increased from 52.2% to 70%. Days 148-189: The average VS removal efficiency was 68.1%. (Figure 5.7d). Abdul-Sattar Nizami 97 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 5.5.6 pH In the unstable/commissioning period (days 1-84), the pH of the UASB varied in the range of 6.3-7.8 while averaging 6.9. The biological process of the UASB was considered stable from day 85 to day 189. The pH range reduced (from 6.57 to 7.73), averaging at 7.43. In the unstable period, the pH of the leachate varied from 5.2 to 8.7, averaging 6.56. In the stable period, the pH ranged from 4.72 to 8.06 with an average of 7.04 (Figure 5.7e). 5.6 Discussion 5.6.1 Biological process During the last 42 days of operation of this system, it is postulated that the UASB was biologically stable and the hydrolysis rates in the leach beds were optimal. The pH of the UASB varied from 6.57 to 7.73, which is within the optimum pH range for UASB, as suggested by Tiwari et al. [9] and Kim et al. [34] of 6.3-7.8. The pH in the leach beds varied from 5 to 8. This is in line with the suggested range for optimal hydrolysis of 5-7 [12,35], 4-6.5 [36], or less than 7 [37,38]. The maximum DS and VS removal was 69% and 70%, respectively. This is in line with the observations made by Nizami et al. [12] who found the maximum removal of 70.6% for both DS and VS from the same grass silage as that employed in this study. The temperatures of the UASB and the leach beds were maintained at around 37 0C, which is stated as the optimum temperature within the mesophilic temperature range (35-40 0C) for anaerobic digestion [39,40]. 5.6.2 Methane production -1 Using the Hohenheim methodology, an upper limit on methane production was suggested at 350 L CH 4 kg VS (Figure 5.8). There is a wide range of values for methane production in the literature for digesting fresh and ensiled grass of different varieties [2,12]. Mähnert et al. [41] employing batch lab-scale digestion of fresh and ensiled grass of 3 -1 various species, observed methane production in the range of 0.31-0.36 m kg VS. Baserga and Egger [42] and 3 -1 KTBL [43] observed biogas production from 0.5 to 0.6 m biogas kg VS added using the same scheme at laboratory 3 -1 scale; at 55% methane, this corresponds to 0.28 0.33 m kg VS. Yu et al. [36] and Cirne et al. [44] observed a 3 -1 3 -1 methane production of 0.165 m CH4 kg VS added and 0.27 m kg VS using leach beds coupled with a high-rate anaerobic filter at lab scale. -1 This SBLB-UASB system yielded a methane production of 305LCH4 kg VS at the highest recirculation rates (100 L -1 d ), which corresponds to the upflow velocity limit of the UASB (Table 5.3). It may thus be stated that the system has obtained a significant degree of efficiency and is in the high range of expected methane production. Abdul-Sattar Nizami 98 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 0.40 0.30 3 -1 CH4 Production (m kg VS) 0.35 0.25 0.20 0.15 Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Sample 6 0.10 0.05 0.00 0 10 20 30 40 Days Figure 5.8. Results of the small batch experiments 5.6.3. Methane content of biogas The observed methane content was 53.8% on average from day 1 to day 84. The methane content increased to 70.2% when the water filter was employed to absorb the incoming foam from the UASB. This would be in line with the significantly higher solubility of carbon dioxide in water as compared with methane. Jung et al. [45] used a two-stage anaerobic process (first stage, acidogenic; second stage, methanogenic) at pilot scale treating pig slurry. The methane content of the biogas was observed in the range of 80-90% as the carbon dioxide was dissolved into the weak effluent wastewater. This observation is line with Sawyer et al. [46] who stated that the methane content of the biogas can be increased by dissolving carbon dioxide into water, as is the case in this research. 5.6.4 Retention times The majority of the methane was observed during the first 3 days after loading each new batch (Figure 5.7c). The UASB efficiency of 80% is recorded on the first day when the leachate contains high COD levels (on average, 16 g -1 L ). In the last 42-day cycle (day 148 to day 189), the system was optimized after eliminating all the start-up problems and the UASB was fed at its highest flow rate. The average observed methane production in the first 4 -1 -1 days after loading the new batch and attaching the UASB was 61 L CH4 kg VS added (4 * 61 = 244 L CH4 kg VS -1 added). In the last 2 days of each batch cycle of 7 days, the average methane production was 22 L CH4 kg VS added -1 -1 (2 * 22 =44LCH4 kg VS added). On average, 11 L CH4 kg VS added was observed in the first 24 h when the Abdul-Sattar Nizami 99 Green Grass: Developing Grass for Sustainable Gaseous Biofuel UASB was not fed and disconnected from the leachate tank (Figure 5.7c). According to Liu et al. [47] green waste can be digested in 12 days, generating high biogas production. After 50 days of grass silage digestion, the reported yield Lehtomäki and Björnsson [48] is 0.39 m3 CH4 kg-1 VS added at 59% VS removal using a two stage system: a leach bed coupled with a high-rate anaerobic filter at lab scale. However, they obtained 85% of total methane production from grass within the first 30 days. 5.6.5 UASB efficiency The UASB efficiency increased with increasing organic loading associated with increasing recirculation rates (Figure 5.7b). During the last 42 days, when the highest upflow rate and associated highest loading rate were applied, an efficiency of 80% was achieved. Mahmoud et al. [19] suggested increasing the organic loading rates through reduced retention times to increase the efficiency of the UASB. In addition, it is suggested that an increased concentration in the influent leads to a higher efficiency in the UASB. -1 The upflow velocity should be of the order of 0.1 m hr ; a constant upflow velocity is necessary for good mixing within the UASB [19]. Low upflow velocity does not generate the required turbulence in the sludge bed, which results in poor mixing. Excessive upflow velocity may lead to wash out of influent solids along with viable biomass. High solid loading rates are not of benefit as it results in a reduced solid retention time (SRT), leading to undegraded proteins and lipids, which affect the physical and chemical conditions of the microbial mass. Thus, the efficiency of the UASB system is controlled by the influent COD concentration and the up-flow velocity [49]. 5.6.6 Improving the efficiency of the system Within this system, it may be an option to use a second pump for hydrolysis, generating a more efficient leaching process producing a higher concentration of COD while maintaining a constant upflow in the UASB of 0.1 m hr -1 using the existing pump. The hydraulic retention time may be reduced to six cycles of 5 days, dropping the overall retention time from 42 days to 30 days. 5.7 Conclusion In designing digester systems, there is a significant difference between lab-scale work and small pilot-scale work. At small-scale pilot, the potential for engineering problems is significant. Often times, the operator can assume a biological problem when the problem is of an engineering nature. As in this research, the absence of gas flow recorded in the gas flow meter was assumed to be due to lack of gas production rather than a gas leak or gas taking an alternative route through the system. The perceived solution to the problem may exasperate the problem; as in this case, increasing loading to create gas flow actually over loaded the digester system. The SLBR-UASB showed itself to be an efficient system when the commissioning was completed. The recorded gas -1 production when the system was operated as designed (days 148-149) was of a high level (305 L CH4 kg VS added) at a retention time of 42 days, effecting a volatile solid reduction of 68%. This may be compared to the suggested -1 upper limit from a batch test of 350 L CH4 kg VS added. Abdul-Sattar Nizami 100 Green Grass: Developing Grass for Sustainable Gaseous Biofuel -1 The first 5 days of the 7-day cycle results in 86% of CH4 production (44 L of the total 305 L CH4 kg VS added is produced in the last 2 days of each 7-day cycle). There is scope to reduce the retention time to six cycles of 5 days (30 day HRT) rather than six cycles of 7 days (42 day HRT). This would suggest that a reduction in size of 29% would lead to a reduction in methane production of 14%. However, this may be improved by using a second pump -1 that has a function of purely recirculating the leachate over the leach beds (at flows in excess of 100 L d , limiting flow of UASB) to improve the rate of destruction of volatiles and increase the production rate of COD. The initial pump may now be used to pump the leachate to the UASB at a set rate corresponding to an upflow velocity of 0.1 m -1 hr . Higher concentration rates of COD would result, leading to higher conversion efficiencies in the UASB. Acknowledgements Research funding was obtained from the Department of Agriculture, Fisheries, and Food (DAFF) Research Stimulas Fund Project “GreenGrass”. We thank: Eddie Appelbe and Erneside Engineering for the fabrication and modification of the reactors: Padraig O‟Kiely and Joe McInerney from Teagasc, Grange, for the supply of grass silage; Richard Kearney from the Cork Institute of Technology (CIT) for help with all maters agriculture. We also thank Beatrice Smyth, Thanasit Thamsiriroj and James Brown for advice, brainstorming sessions, conversations and critiques. References [1] Vandevivere, P.; De Baere, L.; Verstraete, W. Types of anaerobic digester for solid wastes. In: Mata-Alvarez J, editor. Biomethanization of the organic fraction of municipal solid wastes. London: IWA Press; 2003. p. 112– 40. [2] Nizami, A. S.; Murphy, J. D. What is the optimal digester configuration for producing grass biomethane? Renewable and Sustainable Energy Reviews. 2010, 14 (6), 1558–1568. [3] Siciliano-Jones, J.; Murphy, M. R. Specific gravity of various feedstuffs as affected by particle size and in vitro fermentation. J. Dairy Sci. 1991, 74 (3), 896-901. [4] James, S.; Wiles, C.; Swartzbaugh, J.; Smith, R. Mixing in large-scale municipal solid waste-sewage sludge anaerobic digesters. In Biotechnology and Bioengineering Symposium, 1980, Vol. 10, pp. 259-272. [5] Prochnow, A.; Heiermann, M.; Plochl, M.; Linke, B.; Idler, C.; Amon, T.; Hobbs, P. J. Bioenergy from permanent grassland: A review. Bioresource Technol. 2009, 100 (21), 4931–4944. [6] Thamsiriroj, T.; Murphy. J. D. The difficulties associated with mono-digestion of grass as exemplified by commissioning a pilot scale digester. Energy and Fuels. 2010, doi: 10.1021/ef100303. [7] Thamsiriroj, T.; Murphy, J. D. Modelling mono-digestion of grass silage in a 2 stage CSTR using ADM1. Bioresource Technology. 2010, doi:10.1016/j.biortech.2010.09.051. [8] Paravira, W.; Murto, M.; Zvanya, R.; Mattiasson, B. Comparative performance of a UASB reactor and an anaerobic packed-bed reactor when treating potato waste leachate. Renew energy. 2006, 31 (6), 893-903. [9] Tiwari, M. K.; Saumyen, G.; Harendranath, C. S.; Shweta, T. Influence of extrinsic factor on granulation in UASB reactor. Appl. Microbiol. and Biotech. 2006, 71 (2), 145-154. Abdul-Sattar Nizami 101 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [10] Sayed, S. Z.; Lettinga, W. Anaerobic treatment of slaughter house waste using a flocculent sludge UASB reactor. Agric Wastes. 1984, 11 (3), 197-226. [11] Gooden, J.; Finlayson, M.; Low, E. W. A further study of the anaerobic bio-treatment of malt whisky distillery Potale using an UASB system. Bioresource Technol. 2001, 78 (2), 155-160. [12] Nizami, A. S.; Thamsiriroj, T.; Singh, A.; Murphy, J. D. The role of leaching and hydrolysis in a two phase grass digestion system. Energy and Fuels. 2010.doi:10.1021/ef100677s. [13] Ward, A. J,; Hobbs, P. J.; Holliman, P. J.; Jones, D. L. Optimization of the anaerobic digestion of agricultural resources, review. Bioresource Technology. 2008, 99 (17), 7928–40. [14] Nizami, A. S.; Korres, N. E.; Murphy, J. D. A review of the integrated process for the production of grass biomethane. Environmental Science and Technology. 2009, 43 (22), 8496–8508. [15] Esteves, S. R. R.; Wilcox, S. J.; O‟Neill, C.; Hawkes, F. R.; Hawkes, D. L., On-line Monitoring of AnaerobicAerobic Biotreatment of a Simulated Textile Effluent for Selection of Control Parameters. Environmental Technology. 2000, 21 (8), 927–936. [16] Esteves, S. R. R.; Wilcox, S. J.; Hawkes, D. L.; O‟Neill, C.; Hawkes, F. R. The Development of a Neural Network Based Monitoring & Control System for Biological Wastewater Treatment Systems. International Journal of Condition Monitoring & Diagnostic Engineering Management. 2001, 4 (3), 22–28. [17] Lehtomäki, A.; Huttunen, S.; Lehtinen, T. M.; Rintala, J. A. Anaerobic digestion of grass silage in batch leach bed processes for methane production. Bioresour. Technol. 2008, 99 (8), 3267–3278. [18] GonCalves, R. F.; Cha Lier, A. C.; Sammut, F. Primary fermentation of soluble and particulate organic matter for waste water treatment. Water Sci. Technol. 1994, 30 (6), 53–62. [19] Mahmoud, N.; Zeeman, G.; Gijzen, H.; Lettinga, G. Solids removal in upflow anaerobic reactors, a review. Bioresour Technol. 2003, 90 (1), 1–9. [20] Kosaric, N.; Blaszczyk, R.; Orphan, L. Factors influencing formation and maintenance of granules in upflow anaerobic sludge blanket reactors (UASBR). Water Sci Technol. 1990, 22 (9), 275–282. [21] Wiegant, W. M. Experiences and potentials of anaerobic wastewater treatment in tropical regions. Anaerobic digestion of sustainable development-Farewell seminar of Prof. Dr. Ir. Gatze Lettinga, Wageningen-The Netherlands, EP & RC. 2001. Pp. 111-118. [22] Vieira, S. M. M.; Gracia Jr, A. D. Sewage treatment by UASB-reactor. Operation results and recommendations for design and utilization. Water Science and Technology. 1992, 25 (7): 143-157. [23] Souza, M. E. Criteria for the utilization, design and operation of UASB reactors. Water Science and Technology. 1986, 18 (12): 55-69. [24] APHA. Standard methods for the examination of water and wastewater. 20 th ed. American Public Health Association, Washington DC. 1998. [25] Helffrich, D.; Oechsner, H. The Hohenheim biogas yield test. Landtechnik, 2003, pp. 148–149. [26] Forbes, C.; O'Reilly, C.; McLaughlin, L.; Gilleran, G.; Tuohy, M.; Colleran, E. Application of high rate, high temperature anaerobic digestion to fungal thermozyme hydrolysates from carbohydrate wastes. Water Res. 2009, 43 (9), 2531-9. [27] Shin, H.; Paik, B. Improved performance of UASB reactors by operating alternatives. Biotechnol. Lett. 1990, 22 (6) 469-474. Abdul-Sattar Nizami 102 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [28] Cuervo Lopez, F. M.; Martinez, F.; Gutierez Rojas, M.; Noyola, R. A. Effect of N loading rates and carbon sources on denitrification and sludge settleability in UASB reactor. Water Science and Technology. 1999, 40 (8), 123–130. [29] Tchobanoglous, G.; Burton F. L., and Stensel, H. D. Wastewater engineering. treatment and reuse. 4th edition. 2003. McGraw-Hill Companies, New York, USA. [30] Miranda, L.A.S.; Henriques, J.A.P.; Monteggia L.O. A full-scale uasb reactor for treatment of pig and cattle slaughterhouse wastewater with a high oil and grease content. Brazilian Journal of Chemical Engineering. 2005 22 (04), 601-610. [31] Bhatia, D.; Vieth, W. R.; Venkatasubramanian, K. Steady-state and transient behavior in microbial methanification. I: experimental results, Biotechnol. Bioeng. 1985, 27 (8), 1192–1198. [32] Fongastitkul, P.; Mavinic, D. S.; Lo, K. V. A two-phased anaerobic digestion process: concept, process failure and maximum system loading rate. Water environment research. 1994, 66 (3), 243-254. [33] Chua, H.; Hu, W. F.; Yu, P. H. F.; Cheung, M. W. L. Response of an anaerobic fixed-film reactor to hydraulic shock loadings. Bioresource Technology. 1997, 61 (1), 79-83. [34] Kim, J.; Park, C.; Kim, T. H.; Lee, M.; Kim, S.; Kim, S. W.; Lee, J. Effects of various pretreatments for enhanced anaerobic digestion with waste activated sludge. J. Biosci. Bioeng. 2003, 95 (3), 271–275. [35] Veeken, A.; Kalyuzhnyi, S.; Scharff, H.; Hamelers, B. Effect of pH and VFA on hydrolysis of organic solid waste. J. Environ. Eng. 2000, 126 (12), 1076–1081. [36] Yu, H. W.; Samani, Z.; Hanson, A.; Smith, G. Energy recovery from grass using two-phase anaerobic digestion. Waste Manage. 2002, 22 (1), 1–5. [37] Dinamarca, S.; Aroca, G.; Chamy, R.; Guerrero, L. The influence of pH in the hydrolytic stage of anaerobic digestion of the organic fraction of urban solid waste. Water Sci. Technol. 2003, 48 (6), 249–254. [38] Babel, S.; Fukushi, K.; Sitanrassamee, B. Effect of acid speciation on solid waste liquefaction in an anaerobic digester. Water Res. 2004, 38 (9), 2417–2423. [39] Zoetemeyer, R. J.; Arnoldy, P.; Cohen, A.; Boelhouwer, C. Influence of temperature on the anaerobic acidification of glucose in a mixed culture forming part of a two-stage digestion process. Water Res. 1982, 16 (3), 313–321. [40] Choorit, W.; Wisarnwan, P. Effect of temperature on the anaerobic digestion of palm oil mill effluent. Electronic Journal of Biotechnology. 2007, 10 (3), 376-385. [41] Mähnert, P.; Heiermann, M.; Linke, B. Batch- and semi-continuous production from different grass species. Agricultural Engineering International The CIGRE E Journal. 2005. Manuscript EE 05 010, vol. V11. [42] Baserga, U.; Egger, K. Verga¨rungvonEnergiegras zurBiogasgewinnung.Bundesamt fur Energiewirtschaft, Forschungsprogramm Biomasse. Tanikon; 1997, 41 p. [43] KTBL, Gasertrage. Gasausbeuten in landwirtschaftlichen Biogasanlagen. Darmstadt: Association for Technology and Structures in Agriculture, KTBL; 2005. pp. 10–16. [44] Cirne, D. G.; Lehtomäki, A.; Björnsson, L.; Blackall, L. L. Hydrolysis and microbial community analyses in twostage anaerobic digestion of energy crops. Applied Microbiology. 2007, 103 (3), 516–27. [45] Jung, Jin-Young.; Lee, Sang-Min.; Shin, Pyong-Kyun.; Chung, Yun-Chul. Effect of pH on phase separated anaerobic digestion. Biotechnology and Bioprocess Engineering. 2000, 5 (6), 456-459. Abdul-Sattar Nizami 103 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [46] Sawyer, C. L.; McCarty, P. L.; Parkin, G. F. Chemistry for Environmental Engineering 4th ed., 1994. pp. 608616. McGraw-Hill International editions, NY, USA. [47] Liu, G.; Zhang, R.; Hamed, M.; El-Mashad.; Withrow, W. Dong R. Biogasification from kitchen and grass wastes using batch and two-phased digestion. Journal of China Agricultural University. 2006, 11(6), 111–5. [48] Lehtomäki, A.; Björnsson, L. Two-stage anaerobic digestion of energy crops: Methane production, nitrogen mineralization and heavy metal mobilisation. Environmental Technology. 2006, 27 (2), 209–18. [49] Singh, A.; Nizami, A.S.; Korres, N.E.; Murphy, J.D. The effect of reactor design on the sustainability of grass biomethane. Renew Sustain Energy Rev(2010), doi:10.1016/j.rser.2010.11.038 Abdul-Sattar Nizami 104 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 6: An Optimized Two Phase Digestion System for the Production of Gaseous Biofuel from Grass Silage, A High Solids Content Feedstock Abdul-Sattar Nizami 105 Green Grass: Developing Grass for Sustainable Gaseous Biofuel An optimized two phase digestion system for the production of gaseous biofuel from grass silage, a high solids content feedstock Abdul-Sattar Nizamia,b, Jerry D. Murphy,a,b* a Department of Civil and Environmental Engineering, University College Cork, Cork, Ireland b Biofuels Research Group, Environmental Research Institute, University College Cork, Cork, Ireland Abstract Anaerobic digestion of high solid content feedstock for sustainable renewable gaseous fuel production is a relatively new endeavor when compared to the ubiquitous technology of digestion of liquid slurries and sludges for waste treatment. For renewable energy production it is essential to maximize the resource per unit of feedstock (L CH4 kg-1 VS added). For sustainability it is also required to minimize the potential for biodegradability (and greenhouse gas emissions) post anaerobic digestion. This paper examines the optimization of a two phase anaerobic digestion process using grass silage as a feedstock. Previous work on this system had shown the potential to produce 305 L CH4 kg-1 VS added at retention time of 42 days; the theoretical maximum production is 490 L CH 4 kg-1 VS added. The optimised system investigated potential for reduction in retention time and separation of flows to the first stage (leach beds) and the second stage (UASB) through addition of an extra pump to optimize leaching. The optimised system increased the CH4 production by 11.8%, and effected a reduction in size or retention time of 40% (42 days decreased to 30 days retention time). The optimized system is shown to be very stable. The alkalinity and the pH of the system is as should be with reference to the scientific literature. The granules are healthy: the diameter of 2.55mm suggests a UASB subject to high organic loading rates; the populations of bacteria, in particular filamentous bacteria are associated with high COD removal. Keywords: UASB; digestion; grass silage; biogas; biomethane. * Corresponding author. Tel +353 21 4902286 Fax +353 21 4276648 E-mail address: jerry.murphy@ucc.ie Abdul-Sattar Nizami 106 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 6.1 Introduction 6.1.1 Anaerobic reactor design for high solid content feedstock Anaerobic digestion (AD) of organic feedstock with less than 12% dry solids is well studied, well published, mature and implemented successfully at full scale [1,2]. The technology for high solid content substrates (greater than 20% dry solids content) is not as mature or as well published [3]. Dry digestion of high solid content feedstock includes for: leach-bed reactors [4] for the organic fraction of municipal solid waste (OFMSW); plug flow reactors for fruit and vegetable waste [5]; and solid phase reactors with leachate recycling for OFMSW [6]. A good digester should increase solid retention time, minimise hydraulic retention time, decrease reactor size and reduce process energy input [7]. Good research and development (at laboratory and pilot scale) allowing for reactor optimisation is required [8,9] prior to design of a full scale system [10]. 6.1.2 Methane yields from grass biomethane digesters A yield of 306 L CH4 kg-1 VS added was observed by Lehtomäki et al. [11] using a dry batch digester at laboratory scale to digest an ensiled mixture of timothy and meadow fescue. Mähnert et al. [12] achieved methane yields in the range of 310-360 L CH4 kg-1 VS added from a batch experiment using different types of grass. Baserga and Egger [13] and KTBL [14] both using batch lab-scale experiments on grass silage yielded 280 to 330 L CH 4 kg-1 VS added. Thamsiriroj and Murphy15 digested grass silage in 2-stage continuously stirred tank reactor (CSTR) complete with recirculation of liquid digestate at a retention time of 60 days, obtained 455 L of CH4 kg-1 of VS added. Coupling a dry batch digester to an upflow anaerobic sludge blanket (UASB) or another high rate reactor shows distinct advantages over a simple dry batch process [7,16,17]. The dry batch process may be optimised for hydrolysis due to the ability of the batch to process high solid content (up to 50 %DS) [8]. The UASB or high rate digester may be optimised for methanogensis due to its ability to accept organic loading rates up to 20 kg soluble chemical oxygen demand (COD) m-3 reactor d-1[7]. Methane production rates of 0.165 m3 kg-1 VS added and 0.27 m3 kg-1 VS added was observed by Yu et al. [18] and Cirne et al. [19] using lab-scale leach beds connected to a high rate digester (in this case an anaerobic filter). 6.1.3 The requirement for a stable UASB system The successful operation of an UASB is significantly correlated with the development of granules [20]. The granulation is the natural self-immobilization phenomenon of different bacteria [21]. The active biomass concentration of granules together with the influent flow rate and organic loading determines the efficiency of the UASB system. The granulation process is affected by system operating conditions including for: composition and strength of substrate; system hydrodynamics; presence of trace metals, polymers and metal ions; and microbial ecology [22]. The structure of the granule is affected by the nature of organic compounds in the substrate, the kinetics of substrate degradation and the concentration of different microbial species [23]. The UASB granules are dense with a particle size of 1-4 mm [24]. The stability of the UASB is critical due to its long start-up time and its Abdul-Sattar Nizami 107 Green Grass: Developing Grass for Sustainable Gaseous Biofuel sensitivity to shock loading [25]. Higher COD removal efficiency is achieved when the granules are well developed. Nevertheless, the mechanism involve in the process of granulation is unknown [21]. 6.1.4 Digestate: a value-added product The AD process produces an end product (biofuel, biogas or biomethane) and a by-product (digestate). The AD process conserves and mineralises the nutrient value of the original feedstock. Thus the digestate may be used as an effective substitute for conventional fossil fuel based fertilizers; it may also be used as a value added product [26]. The digestate may be separated into liquid and solid digestate [8]. Some portion of liquid digestate may be recirculated back to the AD process to reduce fresh water input and increase process efficiency [27]. The remaining digestate may be further processed into liquid biofertilizer and nutrient rich press juice [26]. The solid digestate can be processed into fibres, which can be applied to land as soil conditioner, be used for thermal combustion or converted into insulation boards [28]. This demonstrates an array of uses (i.e. energy, chemicals and materials) from one commodity, grass silage [29]. Data on chemical composition of grass digestate is not readily available in the scientific literature [30]. 6.1.5 Previous work on two stage grass digestion This paper builds on two previous papers. Nizami et al. [17] highlighted the potential for leaching volatile solids from grass silage to chemical oxygen demand (COD) in a liquid stream. They found that a flow rate of 100 L d -1 sprinkled over a leach bed containing 3.5 kg of baled grass silage (32% dry solids) effected a 70% destruction of volatile solids. Each kg of VS converted to 1.4 kg COD. Theoretically each kg COD should produce 350 L CH4 in a UASB reactor. Thus the maximum conversion of volatile solids to methane is 490 L CH 4 kg-1 VS; 70% destruction should equate to 343 L CH4 kg-1 VS added at 100% UASB efficiency [17]. The second paper [7] outlined the commissioning and design of a sequentially fed leach bed reactor complete with an upflow anaerobic sludge blanket (SLBR-UASB). The system consisted of 6 leach beds fed sequentially every 7 days (Figure 6.1). At a 42 day retention time (6 batches fed every 7 days in series) 68% of volatiles were destroyed and a methane production rate of 305 L CH4 kg-1 VS added was achieved; 86% of methane production occurred in the first 5 days after feeding. Recirculation flow rate over the feedstock was dictated by the upflow velocity of the UASB which cannot exceed 0.1m hour-1. The maximum flow equated to 100 L day-1 which equated to approximately 17 L day-1 over each of the six leach beds. Thus the UASB limited the flow which in turn reduced the potential to leach VS in the grass silage to COD in the liquid stream. 6.1.6 Focus of paper This paper deals with the optimization of a small pilot scale, SLBR-UASB. The retention time will be reduced to 6 periods of 5 days (30 day retention time) as opposed to 6 periods of 7 days (42 day retention time) to assess the effect on methane production. The effect of adding a second pump to the system (Figure 6.1) allowing differentiation between upflow velocity in the UASB and recirculation rate over the leach beds is investigated. Of Abdul-Sattar Nizami 108 Green Grass: Developing Grass for Sustainable Gaseous Biofuel interest in this paper is the maximum production of methane, the retention time and the operation of the UASB system including inspection of the granules. 6.2 Methodology 6.2.1 Characteristics of grass silage The substrate for the digester system was baled grass silage (Table 6.1). It was prepared by the Irish Agricultural Institute „Teagasc‟ at the Grange Research Centre, Dunsany, Co Meath, Ireland. The silage consisted of homogenous perennial ryegrass. The grass was cut at early mature stage, and was field wilted for 24 hours before baling. The bales were wrapped by using polythene stretch-film and stored for five weeks. Small square bales of 25 kg were prepared for experimental use. The dry solids content was established as 30%; the volatile solids content was 92% of the dry solids. The silage was macerated to average particle size of approximately 20 mm using a mobile macerator (approximated capacity of 10 kg of silage h-1) and frozen immediately at -15oC. The samples were thawed overnight at 6oC before feeding to the reactor. Table 6.1. Characteristics of input grass silage in the study Parameters Measuring Unit Values Lactic acid Ethanol Acetic acid Propionic acid Butyric acid VFA Ammonia WSC pH Protein ME DMD or D-value Silage intake or Palatability PAL NDF FME FME/ME ratio Oil C H N DS VS g kg-1DS g kg-1DS g kg-1DS g kg-1DS g kg-1DS g kg-1DS g kg-1DS g kg-1DS 26.95 11.54 3.93 0.25 1.43 5.61 46.18 49.83 4.3 9.5 10 64 89 821 59 8.2 0.81 3.3 43.03 5.82 1.61 30.66 92.46 Abdul-Sattar Nizami % DS MJ kg-1 DS % DS or D-value g kg-1 W0.75 meq kg-1 DS % DS MJ kg-1 DS % DS % DS % DS % DS % % 109 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 6.2.2 Scheme of SLBR-UASB system The SLBR-UASB system (Figure 6.1) combines six leach bed reactors connected to a high rate UASB reactor. Nizami et al. [7] described in detail the design and commissioning of the SLBR-UASB system. In brief the liquor is sprinkled from six cups over the feedstock in the six batch leach beds. A leachate high in COD collects in the leachate holding tank and from there may either be pumped into the UASB at a maximum upflow of 0.1m h -1 or recirculated back to the sprinkling cups for discharge over the leach beds. Biogas is produced over the UASB and the liquor is sent back to the sprinkling cups for discharge over the leach beds. Hydrolysis takes place in the leach beds; 1.4 kg COD is generated per kg of volatile solids (VS) destroyed. Methanogensis takes place in the UASB; 350 L CH4 is generated per kg of COD destroyed [17]. The leach beds are fed sequentially. For a 30 day retention time batch 1 is fed on day 0, batch 2 day 5, batch 3 day 10, batch 4 day 15, batch 5 day 20, batch 6 day 25; batch 1 is emptied and refilled on day 30. 6.2.3 Experimental set-up The first 189 days of operation of this SLBR-UASB is described by Nizami et al. [7]. This period covered designing and commissioning problems, system start-up and problems in achieving steady state. Commissioning was deemed complete by day 147 and a steady state was effected between days 148 – 189. In this period methane production was 305 L CH4 kg-1 VS added at a retention time of 42 days effecting a VS reduction of 68%. It was observed that 86% of CH4 was produced in the first 5 days of the 7 day digester feeding cycle. This paper outlines a reduction of the feeding cycle to 5 days with an overall 30 day hydraulic retention time (HRT). Afterwards a second pump is introduced to differentiate recirculation of leachate from the pumped upflow to the UASB. The experimental period is divided in two: Experimental Scheme 1 (Day 1 to 60): The system is fed with 3.5 kg of grass silage in each leach bed in a 5 day sequential cycle; leading to an overall HRT of 30 days. Initially tap water is used to fill the leachate tank (capacity of 40 L). On the first day of each 5 day cycle (after loading the new batch) the UASB was disconnected and high strength leachate was generated using a high recirculation rate. The pump was set to 1180 L d -1 which recirculated the equivalent of 200 L d-1 over each leach bed. On day 2 to day 5, the leachate was fed to the UASB at flow rate of 100 L d-1; generating an upflow velocity of 0.1 m h-1 in the UASB. The 30 day experiment was duplicated. Experimental Scheme 2 (Day 61 to 120): A second pump was added to allow separate recirculation; a pump rate of 600 L day-1 equating to 100 L day-1 over each leach bed. This matches the hydrolysis experiments by Nizami et al. [17] which effected 70% destruction of volatiles in baled grass silage over a 30 day period. If this destruction of volatiles takes place then methane production should equate to 343 L CH 4 kg-1 VS added (assuming 100% UASB efficiency). The existing pump is continued at 100 L day-1 generating 0.1 m h-1, the maximum allowable upflow velocity in the UASB. This experimental scheme was also duplicated. Thus, this paper presents the last 120 days of SLBR-UASB operation out of a total 309 days operation. This paper seeks to optimise the system. Abdul-Sattar Nizami 110 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Second-phase (UASB) a First-phase (Sequencing Batch Leach bed) b Figure 6.1. Sequentially fed leach bed reactor complete with an upflow anaerobic sludge blanket (SLBR-UASB) (a) Schematic diagram [7,16] (b) Pectoral diagram 6.2.4 Analytical Methods DS and VS contents of the input and output grass silage were analyzed using standard methods [31]. Three samples were taken on loading each new batch of silage to the leach bed. Three samples of the digestate from the previous batch were also taken. The average values were used for DS and VS removal calculations during each loading and Abdul-Sattar Nizami 111 Green Grass: Developing Grass for Sustainable Gaseous Biofuel unloading cycle. The COD concentration was measured every second day of digester operation by a COD analyzer (HACH DRB200 and DR/2800, USA). The total volatile fatty acids (TVFA) and alkalinity of leachate were measured once a week using titration based methods [32]. Before the start up of the experiments, a complete analysis of grass silage based on its feeding value for dairy cattle and digestate value (combined liquid and solid fraction) was carried out by Agri-Food and Biosciences Institute (AFBI), Belfast, Northern Ireland and the Irish Agricultural Institute „Teagasc‟. The C-H-N ratio of grass silage were analyzed by elemental analyzer (CE 440 MODEL) using the ultimate analysis method. The pH of both leachate and UASB effluent were measured digitally every hour using a pH sensor probe (Signet 2754-2757 DryLoc pH/ORP) controlled by the programmable logic controller (PLC) of the system. The measurement of gas flow was carried out on an hourly basis by a digital gas flow meter (FMA-1600A Omega) also controlled by the PLC system. Biogas was analysed for its composition on a daily basis using a portable biogas analyser (PGD3-IR, Status Scientific Controls Ltd). The digestate value of liquid and solid fractions was tested for total kjeldahl nitrogen (TKN), total phosphorous (P) and total potassium (K) analysis. TKN was analysed using QuickChem IC and FIA flow injection analyzer (8000 series, Lachat Instrument). Total-P was analysed using a spectrophotometer (Shimadzu UV-160A). Total-K was analysed using an atomic absorption spectrometer (SpectrAA-300, Varian). The granules of the UASB were observed using electron microscope (FEI Inspect F) using conductive carbon tap at acceleration voltage of 5 kv. The different bacterial types present in the granule were identified based on their morphology. 6.3 Results 6.3.1 Methane production Experimental scheme 1 (Day 1 to 60): The methane production was 306 L CH4 kg-1 VS added for the first 30 day cycle. This increased to 314 L CH4 kg-1 VS added for the second 30 day cycle. The average is taken as 310 L CH 4 kg-1 VS added (Figure 6.2a). Experimental scheme 2 (Day 61 to day 120): With the addition of the second pump methane production increased substantially; to 339 L CH4 kg-1 VS added in the first cycle of 30 days and 344 L CH 4 kg-1 VS added in the second cycle of 30 days. The average is taken as 341 L CH 4 kg-1 VS added (Figure 6.2a). 6.3.2 UASB efficiency The UASB efficiency was calculated by the expression in Equation 1 below: UASB efficiency (%) = 100 × COD in – COD out /COD in Eqn (1) Where, CODin is the COD flowing into the UASB CODout is the COD leaving the UASB. Efficiencies of 90 and 93% were achieved in the first and second experimental scheme respectively (Figure 6.2b). Abdul-Sattar Nizami 112 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Days a 0 20 40 60 80 100 120 75 74 72 71 50 70 69 40 68 67 CH4 Production 66 % CH4 in Biogas 20 2050 0 OLR (g COD d-1) b 20 40 60 80 100 65 100 120 1950 95 1850 90 1750 85 1650 80 1550 75 1450 70 OLR 1350 65 % UASB efficiency 1250 c 60 0 % DS Removal % UASB efficiency 30 % CH4 in Biogas 73 60 20 40 60 80 100 120 79 79 77 77 75 75 73 73 71 71 69 % VS Removal CH4 production (L d-1) 70 69 % VS Removal 67 67 % DS Removal 65 500 0 40 60 80 100 65 700 120 Leachate TVFA (mg L-1) 600 400 550 350 500 300 450 250 400 200 150 8 0 Leachate TVFA Leachate Alkalinity 20 40 60 80 100 300 7.5 7 7 pH - Leachate 350 140 8 120 7.5 e Leachate Alkalinity (mg L-1) 650 450 6.5 6.5 6 6 5.5 5.5 5 5 4.5 4.5 pH - Leachate pH - UASB 4 4 3.5 3.5 0 20 40 60 80 100 Days Figure 6.2. Results of the experimental SLBR-UASB operation Abdul-Sattar Nizami 113 120 pH - UASB d 20 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 6.3.3 Dry solid removal efficiency Day 1 - 60: With a retention time of 30 days the dry solids removal efficiency varied from 68 % to 71% averaging 69% (Figure 6.2c). Day 61 - 120: In the second experimental scheme, the DS removal efficiency increased from 72% to 74% averaging 73% (Figure 6.2c). 6.3.4 Volatile solid removal efficiency Day 1 - 60: The volatile solids removal efficiency varied slightly from 70 % to 71%. Day 61 – 120: The VS removal efficiency increased from 74% to 77% with average value of 75% (Figure 6.2c). 6.3.5 Alkalinity and total volatile fatty acids During both experimental schemes, the alkalinity of the leachate varied from 340 to 680 mg L-1. The total volatile fatty acids in the leachate varied from 208 to 440 mg L-1 (Figure 6.2d). 6.3.6 pH Day 1 - 60: The pH of the leachate varied from 4.5 to 7.0 and averaged 6.0. The pH range of the UASB varied from 6.9 to 7.7 with an average of 7.4. The biological process of the UASB was considered stable (Figure 6.2e). Day 61 – 120: The pH of the leachate varied from 4.5 to 7.0 averaging 5.9. The pH of the UASB ranged from 6.8 to 7.7 with an average of 7.4 (Figure 6.2e). 6.3.7 Morphology and structure of granule The average size of granule was found to be 2.55 mm. On visual examination, the granules were dark grey in color with a spherical shape (Figure 6.3a). The surface of the granule was rough and uneven with irregular projections (Figure 3b). On the surface of the granule, heterogeneous populations of bacteria were present (Figure 6.3c). The large non-staining dark center was seen in the cut cross sectional area of the granule (Figure 3d). The rod-shaped bacteria (Methanothrix) and cocci-shaped bacteria (Methanosaricina) were dominant on the granule surface (Figure 6.3e & g). Filamentous-shaped bacteria (Methanosaetaceae) were seen in abundance on the granule edges (Figure 6.3h). Figure (6.3f) showed the presence of round ended (Methanobacterium/Methanobrevibacter) in core and internal layers of the granules. Abdul-Sattar Nizami 114 short rod bacteria Green Grass: Developing Grass for Sustainable Gaseous Biofuel a b 1mm 500 m c d 30m 500m e f 5m 3m g h 30m 50m Figure 6.3. Morphology and structure of granule Abdul-Sattar Nizami 115 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 6.3.8 Digestate value After completing the last cycle of 30 days, the digestate (combined liquid and solid fraction) was analysed for DS, VS, neutral detergent fibre (NDF; structural components in plant cells such as lignin, hemicellulose and cellulose), acid detergent fibre (ADF; the fibrous, least-digestible portion of roughage, highly indigestible) and total-N (Table 6.2). The DS content of digestate was 10.18%. The value of NDF was still high 55.2% (Table 6.2) in comparison to input substrate (59%; Table 6.1). For a detailed analysis of nutrients (N, P, K) and fibres in the digestate, the digestate was separated into liquid and solid fractions by manual pressing. The solid digestate retained higher quantities of total kjeldahl nitrogen (TKN) (18.57 g kg-1 DS), total-P (2.72 g kg-1 DS) and total-K (36.49 g kg-1 DS) in comparison to liquid digestate (Table 6.3). Table 6.2. Characteristics of combined digestate (liquid + solid) Measuring Unit % % %DS % DS Values 10.18 81.78 18.22 3.71 Neutral Detergent Fibre % DS 55.2 Acid Detergent Fibre % DS 35.8 Sulphur (S) %DS 1.92 Parameters Dry Solid Volatile Solid Ash Total-Nitrogen Table 6.3. Characteristics of liquid and solid digestate Parameters Total Kjeldahl Nitrogen (TKN) Total-P Total-K Liquid digestate (kg m-3) 1.42 0.313 3.36 Solid digestate (g kg-1 DS) 18.57 2.72 36.49 6.4 Discussion of results 6.4.1 Stability of biological process The UASB was biologically stable for the 120 days including for both experimental periods. The pH in the UASB stayed in the range 6.8 to 7.7. This is in the optimal range as suggested by Tiwari et al. [24] 6.6-7.7 and Kim et al. [33] 6.3-7.8. The pH of the leachate varied from 4.5 to 7.0. This is very close to the range suggested for optimal hydrolysis e.g. 5 to 7 [17], 4 to 6.5 [18] or less than 7 [34]. Hydrolysis rates were good. In a previous work on hydrolysis and leaching (without the UASB) Nizami et al. [17] using the same grass silage found DS and VS removal efficiencies of 70.6% when sprinkling 100 L day -1 over each batch. In both experimental periods (with the UASB connected) over 70% destruction of volatiles was effected. Abdul-Sattar Nizami 116 Green Grass: Developing Grass for Sustainable Gaseous Biofuel The addition of the second pump increased the DS and VS removal efficiency to 73 and 75% respectively. The results are in the higher levels as compared with other works such as Ramasamy and Abbasi [35] who observed destruction of VS in the range 37% to 78% during digestion of various lignocellulosic substrates. The biodegradability of grass silage is highlighted by the acid detergent fiber content (35.8 %DS) as compared to the neutral detergent fibre content (55.2%DS) in the digestate (Table 6.2). Nizami et al. [16] explained that in efficient hydrolysis, the cellulosic part of acid detergent fibre is more solubilised than the insoluble carbohydrates of neutral detergent fibre. 6.4.2 Methane production and retention time The SLBR-UASB was optimised during the second experimental scheme with the separation of flow rates for hydrolysis and upflow in the UASB. The system produced 341 L CH 4 kg-1 VS added with 75% removal of VS at a 30 day retention time. The relationship between COD, VS and CH 4 during the second experimental scheme is shown in Table 6.4. There is an excellent correlation between the numbers as exemplified below: Period 1: 70.5% destruction of volatiles @ 1.4 kg COD kg-1 VS destroyed with an efficiency of 90% in the UASB generating 350 L CH4 kg-1 COD removed equates to 311 L CH4 kg-1 VS added. The actual value recorded is 310 L CH4 kg-1 VS added. Period 2: 75% destruction of volatiles @ 1.4 kg COD kg-1 VS destroyed with an efficiency of 93% in the UASB generating 350 L CH4 kg-1 COD removed equates to 342 L CH4 kg-1 VS added. The actual value recorded is 341 L CH4 kg-1 VS added The methane production range is in line with expected methane production from fresh and ensiled grass of different varieties [8,17]. Nizami et al. [7] and Asim et al. [36] suggested an upper limit of 350 and 361 L CH4 kg-1 VS from grass silage using a biomethane potential (BMP) assay. A range of 310-360 L CH4 kg-1 VS added was observed by Mähnert et al. [12] in a batch experiment from fresh and ensiled grass of various species. The methane production associated with the short retention time (30 days) is high when compared to other two phase digestion schemes. Yu et al. [18] and Cirne et al. [19] using leach bed reactors connected with an anaerobic filter obtained 165 and 270 L CH4 kg-1 VS added respectively. Using the same reactor scheme Lehtomäki and Björnsson [37] obtained a methane yield of 390 L CH4 kg-1 VS added but at a 50 day retention time. However, 85% total methane (or 330 L CH4 kg-1 VS) was achieved in the first 30 days of digester operation. This is similar but less than the results recorded here. 6.4.3 Methane content of biogas The average methane content during the 120 day experimental period was 70.6%. For grass digestion a value of 55% is typical [8]. This increased level of methane is associated with the solubility of CO 2; the biogas is passed through a water filter prior to measurement of flows. This increment in methane content is in line with the observations made by Sawyer et al. [38] and Jung et al. [39]. In particular Jung et al. [39] observed a methane content of 80-90% when the biogas was passed through a weak liquor digestate from a two-stage anaerobic process at pilot scale treating pig slurry. Abdul-Sattar Nizami 117 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 6.4. The relationship between COD, biogas and methane production in the second experimental period Feedstock = 21 kg = 6.3 kg DS = 5.8 kg VS = 4.35 kg VS = 30 days COD production COD produced = 1.4 kg COD kg-1 VS = 4.35kg VS *1.4 kg COD kg-1 VS = 6.09 kg COD @ 93% efficiency in UASB = 5.68 kg COD / 30 days = 0.189 kg COD d-1 Methane production Methane produced @ 93% COD removal efficiency of UASB = 0.35 m3CH4 kg-1 COD = 0.35 m3CH4 kg-1 COD *6.09 kg COD *0.93 = 1.98 m3 CH4 = 0.455 m3 CH4 kg-1 VS destructed = 0.341 m3 CH4 kg-1 VS added Biogas and Methane production Methane produced = 1.98 m3 CH4 / 30 days = 0.06 m3 CH4 d-1 = 66 L CH4 d-1 Biogas produced @70.6% methane = 2.8 m3 biogas / 30 days = 93.48 L d-1 Grass silage DS content @ 30% VS content @92 % of DS VS conversion @ 75% Retention time 6.4.4 UASB efficiency The UASB efficiency varied between 67.5 to 93%. High efficiency in a UASB is attributed to the constant upflow velocity (< 0.1 m hour-1), increased influent COD concentration and high quality granules [7,40]. A constant upflow velocity is required for a good mixing in the UASB; low upflow velocity results in poor mixing and turbulence in the sludge bed, while there is a risk of washout of granules at high upflow velocity [40]. The UASB efficiency is increased by increasing the influent concentration [40]. 6.4.5 Alkalinity and TVFA The observed Alkalinity and TVFA of leachate varied from 340 to 680 and 208 to 440 mg L -1 respectively. For granulation process, the alkalinity and TVFA of the leachate are important [41]. According to Singh et al. [42], the optimum alkalinity within the influent to the UASB should be in the range 250-950 mg L-1; whilst the TVFA should be less than 1000-1500 mg L-1 [43]. 6.4.6 UASB granule structure and morphology The average size of the granule was 2.55 mm. Gupta and Gupta [21] stated that the size of granule is greater than 2 mm when the UASB is fed at high OLR. Larger granules remove COD more efficiently, typically converting 95% of COD converted to methane while the remaining 5% is converted to biomass [44]. The shape of the granules varies Abdul-Sattar Nizami 118 Green Grass: Developing Grass for Sustainable Gaseous Biofuel with the operating conditions but usually have a spherical shape [45] as shown in Figure (6.3a). Large dark centers are noted in the centre of the granule as may be noted in the cut cross section of the granule (Figure 6.3d). This does not agree with the initial theory of a layered structure in the UASB granules [23]. The non-layered structure of the UASB granule was also observed and reported by Wu et al. [46]. According to Fang [47] the non-layered structure of the granules is due to the rate-limiting hydrolytic or fermentative stage associated with a high protein content. Grass silage has a high protein content (Table 6.1: 9.5% of DS). Insoluble protein covers the surface of the granule as suspended solids [48] this results in a “puffy” granule as shown in Figure (6.3b). The presence of rod-shaped bacteria (Methanothrix) and cocci-shaped bacteria (Methanosaricina) on the granule surface (Figure 3e & g) is in line with the observation of Gupta and Gupta [21] and Bhatti et al. [49]. Filamentous-shaped bacteria (Methanosaetaceae) were present in abundance on the granule edges (Figure 3h). According to Li et al. [50] high COD removal in the UASB is the result of the high proportion of filamentous bacteria. 6.4.7 Digestate: a valuable product Data on grass digestate is rarely available in the scientific literature. Thus for context, in this study grass digestate is compared to food and manure based digestate. The digestate was separated by manually applied compression into liquid and solid digestate. The high value of total-N in the whole digestate (3.71 %DS; table 6.2) confirmed the view of Nizami and Murphy [8] that grass digestate has a significant role as a soil conditioner and fertilizer. The major plant nutrients (N, P and K) are associated with both the liquid and solid digestate (Table 6.3). Total-P in the liquid digestate (0.31 kg m-3) is slightly below the range of values (0.5 kg m-3 and 1.1 kg m-3) reported from food and manure based digestate respectively [51]. The value of TKN (sum of organic nitrogen and ammonia) which is readily available for plant uptake was found to be high in the solid digestate (18.57 g kg -1 DS). McEniry et al. [52] and Wachendorf et al. [28] used mechanical dehydration with hydraulic pressure and screw press to separate press-cake and juice from grass silage respectively. McEniry et al. [52] obtained 19.9 g N kg-1 DS which matches the results of this study. The fibrous materials associated with neutral detergent fibre (55.2 %DS) facilitates potential to produce value added products. The NDF content of solid digestate could be further increased with good mechanical dehydration. This would allow production of added value products such as insulation boards; it would also facilitate thermal combustion and gasification of dried digestate [28,52]. More efficient pressing could increase the TKN value of the liquid digestate to levels of 7.4 kg m-3 and 4.4 kg m-3 as observed in the digestate of food and manure based digestate [51]. This if the well pressed fibrous element were used for insulation or combustion the liquor digestate could be used as a fertiliser due to the high TKN and total-K values (Table 6.2 & 6.3). 6.5 Conclusions This paper had an objective to optimize a small pilot scale SLBR-UASB. At the start of the experimental period a retention time of 42 days was in place effecting a VS removal rate of 68% and 305 L CH 4 kg-1 VS added [7]. Two aspects were investigated: reduction of retention time to 30 days (experimental system 1); and separation of flows to the UASB and to leach beds through addition of an extra pump to optimize leaching (experimental system 2). Abdul-Sattar Nizami 119 Green Grass: Developing Grass for Sustainable Gaseous Biofuel The optimal system is shown to be experimental system 2. System 2 with the addition of a second pump increased the CH4 production by 10% over system 1 (341 v‟s 310 L CH4 kg-1 VS added). As compared to the end of the commissioning period of the SLBR-UASB described by Nizami et al. [7], system 2 shows an improvement in methane production of 11.8% (343 v‟s 305 L CH 4 kg-1 VS added) and a reduction in size or retention time of 40% (42 days decreased to 30 days retention time). The SLBR-UASB system is shown to be very stable as evidenced by the alkalinity and the pH of the system. The granules also are shown to be healthy with a diameter of 2.55mm and healthy populations of bacteria, in particular a high proportion of filamentous bacteria which are associated with high COD removal. Acknowledgements Research funding was obtained from the Department of Agriculture, Fisheries, and Food (DAFF) Research Stimulas Fund Project “GreenGrass”. We thank: Eddie Appelbe and Erneside Engineering for the fabrication and modification of the reactors: Padraig O‟Kiely and Joe McInerney from Teagasc, Grange, for the supply of grass silage; Richard Kearney from the Cork Institute of Technology (CIT) for help with all maters agriculture. We also thank Beatrice Smyth, Thanasit Thamsiriroj and James Brown for advice, brainstorming sessions, conversations and critiques. References [1] M. Wichern, T. Gehring, K. Fischer, M. Lübken, K. Koch, A. Gronauer and H. Horn. Bioresource Technology., 2009, 100, 1675–1681. [2] R. Rafique, T. G. Poulsen, A. S. Nizami, Z. Z Asam, J. D. Murphy and G. Kiely. Energy., 2010, 35, 4556-4561. [3] V. A. Vavilin, S. V. Rytov, L. Ya. Lokshina, S. G. Pavlostathis and M. A. Barlaz, Biotechnol. Bioeng., 2002, 81, 66–73. [4] S. Ghosh, In: A. A. Antonopoulos, (Ed.), Biotechnological Advances in Processing Municipal Waste for Fuels and Chemicals. Noyes Data Corporation, Park Ridge, NJ, 1987, 303–320. [5] M. Myint, N. Nirmalakhandan and R. E. Speece, Wat. Res., 2007, 41, 323–332. [6] X. Hai-Lou, W. Jing-Yuan and T. Joo-Hwa, Biotechnol. Lett., 2002, 24, 757–761. [7] A. S. Nizami, A. Singh and J. D. Murphy. Energy and Fuels, 2011, 25, 823–834. [8] A. S. Nizami and J. D. Murphy, Renewable and Sustainable Energy Reviews, 2010, 14, 1558–1568. [9] A. Singh, A. S. Nizami, N. E. Korres and J. D. Murphy, Renewable and Sustainable Energy Reviews, 2011, 15, 1567-1574. [10] R. Mulder, T. L. F. M. Vereijken, C. M. T. J. Frijters and S. H. J. Vellinga, Water Sci Technol., 2001, 44, 27–32. [11] A. Lehtomäki, S. Huttunen and J. A. Rintala, Resour. Conserv. Recycl., 2007, 51, 591–609. [12] P. Mähnert, M. Heiermann and B. Linke, Agricultural Engineering International The CIGRE E Journal, 2005, EE 05 010, vol. V11. [13] U. Baserga and K. Egger, Forschungsprogramm Biomasse. Tanikon., 1997, 41 p. [14] KTBL, Gaserträge. Association for Technology and Structures in Agriculture, KTBL, Darmstadt, 2005, 10–16. [15] T. Thamsiriroj and J. D. Murphy, Energy and Fuels, 2010, 24, 4459–4469. Abdul-Sattar Nizami 120 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [16] A. S. Nizami, N. E. Korres and J. D. Murphy, Environmental Science and Technology, 2009, 43, 8496–8508. [17] A. S. Nizami, T. Thamsiriroj, A. Singh and J. D. Murphy, Energy and Fuels, 2010, 24, 4549–4559. [18] H. W. Yu, Z. Samani, A. Hanson and G. Smith, Waste Management. 2002, 22, 1–5. [19] D. G. Cirne, A. Lehtomäki, L. Björnsson and L. L. Blackall, Applied Microbiology, 2007, 103, 516–27. [20] A. S. Bal and N. N. Dhagat, Indian J. Environ. Health, 2001, 43, 1–83. [21] S. K. Gupta and S.K. Gupta, Clean Technol. Environ. Policy, 2005, 7, 203–212. [22] Y. H. Ahn, Y. J. Song, Y. J. Lee and S. Park, Environ Technol., 2002, 23, 889–897. [23] V. Saravanan and T. R, J. Environ Manage., 2006, 81, 1–18. [24] M. K. Tiwari, G. Saumyen, C. S. Harendranath, and T. Shweta, Appl. Microbiol. Biotech., 2006, 71, 145-154. [25] K. Mergaert, B. Vanderhaegen and W. Verstraete, Water Res., 1992, 26, 1025–1033. [26] A. Salter and C. J. Banks, Water Science Technology, 2009, 59, 1053-1060. [27] M. Berglund and P. Borjesson, Biomass and Bioenergy, 2006, 30, 254-266. [28] M. Wachendorf, F. Richter, T. Fricke, R. Graß and R. Neff, Grass and Forage Science, 2009, 64, 132–143. [29] B. Kamm and M. Kamm, Appl. Microbiol. Biotechnol., 2004, 64, 137–145. [30] N. E. Korres, A. Singh, A. S. Nizami and J. D. Murphy, Is grass biomethane a sustainable transport biofuel? Biofuels, Bioproducts and Biorefining, 2010, 4, 310–325. [31] APHA, 20th ed. American Public Health Association, 1998, Washington, DC. [32] L. E. Ripley, W. C. Boyle and J. C. Converse, J. WPCF, 1986, 58, 406–411. [33] J. Kim, C. Park, T. H. Kim, M. Lee, S. Kim, S. W. Kim and J. Lee, J. Biosci. Bioeng., 2003, 95, 271–275. [34] S. Babel, K. Fukushi and B. Sitanrassamee, Water Res., 2004, 38, 2417–2423. [35] E. V. Ramasamy and S. A. Abbasi, Environ.Technol., 2000, 21, 345–349. [36] Z. Z Asam, T. G. Poulsen, A. S. Nizami, R. Rafique, G. Kiely and J. D. Murphy, Applied Energy, 2011, 88, 2013-2018. [37] A. Lehtomäki and L. Björnsson, Environmental Technology, 2006, 27, 209–18. [38] C. L. Sawyer, P. L. McCarty and G. F. Parkin, 4th ed., McGraw-Hill International editions, NY, USA, 1994, 608616. [39] J. Y. Jung, S. M. Lee, P. K. Shin and Y. C. Chung, Biotechnology and Bioprocess Engineering, 2000, 5, 456-459. [40] N. Mahmoud, G. Zeeman, H. Gijzen and G. Lettinga, Bioresour Technol., 2003, 90, 1–9. [41] J. S. Gonzalez, A. Rivera, R. Borja and E. Sanchez, Int Biodeterior Biodegrad., 1998, 41, 127–131. [42] R. P. Singh, S. Kumar and C. S. P. Ojha, Biochem Eng J., 1999, 35–54. [43] E. Alkaya, T. H. Erguder and G. N. Demirer, Eng. Life Sci., 2010, 10, 552–559. [44] H. H. P. Fang, H. K. Chui and Y. Y. Li, Water Sci Technol., 1995, 31, 129–135. [45] L. W. Pol. Hulshoff, J. J. M. van de. Worp, G. Lettinga and W. A. Beverloo, A grownup technology, RAI Halls, Amsterdam, 1986, 89–101. [46] J. H. Wu, W. T. Liu, I. C. Tseng and S.S. Cheng, Microbiology, 2001, 147, 373–382. [47] H. H. P Fang, Water Science & Technology, 2000, 42, 201-208. [48] G. Lettinga, and L. W. Hulshoff-Pol, Wat. Sci. Tech., 1991, 24, 87-107. [49] Z. I. Bhatti, K. Furukawa and M. Fujita, Pure & Appli. Chem., 1997, 69, 2431-2438. [50] J. Y. Li, B. L. Hu, P. Zheng, Q. Mahmood and L. L. Mei, Bioresour. Technol., 2008, 99, 3431–3438. Abdul-Sattar Nizami 121 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [51] M. J. Taylor, A. J. Rollett, D. Tompkins and B. J. Chambers, 15th European Biosolids and Organic Resources Conference, America, 2010, November,15-17. [52] J. McEniry, C. King, J. Finnan and P. O‟Kiely, 2011 Agricultural Research Forum. Tullamore, Co. Offaly, Ireland, 2011, March 14-15, page 133. Abdul-Sattar Nizami 122 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 7: How Can We Optimise Production of Biomethane from Grass Silage? Abdul-Sattar Nizami 123 Green Grass: Developing Grass for Sustainable Gaseous Biofuel How can we optimise production of biomethane from grass silage? Abdul-Sattar Nizamia,b, A Orozcoc, E. Groomc, Jerry D. Murphy,a,b* a Department of Civil and Environmental Engineering, University College Cork, Cork, Ireland b Biofuels Research Group, Environmental Research Institute, University College Cork, Cork, Ireland c School of Chemistry and Chemical Engineering, Queen‟s University Belfast, Northern Ireland Abstract Grass biomethane has been shown to be a sustainable gaseous transport biofuel, with a good energy balance, and significant potential for economic viability. Of issue for the designer is the variation in characteristics of the grass depending on location of source, time of cut and species. Further confusion arises from the biomethane potential tests (BMP) which have a tendency to give varying results. This paper has dual ambitions. One of these is to highlight the various results for biomethane potential that may be obtained from the same grass silage. The results indicated that methane potential from the same grass silage varied from 350 to 493 L CH4 kg-1 VS added for three different BMP procedures. The second ambition is to attempt to compare two distinct digestion systems again using the same grass: a two stage Continuously Stirred Tank Reactor (CSTR); and a Sequentially fed Leach Bed Reactor connected to an Upflow Anaerobic Sludge Blanket (SLBR-UASB). The two engineered systems were designed, fabricated, commissioned and operated at small pilot scale until stable optimal operating conditions were reached. The CSTR system achieved 451 L CH4 kg-1 VS added over a 50 day retention period. The SLBR-UASB achieved 341 L CH4 kg-1 VS added at a 30 day retention time. Keywords: * Biofuel; biomethane; grass silage; UASB; Corresponding author. Tel +353 21 4902286 Fax +353 21 4276648 E-mail address: jerry.murphy@ucc.ie Abdul-Sattar Nizami 124 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 7.1 Introduction Grass biomethane has been shown to be a sustainable gaseous transport biofuel [1]. It has an excellent energy balance; superior to first generation liquid biofuels from temperate climates and similar to tropical biofuel systems [2]. It is also shown to allow economic viability both to the producer and the consumer [3]. Of issue with this paper is the variable nature of grass, the variable data relating to gas production from grass, and the preferred technology to maximise gas production from grass. 7.1.1 Difficulties in comparing digester configurations from scientific literature Data from scientific press on biogas production from grass silage using a variety of systems (dry batch, wet batch, one stage and multi stage systems) are numerous [4-8]. Of issue however is that the characteristic of grass varies with location, soil type, time of harvesting (both date and time of day) and grass species. For example, the level of water soluble carbohydrates is higher in the afternoon than the morning; higher levels of water soluble carbohydrates increases production of methane [9]. The structural and chemical composition of grass changes as the plant matures; grass cut later in the growing season has a higher lignin content and the potential for methane decreases [9]. Furthermore different grass species have different biomethane potentials. Grass may be ensiled using different methodologies: field wilting in continental Europe to obtain 40% dry solids as compared to temperate Europe where pit or clamp silage may be limited to 22% dry solids; baling of silage in plastic as opposed to pit or clamp silage; use or otherwise of silage additives [10]. Thus in comparing biogas or biomethane production of grass silage from different digester designs as assessed from a review of the scientific literature, it is possible to draw inconclusive or incorrect results. To assess the preferable digester configuration there is a need to compare different digester configurations using the same grass silage, cut from the same field, at the same time, ensiled in the same manner. 7.1.2 Comparability of different biomethane potential assays The biomethane potential (BMP) assay has a significant role in digester design, including for economic and management issues. The importance of BMP is highlighted with the significant number of papers in the scientific press relating to biodegradability of different substrates [11]. Despite this there are issues relating to relevance and compatibility of the results. The protocols (ratio of inoculum to substrate; liquid and headspace volumes; pH; headspace pressure; and measurements tools) differ from test to test. It may be stated that the BMP assays can not be assumed to give the exact biomethane rate for a substrate as the results are not comparable [11,12]. 7.1.3 Wet and dry digestion technologies Anaerobic digestion technologies are distinguished and characterized based on the moisture content of the feedstock, number of phases or stages, operating temperature, and method of substrate feeding [13]. Previously wet digestion systems such as the continuously stirred tank reactor (CSTR; Figure 7.1a) were considered problematic for mono-digestion of grass silage, particularly when compared to dry digestion systems. Problems relating to the tendency of grass to float and wrap around agitators/mixers are reported by Thamsiriroj and Murphy [7]. An Abdul-Sattar Nizami 125 Green Grass: Developing Grass for Sustainable Gaseous Biofuel affective mixing mechanism can significantly minimize these problems. Alternatively, dry batch digestion offers a simple robust system. Feedstock is inserted into the batch digester, gas production increases, peaks, decreases and ceases. Half of the substrate is left in the batch as an inoculum for the second cycle of feeding, the batch is filled and the process is repeated [14]. An improvement on this is to couple the dry batch system to an upflow anaerobic sludge blanket (UASB; Figure 7.1b). There is no need to leave half the feedstock in the dry batch reactor as the methane is now produced in the UASB. Thus, the retention time and required digester volume is reduced significantly [9]. 7.1.4 Documented grass biomethane potential Braun et al. [15] stated that the values of methane yield from grass silage vary from 290 to 467 L CH 4 kg-1 volatile solid (VS) added. Stewart et al. [16] reported 310 L CH4 kg-1 VS added from a single stage CSTR using ryegrass plus clover at an organic loading rate of 2.25 kg VS m-3 d-1. A range of methane yields from 320 to 510 L CH4 kg -1 VS added for grass was documented in a review by Nizami and Murphy [13]. A higher range of methane yields (423 627 L CH4 kg-1 VS added) was observed from ryegrass at various stages of maturity by Pouech et al. [17]. Batch lab scale experiments produced methane yields of 0.28 to 0.33 m3 kg-1 VS with 55% methane content [18,19]. A methane yield of 0.165 and 0.27 m3 kg-1 VS added was observed by Yu et al. [20] and Cirne et al. [21] respectively using lab scale leach beds connected with an anaerobic filter. The ranges documented obviously can lead to confusion to a designer of a grass reactor. 7.1.5 Focus of the paper This paper has dual ambitions. One of these is to highlight the various results for biomethane potential that may be obtained from the same sample of grass silage. The second ambition is to attempt to compare two distinct digestion systems (a wet continuous system with a dry batch system connected to a UASB). Two digestion systems were designed and fabricated at small pilot scale. The wet system is a two-stage CSTR with recirculation of liquid digestate, the design and commissioning of which is documented by Thamasiriroj and Murphy [7]. The second system is a sequentially fed leach bed reactor complete with an upflow anaerobic sludge blanket (SLBR-UASB). The design and commissioning of this is documented by Nizami et al. [22]. A series of BMP assays were conducted at small and large scale to determine the upper limit for methane production. The same feedstock (grass silage) was used for both design configurations and the BMP assays. Abdul-Sattar Nizami 126 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Gas flow meter 3 Phase variable speed motor and gearbox Feeding place Gas flow meter Sight glass Sight glass Product front line a Twin mixers Heating element Heating element Digester 1 Digester 2 Pump Second-phase (UASB) First-phase (Sequencing Batch Leach bed) b Leachate tank c d Figure 7.1. Grass silage digestion systems (a) 2-stage CSTR, (b) SLBR-UASB, (c) Large BMP, (d) Small BMP Abdul-Sattar Nizami 127 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 7.2 Methodology 7.2.1 Preparation of grass silage For all experiments, baled grass silage was used as a mono-substrate. The Irish Agricultural Institute („Teagasc‟ Grange Research Centre, Dunsany, Co Meath, Ireland) prepared the silage from homogenous perennial ryegrass. The grass was a first cut (cut at an early mature stage). Before baling, the herbage was wilted in the field for 24 hours. Polythene stretch-film was used to wrap the bales, which were stored for 5 weeks. Small squared bales of 25 kg each were prepared for the experimental work. A mobile macerator (approximated capacity of 10 kg of silage h-1) was used for chopping the grass silage to a particle size of approximately 20 mm. After chopping, the silage was frozen immediately at -15oC and thawed overnight at 6oC before inserting into the 2-stage CSTR and SLBR-UASB systems. Characteristics of the grass silage are reported in Table 7.1. Table 7.1. Characteristics of input grass silage in the study Parameters Measuring Unit Values Lactic acid Ethanol Acetic acid Propionic acid Butyric acid VFA Ammonia WSC pH Protein ME DMD or D-value Silage intake or Palatability PAL NDF FME FME/ME ratio Oil C H N DS VS G kg-1DS G kg-1DS G kg-1DS G kg-1DS G kg-1DS G kg-1DS G kg-1DS G kg-1DS 26.95 11.54 3.93 0.25 1.43 5.61 46.18 49.83 4.3 9.5 10 64 89 821 59 8.2 0.81 3.3 43.03 5.82 1.61 30.66 92.46 % DS MJ kg-1 DS % DS or D-value G kg-1 W0.75 meq kg-1 DS % DS MJ kg-1 DS % DS % DS % DS % DS % % 7.2.2 Experimental set-up The CSTR system: The two stage CSTR system consists of two digesters (312 L effective volume with gas headspace of 160 L each) in series (Figure 7.1a). Feedstock flows between the two digesters via a straight connected pipe at a low level. Some of the liquor digestate from the second digester is recirculated to the inlet of the first Abdul-Sattar Nizami 128 Green Grass: Developing Grass for Sustainable Gaseous Biofuel digester to dilute the dry solid (DS) content of the baled silage (30.7% dry solids). The system is fed on daily basis. The initial commissioning period included for teething problems as documented by Thamsiriroj and Murphy [7]. A scum layer including for undigested grass built-up on the liquid surface of the first digester; this severely hampered the operation of the system [7]. The agitating/mixing system was redesigned to prevent floatation of grass and scum. In the experimental period described here (post commissioning) the organic loading rate (OLR) was increased incrementally from 1.5 to 2.5 kg VS m-3 d-1 at retention time of 50 days at a mesophilic temperature of 37 oC. SLBR-UASB: The SLBR-UASB system consists of 6 leach beds in parallel connected to a UASB reactor (Figure 7.1b). Each leach bed is loaded sequentially with 3.5 kg of grass silage in a 5 day feeding sequence. The overall retention time is therefore 30 days; every 30 days the batch is fully emptied and a new batch is introduced. Initially the leachate tank (capacity of 40 L) is filled with water. The water/leachate is sprinkled over the grass silage in the leach beds via peristaltic pumps. One pump recirculates leachate from the leachate tank to the leach beds in a closed loop. This pumps at a flow rate of 601.8 L d-1 or 100 L d-1 over each leach bed. This is deemed optimal for hydrolysis [14]. The second pump feeds the UASB at a flow rate of 100 L d -1. This flow equates to the maximum upflow velocity of a UASB (0.1 m hr-1). The system is operated at mesophilic temperature (37 oC). Previously, Nizami et al. [22] outlined the commissioning period of the SLBR-UASB system; this involved debugging the system which originally operated at a retention time of 42 days. The system was optimised at a 30 day retention time. This paper describes the results of two periods of 30 days. Micro BMP: These BMP assays were conducted in discontinuous digesters based on the Hohenheim biogas yield test (HBT) principle [23]. For each sample, 500 mg of grass silage was taken as substrate and dried at 60°C over 48 h and grinded to a size of less than 1 mm. In total 6 samples were prepared. The grass silage was digested with 30 mL inoculum in three replicates in a 100 ml glass syringe. The process was carried out at an operating temperature of 37oC. The detail of these experiments was previously described by Nizami et al. [22]. Large BMP assay: In the large BMP assays (Figure 1c), the working volume of the reactor (1.5 L) was filled with 10 g VS at a ratio of 2:1 VS inoculum to VS sample. The reactor was further filled with trace metal solution, sodium hydrogen carbonate (NaHCO3) and stock solution. In each vessel, a magnetic stirrer was used for mixing. The pH was adjusted to 7 using HCl. Silicon grease was applied to stoppers and connectors. The reactor vessels were purged with nitrogen for 2 minutes and were then connected to the gas collecting columns. The operating temperature was 38oC. In total three samples were tested for biomethane potential. Each sample was further repeated to get three replicates of results. Small BMP assay: Small serum bottles were used for the small BMP assays. The working volume of the bottles were 70 ml with 10 g VS (Figure 7.1d). The bottles were filled with substrate and inoculum with a ratio of 2:1 VS of the inoculum to VS of the sample. The inoculum sludge was from a cattle slurry digester in Belfast, Northern Ireland. The inoculum sludge was filtered in a 2 mm sieve prior to the test. NaHCO3 were added along with microminerals to the bottles and the pH was adjusted to 7 using HCL. A manual meat mincer was used to mill the sample to pass a sieve of 3.5 mm. The final volume in the bottles was adjusted with water. N 2 was used to purge the bottles for 2 minutes. The bottles were then sealed with grey butyl rubber stoppers and capped with aluminium crimp seals. Agitation of samples was effected using an incubator shaker at 120 rpm. The operating temperature was Abdul-Sattar Nizami 129 Green Grass: Developing Grass for Sustainable Gaseous Biofuel 38oC. In total five samples were tested for biomethane potential. Each sample was further repeated to get three replicates of results. 7.2.3 Analytical analysis The standard methods [24] were followed to determine the dry solids (DS) and volatile solids (VS) contents. Grass silage was analysed for its feeding value for dairy cattle by the Agri-Food and Biosciences Institute (AFBI), Belfast, Northern Ireland and the Irish Agricultural Institute „Teagasc‟. The grass silage C-H-N ratio were analyzed by elemental analyzer (CE 440 MODEL) using the ultimate analysis method. The pH and gas flow of both digesters of the two stage CTSR were measured every hour using pH sensor probe (Signet 2754-2757 DryLoc pH/ORP) and flow meter (FMA-1600A Omega) respectively digitally controlled by a programmable logic controller (PLC). A portable biogas analyzer (PGD3-IR, Scientific Controls Ltd) was used to measure the biogas composition. Titration based methods were used to measure the alkalinity and total volatile fatty acids (TVFA) [25]. For BMP tests, a gas chromatograph (FID detector, Perkin Elmer Auto System XL, PSS injector, Gas column ZB-FFAP 30m x 0.32 mm) was used. Helium was used as a carrier gas in this gas chromatograph. The Micro BMP assay test was carried out by Burkhart Dietrich in the Department of Agriculture Fisheries and Food (DAFF) funded project “Bio-Grass”. 7.3 Results Detailed results for the two-stage CSTR system are indicated in figure 7.2: OLR and methane production (Figure 7.2a); TVFA and alkalinity (Figure 7.2b); DS and VS content (Figure 7.2c); total ammonia (Figure 7.2d); pH and % CH4 in biogas (Figure 7.2e). The five different grass digestion systems/tests were compared for retention time, % CH4 content in biogas and CH4 production per kg VS added (Table 7.2 & Figure 7.3). 7.3.1 Methane production SLBR-UASB: At stable and optimized performance of the system, the average methane production was 341 L CH 4 kg-1 VS added (Figure 7.3a) at a retention time of 30 days [26]. The results in this paper relate to two periods of 30 days when the system was optimised. CSTR: This paper covers 140 days of stable digestion at organic loading rates (OLR) of 1.5, 2.0 and 2.5 kg VS m -3 d1.    The observed methane yield (Figure 7.2a & 7.3b) was: 50 days at OLR of 1.5 kg VS m_3 d_1 (HRT of 50 days) produced 444 L CH4 kg-1 VS added 50 days at OLR of 2.0 kg VS m_3 d_1 (HRT of 50 days) produced 451 L CH4 kg-1 VS added 40 days at OLR of 2.5 kg VS m_3 d_1 (HRT of 50 days) produced 363 L CH4 kg-1 VS added The system was stopped on day 40 when loaded at an OLR of 2.5 kg VS m-3 d-1 due to mechanical difficulties. The solids concentration had increased to a level such that the agitator/mixing paddle was under powered. The reduction in methane production however would suggest that the optimal performance for the two stage CSTR system produces 451 L CH4 kg-1 VS added at an OLR of 2.0 kg VS m-3 d-1. Micro BMP: The batch experiments produced 350 L CH4 kg-1 VS added for six samples of grass silage (Figure 7.3c). Abdul-Sattar Nizami 130 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Large BMP: The large BMP assays produced 483 to 493 L CH4 kg-1 VS added for the three samples of grass silage (Figure 7.3d). Small BMP: The small BMP assays produced 355 to 419 L kg -1 VS added for the five samples of grass silage (Figure 7.3e). 7.3.2 Alkalinity and total volatile fatty acids CSTR: The total volatile fatty acids (TVFA) increased from 180 to 420 mg L-1 and 100 to 590 mg L-1 in digester 1 & 2 respectively (Figure 7.2b). The alkalinity increased from 900 to 1900 mg L-1 and 960 to 3000 mg L-1 in digester 1 & 2 respectively. SLBR-UASB: The TVFA and alkalinity of the leachate varied from 208 to 440 mg L -1 and 340 to 680 mg L-1 respectively when the system was biologically optimized [26]. 7.3.3 Dry and volatile solid contents CSTR: The average DS and VS content of the feedstock in the digesters were 5.7% and 74.2% of dry solids respectively (Figure 7.2c). The volatile solids were reduced by about 90%. Typically 83% of the total biogas production was produced in digester 1. SLBR-UASB: At optimized conditions, the average DS and VS removal was 73% and 75% respectively [26]. 7.3.4 Ammonia The total ammonia content was only measured in the two stage CSTR system. The total ammonia in g N l-1 increased from 1.49 to 2.35 in digester 1. The total ammonia in digester 2 increased from 1.36 to 2.41 g N l -1. 7.3.5 pH CSTR: In the two-stage CSTR system, the average pH in digester 1 & 2 were 7.64 and 7.73 respectively (Figure 7.2e). SLBR-UASB: In the SLBR-UASB, the average pH of the leachate and UASB was 5.9 and 7.4 respectively when the system was at its optimum performance [26]. Abdul-Sattar Nizami 131 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Days 40 60 80 100 120 140 3 800 2.5 700 600 2 500 1.5 400 1 0.5 OLR 300 Methane Production 200 0 3500 100 0 40 60 80 100 120 140 Alkalinity - Digester 1 600 Alkalinity - Digester 2 500 TVFA - Digester 1 2500 400 TVFA - Digester 2 2000 300 1500 200 1000 100 0 500 % VS c 100 0 90 80 70 60 50 40 30 20 10 0 3 0 20 40 60 80 100 120 140 9 8 7 6 5 Total Ammonia (g N L-1) 4 % VS - Digester 1 20 40 60 80 % VS - Digester 2 3 % DS - Digester 1 2 % DS - Digester 2 1 100 120 0 3 140 2.5 d TVFA (mg L-1) Alkalinity (mg L-1) 3000 20 2.5 2 2 1.5 1.5 1 1 Total Ammonia - Digester 1 0.5 % DS b Methane Production (L d-1) 20 0.5 Total Ammonia (g N L-1) a OLR (kg VS m-3 d-1) 0 Total Ammonia - Digester 2 0 0 8.5 0 20 40 60 80 100 120 140 59 8 57 pH 7.5 55 53 7 51 6.5 pH - Digester 1 pH - Digester 2 % CH4 - Digester 1 % CH4 - Digester 2 6 49 47 45 5.5 0 20 40 60 80 100 120 140 Days Figure 7.2. Results of the experimental 2-stage CSTR operation Abdul-Sattar Nizami 132 % CH4 in Biogas e Green Grass: Developing Grass for Sustainable Gaseous Biofuel Days 40 60 80 100 120 140 400 80 350 75 300 70 250 65 200 CH4 production 150 60 % CH4 100 55 CH4 Production (L CH4 kg-1VS added) 0 b % CH4 in Biogas a 20 20 40 60 80 100 CH4 production - Digester 1 120 &2 % CH4 - Digester 1 % CH4 - Digester 2 750 65 60 650 55 550 50 450 45 350 40 250 35 % CH4 in Biogas CH4 Production (L CH4 kg-1VS added) 0 30 150 0 50 100 150 50 CH4 production-Sample 1 CH4 production-Sample 2 CH4 production-Sample 3 CH4 production-Sample 4 CH4 production-Sample 5 CH4 production-Sample 6 % CH4-Sample 1 % CH4-Sample 2 % CH4-Sample 3 % CH4-Sample 4 % CH4-Sample 5 % CH4-Sample 6 270 220 170 120 70 45 40 35 30 25 20 20 CH4 Production (L CH4 kg-1VS added) 0 d % CH4 in Biogas 320 10 20 30 40 80 500 70 60 400 300 200 CH4 production - Sample 1 50 CH4 production - Sample 2 40 CH4 production - Sample 3 30 % CH4 - Sample 1 % CH4 in Biogas c CH4 Production (L CH4 kg-1VS added) 55 370 20 % CH4 - Sample 2 100 10 % CH4 - Sample 3 0 0 0 5 10 15 20 25 30 55 400 350 45 300 250 CH4 production-Sample CH4 production-Sample CH4 production-Sample CH4 production-Sample CH4 production-Sample % CH4-Sample 1 % CH4-Sample 2 % CH4-Sample 3 % CH4-Sample 4 % CH4-Sample 5 200 150 100 50 35 1 2 3 4 5 25 % CH4 in Biogas e CH4 Production (L CH4 kg-1VS added) 450 15 0 5 0 5 10 15 20 25 Days Figure 7.3. Methane production and methane content of biogas (a) SLBR-UASB (b) 2-stage CSTR, (c) Micro BMP, (d) Large BMP, (e) Small BMP Abdul-Sattar Nizami 133 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Table 7.2. Summary of the results from three various grass digestion systems SLBRCSTR Micro BMP UASB HRT (Days) 30 50 35 CH4 content (% CH4 in Biogas) 71 52 51 CH4 production (L CH4 kg-1 VS added) 341 451 350 Small BMP 22 54 355-419 Large BMP 26 70 483-493 7.4 Discussion 7.4.1 Biological process The average pH of the two stage CSTR system was 7.64 in digester 1 and 7.73 in digester 2 (Figure 7.2e). The SLBRUASB (stable and optimized) had an average pH of 5.9 in the leachate and 7.4 in the UASB [26]. This range of pH in both the CSTR and the UASB is within the optimum pH range (6.3 to 7.8) for anaerobic digestion [27,28]. Both systems operated at 37oC, which is considered optimum for mesophilic anaerobic digestion [29,30]. 7.4.2 Organic loading rate The reduction in methane rate during the last 40 days of the two stage CSTR operation suggests that the system had surpassed its optimum (Figure 7.2a). Thus it is suggested that the optimal loading rate was of the order of 2 kg VS m_3 d_1 at a HRT of 50 days. Nordberg et al. [31] suggested that the optimum OLR for alfalfa silage with and without recirculation was 2.25 to 3 kg of VS m-3 d-1. The lower methane production beyond optimal OLR (Figure 7.2a) is also in line with the observations of Jarvis et al. [32] during the operation of a CSTR. In the SLBR-UASB system, the UASB efficiency (converting chemical oxygen demand (COD) into methane) peaked at 93% when the system was optimised. This efficiency was achieved with the recommended constant upflow velocity (0.1 m h-1) and increased influent COD concentration associated with lower retention times and higher leaching rates [26]. Mahmoud et al. [33] observed increased UASB efficiency at constant upflow velocity of 0.1 m h -1 ensured good mixing in the UASB without washout. Lower velocities result in poor mixing; higher velocities result in turbulence in the sludge bed and washout of granules. The UASB efficiency is also increased by increasing the influent concentration of COD [33]. 7.4.3 Methane production The same grass silage was used in all 5 grass digestion systems. The BMP assays (small, large and micro) suggested an upper limit on methane production of 419, 493 and 350 L CH4 kg-1 VS added respectively (Table 7.2). The methane production of the two stage CSTR (451 L CH4 kg-1 VS added) and the SLBR-UASB (341 L CH4 kg-1 VS added) systems were within the range of BMP assays (Figure 7.3). The two stage CSTR produced higher methane levels than the SLBR-UASB system but at a longer retention time (50 days versus 30 days). This is in line with the findings of Lehtomäki et al. [5] who digested grass in one and two-stage leach bed processes and in a CSTR system. They suggested that the higher methane yield of the CSTR is due to better microbial adaptation to the substrate [5]. Abdul-Sattar Nizami 134 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Mähnert et al. [4] reported the biogas yield from a single stage CSTR to be 80 – 85% of the yield from a batch experiment. On the other hand, Stewart et al. [16] reported a similar biogas yield from a CSTR and a batch experiment. Asim et al. [34] suggested an upper limit of 361 L CH4 kg-1 VS from grass silage using a small batch experiment. The methane production from the two stage CSTR is greater than that of the micro BMP assay test and the small BMP assay; it is 90% of the value of the large BMP assay. The methane production from the SLBR-UASB is of the same order as that of the micro BMP assay but less than the two BMP assays. 7.4.4 Retention times The two stage CSTR system was operated at a HRT of 50 days for different organic loading rates (1.5, 2.0 and 2.5 kg VS m_3 d_1). The HRT was adjusted by the recycling rate of liquor between the two tanks. The system produced 451 L CH4 kg-1 VS added with 90% VS removal at an OLR of 2 kg VS m _3 d_1 (Figure 7.2a). The solid retention time (SRT) was found to be greater than the HRT; this is postulated as the primary reason for the high levels of methane production [35]. The SLBR-UASB system with a solid retention time of 30 days producing 341 L CH 4 kg-1 VS added with VS removal of 75% [26]. Lehtomäki and Björnsson [36] reported a methane yield of 0.39 m3 kg-1 VS added at a retention time of 50 days using a leach bed reactor coupled with an anaerobic filter. The digester produced 85% of the methane in the first 30 days (0.312 m3 kg-1 VS added in 30 days). 7.4.5 Methane content In the two stage CSTR system, the average methane content in digester 1 & 2 was 52.62 and 52.25% respectively (Figure 7.2e). In the SLBR-UASB, the average methane content was 70.6%. The higher level in the SLBR-UASB system was due to the passing of biogas through a water filter which allowed dissolution of some of the carbon dioxide [26]. Jung et al. [37], observed a methane content of 80-90% after dissolving biogas into a weak effluent wastewater using a two stage anaerobic process at pilot scale treating pig slurry. Sawyer et al. [38] documented the decrease in the carbon dioxide content of biogas through dissolving in leachate or water. Thus, in a similar fashion the methane content of the two stage CSTR can be increased by adding water a filter on the biogas line. 7.4.6 Process inhibition In the two stage CSTR system, the total ammonia content increased due to the high ammonia content of the grass silage. The ammonia content must be a concern in particular in mono-digestion of grass silage [7]. The maximum total ammonia of 2.35 and 2.41 g N l-1 was observed in digester 1 & 2 respectively at a maximum OLR of 2.5 kg of VS m-3 d-1 (Figure 7.2d); these levels are lower than the inhibition level (4.7 g N l -1) for grass digestion as suggested by Marin- Perez et al. [39]. The relatively high level of ammonia in the two stage CSTR system is in line with the observation of Marin-Perez et al. [39] that ammonia content increases when treating grass at high OLR. Recirculation of digestate increases ammonia within the system [31]. Thamsiriroj and Murphy [7] suggested that the Abdul-Sattar Nizami 135 Green Grass: Developing Grass for Sustainable Gaseous Biofuel recirculation rate can be a useful management tool for a two stage CSTR system allowing some control in relative levels (between the two digesters) of ammonia, VFA and methane production. High OLR is associated with high ammonia concentration, high VFA (Figure 7.2b) and decreased methane production (Figure 7.2a). 7.4.7 Digestate and value added products The digestate from the CSTR has an average DS content of 5.6% (Figure 7.2c) from a feedstock with a dry solids content of over 30%. The digestate is observed to be a green luminous liquor rich in protein and nutrients and low in fibres. This digestate may be used as a biofertilizer due to the high level of readily available ammonia-N [1,9]. Additionally, it may be used as animal fodder due to the high protein content; alternatively it may be used as a source of chemicals for pharmaceutics [40]. The digestate from the SLBR-UASB system has a DS content of 10.18% [26]. The digestate is more fibrous in nature; this has been suggested to be more suited to insulation board or fibres for combustion and gasification [26]. 7.4.8 Post methanation potential Thamsiriroj and Murphy [7] suggest that full destruction of volatile solids results in approximately 500 L CH 4 kg-1 VS added. The two stage CSTR effected 451 L CH4 kg-1 VS added, equivalent to 90% destruction of volatiles. The average DS and VS content of the digestate was 5.6% and 74% of dry solids (or 4.14% of digestate) respectively (Figure 7.2c). There is little potential for post-methanation. The SLBR-UASB system effected 75% destruction of volatile solids. The digestate has DS and VS value of 10.18% and 81.78% of dry solids (or 8.32% of digestate) respectively [26]. The methane potential of the digestate is higher for the SLBR-UASB system. However the fibrous nature of this digestate has potential for further use. 7.5 Conclusion This paper has dual ambitions. One of these is to highlight the various results for biomethane potential that may be obtained from the same sample of grass silage. The second ambition is to attempt to compare two distinct digestion systems (a wet continuous system with a dry batch system connected to a UASB).  The methane production of the large BMP test is the largest of the five assessed at 493 L CH 4 kg-1 VS added. The two stage CSTR (451 L CH4 kg-1 VS added) is greater than that of the small BMP test (419 L CH4 kg-1 VS added) which is greater than the micro BMP assay (350 L CH4 kg-1 VS added). The least value is obtained  by the SLBR-UASB (341 L CH4 kg-1 VS added). The two engineered systems have different merits. The two stage CSTR system effects 90% destruction of volatiles (451 L CH4 kg-1 VS added) at a retention time of 50 days. The SLBR-UASB however effects 75% destruction of volatiles with a UASB efficiency of 93% generating 341 L CH 4 kg-1 VS added at a retention time of 30 days. Abdul-Sattar Nizami 136 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Acknowledgements Research funding was obtained from the Department of Agriculture, Fisheries, and Food (DAFF) Research Stimulus Funded Project “GreenGrass”. We acknowledge the input of Burkhart Dietrich and the DAFF Research Stimulus Funded Project “Bio-Grass” on the micro assay test. We thank: Eddie Appelbe and Erneside Engineering for the fabrication and modification of the reactors: Padraig O‟Kiely and Joe McInerney from Teagasc, Grange, for the supply of grass silage; Richard Kearney from the Cork Institute of Technology (CIT) for help with all maters agriculture. We also thank Beatrice Smyth, Thanasit Thamsiriroj and James Brown for advice, brainstorming sessions, conversations and critiques. References [1] N. E. Korres, A. Singh, A. S. Nizami and J. D. Murphy, Biofuels, Bioproducts and Biorefining, 2010, 4, 310– 325. [2] B. Smyth, J. D. Murphy and C. O'Brien, Renewable and Sustainable Energy Reviews, 2009, 13, 2349-2360. [3] B. Smyth, H. Smyth and J. D. Murphy, Biofuels, Bioproducts, Biorefinery, 2010, 4, 519–537. [4] P. Mähnert, M. Heiermann and B. Linke, Agricultural Engineering International The CIGRE E Journal, 2005, EE 05 010, vol. V11. [5] A. Lehtomäki, S. Huttunen, T. M. Lehtinen, J. A. Rintala, Bioresour. Technol., 2008, 99, 3267–3278. [6] G. Liu, R. Zhang, M. Hamed, El-Mashad, W. Withrow, R. Dong, Journal of China Agricultural University, 2006, 11, 111–5. [7] T. Thamsiriroj and J. D. Murphy, Energy and Fuels, 2010, 24, 4459–4469. [8] R. Rafique, T. G. Poulsen, A. S. Nizami, Z. Z Asam, J. D. Murphy and G. Kiely. Energy., 2010, 35, 4556-4561. [9] A. S. Nizami, N. E. Korres and J. D. Murphy, Environmental Science and Technology, 2009, 43, 8496–8508. [10] A. Singh, A. S. Nizami, N. E. Korres and J. D. Murphy, Renewable and Sustainable Energy Reviews, 2011, 15, 1567-1574. [11] I. Angelidaki, M. Alves, D. Bolzonella, L. Borzacconi, J. L. Campos, A. J. Guwi and J. B. van Lier, Water Sci Technol., 2009, 59, 927–934. [12] W. Muller, I. Frommert and R. Jorg, Rev. Environ. Sci. Biotechnol., 2004, 3, 141–158. [13] A. S. Nizami and J. D. Murphy, Renewable and Sustainable Energy Reviews, 2010, 14, 1558–1568. [14] A. S. Nizami, T. Thamsiriroj, A. Singh and J. D. Murphy, Energy and Fuels, 2010, 24, 4549–4559. [15] R. Braun, P. Weiland, J. D. Murphy and A. Wellinger, International Energy Agency (IEA) Task 37, 2011. Available In: http://www.iea-biogas.net/ [16] D. J. Stewart, M. J. Bogue, D. M. Badger, N. Z. J. Sci., 1984, 27, 285–294. [17] P. Pouech, H. Fruteau and H. Bewa, Proceeding of the 10th European Conference on Biomass for Energy and Industry; CARMEN:Wurzburg, Germany, 1998. [18] U. Baserga and K. Egger, Forschungsprogramm Biomasse. Tanikon., 1997, 41 p. [19] KTBL, Gaserträge. Association for Technology and Structures in Agriculture, KTBL, Darmstadt, 2005, 10–16. Abdul-Sattar Nizami 137 Green Grass: Developing Grass for Sustainable Gaseous Biofuel [20] H. W. Yu, Z. Samani, A. Hanson and G. Smith, Waste Management. 2002, 22, 1–5. [21] D. G. Cirne, A. Lehtomäki, L. Björnsson and L. L. Blackall, Applied Microbiology, 2007, 103, 516–27. [22] A. S. Nizami, A. Singh and J. D. Murphy. Energy and Fuels, 2011, 25, 823–834. [23] D. Helffrich, H. Oechsner, Landtechnik, 2003, 58, 148–149. [24] APHA, 20th ed. American Public Health Association, 1998, Washington, DC. [25] L. E. Ripley, W. C. Boyle and J. C. Converse, J. WPCF, 1986, 58, 406–411. [26] A. S. Nizami and J. D. Murphy, Energy and environmental sciences, Submitted March, 2011. [27] M. K. Tiwari, G. Saumyen, C. S. Harendranath, and T. Shweta, Appl. Microbiol. Biotech., 2006, 71, 145-154. [28] J. Kim, C. Park, T. H. Kim, M. Lee, S. Kim, S. W. Kim and J. Lee, J. Biosci. Bioeng., 2003, 95, 271–275. [29] R. J. Zoetemeyer, P. Arnoldy, A. Cohen and C. Boelhouwer, Water Res., 1982, 16, 313–321. [30] W. Choorit and P. Wisarnwan, Electronic Journal of Biotechnology. 2007, 10, 376-385. [31] A. Nordberg, A. Jarvis, B. Stenberg, B. Mathisen and B. H. Svensson, Bioresour. Technol., 2007, 98, 104–111. [32] A. Jarvis, A. Nordberg, T. Jarlsvik, B. Mathisen, B. H. Svensson, Biomass Bioenergy, 1997, 12, 453–460. [33] N. Mahmoud, G. Zeeman, H. Gijzen and G. Lettinga, Bioresour Technol., 2003, 90, 1–9. [34] Z. Z Asam, T. G. Poulsen, A. S. Nizami, R. Rafique, G. Kiely and J. D. Murphy, Applied Energy, 2011, 88, 2013-2018. [35] T. Thamsiriroj and J. D. Murphy, Bioresour. Technol., 2011, 102, 948-959. [36] A. Lehtomäki and L. Björnsson, Environmental Technology, 2006, 27, 209–18. [37] J. Y. Jung, S. M. Lee, P. K. Shin and Y. C. Chung, Biotechnology and Bioprocess Engineering, 2000, 5, 456459. [38] C. L. Sawyer, P. L. McCarty and G. F. Parkin, 4th ed., McGraw-Hill International editions, NY, USA, 1994, 608-616. [39] C. Marin- Perez, M. Lebuhn and A. Gronauer, Biogas J., 2009, 4, 72–75. [40] A. Salter and C. J. Banks, Water Science Technology, 2009, 59, 1053-1060. Abdul-Sattar Nizami 138 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Chapter 8: Recommendations Abdul-Sattar Nizami 139 Green Grass: Developing Grass for Sustainable Gaseous Biofuel Recommendations The recommendations of the thesis are: 1) The procedures/technologies adopted at the laboratory/pilot scale must be assessed for upscale potential to allow improved systems at commercial and industrial-scale. 2) The potential for co-digesting grass silage with other substrates (such as slurry) can lead to benefits such as: increased buffering capacity and microbiological stability. Hydrolysis through use of slurry should be investigated. 3) Optimisation of leaching in the leach bed through thermal and enzymatic pretreatment of the grass silage should be investigated. 4) Value added products (e.g. fire logs, insulation board, fertiliser) from digestate may also help in reducing GHG emissions of the whole system and improve financial stability. 5) CO2 removal during upgrading of biogas into enriched biomethane adds to greenhouse gas (GHG) emissions. This CO2 content of the produced biogas can be minimized by dissolution in water; dissolved CO2 forms carbonic acid. This carbonated water can then be used as a leaching medium which can improve dissolution of hemicellulose. Abdul-Sattar Nizami 140

RELATED PAPERS

RELATED TOPICS

Log In

Green Grass: Developing Grass for Sustainable Gaseous Biofuel.pdf

Green Grass: Developing Grass for Sustainable Gaseous Biofuel.pdf

Related Papers

RELATED PAPERS

RELATED TOPICS