|UFDC Home||myUFDC Home | Help|
1 AN INTEGRATED TECHNOLOGY FOR RECOVERY OF ENERGY, NUTRIENTS AND CLEAN WATER FROM CELLULOSIC ETHANOL STILLAGE By GAYATHRI RAM MOHAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFI LLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Gayathri Ram Mohan
3 To my mother and my grandmother who have always been supportive of my dreams
4 ACKNOW LEDGMENTS First and foremost I would like to thank Lord Almighty for bestowing his blessings on my family and me. Secondly, I would like to extend my deepest appreciation to my advisor and chairperson Dr. Pratap Pullammanappallil for giving me such a w onderful opportunity to explore the scientific world. I would like to sincerely thank Dr. Ben Koopman, Dr. Spyros Svoronos and Dr. Arthur Teixeira for inspiring me to grow and sharpen my skills as a professional I would like to thank Dr. Lonnie Ingram for serving on my committee and providing me with resources from his lab. I would like to thank Dr. K.T Shanmugam for his valuable guidance. I would also like to thank Dr. Amir Varshovi for sharing his knowledge and experience with me. I would like to thank Mike Mullinix and Tian Zhouli for rendering technical support and being a good friend when I needed one I would like to thank my friends and lab mates Samriddhi Buxy, Mandu Inyang, Patrick Dube, David Palubin, Paul Ro senberg er Abhay Koppar, Cesar Moreir a, Jaime Chavez and Marco Pazmino for their support and encouragement I would like to thank Steve Feagle, Rudy, Paul Lane and Billy Duckworth for their contribution in making this research project a success. Last but not the least, I would like to thank my parents, sisters, brothers and my extend a special thanks to my fianc for his love and encouragement throughout my studies.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 Background ................................ ................................ ................................ ............. 15 Bioethanol from Lignocellulosic Feedstock ................................ ....................... 16 Conventional Techniques for Handling Distillery W astewaters ......................... 18 Water Usage in Bioethanol Plants ................................ ................................ .... 19 Problem S tatement ................................ ................................ ................................ 21 Project Goal ................................ ................................ ................................ ............ 21 Research Objectives ................................ ................................ ............................... 22 Research A ppr oach ................................ ................................ ................................ 23 Dissertation Organization ................................ ................................ ........................ 25 2 DIGESTION OF CELLULOSIC ETHANOL STILLAGE IN AN ANAEROBIC FLUIDIZED BED REACTOR ................................ ................................ .................. 29 Summary ................................ ................................ ................................ ................ 29 Background ................................ ................................ ................................ ............. 29 Materials and Methods ................................ ................................ ............................ 34 Feedstock ................................ ................................ ................................ ......... 34 Reactor Design ................................ ................................ ................................ 34 Parameters Monitored and Analyzed ................................ ............................... 36 Reactor Operation ................................ ................................ ............................ 39 Results ................................ ................................ ................................ .................... 40 Characterization of Stillage ................................ ................................ ............... 40 Biogassification of Stillage ................................ ................................ ................ 41 Discussion ................................ ................................ ................................ .............. 43 Closing Remarks ................................ ................................ ................................ .... 52 3 RECOVERY OF NUTRIENTS FROM CELLULOSIC ETHANOL STILLAGE BY PRECIPITATION AS STRUVITE ................................ ................................ ............ 62 Summary ................................ ................................ ................................ ................ 62 Background ................................ ................................ ................................ ............. 63 Commercial Technologies for Struvite Recovery ................................ .............. 66
6 Research Objectives ................................ ................................ ............................... 68 Materials and Methods ................................ ................................ ............................ 68 Reactor Design ................................ ................................ ................................ 69 Reactor Operation ................................ ................................ ............................ 69 Results ................................ ................................ ................................ .................... 71 Characterization of Cellulosic Ethanol Stillage ................................ ................. 71 Struvite Sludge from Raw Stillage ................................ ................................ .... 72 Struvite Sludge from Anaerobically Digested Stillage ................................ ....... 73 Discussion ................................ ................................ ................................ .............. 77 Agronomic Efficiency of Processed Sludge Containing Struvite ....................... 77 Amount of Nitrogen and Phosphorus in Leachate ................................ ............ 79 Closing Remarks ................................ ................................ ................................ .... 80 4 PHOTOCATALYTIC TREATMENT FOR FINAL POLISHING OF WASTEWATER FROM A CELLULOSIC ETHANOL PROCESS; EFFECT OF INITIAL COD, PH, DEPTH OF STILLAGE ................................ .............................. 85 Summary ................................ ................................ ................................ ................ 85 Background ................................ ................................ ................................ ............. 85 Materials and Methods ................................ ................................ ............................ 87 Wastewater ................................ ................................ ................................ ...... 87 Titanium dioxide Coated S heets ................................ ................................ ....... 87 UV lamps ................................ ................................ ................................ .......... 87 Photoreactor Assembly ................................ ................................ .................... 87 Results ................................ ................................ ................................ .................... 89 Effect of Initial sCOD Concentration ................................ ................................ 89 sCOD Reduction ................................ ................................ .............................. 91 Extent of Decolorization ................................ ................................ ................... 92 sCOD Absorbance Correlation ................................ ................................ ......... 92 Effect of Initial pH ................................ ................................ ............................. 93 Discussion ................................ ................................ ................................ .............. 94 Closing Remarks ................................ ................................ ................................ .... 98 5 COMPARISON OF PHOTOCATALYTIC EFFICIENCY OF UNDOPED AND SILVER DOPED TIO 2 UNDER UV A, UV C AND SUNLIGHT FOR COD AND COLOR REMOVAL FROM CELLULOSIC ETHANOL STILLAGE ........................ 106 Summary ................................ ................................ ................................ .............. 106 Background ................................ ................................ ................................ ........... 106 Materials and Methods ................................ ................................ .......................... 109 Results ................................ ................................ ................................ .................. 109 Absorbance Profiles ................................ ................................ ....................... 109 UV A 350 nm Light S ource ................................ ................................ .................. 109 UV C 180 nm Light Source ................................ ................................ .................. 110 Solar Radiation ................................ ................................ ............................... 110 sCOD Removal ................................ ................................ .............................. 111 sCOD Absorbance Correlation ................................ ................................ ....... 111
7 Discussion ................................ ................................ ................................ ............ 112 Use of Immobilized TiO 2 Coated Sheets vs. Powdered TiO 2 Catalysts .......... 112 Wastewater Characteristics: Dilution v s. Aerobic ally Treated AD Stillage ...... 112 Effect of Different UV Sources on TiO2 Photocatalysis ................................ .. 115 Effect of Solar Radiation on TiO 2 Photocatalysis ................................ ............ 116 Effect of Silver Doping on TiO 2 Photocatalysis ................................ ............... 118 Closing Remarks ................................ ................................ ................................ .. 119 6 CONCLUSIONS AND FU TUR E WORK ................................ ............................... 126 Conclusions ................................ ................................ ................................ .......... 126 Future Work ................................ ................................ ................................ .......... 130 LIST OF REFERENCES ................................ ................................ ............................. 132 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 142
8 LIST OF TABLES Table page 1 1 Composition of Lignocellulosic Feedstocks (% dry weight) (Limayem and Ricke 2012) ................................ ................................ ................................ ........ 27 1 2 Water Consumption in Bioethanol Plants (Aden and Foust 2009) ...................... 27 2 1 Characterization of raw stillage ................................ ................................ ........... 53 2 2 S teady state values of parameters monitored ................................ .................... 53 3 1 Characterization of Feedstock ................................ ................................ ............ 81 3 2 Characterization of processed stillage fo r use as soil amendment* .................... 81 3 3 Nutrient Recovery from Anaerobically Digested Stillage ................................ .... 82 3 4 Characterization of Processed Stil lage from AD stillage+1% (wet w/v) stillage solids ................................ ................................ ................................ .................. 82 3 5 Characterization of Processed Stillage from AD stillage+10% (wet w/v) stillage solids ................................ ................................ ................................ ...... 82
9 LIST OF FI GURES Figure page 1 1 Schematic of Bioethanol Process Using Sugarcane Bagasse as Feedstock (Nieves, Geddes et al. 2011) ................................ ................................ .............. 28 2 1 Schematic representation of an anaerobic fluidized bed reactor ........................ 54 2 2 Temperature Profile of digester liquid during a 24 hour period. .......................... 55 2 3 Feed Flow rate, Organic loading rate and influent Soluble COD concentrations of digested stillage during entire operation of AFBR .................. 56 2 4 Parameters monitored during the entire operation of fluidized bed reactor (i) Methane Percent in Biogas (ii) Methane Yield (iii) Methane Production Rate, (iv) pH ................................ ................................ ................................ ................. 57 2 5 Concentration of glucose, xylose and arabinose (sugars) in digested stillage .... 58 2 6 Organic acids concentrations in digested stillage effluent ................................ .. 59 2 7 COD balance on digested stillage effluent ................................ .......................... 60 2 8 Succinic acid, Levulinic acid and Furfural compounds (Furfural and Hydroxymethyl Furfural, HMF) concentrations in feed and digested stillage. ..... 61 3 1 Schematic diagram, photograph and working of a Sequential Batch Reactor .... 83 3 2 Settleability of processed stillage from anaerobically digested stillage eff luent seeded with stillage solids. ................................ ................................ ................. 84 4 1 Absorbance and sCOD Profiles of TiO2 treated Diluted Anaerobic Stillage ....... 99 4 2 sCOD an d absorbance profiles for duplicate (Trail 1 and 2) 3 mm depth experiments ................................ ................................ ................................ ...... 100 4 3 sCOD removal in various 3mm depth experiments. (i) Control 1: TiO2 in dark, (ii) UV exposure in the absence o f TiO2 sheets, (iii) Experiment 1: No pH adjustment in aerobically treated samples, (ii) Experiment 2: Initial pH adjusted to 7 in aerobically treated samples ................................ ..................... 101 4 4 Absorbance of samples fro m 1, 3 and 5 mm depth experiments plotted against time of exposure of samples ................................ ................................ 102 4 5 Extent of decolorization of photocatalytically treated stillage (3 mm depth) ...... 103 4 6 sCOD Absorbance Correlation ................................ ................................ ...... 104
10 4 7 pH profile in treated stillage. (i) Experiment 1: Without pH adjustment, (ii) Experiment 2: With initial pH ad justed to neutral. ................................ ............. 105 5 1 Decolorization profiles of photocatalytically treated Cellulosic ethanol stillage 120 5 2 Percent dec olorization and percent sCOD removal from various photocatalytic experiments ................................ ................................ ............... 121 5 3 sCOD Absorbance correlation for experiments 2 and 3 ................................ ... 122 5 4 1st order rate constants from various photocatalytic experiments .................... 123 5 5 Decolorization profile of photocatalytically treated diluted anaerobic stillage (DAS) ................................ ................................ ................................ ................ 124 5 6 Percent Color and sCOD removal from photocatalytically treated diluted anaerobic stillage (DAS) ................................ ................................ ................... 125 6 1 Mass and Energy balance of cellu losic ethanol stillage treatment .................... 131
11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy AN INTEGRATED TECHNOLOGY FOR RECOVERY OF ENERGY, NUTRIENTS AND CLEAN WATER FROM CELLULOSIC ETHANOL STILLAGE By Gayathri Ram Mohan December 2012 Chair: Pratap Pullammanappallil Major: Agricultural and Biological Engineering T he downstream section of a lignocellulosic bio chemical ethanol process produces a waste stream following the distillation of ethanol from the fermented sy rup. This distillery wastewater, referred to as stillage, is a dark colored, acidic, nutrient rich liquid that presents signifi cant problems with its disposal. Unlike conventional distillery wastewater, this stillage may contain recalcitrant products of biomass pretreatment With US EPA tightening the industrial effluent discharge standards accompanied by decreasing land availabil ity for waste discharge, more intens ive treatment approaches must be applied to overcome stillage disposal issues. In this dissertation, an integrated system was developed to treat stillage produced from a sugarcane bagasse feedstock process. The sugarca ne bagasse is first pretreated in a dilute phosphoric acid process followed by enzymatic saccharification and then fermentation. The goals of the integrated treatment system were to recover green energy i n the form of biogas, recover nutrients, specifical ly phosphate via struvite precipitation and finally polish the effluent using advanced oxidation process to recover clean water for possible reuse in the ethanol plant. The aim
12 of this integrated system wa s to reduce the carbon and water foot print associ ated with the ethanol production process Thermophilic anaerobic digestion (55 C) of cellulosic ethanol stillage was investigated using a 5.5 L laboratory scale anaerobic fluidized bed reactor (AFBR). Batches of coarsely separated bagasse stillage (0.425 mm sieve) obtained from the Biofuels pilot plant at the University of Florida with a high and var iable soluble chemical oxygen demand ( COD ) ranging from 11.5 to 63 g sCOD/L was used as feed. The AFBR was operated for 100 days. The methane potential achi eved was 12.9 L CH 4 (L Stillage) 1 0.288 L CH 4 (g sCOD loaded ) 1 Soluble organic matter removal efficiency of 93% was achieved at a hydraulic retention time (HRT) of 7.3 days and an orga nic loading rate of 6.54 g sCOD L 1 d 1 The effluent soluble COD was 3.5 g sCOD/L. Organic acids and sugars were monitored periodically during the entire run. No external nutrients (N and P) were added during the operation of the fluidized bed reactor. Typical dilute acid pretreatment products like furfurals and levulinic acids found in the stillage were degraded in the AFBR. Upon operating the AFBR at a loading rate of 8.8 g sCOD L 1 d 1 volatile organic acids (especially propionic acid) accumulated and methane production was suppressed. These organic acids were consume d within a couple of days after turning off the feed. Thereafter, the AFBR was operated at normal operating conditions (6.54 g sCOD L 1 d 1 ) A novel sequential batch reactor (SBR) technology (with fill, react, settle, decant and drain phases) was us ed to precipitate and recover phosphate from anaerobically digested stillage, in the form of the slow release fertilizer struvite. The use of coarsely separated stillage solids as seed in the SBR to promote settling of struvite was
13 investigated. Results showed that unseeded trials produced a large amount of unsettled fine mineral precipitates, while using 1% (wet w/v) stillage solids as seed material improved settleability of processed sludge. More than 95% of settling occurred within the first 15 minutes of undisturbed settling following the 30 min reaction time. About 99.9% and 56% of orthophosphate phosphorous and ammonia nitrogen respectively, were recovered in the sludge. Seeding also increased the yield of net amount of settled struvite precipitate contain ing sludge by 63%. The struvite containing settled sludge was also tested for its agronomic applicability and nutrient leachability. Results showed markedly improved nutrient uptake by plants and reduced N and P levels in leachate on application of settled sludge as soil amendment for cultivation of sweet sorghum. Anaerobically digested and struvite precipitated effluent was then subjected to photocatalytic oxidation to decolorize and reduce organic matter further. Photocatalytic oxidation was car ried out using TiO 2 coated reusable sheets (TiO 2 loading : 3 mg/cm 2 ) Batch e xperiments were performed to study the effect of initial so luble chemical oxygen demand (sCOD), depth, pH, light source and silver doping of TiO 2 catalyst on decolor i zation and organic matter removal. Organic concentration of 35 00 mg sCOD/L in anaerobically digested effluent was an obstacle for efficient photocatalytic treatment. It was shown that dilution of this effluent did not enhance photocatalytic degradation. However, af ter reducing the soluble COD by biological aerobic t reatment organic matter removal, decolorization was enhanced. The extent of decolorization was 91.3%, 91.5% and 77% after 22, 55 and 80.5 hours for 1, 3 and 5 mm depths respectively at neutral pH The final treated s COD was 8 0 m g/L. Significant decolorization was observed only
14 after the sCOD dropped below 6 00 m g/L during treatment There was very little decolorization (~25%) at pH of 9 or above First order rate constants were determined to compare the kinetics of the photodegradation process. Undoped TiO 2 with UV A lamp and silver doped TiO 2 with UV C lamp had the highest first order rate constants; 0.087 h 1 and 0.063 h 1 respectively. Correspondingly 94%, 99% decolorization and 82%, 91% COD remova l was attained after 41 and 5 9 hours of exposure A mass and energy balance was developed for a 1 million gallon ethanol plant using sugarcane bagasse as feedstock. The results showed biogasification of stillage produces enough biogas to meet the energy de mands in the plant for steam generation and the excess may be used to produce electricity. Phosphates recovered from the process as struvite are en ough to supply ~43 % of the fertilizer needs to cultivate sugarcane for bagasse production. And finally polish ing the wastewater provides a means for recycling water in the plant. This might help reduce the dependence of bioethanol plants on ground water resources.
15 CHAPTER 1 INTRODUCTION Background In the latest announcement by the International Energy Agency ( IEA ), it was speculated that by 2017 producer and by 2030 energy independent ( Reddall 2012 ) However, at this time, f ast diminishing fossil fuel reserves a long with high prices of crude oil imp orted into the United States have led researchers and large scale indus tries to divert their focus on sustainable low cost biomass feedstock s for production of renewable biofuels like ethanol, butanol and biodiesel ( Asif and Muneer 2007 ) In the US and Brazil, eth anol is primarily produced from the food crops : corn and sugarcane respectively. The Environmental Protection Agency EPA, ( www.epa.gov ) has also announced its plans to move forward with the mandate that requires mixing corn ethanol with gasoline, in spite of the drought conditions that have affected corn prices this year. Therefore, the use of food for biofuel production raises ethical concerns related to diverting food crops for fuel production. R esearchers have shown t hat this could cause increase in food prices and therefore more hunger ( Tenenbaum 2008 ) Hence, much attention is being devoted to bioethanol production from lignocellulosic biomass originati ng from agricultural and forest residues Such a process helps over come diverting food crops for ethanol production and at the same time provide s a renewable feedstock for ethanol production ( Limayem and Ricke 2012 ) However, lignocellulosic ethanol process has its own limitations making it less economically favorable from a production point of view. The recalcitrant nature of lignin polymer makes it dif ficult for microorganisms to access the cellulose and hemicellulose polymers in order to extract the sugars that are fermented
16 to ethanol H ence a number of pretreatment options are being studied to make these sugars more readily available for subsequent fermentation Also, large scale bioethanol plants based on conventional feedstocks involve energy intensive processes and use large amounts of water that are currently not recycled in the plant At the downstream end of the bioethanol process, stillage, a wastewater stream is produced following the distillation of ethanol from the fermented liquor. This distillery wastewater is a dark colored, acidic, nutrient rich liquid that presents significant problems with its disposal ( Sheehan and Greenfield 1979 ) With EPA tightening the industrial effluent discharge standards accompanied by decreasing land availability for waste disposal more intense treatm ent approaches must be applied to overcome such hurdles This dissertation presents research d evoted to developing a sustainable model to treat lignocellulosic ethanol stillage while being able to recover green energy in the form of biogas, nutrients in t he form of a slow release fertilizer, ( struvite ) and finally water for reuse in the bioethanol plant. The aim of this work is to reduce the carbon foot print and ground water demands associated with the entire bioethanol process and make it a more sustaina ble process that would negate any harmful impact stillage disposal may otherwise pose to the environment. B ioethanol from Lignocellulosic F eedstock In 2006 bioethanol production was mainly corn starch based in the US and sugarcane and molasses based in B razil, together accounting for 89% of bioethanol pro duced worldwide ( USDA 2006 ) B ioethanol can be produced from fermenting three different sources : (i) sucrose conta ining feedstock, such as sugarcane, (ii) starch based, such as corn and (iii) lignocellulosic feedstock, such as wheat straw. D ue to their
17 abundance and low cost, lignocellulosic feedstock s appear to be a good alternative over food crops for ethanol produc tion ( Balat 2011 ) Table 1 1 lists the composition of lig nocellulosic feedstock Hemicellulose and cellulose make up for about 70% of the dry weight of biomass and are held together by the lignin polymer, which is recalcitrant in nature, thus restricting microbial access to cellulose and hemicelluloses. ( Limayem and Ricke 2012 ) E. c oli K011 that has the capability to ferment hexoses and pentoses to e thanol is used for fermentatio n, after acidic, enzymatic and thermal pretreatment steps. A schematic of the bioethanol production process employed in the Biofuels plant at University of Florida is shown in Figure 1.1. Sugarcane bagasse obtained from Florida Crystals Corporation was use d without any further size reduction. The feedstock was soaked for a period of 4 hours in 1% phosphoric acid and then passed through a screw press to obtain 50% solids. The acid treated feedstock was then fed into a steam vessel and pretreated at 180 C for 10 minutes. The 30% solids content of pretreated b agasse was then fermented using liquefaction plus simultaneous saccharification and co fermentation process ( Nieves, Geddes et al. 2011 ) Following the fermentation process, the ethanol was distilled leaving behind a distillery wastewater called stillage. Although more and more progress is witnessed in the area of lignocellulosic ethanol production, due to the resistant nature of the lignin rich feedstock, and the byproducts produced as a result of harsh pretreatment steps, stillage produced at the end of the process has a high residual organic content meas ured in terms of chemical oxygen d emand, COD Also use of ammonium hydroxide and phosphoric aci d during
18 the pretreatment steps results in the accumulation of ammonia and phosphate in stillage. S uch distillery wastewater rich in organic and nutrient content pose s significant problems to the environment if not properly treated Conventional Techniques f or Handling Distillery Wastewaters Recycle: Studies on handling stillage from cane molasses industries have shown that in distilleries associated with sugar mills, about 30% stillage can be rec ycled as 10 20% of molasses dilution water for a number of cycles without overloading the front end ( Sheehan and Greenfield 1979 ) Dubey (1974) suggested that about 50% of stillage from cane molasses based ethanol industries can be recycled and this would help reduce associated steam consumption and nutrient addition A study on effect of recycling distillery efflu ents to preserve groundwater resources has shown that recycling leads to significant reduction in fermentation activity beyond a certain point. This has been attributed to accumulation of inhibitory compounds such as various organic acids and furfural comp ounds ( Couallier, Payot et al. 2 006 ) In case of lignocellulosic ethanol stillage, the high organic and nutrient content along with low pH add to the inhibitory effect of accumulated organic acids and furfural compounds, thus leading to overload and eventually failure of the entire fer mentation process Direct Land Application / Disposal to water bodies : Sheehan et al. (1979) have stated the advantages of direct land application of cane molasses stillage as irrigation water and as fertilizer They claim that direct land application of stillage is beneficial to crop cultivation as it improves the characteristics of soil including its pH (neutral), water retaining capacity and mineral salts concentrations. However, direct application of c ellulosic ethanol stillage may pose significant pr oblems such as nitrogen immobilization caused by accumulation of excess ive nutrients in soil that may lead to
19 burning of root and shoot tips in crops ( Sachin 2012 ) Phenolic compounds such as tannins, melanoidins produced from Maillard reaction of sugars and proteins, caramels from overheated sugars and furfural compounds from acid hydrolysis contribute towards the dark color of stillage effluent ( Wilkie, Riedesel et al. 2000 ) Hence, disposa l of stillage into water bodies is limited by the dark colored appearance of the wastewater. Fee d Supplement : Corn ethanol and cane molasses stillage have high nutritional values and can be dried and sold as fodder feed or distillers dry grains with solub les (DDGS). Some studies have shown that addition of CaCO 3 for neutralizing the pH of stillage also helped improve the nutritional value of stillage that is otherwise low in calcium content ( Dubey 1974 ) Dried stillage (45 58% total solids) can replace molasses added to cattle feed and may improve cattle milking capacities and act as a good laxative ( Ho dge and Hildebrandt 1954 ; Dubey 1974 ) However, due to the low nutritional value of lignocellulosic ethanol stillage, drying and selling it as animal feed is not a viable option ( Tian 2011 ) Evaporation : Stillage rich in organic carbon content can be evaporated and residue burnt as a fuel or incinerat ed to ash that can be used as fertilizer However, evaporating stillage requires use of multi effect evaporators and this process is also limited by the recurring problems of scaling on the heat transfer surface of evaporators ( Willington and Marten 1982 ) In case of cellulosic ethanol stillage its high moisture content (60%) makes the option of evaporating and combusting it for use as a fuel an unfavorable one ( Tian 2011 ) Water Usage in Bioethanol Plants Consumptive water usage may be defined as the sum total of water input less water that is recycled or reused in the process ( Wu, Mintz et al. January 2009 )
20 Bioethanol plants fall under the category of industries that are high on consumptive water usage. Life cycle inventory and analysis on ethanol production process considers two main stages that are water intensive: (i) irriga tion for crop farming and (ii) e thanol production. Argonne National Laboratory has reported that corn ethanol plants (wet grind ) use 10 gal of water per gal of ethanol produced with 70% for farm irrigation and remaining 30% for ethanol production. Due to the lack of fully functional large scale lignocellu losic ethanol plan ts, calculations based on model simulations have shown that cel lulosic ethanol plants may use a minimum of 6 gal water/gal ethanol primarily during feedstock pretreatment ( Wu, Mintz et al. January 2009 ) Aden (2007) reported results on water usage in commercially operated corn ethanol plants and a simulated cellulosic ethanol plant These results are reported in Table 1 2. Corn e thanol plants operating on dry grind technology use 3 4 gallons of water per gallon of ethanol, while cellulosic ethanol plants operating on biochemical platform could use about 5 6 gallons of water per gallon of ethanol produced. Percent fresh water deman d in individual units, mainly cooling towers and boilers has been list ed in Table 1 2. It has been estimated that cellulosic ethanol plants that operate on thermochemical platform may be able to bring d own their water consumption to as low as 1.9 gal lons w ater per gal lon of ethanol. Thus these estimate s of water usage would result in 250 300 million gallons of water being used in a plant that produces 50 million gallon ethanol per year ( Aden and Foust 2009 ) Most of the water leaves the process as stillage after distillation. Currently, many ethanol plants dispose the distillery wastewater produced at the end of the process without recycling it in the plant, due to various limitations leading to a n excessive dependence on fresh ground water resources T he environmental impact of the process
21 can be greatly minimized alongside improving the economics of the process by implementing technologies for trea ting stillage to such an extent so as to be able to recycle the water within the ethanol plant. Problem Statement In a nutshell, direct disposal of stillage is not an environment friendly procedure and thus it has to be subjected to additional treatment st eps. However, considering the fact that ethanol industries are high on consumptive water usage it would be wise to treat stillage to such an extent so that water recovered in the process may be recycled in the plant. Techno economic studies on pilot scale bioethanol plants have shown that anaerobic digestion of stillage reduces the organic load while producing biogas that can help reduce the dependency on natural gas in the plant and is a viable option in making the process more economical. Although anaero bic digestion reduces the organic load, the nutrients present in stillage are not recovered during the process. Further treatment of the stillage to recover additional byproducts and clean it for reuse has so far not been addressed. To further enhance the sustainability and reduce the carbon footprint, these nutrients could be recovered and utilized for fertilizing the biofuel crop. Finally, t he wastewater could be polished for reuse in the plant. The research presented in this dissertation addresses the d evelopment of an integrated technology to achieve these goals. Project Goal The focus of this project is to develop integrated technologies for biogassification of cellulosic ethanol stillage, recovery of nutrients from stillage and tertiary treatment of the stillage to reuse quality. Incorporation of these technologies would produce a salable product, and reduce water and energy consumption within the plant and could
22 eventual ly lead to decrease in the cost of biofuels produced from lignocellulosic feedsto ck s The technologies proposed are: 1. Anaerobic digestion of stillage for production of biofuel in the form of biogas. Previously, Tian (2011) had shown that stillage can be biogasified in a continuously fed stirred anaerobic digester at thermophilic conditi ons. However, the organic loading rates and hydraulic retention time achieved in this s ystem was only 1.85 2.39 g sCOD L 1 d 1 at 14 22 days retention respectively. A high rate anaerobic digester design that is capable of handling higher organic loading ra tes at low hydraulic retention time may reduce the size of the system and increase the efficiency to treat millions of gallons of stillage in large scale biorefineries. 2. Phosphate recover y from anaerobically digested stillage: Stillage being rich in ammonia phosphorous and other mineral salts is produced by precipitating these nu trients in the form of struvite, a salable fertilizer. This may be a better alternative than subjecting it to conventional wastewater treatment processes. 3. Tertiary treatment for wa ter reuse: Finally, the dark color of stillage along with its high organic matter and nutrient content prevents its direct land application or disposal to water bodies on one hand and the high consumptive water usage of cellulosic ethanol plants on the oth er, subjecting the nutrient deprived anaerobically digested stillage to an extensive oxidation treatment such as photocatalytic degradation may prove useful by decolorizing the effluent and recycling such polished wastewater in the plant. Research Objecti ves The overall goal of this research was to develop an integrated system to treat stillage The research objectives were, 1. To demonstrate the applicability of using a high rate fluidized bed anaerobic treatment system for biogasification of stillage. Speci fic objectives were to investigate the viability of long term operation, response to varying organic loads, and efficiency of organic matter removal and optimum hydraulic retention time. 2. To develop and test a sequential batch reactor for struvite precipita tion and recovery. Specific objectives were to determine the efficiency of soluble phosphate removal, yield of struvite, reaction and settling time and methods for enhancement of settling.
23 3. To study the applicability of titanium dioxide mediated photocatal ytic treatment on decolorization and reduction of organic content in the nutrient recovered stillage stream. Specific objectives were to investigate the effect of time period of exposure, artificial and natural UV sources, depth of treated stillage, pH, in itial organic content, silver doping of TiO 2 particles and wastewater characteristics used in the system on the extent of decolorization and organic matter removal. Research A pproach The goal of this research was to develop an integrated system to treat stillage from the bioethanol process. This section describes the research approach followed in order to meet the overall and specific objectives described above. Lignocellulosic ethanol stillage (11.5 48 g sCOD/L) obtained from the Biofuels plant at the University of Florida was coarse separated using a 0.425 mm sieve. Biogasification of coarse separated stillage was investigated using a high rate anaerobic fluidized bed reactor (AFBR). An AFBR design was developed in collaboration with SRS Scientific, Al achua, FL and the Agricultural and Biological Engineering machine shop, Gainesville, FL. The reactor was built with a cylindrical body and a conical bottom and was operated using activated carbon as the fluidization medium. Fluidization was provided by rec ycling the reactor liquid using a high rate centrifugal pump. Thermophilic conditions were maintained using a heating tape controlled by a CR10 feedback program. The reactor was activated using inoculum from a stirred anaerobic digester that was previously digesting lignocellulosic ethanol stillage. After the initiation phase, the digester was tested for various organic loading rates. The performance of the reactor was assessed based on various parameters that were monitored regularly during the digestion p rocess including: sCOD, pH, organic acids, biogas production rate and methane yield over a period of 100 days. Such an in depth study of the digestion
24 process allowed optimization of hydraulic retention time and organic loading rates for long term operatio n. Following the study on biogassification of stillage, a sequential bed reactor was designed and operated to recover excess soluble phosphate and ammonia present in stillage, in the form of a slow release fertilizer, struvite. A mathematical model progr am based on physi c o chemical equilibrium, mass and charge balance equations previously developed by Gadekar (2010) to determine the yield of struvite from various wastewaters was used in this study. Optimum pH and magnesium requirements were determined usi ng the model. Experiments were carried out on raw and anaerobically digested stillage to determine optimum reaction and settling time. Finally effect of seeding using stillage solids was tested for improving the settleability of struvite containing sludg e. In all the trials mixing was provided by aeration and samples were collected at different heights of the reactor and analyzed for PO 4 P, NH 3 N and total suspended solids content. Subsequent to energy and nutrient recovery, much emphasis was laid on dev eloping a technology for polishing the wastewater. Applicability of TiO 2 /UV mediated photocatalytic treatment for decolorization and organic matter removal from stillage was investigated using a photoreactor set up. The set up was designed using a 13X9 inc h aluminum tray, aluminum foil and three 15 W UV lamps. Reusable TiO 2 coated sheets used in this study were provided by IMDC, Gainesville, FL. Nutrient recovered stillage stream was applied on the TiO 2 coated sheet and exposed to UV for a set period of tim e. Samples were taken at regular intervals and analyzed for absorbance and COD content. These experiments were repeated using UV A, UV C and sunlight, to study the
25 effect of various light sources and at varying depths of treated sample on extent of color a nd organic matter removal The effect of initial pH and organic content was tested during the photocatalytic treatment by carrying out various trials at unmodified and modified initial conditions. Dissertation Organization An outline of the subsequent cha pters is provided in this section. Chapter 2. Anaerobic Digestion of Cellulosic Ethanol Stillage in a Fluidized Bed Reactor This chapter describes the biogasification of stillage under thermophilic conditions in an anaerobic fluidized bed reactor. A detai led discussion on reactor design, operation and parameters monitored during the operation is provided in this chapter. Chapter 3. Recovery and Reuse of Nutrients f r om Cellulosic Ethanol Stillage v ia Struvite Precipitation The third chapter describes the p rocess of recovering excess soluble phosphates and ammonia from stillage in the form of struvite, a slow release fertilizer. The results on struvite potential of raw and anaerobically digested stillage are provided in this chapter. Effect of seeding for im proved settling characteristics of struvite containing sludge has also been reported in this chapter.
26 Chapter 4. Photocatalytic Treatment for Final Polishing of Wastewater from a Cellulosic Ethanol Process Subsequent to energy and nutrient recovery from stillage, the residual recalcitrant organic matter is degraded and the wastewater is decolorized via titanium dioxide mediated photocatalytic treatment. This chapter reports the results on optimization of the photocatalytic treatment step. Chapter 5. Conc lusions and Future Work Based on the results from the present study, overall mass and energy balances were developed for a one million gallon per year cellulosic ethanol plant to determine the extent to which energy and nutrient recovered from stillage may be utilized for various purposes. Recommendations for future work on related areas of research have also been included in this chapter.
27 Table 1 1. Composition of Lignocellulosic Feedstocks (% dry weight) ( Limayem and Ricke 2012 ) Feedstock Hemicelluloses (%) Cellulose (%) Lignin (%) Others (%) Agricultural residues 25 50 37 50 5 15 12 16 Hardwood 25 40 45 47 20 25 0.8 Softwood 25 29 40 45 30 60 0.5 Grasses 35 50 25 40 Switchg rass 30 35 40 45 12 *Agricultural residues such as corn stover, corn stalk, sugarcane bagasse and wheat straw. Table 1 2. Water Consumption in Bioethanol Plants ( Aden and Foust 2009 ) Fresh Water Demands C orn Ethanol Dry Grind Cellulosic Ethanol Biochemical Cellulosic Ethanol Thermochemical Cooling Tower Makeup (% ) 68 71 71 Boiler a nd Process Makeup (%) 32 29 29 Overall Water Demand (Gal H 2 O /Gal Ethanol ) 3 4 6 1.9 Estimates based on Mathematical Mod el
28 Figure 1 1 Schematic of Bioethanol Process Using Sugarcane Bagasse as Feedstock (Nieves, Geddes et al. 2011)
29 CHAPTER 2 DIGESTION OF CELLULOSIC ETHANOL STILLAGE IN A N ANAEROBIC FLUIDIZED BED REACTOR Summary Thermophilic diges tion of cellulosic ethanol stillage was investigated using an anaerobic fluidized bed reactor. Batches of coarsely separated bagasse stillage (0.425 mm sieve) obtained from the Biofuels pilot plant at the University of Florida with a high and variable COD in the range of 11.5 63 g sCOD/L was used as feed. The methane potential achieved was 12.9 L CH 4 (L Stillage) 1 or 0.288 L CH 4 (g sCOD loaded ) 1 The 5.5L digester was operated at a HRT of 7.3 days for a total per iod of 10 0 days at an orga nic loading rat e of 6.54 g sCOD L 1 d 1 The effluent COD at the end of the reactor operation was 3.5 g sCOD/L yielding a soluble organic matter removal efficiency of 93%. This translates to a methane yield of 0.311 L CH 4 (g sCOD removed ) 1 .Operation of the reactor at hig he r loading rates, ie, 8.8 g sCOD L 1 d 1 led to inhibition. pH was maintained between 7.1 8.4 throughout the operational period. Background Lignocellulosic ethanol stillage has a hig h COD content in the range of 11.5 63 g sCOD/L ( Wilkie, Riedesel et al. 2000 ; Tian 2011 ) Although use of stillage as a fuel by directly burning it may seem like an attractive choice; the high moisture content of stillage makes this option an unfavorable one. So, other options have to be considered for managing this waste. As outlined in Chapter 1, recyclin g, land application, processing as feed supplement or evaporation may not be viable and sustainable. However, stillage still have to be treated. The high organic content would exclude aerobic treatment due to large aeration requirements. Anaerobic digestio n of stillage helps overcome this obstacle as a pretreatment step while producing biogas, that could
30 be used as fuel in the plant ( Tian 2011 ) Anaerobic digestion is a multistep biochemical process in which microorganisms break down complex organic substrate to produce biogas which is a mixture of methane (60%) and carbon dioxide (40%) In t he first step called hydrolysis, complex polymeric macromolecules like carbohydrates, proteins and lipids are hydrolyzed to its monomers like sugars, amino acids and fatty acids by extracellular enzymes. These are subsequently metabolized within the cell in to CO 2 H 2 an d volatile organic acids, such as acetic butyric and propionic acid. In the next step referred to as acetogenesis, propionic and butyric acid s are in turn converted to acetic acid. In the final step called methanogenesis, acetic acid and some CO 2 and H 2 a re converted to methane gas ( Bitton 2005 ) So anaerobic digestion may be an option to successfully pretreat stillage while being able to recover gre en energy in the form of biogas ( Wilkie, Riedesel et al. 2000 ) Biogas is an eco friendly biofuel composed mainly of methane that has a calorific value of ~ 32 3 6 MJ/Kg and rest CO 2 ( Shrestha 2001 ) This is also an environment friendly option, as trapping methane and using it as a biofuel, reduces the harmful ef fect of this global warming gas otherwise being releas ed in to ( Paramasivam, Fortenberry et al. 2008 ) Other than converting high COD of the distillery wastewater to methane, anaerobic di gestion also has the advantage of producing much lesser quantities of sludge and having lower nutrient requirements than conventional aerobic treatment for wastewaters ( Wilkie, Riedesel et al. 2000 ) In order to establish the anaerobic digestion process in a reaction vessel, a number of parameters have to be optimized including: temperature, pH, and nutrient concentration s
31 M e thanogens are strict anaerobes. Thus maintaining a completely sealed anaerobic condition is top priority while considering designing an anaerobic digester. An anaerobic digester may be viewed as a cylindrical vessel with a sealed lid on top and a conical bottom equipped with a heating system to maintain the desired temperature (37 C or 55 C), a gas liquid separator and a mixing system. Various reactor configurations have been developed in the past for anaerobic ally digesti n g distillery wastewaters from fe rmentation of sugar or starch based feedstocks such as canesugar or corn ( Moletta 2005 ) These include continuous stirred tank reactors (CSTRs), Upflow anaerobic sludge blanket reactors; downwards stationery fixed film reac tors and fluidized bed reactors ( Callander, Clark et al. 1987 ; Bories, Raynal et al. 1988 ; Fernandez, Montalvo et al. 2008 ; Tian 2011 ) While CSTRs have been shown to be able to digest stillage wastewater with a retention time of 14 22 days, it can be hypothesized that use of a higher end design such a s fluidized bed capable of handling high er organic loading rates may be a more economical alternative for still age digestion at a larger scale ( Tian 2011 ) Suspended culture reactor designs such as anaerobic contact reactors or sludge beds have been used to treat winery and distillery wastewaters ( Vijayaraghavan and Ramanujam 2000 ; Moletta 2005 ) Another technology employed to treat stillage uses microbial aggregates such as biofilm that may be developed on a carrier or as granules that form a sludge blanket. Upflow anaerobic sludge blanket reactors consist of granules that are suspended by the biogas produced in the reactor and the recirculation of the wastewater within the reactor while other designs such as a fixed film reactor or a fluidized bed reactor have microorganisms growing on carriers, in immobilize d and suspended forms respectively
32 ( Wolmarans and de Villiers 2002 ; Moletta 2005 ; Fernandez, Montalvo et al. 2008 ) This research work was focused on designing a fluidized bed reactor to digest cellulosic ethanol stillage at thermophilic conditions. Typically UASB and fluidized bed reactors are cha racterized as high rate systems and have been applied to digest a variety of wastewaters. Not all wastewaters are amenable for high rate anaerobic treatment. The formation and maintenance of granules and biofilms on carriers have been linked to certain cha racteristics of the wastewater ( Moletta 2005 ) Characteristics of wastewater from cellulosic ethanol process could also differ from one another depending on feedstock as well as the pretreatment process used. V ery few studies have dealt with anaerobic digestion of cellulosic ethanol stillage owing to the presence of inhibitory, slowly and poorly degradable organics such as phenolic compounds and furfural derivatives produced during the pretreatment and f ermentation stages Although uncertainties may be faced during digestion, the option of utilizing biogas produced in the process for combined heat and power generation in the plant promises energy efficiency of the ethanol production process. Barta et al. (2010 ) conduct ed techno economic evaluation of spruce to ethanol process based on a SO 2 catalyzed steam pretreatment followed by simultaneous saccharification and fermentation trail using Aspen Plus TM They also investigated the overall energy efficienc y of the process by including various anaerobic digestion configurations and compared these with conventional evaporation of stillage. They reported that stillage yielded significantly higher overall energy efficiencies; 87 92%, as opposed to the conventi onal evaporation of stillage that yielded 81% energy efficiency.
33 Such increased energy efficiency of the ethanol production process may be attributed to production of biogas that can be used in a combined heat and power facility to produce energy or upgrad ed and sold as fuel. This helps compensate for the energy put into the process such as during steam generation for use in pretreatment and distillation. Callander et al., (1987) carried out studies on laboratory scale mesophilic anaerobic digestion of woo They investigated the amenability of anaerobically digesting cellulosic ethanol stillage that was 5 times as dilute as the feed used in the current study in a high rate treating system. However, s ince wood to ethanol stillage was nutrient deficient, N and P were externally added at 240 mg/L and 120 mg/L stillage respectively ( Callander, Clark et al. 1987 ) In another study by Tian (2011), a continuously fed stirred tank reactor which is a low rate technology was used for thermophilic digestion of lignocellulosic ethanol still age. N o external nutrients were added and 90% of methane yield as determined by the biochemical methane potential assay was p roduced within a 14 day retention time resulting in a residual COD of 7 8 g sCOD (L Effluent ) 1 In this work, use of a n anaerobic fluidized bed reactor was investigated for high rate treatment of a higher strength stillage. The specific objectives of this research work were to carry out long term operation of the AFBR to study the presence of inhibitors, degradability of slow degrad ing compounds, COD removal efficiency, maximum OLR, minimum HRT and methane yield from anaerobic digestion of stillage.
34 Materials and Methods Feedstock Cellulosic ethanol stillage was collected from the Biofuels Pilot plant located in the Frazier Rogers H all, at the University of Florida. As explain ed in Chapter 1, raw sugarcane bagasse wa s subjected to dilute phosphoric acid, thermal and enzymatic pretreatment steps and fermented using recombinant E. c oli KO11 Stillage was coarsely separated by passing i t through a 0.425 mm sieve Stillage filtrate was boiled at 80 100 C for an hour to reduce it to half its volume. Deionized water was added to make up the volume to original level. Filtered stillage (digester feed) samples were measured for total solids, volatile solids, ammonia and phosphate. Reactor Design Thermophilic anaerobic digestion of cellulosic ethanol stillage was carried out in a fluidized bed reactor. The schematic diagram of the reactor assembly is shown in Figure 2 1. The reaction vessel wa s made from a modified hollow pyrex glass column, in diameter with a total volume of 7 L The r eactor had a working volume of 5.5 L leaving a headspace of 1.5 L with 6 ports spaced apart that were used for sample collection, feed input, recirculation and thermocouple access. At the upper end, a clamp seal design was used to seal the reactor air tight. A glass flange fitted with an O ring was clamped to a glass lid to ensure a leak proof design. The bottom of the reactor was clamped to a PVC pipe fitted with a ball valve for liquid access. A Teflon ring was used to provide a good seal between the glass reactor and the PVC pipe. A feed tank with a holding capacity of 4 L was refilled periodically with stillage and placed in a refrige rator and continuously stirred. A two headed peristaltic pump was
35 calibrated and set on a timer to feed the digester and remove effluent simultaneously five to ten times a day. The varying feed flow rate used in this study is shown in Figure 2 3 Commerci al p rewashed activated carbon particles (0.5 mm in diameter) were used as carrier in the fluidized bed reactor. The volume occupied by these activated carbon particles amounted to 2 L when fluidized and 1 L when settled. Fluidization was provided by recirc ulating the reactor liquid using a centrifugal pump at a flow rate of 2 L per minute. Use of such high flow rate ensured good mixing in the reactor. In order to prevent clogging in the reactor due to draining of activated carbon particles into the recircul ation line, a screen mesh was inserted be tween the bottom of the reactor and the PVC pipe. Biomass growth was observed on the surface of the activated carbon particles. This design was developed in collaboration with Southern Scientific Inc., located at Mi Engineering machine shop. A modular data logger, CR10 X that is available with a measurement and controller module was used to control the temp erature of the reactor ( Polematidis, Koppar et al. 2010 ) A heating tape was helically wound across the length o f the reactor and con nected to the CR 10 controlled power source The reactor was insulated using an insulation tape 1mm in thickness to reduce heat loss A thermocouple rod (T1) was inserted perpendicular to the surface of the reactor through one of the ports located at a heig for measuring the temperature of the mixed liquor A second thermocouple (T2) was placed on the wall of the reactor near the heating tape
36 to monitor the wall temperature of the glass reactor. An on off protocol was develop ed using the CR10 software to control the process of heating up the reactor liquid to 55 2 C. A provision was made to shut down heating if the wall temperature increased above 100 C in order to prevent burning on outer surface of the reactor. Continuous recirculation prevented formation of temperature gradients inside the reactor. T he is presented in Figure 2 2. A positive displacement U Tube manometer connected to a clicker counter, a solenoid valve and a float s witch was used to measure the volume of biogas produced from the digester. The ga s meter was calibrated at 29 ml /click prior to start of the experiment and checked periodically The CR10 X also record ed the clicks along with time delay made by the gas mete r. The details of the design of gas meter is provided elsewhere ( Koppar and Pullammanappallil 2008 ) Parameters Monitor ed and Analyzed The following on line and off line measurements were made: On line measurements: Biogas production rate was monitored on line. This was calculated by dividing the volume per click of the counter by the time interval between clicks. T empe rature of the digester mixed liquor and the reactor wall temperature was reco r ded on line every 5 seconds. Off line measurements: Biogas composition: The composition of biogas produced was determined by analyzing the samples in a Fisher gas partitioner, model 1200 equipped with a thermal conductivity detector. The gas chromatograph (GC) was calibrated using a gas mixture
37 of 25:45:30 (volume ratio) of N 2 :CH 4 :CO 2 ( Tian 2011 ) The GC was turned on and allowed to heat up for ~15 minutes to reach the set point temperature (50 C). Then 20 ml of the biogas sample collected from the digester (on a daily basis) w as analyzed by purging it through the GC. Digester liquid effluent samples were analyzed daily for pH and periodically for TS, VS, soluble COD volatile organic acids, sugars, other carbohydrates, ammonia and phosphate. pH: The pH of the digester efflu ent was monitored on a daily basis using a n Orion benchtop pH meter. 20 mls of digester effluent sample was collected every day Total (TS) and Volatile Solids (VS): Influent and effluent stillage samples were analyzed for total solids and volatile solids content. Samples were collected and dried in an oven (Fisher Scientific Isotemp model 350G) at 105 C for overnight. The TS content was then calculated using the wet weight, pan weight and the dry weight data. Following the overnight drying, samples are b urnt in a muffle furnace (Fisher Scientific Isotemp) at 550 C for 2 3 hours. The weight of the residue after burning is used to determine the %VS. Soluble COD: Samples were centrifuged in a Fisher Marathon micro H centrifuge for 10 minutes at 8000 RPM an d filtered using a Whatman 0.45 Micron pore size filter paper. Filtered samples were diluted appropriately and analyzed for organic matter content using HACH colorimetric test kit that includes COD reagent vials (20 1500 mg/L COD). The combined mixture was digested for 2 hours at 150 C in a COD digester. The change in color was measured using a HACH colorimeter (model DR/890) to determine the soluble COD of the sample.
38 Organic acids, Furfurals and Sugars: Digester mixed liquor samples were periodically ana lyzed for volatile organic acids such as acetic, propionic and butyric acids using a Shimadzu gas chromatograph (Model GC 9AM) equipped with a flame ionization detector. The GC was allowed to reach the set temperature (column 155C, injector 180C and det ector 200C) after which air, hydrogen and nitrogen gas supplies were turned on and the flame was lit using a multipurpose lighter. 2 microlitres of centrifuged, filtered (0.22 m) and diluted samples were injected through the GC and analyzed for VFA conte nt. A high performance liquid chromatograph (HPLC) was used to analyze the organic acids such as formic, lactic, succinic, levulinic acids, furfurals and hydroxymethyl furfurals and sugar content. Samples analyzed in the HPLC were centrifuged and filtere d using a 0.22 m filter. The HPLC, Infinity 1200 series (Agilent Technologies) equipped with a r efractive index detector was used in this analysis This is a general purpose detection method that can measure concentrations of various components that diffe r in their refractive index. Two different columns (Bio Rad Aminex HPX 87P and Aminex HPX 87H) were used to analyze (i) sugars, furfurals and (ii) organic acid concentrations respectively, in the samples. Organic acids and sugar content of the digested s tillage effluent were measured in the HPLC located at the Biofuels plant in the Agricultural and Biological Engineering department at the University of Florida. Ammonia and Phosphate: Centrifuged and filtered (0.45 m) samples were appropriately diluted a nd analyzed for ammonia and phosphate content. The HACH ammonia nitrogen kit consisting of ammonium salicylate and ammonium cyanurate
39 reagents was used to perform colorimetric analysis to determine NH 3 N concentrations of stillage samples. Although stillag e samples are brown colo red in appearance, appropriate dilution of samples to desired NH 3 N range also helped in eliminating any interference due to sample color. Orthophosphate phosphorous content was measured using the ascorbic acid method as in the Stan dard methods book (18th edition). The change in color was analyzed in a UV Vis spectrophotometer at 880nm in order to determine the PO 4 P concentrations in stillage samples. Reactor Operation The fluidized bed reactor was operated for 90 100 days by feedi ng stillage. Reactor operation was divided into two distinct phases during this period: Phase i. Reactor Start up (days 0 17) Phase ii. Operational phase (days 17 100) Phase ( i ) : Anaerobi c digestion of stillage in the 5.5 L fluidized bed reactor was i nitiated with seed inoculum from a CSTR that was previously used to digest cell ulosic ethanol stillage (Tian 2011). During phase (i) pH of the stillage was adjusted to 7 us ing 5N sodium hydroxide and buffered using 5 g/L of sodium bicarbonate before being fed to the digester. 300 ml of stillage was fed twice every week for the first we ek and then increased to 500 ml twice a week. Macro and micro nutrients used in conventional BMP assays were added to the reactor during the start up phase (Owen, Stuckey et a l., 1979) Reactor was fed daily once the inoculum was activated and acclimated to cellulosic ethanol stillage. This allowed the inoculum to be acclimated to the feedstock and to develop biofilm on the activated carbon carrier particles. Phase ii : The op erational phase of the reactor can be further divided into three phases: (a) elevated loading, (b) normal operation and (c) inhibited phase.
40 During the variable or elevated loading phase, the organic loading rates were g radually increased from 2.11 g sCO D L 1 d 1 to 8.8 g sCOD L 1 d 1 to test the organic handling capacity of the reactor. The digester was subjected to an increased org anic loading rate of 8.8 g sCOD L 1 d 1 for a period of 2 days by increasing the flow rate of the feed which in turn reduced th e HRT. After testing elevated organic loading rates, full fledged operation of the reactor was continued at a loading rate of 6.54 g sCODL 1 d 1 Results Characterization of Stillage Various batches of stillage obtained from the Biofuels plant at the Uni versity of Florida were combined into one and coarse separated for use as feedstock in this study. The total solids content prior to solids separation was 12.5 2.03 g (kg stillage) 1 and after coarse separation was 4.14%. Different batches of stillage obt ained from the Biofuels pilot plant had varying organic content between 11.5 63 g sCOD/L. The combined stillage mixture had a sCOD of 23.5 48.4 g/L. The pH of the influent was 6.42 0.86. T otal nitrogen content of stillage was 65.8 5.68 g (kg stillage) 1 and the individual concentrations of ammonia, nitrate and organic nitrogen were determined to be 23.6 3.01 33.6 3.62 an d 10.1 0.94 g (kg stillage) 1, respectively on a dry weight basis. The total phosphorous content was determined to be 35.2 34.3 g (kg stillage) 1. Elemental composition of whole stillage sample including K, Ca, Mg, Fe, S levels are reported in Table 2 1 These above mentioned analysis were performed on 6 replicates. The mean and standard deviation of the results obtained are listed in Table 2 1
41 Stillage was also characterized for sugar and organic acids content. The results showed that the coarse separated stillage had 1.06 g/L, 18.6g/L and 1.2g/L of glucose, xylose and arabinose respectively. Organic acids such as succinic, lacti c, formic, acetic and levulinic acid were present at concentrations of 9.3g/L, 3.7g/L, 10.1g/L, 5.4g/L and 1.6g/L respectively. The influent s tillage also contained hydroxy methyl furfurals and furfurals at 0.09 and 0.05 g/L respectively. The total COD of the stillage mixture was measured to be 48.4 g sCOD/L, while the above mentioned components add up to a total of 46.8 g sCOD/L. It can be hypothesized that other unknown compounds such as furfurals and hydroxymethyl furfurals may contribute to the remaini ng COD of stillage. Biogassification of Stillage The fluidized bed reactor was run as a batch or semi continuous system for a period of 100 days. During the initial start up phase, ie., between day 0 day 17, the reactor volume was cumulatively built up by adding 300 mls and subsequently 500 mls of stillage twice a week. The pH of the reactor ranged between 7.8 and 8.1. The average residual soluble COD of the digestate was about 1.4 g/L. This phase allowed the microbial consortia to get acclimated to the feedstock. Following the start up phase, performance of the reactor was tested at elevated organic loading rates. The organic loading ra tes used are shown in Figure 2 3 along with the feed flow rate and the sCOD of the influent mixture fed to the digeste r. The reac tor was operated at 2.13 g sCOD L 1 d 1 from day 18 to day 25 following which the organic loading ra te was increased to 6.54 g sCOD L 1 d 1 Figure 2 4 presents the pH, methane percent, methane yield and methane production rate of anaerobically d iges ted stillage effluent. Table 2 2 summarizes the steady state results from operational phase of the fluidized bed reactor. The reactor was operated at a HRT of 7.3 days and the pH remained between 7.1 and
42 8.2 during this period. The biogas composition stabilized around 60 5 % methane and 40 5 % carbon dioxide. The methane yield during this period was 12.9 L CH 4 (L Stillage) 1 or 0.311 L CH 4 (g sCOD removed ) 1 or 0.288 CH 4 (g sCOD loaded ) 1 The sCOD of the feed and the digested effluent were 46.71. 7 and 3.50.2 g sCOD/L respectively, yielding an organic removal efficiency of 93%. Influent and effluent ammonia concentrations were 460, and 320. Phosphate concentrations in the feed and effluent wer e 667 and 527.3 mg/L. Figure 2 5 presents the concent rations of hexose a nd pentose sugars and Figure 2 6 presents various organic acids such as formic, acetic, propionic, butyric, lactic, succinic and levulinic acid levels monitored during the operation of the reactor. The total sugars concentration in the reactor remained below 1.5 g/L with glucose being less than 5% of the total sugars and xylose and arabinose accounting for 45 50% each. As shown in Figure 2 6 acetic, propionic and butyric acid contents were determined to be well below 1.5g/L each. Aroun d day 60, the organic loading rate of the reactor was fur ther increased from 6.54 g sCOD L 1 d 1 to 8.8 g sCOD L 1 d 1 On doing so the concentrations of acetic, propionic and butyric acids were significantly elevated to 2.17, 3.88 and 0.5 g/L respectively. During this period, although pH of the digester effluent remained around 7.5 7.8, the methane content of biogas dropped from 60% to 31%. Also the residual sCOD increased from 3.4 to 10.3 g/L. Such a high effluent COD and increased amounts of organic acids concentrations in the digester were indicative of a dominant acidogenesis and an inhibited methanogenesis. In order to bring the system back to balanced operation and prevent any sustained imbalance in the reactor due to overloading, the reactor was left unfed for a day and sub sequently feeding was
43 resumed to a lower organic loading rate for the rest of the period. On resuming re actor operations at 6.54 g sCOD L 1 d 1 the methane content hiked back up 605% and the residual COD decreased to 7 g/L and event ually back to 3.5 g/L after ~10 days. Total volatile organic acid levels also reduced to below 3 g/L with acetic, propionic and butyric acid levels at 0.282, 2.22, 0.132 g/L respectively. Discussion In this study use of a conventional fluidized bed reac tor was investigated for thermophilic anaerobic digestion of cellulosic ethanol stillage. Since limited data is available from previous research on anaerobic digestion of cellulosic ethanol stillage, results from the present study have been discussed in th e following paragraphs in comparison to two other studies carried out in the same area. T he steady state values of parameters monitored during the digestion process in all three cases, (i) FBR present work, (ii) CSTR Tian (2011) and (iii) UASB Callander e t al., (1987) are presented in Table 2 1 Tian (2011 ) showed that 70% of the methane yield was obtained from digestion of the filtrate and the solids not only contributed a minimal fraction but also were slow degrading. Thus biogasification stillage filtr ate alone was investigated in this study study. The rationale behind the use of thermophilic temperature (55 C) for digestion of stillage was to make use of the high er temperatures (75 80 C) in the stillage discharged from the distillation columns. Instead of applying additional energy to cool the effluent to mesophilic temperature, it was hypothesized that the higher effluent temperatures can be taken advantage of b y carrying out digestion at thermophilic conditions.
44 Callander et al., (1987) carried out studies on laboratory scale mesophilic anaerobic digestion of wood ethanol stillage using a UASB reactor. Pilot scale studies at the New Zealand research institute we re focused on fermenting softwood Pinus radiata to ethanol using a sulfuric acid hydrolysis fermentation distillation scheme in a 220L reaction vessel. Characterization of stillage from this process showed that the main contributors to the high organic con tent of stillage were carbohydrates and lignin degradation products. Acetic acid, levulinic acid and hydroxymethylfurfural from hexose sugar degradation and tra ce amounts of volatile furfurla from pentose sugar degradation were determined to be present wel l below their inhibitory levels in stillage ( Callander, Clark et al. 1986 ) Organic acids, furan derivatives and o ther phenolic compounds are primary inhibitors that affect microbial growth during fermentation as well as anaerobic digestion process. Weak acids enter the cell and dissociate causing a significant drop in the pH, while furan derivatives and phenolic comp ounds tend to hinder functioning of fermentative enzymes and cause membrane disruption ( Mills, Sandoval et al. 2009 ) In case of anaerobic dige stion, unless the mixed culture is comprised of furfural degrading methanogens such as Methanococcus deltae that is capable of detoxifying fulfural to furfuryl alcohol, the process may be inhibited as a result of stunted microbial growth ( Belay, Boopathy et al. 1997 ) B oopathy (2001) has shown that furfurals can be degraded to produce methane using defined mixed cultures under mesophilic conditions. Boopathy et al., (1992) have also in vestigated the biotransformation of furfurals and 5 HMF by 17 different strains of enteric bacteria under mesophilic conditions. They show that although enteric bacterial strains do not grow using furfurals as a sole carbon source, they are capable of co m etabolizing them along with glucose
45 and peptone and the primary means of transformation of furfurals is detoxification followed by its ultimate anaerobic digestion to CH 4 and CO 2 However, research work by de Vrije et al., (2009) on hydrogen production fro m lignocellulosic biomass under thermophilic conditions shows that furfurals (0 4 g/l) were found to inhibit growth of Caldicellulosiruptor saccharolyticus and Thermotoga neapolitana even at low concentrations. Contrary to this information, results obtaine d in the present work show that influent furfurals were degraded, if not completely, atleast to levels below the detection limit of the HPLC. It may be hypothesized based on these results that the mixed culture used in this study comprises of thermophiles capable of degrading furfurals. In this study, stillage obtained from the Biofuels plant at the University of Florida was produced from fermentation of bagasse at a stillage to wood ratio of 4.63 L Stillage (k g bagasse ) 1 indicating that stillage from t his process was ~5 times more concentrated than that used in the study by Callander et al (20 L Stillage (Kg Wood) 1 ). Correspondingly, total reducing sugars concentration and chemical oxygen demand of the influent used in the present study were 10.5 and 1.89 times higher than wood stillage respectively Concentration of acetic acid in the bagasse stillage was 34.3 times higher than wood stillage. Also other organic acids such as levulinic acid were present in much higher concentrations (2.9 times highe r) in bagasse stillage. A review paper on various pretreatment methods to improve the digestibility (saccharification) of lignocellulosic biomass by Hendricks and Zeeman (2009) states that subjecting the feedstock to strong acid pretreatment compared to di lute acid treatment will result in immediate solubilization of hemicellulose leading to production of furfurals and
46 hydroxymethyl furfurals, precipitation of solubilized lignin compounds and production of other volatile organic compounds. Since bagasse st illage was pretreated using dilute phosphoric acid, the amount of acids, furfurals and hydroxymethyl furfurals were present well below their inhibitory levels; 0.5 3, 1.7 2.1 and 8.1 11.5 g/L respectively (Callander 1987). Followi ng the studies on charac terization of stillage, Callander et al., (1987) investigated the possibility of anaerobic ally digesting stillage in a n UASB reactor to study the amenability of the process. At a retention time of 2 days, the UASB reactor was operated for a period of 130 d ays by gradually elevating organic loading rates from 3.6 g COD L 1 d 1 to 16 g COD L 1 d 1 ( Callander, Clark et al. 1987 ) In the present study, the fluidized bed reactor was initiated by feeding stillage at loadi ng rates of 2.11 g sCOD L 1 d 1 and t hen gradually increased to 8.8 g sCOD L 1 d 1 Different batches of stillage received had varying organic content and they were combined into a uniform mixture and used as feed. During the start up phase, the feed had a sCOD of 23.5 g/L and during the operational phase the sCOD content was 46.71.7 g/L. The high COD, total sugars and organic a cid levels in stillage used during the operational period was a result of combining one of the batches of stillage that was from a fermentation trial that failed to reach completion. In the study by Tian on thermophilic anaerobic digestion of stillage in a had a COD of 38.8 g sCOD/L ( Tian 2011 ) This stillage was also from bagasse fermentation and had comparable characteristics with that used in the present study. The fluidized bed reactor was run at a HRT of 7.3 days and an organic loading rate of 6.54 g sCOD L 1 d 1 during the operational period. The methane yield during this
47 period was 12.9 L CH 4 (L Stillage) 1 The methane yield reported by Tian (2011) from stillage digestion using the CSTR was 10 L CH 4 (L Stillage) 1 at a HRT of 22 days and organic loading rate that could be attained was only about one third of what was used in this present work. The higher methane yield in the FBR may be attributed to higher organic handling capacity resulting in greater COD removal. The retention time employed in the AFBR was reduced to one third of that in the CSTR. However, this period was still three times as long as the retention time achieved in the UASB. The reason for longer retention times in the FBR is hypothesized to be a consequence of using a more concentrated influent feeds along with carrying out the digestion at a much higher temperature. As explained in the introduction section thermophilic conditions possess many notable advantages such as increased pathogen destruction, higher rates of solubilization and methane production. However, higher rates of solubilization at thermophilic temperatures lead to higher organic acids production in the reactor ( Chen, Cheng et al. 2008 ) Suc h an observation is consistent with the results obtained from the FBR, CSTR and UASB reactors. Volatile organic acids produced in the FBR and CSTR operated at thermophilic conditions was 20 30 times higher than that produced in the UASB operated at mesophi lic conditions. Fuchs et al., (2003) studied the anaerobic digestion of three different wastewaters: artificial wastewater, wastewater from vegetable processing industry and wastewater from an animal slaughterhouse with a COD of 9700, 5800 20150 and 40700 64600 mg/L respectively. The organic acid levels were 820, 395 4055 and 6430 10940 mg/L respectively. These wastewaters were digested in a stirred tank coupled with a membrane filtration unit at organic loading rates of 20, 8 and 6 8 g CODL 1 d 1 respect ively ( Fuchs, Binder et al. 2003 ) Due to th e
48 7.3 days) in order to obtain more than 90% COD removal and to avoid any inhibition in (2 days), due to use of stillage that had twice the COD content of wood stillage. Us e of d to inhibition due to over production of organic aci ds. The organic removal efficiency was as high as 93 % in the FBR as compared to 86% and 80% during the operational phase of the UASB and CSTR respectively Although the initial COD of the feed used in the FBR was about twice as much as the feed digested in the UASB reactor, the final COD was the same: 3.5 g COD/L. This shows that the fluidized bed reactor was much more efficient in organic matter removal than the UASB reactor. The resultant methane yields in L CH 4 per g COD removed were 0.311, 0.302, and 0.300 from the FBR, UASB and CSTR respectively. The methane yield was also calculated for the three reactor designs based on the amount of organic matter loade d and was determined to be 0.288 0.260, 0.258 L CH 4 per g COD loaded. S ince wood to ethanol sti llage was nutrient deficient, N and P were externally added at concentrations of 240 and 120 mg ( L stillage ) 1 respectively. Also, since their fermentation scheme involved sulfuric acid pretreatment, the influent stillage with high sulfate levels (1800 mg /L) was subjected to barium precipitation technique before feeding the digester. Studies have shown that barium precipitation techniques could also be used to precipitate out phosphates, inorganic and organic, in the form of barium phosphates and nitrogen in the form of barium nitrates Barium phosphates are insoluble in water and
49 may have precipitated out along with sulfates leading to P removal from stillage ( Whittier 1991 ) On the contrary, stillage obtained from the Biofuels pilot plant at the Uni versity of Florida was subjected to dilute phosphoric acid pretreatment. This eliminated the otherwise inhibitory effects of sulfides (> 50 100mg/L) that originate from mineralization of sulfur containing proteins and from reduction of sulfate in the diges ter. Stillage obtained from the biofuels plant was rich in N (460 mg/L) and P (667 mg/L). Nitrogen and phosphorous are macronutrients that are essential for the growth of microorganisms in the digester. The phosphoric acid pretreatment of lignocellulosic feedstock contributes to the high phosphorous content of stillage. Anaerobic digestion of organic materials such as proteins leads to accumulation of ammonia in the system. Although free ammonia is toxic to microbial growth at inhibitory levels of 1500 300 0 mg/L, around neutral pH it is present as ammonium that is less toxic than free ammonia ( Lyberatos and Pullammanappallil 2010 ) At the end of the operationa l phase the effluent ammonia and phosphate levels were 320 mg/L and 527.3 mg/L respectively, indicating that the reactor was not nutrient deprived unlike the UASB where nutrients were externally added. M ore than 90% of incoming glucose and xylose and 55% of arabinose were readily taken up by the mixed culture of microorganisms during the digestion process indicating that the culture was well acclimated to the feed. The influent cellulosic ethanol stillage had a COD of 46.71.7 g/L that was reduced to ~3.5 g sCOD/L in the effluent during the operational phase. This level went up to as high as 9.4 10.3 g/L during the inhibited phase. A COD balance was performed on the effluent as shown in Figure 2 6. The influent stillage contained 9.3, 14.7, 27 and 1.6 g/L of succinic, lactic, formic and
50 levulinic acids respectively. Figure 2 8 presents the plot of succinic, levulinic acids and furfural compounds concentration in the feed and di gested stillage. Furfurals and h ydroxymethyl furfurals measured in digested sti llage were well below the detection limit of the HPLC and hence were not reported in this plot. Although succinic, levulinic acids and furfural compounds are not produced during the anaerobic digestion process, they are introduced into the digester along w ith the influent. Microbial biomass in the digester seems to have been well adapted to the influent feed and hence these compounds were digested along with other sugars and volatile fatty acids. Such an observation is also in agreement with studies by Owen (1979) and Brune et al., (1982) who showed that furfural, hydroxymethyl furfural and levulinic acids produced from fermentation of lignocellulosic feedstock can be completely degraded by methanogenic consortia during anaerobic digestion. Lactic acid, whos e presence is indicative of overloading or dampened functioning of the anaerobic digester, was present in very low quantities during the operational phase of the reactor indicating that the reactor was able to handle higher organic loads. Acetic, propionic and butyric acids were monitored throughout the operation of the FBR. Their concentrations remained below 1.5 g/L throughout the operational phase. This is indicative of a dominant population of active methanogens that were not dominated by the fast growi ng acidogenic bacteria. The COD balance presented in Figure 2 7 presents that about 85% of the organics in the effluent are accounted, leaving behind an average of 15% unaccounted for. These unaccounted organics are shown as unknown compounds in the plot Such a COD balance acts as a control tool in testing the organic matter distribution. All the COD that enters the digester is distributed into methane gas produced, biomass
51 generated and residual organic matter in liquid effluent. In this study since th e amount of methane produced and residual organics were periodically monitored, it is safe to assume that the unaccounted COD may have been used for biomass synthesis (Colon 2012). During the inhibited phase that occurred as a result of further increasing the org anic loading rate to 8.8 g sCOD L 1 d 1 the digester suffered imbalance from high levels of volatile organic acids; 2.17, 3.88 and 0.5 g/L of acetic, propionic and butyric acids respectively. Lactic acid concentration also increased ~10 times fro m 58 mg/L on day 64 to 558 mg/L on day 70. Such a drastic increase in lactic acid levels is a clear indicator of inhibition. Such drastic elevation in organic acid levels led to inhibition of methanogenic biomass that in turn led to a significant drop in t he methane content of biogas produced from the reactor. As a result of reactor inhibition, the residual COD of the reactor effluent gradually increased from 3.4 to 5.7, 9.4 and 10.3 g sCOD/L successively over a period of one week. Although increased feedin g rate was only implemented for two operational days, it took more about 10 days for the biomass to resume methanogenic activity. Studies on anaerobic toxicity of pulp mill waste by Benjamin et al., shows that overall response of methanogenic microbes to t oxic compounds can be categorized into three ranges: (i) where the toxicant is present in very low concentrations and does not affect microbial activity, (ii) toxicant is present in slightly higher levels leading to temporary disturbances in microbial grow th, seen as longer lag period or lower metabolic rates, but ultimately leading to acclimation and renewal of microbial activity and (iii) where the toxicant if present in significant levels leads to ceasing of any microbial activity. In this study, initial ly when inhibitory
52 compounds were present in very low levels, microbial activity in the digester was not affected by their presence. On increasing the loading rates to 8.8 g sCODL 1 d 1 toxicant levels were significantly increased leading to temporary dist urbances in microbial growth. Inhibition of the reactor due to increased loading rates may also be hypothesized to have occurred due to limited population if methanogens that were available to breakdown additional organic matter fed to the digester. Howeve r, since inhibition was detected at early stages, resuming the reactor operation at reduced organic loading rates allowed the biomass to renew methanogenic activity after an extended lag period of more than one week. Closing Remarks Anaerobic digestion of bagasse stillage was successfully carried out at thermophilic conditions in a fluidized bed reactor at HRT of 7.3 days. The FBR was able to successfully handle 6.54 g sCOD L 1 d 1 organic loading rate; however at hig her organic loading (8.8 g sCOD L 1 d 1 ) t he reactor was inhibited. 93% COD removal was achieved at the end of the run. The methane potential at steady state was 12.9 L CH 4 (L Stillage) 1 Total sugars and organic acids content of the digester effluent at the end of the run were 0.317 g/L and 2.64 g/L respectively. Effluent ammonia and phosphorous and sCOD levels were 320 mg/L and 527.3 mg/L and 3.5 g/L respectively, indicating that a nutrient removal step along with tertiary polishing step is required before discharging the effluent to the envir onment or reusing the treated wastewater in the plant.
53 Table 2 1 Characterization of raw stillage Property Raw Stillage pH 6 .42 0.86 Solids (%) 12.5 2.03 Nitrogen (g kg 1 ) 65.8 5.68 Ammonium N (g kg 1 ) 23.6 3.01 Nitrate N (g kg 1 ) 33.6 3.62 Organic N (g kg 1 ) 10.1 0.94 Total iron (g kg 1 ) 0.14 0.02 Total aluminum (g kg 1 ) 0.09 0.01 Total P (g kg 1 ) 35.2 34.3 WEP (g kg 1 ) 6.42 0.79 PWEP (%) 25.4 3.62 Mehlich 3 P (g kg 1 ) 18.4 2.04 Mehlich 3 K (g kg 1 ) 4.63 0.61 Mehlich 3 Ca (g kg 1 ) 1.02 0.03 Mehlich 3 Mg (g kg 1 ) 0.94 1.04 Elemental S (g kg 1 ) 6.72 1.01 Reference: Agyin Birikorang, et al. (2010) Table 2 2 Steady state values of parameters monitored Parameters FBR CSTR UASB pH pH units 8.00.3 7.76 70.15 HRT D ays 7.3 22 2 OLR g COD/L/d 6.6 1.85 16.4 sCOD (influent) g/L 46.71.7 38.8 25.5 sCOD (effluent) g/L 3.50.1 7.5 3.57 Methane Yield L CH 4 (g COD removed ) 1 0.311 0.300 0.302 Methane Yield L CH 4 (g COD loaded ) 1 0.288 0.258 0.260 VFA ( Acetic, Propionic acids) g/L 2.64 2.23 0.07 NH3 N (influent) g/L 0.46 0.3 0.240* NH3 N (effluent) g/L 0.32 0.24 <0.310 PO4 P (influent) g/L 0.667 0.53 0.120* PO4 P (effluent) g/L 0.527 0.19 <0.003 Reference Present work Tian (2011) Callander (1987) *Nutrients were externally supplied
54 Figure 2 1. Schematic representation of an anaerobic fluidized bed reactor
55 Figure 2 2. Temperature Profile of digester liquid during a 24 hour period.
56 Figure 2 3. Feed Flow rate, Org anic loading rate and influent Soluble COD concentrations of digested stillage during entire operation of AFBR
5 7 Figure 2 4 Parameters monitored during the entire operation of fluidized bed reactor (i) Methane Percent in Biogas (ii) Methane Yield (iii) Methane Production Rate, (iv) pH
58 Figure 2 5 Concentration of glucose, xylose and arabinose (sugars) in digested stillage
59 Figure 2 6 Organic acids concentrations in digested stillage effluent
60 Figure 2 7 COD balance on digested sti llage effluent
61 Figure 2 8. Succinic acid, Levulinic acid and Furfural compounds (Furfural and Hydroxymethyl Furfural, HMF) concentrations in feed and digested stillage.
62 CHAPTER 3 RECOVERY OF NUTRIENTS FROM CELLULOSIC ETHANOL STILLAG E BY PRECIPITATION AS STRUVITE Summary Struvite precipitation is a means of recovering nutrients, mainly phosphorous and nitrogen from treated wastewaters. While much research has been carried out in this area on optimizing the conditions for struvite prec ipitation including pH and ratio of chemical s added, less attention has been given to seeding the process to promote the rate of crystal growth and thus improve settleability of precipitated struvite from anaerobically digested effluents. In this study, st ruvite precipitation from raw and anaerobically digested stillage was investigated at a pH of 8.9 and che mical dosage determined using a physi co chemical equilibrium model developed by Gadekar (2010) that predicts the yield of struvite from wastewaters. Stu dies were carried out on settleability of struvite containing sludge obtained from processing unseeded and seeded anaerobically digested stillage. Results showed that unseeded trials produced a large amount of unsettled fine mineral precipitates, while in the subsequent experiments where 1% (wet w/v) stillage solids were used as seed material, improved settleability of processed sludge was witnessed. More than 95% of settling occurred within the first 15 minutes of undisturbed settling following the 30 min reaction time. About 99.9% and 56% orthoph osphate phosphorous and ammonia nitrogen respectively were removed from anaerobically digested cellulosic ethanol stillage via struvite precipitation. Seeding also increa sed the yield of settled sludge containing struvite precipitate by 63%. Struvite containing settled sludge obtained from processing raw stillage was also tested for its agronomic and environmental efficiency and nutrient leachability. Results showed
63 improved nutrient uptake by plants and reduced N and P levels in leachate on application of processed sludge as soil amendment for cultivation of sweet sorghum. Background R emoval of nitrogen (N) and phosphorous (P) from wastewater s is becoming an increasing challenge for operators as regulatory autho rities tighten discharge standards to avoid eutrophication problems in receiving waters (UWWTD, 1991). Currently a number of physical, chemical and biological techniques have been employed to treat high levels of N and P in wastewaters prior to their dispo sal to water bodies. This is a vital step in wastewater treatment as direct discharge of wastewaters rich in nutrients is illegal in many countries and such actions have deleterious effect on aquatic life (Doyle, 2002). In this work, emphasis has been laid on phosphorous recovery due to the rapid depletion of mineral phosphate resources which are the primary source of phosphorous for fertilizer, detergent and insecticide industries (Steen, 2004). While on one hand there is shortage of naturally available mi neral phosphates, on the other phosphate present in waste side streams such as agricultural run offs and industrial effluents disposed off into water bodies or land applied are posing a threat to the environment (Schipper, 2001). Thus, it is wise to recove r and reuse phosphate from wastewaters in order to reduce our dependence on mineral phosphate resources as well as avoid any negative impact on the environment. Phosphate removal techniques presently used in large scale i ndustries include (Strom, 2006),
64 (i) P hysical: Filtration, Membrane technologies (ii) Chemical: Precipitation, Adsorption (iii) Biological: Assimilation, Enhanced Biological Phosphorous Removal Filtration is a conventional physical treatment for removal of particulate phosphate using sand filters or ot her media that can trap sediments and particulate matters (Reardon, 2006). However, conventional filtration techniques fail to completely remove suspended matter thus allowing high levels of residual phosphorous in wastewater effluents. Use of technologie s such as membrane bioreactors, provides a way of introducing a physical barrier that is capable of capturing all the suspended solids from treated wastewaters. Studies employing submerged membrane bioreactors for nutrient removal have shown that when used in combination with chemical or biological phosphorous removal techniques, membrane technology can help remove upto 70% and 93% of N and P from treated wastewaters respectively (Koch 2009). In chemical precipitation, metal salts such as alum, ferric chlo ride or calcium are added to precipitate out phosphorous present in treated waters (Storm, 2006). Application of such chemical precipitation techniques suffer from limitations such as added costs associated with use of chemicals and sludge processing and h ence they are less suitable as a primary wastewater treatment method. Therefore, such techniques are used for tertiary treatment post biological treatment of wastewaters (Neethling and Gu, 2006). Various adsorbants such as granulated ferric hydroxide, ac tivated aluminum oxide, biochar and activated carbon have also been used for phosphorous removal from biologically treated wastewaters ( Genz, Kornmuller et al. 2004 ; Yao, Gao et al. 2011 ; Zhang, Wan et al. 2011 ) Such methods are useful as polishing agents that reduce the
65 final nutrient level to meet the regulato ry standards post biological or chemical nutrient removal. However, use of such treatment technologies as primary means of nutrient removal may not be economical. Biological phosphate removal technology involves the use of phosphate conditions. These organisms release phosphate during the anaerobic phase alongside production of volatile fatty acids which then are broken down alongside phosphate uptake during the aerobic stage ( de Bashan and Bashan, 2004). Such a treatment can help reduce phosphate levels in treated wastewaters by increasing the P uptake capability of microorganisms from 1.5% to 5% (Storm, 2006). Although the physical and biological techniques mentioned above po ssess a number of advantages, one common disadvantage these methods suffer from is their focus on phosphate removal rather than recovery in a reusable form. In this study, chemical precipitation in the form of struvite, a slow release fertilizer, has been investigated for recovery and reuse of nitrogen and phosphorous from cellulosic ethanol stillage. The key feature of this nutrient recovery technique is the combined removal of ammonium (NH 4 + ), phosphate (PO 4 ) and magnesium (Mg 2+ ) from supersaturated s tillage in the form of magnesium ammonium phosphate hexahydrate (MgNH 4 PO 4 stable, white, orthorhombic crystals in the stoichiometric ratio (Mg 2+ :NH 4 + N:PO 4 3 P = 1:1:1) according to follo w ing reaction (Li et al., 1999) Mg 2+ + NH 4 + + PO 4 3 + 6H 2 4 PO 4 2 (1)
66 Several researchers have addressed the recovery of NH 4 + or PO 4 as struvite from various types of wastewaters, including landfill leachate (Li et al., 1999; Li and Zhao, 2003; Kim et al., 2007 ), swine wastewater (Oh et al., 2005; Suzuki et al., 2007), source separated human urine (Basakcilardan Kabakci et al., 20 06; Ronteltap et al., 2007 ), industrial was tewater (Celen and Turker, 2001 ), anaerobically pretreated domestic wastewater (Altinbas et al., 2002), slaughterhouse wast ewater (Kabdasli et al., 2003), filtered pig manure wastewater (Kalyuzhnyi et al., 2002), anaerobic swine lagoon liquid (Nelson et al., 2003), and supernatant of anaerobically digested sludge (Battistoni et al., 1997). Commercial Technologies for Struvite Recovery A number of reactor designs have been studied in the past for precipitation of struvite from wastewaters. Large scale plants such as those in Italy (Treviso), USA (Crystalactor) and Canada (Ostara) primarily employ a fluidized bed reactor for stru vite precipitation. The two important stages in formation of mineral prec ipitates from wastewater are nucleation and crystallization. Struvite precipitation process has the following common scheme that has been tailored in various ways by different researc hers (Gadekar 2010), A ddition of required dosage of chemicals for supersaturation in order to promote nucleation. A djustment of pH either by addition of alkali or aeration in order to provide optimal conditi ons for struvite precipitation. A ddition of a see d material or provision of a surface for crystal growth that causes settling. the form of pellets, 1.5 3.5mm in diameter. The nutrient rich feed is fed to the reactor and r equired chemicals are added and the pH is adj usted by addition of alkali (NaO H).
67 Struvite fines are formed that act as seed material which are fluidized in the reactor till they grow in size (1 3.5 mm in diameter) following which they are harvested from th e bottom end of the reactor. struvite. The first zone allows crystal nucleation while the second one is used for crystal growth and settling. The fines suspended in the settling zone are recycled back to zone one to act as seed material to enhance the nucleation process. In the Crystalactor plant and the Treviso sewage works plant, sand particles are used as seed to allow growth of struvite crystals and improved settling in the reacto r. Fluidized bed reactors have been preferred for chemical precipitation due to their ability to promote crystal growth and control particle size distribution. However, these designs suffer from some significant drawbacks, It is critical to control the hy drodynamic behavior such as initial contacting and hindering complete mixing conditions inside the reactor (Doino, 2011). Secondly, large scale P recovery processes such as the Phosn ix process employed in Japan, use long hydraulic retention times in the range of 10 14 days to obtain large crystals that cause faster settling and more than 90% P removal. Finally, studies have shown that use of s horter hydraulic retention time results i n fine suspended particles being entrained in the reactor liquid and removed before crystallization occurs. Studies by Wang and Burken (2003) have shown that use of appropriate seeding material and right mixing strength will promote crystal growth and imp rove settleability. Experiments were carried out at pH 9.0 using various mixing intensities provided by aeration (200, 300, 400 ml/min of air bubbles) and 0.5g seed. It was concluded that struvite fines, 75 150 m in size, were most effective as a seeding material for struvite precipitation. However, it was also shown that use of larger sized seeding material
68 resulted in higher P removal. They suggested that though smaller seed material provided more surface area, due to its lower specific gravity, they ten d to remain suspended in solution than settle down. In this present study, use of a sequential batch reactor has been investigated for recovery of nutrients via struvite precipitation from cellulosic ethanol stillage. During reaction period mixing is provi ded via aeration, and the reacted products allowed are to settle in a quiescent state with aeration being turned off. Secondly, in order to enhance size of the struvite precipitate, the effect of seeding using stillage fibrous solids has been investigated. It is also hypothesized that the struvite fines may be swept along with the stillage solids used as seed and this in turn decrease s the settling time from days to minutes. Research Objectives The ov erall objective s of this study were to 1. assess the feasib ility of struvite precipitation from raw and anaerobically digested cellulosic ethanol stillage, 2. develop a sequential batch reactor process incorporating stillage solids to enhance settling of struvite particles, and 3. study the leachability of nutrients fro m processed stillage containing struvite after application as soil amendment, Materials and Methods Feed Wastewater: Struvite precipitation and recovery were carried out on two wastes, 1. Raw cellulosic ethanol stillage obtained from the Biofuels pilot plan t at the University of Florida, and 2. Anaerobically digested stillage.
69 As described in C hapter 2, stillage was coarsely sieved for solid liquid separation following which it was anaerobically digested in a fluidized bed reactor under thermophilic conditions Digested stillage effluent was used as feedstock in part 2 Reactor Design Struvite precipitation from stillage was carried out in a cylindrical pyrex glass reactor (7L) with a conical base (2 L). The schematic representation and a photograph of the seq uential bat ch reactor is presented in Figure 3 respectively for sample col lection. The reactor was covered with a plastic lid equipped with 3 ports for air inlet, feed supply and sample collection. The settled solids were drained from the bottom of the reactor that was attached to a PVC pipe fitted to a ball valve via an O ring. Reactor Operation As shown in Figure 3 1 (b) in a sequential batch reactor ( SBR ) all four stages; fill, react, settle and draw occur in one single reactor in a sequential manner. Stillage was processed in the sequential batch reactor shown in Figure 3 2. The reactor was filled with 8 L of cellulosic ethanol stillage (raw or digested) and the pH was adjusted using 5N sodium hydroxide to 8.9, the optimum for struvite precipitation (Gadekar and Pullammanappallil, 2010). The dosage requirements of magnesi um chloride and sodium hydroxide were determined using a chemical equilibrium model developed to determine the yield and purity of struvite precipitated from various wastewaters (Ram Mohan et al. 2011). The contents were mixed by sparging air through the bottom of the reactor and struvite precipitation reaction was allowed to
70 proceed for 30 minutes. At the end of the reaction time, air was turned off and the precipitate and other solid organic ma tter in the stillage were allowed to settle. When raw stil lage was used as feedstock, following the 30 minute reaction time solids were allowed to settle down for 24 hours and then the supernatant was decanted and the settled sludge was used for agronomic studies carried out in collaboration with the Soil and Wat er Science department at the University of Florida. In this experiment, stillage solids were not separated from the liquid prior to struvite precipitation. S o the resulting settled material included organic materials and other constituents contained in th e raw stillage. Therefore, the settled sludge does not qualify as a pure struvite, but rather as a struvite throughout this work. Struvite precipitation from anaerobically digested stillage was carried out in three ways. F rom coarsely sieved stillage (devoid of stillage solids) that was anaerobically digested F rom anaerobically digested stillage seeded with 1% (wet w/v) stillage solids and F rom anaerobically digested stillage seeded with 10 % (wet w/v ) stillage solids. T he stillage solids collected after sieving prior to anaerobic digestion were used in the second and third treatments The coarse solids were added to the reactor along with stillage prior to aeration. After 30 minutes of reac tion time, the processed sludge was allowed to settle. Samples were collected at 0, 0.25, 4, 24 and 72 hour periods from the top of the reactor and just above the interface of settled solids and liquid and analyzed for ammonia, phosphate and total suspend ed solids. Struvite containing sludge collected from the bottom of the reactor was also analyzed for total suspended
71 solids content. Ammonia nitrogen was analyzed using the Hach test kit and phosphate was analyzed using the ascorbic acid method explained i n the 18 th edition of Standard Methods book (APHA, 1992). Total solids content was determin ed by drying the samples at 105 C in an oven for overnight. The pH was measured using an Orion benchtop pH meter. All the above experiments were carried out in dup licates and the reported values are average of the results obtained from the repeat experiments. Results Characterization of Cellulosic Ethanol Stillage Raw stillage and anaerobically digested stillage were analyzed for ammonia nitrogen and orthophosphat e phosphorous content. Raw stillage obtained from the Biofuels plant had a pH of 6.2 and NH 3 N and PO 4 P concentrations of 265 35 and 779.5 29.5 mg/L respectively (Table 3 1) The total solids concentration of raw stillage was 12 2 .03 %. This was used in part 1 of this study as whole stillage without solid separation. In the agronomic studies carried out by Agyin Birikorang et al. (2010) on processed sludge (which was the settled sludge containing organic compounds as well as the mineral precipitates) w as collected together and used as soil amendment. Although struvite precipitation from raw stillage helps recover N and P, the high soluble COD of the waste is not affected in the process. Thus additional biological treatment methods such as anaerobic o r aerobic treatment may be required post struvite precipitation. Such a post biological treatment would also require addition of nutrients for microbial growth. Therefore, it would be logical to anaerobically digest the high strength wastewater prior to re covery of nutrients. Anaerobic digestion of coarse separated stillage was carried out in a fluidized bed reactor as explained in Chapter 2,
72 to convert the soluble COD into biogas. The anaerobic effluent had a pH of 7.70.8 and NH 3 N and PO 4 P concentration s of 302 17.5 and 559 32 mg/L respectively. The total solids content of the coarse separated stillage influent and anaerobically digested effluent were 40.1 % and 20.4 % respectively. Struvite Sludge from Raw S tillage Struvite precipitation was carr ied out using 8L of raw whole stillage in a SBR. After allowing the processed sludge to settle for a period of 24 hours, the supernatant was decanted and the solids were collected form bottom of the reactor. Mineral composition of raw stillage and processe d sludge at the Soil and Water Science department at the University of Florida. Stillage obtained from the Biofuels pilot plant was acidic in nature with a pH of 6.42 0.86. Gadekar (2010) studied the significance of pH to obtain higher yields of struvite from various wastewaters. Based on this work, the pH was raised from 6.42 0.86 to 8.9 using 5N sodium hydroxide. A chemical equilibrium model based on charge and mass balances developed by Gadekar (2010) to predict the yield and purity of struvite precip itated from wastewater streams was used to determine the amount of alkali and magnesium required to recover N and P as struvite. Using this model it was determined that 2.3 g/L of NaOH and 350 mg/L of magnesium would be required to precipitate N and P from stillage as struvite (Ram Mohan et al., 2010). The initial N and P concentrations in raw stillage were 65.85.68 and 35.234.3 g (Kg stillage) 1 respectively. Among the total N, individual ammonia, nitrate and organic N concentrations were determined to be 23.63.01, 33.63.62 and 10.10.94 g Kg 1 of raw stillage respectively. Elemental analysis was also carried out on cellulosic ethanol stillage by Agyin Birikorang et al. (2010 ). Iron,
73 Aluminum, Potassium, Calcium, Magnesium and Sulfur content of raw s tillage is listed in Table 3 2. Following the characterization of waste feedstock, stillage was subjected to struvite precipitation in a SBR. After 30 minutes of reactor time, the processed sludge was allowed to settle for ~1 hour. Samples were collected from the settled sludge that occupied about 35% of the total feed wastewater volume and were analyzed for percent solids, N, P, ammonia, nitrate, organic N and elemental composition. The total suspended solids content of settled processed sludge collected from the bottom of the reactor was 28 3.14 %. A mass balance on NH 3 N and PO 4 P shows that about 76% (NH 3 ) and 86.5% (PO 4 ) were recovered in the processed sludge leaving behind a residual 23% (NH 3 ) and 14.5% (PO 4 ) nutrient concentration in the supernatant Although 86.5% of P was recovered in the process, the major obstacle in carrying out this procedure lies in obtaining complete mixing via aeration. Due to the presence of stillage solids, much difficulty was faced with during aeration as well as sludge c ollection and reactor clean up as the sludge tends to clog up the sample collection ports. The processed sludge collected at the end of 1 hour settling period was used in a greenhouse study conducted by Agyin Birikorang (2010), at the Soil and Water Scien ce department at the University of Florida on sweet sorghum, a bioenergy crop This study was carried out to investigate the agronomic and environmental effectiveness of application of raw stillage versus processed sludge as soil amendment. Struvite Slud ge from Anaerobically Digested Stillage Tian (2011) showed that 70% of the methane potential of whole stillage is generated from stillage filtrate that is obtained post separation of the fibres or solids
74 from whole stillage. Therefore, in this study stilla ge solids were separated by passing whole stillage through a 0.425 mm sieve and the solids retained in the sieve were stored for later use. 2L of anaerobically digested stillage was filled in the SBR, pH was adjusted from 7.7 to 8.9 using 2.3 g/L of NaOH, 350 mg /L of Mg was added as MgCl2. 6 H2O (determined using the model) and the mixture was aerated for 30 minutes and allowed to settle for 24 hours prior to sample collection and analysis. The supernatant had 55.851.2 and 41.93.7 mg/L of ammonia and phosph ate respectively. The processed sludge contained precipitated struvite and was collected at the bottom of the reactor for total suspended solids analysis. About 2.25 0.2 g/L struvite containing sludge was obtained from these experiments. The results from the model used to determine the yield of struvite agreed well with the experimental results. The error between the model predictions and measured total suspended solids was 12.6%. However, one obstacle that had to be overcome in this process was the produ ction of fines that caused hindered settling and therefore need for an additional separation phase. In previous studies, use of quartz, granite chips and struvite fines as seed material has been investigated to study their effect on crystal growth and set tling. Wang and Burken (2003) showed that addition of seed material enhanced P removal via struvite precipitation up to 23 83%. While granite performed poorly at low mixing intensities (200 ml air/min), sand and struvite fines were shown to have performed equally good at any given mixing intensity. Contrary to the idea that large seeds perform poorly as a seeding material due to smaller available surface area, it was shown that P removal was higher with larger seed size at any given mixing intensity.
75 In th is study addition of small amounts of stillage solids as seed was tested for improved settling of processed sludge. The top port of the reactor was used to collect samples of the supernatant, the bottom port was used to collect liquid samples from the liqu id settled solids interface and processed sludge was collected at the bottom end of the reactor. The first experiment was conducted with no seeding. In t he second experiment, 1% (wet w/v ) stillage solids were used as seed in struvite precipitation trials. In the third set of experiments, 10% (wet w/v ) solids were used as seeding material. In all the trials, samples were collected at 0, 0.25, 4, 24 and 72 hours and analyzed for residual NH 3 N, P O 4 P and TSS. Table 3 3 list s the results of residual NH 3 N, P O 4 P and TSS measured at the end of struvite precipitation process for 0%, 1% and 10% seeded experiments, respectively. The total suspended solids settled at the bottom end of the reactor was 63 3.5 % higher in case of seeded experiments as compared to pro cessed sludge from unseeded AD stillage trials. The percent increase in solids was calculated after subtracting the amount of stillage solids added to AD stillage. Thus any increase in TSS would correspond to precipitated minerals. Correspondingly, residua l phosphate content in the supernatant was reduced to 41.93.7%, 0.30.05%, 2.70.2% in the 0%, 1% and 10% seeded experiments. Results on P removal show that about 99.70.2 % of phosphate was removed at the end of struvite precipitation process on adding stillage solids compared to 94.6% P removal from AD stillage without solids. Results obtained in the present study were in good agreement with observations made by Wang and Burken (2003) on seeding the struvite precipitation process. In the present study, du e to production of fines that did not settle under gravity, a post filtration step was required even after 24 hours of settling
76 time. In the subsequent experiments, use of stillage solids that were obtained from coarse separation prior to anaerobic diges tion, were used tested for use as seed material. Although mixing strength was maintained constant, the amount of seed material was varied to determine an optimum amount of seed that could promote crystal growth and settleability. Results from Table s 3 4 a nd 3 5 present the data on N and P removal. It can be seen that the difference in P or N removal were not significantly different for 1% and 10% (wet w/v) seeded experiments. Supernatant samples collected from the top port in case of both 1% and 10% seede d experiments had a TSS content of ~1 1.2% after 15 min of settling and showed no further change even after 3 days. Samples collected from the bottom port near the liquid settled sludge interface also had a TSS content of ~10.1% in samples analyzed at 15m in, 4hr, 24hr and 3 days. This showed that about 95% of settling occurred within the first 15 minutes showing that use of seed material improved settleability of the processed sludge. Data in Table s 3 4 and 3 5 also list the residual NH 3 N and PO 4 P conce ntrations of the supernatant samples collect from the two ports at different settling times. In 1% seeded experiments, residual phosphate and ammonia levels were reduced to <0.1 mg/L and 160 mg/L respectively, after 15 minutes of settling and remained unc hanged throughout. While in case of 10% seeded experiments, although the final phosphate and ammonia levels were comparable with the former, it took about 3 days to achieve these concentrations. Kim et al., (2006) studied the use of varying quantities of seed for enhancing struvite precipitation. Preformed struvite crystals at concentration varying from 0 g/l to 40 g/l were tested for use as seed. It was speculated that while at lower seed concentrations, struvite precipitation proceeded via crystal nuclea tion and crystal
77 growth, at higher seed concentration, crystal growth was found to be more dominant. So, it can be hypothesized that while nucleation followed by crystal growth is the mechanism of struvite precipitation in 1% seeded experiments, the proces s shifts more towards crystal growth in case of 10% seeded experiments. Although no significant difference was observed with respect to settling in either case, P recovery was faster in 1% seeded experiments. Therefore, it can be concluded that addition of 1% (wet w/v) seed was enough to cause 99.9% P removal, 56% N removal and promote crystal growth and faster settling (95%) within 15 minutes. Discussion Agronomic Efficiency of Processed Sludge Containing Struvite Results from greenhouse studies conducted out by Agyin Birikorang et al., (2010 ) on application of raw stillage vs. processed sludge as soil amendment for cultivation of sweet sorghum has been discussed in this section. The agronomic and environmental effectiveness of the raw stillage and process ed sludge as potential plant nutrient sources were compared to inorganic fertilizers (ammonium nitrate + tri ple superphosphate), biosolids and manure. Two biosolids (G ainesville R egional U tilities (GRU) biosolids and Milorganite biosolids) of different phy toavailability, and poultry manu re were selected for the study (Agyin GRU biosolids was obtained from the water reclamation facilities of the Gainesville Regional Utiliti es (Gainesville, FL) and was produced through aerobic digestion. The GRU biosolids had a high soluble P content (~7 g kg 1 ). Milorganite biosolids, obtained from Milwaukee Metropolitan Sewerage District, Milwaukee, WI, was generated from anaerobically dige sted material that was heat dried and pelletized. Milorganite has a low
78 soluble P content (~0.1 g kg 1 ). Milorganite biosolids was stabilized with iron salts to decrease its P solubility. Poultry manure utilized for the study was obtained from an egg produ cing farm in Indiantown, FL. The nutrient sources were mixed with 4 kg of A horizon Immokalee soil at the recommended N application rate of 150 kg PAN ha 1 for sweet sorghum (Mylavarapu et al., 2007). The plant available N (PAN) was calculated based on the inorganic N content of the nutrient sources, and an assumed mineralization of the organic N content. The stillage manure and biosolids amended soils were equilibrated (~80% water holding capacity) in zip lock plastic bags for 2 weeks in the laboratory p rior to use in the greenhouse. Nitrogen uptake efficiency was found to be similar between the treatments with inorganic fertilizer (0.35 g g 1 ) and the processed stillage (0.36 g g 1 ). The N uptake efficiency of both treatments was shown to be greater than that of the manure (0.31g g 1 ), the Milorganite biosolids (0.29g g 1 ) and GRU biosolids (0.28 g g 1 ). The least N uptake efficiency of ~0.27 g g 1 occurred in the treatment with the raw stillage material. Results from this study showed that p rocessing raw stillage increased N uptake efficiency by an average of ~0.95 g g 1 compared to raw stillage when applied as soil amendment Phosphorus uptake efficiency of the processed stillage treatment (~0.43 g g 1 ) was statistically similar to the P uptake efficienc y of the inorganic P fertilizer and the organic nutrient sources (ranging between ~0.40 g g 1 and ~0.44 g g 1 ). Similar to N uptake efficiency, P uptake efficiency of the processed s ludg e containing struvite was significantly greater than that of the raw s tillage. Processing the raw stillage, thus,
79 improved the efficiency of the material in releasing nutrients to the plant s, compared to the raw stillage. Amount of Nitrogen and Phosphorus in Leachate Improved N and P uptake efficiency, through the use of pr ocessed sludge significantly reduced masses of N and P los t via leaching. Without processing, ~55 mg of the N applied from t he raw stillage leached out. Processing of stillage however reduced the mass of N leached from ~55 mg to ~32 mg. These results are i n agreement with work done by Ren et al. (2010) who showed that total N loss was reduced from 35% to 12%, when struvite was used as a nutrient source relative to unamended control. C onsistent with P uptake a significant reduction in mass of P leached also was observed when the processed sludge was used as the nutrient source. With the exception of the inorganic P source, application of the other nutrient sources (stillage, biosolids, and manure) at N based rates resulted in differential total P application rates (189 kg P ha 1 Raw and 168 kg P ha 1 for processed stillage; 154 kg P ha 1 for GRU biosolids; 133 kg P ha 1 for Milorganite biosolids; and 113 kg P ha 1 for manure). The greatest percent P leached (~51%), occurred with the raw stillage, followed by the inorganic nutrient source (49%), GRU biosolids (~45%), a nd manure (~36%) respectively. The least percentage of the applied P measured in the leachate (~31%) occurred in the treatments that received processed sludge and Milorganite biosolids. The P le aching data were consistent with the c alculated (PWEP) values (Table 3 2 ), and also consistent with the observations of Brandt et al. (2004) and Agyin Birikorang et al. (2008) that PWEP values of P sources are strongly related to off site P losses.
80 Closing Remarks Potential for recovery of N and P as struvite from raw and anaerobically digested cellulosic ethanol stillage was successfully investigated in a sequential batch reactor. The struvite precipitation from raw and anaerobically digested stillage was investigated at a pH of 8.9 by allowing the reaction to proceed for 30 minutes after dosing the reactor with required alkali and Mg source and finally allowing the sludge to settle for a set period of time. While 86.5% P recovery was obtained via struvite precipitation from raw stillage, it posed clogging problems due to presence of higher amounts of stillage solids. On the other hand, while 97% P removal was obtained from anaerobically digested stillage via struvite precipitation, lack of any stillage sol ids led to production of unsettled fines. Studies were carried out to investigate the effect of seeding the process to improve struvite crystal growth and settleability. Results showed that use of 1% stillage solids as seed improved N and P removal, increa sed the yield of of struvite containing sludge and positively impacted settling of processed sludge with more than 95% of settling occurring within the first 15 minutes. Processing raw stillage also resulted in improved nutrient uptake efficiencies of N a nd P: 0.36 g/g and 0.4 g/g respectively that were comparable with that of inorganic fertilizers and much higher than results obtained from application of raw stillage as nutrient source for cultivation of sweet sorghum, a bioenergy crop. Application of ra w stillage vs. processed sludge as soil amendment reduced the mass of N and P leached out by 42% and 20% respectively.
81 Table 3 1. Characterization of Feedstock Parameter Raw Stillage Anaerobically Digested Stillage pH 6.2 0.3 7.7 0.8 Ammonia Nit rogen (mg/L) 265 35 302 17.5 Orthophosphate Phosphorous (mg/L) 779.5 29.5 559 32 Mg Required (mg/L) 350 450 75 Alkali Required (g/L) 2.3 1.06 *Determined using mathematical model. Table 3 2. Characterization of processed stillage for use as s oil amendment Property Processed stillage pH 8.71 0.38 Solids (%) 28.0 3.14 Nitrogen (g kg 1 ) 69.2 7.24 Ammonium N (g kg 1 ) 25.2 3.14 Nitrate N (g kg 1 ) 35.3 4.02 Organic N (g kg 1 ) 9.74 1.02 Total iron (g kg 1 ) 0.21 0.03 Total alumi num (g kg 1 ) 0.11 0.02 Total P (g kg 1 ) 34.3 1.84 WEP (g kg 1 ) 0.29 0.01 PWEP (%) 0.16 0.02 Mehlich 3 P (g kg 1 ) 18.2 1.98 Mehlich 3 K (g kg 1 ) 4.58 0.62 Mehlich 3 Ca (g kg 1 ) 0.96 0.04 Mehlich 3 Mg (g kg 1 ) 8.98 1.02 Elemental S (g kg 1 ) 7.04 0.92 Reference: Agyin Birikorang, et al. (2010)
82 Table 3 3. Nutrient Recovery from Anaerobically Digested Stillage Parameter AD Stillage+ 0% seed AD Stillage+ 1% seed AD Stillage+ 10% seed Residual NH 3 N (mg/L) 55.81.25 1536.2 1 607.2 Residual PO 4 P (mg/L) 41.93.7 0.30.05 2.70.2 Total Suspended Solids (g/L) 2.250.2 5.60.25 6.80.75 Table 3 4. Characterization of Processed Stillage from AD stillage+1% (wet w/v) stillage solids Settling time (hrs) Sampling location NH 3 N (mg/L) PO 4 P (mg/L) TSS (%) 0 Top 180 38.5 1.16 0.25 Top 160 0.3 1.01 4 Top 140 0.593 1.001 4 Bottom 140 <0.1 1.001 24 Top 160 0.297 0.85 24 Bottom 150 <0.1 1.03 72 Top 140 <0.1 0.97 72 Bottom 160 <0.1 1.01 The values presented here are average o f duplicate runs with st andar d dev iation was <10%. Table 3 5. Characterization of Processed Stillage from AD stillage+10% (wet w/v) stillage solids Settling time (hrs) Sampling location NH 3 N (mg/L) PO 4 P (mg/L) TSS (%) 0 Top 170 61.9 1.61 0.25 Top 210 77.7 1.22 4 Top 180 23.1 1.14 4 Bottom 190 30.1 1.16 24 Top 180 18.8 1.15 24 Bottom 190 16.2 1.11 72 Top 130 2.7 1.07 72 Bottom 160 2.95 1.12 The values presented here are average of duplicate runs with st andar d dev iation was <10%.
83 Figure 3 1. (a) Schematic diagram (left) and Photograph (right) of a Sequential Batch Reactor Figure 3 1. (b) Working of a Sequential Batch Reactor
84 Figure 3 2. Settleability of p rocessed stillage from anaerobically digested stillage effluent seeded with stillage solids.
85 CHAPTER 4 PHOTOCATALYTIC TREATMENT FOR FINAL POLISHING OF WASTEWATER FROM A CELLULOSIC ETHANOL PROCESS; EFFECT OF INITIAL COD, PH, DEPTH OF STILLAGE Summary Ad vanced photocatalytic oxidation of lignocellulosic ethanol stillage was studied, using TiO 2 coated reusable sheets (TiO 2 loading : 3 mg/cm 2 ) activated by 180 nm ultraviolet light, for simultaneous reduction of organics and decolorization. Experiments were performed to study the effect of so luble chemical oxygen demand (sCOD), depth and pH. Organic concentration of 3000 mg sCOD/L in anaerobically digested effluent was an obstacle for efficient photocatalytic treatment. Therefore, anaerobically digested still age was further treated in an aerobic biological process to reduce sCOD to 1350 mg/L The extent of decolorization was 91.3%, 91.5% and 77% after 22, 55 and 80.5 hours for 1, 3 and 5 mm depths respectively. The final treated s COD was 8 0 m g/L. Significant decolorization was observed only after the sCOD dropped below 6 00 m g/L. There was very little decolorization (~25%) at pH of 9 or above. If pH was adjusted and maintained at neutral value, 92% decolorization was observed. Background As discussed in Chapt er 2, most of the organic carbon in the liquid fraction of the lignocellulosic ethanol stillage can be removed through anaerobic biological treatment. However, even after this the effluent is dark colored and has an organic content of 3500 mg sCOD/L. Adv anced oxidation processes have been actively studied for decolorization of textile wastewaters rich in colored dyes, dark colored high strength distillery wastewaters, wastewaters rich in phenolic compounds, and pulp and paper industry effluents ( Kiriakidou, Kondarides et al. 1999 ; A. K Rajvanshi and Nimbkar 2004 ;
86 Boyd and Almquist 2004 ; Fotiadis, Xekoukoulotakis et al. 2007 ; Baransi, Dubowski et al. 2012 ) The objective of an AOP des ign is to generate OH radicals that non selectively attack compounds that may otherwise be recalcitrant to conventional oxidation ergy sources such as ultraviolet radiations ( Radjenovic, Sirtori et al. 2009 ) However, due to the high costs associated with ozone generation, and use of H 2 O 2 economica lly favorable for wastewater treatment. Photocatalysis is a process where a semiconductor such as titanium dioxide acts as a catalyst in producing free radicals, mainly hydroxyl (OH ), upon absorption of photons from UV radiation. The hydroxyl radicals pa rticipate in the decomposition of organics in wastewaters. This process has a number of advantages including low capital costs, ambient temperature operation, and eco friendliness since it can be mediated using sun light or near UV irradiation. TiO 2 posses ses other advantages such as being very stable in aqueous environment, non toxic and insoluble in dark as well as illuminated environments and can therefore be recovered at the end of the process as it does not undergo any change by itself, but rather acts as a catalyst ( Lakshmi, Renganathan et al. 1995 ) In this work use of TiO 2 photocatalysis has been investigated for decolorization and degradation of recalcitrant organics from biologically treated cellulosic ethanol stillage.
87 Materials and Methods Wastewater The decanted wastewater obtained after struvite precipi tation (Chapter 3) was further treated usi ng a photocatalytic process It should be noted that this wastewater was anaerobically digested (Chapter 2) before being used for struvite recovery. The decanted wastewater was used in three ways for photocatalytic treatment. First the decanted wastewater was used without any further processing for photocatalytic treatment. In the second approach, the wastewater was diluted to reduce the COD and then subjected to photocatalytic treatment. In the third approach, the wastewater was anaerobically treated to reduce the COD and then used for photocatalytic treatment. Titanium dioxide coated sheets In the literature advanced oxidation studies using TiO 2 have mostly been done with powdered catalyst suspended and agitated in the wastewater ( Fotiadis, Xe koukoulotakis et al. 2007 ) However, s urfaces coated with TiO 2 may be a better system for easily scaling up the technology. Hence in the present study TiO 2 coated on flexible plastic sheets using a proprietary technology developed and supplied by Innova tive Materials Development Company (IMDC), Gainesville, Florida were used. The TiO 2 content was 3 mg TiO 2 /sq cm of the sheet. UV lamps UV lamps were also provided by IMDC. Three 15 W, 180 nm ultraviolet lamps, Photoreactor As sembly The three lamps spaced 5 cm apart was assembled together using a standard size holder. A 33 cm x 23 cm tray was used to hold the TiO 2 coated sheet and the
88 treated stillage. The entire assembly (lamps and tray) was placed inside a box. The distance between the UV lamps and surface of tray was about 5 cm and this was kept constant throughout the study. A weighing balance was used to measure the mass of the tray holding the TiO 2 shee ts and stillage effluent. All photocatalytic experiments were carried out in a batch mode. A TiO 2 coated sheet was placed on the tray. A known volume of the treated stillage was poured over the sheet and allowed to equilibrate in the dark for an hour prior to UV exposure. The pH of the treated stillage was adjusted by drop wise addition of 5N sulfuric acid or 5N sodium hydroxide and monitore d using Orion Benchtop pH meter The progress of organics degradation and decolorization of the sample was monitored by analyzing samples exposed for varying time periods for sCOD and abs orbance values. Control experiments were carried out with (i) TiO 2 coated sheets in dark, as well as with (ii) only UV lamps in the absence of TiO 2 coated sheets. Samples in these experiments were exposed for 60 hours. The experiments were repeated at th ree different depths: 1, 3 and 5 mm of stillage. Mass of the entire set up was maintained constant by adding DI water to make for water evaporated during experiment. Chemical oxygen demand is the total quantity of oxygen required to oxidize organic and som e inorganic matter in a given wastewater sample, whilst total organic carbon measures the organic matter that can be completely oxidized into CO 2 Though TOC tests can be performed in the order of few minutes, they require sophisticated equipment. Many r esearchers have studied COD/TOC correlation, to determine the prospects of substituting COD in place of TOC. It has been shown that influent and
89 effluent wastewater samples have a consistent average COD:TOC ratio of 3 ( Dubber and Gray 2010 ) In this paper soluble COD (sCOD) measurements were used to determine organic carbon content. sCOD was measured colorimetrically using a HACH kit on filtered s amples and absorbance was measured using UV VIS spectrophotometer. The stillage sample was diluted appropriately and tested for maximum absorbance by running a UV VIS scan between 200 and 800 nm wavelength. Maximum absorbance ( was recorded at 350 nm an d this wavelength was chosen for measuring color intensity of samples. Percent decolorization and COD removal were determined using the following equation: % decolorization = 100* (A o A)/A o % COD removal = 100* (C o C)/C o where A o A and C o C are the absorbance and sCOD measurements at t=0 and at time t respectively. Results Effect of Initial sCOD Concentration At first photocatalysis experiments were carried out on anaerobically digested stillage effluent. The effluent from the anaerobic digester h ad a residual sCOD of 3000 mg sCOD/L. Photocatalytic degradation studies were performed using anaerobically digested stillage effluent samples filled to a depth of 1 mm for a period of 60 hours. In the first 20 hours, an exponential degradation pattern was observed removing only 24.6% of the initial sCOD. After this period no further sCOD decrease was observed even after 60 hours. The failure of photocatalytic oxidation to considerably reduce sCOD was hypothesized to be due to the high initial sCOD contri buted by various organic molecules.
90 In the second approach, photocatalytic expe riments were carried out using 3 mm depth of diluted anaerobically digested stillage (whose initial sCOD was reduced from 3500 mg/L to 1000 mg/L) for a pe riod of 55 hours. Fi gu re 4 1 presents the decolorization and sCOD profile s of photocatalytically treated diluted AD stillage. In the first 3 0 hours, an exponential degradation pat tern was observed removing about 60 % of the initial sCOD and causing about 90% decolorization S ubsequently, the photocatalytic experiment was continued for another 25 hours. H owever no significant color or sCOD removal was observed during this period Although dilution reduced the concentration of color causing organic compounds; lack of transforma tion of these compounds to faster degrading intermediates led to a decrease in percent COD removed from the treated samples Many researchers have studied the effect of initial concentration of organics on photocatalytic degradation of various wastewaters such as pulp and paper mill, olive mill, cottonseed processing, textile and distillery wastewaters. ( Kiriakidou, Kondarides et al. 1999 ; A. K Rajvanshi and Nimbkar 2004 ; Boyd and Almquist 2004 ; Fotiadis, Xekoukoulotakis et al. 2007 ; Baransi, Dubowski et al. 2012 ) It has been observed that at higher concentrations the rate of degradation is adversely affected. This phenomenon has been explained on the basis of availability of catalyst surfa ce for redox reactions. Increasing concentrations causes multilayer formation of adsorbed molecules which in turn limits their contact with photo induced holes or hydroxyl radicals. W hen photocatalytic degradation is carried out under lower organic concen trations, no multilayers are formed and the reaction rate is higher Typically the wastewater is diluted to reduce concentration of organics prior to photocatalytic
91 treatment. This approach has shown to improve degradation of organics in certain type of w astewaters generally textile effluents containing dyes. However, dilution does not improve degradation of complex mixture of organics, like for example anaerobically digested effluents (Raj vanshi and Nimbkar 2004) Therefore, it appears in addition to co ncentration of organics the nature of these organic compounds would also affect its degradation during photocatalytic treatment. Apart from dilution aerobic biological treatment can also reduce organics content in anaerobically digested effluents. In thi s study the effect of dilution was compared with aerobic treatment as a means of reducing organic content to enhance subsequent degradation of organics and color by photocatalysis. In the third approach, a naerobically treated stillage effluent was further treated in a short aerobic biological treatment step using activated sludge obtained from the Water Reclamation facility at University of Florida. Aerobic post treatment of anaerobically digested effluent is typically employed in wastewater treatment pla nts to lower sCOD prior to disposal. A number of aerobic treatment trials were performed to generate treated wastewater samples for the photocatalytic polishing step. Aerobic treatment reduced sCOD from 3000 mg/L to between 1000 and 1400 mg/L. sCOD Reduct ion Figure 4 2 shows the sCOD profile from 3mm depth experiments. The average of measured sCOD values from duplicate experiments is plotted. The sCOD of sample was reduced by 91% within 55 hours of exposure. Similarly greater than 90% COD removal wa s obtained in the experiments using treated samples filled to depths of 1 and 5 mm (sCOD data not shown), however the duration of exposure to remove this sCOD was different. Two different control experiments were performed including Control 1:
92 where the sa mple was treated using the TiO 2 coated sheets in dark and Control 2: where the sample was exposed to UV radiation in the absence of TiO 2 sheets. Figure 4 3 presents the percent COD removal from the control experiments with 3 mm depth of stillage. The cont rol experiments led to 6 and 17 percent sCOD removal respectively. sCOD removed in the dark control experiment can be due to adsorption of stillage onto the TiO 2 coated sheets rather than sCOD destruction. UV light by itself appear s to remove some sCOD. It can be inferred from the results that photocatalysis experiments produced much higher COD and color removal as compared to the control experiments, thus showing the need for both TiO 2 and UV for successful sCOD degradation. Extent of D ecolorization Ce llulosic ethanol stillage is a dark colored effluent and hence color removal is a vital step to recycle water recovered from the process. Even after anaerobic digestion and aerobic treatment, the dark color persisted in the samples. Extent of decolorizat ion was measured as change in absorbance of the treated samples obtained after various exposure times. Negligible color removal was observed in both the control experiments. Figure 4 1 presents the average absorbance values of samples from duplicate 3 mm depth experiments. Figure 4 3 presents the decolorization profile of samples from 1, 3 and 5 mm depth experiments and Figure 4 4 presents the percent decolorization vs. time period of photocatalysis for 3 mm depth of stillage samples An exponential t rend was seen in the decolorization pattern. For 1, 3 and 5 mm depth trials, 91.2, 91.5 and 77 % decolorization was obtained after 22, 55 and 80.5 hours of exposure respectively. sCOD Absorbance C orrelation To polish stillage for water reuse, it is necess ary to reduce both the organic content and the color. Even though color contributing organic molecules also exert a
93 sCOD, it is possible that not all organic molecules in the treated stillage contribute to color. To verify this, sCOD values were plotted against absorbance measured in the same samples (Figure 4 5). At lower sCOD values indicated as zone 4(< 150 mg/l), a direct variation is evident with a correlation coefficient of R 2 =0.9, showing that compounds that contribute to COD may also contribute to the colored appearance of the sample. However, at higher COD values the direct variation is not applicable even though sCOD and absorbance may be linearly correlated. This trend may have occurred as a result of different groups of molecules being degraded at different time periods. Soluble small chain organic molecules degrade earlier in the photocatalytic experiments leading to a noticeable decrease in sCOD (Zone 1in Figure 4 5) in the first two hours that may not necessarily contribute significantly to d ecolorization as seen by a steep change in COD with very little change in absorbance. After the first 2 hours, a significant decrease in color is seen, (Zone 2, Figure 4 5) while the sCOD only reduced by 200mg/l. TiO 2 treatment transforms high molecular we ight compounds to smaller chain intermediates before they undergo complete oxidation. The profile in zone 3 is similar to that of zone 1 where a large decrease in sCOD is accompanied by a small change in absorbance. Long chain phenolic compounds contribut ing to color degrade at a later time period in the study due to their recalcitrant nature. In summary for low values of absorbance the sCOD of the treated sample is also low and a direct variation exists between sCOD and absorbance. Effect of I nitial pH The pH of the wastewater samples i s a critical factor in all photocatalytic experiments as it influences the rates of reaction ( Lakshmi, Renganathan et al. 1995 ; Ehrampoosh, Moussavi et al. 2011 ) 1 mm depth trials with treated stillage were carried
94 out with and without pH adjustment. Figure 4 6 present s the change in pH with time for both the trials. T he percentage sCOD removed in these experiments is shown in Figure 4 2. Results show that performing photocatalytic trials on stillage effluents at a basic pH value of 9.6 or higher lowers the rates of organics degradation. Only about 25% of sCOD was deg raded after 50 hours of exposure. However, on adjusting the initial pH to about neutral, much higher percentage of organics were degraded (92%). This trend was applicable for all the depths studied. Discussion A first order model was fit to the absorbanc e data shown in Figure 4 3. The goodness of fit of the first order model to the absorbance data was studied on the basis of R 2 value. R 2 values were determined to be 0.83, 0.96 and 0.95 for 1, 3 and 5 mm depth studies, indicating a good fit of data. The first order model predicted the reduction in absorbance with time of exposure. The first order rate constants; k 1 (1 mm depth), k 3 (3 mm depth) and k 5 (5 mm depth) were 0.15, 0.034, 0.08 h 1 respectively. Results showed that the reaction constant (k) va lues were inversely related to the depth used in the study. For example, k 3 was one third the value of k 1 and k 5 was approximately one half of k 3 The experimental results obtained in this study using AAS were compared to that reported in the literature f or other types of wastewater. It should be noted that majority of the research work reported in the literature used powdered TiO 2 agitated together with wastewater sample. Ehrampoosh et al., obtained 98% decolorization of methylene blue from a 15 mg/l sam ple within a period of 56 minutes ( Ehrampoosh, Moussavi et al. 2011 ) In another study, ninety three percent color removal was obtained within 4 hours with 0.5 g/L TiO 2 in textile wastewater samples containing dyes like remazol black, red,
95 golden yellow, cibacro n black, drimaren red, scarlet and yellow with an initial sCOD of 404 mg/l ( Pekakis, Xekoukoulotakis et al. 2006 ) Similar extent of color removal was obtained in this work, but after a longer time of exposure. This was due to a higher initial organics concentration. Watts et al., studied the effect of pH, COD, ammonia nitrogen and suspended solids for the treatment of 1,2 bis(2 chlo roethoxy)ethane (BCEE) in secondary wastewater effluent using TiO 2 mediated photocatalysis ( Watts, Kong et al. 1994 ) Ammonia nitrogen concentrations up to 70 mg/L and suspended solids concentration up to 90 mg/L did not affect the photocatalysis process. Rapid rate of degradation were obtained at pH 4. The photocatalytic rates were inversely related to the sCOD concentrations of BCEE up to 164 mg/L. However, at sCOD levels higher than 164 mg/L, the reaction seemed to be completely quenched, showing that the initial organics concentration played a vital role in determining the rate and extent of photocatalytic degradation of organics in the treated sample The si gnificance of initial organic carbon concentration in sample was also studied by Huang et al., ( Huang, Le al et al. 2008 ) on measured as total organic carbon (TOC) and experiments were carried out at 5 and 10 mg/l TOC concentrations. Results from this study showed that ini tial TOC loading affected the rate of photocatalysis thus affecting the time period required to attain high percent of organic degradation. Fotiadis et al., investigated treatment of wastewater from cottonseed processing and showed that the initial COD con centrations played a significant role in determining the lag period of the reaction. In their work, photocatalysis was carried out for a period of 240 min at a TiO 2 catalyst dosing of 100 1500 mg/l in
96 powder form, on samples with initial COD concentration of 8000 mg/l that was diluted 10 fold to 800 mg/l ( Fotiadis, Xekoukoulotakis et al. 2007 ) Results reported on treated samples show that on diluting samples to a starting COD of 800 mg/l, complete mineralization was achieved leading to 95% COD removal. This study is indicative of the significance of initial COD loading in the photoreactors that determine the lag phase of the reaction. The authors of this work also suggested dilution as an option to overcome high COD levels. This is consistent with the trend attained in our system. Lik ewise in this study, anaerobic effluent samples with initial COD of 3000 mg/l were not easily degraded by photocatalytic treatment due to high initial loading of organic carbon. Hence stillage was diluted to reduce the initial COD and photocatalytically tr eated. Result s obtained in the present study we re in agreement with results obtained from a pilot scale study on photocatalytic treatment of diluted anaerobically digested distillery wastewater carried out by Rajvanshi and Nimbkar (2004) Their system was designed to degrade wastewater using solar irradiated TiO2 catalyst. 200 L/d of an aerobically digested distillery wastewater, diluted 5 times was photocatalytically treated. R esults showed that in spite of lowering the initial COD from 30000 mg/L to 6000 m g/L by dilution; TiO2 alone was not effective in causing organic removal. However, a combination of TiO2 and a coagulant resulted in 87% COD removal ( A. K Rajvanshi and Nimbkar 2004 ) As an alternative option, samples were exposed to a short aerobic treatment. C omple x organic molecules in the anaerobic effluent were transformed to intermediates easily degraded via photocatalysis and high rates of COD removal were witnessed within 55 hours of exposure to TiO 2 /UV treatment.
97 The pH of the sample has an effect on the ef ficiency of treatment. The zero point charge (pH zpc ) on semiconductor particles may be defined as the pH at which concentrations of protonated and deprotonated surface groups occur at equal concentrations. Studies by Khatamian et al., showed that the zero point charge of TiO 2 (Degussa powder) is 6.5 ( Khatamian, Daneshvar et al. 2010 ) As the pH zpc of TiO 2 molecules is about 6.5, at lower pH values, these molecules carry a positive charge while at pH > 7 they carry a negative charge. Lakshmi et al., investigated the degradation of methylene blue using TiO 2 mediated photocatalysis and showed that highest rates of degradation were obtained when the process was carried out at neutral pH ( Lakshmi, Renganathan et al. 1995 ) Gerischer et al., investigated the degradation of chloroethylamine organic molecule containing wastewaters and showed that a t a pH of 7, high organic degradation rates were achieved ( Gerischer and Heller 1991 ) This can be explained on the basis of the charge ca rried on TiO 2 molecules at neutral condition. Organic molecules such as methylene blue and chloroethylamine carry a positive charge that is readily accommodated in negatively charged sites on TiO 2 molecules at neutral pH following which OH radicals generat ed by photon induced electron hole pairs carry out oxidation of organic molecules. This phenomenon is described as an electrostatic model that explains the pH dependency of TiO 2 mediated photocatalysis. In the present study, effect of pH was studied basic and neutral conditions and it was shown that neutral conditions accelerated photocatalytic rates of organics destruction. A novelty of this work is in the use of reusable TiO 2 coated sheets, thus overcoming the obstacle of using and recovering large quanti ties of TiO 2 powder that
98 could enable easier scale up of this technique for large scale wastewater treatment. For systems that use TiO2 in powder form, there is a limit to the concentration of TiO 2 in suspension. High concentrations of TiO 2 lead to difficu lties in stirring the suspension. The concentrations reported in literature are in the range of 100 1500 mg/ L. In our system, the equivalent concentration of TiO 2 was 27, 9, 5 g/ L for 1, 3 and 5 mm depth of treated stillage. Closing Remarks The complet e photocatalytic treatment (decolorization and organics reduction) of anaerobically digested stillage was investigated using novel reusable TiO 2 coated she ets activated by 180 nm UV The initial sCOD of anaerobically treated stillage was too high for sign ificant reduction of organics. Therefore, the initial COD of the sample for treatment was further reduced to 1350 mg/L by aerobic biological treatment rather than dilution. Decolorization and sCOD reduction of anaerobically aerobically treated cellulosic ethanol stillage was successfully carried out. The sCOD was reduced to as low as 80 mg/L which was accompanied by greater than 90% decolorization. The degradation followed a first order kinetics. The first order rate constant was inversely related to the depth of sample. The pH of the samples also influenced the rate and extent of degradation. Neutral pH was best. Future work will investigate the kinetics of photocatalytic treatment using solar and UV (350nm) activated TiO 2 Additionally the effect of other composites of TiO 2 example silver doped TiO 2 will also be studied.
99 Figure 4 1. Absorbance and sCOD Profiles of TiO2 treated Diluted Anaerobic Stillage
100 Figure 4 2 sCOD and absorbance profiles for duplicate (Trail 1 and 2) 3 mm depth experimen ts
101 Figure 4 3 sCOD removal in various 3mm depth experiments. (i) Control 1: TiO2 in dark, (ii) UV exposure in the absence of TiO2 sheets, (iii) Experiment 1: No pH adjustment in aerobically treated samples, (ii) Experiment 2: Initial pH adjusted to 7 i n aerobically treated samples
102 Figure 4 4 Absorbance of samples from 1, 3 and 5 mm depth experiments plotted against time of exposure of samples
103 Figure 4 5 Extent of decolorization of photocatalytically treated stillage (3 mm depth)
104 Figure 4 6 sCOD Absorbance Correlation Zone 1, 3 Zone 2 Zone 4
105 Figure 4 7 pH profile in treated stillage. (i) Experiment 1: Without pH adjustment, (ii) Experiment 2: With initial pH adjusted to neutral.
106 CHAPTER 5 COMPARISON OF PHOTOCATALYTIC EFFICIENCY OF UNDOPE D AND SILVER DOPED TIO 2 UNDER UV A, UV C AND SUNLIGHT FOR COD AND COLOR REMOVAL FROM CELLULOSIC ETHANOL STILLAGE Summary Photocatalytic degradation of anaerobic aerobic treated cellulosic ethanol stillage has been studied using undoped and silver doped TiO 2 catalyst under UV A, UV C (15W lamps) and sunlight conditions. The reactions were carried out under pH and temperature controlled conditions (pH=7, Temperature= 25 C). Photocatalytic efficiency of the catalyst was studied to determine the extent of decol orization and organic matter removal caused upon exposure to TiO 2 /UV photoreactor set up. 1 st order rate constants were determined to compare the kinetics of the photodegradation process. Undoped TiO 2 / UV A and silver doped TiO 2 /UV C experiments had the hi ghest 1 st order rate constants; 0.087 hr 1 and 0.063 hr 1 with 94%, 99% decolorization and 82%, 91% COD removal after 41 and 59 hours of exposure respectively. Background Research on water reuse technologies has been gaining increasing attention in recen t times due to shortage of potable water and added costs for industries such as bioethanol plants that use millions of gallons of water every year. Seventy percent of untreated wastewaters from industries are being discharged into water bodies without bein g reused. Given that industries alone account for 22% of water usage in the world, such discarded industrial effluents amount to a significant quantity of water being wasted. Semiconductor photocatalysis, a type of advanced oxidation process, is attracting extensive research as a tertiary treatment to wastewater treatment process.
107 Following biological treatments such as anaerobic and aerobic digestion that remove a major percentage of the organic content from wastewaters, advanced oxidation process can be u sed as a polishing step to decolorize effluents such as textile, distillery, paper and pulp industrial wastewaters ( Vandevivere, Bianchi et al. 1998 ; Hai, Yamamoto et al. 2007 ; Oller, Malato et al. 2011 ; Ruas, Rodriguez Chaparro et al. 2012 ) In TiO 2 med iated photocatalysis, electron hole pairs are generated on irradiating the catalyst wit ( Xu, Chen et al. 2008 ) Such photo generated electron hole pairs participate in surface redox reactions that in turn generate highly reactive hydroxyl radicals. OH* radicals act as primary oxidants and degrade any residual organic pollutants or dyes in wastewaters, as well as decolorize them to a great extent. Though TiO 2 photocatalysis has a number of advantages s uch as low cost, non toxic and photostable, it suffers from low quantum efficiency of photocatalysis. The faster kinetics of electron hole pair recombination compared to slower surface redox reactions is a hurdle that has to be overcome for successful phot ocatalytic degradation of organics from wastew aters ( Kris hna, Noguchi et al. 2006 ) In order to improve the efficiency of TiO 2 photocatalysis, TiO 2 has been doped with vari ous metals, non metals and nano particles. Such added impurities enhance TiO 2 photocatalytic performance by shifting the absorption spectra t o lower energy regions ( Tan, Wong et al. 2011 ) Li and Li (2002) show ed that gold ion doped TiO 2 and gold doped TiO 2 have higher photo activity than plain TiO 2 catalyst when used to study the degradation of methylene blue. This was quantified using photoluminescence, XRD and XPS studies. Lower electron hole recombination wa s shown to be a function of the dosage,
108 distribution and electron structure impurity energy level of gold ions or gold particles used to dope TiO 2 ( Li and Li 2002 ) Behnajady et al., studied the effect of doping TiO2 nanoparticles with silver for photocatalytic degradation of AR88 dye (C.I Acid Red). About 2% and 0.5% silver deposition on TiO 2 surface using liquid impregnati on and photodeposition respectively, were determined to be the optimal values that helped increase the photocatalytic activity of TiO 2 The increase in photocatalytic activity can be attributed to addition of silver that acts as an electron scavenger and del ays electron hole recombination ( Behnajady, Modirshahla et al. 2008 ) Krishna et al., showed that doping TiO 2 catalyst in powder form (30 100 mg/L) with polyhydroxy fulleren es enhance d the photocatalytic efficiency of TiO 2 Fullerenes have very high electron affinity and when dosed at an optimal weight ratio of 0.001 (PHF:TiO 2 ) they scavenge the photo generated electrons from TiO 2 particles, thus improving the photocatalytic efficiency of the catalyst ( Krishna, Noguchi et al. 2006 ) Ultraviolet light source is required to activate TiO 2 to photocatalytically degrade organic pollutants. However, since sunlight is a natural form of UV source and is available in abundance at no ad ded costs, it may be a cost effective way to irradiate semiconductor particles. The purpose of this study was to investigate the degradation of residual organics and subsequent decolorization of biologically treated lignocellulosic ethanol stillage for p ossible reuse of recovered water in the plant. The obj ective of this study was to investigate enhancement in photocatalytic activity of silver doped TiO 2 versus un doped TiO 2 and
109 s tudy the effect of using various light sources: UV A ( 350 nm ) UV C ( 180nm ) and solar radiation on extent of TiO 2 mediated photocatalytic degradation of treated wastewaters. Materials and Methods Materials and methods used in this study have been described in Chapter 4. The additional materials used include d silver doped TiO 2 c oated sheets provided by IMDC Inc., and UV C (180nm) wavelength emitting lights. All the photocatalytic experiments reported in this study were performed in duplicates. The results reported are an average of the two sets of data. The st andard deviation was less than 1 0% in all cases. Results Absorbance Profiles Trends in absorbance profiles of photocatalytically degraded stillage samples versus time period of exposure were obtained from experiments using silver doped and plain TiO 2 coated reusable sheets a ctivated using three different UV sources: UV A, UV C emitting lamps and solar radiation. Absorption measurements of TiO 2 treated samples were taken at 350 nm to quantify the percent decolorization caused by photocatalysis. UV A 350 nm S ource Experiments 3 and 4 were carried out using plain and silver doped TiO 2 catalysts respectively, under UV A (350 nm) light source Figure 5 1 presents the absorbance profiles of photocatalytically treated samples in experiments 3 and 4. Under 350nm UV source, silver do ped catalysts seemed to have performed poorly compared to plain TiO 2 As shown in F igure 5 4 the rate coefficient for experiment 3 was 1.74
110 times that of experiment 4. In experiments 3 and 4, 94% and 89.5% decolorization was obtained after 41 and 50 hours of exposure. UV C 180 nm Light Source Figure 5 1 presents the trend in absorbance of (i) e xperiment 1: plain TiO 2 treated samples and (ii) e xperiment 2: silver doped TiO 2 treated samples. Using UV C lamps, silver doped TiO 2 catalysts seem to be more phot ocatalytically active than plain TiO 2 catalyst. Since the photocatalytic degradation of stillage follows 1 st or der reaction kinetics, the rate of reactions can be compared on the basis of rate coefficients. Rate coefficients were determined by fitting a 1 s t order model to the data and minimizing errors us ing least squares method. From F igure 5 4 it can be seen that the rate coefficient for experiment 2 was 1.8 times that of experimen t 1. Also, it can be seen from Figure 5 2 that 99% decolorization was obta ined after 59 hours of exposure in experiment 2, compared to 91.5% decolorization in experiment 1 after 55 hours. Solar Radiation Solar radiation is a naturally available source of ultraviolet radiation that can be used to energize semiconductor catalysts such as TiO 2 for photocatalytic degradation of organic compounds. In experiments 5 and 6, plain and silver doped TiO2 coated sheets were used to study photocatalytic degrad ation of stillage. As shown in F igure 5 1c silver doped TiO 2 catalysts performed better under sunlight leading to 90% decolorization of stillage within 43 hours compared to 80% decolorization using plain TiO 2 catalysts after the same time period of exposure. Rate coefficient from experiment 6 was 1.76 times that from experiment 5.
111 sCO D Removal Organic removal from treated wastewaters was studied in terms of soluble COD content. Figure 5 2 presents the percent sCOD degradation obtained in various photocatalytic experiments carried out in this study. Experiments 3 and 4 carried out usin g anatase and silver doped TiO 2 / UV A, resulted in 94% and 90% sCOD removal after 41 hours and 50 hours respectively. Experiments 1 and 2 carried out using TiO 2 and silver doped TiO2 using UV C lamps resulted in 92 and 99 percent sCOD reduction after 55 an d 59 hours respectively. On carrying out experiments 5 and 6 using TiO 2 and silver doped TiO 2 coated sheets under sun light, 80 % and 90 % organic matter removal was obtained after 43 hours of exposure respectively. sCOD Absorbance Correlation The relatio nship between the concentration of organic compounds and color producing compounds during the course of the experiment was studied by plotting sCOD Vs absorbance measured at any given time period of exposure. Based on the 1 st order rate constants, two expe riments with highest k values were chosen as optimum photocat alytic experimental conditions: experiments 3 and 2. The sCOD and absorbance values of the se experiments were plotted in F igure 5 3. The correlation between these sCOD and absorbance values can b e separated into 3 distinct zones. In zone 1, a significant drop in COD, 31% and 29% was obtained within 2 hours of treatment in experiments 2 and 3 respectively. However, this wa s not accompanied by a significant drop in color. In zone 2, after 20 hours of treatment, about 73% and 67% decolorization was obtained. This visible drop in color was however not accompanied by a significant drop in COD. Finally in zone 3, when the sCOD dropped below 600 mg/L, a linear correlation seemed to exist between sCOD and absorbance. The R 2 value from
112 linear regression plots of COD and absorbance in this zone was determined to be 0.9967. Such a high R 2 value is indicative of a highly linear relationship between COD and absorbance in zone 3. Discussion Use of Immobilized T iO 2 Coated S heets vs Powdered TiO 2 C atalysts In photocatalytic experiments TiO 2 is usually used in two forms: (i) p owder or (ii) i mmobilized form. Powder TiO 2 in suspended form is used in slurry photoreactors to degrade organic wastes such as dyes. TiO 2 concentrations in the range of 100 1500 mg/L have been used to treat wastes in slurry reactors ( Pekakis, Xekoukoulotakis et al. 2006 ) Though easy to operate, slurry reactor effluents usually have to be post treated in a liquid solid separator in order to remove the catalyst particles from treated effluents. Additional costs associated with installation and operation of such reactors m akes the process economically less favorable. Immobilized reactors that have been used in research studies earlier have been made by immobilizing TiO 2 on rigid surfaces such as ceramic membranes or glass matrix ( Haenel, Moren et al. 2010 ) Though immobilized TiO 2 reactors are favored in large scale set ups due to their reuse capabilities, they are limited by the thickness of catalyst coating on the surface that come s in contact with the bulk liquid ( Mukherjee 2011 ) In th is study reusable plastic sheets coated with TiO 2 powder have been used for photocatalytic degradation of l ignocellulosic ethanol stillage. Use of such cost effective, reusable, photostable TiO 2 coated sheets constitute towards novelty of this study. Was tewater Characteristics: Dilution vs Aerobic ally Treated AD Stillage Initially photocatalytic experiments were carried out using 1mm depth of anaerobically digested stillage that had an initial sCOD of 3000 4000 mg/L for a period
113 of 60 hours. In the first 20 hours, an exponential degradation pattern was observed but removing only 24.6% of the initial sCOD. After this period no further sCOD decrease was observed even after 60 hours. It is likely that photocatalytic treatment failed due to the high initial COD of anaerobically treated stillage COD was reduced in two ways 1) by dilution and 2) by aerobic treatment COD content of anaerobic aerobic stillage effluent was reduced to ~ 1000 1350 mg/L after 24 hours No further reduction in COD was obtained on pr olonged aerobic treatment. Therefore, it was decided to dilute the anaerobically digested stillage to achieve similar COD concentrations for comparison of the effect of dilution Vs. aerobic treatment for reduction in initial COD of treated wastewater S tu dies were carried out under UV A and UV C sources using undoped and silver doped TiO2 coated sheets. Figure 5 5. presents the absorbance profiles during photocatalytic treatment of diluted AD effluent. Results from photodegradation of diluted AD effluent s how that about 63%, 64%, 91%, 90.8% decolorization was obtained from TiO2/UV A, S TiO2/UV A, TiO2/UV C, S TiO2/UV C conditions respectively after 50 55 hours of exposure. Decolorization occurred to a greater extent under UV C light source using both undo ped and silver doped catalyst. Figure 5 6. presents the percent COD degrad ed during photocatalytic treatment of dilute d anaerobically digested stillage It can be seen that 68.6%, 59%, 68.8%, and 3.9% COD removal was obtained from TiO2/UV A, S TiO2/UV A, TiO2/UV C, and S TiO2/UV C experiments respectively. The percent COD removed was much lower than the extent to which the wastewater was decolorized. Although more than 90%
114 decolorization occurred under UV C source, COD removal was much lower only 3.9 % w hen silver doped TiO 2 was used and about 68.8% with undoped TiO 2 Kiriakidao et al., studied the effect of initial dye concentration on photocatalytic degradation of Acid Orange 7 (AO7), a non biodegradable azo dye seen in textile wastewaters. Experiments were carried out at pH 6 with 25 100 mg/L and at pH 12 with 25 600 mg/L dye concentrations. Results from this study showed that reaction rates were constant at concentrations below 100 mg/L at pH 6 and below 600 mg/L at pH 12. However, when dye concentra tions were increased beyond 100 mg/L or 1000 mg/L at pH 6 or 12 respectively, drop in reaction rate was observed ( Kiriakidou, Kondarides et al. 1999 ) Although the dye was completely degraded at reduced initial concentrations within minutes of exposure to TiO2/UV A set up, the COD curve seemed to have a significant lag and much slower rates o f reaction. Such an observation can be attributed to the uncolored intermediates produced during the degradation of AO7 that contribute towards residual COD. Due to slower rates of reaction longer periods of exposure are required to obtain higher levels o f COD removal. Boyd and Almquist (2004) showed that the initial COD of pulp and paper mill wastewater should be less than 500 mg/L for a successful photodegradation and to obtain a COD removal of greater than 90%. ( Baransi, Dubowski et al. 2012 ) ) carried out studies on photocatalytic degradation of anaerobically digested o live mill wastewater rich in phenolic compounds. T he effluent with a COD of 20,000 mg/L was diluted 10 times (1960 mg/L) and exposed to photocatalytic treatment. In spite a 10 fold dilution; 24 hr exposure to TiO2/UV A set up resulted in a COD removal of only 36.3% and color removal of 66%. A combination
115 of photocatalysis and adsorption process using (TiO2 + Powdered Activated Carbon) was use d as an alternative to treat olive mill wastewater. The final COD and color removal from TiO2+PAC experiments were 5 7.9% and 83% respectively. Similar results were obtained from a pilot scale study on treatment of diluted anaerobically digested distillery wastewater carried out by Rajvanshi and Nimbkar. Their system was designed to degrade wastewater using solar irradi ated TiO2 catalyst. An aerobically digested distillery wastewater was diluted 5 times before exposure to photoreactor set up. However, results showed that in spite of lowering the initial COD by dilution; TiO 2 photocatalysis alone was not effective in causi ng organic removal. TiO 2 photocatalysis was then used in combination with a coagulant that resulted in a COD removal of 87%. ( A. K Rajvanshi and Nimbkar 2004 ) Effect of Dif ferent UV Sources on TiO2 Photocatalysis Since TiO 2 molecules absorb light below visible region light of wavelength less than or equal to 384 nm is required to activate TiO 2 molecules. Typically photocatalysis studies carried out on dyes and wastewater samples reported in literature use UV A emitting light sources (315 and 400 nm) ( Pekakis, Xekoukoulotakis et al. 2006 ; Agustina, Ang et al. 2008 ; Badaw y, Gohary et al. 2009 ) Such UV light can be produced from either artificial UV lamps or obtained from solar radiation. Low, medium and high pressure mercury lamps may be used to emit UV A, UV B (280 315 nm) or UV C (100 280 nm) respectively ( Chong, Jin et al. 2010 ) Most transparent photoreactor materials except a few such as quartz, tend to cut off medium and short UV (UV B and UV C) radiations, allowing UV A to pass through. Results from F igure 5 2. list the % decolorization of TiO 2 treated stillage samples. Upon shining UV C radiations on the photoreactor set up, 91.5% decolorization was obtained after 55 ho urs of exposure
116 compared to 94% decolorization obtained on activating TiO 2 with UV A radiations. The 1 st order r ate coefficient from experiment 3 was found to be 2.5 times t hat of experiment ; showing that the photo degradation of stillage under UV A was 2. 5 times faster than under UV C. Short wavelength UV radiations are more often used for sterilization and microbial inactivation due to their strong germic idal properties ( Benabbou, Derriche et al. 2007 ) Although some studies i n literature have attributed the positive effect of increased rates of COD degradation using short wavelength UV radiations to direct photolysis and higher energy per unit area limited research has been carried out in this area using UV B or UV C radiatio ns ( Aliabadi and Sagharigar 2011 ) In a study on photodegradation of formaldehyde using immobilized TiO 2 catalyst, it was shown that the pseudo 1 st order rate coefficients from TiO2/UV A experiment was 1.32 times that from TiO2/UV C studies. This trend is in agreement with findings in our present study where faster kinetics were obtained with UV A radiations compared to UV C for decolorization of wastewater samples. Earlier studies have suggested that use of UV C that produces more energy per unit area leads to generation of more electron hole pairs. These electron hole pair s due to their large number in turn tend to recombine at a faster rate, leading to lesser organic removal. ( Neppolian, Jung et al. 2007 ) Effect of Solar R adiation on TiO 2 Photocatalysis Though sun light is a cheap and abundant source of naturally available UV radiations, only 5% of solar radiations possess the optimum energy for the band gap excitation of electrons ( Neppolian, Jung et al. 2007 ) Two types of systems have been used in the past that utilize solar radiations for photocatalytic t r eatment: concentrating and non concentrating systems. Concentrating systems such as parabolic trough collectors that were initially designed for solar thermal applications have also been used
117 in photocatalytic studies due to some noted advantages such as c ompact in size and more efficient usage of photons captured from solar radiations. However, due to various disadvantages such as water overheating, and higher costs, researchers have moved to non c oncentrating systems that are cost effective and can operat e using dir ect or diffuse solar radiations ( Blanco Galvez, Fernandez Ibanez et al. 2007 ) In this study, non concentrating systems were used to carry out solar activated photocatalytic degradation of stillage. Neppolian et al., (2007) carried out research on comparison of solar Vs. UV induced photocatalyti c degradation of 3 textile dyes; Reactive Yellow 17, Reactive Red 2 and Reactive Blue 4. It was shown that although sunlight induced mineralization of textile dyes was a cost effective alternative to UV (2 54nm) induced process, it had much slower kinetics and took longer time periods for degradation of dyes. Reactive Yellow 17 (5x10 4 M) was almost completely degraded within 6 hours under UV induced TiO2 (100 mg TiO 2 /100ml) mediated photocatalysis which too k more than 10 hours under solar induced photocatalysis. In this study, a similar trend was observed on comparing photocatalytic degradation of stillage effluent using UV A and solar radiation. About 80% decolorization of stillage effluent was achieved wit hin 30 hours of exposure using TiO 2 (3 mg/cm 2 )/ UV A set up compared to 43 hours under sunlight. COD removal was 70% and 47% respectively. This difference in rate of photocatalytic degradation of organic pollutants can be attributed to the difference in in put energy under UV and sunlight induced conditions. Since UV radiations contain more energy input, the rate of electron hole recombination is delayed to a longer period. Since only 5% of solar
118 radiation contains optimum energy for excitation of electrons in the valence band of TiO 2 this process takes longer time period for mineralization of organics. Effect of Silver D oping on TiO 2 P h otocatalysis Though ideally metal oxides with low band gap energy are preferred for photocatalytic reactions, they are rar ely used due to their photocatalytic instability. Semiconductor catalysts such as TiO 2 that have larger band gap energies are instead used because of their photostability. However, in order to improve the photocatalytic efficiency of such catalysts, the ra te of electron hole recombination must be slowed down. In the past many research studies have been carried out on doping metal oxides such as TiO 2 with group VIII and transition metals such as Au, Ag, Pt, Pd to improve electron hole separation and to exten d the light absorption from UV region to visible re gion, especially sunlight range ( Phoacharern 2006 ) In this study effect of silver doping on enhancement of TiO2 photocatalytic e fficiency has been studied under UV A, UV C and sunlight conditions The rate constant obtained by fitting a 1 st order model to photocatalytic degradation of stillage effluent were used to compare the efficiency of TiO 2 and silver doped TiO 2 mediated photocatalysis. Experiments 1 and 2 were carried out using UV C source, 3 and 4 using UV A and 5 and 6 using sunlight with TiO 2 and silver doped TiO 2 respectively. Under both UV C and sunlight induced conditions, silver doped TiO 2 catalyzed rates of mineralization were1.8 times that of un doped TiO 2 This observation has been made by many researchers in the past on various model compounds or simulated wastewaters. Behnajady et al., (2008) showed that silver doping enhanced the photocatalytic efficiency of TiO 2. They tested the mineralization of monoazo textile dyes using UV C light source with undoped and silver doped TiO 2 and showed that adding silver at 2%
119 (wt/wt) concentration to TiO 2 enhanced the photocatalytic efficiency of TiO2. However, in this present study, under UV A radiations, TiO 2 catalyst performed 1.7 times fast er than silver doped TiO 2 Such an observation may have occurred as a result of positively charged silver attracting the electrons and serving as electron hole recombination centers. Closing Remarks Photocatalytic degradation of residual organics and remo val of color from biologically treated cellulosic ethanol stillage was successfully carried out using TiO 2 coated reusable nonwoven media at a constant pH of 7. The effect of silver doping and use of UV A, UV C and sunlight on enhancing the rates of photoc atalytic degradation of stillage were studied. Rate coefficients from 1 st order modeling of stillage degradation were used to determine the optimal conditions for photocatalysis. Undoped TiO 2 / UV A and silver doped TiO 2 /UV C mediated photocatalysis experim ents had the highest rate coefficients: 0.087 hr 1 and 0.063 hr 1 and produced 94%, 99% decolorization and 82%, 91% COD removal after 41 and 59 hours of exposure respectively.
120 Figure 5 1 Decolorization profiles of photocatalytically treated Cellulosi c ethanol stillage
121 Figure 5 2 Percent decolorization and percent sCOD removal from various photocatalytic experiments
122 Figure 5 3 sCOD Absorbance correlation for experiments 2 and 3
123 Figure 5 4. 1st order rate constants from various photocataly tic experiments
124 Figure 5 5. Decolorization profile of photocatalytically treated diluted anaerobic stillage (DAS)
125 Figure 5 6 Percent C olor and sCOD removal from photocatalytically treated diluted anaerobic stillage (DAS)
126 CHAPTER 6 CONCLUSIONS AND FU TURE WORK Conclusions This research work was carried out to develop an integrated system to treat as well as recover resources from cellulosic ethanol stillage. Experiments were performed using model bench scale apparatuses on each compo nent of this integrated system to optimize operational parameters as well as to analyze the kinetics and yields form the processes. The conclusions of this present work are presented in three parts addressing recovery and reuse of (i) energy, (ii) nutrient s, and (iii) water. Continuous anaerobic digestion of cellulosic ethanol stillage was successfully carried out in a fluidized bed reactor. Such long term bench scale digestion of stillage was useful in determining the feasibility of the process, biochemic al methane potential of stillage and finally, various parameters required to design a large scale digester, including HRT, OLR, pH, VFA and COD levels. Following energy recovery in the form of biogas, stillage effluent rich in N and P was subjected to str uvite precipitation in a sequential batch reactor. Results from this study on nutrient recovery from stillage via struvite precipitation provided a way to determine the struvite potential of stillage. Finally, the remaining effluent ( 3.5 g sCOD/L ) was pa ssed through a TiO2/UV photoreactor set up for final polishing. Such a tertiary treatment allowed recovery of water from the process that can be recycled in the plant. Figure 6 1 represents the schematic of the integrated cellulosic ethanol stillage trea tment process. Stillage obtained from the Biofuels pilot plant at the University of Florida was a dark colored effluent which was first coarse separated resulting in a
127 reduction in total solids content from 10% to 4%. Stillage filtrate was continuously dig ested in an anaerobic fl uidized bed reactor (AFBR) for 100 days at thermophilic conditions. The methane yield was determined to be 12.9 L CH 4 @STP (L Stillage) 1 The AFBR was successfully operated at a HRT of 7.3 days which was down from the 22 days used in previous study (Tian, 2011), on stillage digestion in a CSTR. The organic loading rate (OLR) was inc reased to as high as 6.6 g sCOD L 1 d 1 which is about 3. 5 times higher than the CSTR study. The improved performance of this design was also witnessed as an increase in COD removal from 80% in the CSTR, to 93% along with a 23% increase in methane yield in the FB R. The pH was maintained between 7.1 8.5 during the operation of the digester. Organic acids and sugar levels were also monitored during this proce ss. Effluent ammonia and phosphorous concentrations were determined to be 320 and 527 mg/L showing that no external addition of nutrients was required. Following anaerobic digestion of stillage, the effluent was subjected to struvite precipitation for rec overy and reuse of nutrients. Struvite potential of raw stillage was 1.4 g/L and anaerobically digested stillage was 2.25 g/L. Results from application of raw and processed raw stillage containing precipitated struvite as soil amendments showed that, appl ication of the latter improved nutrient uptake efficiency of the plants and reduced the mass of P leached out by 20% compared to the former. Final part of this study dealt with studying the effect of seeding the process with stillage solids in order to imp rove the settleability of precipitated minerals. Results showed that addition of 1% ( wet w/v ) stillage solids as seed increased the yield of mineral precipitate by 60% and allowed more than 95% of settling to occur within 15 minutes.
128 Final part of this re search work was focused on investigating TiO2/UV mediated photocatalysis for polishing the biologically treated wastewater, to reuse water obtained from the process. In this work, the effect of a number of parameters including: depth of the influent, time period of exposure, initial pH, initial COD, form of the catalyst used (undoped, silver doped), UV light source (artificial, sun) and need for an aerobic pretreatment were studied. Results showed that undoped titanium dioxide under UV A conditions (TiO 2 /UV 350 ) and silver doped titanium dioxide under UV C conditions, (Silver TiO2/UV 180 ) with initial pH and COD adjusted to 7 and ~1000 mg/L respectively, via aerobic pretreatment were the conditions that had the fastest kinetics of degradation, resulting in mo re than 90% decolorization and COD removal. Water released from this process (COD <80 mg/L) may be readily reused in the plant. Mass and energy balances were calculated for a p lant producing 1 million gallon of ethanol per year (Figure 6 1). The mass bala nce (Figure 6 1) shows that about 75122 kg/d raw bagasse would be required to produce 8182 kg/d of ethanol. Such an ethanol production proces s will produce 103698.6 kg/d of stillage post distillation. Assuming all the stillage produced is digested in an AF BR at 7.3 days HRT and 6.54 OLR, it would produce 2834.5 m3 of methane per day. The remaining 107997 kg/d effluent will then be subjected to struvite precipitation for phosphate recovery. This process will produce 615 kg/d of struvite containing processed sludge. The remaining dark colored, nutrient deprived stream, on exposure to advanced oxidation via titanium dioxide mediated photocatalysis will be decolorized and residual organic content will be degraded. This process has the potential to yield 107381.5 kg/d of clean water that can be recycled in the plant.
129 The energy equivalents for respective amounts of steam used in this process were obtained from study by Tian (2011). Steam required to pretreat raw bagasse (50% TS) was determined to be 0.5 g steam/g bagasse. This amounted to a total of 37.5 kg steam/kg bagasse/day for pretreatment. The amount of steam required for distillation was calculated on the assumption that 15 lbs steam is required to distill out 1 gallon of ethanol from a 10% ethanol contain ing mixture. This amounted to be 20.3 kg steam/kg bagasse/day. The total input energy required to generate steam is less than 2% of the total output generated energy from combustion of biogas. Shapouri et al., (2002) released an update on energy balance of corn ethanol on behalf of USDA. Data showed that about 51779 Btu/gal of energy is used in the ethanol conversion process. The amount of energy produced from biogas produced by anaerobically digesting stillage A mass balance was also carried out on the ove rall ortho phosphate phosphorous released in the process. Results from the present work show that, about 70% of phosphate content of stillage comes from the acid pretreatment step with the rest 30% released from the feedstock itself. In this work, N and P h ave been recovered as struvite containing sludge. Other than the phosphate that is used for biomass growth in the digester, there is no loss of phosphate throughout the process. About 99% pho sphate is recovered as struvite containing sludge that can be use d as a fertilizer. Studies have shown that every 1 ton of sugarcane produces 0.3 tons of bagasse ( Cardona, Quintero et a l. 2010 ) In the present work, the amount of sugarcane required to meet the bagasse requirements in a 1 Mgal/yr etha nol plant was determined to be 250 tons. Assuming an average yield of 65 tons of sugarcane per hectare, a total land area of 1392 ha is
130 re quired. Studies have shown that for P limited soils such as Everglades soil in Florida, recommended phosphate dosage is ~36 kg P/ha ( McCray, Rice et al. 2010 ) Based on this information, it can be seen that phosphate precipitated as struvite contain ing sludge can help meet up to 43 % of the phosphate requirements to cultivate the required amount of sugarcane crop to p roduce adequate amount of bagasse feedstock for bioethanol production Future Work The overall objective of this work was to develop a sustainable integrated system that advocates simultaneous resource recovery and wastewater treatment. This objective has been successfully met by subjecting stillage to anaerobic digestion, struvite precipitation and TiO2/UV mediated photocatalytic polishing. This research work has also opened up more opportunities for further research: 1. Characterization and amenability of a naerobic digestion of hardwood stillage from eucalyptus fermentation may be investigated in an AFBR 2. C ontrol strategies for feed input to further increase organic loading rates in the AFBR and prevent imbalance or inhibition of the digestion process 3. Effe ct of using different seeding materials including nanoparticles that improve settleability and recovery of struvite containing processed sludge could be investigated. Initial trials carried out (not reported in this study) by the author showed struvite for mation and growth around magnetite nanoparticles and their ease in recovery by applying magnetic field. 4. TiO2 catalysts with appropriate dopants to improve photocatalytic rates 5. D isinfection of wastewaters using TiO2 mediated photocatalysis. 6. A well mixed sy stem design could be used to photocatalytically degrade treated stillage effluent. This may shorten the period of exposure. This design should be developed keeping in mind the ease in scaling up the reactor design.
131 X Overa ll Mass balance X Energy Balance X Phosphate Mass Balance Figure 6 1. Mass and Energy balance of cellulosic ethanol stillage t reatment
132 LIST OF REFERENCES Aden, A. and T. Foust (2009). "Technoeconomic analysis of the dilute sulfuric acid and enzymatic hydrolysis process for the conversion of corn stover to ethanol." Cellulose 16(4): 535 545. Agustina, T. E., H. M. Ang, et al. (2008). "Treatment of winery wastewater using a photocatalytic/photolytic reactor." Chemical Engineering Journal 135(1 2): 151 1 56. Agyin water treatment residual effects on the phosphorus status of field soils amended with biosolids, manure, and fertilizer. Commun. Soil Sci. Plt. Anal. 39:1700 1719. Aliabadi, M. and T. Sagharigar (2011). "Photocatalytic Removal of Rhodamine B from Aqueous Solutions using TiO2 Nanocatalyst." Journal of Applied Environmental and Biological Sciences 1(12): 620 626. Altinbas, M., C. Yangin, and I. Ozturk.2002. Struvite precipitati on from anaerobically treated municipal and landfill wastewaters, Water Sci. Technol. 46:271 278. Asif, M. and T. Muneer (2007). "Energy supply, its demand and security issues for developed and emerging economies." Renewable & Sustainable Energy Reviews 11 (7): 1388 1413. Badawy, M. I., F. E. Gohary, et al. (2009). "Enhancement of olive mill wastewater biodegradation by homogeneous and heterogeneous photocatalytic oxidation." Journal of Hazardous Materials 169(1 3): 673 679. Balat, M. (2011). "Production of bioethanol from lignocellulosic materials via the biochemical pathway: A review." Energy Conversion and Management 52(2): 858 875. Baransi, K., Y. Dubowski, et al. (2012). "Synergetic effect between photocatalytic degradation and adsorption processes on th e removal of phenolic compounds from olive mill wastewater." Water research 46(3): 789 798. Barta, Z., K. Reczey, et al. (2010). "Techno economic evaluation of stillage treatment with anaerobic digestion in a softwood to ethanol process." Biotechnology fo r Biofuels 3. Basakcilardan 44.
133 r 2929. Behnajady, M. A., N. Modirshahla, et al. (2008). "Enhancement of photocatalytic activity of TiO2 nanoparticles by silver d oping: Photodeposition versus liquid impregnation methods." Global Nest Journal 10(1): 1 7. Belay, N., R. Boopathy, et al. (1997). "Anaerobic transformation of furfural by Methanococcus deltae Delta LH." Applied and Environmental Microbiology 63 (5): 2092 2 094. Beltran, F. J., P. M. Alvarez, et al. (2001). "Treatment of high strength distillery wastewater (Cherry stillage) by integrated aerobic biological oxidation and ozonation." Biotechnology Progress 17(3): 462 467. Benabbou, A. K., Z. Derriche, et al. (2 007). "Photocatalytic inactivation of Escherischia coli Effect of concentration of TiO2 and microorganism, nature, and intensity of UV irradiation." Applied Catalysis B Environmental 76(3 4): 257 263. Bitton, G. (2005). Anaerobic Digestion Of Wastewater And Biosolids. Blanco Galvez, J., P. Fernandez Ibanez, et al. (2007). "Solar photocatalytic detoxification and disinfection of water: Recent overview." Journal of Solar Energy Engineering Transactions of the Asme 129(1): 4 15. Boopathy, R., H. Bokang, et a l. (1993). "Biotransformation of Furfural and 5 Hydroxymethyl Furfural by Enteric Bacteria." Journal of Industrial Microbiology 11 (3): 147 150. Bories, A., J. Raynal, et al. (1988). "Anaerobic Digestion Of High Strength Distillery Waste Water (Cane Molasse s Stillage) In A Fixed Film Reactor Biological Wastes 23(4): 251 267. Boyd, L. K. and C. B. Almquist (2004). "The application of photocatalysis on TiO2 for degrading COD in paper mill wastewaters." Tappi Journal 3(9): 9 15. Brune .G S.M. Schoberth, and H. Sahm, Anaerobic treatment of an industrial wastewater containing acetic acid furfural and sulfite. Process Biochem., 17, 20 25 (1982). Bridger, G. L., M. L. Salutsky, et al. ( 1962). "Micronutrient sources. m etal ammonium phosphates as fertilizers." Jo urnal of Agricultural and Food Chemistry 10: 181 188. Buchanan, J. R. (1993). Struvite Control in Flush Water Recycle Components of Livestock Waste, University of Tennessee, USA.
134 Buhr, H. O. and J. F. Andrews (1977). "Thermophilic Anaerobic Digestion Proce ss." Water research 11(2): 129 143. Callander, I. J., T. A. Clark, et al. (1986). "Anaerobic Digestion Of Stillage From A Pilot Scale Wood To Ethanol Process .1. Stillage Characterization." Environmental Technology Letters 7(6): 325 334. Callander, I. J., T. A. Clark, et al. (1987). "Anaerobic Digestion Of Wood Ethanol Stillage Using Upflow Anaerobic Sludge Blanket Reactor" Biotechnology and Bioengineering 30(7): 896 908. Cardona, C. A., J. A. Quintero, et al. (2010). "Production of bioethanol from sugarcan e bagasse: Status and perspectives." Bioresource Technology 101(13): 4754 4766. Celen, I., M. Turker. 2001. Recovery of ammonia as struvite from anaerobic digester effluents, Environ. Technol. 22:1263 1272. Chen, Y., J. J. Cheng, et al. (2008). "Inhibitio n of anaerobic digestion process: A review." Bioresource Technology 99(10): 4044 4064. Chong, M. N., B. Jin, et al. (2010). "Recent developments in photocatalytic water treatment technology: A review." Water research 44(10): 2997 3027. Couallier, E. M., T Payot, et al. (2006). "Recycling of distillery effluents in alcoholic fermentation Role in inhibition of 10 organic molecules." Applied Biochemistry and Biotechnology 133(3): 217 237. de Bashan, L. E. and Y. Bashan (2004). "Recent advances in removing phosphorus from wastewater and its future use as fertilizer (1997 2003)." Water research 38 (19): 4222 4246. de Vrije, T., R. R. Bakker, et al. (2009). "Efficient hydrogen production from the lignocellulosic energy crop Miscanthus by the extreme thermophil ic bacteria Caldicellulosiruptor saccharolyticus and Thermotoga neapolitana." Biotechnology for Biofuels 2 Doino, V, Mozet, K, Muhr, H andPlasari, E (2011). "Study on Struvite Precipitation in a Mechanically Stirring Fluidized Bed Reactor". Italian Assoc iation of Chemical Engineering. Doyle, J. D. and S. A. Parsons (2002). "Struvite formation, control and recovery." Water research 36(16): 3925 3940. Dubber, D. and N. F. Gray (2010). "Replacement of chemical oxygen demand (COD) with total organic carbon ( TOC) for monitoring wastewater treatment performance to minimize disposal of toxic analytical waste." Journal of
135 Environmental Science and Health Part a Toxic/Hazardous Substances & Environmental Engineering 45(12): 1595 1600. Dubey, R. S. (1974). "Distill ery effluents treatments and disposal." Sugar News Annual 6: 9 26. Ehrampoosh, M. H., G. R. Moussavi, et al. (2011). "Removal Of Methylene Blue Dye From Textile Simulated Sample Using Tubular Reactor And Tio2/Uv C Photocatalytic Process." Iranian Journal of Environmental Health Science & Engineering 8(1): 35 40. Fernandez, N., S. Montalvo, et al. (2008). "Performance evaluation of an anaerobic fluidized bed reactor with natural zeolite as support material when treating high strength distillery wastewater. Renewable Energy 33(11): 2458 2466. Fotiadis, C., N. P. Xekoukoulotakis, et al. (2007). "Photocatalytic treatment of wastewater from cottonseed processing: Effect of operating conditions, aerobic biodegradability and ecotoxicity." Catalysis Today 124(3 4 ): 247 253. Fuchs, W., H. Binder, et al. (2003). "Anaerobic treatment of wastewater with high organic content using a stirred tank reactor coupled with a membrane filtration unit." Water research 37(4): 902 908. Gadekar. S (2010 ). Process development for recovery of nutrients as struvite and struvite based products. Doctoral Dissertation, Agricultural and Biological Engineering. Gainesville, University of Florida. Geddes, C. C., I. U. Nieves, et al. (2011). "Advances in ethanol production." Current Opinion in Biotechnology 22(3): 312 319. Genc, N. and E. Can Dogan (2006). "Photooxidation: A decolorization procedure and a pre treatment step for biodegradation of reactive azo dye." Polish Journal of Environmental Studies 15(1): 73 79. Genz, A., A. Kornmuller et al. (2004). "Advanced phosphorus removal from membrane filtrates by adsorption on activated aluminium oxide and granulated ferric hydroxide." Water research 38(16): 3523 3530. Gerischer, H. and A. Heller (1991). "The Role of Oxygen in Photooxidation of Organic Molecules on Semiconductor Particles." Journal of Physical Chemistry 95(13): 5261 5267. Haenel, A., P. Moren, et al. (2010). Photocatalytic Activity of Tio2 Immobilized on Glass Beads ." Physicochemical Problems of Mineral Processing(45): 49 56.
136 Hai, F. I., K. Yamamoto, et al. (2007). "Hybrid treatment systems for dye wastewater." Critical Reviews in Environmental Science and Technology 37(4): 315 377. Hodge, H. M. and F. M. Hildebrandt (1954). "Alcoholic Fermentation of Molasses in Industrial Fe rmentations Chemical Publisliing Co., New York.: 73 89. Huang, X., M. Leal, et al. (2008). "Degradation of natural organic matter by TiO2 photocatalytic oxidation and its effect on fouling of low pressure membranes." Water research 42(4 5): 1142 1150. Ka bdasli, I., P. Ozcan, and O. Tunay. 2003. Nitrogen removal by magnesium ammonium phosphate precipitation in slaugtheryhouse wastewater, Su Kirlenmesi Kontrolu Dergisi 13:13 18. Khatamian, M., N. Daneshvar, et al. (2010). "Heterogeneos Photocatalytic Decolo rization of Brown NG by TiO2 UV Process." Iranian Journal of Chemistry & Chemical Engineering International English Edition 29(3): 19 26. Koch, J. (2009). Membrane Bioreactor Technology to Remove Phosphorous. and systems. Kiriakidou, F., D. I. Ko ndarides, et al. (1999). "The effect of operational parameters and TiO2 doping on the photocatalytic degradation of azo dyes." Catalysis Today 54(1): 119 130. Krishna, V., N. Noguchi, et al. (2006). "Enhancement of titanium dioxide photocatalysis by water soluble fullerenes." Journal of Colloid and Interface Science 304(1): 166 171. Kalyuzhnyi, S., V. Sklyar, A.E.I. Arkhipchenko, I. Barboulina, O. Orlova, and A. Klapwijk. 2002. Combined biological and physico chemical treatment of filtered pig manure waste water: pilot investigations, Water Sci. Technol. 45:79 87. Kim, D., H D. Ryu, M S. Kim, J. Kim, and S Ill. Lee. 2007. Enhancing struvite precipitation potential for ammonia nitrogen removal in municipal landfill leachate, J. Hazard. Mater. 146:81 85. Laksh mi, S., R. Renganathan, et al. (1995). "Study on Tio2 Mediated Photocatalytic Degradation of Methylene Blue." Journal of Photochemistry and Photobiology a Chemistry 88(2 3): 163 167. Li, X.Z., and Q.L. Zhao. 2003. Recovery of ammonium nitrogen from landfil l leachate as a multi nutrient fertilizer, Ecol. Eng. 20:171 181. Li, X.Z., Q.L. Zhao, and X.D. Hao. 1999. Ammonium removal fromlandfill leachate by chemical precipitation,Waste Manag. 19:409 415.
137 Li, F. B. and X. Z. Li (2002). "Photocatalytic properties o f gold/gold ion modified titanium dioxide for wastewater treatment." Applied Catalysis a General 228(1 2): 15 27. Limayem, A. and S. C. Ricke (2012). "Lignocellulosic biomass for bioethanol production: Current perspectives, potential issues and future pros pects." Progress in Energy and Combustion Science 38(4): 449 467. Liu, C. C., Y. H. Hsieh, et al. (2006). "Photodegradation treatment of azo dye wastewater by UV/TiO2 process." Dyes and Pigments 68(2 3): 191 195. Lyberatos, G. and P. C. Pullammanappallil (2010). Anaerobic Digestion in Suspended Growth Bioreactors, Humana Press Inc, 999 Riverview Dr, Ste 208, Totowa, Nj 07512 1165 USA. McCray, J. M., R. W. Rice, et al. (2010). "Sugarcane Response to Phosphorus Fertilizer on Everglades Histosols." Agronomy J ournal 102(5): 1468 1477. Mills, T. Y., N. R. Sandoval, et al. (2009). "Cellulosic hydrolysate toxicity and tolerance mechanisms in Escherichia coli." Biotechnology for Biofuels 2. Moletta, R. (2005). "Winery and distillery wastewater treatment by anaero bic digestion." Water Science and Technology 51(1): 137 144. Molina, F., G. Ruiz Filippi, et al. (2007). "Winery effluent treatment at an anaerobic hybrid USBF pilot plant under normal and abnormal operation." Water Science and Technology 56(2): 25 31. Mu kherjee, D. (2011). Development of a Novel TiO2 polymeric Film Photocatalyst for Water Purification both under UV and Solar Illuminations. Chemical and Biochemical Engineering. London, Ontario, Canada, The University of Western Ontario. Ph.D. Munch, E. V. and K. Barr (2001). "Controlled struvite crystallisation for removing phosphorus from anaerobic digester sidestreams." Water research 35(1): 151 159. Neethling, J.B. and A. Gu. 2006. Chemical phosphorus removal constraints Introduction. Session P2 in WERF, 2006. Neppolian, B., H. Jung, et al. (2007). "Photocatalytic degradation of 4 chlorophenol using TiO2 and Pt TiO2 nanoparticles prepared by sol ge l method." Journal of Advanced O xidation Technologies 10(2): 369 374. Nieves, I. U., C. C. Geddes, et al. (2011). "Effect of reduced sulfur compounds on the fermentation of phosphoric acid pretreated sugarcane bagasse by ethanologenic Escherichia coli." Bioresource Technology 102(8): 5145 5152.
138 Noethe, T., H. Fahlenkamp, et al. (2009). "Ozonation of Waste water: Rate of Ozone Consumption and Hydroxyl Radical Yield." Environmental Science & Technology 43(15): 5990 5995. Ogi, H., M. Hirao, et al. (2002). "Activation of TiO2 photocatalyst by single bubble sonoluminescence for water treatment." Ultrasonics 40(1 8): 649 650. O'Connor GA, Elliott HA (2006) The agronomic and environmental availability of biosolids P (Phase II). Rep. 99 PUM 2T Water Environment Research Foundation, Alexandria, VA. Oller, I., S. Malato, et al. (2011). "Combination of Advanced Oxidati on Processes and biological treatments for wastewater decontamination A review." Science of the Total Environment 409(20): 4141 4166. Omolola, A. M. (2007). Anaerobic Digestion Of Ethanol Distillery Waste Stillage For Biogas Production. Environmental Eng ineering Hgskolan I Bors. M.sc. Owen W. O. (1979). Autohydrolysis for improving methane yield from fermentation of lignocellulose. Ph.D. Thesis, Stanford University, Stanford, C.A Paramasivam, S., G. Z. Fortenberry, et al. (2008). "Evaluation of em ission of greenhouse gases from soils amended with sewage sludge." Journal of Environmental Science and Health Part a Toxic/Hazardous Substances & Environmental Engineering 43(2): 178 185. Pekakis, P. A., N. P. Xekoukoulotakis, et al. (2006). "Treatment of textile dyehouse wastewater by TiO2 photocatalysis." Water research 40(6): 1276 1286. Applied for Wastewater and Drinking Water Treatment. Elimination of Pharmaceuticals." T he Holistic Approach to Environment 1(2): 63 74. Phoacharern, P. (2006). Photocatalytic Degradation of Trypan Blue using Gold/Titanium Dioxide Chemistry, Kasetsart University. Master of Science. Polematidis, I., A. Koppar, et al. (2010). "Biogasification p otential of desugarized molasses from sugarbeet processing plants." Journal of Sugar Beet Research 47(3/4): 89 104. Radjenovic, J., C. Sirtori, et al. (2009). "Solar photocatalytic degradation of persistent pharmaceuticals at pilot scale: Kinetics and char acterization of major intermediate products." Applied Catalysis B Environmental 89(1 2): 255 264. Rajvanshi. A.K and Nimbkar. N (2004). Solar Detoxification of Distillery Waste. Maharashtra, India, Nimbkar Agricultural Research Institute
139 Rammohan, G., Ga dekar S and Pullammanappallil. P (2011). "Development of a ProcessModel for Recovery of Nutrients fromWastewater by Precipitation as Struvite." Florida Water Resources Journal(Jan 2011). Reardon, R. 2006. Technical introduction of membrane separation pr ocesses for low TP limits. Session P3 in WERF, 2006. Predicted US oil glut a boon to those who move Reddy, S. S. and B. Kotaiah (2005). "Decolorization of simulated spent reactive dye bath using solar/Ti O 2/H 2O 2." International Journal of Environmental Science and Technology 2(3): 245 251. Ren, L., F. Schuchardt, Y. Shen, G. Li, and C. Li. 2010. Impact of struvite crystallization on nitrogen losses during composting of pig manure and cornstalk. Waste M anagement 30 (2010) 885 892 Ronteltap, M., M. Maurer, and W. Gujer. 2007. Struvite precipitation thermodynamics in source separated urine, Water Res. 41:984 997. Ruas, D. B., T. Rodriguez Chaparro, et al. (2012). "Advanced oxidation process H2O2/UV combin ed with anaerobic digestion to remove chlorinated organics from bleached kraft pulp mill wastewater." Revista Facultad De Ingenieria Universidad De Antioquia(63): 43 54. Ruiz, C., M. Torrijos, et al. (2002). "Treatment of winery wastewater by an anaerobic sequencing batch reactor." Water Science and Technology 45(10): 219 224. Sampaio, M. A., M. R. Goncalves, et al. (2011). "Anaerobic digestion challenge of raw olive mill wastewater." Bioresource Technology 102(23): 10810 10818. Sarkar, A. K. (1990). Phos phate Cement Based Fast Setting Binders ." American Ceramic Society Bulletin 69(2): 234 238. Satyawali, Y. and M. Balakrishnan (2008). "Wastewater treatment in molasses based alcohol distilleries for COD and color removal: A review." Journal of Environmenta l Management 86(3): 481 497. Schuiling, R. D. and A. Andrade (1999). "Recovery of struvite from calf manure." Environmental Technology 20(7): 765 768. Schipper, W., Klapwijk, B., Potjer, B., Rulkens, W.,Temmink, H., Kiestra, F., Lijmbach,D. (2001): Phospha te recycling in the phosphorus industry. 2 nd International
140 conference on the recovery of phosphorus from sewage and animal waste, Nordwijkerhout. Shapouri, H., Duffield, J.A et al. (2002). The Energy Balance of Corn Ethanol: An Update, U.S. Department of Agriculture. Sheehan, G. J. and P. F. Greenfield (1979). "Utilisation, treatment and disposal of distillery wastewater." Water Research 14(3): 257 to 277. Shin, C., J. Bae, et al. (2012). "Lower operational limits to volatile fatty acid degradation with d ilute wastewaters in an anaerobic fluidized bed reactor." Bioresource Technology 109: 13 20. Shrestha, J. N. (2001). Efficiency Measurement of Biogas, Kerosene and LPG Stoves. Lalitpur, Nepal, Tribhuvan University. Sten, I., ( 2004 ) Phosphorus recovery in the context of industrial use. In: Phosphorus in Environmental Technologies, principles and applications, Valsami Jones, E. (Ed), pp. Strom, P.F., H .X. Littleton, & G.T. Daigger. ( 2004 ). Characterizing Mechanisms of Simultaneous Biological Nutrient Remova l during Wastewater Treatment. Water Environment Research Foundation, Alexandria, VA. Tan, Y. N., C. L. Wong, et al. (2011). "An Overview on the Photocatalytic Activity of Nano Doped TiO2 in the Degradation of Organic Pollutants." ISRN Materials Scienc e: 18. Tao, L., A. Aden, et al. (2011). "Process and technoeconomic analysis of leading pretreatment technologies for lignocellulosic ethanol production using switchgrass." Bioresource Technology 102(24): 11105 11114. Tenenbaum, D. J. (2008). "Food vs. fu el: Diversion of crops could cause more hunger." Environmental Health Perspectives 116(6): A254 A257. Tian, Z. (2011). Anaerobic Digestion Of Biofuel Production Residues. Agricultural and Biological Engineering. Gainesville, University of Florida. Ph.D. US DA (2006). The Economic Feasibility of Ethanol Production From Sugar in the United States UWWTD (1991). Official Journal of the Europea Community Series L, 135/40. Vandevivere, P. C., R. B ianchi, et al. (1998). "Treatment and reuse of wastewater from the textile wet processing industry: Review of emerging technologies." Journal of Chemical Technology and Biotechnology 72(4): 289 302.
141 Wang, J., Y. Song, et al. (2006). "Modeling the crystalli zation of magnesium ammonium phosphate for phosphorus recovery." Chemosphere 65(7): 1182 1187. Watts, R. J., S. Kong, et al. (1994). "Titanium Dioxide Mediated Photocatalysis of a Biorefractory Chloroether in Secondary Waste Water Effluent." Environmental Technology 15(5): 469 475. Wilkie, A. C., K. J. Riedesel, et al. (2000). "Stillage characterization and anaerobic treatment of ethanol stillage from conventional and cellulosic feedstocks." Biomass & Bioenergy 19(2): 63 102. Willington, I. P. and G. G. Mar ten (1982). "Options For Handling Stillage Waste From Sugar Based Fuel Ethanol Production." Resources and Conservation 8(2): 111 129. Whittier, A. C. (1991). Investigations on the Estimation of Inorganic Phosphorous in Animal Tissues. The Journal of Indust rial and Engineering Chemistry American Chemical Society. Wolmarans, B. and G. H. de Villiers (2002). "Start up of a UASB effluent treatment plant on distillery wastewater." Water Sc 28(1): 63 68. Wu, M., M. Mintz, et al. ( 2009). Consumptive Water Use in the Production of Ethanol and Petroleum Gasoline, Argonne National Laboratory. Xu, J., Z. D. Chen, et al. (2008). "Photocatalytic killing effect of gold doped TiO2 nanocomposites on human colon carcinoma LoVo cells." Acta Chimica Sinica 66(10): 1163 1167. Yao, Y., B. Gao, et al. (2011). "Biochar derived from anaerobically digested sugar beet tailings: Characterization and phosphate removal potential." Bioresource Technology 102(10): 6273 6278. Zhang, L., L. Wan, et al. (2011). "Removal of phosphate from w ater by activated carbon fiber loaded with lanthanum oxide." Journal of Hazardous Materials 190(1 3): 848 855.
142 BIOGRAPHICAL SKETCH G ayathri Ram Mohan received her b iotechnology from SRM University, India in May 2008. Following that she decided to pursue hi gher education in the field of biological e ngineering and joined the University of Florida. She received merit scholarship s for academic excellence in her undergraduation and research assistantship f or her doctoral studies. She has received certificate s of academic excellence from the University of Florida for three consecutive years. She is also a member of many prestigious h onor societies including Gamma Sigma Delta and Phi Kappa Phi. Upon receiving her doctorate she plans on jo ining the Wastewater treatment industry to explore career opportunities.