UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Pilot scale cofermentation of the xylose and hexose fractions of spent sulfite pulping liquor using recombinant… Guild, Jeffrey Joseph 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-0223.pdf [ 38.48MB ]
Metadata
JSON: 831-1.0058762.json
JSON-LD: 831-1.0058762-ld.json
RDF/XML (Pretty): 831-1.0058762-rdf.xml
RDF/JSON: 831-1.0058762-rdf.json
Turtle: 831-1.0058762-turtle.txt
N-Triples: 831-1.0058762-rdf-ntriples.txt
Original Record: 831-1.0058762-source.json
Full Text
831-1.0058762-fulltext.txt
Citation
831-1.0058762.ris

Full Text

PILOT SCALE COFERMENTATION OF THE XYLOSE AND HEXOSE FRACTIONS OF SPENT SULFITE PULPING LIQUOR USING RECOMBINANT STRAINS OF SACCHAROMYCES CEREVISIAE by J E F F R E Y JOSEPH G U I L D B A S c , The University of British Columbia, 1998 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S FOR T H E D E G R E E OF M A S T E R OF A P P L I E D S C I E N C E in T H E F A C U L T Y OF G R A D U A T E STUDIES (Department of Chemical and Biological Engineering) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF BRITISH C O L U M B I A April 2004 © Jeffrey Joseph Guild, 2004 Library Authorization In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Name of Author (please print) L ^ d d f r ^ Title of Thesis: f ^ t ±a\UL ce f -^IL/^TTVrxsrO c f -rv\C- < ^ U ^ L ft^ft AjKQ^ £ f r A c - r i p / ^ g f sp£~>T~ S W N L F M T & f ^ u f U G - n ^ ^ L Degree: /Y\ASTfc & A fP<~\ £0 ^c\}JAc<i Y E A R : >^)0 i f Department of C\fcr<\\ (SAL fi^O &vaU?£\c*M. £ J 6 itt£&(L \ r$G The University of British Columbia Vancouver, BC Canada 11 ABSTRACT At the Tembec sulfite pulp mill, an existing ethanol fermentation facility currently ferments the hexose sugars from spent sulfite liquor (SSL) to ethanol using an industrial strain of Saccharomyces cerevisiae. While this industrial yeast strain achieves high fermentation efficiency on the hexose fraction, it is incapable of fermenting the pentose fraction (primarily comprised of xylose) present in the SSL. The plant Saccharomyces strain exhibits high ethanol tolerance, fast fermentation rates, fermentation at low pH, and resistance to inhibitory substances. Natural xylose-fermenting yeasts such as P. stipitis exhibit high ethanol yield on xylose, however, fermentation rates are slow, and tolerance towards inhibitors in SSL is low. A recombinant strain possessing both the advantageous characteristics of S. cerevisiae and the xylose-fermenting capability of P. stipitis could significantly increase the efficiency of ethanol production from lignocellulosic hydrolysate. Xylose fermentation would have a large impact on the economics of ethanol production in the facility. Tembec Inc. could increase the production of high-quality industrial ethanol by approximately 30% if the xylose present in softwood and hardwood SSL were cofermented with hexose sugars. Xylose-fermenting recombinant Saccharomyces yeasts (plasmid-bearing and chromosomal transformants of robust industrial yeast strains) were constructed by Prof. N . Ho, Purdue University, and tested on xylose-rich SSL at pilot scale. The pilot fermentation plant is an automated 1/1000-scale emulation of the full-scale fermentation operation at Tembec and operates continuously in complete yeast recycle mode under typical industrial (aseptic) conditions. While up to 100% of the xylose present was consumed in extended pilot trials with the plasmid-bearing strain (LNH32), no appreciable increase in ethanol yield over the current plant U l strain was demonstrated. The fermentation efficiency of LNH32 was comparable to results with the plant yeast strain, despite the fact that LNH32 utilized only 50% of the galactose. Incomplete fermentation of galactose by the LNH32 would seem to explain the lack of increased ethanol production, and due to xylose fermentation by the strain, a 15% increase in fermentation efficiency (based on available hexose sugars) would be realized over the plant yeast strain if LNH32 were capable of fermenting galactose. Finally, in laboratory trials with LNH-ST, a stable xylose-fermenting chromosomal transformant of a galactose-positive host strain, a 20% increase in ethanol yield over the parental strain and plant yeast strain was obtained in both hardwood and softwood spent sulfite liquors. iv T A B L E OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES viii LIST OF FIGURES x LIST OF ABBREVIATIONS AND ACRONYMS xv LIST OF UNITS xvii ACKNOWLEDGEMENTS xviii 1. INTRODUCTION 1 1.1 The Situation for Tembec Inc 1 1.2 Thesis Organization 3 2. LITERATURE REVIEW 6 2.1 Ethanol Production from Biomass 6 2.1.1 The Sugars in Biomass 6 2.1.2 Theoretical Ethanol Yields from Hexose and Pentose Sugars 7 2.1.3 Selection of Organisms for Fermentation 9 2.1.4 Fermentation Configurations for Ethanol Production 12 2.1.5 Operating Conditions for Fermentation with Saccharomyces cerevisiae 17 2.1.6 The Economics of Ethanol Production 18 2.2 Conversion of Spent Sulfite Liquors to Ethanol 18 2.2.1 Description of the Sulfite Pulping Process 19 2.2.2 Composition of Hardwood and Softwood Spent Sulfite Liquors 23 2.2.3 Description of the Tembec Alcohol Plant 24 2.3 Xylose Fermentation by Naturally Occurring Strains 28 2.3.1 Xylose Fermentation by Pichia stipitis 28 2.3.2 Xylose Fermentation in Natural Strains of Saccharomyces cerevisiae 32 2.4 Tembec Pilot Plant - Original Research Studies 33 2.4.1 Baseline Fermentation Testing with the Tembec Plant Strain (T2) 33 2.4.2 Baseline Fermentation Testing with Pichia stipitis 35 2.4.3 Baseline Pilot Plant Fermentation Testing - Conclusions 37 2.5 Design of Pentose-Fermenting Saccharomyces cerevisiae Strains 38 2.5.1 General Microbial Genetics and Techniques in Recombinant Engineering 38 2.5.2 Genetic Engineering of Xylose Metabolism in Saccharomyces cerevisiae 42 2.5.3 Recombinant Pentose-Fermenting Strains used at Tembec 46 2.6 Laboratory Shake Flask Experiments with LNH32 on SSL 53 V T A B L E OF CONTENTS 2.6.1 Ethanol Production by 1400, LNH32, and the Tembec Strain 53 3. RESEARCH OBJECTIVES 55 4. MATERIALS AND EXPERIMENTAL METHODS 57 4.1 Inoculum Storage 57 4.1.1 Inoculum Storage in Liquid Nitrogen 57 4.1.2 Inoculum Storage on Agar Slopes 58 4.1.3 Stock Mother Cultures of Inoculum for Trials 58 4.2 Propagation of LNH32 in Shake Flasks 59 4.3 Propagation of LNH32 in the 16 L Lab Fermenter 59 4.4 Propagation of LNH32 in the Pilot Plant 61 4.5 Shake Flask Testing with 259A (LNH-ST) 62 4.6 Operation of the Pentose Pilot Plant 63 4.6.1 Process Description & Control Strategy 64 4.6.2 Plant Control System and Interface 68 4.7 Sampling, Analysis and Data Acquisition 70 4.7.1 Dissolved Solids Concentration Measurement 71 4.7.2 Ethanol Analysis by Gas Chromatograph 71 4.7.3 Sugar Analysis by HPLC 72 4.7.4 Analysis of Yeast Volume and Viability 73 4.7.5 Data Acquisition and Trending 75 4.8 Spent Sulfite Compositions for the Pilot Trial Runs 75 4.8.1 Pilot Plant Trial #1, LNH32 on SWD SSL 75 4.8.2 Pilot Plant Trial #2, LNH32 on H WD SSL 76 4.8.3 Pilot Plant Trial #3, LNH32 on Blended SSL 77 4.8.4 Pilot Plant Trial #4, LNH32 on Blended SSL 77 4.8.5 Pilot Plant Trial #5, LNH-ST on Hardwood SSL 78 4.9 Sterilization Techniques in the Laboratory and Pilot Plant 79 4.10 Plasmid Verification Test Procedures 80 4.10.1 The N . Ho Test for Confirmation of Ethanol Production from Xylose 80 4.10.2 Differential Plate Counts to Test Growth on Xylose as the Sole Sugar 84 4.10.3 Durham Tube Test to Test Gas Production with Xylose as the Sole Sugar ..88 5. RESULTS AND DISCUSSION , 90 5.1 Bench Scale (16 L Lab Fermenter) Testing with LNH32 90 5.1.1 Bench Scale Trial #1 - Propagation of LNH32 on Softwood SSL 90 5.1.2 Bench Scale Trial #2 - Batch Fermentation of Softwood SSL by LNH32 96 5.1.3 Bench Scale Trial #3 - Propagation of LNH32 on Hardwood SSL 98 5.1.4 Bench Scale Trial #4 - Batch Fermentation of Hardwood SSL by LNH32 .... 103 vi T A B L E OF CONTENTS 5.1.5 Experimental Plan for Pilot Plant Testing with LNH32 105 5.2 Pilot Scale Testing with LNH32 105 5.2.1 Pilot Plant Trial #1 - Softwood Spent Sulfite Liquor 106 5.2.2 Pilot Plant Trial #2 - Hardwood Spent Sulfite Liquor 122 5.2.3 Pilot Plant Trial #3 - An Extension of Pilot Plant Trial #2 on Blended SSL ..135 5.2.4 Pilot Plant Trial #4 - An Extension of Pilot Plant Trial #2 on Blended SSL ..143 5.2.5 Pilot Plant Trials #2 to #4 - Summary of Long Term Fermentation Trends ... 150 5.3 Shake Flask Testing with 259A(LNH-ST) 160 5.3.1 Softwood and Hardwood Trials with 259A(LNH-ST) 160 5.3.2 Experimental Plan for Pilot Plant Trial #5 with LNH-ST 163 5.4 Pilot Scale Testing with LNH-ST 163 5.4.1 Pilot Plant Trial #5 - LNH-ST on Hardwood & Softwood SSL 163 6. CONCLUSIONS 168 6.1 Bench & Pilot Scale Trials with LNH32 168 6.1.1 Plasmid Retention and Strain Growth 168 6.1.2 Xylose Metabolism and Fermentation 169 6.1.3 Fermentation Efficiency and Ethanol Productivity 170 6.2 Shake Flask Trial with LNH-ST 171 6.3 Pilot Scale Trial with LNH-ST 172 6.3.1 Strain Confirmation 172 6.3.2 Xylose Metabolism and Fermentation 172 6.3.3 Fermentation Efficiency and Ethanol Productivity 173 6.4 Implications for the Tembec Full-scale Operation 173 7. RECOMMENDATIONS 174 7.1 Extended Pilot Plant Trials with 1400(pLNH32) 174 7.2 Further Testing with 259A(LNH-ST) 175 7.3 Genetic Modification of the Tembec Strain (T2) 175 7.4 Yeast Purges with the Tembec Strain (T2) and Recombinant Strains 176 7.5 Monitoring of Xylitol Production with Recombinant Strains 177 7.6 Alteration of Gene Expression and Enzyme Ratios in Recombinant Strains 178 8. BIBLIOGRAPHY 179 APPENDIX A. RECIPES USED FOR LAB CULTURE CULTIVATION 184 A. 1 Recipes for Yeast Growth on Softwood SSL 184 A.2 Recipes for Yeast Growth on Hardwood SSL 185 A.3 Recipes for YPD, YPX, and Lysine Agar Plates 187 vii T A B L E OF CONTENTS APPENDIX B. ADDITIONAL GRAPHS FROM PILOT PLANT TRIALS 188 B.l Pilot Plant Trial #1 - Softwood Spent Sulfite Liquor 188 B.2 Pilot Plant Trial #2 - Hardwood Spent Sulfite Liquor 194 B.3 Pilot Plant Trial #3 - An Extension of Pilot Plant Trial #2, Blended SSL 200 B.4 Pilot Plant Trial #4 - An Extension of Pilot Plant Trial #2, Blended SSL 206 B. 5 Pilot Plant Trial #5 - LNH-ST on Blended SSL 212 APPENDIX C. RAW DATA FROM BENCH & PILOT PLANT TRIALS 218 C l Bench Trial #1 - Softwood Spent Sulfite Liquor, Continuous 218 C. 2 Bench Trial #2 - Softwood Spent Sulfite Liquor, Batch 219 C.3 Bench Trial #3 - Hardwood Spent Sulfite Liquor, Continuous 220 C.4 Bench Trial #4 - Hardwood Spent Sulfite Liquor, Batch 221 C.5 Pilot Plant Trial #1 - Softwood Spent Sulfite Liquor 222 C.6 Pilot Plant Trial #2 - Hardwood Spent Sulfite Liquor 223 C.7 Pilot Plant Trial #3 - An Extension of Pilot Plant Trial #2, Blended SSL 224 C.8 Pilot Plant Trial #4 - An Extension of Pilot Plant Trial #2, Blended SSL 225 C. 9 Pilot Plant Trial #5 - LNH-ST on Blended SSL 226 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA 227 D. l Bench Scale Trial #1, LNH32 on SWD SSL 227 D.2 Bench Scale Trial #3, LNH32 on HWD SSL 228 D.3 Pilot Plant Trial #1, LNH32 on SWD SSL 230 D.4 Pilot Plant Trial #2, LNH32 on HWD SSL 232 D.5 Pilot Plant Trial #3, LNH32 on Blended SSL 234 D.6 Pilot Plant Trial #4, LNH32 on Blended SSL 235 viii LIST OF TABLES Table 2-1 Typical composition of softwood and hardwood liquor (Cameron, 2000) 24 Table 2-2 Full-scale ethanol production on softwood and hardwood SSL (St. Onge, 1996) 27 Table 2-3 Yeast strains used in laboratory and pilot plant trials 46 Table 4-1 Average incoming SSL composition for Pilot Plant Trial #1 76 Table 4-2 Average incoming SSL composition for Pilot Plant Trial #2 76 Table 4-3 Average incoming SSL composition for Pilot Plant Trial #3 77 Table 4-4 Average incoming SSL composition for Pilot Plant Trial #4 78 Table 4-5 Average incoming SSL composition for Pilot Plant Trial #5 79 Table 4-6 Ethanol production by cofermentation of dextrose and xylose 83 Table 4-7 Yeast colony count - July 30 t h lab fermenter sample 85 Table 4-8 Summary of the July 30 th yeast colony count 86 Table 4-9 Yeast colony count - August 17th pilot plant sample 86 Table 4-10 Summary of the August 17th yeast colony count 87 Table 5-1 Differential plate counts at the beginning and end of the trial 113 Table 5-2 Sugars utilization on day 13 of the softwood trial 119 Table 5-3 Fermentation efficiency calculation 119 Table 5-4 Differential plate counts at the beginning and end of the trial 126 Table 5-5 Average sugar uptake in Pilot Plant Trial #1 & #2 136 Table 5-6 Average sugar uptake in Pilot Plant Trial #1, #2, and #3 143 Table 5-7 Average sugar uptake in Pilot Plant Trials #1, #2, #3, and #4 150 Table 5-8 Average sugar uptake in Pilot Plant Trial #5 165 Table D-1 Yeast Colony Count - May 28 t h Lab Fermenter Sample 227 Table D-2 Summary of the May 28 t h Yeast Colony Count 227 Table D-3 Yeast Colony Count - June 11 th Lab Fermenter Sample 228 Table D-4 Summary of the June 11 th Yeast Colony Count 228 Table D-5 Yeast Colony Count - June 15th Lab Fermenter Sample 229 Table D-6 Summary of the June 15th Yeast Colony Count 229 Table D-7 Yeast Colony Count - July 30 t h Lab Fermenter Sample 230 Table D-8 Summary of the July 30 t h Yeast Colony Count 230 Table D-9 Yeast Colony Count - August 17th Pilot Plant Sample 231 Table D-10 Summary of the August 17th Yeast Colony Count 231 Table D-l 1 Yeast Colony Count - September 22 n d Lab Fermenter Sample 232 ix LIST OF TABLES Table D-12 Summary of the September 22 n d Yeast Colony Count 232 Table D-13 Yeast Colony Count - September 25 t h Pilot Plant Sample 232 Table D-14 Summary of the September 25 t h Yeast Colony Count 233 Table D-15 Summary of the October 10th Yeast Colony Count 233 Table D-16 Yeast Colony Count - October 20 th Pilot Plant Sample 234 Table D-17 Summary of the October 20 t h Yeast Colony Count 234 Table D-18 Yeast Colony Count - October 30 th Pilot Plant Sample 235 Table D-19 Summary of the October 30 t h Yeast Colony Count 235 X LIST OF FIGURES Figure 2-1 Theoretical ethanol yield from glucose (reproduced from Ingledew, 1995) 8 Figure 2-2 Theoretical ethanol yield from xylose (reproduced from Spencer et al, 1983) 8 Figure 2-3 Growth of a typical microbial culture in batch conditions (reproduced from Stanbury etal, 1995) 14 Figure 2-4 Simplified sulfite pulp mill flowsheet 20 Figure 2-5 Simplified Tembec flowsheet 21 Figure 2-6 Basic solids balance in the Tembec pulp mill (Cameron, 2000) 22 Figure 2-7 Simplified Tembec alcohol plant flowsheet 25 Figure 2-8 Tembec alcohol plant strain of Saccharomyces cerevisiae (lOOx magnification) 27 Figure 2-9 Initial steps of xylose pathway in yeasts and filamentous fungi. XYL1 codes for xylose (aldose) reductase; XYL2 codes for xylitol dehydrogenase; XYL3 codes for xylulokinase (reproduced from Jeffries et al, 1998) 30 Figure 2-10 Ethanol production by the Tembec alcohol plant Saccharomyces strain on softwood SSL in the pilot plant (Cameron, 2000) 34 Figure 2-11 Ethanol production and xylose uptake by Pichia stipitis on softwood SSL in the pilot plant (Cameron, 2000) 36 Figure 2-12 Xylose metabolic pathways in organisms (Ho et al, 1998) 43 Figure 2-13 Construction of high-copy-number yeast-£. coli shuttle plasmid LNH32, containing the XYL three-gene cassette K K - A R - K D . The Xhol DNA fragment containing K K - A R - K D was inserted into pUCKmlO at its Sail site (reproduced from Ho et al, 1998) 48 Figure 2-14 Comparison of the abilities of recombinant Saccharomyces strain 1400(pLNH32) (left) and parent strain 1400 (right) to coferment glucose and xylose (reproduced from Ho et al, 1998) 49 Figure 2-15 Chromosome manipulation - LNH-ST (Ho, 1999) 51 Figure 2-16 Ethanol production by the stable strain - LNH-ST (Ho, 1999) 52 Figure 2-17 Comparison of shake flask ethanol production by Saccharomyces 1400, LNH32, and the Tembec Strain (reproduced from Cameron, 1995) 54 Figure 4-1 Mother culture of LNH32 in 50 mLs of YPX medium 58 Figure 4-2 Autoclave, centrifuge, and shaker bath 59 Figure 4-3 16 L New Brunswick lab fermenter 60 Figure 4-4 Hardwood and softwood compositions for shake flask fermentation by T2 -the Tembec plant strain, 259A - the transformant parental strain, and 259A(LNH-ST) - the stable transformant strain (Chenier, 1999) 63 Figure 4-5 Pilot plant process and instrumentation drawing 64 xi LIST OF FIGURES Figure 4-6 Five 1000 L pilot plant fermentation tanks in series 67 Figure 4-7 Pilot plant solid bowl centrifuge 68 Figure 4-8 Pilot plant overview display screen 69 Figure 4-9 Spent sulfite liquor feed display screen 69 Figure 4-10 Fermentation tanks display screen 70 Figure 4-11 Yeast recycle display screen 70 Figure 4-12 Gas chromatograph for ethanol analysis 72 Figure 4-13 Dionex DX-300 with a CarboPac PA1 column for sugar analysis 73 Figure 4-14 Yeast budding and oxalate crystals (lOOx magnification) 75 Figure 4-15 Streak plating for colony isolation & plasmid verification 81 Figure 4-16 Ethanol production by cofermentation of dextrose and xylose 83 Figure 4-17 Spread plating on YPX to identify LNH32 cells 85 Figure 4-18 Durham tube test for xylose utilization 89 Figure 5-1 Bench Scale Trial #1 - LNH32 growth curve when utilizing softwood spent sulfite liquor as the primary substrate in the 16 L lab fermenter 91 Figure 5-2 Plasmid-bearing strain LNH32 (lOOx magnification) 93 Figure 5-3 Bench Scale Trial #1 - Ethanol production by LNH32 grown on softwood spent sulfite liquor in the 16 L lab fermenter 94 Figure 5-4 Bench Scale Trial #1 - Sugar consumption and ethanol production by LNH32 grown on softwood spent sulfite liquor in the 16 L lab fermenter 95 Figure 5-5 Bench Scale Trial #1 - Ethanol production by LNH32 grown on softwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived 96 Figure 5-6 Bench Scale Trial #2 - Sugar consumption and ethanol production by LNH32 in a batch fermentation of concentrated softwood SSL in the 16 L lab fermenter 97 Figure 5-7 Bench Scale Trial #3 - LNH32 growth curve when utilizing hardwood spent sulfite liquor as the primary substrate in the 16 L lab fermenter 99 Figure 5-8 Bench Scale Trial #3 - Ethanol production by LNH32 grown on hardwood spent sulfite liquor in the 16 L lab fermenter 100 Figure 5-9 Bench Scale Trial #3 - Sugar consumption and ethanol production by LNH32 grown on softwood spent sulfite liquor in the 16 L lab fermenter 101 Figure 5-10 Bench Scale Trial #3 - Ethanol production by LNH32 grown on hardwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived 102 xu LIST OF FIGURES Figure 5-11 Bench Scale Trial #4 - Sugar consumption and ethanol production by LNH32 in a batch fermentation of concentrated hardwood SSL in the 16 L lab fermenter 104 Figure 5-12 Pilot Plant Trial #1 - LNH32 growth curve on softwood SSL in the 16 L lab fermenter prior to pilot plant inoculation 106 Figure 5-13 Pilot Plant Trial #1 - Ethanol production by LNH32 on softwood SSL in the 16 L lab fermenter prior to pilot plant inoculation 107 Figure 5-14 Pilot Plant Trial #1 - Sugar consumption and ethanol production by LNH32 on softwood SSL in the 16 L lab fermenter prior to pilot plant inoculation 108 Figure 5-15 Pilot Plant Trial #1 - Ethanol production by LNH32 grown on softwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived 109 Figure 5-16 Pilot Plant Trial #1 - Viable cell concentration of LNH32 in fermenter #1 and #2 on softwood SSL .110 Figure 5-17 Pilot Plant Trial #1 - A decrease in the viability of LNH32 in fermenter #1 and #2 over time on softwood SSL 111 Figure 5-18 Pilot Plant Trial #1 - Utilization of xylose by LNH32 on softwood SSL 112 Figure 5-19 Pilot Plant Trial #1 - Utilization of galactose by LNH32 on softwood SSL 114 Figure 5-20 Pilot Plant Trial #1 - Comparison of galactose use by LNH32 and the Alcohol Plant Strain on softwood SSL (Alcohol Plant data reproduced from Cameron, 1995) 114 Figure 5-21 Pilot Plant Trial #1 - Comparison between ethanol produced by LNH32 on softwood SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized 115 Figure 5-22 Pilot Plant Trial #1 - Sugar utilization and ethanol production by LNH32 on softwood SSL (t - day 9) 116 Figure 5-23 Pilot Plant Trial #1 - Fermentation efficiency by LNH32 on softwood SSL based on total hexose and sugar available, as well as hexose and sugar utilized 118 Figure 5-24 Pilot Plant Trial #1 - Fermentation efficiency comparison between LNH32 and the Alcohol Plant Strain from a previous baseline trial on softwood SSL 121 Figure 5-25 Pilot Plant Trial #2 - LNH32 growth curve on hardwood SSL in the 16 L lab fermenter prior to pilot plant inoculation 123 Figure 5-26 Pilot Plant Trial #2 - Ethanol production by LNH32 on hardwood SSL in the 16 L lab fermenter prior to pilot plant inoculation 124 Figure 5-27 Pilot Plant Trial #2 - Sugar consumption and ethanol production by LNH32 on hardwood SSL in the 16 L lab fermenter prior to pilot plant inoculation 124 Figure 5-28 Pilot Plant Trial #2 - Ethanol production by LNH32 grown on hardwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived 125 X l l l LIST OF FIGURES Figure 5-29 Pilot Plant Trial #2 - Viable cell concentration of LNH32 in fermenter #1 and #2 on hardwood SSL 127 Figure 5-30 Pilot Plant Trial #2 - A decrease in the viability of LNH32 in fermenter #1 and #2 overtime on hardwood SSL 128 Figure 5-31 Pilot Plant Trial #2 - Utilization of xylose by LNH32 on hardwood SSL 129 Figure 5-32 Pilot Plant Trial #2 - Utilization of galactose by LNH32 on hardwood SSL 130 Figure 5-33 Pilot Plant Trial #2 - Utilization of hexose by LNH32 on hardwood SSL 130 Figure 5-34 Pilot Plant Trial #2 - Comparison between ethanol produced by LNH32 on hardwood SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized 131 Figure 5-35 Pilot Plant Trial #2 - Sugar utilization and ethanol production by LNH32 on hardwood SSL (t = day 11) 132 Figure 5-36 Pilot Plant Trial #2 - Fermentation efficiency by LNH32 on hardwood SSL based on total hexose and sugar available, as well as hexose and sugar utilized 133 Figure 5-37 Pilot Plant Trial #2 - Fermentation efficiency comparison between LNH32 and the Alcohol Plant Strain from a previous baseline trial on hardwood SSL.... 134 Figure 5-38 Pilot Plant Trial #3 - An increase in the viable cell concentration of LNH32 in fermenter #1 and #2 on blended SSL 137 Figure 5-39 Pilot Plant Trial #3 - Utilization of xylose by LNH32 on blended SSL 138 Figure 5-40 Pilot Plant Trial #3 - Utilization of galactose by LNH32 on blended SSL 139 Figure 5-41 Pilot Plant Trial #3 - Comparison between ethanol produced by LNH32 on blended SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized 140 Figure 5-42 Pilot Plant Trial #3 - Sugar utilization and ethanol production by LNH32 on blended SSL (t = day 24) 141 Figure 5-43 Pilot Plant Trial #3 - Fermentation efficiency by LNH32 on blended SSL based on total hexose and sugar available, as well as hexose and sugar utilized 142 Figure 5-44 Pilot Plant Trial #4 - Viable cell concentration of LNH32 in fermenter #1 and #2 on blended SSL. Also shown are the ethanol profile, and the point at which a yeast purge was attempted 144 Figure 5-45 Pilot Plant Trial #4 - Viability and budding of LNH32 in fermenter #1 and #2 on blended SSL. Also shown are the ethanol profile, and the point at which a yeast purge was attempted 146 Figure 5-46 Pilot Plant Trial #4 - Utilization of xylose by LNH32 on blended SSL 147 xiv LIST OF FIGURES Figure 5-47 Pilot Plant Trial #4 - Comparison between ethanol produced by LNH32 on blended SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized 148 Figure 5-48 Pilot Plant Trial #4 - Sugar utilization and ethanol production by LNH32 on blended SSL (t = day32) 148 Figure 5-49 Pilot Plant Trial #4 - Fermentation efficiency by LNH32 on blended SSL based on total hexose and sugar available, as well as hexose and sugar utilized 149 Figure 5-50 Pilot Plant Trials #2-4 - Viable cell concentration of LNH32 in fermenter #1 and #2 on varied grades of SSL. Also shown is where a yeast purge was attempted 151 Figure 5-51 Pilot Plant Trials #2-4 - Viability and budding of LNH32 in fermenter #1 and #2 on varied grades of SSL. Also shown is where a yeast purge was attempted 152 Figure 5-52 Pilot Plant Trials #2-4 - Long-term utilization of xylose by LNH32 on varied grades of SSL 154 Figure 5-53 Pilot Plant Trials #2-4 - Long-term utilization of galactose by LNH32 on varied grades of SSL 155 Figure 5-54 Pilot Plant Trials #2-4 - Comparison between ethanol produced by LNH32 on varied grades of SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized 156 Figure 5-55 Pilot Plant Trials #2-4 - Sugar utilization and ethanol production by LNH32 on varied grades of SSL 156 Figure 5-56 Pilot Plant Trials #2-4 - Fermentation efficiency by LNH32 on varied grades of SSL based on total hexose and sugar available, as well as hexose and sugar utilized 158 Figure 5-57 Comparison of shake flask ethanol production by T2, 259A, and LNH-ST (Chenier, 1999) 161 Figure 5-58 Comparison of sugar utilization by T2, 259A, and LNH-ST (reproduced from Chenier, 1999) 162 Figure 5-59 Pilot Plant Trial #5 - Sugar utilization and ethanol production by LNH-ST on softwood SSL (t = day 15) 164 Figure 5-60 Pilot Plant Trial #5 - Fermentation efficiency comparison between LNH-ST and the Alcohol Plant Strain from a previous baseline trial on hardwood SSL.... 166 LIST OF ABBREVIATIONS AND ACRONYMS xv 259A The Saccharomyces parental strain used for chromosomal transformation 259ST Chromosomal transformed Saccharomyces cerevisiae strain, 259A(LNH-ST) AUR1-C Dominant selectable marker; confers resistance to Aureobasidin A BC British Columbia BOD Biological oxygen demand BUT n-butylamine CFU Colony forming unit COD Chemical oxygen demand COFI Council of Forest Industries of British Columbia EMP Embden-Meyerhof Pathway ETOH Ethanol FCC Flux control coefficient GC Gas chromatography HMW High molecular weight HPLC High performance liquid chromatography HWD Hardwood 1400 The Saccharomyces parental strain used for plasmid transformation LNH32 Plasmid transformed Saccharomyces cerevisiae strain, 1400(pLNH32) LNH-ST Chromosomal transformed Saccharomyces cerevisiae strain, 259A(LNH-ST) NSERC Natural Sciences and Engineering Research Council of Canada PPP Pentose Phosphate Pathway SWD Softwood SSL Spent Sulfite Liquor T2 The Saccharomyces strain used in Tembec's full-scale alcohol plant X5P Xylulose 5-phosphate X D H Xylitol dehydrogenase XI Xylose isomerase X K Xylulokinase XKSI Gene encoding the enzyme X K XR Xylose reductase LIST OF ABBREVIATIONS AND ACRONYMS XYL1 Gene encoding the Pichia stipitis enzyme XR XYL2 Gene encoding the Pichia stipitis enzyme XDH XYL3 Gene encoding the Pichia stipitis enzyme X K XylA Gene encoding for the Thermus thermophilus enzyme XI YPD Yeast Extract, Peptone, and Dextrose YPX Yeast Extract, Peptone, and Xylose ZEO R Dominant selectable marker; confers resistance to zeocin Note: Al l genes are from Saccharomyces cerevisiae unless stated otherwise. xvii LIST OF UNITS A Angstrom adt, (t) Air dried tonne, (tonne) cm 3, (m3) Cubic centimeter, (cubic meter) Da, (kDa) Dalton, (kilodalton) d Day eq, (meq) Equivalent, (milliequivalent) g, (pg), (mg), (kg) Grams, (microgram), (milligram), (kilogram) h, hr Hour K Kelvin kV Kilovolts L, (u,L), (mL) Litre, (microlitre), (millilitre) lb Pound m Square meter mM, (M) Millimolar, (molar) mA Milliampere min Minute mm, (nm), (pm) Millimeter, (nanometer), (micrometer) N Normal °C Degrees Celcius ppm Parts per m i 11 ion rpm Revolution per minute s Second v/v Volume by volume w/v Weight by volume XVIll ACKNOWLEDGEMENTS I would like to gratefully acknowledge the contributions of the many people involved in the completion of this research. First and foremost, I owe thanks to my two supervisors. 1 am sorry to say that during the course of my studies, we suffered the loss of my industrial adviser, Dr. David R. Cameron (Technical Director, Tembec Inc., Chemical Division). 1 am very grateful to Dave for his guidance and encouragement during my time at Tembec. Not only did I learn a great deal as an academic about the potential of recombinant fermentation technology, 1 also learned about the realities of implementing new technology at full-scale in an industrial setting. I would also like to offer sincere thanks to my faculty adviser, Dr. Sheldon J.B. Duff, for giving me the opportunity to carry out this research, and for his patience while waiting for a finished product. I owe a debt of gratitude to Annie Chenier (Tembec Inc., Chemical Division) who provided invaluable technical assistance through all stages of the pilot studies, and for countless HPLC and GC analyses. I would like to thank Bob Benson (Tembec Inc., Chemical Division) for his ongoing support of the piloting studies at Tembec. A special thanks to Daniel Boucher (Tembec Inc., Chemical Division QA/QC Laboratory) and the rest of the laboratory staff for their assistance. Last, but not least, I'd like to thank all of the great people that I met in the Alcohol Division for always making me feel at home. This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC) through an industrial post-graduate scholarship. Additional funding was provided for this research by the Natural Resources Canada Bioenergy Development Program (NRC), who funded the original construction of the pilot plant and the Council of Forest Industries of British Columbia (COFI) through the NSERC/COFI Industrial Research Chair in Forest Products Waste Management. CHAPTER 1. INTRODUCTION 1 1. INTRODUCTION 1.1 The Situation for Tembec Inc. At the Tembec sulfite pulp mill, an existing ethanol fermentation facility currently ferments the hexose sugars from spent sulfite liquor to ethanol using an industrial strain of Saccharomyces cerevisiae. While this industrial yeast strain achieves a high level of fermentation efficiency on the hexose fraction, it is incapable of fermenting the pentose fraction (primarily comprised of xylose, with a small amount of arabinose) present in the spent sulfite liquor. The proportion of xylose and hexose sugars varies widely in the Temiscaming mill; as many as 10 different grades of softwood, hardwood, or blended pulps can be produced. Liquors from these different grades of pulp not only have different concentrations of sugars, but will also have variable pH, osmotic strength, and concentrations of toxic compounds (such as SO2, acetic acid, and furfural). The proportion of sugars typically found in softwood spent sulfite liquor is approximately 75% hexose and 25% xylose, at concentrations of roughly 30 g/L and 10 g/L respectively. In hardwood SSL the balance differs, in that the ratio of xylose to hexose is approximately 50/50 (Cameron, 1998). It would be desirable to identify a yeast strain capable of utilizing the large concentration of currently unfermented sugars present in both softwood and hardwood liquors. Fermentation of the xylose fraction would have a large impact on the economics of ethanol production in the facility. Tembec could increase the production of high quality industrial ethanol by approximately 30% (5,000,000 litres per year) if the xylose present in softwood and hardwood spent liquors were cofermented with hexose sugars (Cameron, 1998). In CHAPTER L INTRODUCTION 2 addition, if the xylose fraction were cofermented in the existing alcohol plant by a xylose-fermenting recombinant Saccharomyces yeast, the additional ethanol production would represent pure profit to Tembec, as little or no capital investment would be required. Tembec Inc. has a pentose pilot plant, which is a 1/1000-scale emulation of the full-scale fermentation operation. The design of the plant is very flexible, making it relatively easy to operate under various conditions, retention times, and with different organisms. Most of the trials that have been completed in the pilot plant to date have involved the alcohol plant yeast strain to establish baseline results on both softwood and hardwood liquors. These trials also refined the process by which the pilot plant was inoculated, the yeast was propagated, and the timing of changes to process variables such as feed rates and aeration requirements. In addition to determining the optimum fermentation conditions for the yeast presently being used in the alcohol plant, this pilot plant was designed to test new yeast strains and their ability to coferment xylose and hexose sugars to ethanol. While there are naturally occurring yeast strains, such as Pichia stipitis, that are capable of fermenting xylose to ethanol, in laboratory and pilot plant trials these strains have had low fermentation efficiencies compared to the strain currently used in the alcohol plant (Cameron, 1998). Dr. Nancy Ho of Purdue University provided a recombinant Saccharomyces strain, 1400(pLNH32), for research purposes. A host Saccharomyces strain (1400, a fusion strain) was transformed using genes from Pichia stipitis. These genes, the genes necessary for xylose metabolism to ethanol, were cloned on a high copy-number plasmid that was inserted into the parent yeast fusion 1400. In laboratory shake flask experiments with both well-defined and typical alcohol plant media, this recombinant yeast strain consistently produced more ethanol than the yeast strain currently utilized in the alcohol plant (Cameron, 1998). CHAPTER 1. INTRODUCTION 3 A previous trial in the pilot plant (July '97) using the recombinant Saccharomyces strain 1400(pLNH32) was unsuccessful as the plant became contaminated with yeast from the alcohol plant after only 7 days. By the end of this trial, the plasmid enabling the strain to ferment xylose was no longer present. These plasmids, although introduced in large numbers to the host cells, can be lost gradually during cell replication on a non-selective medium (i.e. if too many generations are produced while not utilizing xylose). These results indicated that maintaining the plasmid could be the major hindrance to overcome in future trials (Cameron, 1998). Trials conducted in the summer of 1998 established the optimum growth conditions for LNH32 on both softwood and hardwood spent sulfite liquors in the 16 L laboratory fermenter. The yeast was successfully generated in sufficient quantity to inoculate the pilot plant, and the xylose-fermenting capabilities of the recombinant strain were maintained throughout the trials. Expanding on the results found in these trials, pilot scale trials were conducted using xylose-fermenting recombinant Saccharomyces strains on actual industrial substrates to determine ethanol production capabilities. The ideal yeast for Tembec will exhibit stable performance under typical alcohol plant conditions, while cofermenting the xylose and hexose sugars in the spent liquor from the sulfite pulp mill. 1.2 Thesis Organization Chapter one introduces the specific considerations for Tembec Inc. in exploring xylose fermentation. Details are provided on the full-scale alcohol plant and the motivation for construction of the pentose pilot plant. Finally, pilot plant trials performed to date involving both the alcohol plant strain and recombinant strains of Saccharomyces, are summarized. CHAPTER L INTRODUCTION 4 Chapter two first presents an overview of considerations in the fermentation of hemicellulose hydrolysates, summarizing some of the processing challenges faced by ethanol producers that utilize these feedstocks. Major considerations behind the use of ligno-cellulose as a renewable feedstock for ethanol production are introduced, as are the economic implications of improving upon current levels of ethanol production. Specific emphasis is given in this chapter to the particular challenges faced in the conversion of spent sulfite liquors to ethanol. The full-scale alcohol plant at Tembec is discussed, with details provided regarding baseline ethanol production efficiency on both softwood and hardwood spent sulfite liquor, and the characteristics of the Saccharomyces strain used in the full-scale operation. Some of the findings from original studies performed in the pentose pilot plant are presented, including a comparison of the fermentation efficiency of the alcohol plant strain (T2), and Pichia stipitis, a strain known to have xylose-fermenting capability. A detailed literature review of research work conducted in the areas of xylose fermentation, and the design of pentose fermenting Saccharomyces strains is presented. Finally, the construction of LNH32 and LNH-ST, the plasmid bearing and stable transformant strains, respectively, is discussed, and laboratory shake flask experiments involving fermentation of well-defined media with LNH32 are presented. Chapter three outlines the research objectives and major motivation behind this research. Specific goals are set for the plasmid-bearing and chromosomal recombinant Saccharomyces strains, with the major objective being the successful cofermentation of the xylose and hexose sugars present in the spent sulfite liquor. Chapter four first describes the methodologies used for inoculum storage and preservation at Tembec, and details the techniques used to propagate yeast in shake flasks, the 16 L bench fermenter, and in the pilot plant. The pilot plant is described with respect to its CHAPTER 1. INTRODUCTION 5 operation and control strategy, and the daily sampling and testing procedures are detailed. Ana ly t ica l techniques for ethanol, sugar, and yeast analysis are provided. The compositions o f the spent sulfite liquors used in the pilot plant trials are provided. F ina l ly , the techniques used for sterilization in the laboratory and pilot plant, verification o f plasmid presence and contaminant identification are described. Chapter five summarizes shake flask, 16 L bench scale, and pilot scale fermentation results obtained at the Tembec m i l l during the course o f this research. A total o f 5 pilot plant trial runs are presented that were conducted from M a y to December o f 1998. The majority o f the work presented focuses on fermentation trials with L N H 3 2 , as the stable Saccharomyces transformant L N H - S T was obtained late in the study period. A shake flask experiment, and a single pilot plant run ut i l iz ing L N H - S T are presented. Chapter six reviews the major conclusions o f this work with respect to the fermentation ability o f both L N H 3 2 and L N H - S T , and implications for the Tembec full-scale operation. F ina l ly , Chapter seven provides a list o f recommendations for future research in this area, specifically focussed on short and long-term studies that may be o f value for Tembec Inc. Appendix A presents the recipes used for lab culture cultivation on media containing softwood or hardwood S S L , as wel l as the recipes used for Y P D , Y P X , and Lysine agar plates. Appendix B provides additional graphs o f the results from the pilot plant trials, providing the same data that was summarized in Chapter 5, but graphed in a different light (for example, graphing the concentration changes o f a given sugar through the different unit operations o f the pilot plant, rather than simply summarizing it as "influent and effluent"). Appendix C presents the raw data from bench and pilot plant trials. Lastly, Append ix D summarizes al l o f the differential plate count data from both bench and pilot plant trials. CHAPTER 2. LITERATURE REVIEW 6 2. LITERATURE REVIEW 2.1 Ethanol Production from Biomass A relatively small fraction of commodity chemicals are produced by fermentation, with practically all being manufactured from petroleum and natural gas. However, fluctuating costs of petroleum and unease about the rate of depletion of non-renewable resources have increased interest in the use of non-petroleum feedstocks for ethanol production (Ward, 1989; Crueger and Crueger, 1990). There are several advantages to the use of fermentation processes for production of industrial alcohol (Ward, 1989): 1. Fermentation technology currently exists 2. Alcohols can be produced from renewable resources 3. Wastes and inedible agricultural products may be used as substrates 4. Recovery processes are relatively simple 5. Alcohol fuels burn more cleanly than gasoline fuels The inherent disadvantage with biomass feedstocks is that their energy yield per unit weight is approximately one-third that of petroleum (Ward, 1989). Consequently, the efficiency and economics of the conversion of biomass to ethanol varies considerably depending upon the availability and composition of the raw materials, the organism used for the fermentation, and the process configuration. 2.1.1 The Sugars in Biomass Plant and woody biomass, or lignocellulose, is the most abundant form of biomass available for fermentation to ethanol (Johansson, 2001). Lignocellulose contains cellulose, CHAPTER 2. LITERATURE REVIEW 7 hemicellulose, and lignin, where the cellulose is a linear glucose polymer linked via P-1,4 linkages, and hemicellulose is a mixture of polymers comprised of pentose sugars (xylose and arabinose) and hexose sugars (mannose, glucose, and galactose) (Johansson, 2001). The sugar polymers in lignocellulose are relatively easily recovered by acid or enzymatic hydrolysis, with the resulting hydrolysate containing mainly mannose, glucose, galactose, xylose and arabinose in varied proportions (Johansson, 2001). The available sugars in lignocellulose hydrolysate are dependent on the starting raw material, with softwood species (i.e. pine) containing relatively high ratios of hexose to xylose, and hardwood species (i.e. birch) containing proportionately higher concentrations of xylose (Johansson, 2001). Thus, hexose sugars (mannose, glucose, and galactose - the 6-carbon sugar molecules) and xylose are the major fermentable substrates present in lignocellulose hydrolysate. 2.1.2 Theoretical Ethanol Yields from Hexose and Pentose Sugars The first step in the utilization of sugar by yeast is usually either its intake across the cell membrane, or an initial hydrolysis outside the membrane followed by entry into the cell by some or all of the hydrolysis products (Spencer et al, 1983). S. cerevisiae is constitutive for the ability to metabolize glucose and mannose, two of the major sugars available in both softwood and hardwood SSL. Uptake of galactose has been described as an inducible system (Spencer et al, 1983). The transport of these hexose sugars by the constitutive and inducible systems in S. cerevisiae is passive, and thus does not require energy (Spencer et al, 1983). Because the uptake of hexose sugars occurs via facilitated diffusion (Barnett, 1997), virtually no energy is expended in the metabolism of these sugars, resulting in high theoretical conversion efficiency to ethanol. For each gram of hexose sugar metabolized by Saccharomyces, CHAPTER 2. LITERATURE REVIEW 8 0.5 lg of ethanol are theoretically produced according to the following stoichiometry (conversion of glucose is shown) (Spencer et al, 1983). C6H,206 • 2 C 2H 5OH + 2 C0 2 Glucose Ethanol Figure 2-1 Theoretical ethanol yield from glucose (reproduced from Ingledew, 1995). While the theoretical stoichiometric yield is 0.5lg EtOH / g glucose, in practice approximately 10% of the hexose is converted to biomass, thus the yield of ethanol may only reach 90% of theoretical values (Ward, 1989). In yeasts that can metabolize xylose, uptake occurs via active transport (proton symport), which results in energy expenditure for the yeast to utilize the sugar (Jeffries et al, 1998). For each gram of xylose sugar metabolized by Saccharomyces, 0.5lg of ethanol are theoretically produced according to the following stoichiometry (Spencer et al, 1983). CsFbtA ^ 5/3 C 2H 5OH + 5/3 C0 2 D-Xylose Ethanol Figure 2-2 Theoretical ethanol yield from xylose (reproduced from Spencer et al, 1983). While the theoretical stoichiometric yield is 0.5lg EtOH / g xylose, it has been reported that only 60% of this maximal fermentation yield is attainable (i.e. 0.3g EtOH / g xylose metabolized) (Spencer et al, 1983). If it is assumed that 10% of the xylose was consumed in the formation of biomass, then the supposition is that 30% of the xylose is used in cell maintenance and/or the formation of undesirable metabolic products (Spencer et al, 1983). The over-simplified stoichiometry provided above for ethanol production summarizes the result of the reactions in the glycolytic or Embden-Meyerhof pathway (EMP). This pathway CHAPTER 2. LITERATURE REVIEW 9 is used by yeasts in the breakdown of sugars into energy, intermediates required for cell growth, and large amounts of the major end-products of fermentation, namely ethanol and CO2 which are excreted (Ingledew, 1995). These pathways are described in great detail in the literature (Ingledew, 1995; Spencer et al, 1983; Ward, 1989), and thus wil l not be discussed here. 2.1.3 Selection of Organisms for Fermentation 2.1.3.1 Single Organism Fermentations with Naturally Occurring Strains Saccharomyces is the dominant organism used for fermentation protocol (Crueger and Crueger, 1990; Stanbury et al, 1995; Demain and Solomon, 1985). The reasons for this are numerous. It can grow on a wide range of sugars, and is capable of fermenting glucose, galactose and mannose to ethanol (Ho et al, 1998). Unfortunately, the fungi are incapable of naturally fermenting pentoses (Ho et al, 1998; Demain and Solomon, 1985). Saccharomyces exhibits significant tolerance to ethanol, with metabolic functions largely unaffected at concentrations as high as 10% depending on carbohydrate feeding strategy and the period of adaptation to high ethanol concentrations (Crueger and Crueger, 1990). O f major importance is the level of understanding of the metabolic pathways in the different strains of Saccharomyces yeast. Saccharomyces has been used in various fundamental biotechnological applications for literally thousands of years, and the understanding of its regulation is second perhaps only to Escherichia coli. Saccharomyces is capable of producing ethanol at a maximal rate of 82 g/L/hr in a typical stirred-tank reactor (Crueger and Crueger, 1990). Over 96% of fermentation ethanol is produced by strains of Saccharomyces cerevisiae or related species (Ward, 1989). Other organisms, particularly bacteria, but including other yeast, are receiving increased attention for use in fermentation for ethanol production. One of the most promising of CHAPTER 2. LITERATURE REVIEW 10 the "alternative fermenters" is Zymomonas mobilis, a bacterium (Crueger and Crueger, 1990; Stanbury et al, 1995). While this bacterium has demonstrated distinct advantages over Saccharomyces, there are also additional complications. Advantages include a substantial osmotic tolerance to sugars, allowing Zymomonas mobilis to withstand concentrations up to 400 g/L (Crueger and Crueger, 1990). The species is also more ethanol tolerant than Saccharomyces, resisting concentrations up to 13%, compared to a more conservative 10% average for the yeast (Crueger and Crueger, 1990). Zymomonas mobilis also has a substantially faster specific growth rate (0.27 h"1) and ethanol production rate than Saccharomyces (Crueger and Crueger, 1990). Zymomonas mobilis is capable of producing ethanol at a maximal rate of 120 g/L/h (Crueger and Crueger, 1990). While it would seem that this bacteria is a worthwhile alternative to Saccharomyces for the fermentation of lignocellulosic residues, unfortunately this is not the case. Though specific growth rates and production rates are high, wild strains of Zymomonas are capable of growth on only three carbon sources: glucose, sucrose and fructose (Crueger and Crueger, 1990). This makes it an impractical choice for fermentation of lignocellulosic residues. Although the exact chemical composition differs from species to species, sugars in the hemicellulose fraction of softwoods are primarily galactoglucomannans, and pentoses such as arabinose and xylose (Cameron, 1998). Of these 5 sugars (glucose, mannose, galactose, xylose, and arabinose) only one is fermentable to ethanol by Zymomonas. This decreases observed yields by Zymomonas dramatically compared to Saccharomyces, which is capable of growth on all the sugars present in the hydrolysate except the xylose fraction (Crueger and Crueger, 1990). Lost productivity due to the inability of Saccharomyces to ferment pentoses is very small on softwood hydrolysate, as pentoses only make up 5-8% of the hemicellulose fraction of softwoods. Recently, CHAPTER 2. LITERATURE REVIEW 11 experimentation with a genetically engineered Z. mobilis strain, designed to express a xylose metabolic pathway, has indicated that Z. mobilis may be of commercial interest in the future (Jeffries and Himmel, 1998). Still other organisms are available, including, but not limited to Escherichia coli (bacteria), Pachysolen tannophylus (yeast - suitable for pentose fermentation), Pichia stipitis (another yeast - also suitable for pentose fermentation), and Mucor spp. (fungi) (Crueger and Crueger, 1990). Often there are applications where one organism is better suited to a particular fermentation process, due to the presence of inhibitors, acidity, and the specific sugar composition of the substrate. However, this does not mean that they can be employed in every situation. Though some species demonstrate reasonable yields of ethanol, they are generally not used because of decreased productivity (Crueger and Crueger, 1990). Saccharomyces is a very robust organism, and as such, it can be employed in various applications. 2.1.3.2 Multi-organism Fermentations The previous section considered microorganism selection when only a single organism would be used to ferment the hemicellulose hydrolysate. Typically a single organism fermentation process of this kind would be carried out in a single reactor. Due to the aforementioned issue regarding the inability of Saccharomyces to ferment both the hexose and pentose fractions of hemicellulose hydrolysate, some consideration has been given to the use of two reactors in series (Grootjen et al, 1991). Doing so would allow the utilization of 2 species, each optimized for its own fermentation of hexoses or pentoses. The efficient use of both pentoses and hexoses is especially important for the bioconversion of hydrolysate from hardwood species, where the percentage of pentoses is much CHAPTER 2. LITERATURE REVIEW 12 h i g h e r ( a p p r o x i m a t e l y 2 5 % o f the h e m i c e l l u l o s e f r ac t ion ) . F o r s o f t w o o d s , v i r t u a l l y a l l o f the sugars present are f e rmen tab le b y Saccharomyces cerevisiae, so the s m a l l pentose f r ac t i on w o u l d not j u s t i f y the c o m p l e x i t y o f a m u l t i - o r g a n i s m fe rmen ta t ion p roces s . U n f o r t u n a t e l y , the c h a l l e n g e s are s i g n i f i c a n t w i t h o p e r a t i n g t w o fe rmente rs i n ser ies w i t h d i f fe ren t o r g a n i s m s . In the s tudies p e r f o r m e d b y G r o o t j e n et a l . ( 1 9 9 1 ) , t w o reactors w e r e opera ted i n ser ies w i t h Saccharomyces cerevisiae i n the f i rs t reac tor , a n d Pichia stipitis i n the s e c o n d . In these c o n t i n u o u s f e rmen ta t ion s tudies , the researchers f o u n d that o n l y 2 0 % o f the x y l o s e w a s c o n s u m e d i n the s e c o n d reactor , due to c o m p e t i t i o n b e t w e e n Saccharomyces a n d Pichia fo r o x y g e n in the s e c o n d reactor ( G r o o t j e n et a l , 1991) . T h i s c o m p e t i t i o n resu l t ed due to the c a r r y o v e r o f r e s i d u a l hexose f r o m fe rmente r #1 to fe rmente r #2, d e c r e a s i n g the e f f i c i e n c y o f Pichia i n the s e c o n d stage ( G r o o t j e n et a l , 1991) . A s a resul t , G r o o t j e n et a l . ( 1 9 9 1 ) c o n c l u d e d that a c o m b i n e d s y s t e m w i t h Saccharomyces cerevisiae a n d Pichia stipitis w o u l d o n l y be c o m p e t i t i v e i f the yeasts w e r e c o m p l e t e l y separated ( G r o o t j e n et a l , 1991 ) . D u e to the p r a c t i c a l d i f f i c u l t i e s w i t h a c o n f i g u r a t i o n i n v o l v i n g the c o m p l e t e separa t ion o f t w o spec i e s i n a f e rmen ta t ion p rocess , it is de s i r ab l e to f i n d (o r create) , a s i n g l e spec ies that is c a p a b l e o f f e r m e n t i n g bo th the hexose a n d pentose f rac t ions o f h e m i c e l l u l o s e h y d r o l y s a t e . 2.1.4 Fermentation Configurations for Ethanol Production T h e r e are seve ra l op t i ons for the c o n f i g u r a t i o n used i n the f e rmen ta t i on o f h e m i c e l l u l o s e h y d r o l y s a t e . C o n f i g u r a t i o n o p t i o n s fo r the f e r m e n t a t i o n p rocess i n c l u d e ba tch ( w i t h o r w i t h o u t r e c y c l e ) , f ed -ba tch , a n d c o n t i n u o u s ( w i t h o r w i t h o u t r e c y c l e ) . CHAPTER 2. LITERATURE REVIEW 13 2.1.4.1 Batch Fermentation Configuration for Ethanol Production Batch fermentation is a closed culture system that contains an initial amount of nutrients and substrate, which undergo bioconversion into the desired product(s). In batch fermentation, the accumulation of ethanol in the reactor will inhibit productivity at a certain concentration, and unchecked, it will eventually stop the fermentation completely (Crueger and Crueger, 1990). The ultimate amount of product produced is also limited by the concentration of initial substrate the microorganism can tolerate (Crueger and Crueger, 1990; Stanbury et al, 1995). Batch fermentations have the possible advantage of selecting for strains that are tolerant to a higher initial concentration of substrate. This is particularly applicable to batch fermentations in which the biomass is recycled into a subsequent batch reaction (Stanbury et al, 1995). In the batch reaction the culture undergoes several metabolic stages, some of which are inherent but not desirable in terms of process productivity. A batch reactor undergoes a lag phase (a period in which the microorganism adjusts to the environment) a logarithmic growth phase (where the microorganism population undergoes the maximum growth rate), a declining growth phase (where the population growth begins to decline), and the stationary phase (where population growth is negligible) (Crueger and Crueger, 1990; Shuler and Kargi, 1992; Stanbury etal, 1995). CHAPTER 2. LITERATURE REVIEW 14 Lag Phase Log or Exponential Phase D Stationary Phase e e e a o n P h a e Time Figure 2-3 Growth of a typical microbial culture in batch conditions (reproduced from Substrate conversion to product is related to the reaction phase, as the metabolic state of the microorganisms governs the allocation of resources. During the lag phase the production of ethanol is negligible (Asenjo et al, 1991; Stanbury et al, 1995), as the inoculum adjusts to the environment. During the logarithmic phase the yeast is producing ethanol at the maximum rate as a by-product of the maximum growth occurring during this phase. Minimal ethanol production occurs during the stationary phase as the net growth rate is zero and minimal substrate is available in this phase (Stanbury et al, 1995). Batch fermentation offers the advantages of requiring lower technical skills, it is relatively easier to keep the system free of contamination, and the process is not integrally dependent on the successful operation of upstream or downstream processes. Batch reaction processes are proven, and because of their simplicity, they are well accepted in industry. This process configuration also offers the advantage of versatility, in that the fermentation conditions Stanbury et al, 1995). CHAPTER 2. LITERATURE REVIEW 15 can be easily changed from batch to batch. This flexibility increases the chance that high plant efficiencies will be achieved, even when the production of ethanol must occur from a mixture of sugar sources, and operating conditions are varied from batch to batch. 2.1.4.2 Fed-batch Fermentation Configuration for Ethanol Production Another configuration variation is a fed-batch fermentation reaction. In a fed-batch fermentation the culture is subjected to tolerable concentrations of substrate initially, and then the process is fed a series of substrate additions which both feed the culture and simultaneously dilute the product concentration in the reactor (Crueger and Crueger, 1990; Shuler and Kargi, 1992; Stanbury et al, 1995). This method of feeding allows rough control over the microorganisms in terms of their phase (preferentially the log phase) in which they are metabolizing the substrate to product at a maximal rate. 2.1.4.3 Continuous Fermentation Configuration for Ethanol Production The use of a continuous fermentation configuration can eliminate the problems encountered in a batch configuration associated with the accumulation of product or lack of substrate. The reactor is continuously supplied with fresh substrate (media) and product is continuously removed. The rate at which substrate is added and product is removed determines the metabolic phase of the microorganisms (Crueger and Crueger, 1990; Shuler and Kargi, 1992; Stanbury et al, 1995). Once the rates of substrate addition and product removal are optimized, the yeast can be held in a phase of growth where the bioconversion is optimized for maximum ethanol production and minimal substrate losses. In theory, a continuous fermentation process can proceed indefinitely, so long as washout or contamination does not occur. Recycling of the cells in a continuous reactor is typical CHAPTER 2. LITERATURE REVIEW 16 to help ensure a high fermentation rate, and that washout does not occur (Stanbury et al, 1995). Immobilized continuous reactors utilize entrapped cells to ensure that washout does not occur. The entrapped cells are subjected to minimal shear stresses, but due to the mass transfer limitations imposed by such a system the conversion of substrate into product is not optimal, thus the additional cost of capsulation is not economic (Crueger and Crueger, 1990). Continuous reactors have been the focus of many studies, which have revealed that they are superior to batch reactors in terms of throughput, but there are inherent problems to using a continuous reactor. The continuous supply of sterile substrate media is one of the largest problems (Shuler and Kargi, 1992). Contamination of a continuous process is problematic because the process must be completely shutdown, and extensive plant sterilization must occur. A plant shutdown for sterilization would result in lost production during the sterilization process. In addition, the re-establishment of steady-state fermentation operation may take an extended period of time (> 4 volume changes). The process is further complicated by the integral dependence of continuous upstream facilities and maintenance of exact process control, thus the establishment of steady-state operation may require relatively more technical skill. In some cases, the cost of the technical expertise required in the process may offset any gains experienced with greater production rates. Finally, there are increased difficulties regarding distillation of the fermentation broth to a concentrated ethanol product. Since the steady-state concentration in a continuous fermentation is lower than the final product concentration of batch fermentation, the cost for the downstream distillation of the product can increase significantly (Stanbury et al, 1995). CHAPTER 2. LITERATURE REVIEW 17 2.1.5 Operating Conditions for Fermentation with Saccharomyces cerevisiae 2.1.5.1 Aeration Requirements for Cell Growth and Fermentation Yeasts are considered facultative anaerobes, meaning they have the ability to grow in the presence or absence of oxygen (Ingledew, 1995). Under aerobic conditions, where a typical aeration rate would be 1 vvm (1 volume of air per volume of medium per minute), yeasts are capable of growing rapidly to high cell yields (0.5g yeast / g sugar) (Ingledew, 1995). In order to maximize the cell yield in this respirative mode of growth, substrate concentrations are ideally kept low (<0.1% w/v) through incremental feeding of media. This ensures that all sugars are metabolized to CO2 and water, which maximizes the conversion to biomass by eliminating by-product formation such as ethanol (desirable during yeast propagation) (Ingledew, 1995). Under anaerobic conditions, cell yield is minimal (less than 0.05g yeast / g sugar) as virtually all of the substrate is converted to ethanol and other fermentation products (0.5g EtOH / g sugar). While alcoholic fermentations are conducted under anaerobic conditions, a small amount of oxygen is required for the yeast to properly synthesize some sterols and unsaturated fatty acid membrane components (Owen, 1989). 2.1.5.2 pH Requirements in Fermentation Yeasts favour pH values between 4-6 for both growth and fermentation activity. Fermentation generally proceeds faster at higher pH, and there is a noticeable lag in fermentation activity between a pH of 3-4 (Owen, 1989). The ability for yeast to strive and produce high levels of ethanol at such low pH levels is advantageous from an industrial perspective, because most bacteria prefer more neutral pH values, thus, the chances for contamination are somewhat reduced (Owen, 1989). CHAPTER 2. LITERATURE REVIEW 18 2.1.5.3 Temperature Requirements in Fermentation Most yeast will ferment with good efficiency in a temperature range of 15-35°C, with the fermentation rate generally increasing with increased temperature (Owen, 1989). From an industrial operating standpoint, the higher the temperature that the fermentation organism can tolerate and still produce high levels of ethanol, the better. The ability to operate at higher temperatures means lower cooling requirements for the plant, and because more species of bacteria thrive in the mid-mesophilic temperature range (15-25°C), the chances for contamination are reduced at elevated temperatures. 2.1.6 The Economics of Ethanol Production Since many lignocellulose-based substrates contain a substantial concentration of pentose sugars (primarily xylose), the ability to coferment xylose is considered by some to be essential for the economic production of ethanol from these feedstocks (Jeffries et al, 1998). Despite concentrated studies on various aspects of xylose fermentation in many laboratories in recent years, the fermentation of xylose is not yet commercial (Jeffries et al, 1998). Presently, large-scale cofermentation of xylose with wild-type yeast species and genetically modified yeast strains is not economic because these strains pose the problems of low yield and instability, respectively (Jeffries et al, 1998). A method for efficiently converting xylose into ethanol would have a dramatic effect on the economics of ethanol production, as xylose can account for 20-50% of the sugars in lignocellulose hydrolysate. 2.2 Conversion of Spent Sulfite Liquors to Ethanol Any medium used in a fermentation process must meet certain basic microbial requirements by having water, carbon, nitrogen, mineral elements and vitamins, and some CHAPTER 2. LITERATURE REVIEW 19 amount of oxygen (depending upon the growth phase) readily available to the microbes. Ideally, the medium is also optimized to produce the maximum yield of product from the substrate, at a maximal rate with undesirable by-product formation held to a minimum (Stanbury et al, 1995). In the conversion of spent sulfite liquors to ethanol, the medium is anything but ideal. Spent sulfite liquor will not only have variable concentrations of sugars (due to the production of different grades of pulp), but will also have variable pH, osmotic strength, and concentrations of toxic compounds (such as SO2, acetic acid, furfural, hydroxymethylfurfural or HMF, and cinnamaldehyde). The effects on yeast metabolism of these variable substrates and fermentation inhibitors, which are inherent to the feed streams, are complex to say the least (Cameron, 1998). This section describes the sulfite pulping process, and highlights specific aspects of the Tembec operation. Typical compositions for the hardwood and softwood spent sulfite liquors in the plant are provided, as well as information on the baseline fermentation efficiency achieved in the full-scale plant. Finally, details are provided on some original testing conducted with the Tembec plant strain (T2) and Pichia stipitis, in the pentose pilot plant. 2.2.1 Description of the Sulfite Pulping Process The sulfite pulping process involves the digestion of wood with a calcium acid sulfite cooking liquor, which is a mixture of calcium bisulfate and excess sulfurous acid. Liquor recovery from the sulfite pulping process is relatively difficult as compared with that in the chlorine or sulphate-based processes, resulting in significant losses and therefore a higher strength waste stream from the process (Smook, 1992). In general, the sulfite pulping process results in superior lignin removal and therefore the papermaking fibers are naturally whiter than those from a sulfate process. The natural CHAPTER 2. LITERATURE REVIEW 20 whiteness of the fibers facilitates them being bleached to high brightness with relatively low chemical use. A downside of the sulfite pulping process is that the paper made from sulfite fibers is not as strong as that from sulfate pulp (Smook, 1992). 2.2.1.1 Typical Sulfite Pulp Mi l l Waste Stream Flowsheet In a sulfite pulp mill, large quantities of spent sulfite liquor (SSL) are produced as a by-product of the sulfite pulp process (Bjorling and Lindman, 1989). The losses in liquor recovery, combined with the superior lignin removal in the pulping process, result in a high strength SSL waste stream containing hexose and pentose sugars. In sulfite pulp mills, much of the SSL is sent to recovery boilers. Still, a significant portion is routed to wastewater treatment, where a biological treatment process oxidizes the sugar "contaminants" in the waste stream (Figure 2-4). Fiber Supply Sulfite 1 Pulp Mi l l Spent Sulfite Liquor I Market Pulp Secondary Wastewater 1 Treatment Effluent Figure 2-4 Simplified sulfite pulp mill flowsheet. 2.2.1.2 Tembec Inc. Waste Stream Flowsheet At the Tembec Inc. sulfite pulp mill in Temiscaming, Quebec, a more integrated approach is taken to waste management. Rather than simply wasting excess SSL directly to a treatment plant, the SSL "waste stream" is utilized, with value added products extracted from the stream. Prior to wastewater treatment, the SSL is processed by the Specialty Chemicals Group, CHAPTER 2. LITERATURE REVIEW 21 which results in the development of commercial products from the pulp mill waste products streams (Figure 2-5). The Specialty Chemicals Group at Tembec consists of alcohol production, as well as lignin and resin recovery. Fiber Supply (4-Sulfite Pulp Mill Market Pulp Spent Sulfite Liquor Specialty Chemicals Group Secondary Wastewater Treatment Effluent Figure 2-5 Simplified Tembec flowsheet. The alcohol group at Tembec produces high purity ethanol (96%) from the fermentation of spent sulfite liquor. The ethanol is used in chemical, cosmetic, and pharmaceutical applications, as well as detergents, food, and industrial products. Tembec is a major supplier of industrial alcohol in eastern Canada, with annual production averaging approximately 18 million Liters/year (Cameron, 1998). The lignin group produces ammonium and sodium lignosulfonates, which are used as binders and surfactants. Because the sulfite pulping process is superior in lignin removal, a large amount of lignin is available in the SSL waste stream. Tembec produces approximately 170,000 tonnes per year of liquid lignosulfonate product (Tembec, 1998). The lignin group gets its SSL directly from the alcohol group, and as such, the fermentation of the xylose fraction would directly affect operations (Cameron, 1998). Lower xylose content in the SSL would mean lower caustic requirements by the lignin group (which is used to raise the pH, volatilizing CHAPTER 2. LITERATURE REVIEW 22 ammonia, which is then removed by steam stripping) (Cameron, 1998). Not only would there be chemical savings, lower caustic additions would also reduce the ash content which is a quality parameter for products from the lignin group. Lower ash content could result in more and different markets for products (Cameron, 1998). The resin group produces phenol-formaldehyde resins, which are adhesives that are designed for the oriented strand board (OSB) industry. Tembec produces 26,000 tonnes of powdered resins per year, and 45,500 tonnes of liquid resins per year (Cameron, 1998). The following figure shows a basic solids balance in the Tembec pulp mill . Wood chips 12001. Spent Sulphite Liquor 700 t. Pulp 500 t. JUL Boiler 485 t. > > f > f Ethanol Lignosulfonate Waste Sales Sales Treatment 70 t. 110t. 35 t. Figure 2-6 Basic solids balance in the Tembec pulp mill (Cameron, 2000). As shown in Figure 2-6, approximately 70% of the spent sulfite liquor is ultimately burned in a recovery boiler to make steam on site after having passed through the processes in the Specialty Chemicals Group. The fermentation of the xylose fraction of SSL would also directly affect the recovery boiler, in that lower xylose content in the SSL would result in a loss CHAPTER 2. LITERATURE REVIEW 23 of recovery boiler efficiency due to a lowered heating value of the waste stream (Cameron, 1998). Finally, approximately 5% of the spent sulfite liquor is ultimately treated in a biological wastewater treatment plant after having passed through the Specialty Chemicals Group. The removal of xylose from the SSL would decrease the BOD load on the treatment plant, decreasing the aeration requirements (which would save some operating cost). More important than the potential operating cost savings is the fact that the wastewater treatment plant at Tembec is already overloaded, so any decrease in BOD load would be beneficial. Tembec's production of silvichemicals (chemicals produced from wood, i.e. alcohol and lignosulfonates) represents a sustainable approach pulp manufacture, and demonstrates environmental stewardship through the responsible utilization of forest resources while supporting environmental goals. 2.2.2 Composition of Hardwood and Softwood Spent Sulfite Liquors The proportion of xylose and hexose sugars varies widely in the Temiscaming mill; as many as 10 different grades of softwood, hardwood, or blended pulps can be produced. The proportion of sugars typically found in softwood spent sulfite liquor is approximately 75% hexose and 25% xylose, at concentrations of roughly 30 g/L and 10 g/L respectively. In hardwood SSL the balance differs, in that the ratio of xylose to hexose is approximately 50/50 (Cameron, 1998). Table 2-1 summarizes the typical composition of softwood and hardwood spent sulfite liquor. CHAPTER 2. LITERATURE REVIEW Table 2-1 Typical composition of softwood and hardwood liquor (Cameron, 2000). 24 Parameter Softwood (g/L) Hardwood (g/L) Dissolved Solids 240 240 Acetic Acid 4 10 Hexose Sugars Mannose 19 9.5 Glucose 6 4 Galactose 5 2.5 Total Hexoses 30 16 Theoretical E tOH (Hexose) 15 8 Pentose Sugars Xylose 8 14 Arabinose 1 1 Total Pentoses 9 15 Theoretical E tOH (Hexose + Pentose) 19.5 15.5 The pH in the spent sulfite liquor from the pulping process is approximately 2, and is raised to 5 prior to fermentation. Dissolved solids in solution typically average approximately 24% w/w. Acetic acid concentrations vary between 4-10 g/L depending upon the grade of pulp being produced, and furfural averages 1 g/L. Other toxic compounds such as SO2 are also present in varied concentrations. 2.2.3 Description of the Tembec Alcohol Plant The full-scale alcohol plant processes SSL that has been stripped of its excess of SO2, water, acetic acid, and other volatiles by the upstream sulfite mill evaporators. SSL is fed to the plant at a rate of approximately 2,000 L/min. The SSL is first fed to a settling tank, which feeds liquor forward to two, one million litre fermenters operated in series. The pH adjustment from 2 to 5 with ammonium hydroxide occurs in the settling tank, which encourages the precipitation of calcium oxalate before it can get into the fermentation system. The fermenters are operated on CHAPTER 2. LITERATURE REVIEW 25 a continuous basis, and provide an average combined H R T of approximately 12.5 hrs. The fermentation process is controlled for pH and temperature (set points for these two variables are 5.0 and 28°C, respectively), and minimal aeration is provided continuously in both fermenters. A surge tank after the fermenters provides a consistent feed to two centrifuges, which separate the yeast from the beer on a continuous basis. The yeast is recycled back to the first active fermenter, and the beer is sent to a series of distillation columns that purify the product from approximately 1.5% to 96%. Figure 2-7 shows a simplified schematic of the Tembec alcohol plant flowsheet. SSL from , evaporators 2,000 L/min 9 6 % Ethanol Fermentation Distillation Yeast 4—i- Beer C ^ c ^ Settling Fermenter Fermenter Tank #1 #2 Beer Separation Figure 2-7 Simplified Tembec alcohol plant flowsheet. 2.2.3.1 Current Alcohol Production with the Tembec Alcohol Plant Saccharomyces Strain The Saccharomyces strain utilized in the Tembec alcohol plant was originally a strain of Saccharomyces cerevisiae used in wine making (Cameron, 1998). Over time, this strain has adapted and acclimatized to the SSL feedstock, and has subsequently mutated from its original form. The plant strain (also called T2) exhibits high ethanol tolerance and fast fermentation rates, CHAPTER 2. LITERATURE REVIEW 26 quickly fermenting all hexose sugars (mannose, glucose, and galactose) to completion and even taking up a small amount of xylose. The strain is considered very robust, and is capable of operating at low pH, helping to minimize the chance of contamination under the aseptic conditions. Perhaps the most important aspect of T2 for the Tembec alcohol group is the strain's resistance to inhibitory substances well outside the accepted "tolerable range" cited in the literature for Saccharomyces cerevisiae (Cameron, 1998). For example, in the presence of high osmotic strength (with dissolved solids >24% w/w), coupled with low pH (< 4.5), and toxins (such as SO2, acetic acid, hydroxymethyl furfural and resin acids), T2 can still achieve acceptable fermentation efficiencies (>80% on softwood, and >70% on hardwood). This is important from an operability standpoint, because the variability in the feedstock with respect to sugar and toxin concentrations is in essence "buffered" by the stability of this robust Saccharomyces strain. Figure 2-8 is a photomicrograph of the Tembec alcohol plant Saccharomyces strain. The photo illustrates a common characteristic of this yeast, which is "balling" or "clumping" (Cameron, 1998). This pattern of cell aggregation occurs during all growth stages, and during steady-state operation. This pattern of growth is somewhat unique, because in general, the clumping and binding together of cells is an indicator of stress on the microbial population (Cameron, 1998). CHAPTER 2. LITERATURE REVIEW 27 Figure 2-8 Tembec alcohol plant strain of Saccharomyces cerevisiae (lOOx magnification). For all of its advantages, T2 still leaves the pentose sugars (xylose and arabinose) unconsumed, amounting to a 20-50% loss through unfermented sugars in the process. 2.2.3.2 Ethanol Production by the Tembec Strain on Softwood and Hardwood SSL Typical levels of ethanol production by T2 in the full-scale alcohol plant on both softwood and hardwood SSL are shown in Table 2-2. Also shown are the typical fermentation efficiencies on the SSL substrates (St. Onge, 1996). Table 2-2 Full-scale ethanol production on softwood and hardwood SSL (St. Onge, 1996). Typical Hexose [EtOH) [EtOH] Fermentation SSL Sugars theoretical* actual Efficiency** (g/L) (g/L) (g/L) (%) Softwood 45.0 23.0 18.0 78.3 Hardwood 30.0 15.3 10.0 65.4 * calculated as [EtOH] theoretical = [Hexose Sugars] x 0.51 ** calculated as Efficiency (%) = ([EtOH] actual + [EtOH] theoretical) x 100 CHAPTER 2. LITERATURE REVIEW 28 The full-scale alcohol plant achieves acceptable fermentation efficiency on softwood, and a less-than-optimal result when fermenting hardwood hydrolysate. In general, lower fermentation efficiency is expected during hardwood fermentation, as there are proportionately more inhibitory compounds present. Studies in the mill on the effects of pH, osmotic pressure, organic acid production, and the toxic effects of acetic acid, furfural, and sulfite, have aimed at improving the fermentation efficiency in the plant (St. Onge, 1996). Of course, if the pentose sugars that currently escape fermentation were converted to ethanol, the fermentation efficiency could theoretically exceed the stoichiometric maximum from hexoses alone. 2.3 Xylose Fermentation by Naturally Occurring Strains It has been established that currently, while achieving high fermentation efficiency on hexose, the Saccharomyces strain used in the Tembec alcohol plant is incapable of utilizing the xylose fraction present in spent sulfite liquor. Ultimately, the goal is to identify (or create) a xylose-fermenting, mutant yeast strain capable of increased ethanol production. This section explores the major differences that enable naturally occurring species and strains of yeast to utilize xylose. Specifically, experience with Pichia stipitis is cited, as it is one of the most researched, and successful yeasts for xylose fermentation. Finally, pathways related to xylose fermentation that are inherent to Saccharomyces cerevisiae are discussed. 2.3.1 Xylose Fermentation by Pichia stipitis There are many naturally occurring species and strains of xylose-fermenting yeast, which have been rigorously studied and compared under different fermentation conditions (Jeffries and Kurtzman, 1994; Toivola et al, 1984; Bjorling and Lindman, 1989). One of the CHAPTER 2. LITERATURE REVIEW 29 most researched, and successful, yeasts with the ability to naturally ferment xylose, is Pichia stipitis (Jeffries et al, 1998). 2.3.1.1 Sugar Metabolism in Pichia stipitis Unlike Saccharomyces, which for the most part takes up sugars by facilitated diffusion, Pichia stipitis uses both high-affinity and low-affinity proton symport mechanisms which require the expenditure of metabolic energy (Kilian and van Uden, 1988; Does and Bisson, 1989). Studies have shown that the presence of glucose inhibits the transport of xylose by P. stipitis in a competitive manner, which indicates the same transport system is used for these sugars (Jeffries et al, 1998; Kilian and van Uden, 1988; Does and Bisson, 1989; Prior and Kotter, 1997). The enzymes for xylose metabolism in Pichia stipitis are inducible by xylose and repressible by glucose (Jeffries et al, 1998; Bicho et al, 1988). Xylose metabolism in Pichia begins with the conversion of xylose to xylitol by xylose reductase (XR), which is encoded for by a gene with the designation XYL1. The reduction of xylose to xylitol by XR uses NADH or NADPH, and NADPH is thought to be used preferentially (Johansson, 2001; Rizzi et al, 1989). In P. stipitis, XYL1 is induced by growth on xylose (Jeffries et al, 1998). The gene for xylose reductase has been cloned from P. stipitis by at least three independent laboratories (Jeffries et al, 1998; Takumaetal, 1991; Amore et al, 1991; Hallborn et al, 1991). Xylitol is oxidized to xylulose by xylitol dehydrogenase (XDH), which is encoded for by a gene with the designation XYL2 (Johansson, 2001; Rizzi et al, 1989). Unlike XR, the enzyme XDH is always specific for the cofactor N A D + (Johansson, 2001; Rizzi et al, 1989). Finally, energy is consumed in the conversion of xylulose to xylulose-5-P by xylulokinase (XK), which is encoded for by a gene with the designation XYL3 (Jeffries et al, CHAPTER 2. LITERATURE REVIEW 30 1998). Xylulose-5-P then enters the Pentose Phosphate Pathway (PPP) for conversion to ethanol. The major stages and responsible enzymes in the metabolism of xylose by Pichia stipitis are shown in Figure 2-9. Xylose XYLl (XR) c Xylitol c XYL2 (XDH) Xylulose XYL3 (XK) Xylulose-5-P NAD(P)H NAD(P) + N A D + NADH c ATP ADP Figure 2-9 Initial steps of xylose pathway in yeasts and filamentous fungi. XYLl codes for xylose (aldose) reductase; XYL2 codes for xylitol dehydrogenase; XYL3 codes for xylulokinase (reproduced from Jeffries et al, 1998). 2.3.1.2 Operational Considerations with Pichia stipitis The particular sensitivity of Pichia stipitis to the rate of aeration and oxygen availability has been well documented (Jeffries et al, 1998; Delegenes et al, 1986; Ligthelm et al, 1988; du Preez et al, 1988; Dellweg et al, 1989; Grootjen et al, 1990; Guebel et al, 1991). Studies have indicated that while the levels of xylose reductase, xylitol dehydrogenase, and their associated cofactors do not change appreciably under anoxic fermentation conditions, enzyme activity decreased significantly, possibly contributing to sub-optimal ethanol production (du CHAPTER 2. LITERATURE REVIEW 31 Preez et al, 1989). Thus, for Pichia stipitis the fermentation is regulated by oxygen availability, and low amounts of well controlled oxygen are required for optimal performance (Skoog and Hahn-Hagerdal, 1990). Fermentation efficiency is optimal for Pichia at a temperature of approximately 25°C, and a pH between 4-5 (Slininger, 1990; du Preez et al, 1986). 2.3.1.3 Ethanol Production and Fermentation Efficiency oi Pichia stipitis Pichia stipitis has exhibited a specific productivity of 0.20g EtOH / g xylose / hr, and an ethanol yield of 0.48g EtOH / g xylose, which represents a fermentation efficiency of approximately 94% (Skoog and Hahn-Hagerdal, 1990). While this level of conversion efficiency is impressive, due to a relatively slow rate of fermentation, the overall productivity is only 0.38g EtOH / g biomass / hr (c.f. ethanol productivity by Saccharomyces cerevisiae on glucose is almost five times higher at approximately 2g EtOH / g biomass / hr) (Johansson, 2001). 2.3.1.4 Genetic Engineering oi Pichia for enhanced Fermentation Attempts have been made to genetically modify Pichia stipitis to exhibit some of the more favourable fermentation traits of Saccharomyces. Due to the inherent constraints with Pichia related to strict aeration requirements, the ability of Saccharomyces to ferment under anaerobic conditions is of particular interest. Research has found that Saccharomyces has a particularly unique adaptation for anaerobic growth, in its ability to produce uracil under anaerobic conditions, which is usually tied to respiratory metabolism (Jeffries and Himmel, 1998). By transforming Pichia stipitis with the gene that regulates the novel form of the enzyme responsible for this uracil production, it is hoped that the ability to ferment under anaerobic CHAPTER 2. LITERATURE REVIEW 32 conditions will be conferred, improving the stability of fermentation with Pichia (Jeffries and Himmel, 1998). 2.3.1.5 Summary on Fermentation using Pichia stipitis The major advantage with P. stipitis is the ability to ferment xylose to ethanol with very high fermentation yields. Because favourable mutant strains (i.e. non-engineered organisms) of P. stipitis are completely stable, they are advantageous from a commercial standpoint. In addition, because no foreign DNA is present, naturally occurring strains of P. stipitis are not considered "recombinant organisms" and thus do not have to be regulated or contained (Jeffries et al, 1998). Major disadvantages with P. stipitis include a relatively low ethanol tolerance, slow rates of fermentation, and sensitive aeration requirements (Skoog and Hahn-Hagerdal, 1990; Jeffries et al, 1998). Perhaps most important to Tembec is the fact that P. stipitis exhibits a low tolerance to harsh environments (Skoog and Hahn-Hagerdal, 1990). 2.3.2 Xylose Fermentation in Natural Strains of Saccharomyces cerevisiae While strains of Saccharomyces cerevisiae are incapable of naturally fermenting D-xylose (Ho et al, 1998; Demain and Solomon, 1985), they are able to metabolize D-xylulose, an intermediate of D-xylose metabolism, and good growth is attained (Demain and Solomon, 1985). Early research centered on the conversion of D-xylose to D-xylulose by inclusion of a glucose isomerase in the medium, which would effectively isomerase D-xylose to D-xylulose (Demain and Solomon, 1985). In a fermentation configuration that employed immobilized glucose isomerase and Saccharomyces yeast cells in separate columns, S. cerevisiae was capable of producing ethanol at high concentrations (Demain and Solomon, 1985). CHAPTER 2. LITERATURE REVIEW 33 Unfortunately, the costs associated with a two-stage fermentation reaction of this nature would be prohibitive (Demain and Solomon, 1985). Ideally, a Saccharomyces organism would be custom designed with the xylose-fermenting capabilities of Pichia stipitis, while maintaining the inherent hexose fermenting capabilities of the host organism. 2 . 4 Tembec Pilot Plant - Original Research Studies Most of the trials that have been completed in the pilot plant up to 1997 have involved the alcohol plant yeast strain to establish baseline results on both softwood and hardwood liquors. These trials also refined the process by which the pilot plant is inoculated, the yeast is propagated, and the timing of changes to process variables such as feed rates and aeration requirements. Ultimately though, this pilot plant was designed to test new yeast strains and their ability to coferment xylose and hexose sugars to ethanol. The following sections detail baseline testing conducted with the Tembec plant strain (T2), and testing with Pichia stipitis, a species known to have xylose-fermenting capability, in the pilot plant. The pentose pilot plant is described in great detail in Section 4.5 of Chapter 4, Materials and Experimental Methods, and won't be described here. 2.4.1 Baseline Fermentation Testing with the Tembec Plant Strain (T2) A pilot plant trial with the alcohol plant strain was conducted on softwood SSL. Because the plant strain (T2) is known to have a relatively fast rate of fermentation, only two fermenters were used in series for this trial with a feed rate of 2 L/min, providing an overall HRT in the fermenters of approximately 12 hrs (6 hrs per fermenter). CHAPTER 2. LITERATURE REVIEW ^ The steady-state sugar utilization through the pilot plant is shown in Figure 2-10. This graph illustrates how quickly hexose is fermented to ethanol by T2, as no hexose sugars remained after the first stage of the system. The results also show no uptake of xylose by T2. 50 S S L in Ferm. #1 Ferm. #2 Fe rm. #3 Fe rm. #4 Figure 2-10 Ethanol production by the Tembec alcohol plant Saccharomyces strain on softwood SSL in the pilot plant (Cameron, 2000). Using the ethanol and sugar concentrations for this day, the fermentation efficiency can be calculated, based on a maximum theoretical yield of 0.5 l g ethanol per gram of sugar utilized in the system. On this basis, T2 produced 0.4g EtOH / g sugar utilized, for a fermentation efficiency of approximately 80%. Assuming a cell concentration of 1.5 g/L (which would amount to 1.2 kg of cells in an 800 L active fermenter volume) the productivity of the system can be estimated. Ethanol production in the system was 2 kg EtOH/hr (16.5 g/L x 2 L/min), so the productivity in the system is estimated at 1.7g EtOH / g biomass / hr (2 kg EtOH/hr + 1.2 kg cells*, *all fermentation occurred in a single fermenter). CHAPTER 2. LITERATURE REVIEW 35 This represents a very fast rate of fermentation, which is advantageous in a commercial operation. In addition, this unit productivity may be conservative, as the fermentation had already reached completion within fermenter #1, so it is reasonable to assume that the system could have processed more substrate. These results attained with T2 correlate very well with the fermentation efficiency achieved in the full-scale plant on similar substrates, eliminating any concerns related to scale-up relevance of the results. 2.4.2 Baseline Fermentation Testing with Pichia stipitis An early pilot plant test involved Pichia stipitis due to its known capacity for xylose fermentation. The trial was conducted using softwood SSL, for comparison to earlier results obtained using T2 on softwood SSL in the plant. Because Pichia is known to have a relatively slow rate of fermentation, all four fermenters were used in series for this trial with a feed rate of 2 L/min, providing an overall HRT in the fermenters of approximately 24 hrs (6 hrs per fermenter). The steady-state sugar utilization through the pilot plant is shown in Figure 2-11. This graph illustrates the slow rate of fermentation by Pichia, as only 40% of the incoming hexose sugars are metabolized in the first fermenter. Steady uptake of xylose occurs after fermenter #2, and 70% xylose uptake results through the system when the hexose concentrations are suitably low to eliminate feedback repression. CHAPTER 2. LITERATURE REVIEW 36 SSL in Ferm. #1 Ferm. #2 Ferm. #3 Ferm. #4 Figure 2-11 Ethanol production and xylose uptake by Pichia stipitis on softwood SSL in the pilot plant (Cameron, 2000). Using the ethanol and sugar concentrations for this day, the fermentation efficiency can be calculated, based on a maximum theoretical yield of 0.5 l g ethanol per gram of sugar utilized in the system. On this basis, Pichia produced 0.35g EtOH / g sugar utilized, for a fermentation efficiency of approximately 70%. Assuming a cell concentration of 1.5 g/L (which would amount to 1.2 kg of cells in an 800 L active fermenter volume) the productivity of the system can be estimated. Ethanol production in the system was 1.5 kg EtOH/hr (12.3 g/L x 2 L/min), so the productivity in the system is estimated at 0.42g EtOH / g biomass / hr (1.5 kg EtOH/hr -H (3 x 1.2 kg cells)*, *fermentation complete by fermenter #3). In general, this pilot testing at Tembec with Pichia stipitis was unsuccessful. The low level of productivity represents a very slow rate of fermentation by Pichia, with the T2 producing ethanol at nearly 4 times the rate. This slow rate of fermentation is problematic because in the full-scale operation the retention time would be only 12 hrs, half the time CHAPTER 2. LITERATURE REVIEW 37 provided in this trial. At the 12-hr mark of this trial (after fermenter #2) little xylose had been consumed, and nearly 10 mg/L of hexose remained unfermented. These results are supported by the literature, as Bjorling and Lindman (1989) concluded that the specific ethanol production rate of P. stipitis was low compared to S. cerevisiae. In addition, there were indications that the organism was very sensitive to the inhibitory compounds found in lignocellulosic hydrolysates, and consequently the maximum yield of ethanol was low in relation to the initial quantity of hexose sugars (Cameron, 1998). Finally, Bjorling and Lindman (1989) determined that the relatively high pH requirements for Pichia would make bacterial infections a serious problem in a continuous fermentation process such as Tembec's. These factors preclude the use of Pichia stipitis in the Tembec alcohol plant. 2.4.3 Baseline Pilot Plant Fermentation Testing - Conclusions Baseline pilot plant testing with the alcohol plant strain (T2) produced ethanol at 80% fermentation efficiency on softwood SSL compared to a fermentation efficiency of 70% attained with Pichia stipitis. T2 had a net productivity of 1.7g EtOH / g biomass / hr (which may have actually been conservative), whereas Pichia had a very low level of productivity at 0.42g EtOH / g biomass / hr (less the one-quarter of that attained by T2). In general, Pichia s low levels of productivity, as displayed in this trial on softwood SSL, combined with its other known disadvantages, make it undesirable from an industrial standpoint for the conversion of SSL to ethanol. The ability of Pichia to metabolize more than 70%> of the xylose during this trial on softwood SSL was excellent. A synergy between the characteristics of T2, and the xylose-fermenting capability of Pichia would be very advantageous for the commercial operation. CHAPTER 2. LITERATURE REVIEW 38 2.5 Design of Pentose-Fermenting Saccharomyces cerevisiae Strains While Saccharomyces cerevisiae can readily ferment the hexose fraction of lignocellulose hydrolysate, it cannot use the xylose-rich pentose fraction, due to a lack of appropriate enzymes, and efficient transport and pathways for xylose (Jeffries and Himmel, 1998). In order to overcome the inherent limitations on xylose fermentation of naturally occurring Saccharomyces strains, strains can be genetically modified using techniques in metabolic and enzyme engineering. The hope is that these genetic modifications will result in the development of new yeast strains that are capable of producing ethanol from both the hexose and pentose fractions of lignocellulose hydrolysate. Increasingly, genetically engineered organisms with altered traits are being integrated into processes (Jeffries and Himmel, 1998). This section introduces general concepts in microbial genetics with respect to recombinant technology, and provides background on the current state of research on the genetic engineering of xylose metabolism in Saccharomyces cerevisiae. Finally, some specifics are provided on the construction and properties of the pentose-fermenting strains provided to Tembec by Dr. Nancy Ho, from Purdue University. 2.5.1 General Microbial Genetics and Techniques in Recombinant Engineering Typically, natural isolates or wild strains of industrial microorganisms produce a final product in relatively low concentrations and therefore attempts are made to increase the productivity of the organism (Stanbury et al, 1995). While increased yields can be achieved through the modification and optimization of the medium and operating conditions, the maximal synthesis of a product is still limited by the organism (Stanbury et al, 1995). Productivity is CHAPTER 2. LITERATURE REVIEW 39 ultimately governed by the genome of an organism, and as such, modifications to the genome are required in order to realize appreciable gains in productivity (Stanbury et al, 1995). Genetic engineering involves the transfer of DNA between different species in order to transfer desirable traits from one organism into another (Stanbury et al, 1995). Genetic engineering can provide a cell with a selective advantage by making a host more resistant -allowing it to outlive other cells or by providing more opportunities for food uptake. Ideally the genetic modifications to an organism are retained, and able to transfer to the entire population. The use of recombination techniques for the improvement of industrial strains has increased significantly in recent years. This is due to advances in recombinant DNA technology, and decreased returns from programs dedicated solely to strain mutation and selection (Stanbury etal, 1995). 2.5.1.1 Genetic Transformation using Extra-chromosomal DNA (plasmids) One recombination technique for the transformation of Saccharomyces strains involves the insertion of plasmids carrying cloned genes. Virtually all essential genes for an organism (i.e. the genes involved in biosynthesis, energy metabolism, DNA metabolism, etc.) are located on the organism's chromosome to ensure that the capability to perform these critical functions is not lost during cell replication (Stanbury et al, 1995). While these most important genes are located on the chromosome, a wide range of non-essential genes can be carried in an organism on extrachromosomal DNA or plasmids (Stanbury et al, 1995). The genes located on plasmids are not essential for viability or growth, but can confer on the cell a trait that gives it a selective advantage under particular environmental conditions (Stanbury et al, 1995). CHAPTER 2. LITERATURE REVIEW 40 Plasmids are closed circles of double stranded DNA that can replicate and transfer within a population. It has been estimated that as many as 50-60% of natural cells in the wild may harbour plasmids (Stanbury et al, 1995). Plasmids exert a very low metabolic cost for a population, as they are only expressed when a cell will gain a selective advantage by doing so (Cameron, 1998). Sophisticated experimental techniques have been developed by which the DNA from any biological source, both prokaryotic and eukaryotic, can be incorporated into bacterial plasmids. The basic requirements for the transfer and expression of foreign DNA in a host microorganism are as follows (Stanbury et al, 1995): 1. A 'vector' DNA molecule (plasmid or phage) capable of entering the host cell and replicating within it. Ideally the vector should be small, easily prepared and must contain at least one site where integration of foreign DNA will not destroy an essential function. 2. A method of splicing foreign genetic information into the vector. 3. A method of introducing the vector/foreign DNA recombinants into the host cell and selecting for their presence. Commonly used simple characteristics include drug resistance, immunity, plaque formation, or an inserted gene recognizable by its ability to complement a known auxotroph. 4. A method of assaying for the 'foreign' gene product of choice from the population of recombinants created. Inherent inefficiencies in the transformation process necessitate the incorporation of selectable genes into the vector DNA such that the transformed cells can be cultured preferentially from a mixture of transformed and parental cells (Stanbury et al, 1995). Plasmids are not a permanent part of the cell. Plasmids are inserted in the cytoplasm of a cell, and only increase in numbers if the cell deems it necessary to replicate the plasmid. If the plasmid is not replicated, the original number of plasmids will get halved with each new cell generation, and will eventually be lost (Cameron, 1998). Regular gene expression by the CHAPTER 2. LITERATURE REVIEW 41 application of selective pressures (i.e. the presence of xylose at all times) can be used to ensure replication of the plasmids (Cameron, 1998). 2.5.1.2 Genetic Transformation using Chromosome Manipulation Another recombination technique for the transformation of Saccharomyces strains involves the direct insertion of cloned genes on the chromosome of the organism. Genes that are inserted on the chromosome will be expressed along with the essential genes for the organism that are responsible for functions such as biosynthesis and energy metabolism. Using this approach, the stability of the transformed host is ensured, because the organism has no choice in the replication of the inserted genes (Cameron, 1998). Because there is no concern with respect to the replication of the integrated genes, the application of selective pressure is not required (Cameron, 1998). While the stability of the transformed host is ensured with respect to maintaining the cloned genes that were inserted, nothing in nature is guaranteed to be completely "stable", and stability may change with regard to more or less expression of the genes over time (Cameron, 1998). There are potential complications in the manipulation of the chromosome to create a stable transformant strain. During transformation, multiple copies of the genes are inserted into the chromosome. The random nature of the insertion results in chance insertion points on the chromosome, which can inadvertently interrupt, change, or affect other cell functions (Cameron, 1998). The potential adverse effects if cell functions are changed could be critical (i.e. if a vital cell function is affected, for example, biosynthesis impairment resulting in retarded cell replication), or undesirable (i.e. if a key characteristic or trait of the host strain is lost, for example, the loss of high tolerance to inhibitory substances) (Cameron, 1998). For example, both CHAPTER 2. LITERATURE REVIEW 42 the plates used to confirm cell gene expression and growth, and the rich media used to test fermentation capacity usually contain vitamins. If a transformant had it's promoter for the genes for vitamin synthesis, or the genes themselves disturbed during insertion, the yeast may not perform optimally in an SSL-based, non-enriched medium (Cameron, 1998). The fact that yeast cells are diploid helps mitigate this risk, because if a critical cell function is destroyed or impaired in one chromosome, the other can compensate (Cameron, 1998). 2.5.2 Genetic Engineering of Xylose Metabolism in Saccharomyces cerevisiae S. cerevisiae exhibits many traits, which give the yeast a demonstrated advantage in industrial ethanol production (Johansson, 2001). S. cerevisiae effectively ferments the hexose fraction of SSL. Pichia stipitis has a demonstrated ability to ferment the xylose fraction of SSL. Therefore, the goal of some researchers has been to use the Pichia genes responsible for xylose fermentation to transform a Saccharomyces host strain. The physiology, genetics, and genetic engineering of Saccharomyces are well known, and the entire genome sequence is available, making Saccharomyces a good host for the introduction of genes from Pichia stipitis (Johansson, 2001). The genes and corresponding enzymes responsible for xylose fermentation in Pichia stipitis were described in Section 2.3.1.1, and are summarized in Figure 2-12. CHAPTER 2. LITERATURE REVIEW 43 D-Xylulose-5-Phosphate ill Pentose Phosphate Pathway The xylose metabolic pathways In microorganisms. . ^ . a . . Xylose non-utilizing yeasts (Saccharomyces cerevisiae, Schlzosaccharomyces pombe, etc.) Xylose utilizing yeasts (Candida ahehatae, Pichia stipitis, Pachysolen tannophllus, etc.) Bacteria ( £ coll, Bacillus species, Streptomyces species, etc.) Figure 2-12 Xylose metabolic pathways in organisms (Ho et al, 1998) Because S. cerevisiae is naturally capable of fermenting xylulose to ethanol, early experiments focused on the transformation of Saccharomyces with XylA, the gene coding for D-xylose isomerase (XI), an enzyme found in bacteria, which can convert xylose to xylulose in a single metabolic step (Johansson, 2001). This approach was eventually abandoned, as several researchers were unable to successfully express xylose isomerase in yeast (Johansson, 2001). Several laboratories have engineered a xylose-fermenting S. cerevisiae through the expression of XYL1 or both XYLl and XYL2 (Jeffries et al, 1998). Expression of'XYL1 alone has CHAPTER 2. LITERATURE REVIEW 44 been shown to enable S. cerevisiae to produce xylitol from xylose at high yield (Jeffries et al, 1998; Hallborn et al, 1991; Hallborn et al, 1994). Co-expression of both XYLl and XYL2 has been the main focus in the genetic modification of Saccharomyces. Kotter et al. (1990) were the first to report construction of an S. cerevisiae strain expressing both XYLl and XYL2 (Jeffries et al, 1998). While expression of both XYLl and XYL2 made it possible for S. cerevisiae to grow on xylose, ethanol formation was poor, with conversion of approximately half of the xylose present in the medium into roughly equimolar amounts of xylitol and ethanol (Jeffries et al, 1998; Kotter et al, 1993). This loss of fermentation efficiency due to excess xylitol production has received attention by several laboratories. Several researchers (Jeffries et al, 1998; Kotter et al, 1993; Eliasson, 2000; Johansson, 2001) in the literature have cited the production of excess xylitol as a major challenge in the fermentation or cofermentation of xylose by recombinant strains of Saccharomyces. Also cited is a diminished consumption of xylose under anaerobic conditions (Jeffries and Himmel, 1998). It is thought that both of these factors could be related to redox imbalances arising from the specific cofactor preferences for xylose reductase and xylitol dehydrogenase (Jeffries and Himmel, 1998). Xylose reductase (from Pichia stipitis) is capable of using either NADH or NADPH for the reduction of xylose to xylitol, with NADPH strongly favoured (Johansson, 2001; Rizzi et al, 1989). Unfortunately, the next enzyme in the pathway, xylitol dehydrogenase (native to Pichia), uses NADH exclusively. Therefore, the excess consumption of NADPH can lead to elevated levels of xylitol while simultaneously blocking the assimilation of xylitol into the fermentation pathways (Jeffries and Himmel, 1998). Some work has focused on the genetic engineering of xylose reductase to use NADH preferentially, which may eliminate this imbalance (Jeffries and Himmel, 1998). Other work has focussed on the over-expression of CHAPTER 2. LITERATURE REVIEW 45 XKS1, which codes for xylulokinase (XK) in Saccharomyces, to enhance the conversion of xylulose to xylulose-5-P (Ho et al, 1998). Striking an ideal balance in a recombinant Saccharomyces strain is a difficult task to say the least, and has not yet resulted in reliable, efficient production of ethanol from xylose (Jeffries et al, 1998). The genes that are chosen for insertion, the level of desired expression, the appropriate co-factor balance, and a host of other factors, all influence the resulting xylose metabolism in a very complicated interaction. Any successes that have been realized to date in the fermentation of pentose sugars using recombinant Saccharomyces strains have been on a relatively small scale, and usually carried out under favourable "laboratory" conditions, which in most cases don't reflect typical industrial conditions. Ultimately, for any recombinant Saccharomyces strain to be considered a success from an industrial perspective, it must produce increased levels of ethanol over conventional strains on relative media, while exhibiting resistance to the toxic compounds present under industrial fermentation conditions. While increasing the yield of ethanol is important and is the ultimate goal of recombination technology, an increase in productivity at the expense of process stability would be unacceptable. Adverse effects of recombination can include decreased strain stability, diminished tolerance to the medium, differences in the response to dissolved oxygen, foam production, or the production of undesirable by-products (Stanbury et al, 1995). Any strain improvement program should use selection techniques in combination with recombination in order to select for organisms with improved stability and increased resistance to inhibitors, while simultaneously increasing ethanol productivity (Stanbury et al, 1995). CHAPTER 2. LITERATURE REVIEW 46 2.5.3 Recombinant Pentose-Fermenting Strains used at Tembec The following sections describe some of the specific details relating to the recombinant pentose-fermenting strains LNH32 and LNH-ST, which were tested during the course of these studies in the Tembec pentose pilot plant. Background is provided on the productivity of the transformed strains at the bench scale on well-defined media. For detailed information on the creation of the plasmid, and the methodology for yeast transformation, refer to: Ho et al (1998). Table 2-3 details the relevant genotypes and phenotypes of the strains used during the course of these studies, the promoters used to control the respective genes, and the vectors used in the transformed recombinant S. cerevisiae strains. Table 2-3 Yeast strains used in laboratory and pilot plant trials. Strain Description (relevant genotype or phenotype) Promoter / Enzyme Reference T2 Adapted Tembec alcohol plant strain, originally a Saccharomyces strain used in wine-making. n/a (Cameron, 1998) 1400 Saccharomyces parental strain used for plasmid transformation. A fusion product of Saccharomyces diastaticus and Saccharomyces uvarum. n/a (Ho etal., 1998) 1400 (pLNH32) Expresses XYLl, XYL2 and XKS1. Vector 2u. plasmid for all genes ADH11 XR PYK11 XDH PYK11 X K (Ho etal, 1998) 259A Saccharomyces parental strain used for chromosomal transformation n/a (Ho, 1999) 259A (LNH-ST) Expresses XYLl, XYL2 and XKS1. Vector Integration into the chromosome of the host strain 259A ? / X R ? 1 XDH ? / X K (Ho, 1999) CHAPTER 2. LITERATURE REVIEW Al 2.5.3.1 LNH32 - A Plasmid-bearing Saccharomyces Transformant The recombinant Saccharomyces yeast used in the majority of these trials, 1400(pLNH32) or LNH32, was transformed by insertion of a high copy-number plasmid containing 2 key xylose-metabolizing genes from Pichia stipitis and over-expression of a native gene to Saccharomyces into the host Saccharomyces 1400 (Ho et al, 1998). The genes inserted from Pichia stipitis on the plasmid were XYL1 and XYL2, encoding for xylose reductase (XR) and xylitol dehydrogenase (XDH). The gene inserted from Saccharomyces on the plasmid was XKS1, encoding for the enzyme xylulokinase (XK). The Pichia genes enable Saccharomyces to synthesize xylose-metabolizing enzymes, and X K was over-expressed in LNH32 because it has been demonstrated that some strains of Saccharomyces contain very low levels of X K activity (Ho et al, 1998). In addition to the three xylose-metabolizing genes, geneticin and ampicillin resistance genes were inserted that served as dominant selectable markers (Ho et al, 1998). The approach by Ho et al. (1998) was to fuse the cloned XR, XDH, and X K genes to highly effective glycolytic promoters (or expression signals), which significantly improved the recombinant yeast strain (Ho et al, 1998). This strategy resulted in a transformant that did not require the presence of xylose for induction of the genes, and also the expression of the cloned genes was not repressed by the presence of glucose in the medium (Ho et al, 1998). This allowed LNH32 to effectively coferment glucose and xylose without a lag period (Ho et al, 1998). The parental yeast strain used in the transformation, Saccharomyces 1400, is a fusion product of Saccharomyces diastaticus and Saccharomyces uvarum (Krishnan et al, 1995). It exhibits high ethanol and temperature tolerance and a high fermentation rate (Ho et al, 1998). Saccharomyces 1400 is capable of using starch (oligomers, short chained glucose molecules) to make ethanol, but there should be little or no starch present in the SSL process (Cameron, 1998). CHAPTER 2. LITERATURE REVIEW 48 While plasmids are capable of autonomous replication within cells when appropriate selective pressure is applied, they are not physically part of the cell's chromosome, and thus can be lost during cell replication (Chenier, 1998). If plasmids were lost by LNH32 during the fermentation process, metabolism of xylose would cease. In order to ensure that the plasmids are not lost during cell replication, selective pressure is applied by growing the strain in the presence of xylose at all times (Chenier, 1998). Applying this selective pressure is not a burden, because SSL feedstocks always contain some amount of xylose. Even xylose-fermenting yeasts such as Pichia stipitis require the presence of xylose for induction, and can be repressed by glucose presence (Ho et al, 1998). Theoretically, due to the high copy number inserted on the plasmid, LNH32 should be able to multiply through 5-6 generations of cells in non-selective media (i.e. media without xylose exerting selective pressure) before losing the plasmid (Chenier, 1998). Figure 2-13 illustrates how the genes were inserted on the vector pUCKmlO plasmid. Figure 2-13 Construction of high-copy-number yeast-£. coli shuttle plasmid LNH32, containing the XYL three-gene cassette KK-AR-KD. The Xhol DNA fragment containing KK-AR-KD was inserted into pUCKmlO at its Sail site (reproduced from Ho et al, 1998). CHAPTER 2. LITERATURE REVIEW 49 2.5.3.2 Bench Testing of LNH32 on Defined Media In bench scale testing with LNH32 on well-defined media, xylose was effectively fermented to ethanol as a sole substrate, and was efficiently cofermented to ethanol with glucose (Ho et al, 1998). Studies of glucose-xylose mixtures indicated that the recombinant yeast ferments the glucose simultaneously with xylose, indicating that xylose synthesis is not repressed by the presence of glucose in the medium (Ho et al, 1998). Ethanol produced by the transformant strain showed a significant increase in the fermentation efficiency above the stoichiometric maximum that could be achieved by the fermentation of glucose alone (Ho et al, 1998). Fermentation of a 90 g/L glucose and 43 g/L xylose mixture gave an ethanol concentration of 62 g/L after 48 hrs. This represents a 135% fermentation efficiency based on the glucose fraction, and a 91% fermentation efficiency based on the starting xylose and glucose concentrations (see Figure 2-14) (Ho et al, 1998). 0 20 40 60 0 20 40 60 Fermentation Time (hr) Fermentation Time (hr) Figure 2-14 Comparison of recombinant Saccharomyces strain 1400(pLNH32) (A) and parent strain 1400 (B) to coferment glucose and xylose (from Ho et al, 1998). CHAPTER 2. LITERATURE REVIEW 3U Also shown in Figure 2-14 is the low amount of xylitol production by LNH32. This is an important accomplishment, because as discussed earlier, many researchers have been challenged by the production of excess xylitol in the fermentation or cofermentation of xylose by recombinant strains of Saccharomyces. 2.5.3.3 LNH-ST - A Stable Chromosomal Saccharomyces Transformant While most work by researchers has focused on the insertion of Pichia genes using a plasmid as a vector (as in the case of LNH32), an ideal transformant would be completely stable and not require the use of selection pressure at any stage of growth or fermentation (Ho et al, 1998). The Saccharomyces parental host strain used for this chromosomal transformation into a stable strain was 259A (Ho, 1999). The recombinant Saccharomyces yeast 259A(LNH-ST) or LNH-ST, was transformed by insertion of 2 key xylose-metabolizing genes from Pichia stipitis and over-expression of a native gene to Saccharomyces (Ho, 1999). These genes were all inserted directly into the chromosome of the host Saccharomyces 259A (Ho, 1999). As in the case of LNH32, the inserted genes from Pichia stipitis were XYLl and XYL2, encoding for xylose reductase (XR) and xylitol dehydrogenase (XDH), and the gene inserted from Saccharomyces was XKS1, encoding for xylulokinase (XK) (Toon et al, 1997). The approach by Ho et al. (1999) to insert multiple copies of the three genes coding for XR, XDH, and X K directly into the chromosome results in constitutive expression of the genes. As a result this transformant did not require the presence of xylose for induction of the genes, demonstrated by over 40 successive generations in non-selective media prior to simultaneous cofermentation of a glucose-xylose mixture to ethanol (Ho, 1999). CHAPTER 2. LITERATURE REVIEW *' Successful transformation was primarily based on two factors: the expression of the genes inserted, and the utilization of xylose by the transformant (Cameron, 1998). The growth rate of the transformant and many other factors are unknowns in the equation (Cameron, 1998). Figure 2-15 illustrates how the genes were inserted on the host chromosome. Xhol KK AR KD pLNH - ST Figure 2-15 Chromosome manipulation - LNH-ST (Ho, 1999) 2.5.3.4 Bench Testing of LNH-ST on Defined Media In bench scale testing with LNH-ST on well-defined media, xylose was effectively fermented to ethanol as a sole substrate, and was efficiently cofermented to ethanol with glucose (Ho, 1999). As with LNH32, studies of glucose-xylose mixtures indicated that LNH-ST ferments glucose simultaneously with xylose, indicating no repression by the presence of glucose in the medium (Ho, 1999). Ethanol produced by LNH-ST exceeded the stoichiometric maximum achievable by fermentation of glucose alone (Ho, 1999). Fermentation of a 75 g/L glucose and 39 g/L xylose CHAPTER 2. LITERATURE REVIEW 52 mixture gave an ethanol concentration of 53 g/L after 33 hrs. This represents a 140% fermentation efficiency based on the glucose fraction, and a 91% fermentation efficiency based on the combined xylose and glucose concentrations (see Figure 2-14) (Ho, 1999). Fermentation Time (hr) Cofermentation of Glucose and Xylose by 259A(LNH-ST) after Being Cultured In Non-selective Medium (YEP+Glucose) for More than 40 Generations Figure 2-16 Ethanol production by the stable strain - LNH-ST (Ho, 1999). Because LNH-ST did not require the presence of xylose for induction of the genes, and expression of the genes was not repressed by the presence of glucose in the medium, LNH-ST effectively cofermented glucose and xylose without a lag period (Ho, 1999). CHAPTER 2. LITERATURE REVIEW 53 2.6 Laboratory Shake Flask Experiments with LNH32 on SSL While the xylose cofermentation results obtained with LNH32 by Ho et al (1998) were very impressive, these tests were conducted on a well-defined medium using laboratory grade glucose and xylose. For this transformant strain to be of interest to Tembec, it must be capable of cofermenting the xylose and hexose sugars in SSL under typical alcohol plant conditions. Testing was conducted to compare the performance of LNH32 - the plasmid-bearing transformant, Saccharomyces 1400 - the parental strain that lacks the genes for xylose utilization, and T2 - the yeast strain currently used in the full-scale alcohol plant. 2.6.1 Ethanol Production by 1400, LNH32, and the Tembec Strain Shake flask experiments were conducted using softwood SSL (grade: Alpha) (Cameron, 1995). Media were supplemented with diammonia phosphate (DAP), and pH was adjusted to 5.8. Testing was conducted with yeast concentrations of 0.2 g/L and 2.0 g/L, which produced nearly identical amounts of ethanol (Cameron, 1995). Ethanol production was monitored from 100 mL of softwood SSL, and analysis was carried out on the residual sugar in the supernatant (Cameron, 1995). Three transfers of the yeast into fresh SSL were conducted to permit acclimation (Cameron, 1995). The results of shake flask testing with LNH32, 1400, and the Tembec plant strain, are presented in Figure 2-17. CHAPTER 2. LITERATURE REVIEW 54 16 14 12 d 10 3 O 8 c ro .c LU • T2 - Tembec Plant Strain • 1400 - Transformant Parental Strain • 1400(pLNH32) - Plasmid-bearing Transformant Strain S W D # 1 S W D #2 Liquor Type and Transfer Number S W D #3 Figure 2-17 Comparison of shake flask ethanol production by Saccharomyces 1400, LNH32, and the Tembec Strain (reproduced from Cameron, 1995). The maximum amount of ethanol produced by Saccharomyces 1400 and LNH32 was approximately 20-24% higher than that produced by the yeast strain currently used in the alcohol plant (Cameron, 1995). The parental host Saccharomyces 1400 is capable of fermenting polymeric sugars including dextrins and melibiose, which may have accounted for the additional ethanol from host (Cameron, 1995). LNH32 out-performed the alcohol plant strain, despite the fact that there was little evidence of significant xylose metabolism by LNH32, and residual galactose remained for both Saccharomyces 1400 and LNH32 at the end of the fermentation (Cameron, 1995). The success of LNH32 in these laboratory experiments on SSL, justified further optimization and scale-up studies at the bench, and in continuous fermentation trials in the pentose pilot plant. CHAPTER 3. RESEARCH OBJECTIVES 55 3. RESEARCH OBJECTIVES At the Tembec sulfite pulp mill, an existing ethanol fermentation facility currently ferments the hexose sugars from spent sulfite liquor to ethanol using an industrial strain of Saccharomyces cerevisiae. This industrial yeast strain is incapable of fermenting the pentose fraction (primarily comprised of xylose, with a small amount of arabinose) present in the spent sulfite liquor. Fermentation of the xylose fraction would have a large impact on the economics of ethanol production in the facility. Tembec could increase the production of high quality industrial ethanol by approximately 30% if the xylose present in softwood and hardwood spent liquors were cofermented with hexose sugars. One goal of these studies was to consistently propagate the plasmid and chromosomal recombinant Saccharomyces strains, 1400(pLNH32) and 259A(LNH-ST) respectively, in the lab at bench scale. While these plasmid and chromosomal transformant strains have been effectively propagated using well defined media in bench scale experiments, testing was required to ensure that the xylose-fermenting capability of the strains is not hampered or lost when propagated on softwood and hardwood spent liquors which contain toxic compounds. Industrial spent sulfite liquors have low pH, high concentrations of dissolved solids, and contain undesirable compounds such as acetic acid, furfural, sulphur dioxide, and resin acids. The stability of the strains was verified by several different analyses during both lab and pilot trials under industrial conditions. The pilot fermentation plant is an automated 1/1000 scale model of the Tembec Chemical Group's Alcohol plant and operates continuously, complete with yeast recycle. While pilot scale fermentation of xylose has been done previously under sterile conditions, ethanol CHAPTER 3. RESEARCH OBJECTIVES 56 production is more often an aseptic process in an industrial setting. Therefore, another goal of these pilot studies was the successful testing of both plasmid and chromosomal transformants of robust industrial Saccharomyces strains under typical industrial (aseptic) conditions. The major purpose of these studies was to test the capabilities of xylose-fermenting recombinant Saccharomyces yeast strains on xylose-containing spent sulfite liquors at pilot scale. The measure of success in these trials was the degree of sugar utilization and ethanol production by the strains under standard industrial operating conditions. The ability of the strains to coferment xylose and hexose sugars in the spent sulfite liquor was monitored, and the fermentation efficiency of these strains was compared with baseline results using the current alcohol plant strain on various spent sulfite liquor substrates. Throughout various stages of the trials, the presence of the plasmid was verified by several different analytical methods. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 57 4. MATERIALS AND EXPERIMENTAL METHODS In this chapter, a description of all materials and experimental methods used in this research is presented. Sections 4.1 to 4.4 describe considerations for inoculum storage, as well as propagation techniques used for shake flask, bench scale 16 L lab fermenter, and pilot plant tests. Section 4.5 details the experimental methodology used in shake flask experiments with LNH-ST. Section 4.6 describes the operation of the pentose pilot plant, and Section 4.7 describes the procedures used for sampling, analysis and data acquisition. Section 4.8 details the composition of spent sulfite liquors used in pilot plant testing. Section 4.9 describes sterilization protocols in the laboratory and pilot plant, and finally, Section 4.10 details the techniques used for plasmid verification and contaminant identification. 4.1 Inoculum Storage The importance of ensuring the preservation of stock cultures for the Tembec Saccharomyces strain, T2, the plasmid bearing Saccharomyces strain, 1400(pLNH32), and the chromosomal transformed Saccharomyces cerevisiae strain, 259A(LNH-ST), can not be overstated. It is essential that all these strains not only retain their respective characteristics, but that they are viable and free of contamination (Stanbury et al, 1995). 4.1.1 Inoculum Storage in Liquid Nitrogen In order to ensure virtually complete inactivation of the metabolic activity of microorganisms, storage at very low temperatures (-150°C to -196°C) is common in liquid nitrogen refrigerators. While there can be some loss of viability during the freezing and thawing stages, there is practically no loss during the storage period, enabling storage for extended CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 58 periods (Stanbury et al, 1995). Tembec keeps many different mutant strains that have been isolated from the alcohol plant stored in liquid nitrogen. 4.1.2 Inoculum Storage on Agar Slopes Cultures were grown on Y P D and Y P X agar slopes, and stored by refrigeration at 5°C and freezing at -20°C. New cultures were periodically sub-cultured from the agar slopes (approximately every 3 months) to develop new stock mother cultures. 4.1.3 Stock Mother Cultures of Inoculum for Trials Stock mother cultures were generated regularly from isolated, pure colonies on Y P D or Y P X agar plates ( Y P X in the case of the xylose-fermenting strains). A single colony of the strain was used to form a mother culture by inoculating a shake flask containing Y P X medium (1% yeast extract, 2% peptone, 2% xylose), and then incubating at 30°C with agitation (around 220 rpm) for 24 hrs. After verification that a mother culture displayed typical growth patterns and levels of productivity on defined media, they were kept refrigerated in Y P X medium to serve as an inoculum for laboratory experimentation, and grow-ups for pilot plant trials. Figure 4-1 Mother culture of LNH32 in 50 mLs of Y P X medium. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 59 4.2 Propagation of LNH32 in Shake Flasks To begin a trial, a pure culture of the recombinant Saccharomyces strain 1400(pLNH32) was propagated in 50 mLs of YPX medium, providing the mother culture for the run. This mother culture was used to inoculate a 200 mL batch of enriched medium (see detailed recipe in Appendix A) containing trace minerals and vitamins in addition to a ratio of 2 parts xylose to 1 part hexose sugars. A portion of the hexose sugars introduced came from 30 mLs of softwood or hardwood liquor, which enabled the strain to habituate to the spent sulfite liquor. This medium was placed in a shake bath at 30°C and agitated (« 220 rpm) for a 24-hr grow-up period, and then used as the inoculum for the 16 L New Brunswick laboratory fermenter. Figure 4-2 Autoclave, centrifuge, and shaker bath. 4.3 Propagation of LNH32 in the 16 L Lab Fermenter To begin a batch, the lab fermenter started with an enriched SSL-based medium with a ratio of 1.2 parts xylose to 1 part hexose, and additional minerals and vitamins (see detailed recipe in Appendix A). Following an initial batch growth mode in the fermenter, a sterilized medium containing a mixture of xylose, softwood or hardwood SSL, and molasses at a CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 60 proportion of 1.2 parts xylose to 1 part hexose (see detailed recipe in Appendix A) was used in either a fed-batch or continuous mode of feeding to the fermenter, depending upon the growth phase of the yeast. The temperature set point was 30°C, and the pH was maintained in the range of 5.0 to 5.2 by the automatic addition of 5% aqueous ammonia. Ample agitation (« 400 rpm) and aeration (« 8 L/min) were maintained in the 16 L lab fermenter throughout the propagation. The 16 L New Brunswick fermenter from the lab is shown in Figure 4-3. This fermenter was utilized both for bench scale fermentation trials, and for yeast propagation from pure mother cultures to provide sufficient yeast volumes to inoculate the pilot plant. The fermenter has a completely controlled environment, with controls for the mixing speed, aeration (volume/volume basis), pH (on feedback control using sulfuric acid and ammonium hydroxide addition), condenser, temperature, and the ability to sterilize the fermenter in-situ. A l l feedstocks were autoclaved prior to addition, along with the tubing used for the feed addition, stoppers, pipets, etc., to ensure that the fermentation would remain sterile. Figure 4-3 16 L New Brunswick lab fermenter. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 61 4.4 Propagation of LNH32 in the Pilot Plant For pilot plant trial #1 (SWD), fermenter #1 was filled with 200 L of softwood SSL, which was then diluted to 1000 L with water, resulting in a dissolved solids concentration of approximately 4%. The pH of this medium was adjusted to approximately 5.0 with concentrated aqua ammonia from the mill, and was then sterilized by heating with steam to 96°C for 2 hrs. The sterilized medium was allowed to cool to a temperature of 35°C at which time the contents of the lab fermenter were inoculated into pilot plant fermenter #1. For pilot plant trial #2 (HWD) and pilot plant trial #5, fermenter #1 was filled with 800 L of hardwood SSL, which was then diluted to 1000 L with water, resulting in a dissolved solids concentration of approximately 16%. The rest of the procedure in the hardwood trials was identical to that in the softwood trial. During this period of batch growth, maximum aeration was provided, the pH was controlled at a set point of 5.0, and the temperature at 28°C. The growth phase of the yeast was monitored closely by checking the concentration of yeast cells, viability, and budding on a daily basis. In addition to these parameters, ethanol production and sugar consumption were also recorded. Ancillary minerals and vitamins were added to fermenter #1 in the form of lOOg of diammonia phosphate and 50g of Fermaid K. At this time the SSL was fed into the pilot plant, first filling the settling tank quickly, and then at a continuous feed rate of 1 L/min in order to fill the fermenters over night, increasing the volume of the system to approximately 3000 L. After 60 hrs of cell multiplication, the concentration of viable cells had increased to about 0.5 g/L with a viability of approximately 80%, and the system was full. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 62 4.5 Shake Flask Testing with 259A (LNH-ST) For this testing, each strain was first grown up in 50 mL of nutrient rich YPD medium (1% yeast extract, 2% peptone, 2% dextrose). The shake flasks containing the YPD medium were inoculated with 2 mLs of mother culture for each of the respective strains, which were cultured from isolated, pure colonies on YPD agar plates. Each strain was incubated at 30°C with agitation at approximately 150 rpm. After 24 hrs of growth, the cells were collected from the shake flasks by centrifugation (10,000 g, 10 min) in sterile tubes (Chenier, 1999). For each type of yeast, the recovered cells were re-suspended in 50 mLs of either S WD or HWD SSL adjusted to pH 5.2 with NH 4 OH. The cultures were incubated at 30°C under maximum agitation in the shaker bath (approximately 220 rpm) for between 16-18 hrs, after which the cells were again recovered by centrifugation (10,000 g, 10 min). This process was repeated 6 times in total for each yeast strain, with the recovered cells transferred into 50 mL of fresh SSL. The first three transfers served to acclimatize the microorganisms to the SSL. After acclimation through three successive transfers, the following three transfers in both SWD and HWD were analyzed for ethanol and residual sugar concentrations in the supernatant (Chenier, 1999). Figure 4-4 details the composition of the softwood and hardwood SSL used in these shake flask experiments. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 63 45.0 Arabinose Galactose Glucose Xylose Mannose Total Hexose F i g u r e 4-4 H a r d w o o d and softwood composi t ions for shake f lask fe rmenta t ion by T 2 -the T e m b e c plant s t ra in , 2 5 9 A - the t r ans fo rman t pa ren ta l s t r a in , and 2 5 9 A ( L N H - S T ) - the stable t rans formant s t r a in (Chen i e r , 1999). 4.6 Operation of the Pentose Pilot Plant Designed as a 1/1000-scale replica model of the Tembec Chemical Group's Alcohol production plant, the pilot plant was built to emulate all aspects of the design of the full-scale alcohol plant. With full automation enabling quick changes to process variables with relative ease, the pilot plant is able to reproduce operating conditions that closely approximate the full-scale process. Figure 4-5 is a detailed process and instrumentation drawing for the pentose pilot plant. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 64 Figure 4-5 Pilot plant process and instrumentation drawing. 4.6.1 Process Description & Control Strategy The pentose pilot plant is equipped with SSL storage tanks to provide plant substrates for trial runs, and ancillary tankage for the supply of additional media to the process if desired. A series of five fermenters form the heart of the process, with the first fermenter in the series acting as a settling tank. A surge tank located downstream from the fermenters provides a consistent feed rate to the centrifuge, which separates the yeast from the beer on a continuous basis. The yeast is recycled back to the first active fermenter, and the beer is sent to the plant evaporators for disposal. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 65 Spent sulfite liquor to be used as feed to the pilot plant was collected from the upstream sulfite mill evaporators, and stored in a 68,000 L SSL storage tank. This storage tank enabled approximately 14 days of continuous operation in the pilot plant on a constant substrate. The SSL in the storage tank was left at pH 2 to prevent any microbial activity. The temperature of the SSL in the feed reservoir was also kept at 60°C, to kill off any yeast contaminants in the SSL feed. SSL from the 68,000 L storage tank was regularly fed to a 6,000 L feed tank, which provided the feed to the pilot plant. An in-line magmeter controlled the flowrate to the plant via a small centrifugal pump, typically at a rate of approximately 2 L/min (1/1000th of the SSL flow in the full-scale operation). Each day, enough ammonium hydroxide was added to the softwood or hardwood liquor in the SSL feed tank to raise the pH from approximately 2, to between 3-4. This encouraged the precipitation of calcium oxalate in the SSL feed tank and in the settling tank. Precipitation of the calcium oxalate helped prevent fouling of the centrifuge. Maintaining a pH between 3-4 was important to prevent any fermentation or sugar degradation in the SSL reservoir (most bacteria and yeast that could be present are not metabolically active at such a low pH). Ancillary tankage (labelled "sugar tanks" on the schematic) was available for the supply of additional substrates to the process if desired. These tanks had been used in previous trials for the supplementing the carbon source with molasses. These tanks were not utilized during the course of the studies presented in this work. The first fermenter in the series acted as a settling tank in the process, to encourage additional calcium oxalate removal through precipitation and sedimentation. The presence of CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 66 calcium oxalate in the process is due to the chemistry of the upstream pulping process (which utilizes a calcium acid sulfite cooking liquor). Calcium oxalate had contributed to operating problems in previous pilot plant trials (primarily with centrifuge operation), thus, for subsequent trials the first fermenter in the series was utilized as a settling tank. After the settling tank, four fermentation tanks are connected in series. The fermenters have top-mounted mixers for agitation and aeration was controlled with manual valves in the plant, which piped the air to distribution manifolds at the bottom of the fermenters. Aeration was provided continuously in both fermenters, at a maximum during the growth of the yeast at the start of a trial, and at a minimum to induce fermentation later in a trial. The aeration rate in fermenter #2 was minimized as much as possible in order to prevent any ethanol consumption. The pH and temperature control for the system took place in fermenter #1, and the set points for these two variables were 5.0 and 28°C, respectively. Aqueous ammonia was used for the pH control in the system. The fermenters were also equipped with steam piping, controlled by manual valves, for sterilization of the plant, and multiple ports for sampling. In order to provide a retention time similar to that in the full-scale operation, a feed rate of 2 L/min was typically used, and only two fermenters were brought on-line. This feed rate, combined with the yeast recycle back to fermenter #1, provided a retention time in the fermenters of approximately 12.5 hrs. A surge tank following the fermenters served as a sampling point to determine the final ethanol and sugar concentrations in the beer from the process. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 67 Figure 4-6 Five 1000 L pilot plant fermentation tanks in series. A solid bowl (disc-plate style) centrifuge was used to separate the yeast from the beer on a continuous basis. From the centrifuge the yeast was recycled back to fermenter #1, and the beer product was either sent to the plant evaporators for disposal, or recycled back to the surge tank to maintain a liquid feed to the centrifuge. The centrifuge was controlled with a regular backwash cycle that would prevent tripping due to high vibration. O f all the equipment in the pilot plant, the centrifuge required the most maintenance and attention, as it was periodically taken out of service for cleaning. Periodic cleaning of the separator could normally be completed during an 8-hr shift, with the unit reassembled and started up as quickly as possible to avoid extended down time. These occasional interruptions had a minimal affect on the process. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 68 Figure 4-7 Pilot plant solid bowl centrifuge. 4.6.2 Plant Control System and Interface T h e s y s t e m w a s opera ted v i a P L C ( P r o g r a m m a b l e L o g i c C o n t r o l l e r ) w i t h a P C - b a s e d F i x D M A C S c o n t r o l s y s t e m . T h i s s y s t e m enab l ed access to h i s t o r i c a l da ta , a n d t r ended a l l p rocess v a r i a b l e s i n the p lan t . T h e sy s t em w a s a l so e q u i p p e d w i t h r e m o t e m o n i t o r i n g c a p a b i l i t y , a l l o w i n g d i a l - u p access a n d c o n t r o l f r o m outs ide o f the p lan t . T h e f o l l o w i n g f igures s h o w d i s p l a y sc reen cap tu res f r o m the p lant . CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 69 Figure 4-8 Pilot plant overview display screen. SSL & SUGAR TANKS HEB2g BBgBEE Figure 4-9 Spent sulfite liquor feed display screen. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 70 Figure 4-10 Fermentation tanks display screen. Figure 4-11 Yeast recycle display screen. 4.7 Sampling, Analysis and Data Acquisition During the propagation of yeast, laboratory bench trials, and trials in the pilot plant, sampling was performed regularly. In the pilot plant, daily sampling involved taking 250 mL CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 71 samples from the SSLin line, settling tank, fermenter #1, fermenter #2, and the surge tank. After centrifuging 50 mL of each of the samples (to separate the yeast from the beer), the supernatant was tested for dissolved solids, ethanol and sugar concentrations. The yeast pellet was used to determine the sludge volume, concentration of viable yeast cells, percent viability, and percentage of budding yeast. 4.7.1 Dissolved Solids Concentration Measurement Dissolved solids measurements were made on the supernatant of all samples, calculated on a w/w basis. Using a drying oven at 105°C, water was driven off from a sample of known weight, providing the weight of dissolved solids in solution. Previous testing in the full-scale and pilot plant has indicated that the dissolved solids concentration in the feed should not exceed approximately 20%, or the fermentation efficiency could be adversely affected. 4.7.2 Ethanol Analysis by Gas Chromatograph Testing of ethanol concentrations was performed using a gas chromatograph (GC). Testing for ethanol was performed on all samples, with the exception of the SSL entering the plant. This testing involved the addition of an internal standard (lOppm of 2-butanol) to a 1:10 dilution of the sample. The supernatant of a centrifuged (10,000 g, 10 min) sample was filtered (0.45 pm) into an eppendorf centrifuge tube, centrifuged (10,000 g, 10 min) for a second time, and then filtered (0.45 u.m) a final time before analysis by gas chromatography. Every time that a sample was tested on the GC, a standard fermented SSL sample with a known ethanol concentration was tested to confirm valid performance of the GC. Approximately 5% error is accepted on GC measurements. CHAPTER 4. MA TERIALS AND EXPERIMENTAL METHODS 72 Figure 4-12 Gas chromatograph for ethanol analysis. 4.7.3 Sugar Analysis by H P L C Sugar analysis was carried out using high-pressure anion exchange chromatography with pulsed amperometric detection using a Dionex unit (Model: DX-300 with a CarboPac PA1 column). A l l preparation of Dionex samples, standards, variation graphs, etc. were performed by lab assistant, Annie Chenier. A carbohydrate mother solution was prepared with a concentration of 2 g/L (containing the following sugars in equal proportion: arabinose, galactose, glucose, xylose and mannose). A fucose mother solution was also prepared to a concentration of 0.002 g/L (containing only fucose), which served as an internal standard. Using set proportions of the carbohydrate mother solution and the fucose mother solution, a more dilute carbohydrate working solution was derived (0.02 g/L) for day-to-day use as a standard to check the Dionex unit variance. An internal standard (0.5 mLs volume) was added to each sample prior to testing, in order to verify that a test result was accurate. Finally, an SSL solution with known sugar concentrations was CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 73 also regularly tested to ensure the accuracy of the Dionex unit. Approximately 5% error i accepted on Dionex measurements. is Figure 4-13 Dionex DX-300 with a CarboPac PA1 column for sugar analysis. 4.7.4 Analysis of Yeast Volume and Viability Using the pellet from the centrifuged sample, the sludge volume was measured, and then a yeast count was performed to determine the concentration of viable yeast cells, percent viability, and percentage of budding yeast. Yeast counts were carried out after washing the yeast sample with water (to improve contrast), and staining of the cells with aqueous methylene blue dye (1ml of 1% solution per 100 mL of diluted yeast tank sample). After staining, the live cells were easily distinguished from the dead cells by their appearance, colourless and bright blue, respectively. Dead cells have no electrons present to maintain the cell wall integrity, and thus are permeable to the methylene blue dye (Cameron, 1998). Regardless of the dilution used, for the yeast count to be statistically significant, between 100-200 total cells must be counted for a given sample (Cameron, 1995). CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 74 After a yeast count was complete, the % viability, concentration of viable cells, and % budding (an indicator of growth phase) were reported using the following equations. ft/ , ... (Average # Live Cells) t n n % Viability = -, * - r x 100 (Average # Total Cells (Dead + Live)) o/u AA- (Average # Budding Cells) % Budding - —. - r-^  x 100 (Average # Live Cells) ^ „ „ . . / T \ (Average* Live Cells) x D F x W x C F Cell Viability (g/L) = ^  ^ Where, V = Volume of Sample in Counting Chamber (5 squares = 2x10" 5ml), DF = Dilution Factor, W = Weight of a single yeast cell (0.4 x 10"10 g/cell) (Cameron, 1995), and CF = Conversion Factor (1000 mL/L). The following picture shows the plant yeast at lOOx magnification (although yeast counts were performed at 40x magnification). The picture shows the basic morphology of the strain, a budding yeast cell, and an oxalate crystal, which were often present in samples and could occasionally make yeast counts difficult. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 75 Figure 4-14 Yeast budding and oxalate crystals (lOOx magnification). 4.7.5 Data Acquisition and Trending During yeast propagation periods, bench scale trials, and pilot plant runs, measurements were also made of the pH and temperature of the fermentation broth. In both the lab fermenter and the pilot plant, pH and temperature were controlled (with concentrated aqua ammonia and a cooling loop, respectively), and in the pilot plant, real-time measurement enabled trending of the data. 4.8 Spent Sulfite Compositions for the Pilot Trial Runs 4.8.1 Pilot Plant Trial #1, LNH32 on SWD SSL The SSL storage tank was filled with softwood liquor twice during this trial. A combination of two softwood grades, "Best" and "Solv" (« 21.7% w/w solids), provided the feed for the first 11 days of the run. Due to a shortage of SSL to complete the trial, the storage tank was topped up with the softwood grade "Alfa", resulting in a feed of slightly different CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 76 composition (« 22.4% w/w solids) for the last 3 days of the trial. Table 4-1 summarizes the composition of the incoming SSL throughout the trial. Table 4-1 Average incoming SSL composition for Pilot Plant Trial #1. I Sugars (g/L) Days 0-11 Days 12-14 I Arabinose 1.62 1.24 | | Glucose 6.48 6.29 Xylose 7.18 6.01 Galactose 4.97 4.18 Mannose 18.85 16.73 Total Hexose 30.30 27.20 4.8.2 Pilot Plant Trial #2, LNH32 on HWD SSL The SSL storage tank was fdled at the start of this trial with liquor that consisted primarily of the hardwood grade "Film", and may have had some residual " H V " grade softwood liquor. This liquor had approximately 21.2% w/w solids, and provided the feed for the entire 17 days of this pilot run. Table 4-2 summarizes the composition of the incoming SSL. Table 4-2 Average incoming SSL composition for Pilot Plant Trial #2. Sugars (g/L) Days 0-17 || Arabinose 0.98 | Glucose 4.24 Xylose 16.03 I Galactose 2.03 | Mannose 7.69 1 Total Hexose 13.96 J CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 11 4.8.3 Pilot Plant Trial #3, LNH32 on Blended SSL The SSL storage tank was filled with liquor that consisted primarily of the softwood grade "Solv". This liquor was added to the remaining hardwood liquor from the previous run. This resulted in a blended SSL with approximately 18.2% w/w solids, and provided the feed for the 8 days following the transition from the hardwood SSL to the blended SSL. Table 4-3 summarizes the SSL composition throughout the trial. Table 4-3 Average incoming SSL composition for Pilot Plant Trial #3. Sugars (g/L) Days 0-2 (Days 18-20 of extended run) 1 Days 3-10 (Days 21-28 of extended run) Arabinose 1.02 1.61 Glucose 4.08 4.66 Xylose 13.71 7.49 Galactose 2.42 3.81 Mannose 9.25 13.00 Total Hexose 15.75 21.47 4.8.4 Pilot Plant Trial #4, LNH32 on Blended SSL The SSL storage tank was filled with liquor on a day when the alcohol plant was in transition from "TCF" grade hardwood SSL to "Solv" grade softwood SSL. This resulted in a blended SSL with approximately 23% w/w solids. The osmotic pressure imposed on the yeast cells by this high a concentration of solids can be detrimental to ethanol production. The SSL feed tank was filled daily as usual, and then the contents were diluted with water to an average concentration of 20.2%. This liquor provided the feed for 10 days following the transition from CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 78 the last trial to the fresh blended SSL. Table 4-4 summarizes the changes in feed composition during the trial. Table 4-4 Average incoming SSL composition for Pilot Plant Trial #4. Sugars (g/L) DaysO (Days 28 of extended run) Days 1-3 (Days 29-31 of extended run) Days 4-10 (Days 32-38 of extended run) Arabinose 1.61 1.47 1.28 Glucose 4.66 6.53 5.90 Xylose 7.49 8.90 8.11 Galactose 3.81 4.42 3.95 Mannose 13.00 17.30 16.19 Total Hexose 21.47 28.25 26.04 4.8.5 Pilot Plant Trial #5, LNH-ST on Hardwood SSL The SSL storage tank was filled with "TCF" grade hardwood liquor to begin this trial, and had an average dissolved solids concentration of approximately 20.2% w/w solids. This provided the feed to the plant for the first 9 days of the trial. For the final 6 days of the trial, a softwood grade of liquor, "Alfa", was used to test the fermentation efficiency of the LNH-ST transformant strain on a higher hexose concentration. The softwood liquor had an average dissolved solids concentration of approximately 22.3% w/w solids, which previous testing had shown to be too high for feed to the pilot plant (Cameron, 1998). To decrease concentration of the solids (and thus the osmotic pressure of the feed), the SSL feed tank was filled daily as usual, and then the contents were diluted with water to an average concentration of 20.3% with water. Table 4-5 summarizes the composition of the incoming SSL throughout the trial. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS Table 4-5 Average incoming SSL composition for Pilot Plant Trial #5. Sugars (g/L) Days 0-9 Days 10-15 Arabinose 1.21 1.50 Glucose 4.41 6.32 Xylose 19.46 8.01 Galactose 2.22 4.08 Mannose 9.42 16.85 Total Hexose 16.05 27.26 4.9 Sterilization Techniques in the Laboratory and Pilot Plant Procedures were in place to ensure a completely sterile environment when working in the laboratory. The following sterile techniques were used during yeast propagation, plating, fermenter inoculation, sampling, etc. • Sterilization of all media used for feeding shake flasks, or the 16 L fermenter, took place in a bench scale autoclave at 15psi. • Any sugars that were part of a medium recipe were sterilized separately in the autoclave at 15psi. When molasses was part of a medium, it would be sterilized twice prior to use in order to ensure that no spore-forming bacteria would survive the sterilization process. • Al l laboratory equipment, including: glassware, stoppers, pipets, tubing for feedstocks, agar plates, and sample bottles were sterilized before and after use in the autoclave at 15psi. • Al l work areas used for the propagation of yeast (such as counter tops, and a stainless steel fume hood) were cleaned before and after use with a concentrated ethanol solution. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 80 • Aseptic lab techniques were used for regular procedures such as transfers of cells from the mother culture to shake flasks, media additions to the fermenter, and during both streak and smear plating on agar plates for contamination testing. All work that could be, was performed in a stainless steel fume hood, with all glassware used being dipped in ethanol, and run through a Bunsen burner flame for sterilization. • The 16 L lab fermenter was fully equipped for in-situ sterilization, which was performed before and immediately after each and every use. For sterilization, the fermenter was cleaned thoroughly, fdled with water (or media ingredients if beginning a trial), and steam sterilized to a temperature of 121 °C. After being held at temperature for 30 minutes to an hour, the fermenter was cooled, and ready for operation. • The 1000 L pilot plant fermenters were also fully equipped for in-situ steam sterilization, which was performed before and after each pilot plant trial. For sterilization, the fermenters were cleaned thoroughly, fdled with water (or media ingredients if beginning a trial), and sterilized with live steam injection, which raised the temperature to 96°C. After being held at 96°C for 2 hrs, the fermenters were cooled, and ready for operation. 4.10 Plasmid Verification Test Procedures 4.10.1 The N. Ho Test for Confirmation of Ethanol Production from Xylose The recombinant Saccharomyces yeast used in the majority of these trials was transformed by insertion of a plasmid containing the 3 key xylose-metabolizing genes from Pichia stipitis. While plasmids are capable of autonomous replication within cells when appropriate selective pressure is applied, they are not physically part of the cell's chromosome, and thus can be lost during cell replication. Since the plasmid can be lost - and the morphology CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 81 of the parent and transformed strains are too similar to determine the presence of LNH32 by simple microscopic examination - the N . Ho test procedure (developed by N . Ho, summarized by Annie Chenier) was used to verify the presence of the plasmid bearing strain (Chenier, 1998). 4.10.1.1 Preparation of a Mother Culture A pure colony was transferred from a Y P X agar plate or 1 mL of a mother culture to a 300 mL Erlenmeyer flask containing 50 mLs of Y E P X (1% yeast extract, 2% peptone, 2% xylose). The flask was incubated at 30°C with maximum agitation (approximately 220 rpm). Simultaneously, a second 300 mL Erlenmeyer containing Y E P only (1% yeast extract, 2% peptone) and a third flask containing Y E P + LNH32 (1% yeast extract, 2% peptone, 1 mL mother culture) were also incubated. These Erlenmeyers served as contamination controls for solutions. The Y E P control was a test of the sterility of the ingredients, while the Y E P + LNH32 control determined whether the yeast starter was contaminated by bacteria or yeasts other than Saccharomyces (Chenier, 1998). Any yeast growth in these controls would indicate contamination of the xylose and/or Y E P solutions (Chenier, 1998). Figure 4-15 Streak plating for colony isolation & plasmid verification. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 82 After 24 hours of incubation, a visual inspection was performed to verify an increase in turbidity (indicating yeast growth) in the YEPX + LNH32 flask. If turbidity was observed, 2 mLs from the YEPX + LNH32 flask was added to a fresh 50 mLs of YEPX. This flask was incubated at 30°C with maximum agitation for 24 hours. After incubation, the yeast propagated in this flask was used as the new mother culture. At the same time as the turbidity inspection in the YEPX + LNH32 flask, 2 mLs of xylose was added to the flask containing solely YEP, and it was incubated at 30°C with maximum agitation for 24 hours (Chenier, 1998). 4.10.1.2 Plasmid Presence Verification In a 300 mL Erlenmeyer flask containing 100 mLs of fresh YEP liquid media, 4 mLs of 50% dextrose (w/w) solution was added. The flask was inoculated with 2 mLs of fresh mother culture and mixed well. After sealing with aluminum foil, the flask was incubated for 24 hours at 30°C with maximum agitation (approximately 220 rpm). After 24 hours, 20 mLs of 50% dextrose (w/w) and 10 mLs of 50% xylose (w/w) were added, bringing the final concentrations to approximately 8% dextrose and 4% xylose. Approximately 1.5 mLs of the solution was pipeted into an eppendorf tube to serve as the time zero sample. The flask was sealed with parafilm (to provide fermentation conditions) and incubated at 30°C with maximum agitation for 48 hours, with samples taken at 24 and 48 hours for ethanol analysis (Chenier, 1998). In a 300 mL Erlenmeyer flask containing 100 mLs of fresh YEP liquid media, 20 mLs of 50% dextrose (w/w) and 10 mLs of 50% xylose (w/w) solution was added. The final concentrations were then approximately 8% dextrose and 4% xylose. 2 mLs of fresh mother culture was added, and the solution was mixed well. 1.5 mLs of the solution was pipeted into an CHAPTER 4. MA TERIALS AND EXPERIMENTAL METHODS 83 eppendorf tube to serve as the time zero sample. The Erlenmeyer was sealed with parafilm, and incubated at 30°C with maximum agitation for 48 hours, taking samples at 24 and 48 hours for ethanol analysis (Chenier, 1998). 4.10.1.3 Ethanol Production by an LNH32 Mother Culture Table 4-6 and graph illustrate the ethanol production from an LNH32 mother culture, both pre-grown in Y E P D medium, and inoculated straight from a mother culture for testing. Table 4-6 Ethanol production by cofermentation of dextrose and xylose. Time (hours) Ethanol Concentration (g/L) Pre-grown in YEPD From Mother Culture 0 3.60 0.18 24 38.60 37.01 48 53.31 52.60 70 24 48 Time (hrs) Figure 4-16 Ethanol production by cofermentation of dextrose and xylose. CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 84 The anaerobic cofermentation of dextrose and xylose by LNH32 was a reliable test to verify the presence of the plasmid-bearing strain, LNH32. Theoretically, 1 gram of sugar can be fermented to 0.5lg of ethanol. The parent strain, 1400, is effectively a non-xylose-fermenting yeast, meaning that if the plasmid was lost, a maximum of approximately 40 g/L of ethanol would be expected. With the additional utilization of xylose to produce alcohol, a maximum of approximately 60 g/L of ethanol could be theoretically achieved, with approximately 53 g/L of ethanol being typical for LNH32. Any ethanol produced in excess of 40 g/L would come from the fermentation of xylose, so an ethanol concentration lower than 53 g/L but greater than 40 g/L could be interpreted as a partial loss of the plasmid in the yeast population tested (Chenier, 1998). . This method for plasmid presence verification provides a reliable indication of the state of the recombinant yeast, whether it comes from a shake flask, the 16 L fermenter or the pilot plant. 4.10.2 Differential Plate Counts to Test Growth on Xylose as the Sole Sugar In addition to the use of the N. Ho Test for regular plasmid verification, differential agar plate counts were employed, both for plasmid verification and contaminant identification. 4.10.2.1 Plasmid Verification through YPD and YPX Differential Plate Counts In order to obtain a relative measure of how many cells had maintained the plasmid, differential plate counts were conducted throughout the trials. A differential plate count was conducted with spread plates, where a yeast sample from the 16 L fermenter or pilot plant was spread (at various dilutions) on YPX, YPD, and Lysine plates. Comparing the amount of colony forming units/mL (CFU/mL) from a selective media spread plate count on YPX (where only CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 85 LNH32 thrives) to a non-selective media spread plate count on Y P D (where wild strains, the alcohol plant yeast, the parent strain 1400, and LNH32 can all thrive), provided a "plate differential". Any growth on Lysine plates is associated with contaminants, as wild (i.e. non-Saccharomyces) yeasts have the ability to use lysine as the sole source of nitrogen for growth. The following picture shows a Y P X agar spread plate, and shows how easy it is to identify plasmid bearing LNH32 cells on the plate. Figure 4-17 Spread plating on YPX to identify LNH32 ceils. The following tables provide the results from two differential plate counts, performed on yeast samples collected from the pilot plant at the beginning and end of trial #1. Table 4-7 Yeast colony count - July 30th lab fermenter sample. Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX 10'5 .5 125 2.5 x l O 7 YPX IO'5 .1 32 3.2 x l O 7 YPX 10"6 .5 27 5.4 x l O 7 YPX 10 6 .1 13 1.3 x l O 8 CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS YPD 10 s .5 100 2.0 x 107 YPD IO -5 .1 48 4.8 x 107 YPD IO"6 .5 25 5.0 x 107 YPD 10"6 .1 8 8.0 x 107 Lysine 10° .5 0 0 Lysine 10"3 .1 0 0 Lysine IO"4 .5 0 0 Lysine IO"4 .1 0 0 Table 4-8 Summary of the July 30 yeast colony count. Agar Plate Average CFU/mL Y P X 3.7 x 107 YPD 3.9 x 107 Lysine 0 *Note: plate count averages only include plates with 25-30 colonies. Table 4-9 Yeast colony count - August 17 pilot plant sample. Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL Y P X 10 s .5 160 3.2 x 107 Y P X 10 s .1 46 4.6 x 107 Y P X IO"6 .5 38 7.6 x 107 Y P X 10"6 .1 9 9.0 x 107 YPD 10 s .5 265 5.3 x 107 YPD 10 s .1 49 4.9 x 107 YPD 10"6 .5 46 9.2 x 107 YPD 10"6 .1 22 2.2 x 108 Lysine IO"4 .5 44 8.8 x 105 Lysine IO"4 .1 24 2.4 x 106 Lysine 10 s .5 13 2.6 x 106 Lysine 10"5 .1 9 9.0 x 106 CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS Table 4-10 Summary of the August 17 t h yeast colony count. Agar Plate Average CFU/mL Y P X YPD Lysine 5.1 x 107 6.5 x 107 8.8 x 10s Looking at the difference between the CFU/mL on YPX and YPD agar plates indicates that plasmid-bearing cells made up approximately 95% ([3.7 x 107/3.9 x 107] x 100) of the yeast population at the beginning of the trial, decreasing to 79% ([5.1 x 107/6.5 x 107] x 100) by the end of the trial run. This indicates that LNH32 was the dominant yeast strain in the plant for the duration of the trial (complete differential plate count data for all trials is provided in Appendix D). 4.10.2.2 Contaminant Identification and Significance As mentioned above, any growth on Lysine plates is associated with contaminants, as wild (i.e. non-Saccharomyces) yeasts have the ability to use lysine as the sole source of nitrogen for growth. Looking at the differential plate count data above, these contaminant yeast strains were non-existent at the start of the trial and constituted approximately 1 in 100 cells at the end. Therefore, of the total yeast population in the plant at the end of the trial, LNH32 was the dominant strain. While a pure culture would be ideal for a number of reasons in a fermentation process, this is virtually impossible in an industrial setting (especially in a continuous process). There are always contaminants present at full-scale and in the pilot plant. If contaminant strains comprise < 1% of the viable cell population, the culture is "pure" from an industrial perspective (Cameron, CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 88 1998). That said, in the pilot plant, a completely pure culture is required in the mother culture, shake flasks used for grow-up, lab fermenter, and at the start of a trial, for the desired culture to out compete any wild yeast strains or bacteria (Cameron, 1998). 4.10.3 Durham Tube Test to Test Gas Production with Xylose as the Sole Sugar While plasmid verification testing using the N. Ho Test and differential plate counts were accurate methods for determining ethanol production and the relative amounts of plasmid-bearing, parent, and contaminant strains, neither test provides information about which strains were responsible for the utilization of xylose. In order to provide a definitive answer as to whether or not LNH32 was solely responsible for the disappearance of xylose in the pilot plant, a relatively simple and quick test (a Durham Tube Test) involving the evolution of CO2 from xylose was used. The sample yeast for testing could be a washed yeast sample from the pilot plant (for a quick check of xylose utilization) or a pure colony from agar plates (to compare xylose use by different strains). After streak plating samples, the pure colonies that grew (each formed from a single cell) would have a different appearance for different strains (i.e. smooth vs. rough shaped colonies, and differences in colour). Microscopic examinations were used to confirm the morphology of contaminant strains, and the parent strain was differentiated from LNH32 on selective and non-selective media. Once the various strains in the pilot plant were isolated on agar media, a given strain could be tested for its ability to utilize xylose as a sole sugar for metabolism using the Durham Tube Test. A Durham tube (basically a test tube), filled with 4.5 mLs of YEPX medium (providing xylose as the sole sugar) was inoculated with the yeast from a homogeneous, isolated colony CHAPTER 4. MATERIALS AND EXPERIMENTAL METHODS 89 from an agar plate, right at the bottom of the tube. A glass cap was submerged inverted inside the tube, directly over the inoculated yeast. The tube was covered with parafilm (to create anoxic conditions), and incubated at 30°C with maximum agitation (approximately 220 rpm). A control was run simultaneously to confirm the sterility of the medium. The control consisted of the same sugar medium, however, without yeast inoculation, incubated under the same conditions as the samples containing yeast. The control was centrifuged (10,000 g, 10 min) after incubation to check for a yeast pellet, which would indicate contamination of the medium. Medium C02 Inverted Cap Yeast Figure 4-18 Durham tube test for xylose utilization. The Durham tube method is a qualitative method, where the evolution of CO2 (apparent by visual inspection of gas trapped inside the inverted glass cap) indicates the ability of the strain to metabolize xylose under anaerobic conditions. CHAPTER 5. RESULTS AND DISCUSSION 90 5. RESULTS AND DISCUSSION This section presents a thorough discussion of results and data collected during the course of this research. Results are presented in four main sections: (1) bench scale (16 L Lab fermenter) testing with LNH32, (2) pilot scale testing with LNH32, (3) shake flask testing with 259A(LNH-ST), and (4) pilot scale testing with LNH-ST. A total of 4 pilot plant runs were conducted with LNH32 during the course of this research, and a single pilot plant run utilizing LNH-ST. The results from these trials, as well as the results of laboratory testing with LNH-ST, are presented here. 5.1 Bench Scale (16 L Lab Fermenter) Testing with LNH32 In order to conduct pilot scale trials with LNH32, the strain had to be grown in sufficient quantity to inoculate the plant without losing the plasmid during the propagation. Bench scale trials were conducted to verify the conditions required to reliably produce LNH32 while retaining the plasmid. In order to ensure the retention of the plasmid, the strain was forced to grow with xylose as the principle source of sugar through each stage of propagation, from shake flask to 16 L lab fermenter, and ultimately to the pilot plant. The following sections detail bench scale testing of LNH32 in the 16 L lab fermenter. These trials involved both continuous feeding conditions for the propagation of LNH32, and batch feeding conditions to test ethanol productivity on both softwood and hardwood SSL. 5.1.1 Bench Scale Trial #1 - Propagation of LNH32 on Softwood SSL Prior to inoculating the pilot plant, a yeast strain was first grown-up in the 16 L laboratory fermenter using an enriched medium (see detailed recipe in Appendix A). Testing was CHAPTER 5. RESULTS AND DISCUSSION 91 conducted with LNH32 using batch, fed-batch, and continuous feeding of softwood SSL resulting in the growth of the yeast to a viable cell concentration of approximately 7.5 g/L (see Figure 5-1). 8.0 7.0 _i ^ O) 6.0 </> "T W c 5 0 >- Ti 0) 5 4 0 ra § 3.0 !> ° § 2.0 O 1.0 0.0 --—^Viable Cells * Sludge Volume ----3.5 2.5 CD E o 2.0 > 0) 1.5 g 1.0 w f 0.5 0.0 10 20 30 40 50 Time (hours) 60 70 80 Figure 5-1 Bench Scale Trial #1 - LNH32 growth curve when utilizing softwood spent sulfite liquor as the primary substrate in the 16 L lab fermenter. Figure 5-1 shows the LNH32 growth curve when utilizing softwood spent sulfite liquor as the primary substrate. Under the feeding conditions, LNH32 exhibited excellent growth in the fermenter virtually immediately after being introduced from shake flasks. Also shown is the correlation between the viable cell concentration of LNH32 and the percent sludge volume * 2 obtained by centrifuging a fermenter sample. This positive correlation (R = 0.9082) is beneficial, as a sludge volume test takes a fraction of the time required to perform a proper microscope examination to determine viability. Typically the grow-up period for softwood SSL is between 70-140 hrs in order to achieve a viable cell concentration of 8-14 g/L. The yeast can be inoculated into the pilot plant any time during these ranges, so long as the yeast is in exponential growth phase. CHAPTER 5. RESULTS AND DISCUSSION 92 Sludge volume measurements at different times in the propagation allow calculation of specific growth rate and doubling time for the yeast according to the relationship, N t = N 0eM t , where: N t is the amount of yeast at time t, N 0 is the initial amount of yeast, and u, is the specific growth rate. In this propagation, the sludge volume at 20 hrs was 0.47 mLs/50 mLs, and at 63 hrs was 1.55 mLs/50 mLs. With N63 = 1.55, N20 = 0.47, and t = 43 hrs, the specific growth rate for LNH32 on softwood is calculated as u, = 0.028 hr"1. Using this specific growth rate with N t = 2 and N 0 = 1, the doubling time for LNH32 on softwood is calculated as trouble = 25 hrs. As a comparison, the specific growth rate for Saccharomyces cerevisiae and Saccharomyces 1400 are approximately equal to 0.018 hr"' and 0.014 hr"1, respectively, with doubling times equal to 38 hrs and 50 hrs, respectively (Cameron, 1998). This indicates that LNH32 exhibits excellent growth on softwood SSL. In this early stage of yeast propagation, it is vital to perform rigorous testing for contamination. A completely pure culture was required in the mother culture, in the shake flasks used for grow-up, and in the lab fermenter or the desired yeast can too easily become contaminated by wild yeast strains or bacteria under the aseptic conditions in the pilot plant (especially during the early growth phase in the pilot plant when aerobic conditions were employed, and yeast concentrations were low). During propagation and trial runs, microscope examinations were performed to determine viability and to examine the morphology of the yeast cells. Figure 5-2 is a photomicrograph of the plasmid-bearing strain LNH32 grown in the 16 L lab fermenter. CHAPTER 5. RESULTS AND DISCUSSION 93 Figure 5-2 Plasmid-bearing strain LNH32 (lOOx magnification). These cells look very similar to the Tembec alcohol plant strain (see Figure 2-8). While LNH32 exhibited slightly different growth characteristics, reliable identification of LNH32 or its host strain, Saccharomyces 1400, was impossible with a simple microscope examination. This necessitated the use of other methods such as the N.Ho Test, streak plating, and Durham tube testing, for reliable strain identification (methodology for these test procedures is discussed in Section 4.10). The ethanol concentration in the lab fermenter was closely monitored to ensure that optimum conditions were maintained for cell growth (not ethanol production). With the aeration rate set at one volume of air per volume of medium per minute (approximately 6 L/min), and a low rate of continuous feeding, the production of ethanol was minimized and nearly all sugars were metabolized to create new cells. Figure 5-3 displays the ethanol production over the course of the LNH32 grow-up, as well as the frequency and mode of feeding throughout the trial. The feeding began with a batch fermentation to allow substantial initial growth of the yeast, followed by fed-batch and continuous feeding to minimize by-product formation. CHAPTER 5. RESULTS AND DISCUSSION 94 3.5 Time (hours) Figure 5-3 Bench Scale Trial #1 - Ethanol production by LNH32 grown on softwood spent sulfite liquor in the 16 L lab fermenter. During the trial, the concentrations of the various sugars were monitored to determine which were being consumed by LNH32 during growth phase. These concentrations, along with the ethanol produced, are shown in Figure 5-4. Al l hexose sugars were completely utilized, while effectively no arabinose was used at all. The utilization of xylose increased toward the end of the propagation when residual hexose sugars and ethanol were at a minimum, as shown in Figure 5-4. CHAPTER 5. RESULTS AND DISCUSSION 95 12.0 10.0 3 c o '•4-1 ro >_ •*-> c O c o o 30 40 50 Time (hours) Figure 5-4 Bench Scale Trial #1 - Sugar consumption and ethanol production by LNH32 grown on softwood spent sulfite liquor in the 16 L lab fermenter. Following the grow-up in the laboratory, a comparison of the ethanol production by the LNH32 cells grown in the 16 L lab fermenter and the LNH32 mother culture from which it was derived, was performed in a shake flask experiment. The methods used in this plasmid verification test were recommended by Dr. N.Ho (see Section 4.10). Results of the test are presented in Figure 5-5. CHAPTER 5. RESULTS AND DISCUSSION 96 70.0 _J o> 60.0 c o T> 50.0 ro c 40.0 0> o c 30.0 o o 20.0 o c ro -C 10.0 UJ 0.0 24 Time (hours) Figure 5-5 Bench Scale Trial #1 - Ethanol production by LNH32 grown on softwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived. As shown in Figure 5-5, the fermenter culture produced as much ethanol as the mother culture, exceeding the theoretical maximum from hexose alone (which is 40 g/L EtOH). This proved that xylose was fermented to ethanol, and that the plasmid was retained through the propagation. The next trial involved the batch fermentation of 6 L of concentrated softwood SSL using the yeast that had been propagated in bench scale trial # 1. 5.1.2 Bench Scale Trial #2 - Batch Fermentation of Softwood SSL by LNH32 Bench scale trial #2 was a 24-hr batch fermentation of softwood SSL to test the ethanol productivity and sugar consumption by LNH32. The yeast population that had been propagated in Bench scale trial #1 on softwood SSL was recovered by centrifugation (in 50 mL tubes, centrifuged at 10,000 g for 10 minutes), and served as the inoculum for the experiment. Enough yeast was recovered in order to start the experiment with 3-4 g/L of viable cells. Because the CHAPTER 5. RESULTS AND DISCUSSION 97 yeast had already been grown on the same softwood SSL for more than 70 hrs, it was well acclimated to the substrate. For this trial, 6 L of concentrated softwood SSL was the only substrate provided to the yeast. During the 24-hr batch trial, sugar consumption was monitored to determine which were being metabolized during ethanol production by LNH32. Sugar concentration profiles, along with the ethanol production, are shown in Figure 5-6. 35.0 30.0 Jv *T 25.0 3 c o *3 ro i_ c CD o c o o 20.0 15.0 10.0 12 Time (hours) Figure 5-6 Bench Scale Trial #2 - Sugar consumption and ethanol production by LNH32 in a batch fermentation of concentrated softwood SSL in the 16 L lab fermenter. Ethanol was produced at a very high rate initially, and tapered off as the concentration of available hexose became depleted. The maximum ethanol concentration achieved in the process was 12.6 g/L EtOH, which represents an 81% fermentation efficiency based on the CHAPTER 5. RESULTS AND DISCUSSION 98 maximum potential ethanol possible from the hexose fraction (15.5 g/L EtOH). A fermentation efficiency of approximately 80% had been achieved in similar experiments conducted with the Tembec plant strain (Cameron, 1998). With the exception of galactose, all hexose sugars were utilized somewhere between the 7 and 16-hr mark of the batch fermentation. Galactose was metabolized to completion by the end of the trial. Approximately 2.0 g/L or 25% of the xylose was metabolized in the batch fermentation. No arabinose was used during the course of this trial, indicating that LNH32 was incapable of metabolizing this sugar. 5.1.3 Bench Scale Trial #3 - Propagation of LNH32 on Hardwood SSL This trial was conducted in a similar manner to bench scale trial #1, with the exception that hardwood SSL was used in the grow-up of LNH32 in the 16 L laboratory fermenter. During the grow-up of LNH32 in the 16 L laboratory fermenter, an enriched hardwood SSL medium was used for the feed (see detailed recipe in Appendix A). Batch, fed-batch, and continuous feeding modes were used in growing the yeast to a viable cell concentration of approximately 15 g/L (as illustrated by Figure 5-7). Due to the high, and relatively constant, viability of the yeast population in the lab fermenter, the sludge volume correlated well (R2 = 0.894) with the viable concentration of yeast. CHAPTER 5. RESULTS AND DISCUSSION 99 0 20 40 60 80 100 120 140 160 180 Time (hours) Figure 5-7 Bench Scale Trial #3 - LNH32 growth curve when utilizing hardwood spent sulfite liquor as the primary substrate in the 16 L lab fermenter. As performed for the softwood SSL propagation trial, sludge volume measurements were used to calculate the specific growth rate and doubling time for LNH32 on hardwood SSL. In this propagation, the sludge volume at 93.5 hrs was 0.69 mLs/50 mLs, and at 141 hrs was 1.75 mLs/50 mLs. With N^i = 1.75, N93.5 = 0.69, and t = 47.5 hrs, the specific growth rate for LNH32 on hardwood is calculated as u. = 0.020 hr"1. Using this specific growth rate with N t = 2 and N 0 = 1, the doubling time for LNH32 on hardwood is calculated as tdoubie = 35 hrs. As a comparison, the specific growth rate for LNH32 on softwood, Saccharomyces cerevisiae and Saccharomyces 1400 are approximately equal to 0.028 hr"', 0.018 hr"1 and 0.014 hr"', respectively, with doubling times equal to 25 hrs, 38 hrs and 50 hrs, respectively (Cameron, 1998). This validated that LNH32 exhibits excellent growth on hardwood SSL. The ethanol concentration in the lab fermenter was monitored so operating conditions could be adjusted to encourage cell growth and not ethanol production. During this propagation, ethanol production was not minimized optimally. As a result, an appropriate viable cell CHAPTER 5. RESULTS AND DISCUSSION 100 concentration for inoculation was not reached until nearly 160 hrs. Better control strategies for the feeding and aeration rates were devised as a result for subsequent propagations in hardwood SSL. Figure 5-8 displays the ethanol production over the course of the LNH32 grow-up, as well as the frequency and mode of feeding throughout the trial. Continuous Feed @ 0.7ml/min _[ Feed Stopped |. 20 40 60 80 100 Time (hours) 120 140 160 180 Figure 5-8 Bench Scale Trial #3 - Ethanol production by LNH32 grown on hardwood spent sulfite liquor in the 16 L lab fermenter. During the trial, the concentrations of all sugars were monitored. These concentrations, along with the ethanol produced, are shown in Figure 5-9. Al l sugars, with the exception of xylose, were completely utilized by the end of the propagation. The hexoses in the fermenter were used very quickly; after the first 20 hrs of the grow-up, hexose was no longer detectable in the system. Xylose uptake was incomplete due to the nearly continuous presence of ethanol. This illustrates again that with the yeast in a respirative growth mode, ethanol was preferentially metabolized before the utilization of xylose would occur. Near the end of the trial (after around 120 hrs), xylose consumption improved as the ethanol concentration dropped to zero. CHAPTER 5. RESULTS AND DISCUSSION 101 30.0 Time (hours) Figure 5-9 Bench Scale Trial #3 - Sugar consumption and ethanol production by LNH32 grown on softwood spent sulfite liquor in the 16 L lab fermenter. Following the grow-up in the laboratory, a comparison of the ethanol production by the LNH32 cells grown in the 16 L lab fermenter and the LNH32 mother culture from which it was derived, was performed in a shake flask experiment. Results of the test are presented in Figure 5-10. CHAPTER 5. RESULTS AND DISCUSSION 102 70.0 Time (hours) Figure 5-10 Bench Scale Trial #3 - Ethanol production by LNH32 grown on hardwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived. As shown in Figure 5-10, the fermenter culture displayed lower ethanol production compared to the mother culture. However, the fact that the fermenter culture produced more than 40 g/L EtOH confirmed that xylose was still being fermented to ethanol, and that the plasmid was retained in some portion of the population through the propagation. The next trial involved the batch fermentation of 6 L of concentrated hardwood SSL using the yeast that had been propagated in Bench scale trial #3. CHAPTER 5. RESULTS AND DISCUSSION 103 5.1.4 Bench Scale Trial #4 - Batch Fermentation of Hardwood SSL by LNH32 This trial was conducted in a similar manner to bench scale trial #2, however, 6 L of concentrated hardwood SSL was used as the substrate for the batch fermentation by LNH32 in the 16 L laboratory fermenter. Bench scale trial #4 was a 24-hr batch fermentation of hardwood SSL to test the ethanol productivity and sugar consumption by LNH32. The yeast population that had been propagated in Bench scale trial #3 on hardwood SSL was recovered by centrifugation (in 50 mL tubes, centrifuged at 10,000 g for 10 minutes), and served as the inoculum for the experiment. Enough yeast was recovered to start the experiment with 3.55 g/L of viable cells. Because the yeast had already been grown on the same hardwood SSL for more than 160 hrs, it was well acclimated to the substrate. During the 24-hr batch trial, sugar consumption was monitored to determine which were being metabolized during ethanol production by LNH32. Sugar concentration profiles, along with the ethanol production, are shown in Figure 5-11. CHAPTER 5. RESULTS AND DISCUSSION 104 25.0 0 4 8 12 16 20 24 Time (hours) Figure 5-11 Bench Scale Trial #4 - Sugar consumption and ethanol production by LNH32 in a batch fermentation of concentrated hardwood SSL in the 16 L lab fermenter. Ethanol concentrations increased steadily for the first 12 hrs of the fermentation, and tapered off as the concentration of available hexose became depleted. The maximum ethanol concentration achieved in the process was 4.2 g/L EtOH, which represents a 43% fermentation efficiency based on the maximum potential ethanol possible from the hexose fraction (9.7 g/L EtOH). A fermentation efficiency of approximately 60% had been achieved in similar experiments conducted with the Tembec plant strain (Cameron, 1998). It is possible that the decreased fermentation efficiency was due to the presence of above average concentrations of inhibitory substances related to the grade of hardwood used (DV2). During the course of the fermentation, it was noted that the viable cell concentration decreased steadily from 3.55 g/L at CHAPTER 5. RESULTS AND DISCUSSION 105 the start of the trial to 0.5 g/L at the end. Decreasing viability with time is often observed in the full-scale plant during the fermentation of hardwood SSL (Cameron, 1998). With the exception of galactose, all hexose sugars were utilized by the 12-hr mark of the batch fermentation. Approximately 50% of the galactose was utilized by LNH32, which is typically the maximum amount of galactose that can be consumed (Cameron, 1998). Approximately 3.9 g/L or 18% of the xylose was metabolized in the batch fermentation. No arabinose was used during the course of this trial, as further proof that LNH32 was incapable of metabolizing this sugar. 5.1.5 Experimental Plan for Pilot Plant Testing with LNH32 The next stage of experimentation with LNH32 involved scale-up studies in continuous fermentation trials in the pentose pilot plant. The fermentation capacity of LNH32 utilizing softwood SSL as the sole substrate was first tested, followed by trials with hardwood SSL as the fermentation medium in pilot plant trials. 5.2 Pilot Scale Testing with LNH32 A total of 4 pilot plant runs were conducted with LNH32 during the course of this research. Pilot plant trial #1 tested the ethanol productivity and sugar consumption by LNH32 on softwood SSL. Pilot plant trial #2 was conducted in a similar manner to pilot plant trial #1, with hardwood SSL used as the substrate in place of softwood SSL. Pilot plant trial #3 and #4 were both extensions of pilot plant trial #2 to test the long-term xylose fermentation capacity of LNH32 on varied grades of softwood and hardwood SSL. CHAPTER 5. RESULTS AND DISCUSSION 106 5.2.1 Pilot Plant Trial #1 - Softwood Spent Sulfite Liquor 5.2.1.1 LNH32 Propagation in the 16 L Laboratory Fermenter Prior to inoculating the pilot plant, the recombinant Saccharomyces strain 1400(pLNH32) was grown-up in the 16 L laboratory fermenter using an enriched softwood SSL medium (see Appendix A for detailed recipe). Batch, fed-batch, and continuous feeding modes were used in growing the yeast to a viable cell concentration of approximately 14 g/L (as illustrated by Figure 5-12). The mass of viable cells increased from l g to lOOg over a 140-hr growth period in the laboratory fermenter. Time (hours) Figure 5-12 Pilot Plant Trial #1 - LNH32 growth curve on softwood SSL in the 16 L lab fermenter prior to pilot plant inoculation. With the aeration set at one volume of air per volume of medium per minute (approximately 6 L/min), and a low rate of continuous feeding, the production of ethanol was minimized and nearly all sugars were converted into new cell biomass. Figure 5-13 displays the ethanol production over the course of the grow-up, as well as the feeding schedule. CHAPTER 5. RESULTS AND DISCUSSION 107 3.5 Time (hours) Figure 5-13 Pilot Plant Trial #1 - Ethanol production by LNH32 on softwood SSL in the 16 L lab fermenter prior to pilot plant inoculation. The concentrations o f the various sugars were monitored to determine which were being consumed by L N H 3 2 during growth phase. These concentrations, along with the ethanol produced, are shown in Figure 5-14. A l l hexose sugars were ut i l ized, whi le effectively no arabinose was used at a l l . Based on a mass balance, approximately 7 5 % o f the xylose supplied was utilized by L N H 3 2 . CHAPTER 5. RESULTS AND DISCUSSION 108 18.0 16.0 14.0 12.0 c o 10.0 fo c 8.0 d> o c 6.0 o o 4.0 2.0 0.0 - - - - i £3 •Ethanol •Arabinose •Glucose Xylose •Galactose •Mannose "Total Hexose 20 40 60 80 Time (hours) 100 120 140 Figure 5-14 Pilot Plant Trial #1 - Sugar consumption and ethanol production by LNH32 on softwood SSL in the 16 L lab fermenter prior to pilot plant inoculation. Following the grow-up in the laboratory, the ethanol production from the recombinant Saccharomyces strain 1400(pLNH32) grown in the 16 L lab fermenter, and the LNH32 mother culture from which it was derived, were compared in a shake flask experiment. While it was found that the fermenter culture produced less ethanol than the mother culture, production did exceed the theoretical maximum from hexose alone (40 g/L EtOH), confirming that the plasmid was retained by at least a portion of the yeast population. CHAPTER 5. RESULTS AND DISCUSSION 109 70.0 3> 60.0 c 50.0 o ro fc 40. e o § 30. O g 20. ro UJ 10.0 0.0; A Mother Culture • L a b Fermenter 12 24 Time (hours) 36 48 Figure 5-15 Pilot Plant Trial #1 - Ethanol production by LNH32 grown on softwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived. While the fermenter culture plasmid confirmation was inconclusive, agar plating on Y P X and Y P D plates indicated that plasmid-bearing cells made up approximately 95% of the yeast population (see Table 5-1). These results, combined with the fact that 75% of the xylose supplied was metabolized by LNH32, supported the assessment that the plasmid was retained by some portion of the yeast population, thus, this batch of yeast was subsequently inoculated into the pilot plant. 5.2.1.2 Sugar Utilization and Fermentation Efficiency in Pilot Plant Trial #1 Once the pilot plant was inoculated, the concentration of viable cells increased from approximately 0.04 g/L to 0.36 g/L in a period of 36 hrs. The plant was typically operated at a feed rate of 2 L/min, and the viable concentration of yeast constantly increased over the course CHAPTER 5. RESULTS AND DISCUSSION 110 of the trial (see Figure 5-16). By the end of the softwood run, the concentration of viable cells was approximately 3.2 g/L in both fermenter #1 and #2. 3 c o ra i_ c a> o c o o -*-» (A ro Oi > ro > 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 • • / A • A • Viable, Fermenter #1 A Viable, Fermenter #2 i 0.0 2.0 4.0 6.0 8.0 10.0 Time (days) 12.0 14.0 16.0 Figure 5-16 Pilot Plant Trial #1 - Viable cell concentration of LNH32 in fermenter #1 and #2 on softwood SSL. The relative viability of the yeast over the course of the trial is presented in Figure 5-17. Viability steadily decreased from approximately 85% to approximately 45% at the end of the trial. Viability in the plant is an important parameter to monitor, as a strong correlation was found between viability and fermentation efficiency in baseline trials (St. Onge and Cameron, 1997). These results were consistent with the results from all baseline trials to date, in that the viability of the yeast population decreases over time (Cameron, 1998). CHAPTER 5. RESULTS AND DISCUSSION 100 80 o > & 60 re (A re a> >-40 20 0 • • • • A • • 4 • • Viable, Fermenter #1 A Viable, Fermenter #2 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Time (days) Figure 5-17 Pilot Plant Trial #1 - A decrease in the viability of L N H 3 2 in fermenter #1 and #2 over time on softwood SSL . The decreasing viability with time was attributed to the increasing yeast volume in the system over time. As the amount of sludge to be separated increases, so do the losses in the beer from the centrifuge. The cells that are lost are usually the smaller, freshly budded, and highly viable cells (St. Onge and Cameron, 1997). The cells that are retained in the system are larger and older, resulting in an overall decrease in the % viability of the population (St. Onge and Cameron, 1997). Determining the ability of this strain to metabolize xylose was the primary focus of this trial. Figure 5-18 presents the utilization of xylose over time. Note that on day 12 of the trial, the softwood SSL supply was replenished, resulting in a decreased xylose concentration in the feed (see Section 4.8 for details on the average influent sugar concentrations throughout the trial). CHAPTER 5. RESULTS AND DISCUSSION 112 8.0 10.0 12.0 Time (days) 14.0 16.0 Figure 5-18 Pilot Plant Trial #1 - Utilization of xylose by LNH32 on softwood SSL. Once the system had reached steady-state (around day 6 of the trial), the uptake of xylose was fairly constant. A decrease to the retention time (12.5 hrs down to 6.2 hrs) occurred on day 10, and resulted in a drop from approximately 28% average usage, to 17% average usage at the end of the trial. These results indicate that an increased retention time could possibility enhance the xylose metabolism in the system, possibly due to increased expression of the cloned genes. A n overall average of approximately 21% of the incoming xylose was metabolized in the system at steady-state. The majority of the xylose metabolism (approximately 60%) occurred in fermenter #1. Table 5-1 summarizes the results of differential plate counts conducted at the beginning and end of the trial. These plate counts were used as an indicator of the proportion of plasmid-bearing yeast in the system. CHAPTER 5. RESULTS AND DISCUSSION Table 5-1 Differential plate counts at the beginning and end of the trial. Agar Plate Beginning of Trial, Average CFU/mL End of Trial, Average CFU/mL YPX 3.7 x 107 5.1 x 107 YPD 3.9 x 107 6.5 x 107 Lysine 0 8.8 x 105 These results show little difference between the amount of colony forming units (CFU) on YPD and YPX at the beginning and end of the trial. Looking at the difference between the CFU/mL on YPX and YPD agar plates indicates that plasmid-bearing cells made up approximately 95% ([3.7 x 107/3.9 x 107] x 100) of the yeast population at the beginning of the trial, decreasing to 79% ([5.1 x 107/6.5 x 107] x 100) by the end of the trial run. These results indicate that the majority of the yeast population had retained the plasmid. Wild (i.e. non-Saccharomyces) yeasts have the ability to use lysine as the sole source of nitrogen for growth. These contaminant yeast strains were non-existent at the start of the trial and constituted approximately 1 in 100 cells at the end. Therefore, of the total yeast population in the plant at the end of the trial, LNH32 was the dominant strain. Early shake flask trials and lab fermenter grow-ups (including bench scale trial #2) indicated that LNH32 was capable of fermenting hexose sugars in softwood SSL to ethanol. The uptake of hexose in this trial was very quick (approximately 90% of the utilized hexose was metabolized in fermenter #1), and once the system was at steady-state, the only hexose sugar left unmetabolized was galactose. Figure 5-19 illustrates the steady-state utilization of galactose in the pilot plant. Galactose uptake was fairly constant, averaging approximately 50%. CHAPTER 5. RESULTS AND DISCUSSION 114 c o ro 1— +"> c CD O c o o CD CO O o ro ro O 7.0 6.0 5.0 3 4 - 0 5 3.0 2.0 1.0 0.0 4.0 1 i • J • • E ffluent ^ • Influent - 2 per. Mov. Avg. 1 1 1 6.0 8.0 10.0 T i m e ( d a y s ) 12.0 14.0 16.0 Figure 5-19 Pilot Plant Trial #1 - Utilization of galactose by LNH32 on softwood SSL. The Tembec alcohol plant Saccharomyces strain is capable of utilizing much more galactose than LNH32. Figure 5-20 compares the utilization of galactose in softwood SSL by LNH32 in this trial with baseline results using the Tembec plant strain (Cameron, 1995). Incoming Effluent Incoming Effluent Saccharomyces S t r a i n 1 4 0 0 ( p L N H 3 2 ) A l c o h o l P l a n t S t r a i n Figure 5-20 Pilot Plant Trial #1 - Comparison of galactose use by LNH32 and the Alcohol Plant Strain on softwood SSL (alcohol plant data from Cameron, 1995). CHAPTER 5. RESULTS AND DISCUSSION 15 The inability of LNH32 to completely utilize galactose was unfortunate, as this essentially represents a loss of approximately 1-3 g/L of potential ethanol production from the process. These results confirm baseline trials in shake flasks, which had indicated slow and incomplete usage of galactose by LNH32 (Cameron, 1997). Throughout the course of the trial, ethanol concentrations through the system were monitored. The ethanol concentration in the settling tank was tested daily, confirming that no fermentation was occurring in either the feed tank or the settling tank. The ethanol profile in the effluent beer from the system is shown in Figure 5-21. A maximum of 14.9 g/L was achieved, followed by a steady-state concentration of approximately 12 g/L. 3 c o ro c Q) O c o o o c ro .c LU 18.0 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 0.0 A Settling Tank • Effluent Theoretical Max. (hexoses utilized) ^"Theoret ica l Max. (total hexoses) 2.0 4.0 6.0 8.0 10.0 Time (days) 12.0 14.0 16.0 Figure 5-21 Pilot Plant Trial #1 - Comparison between ethanol produced by LNH32 on softwood SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized. CHAPTER 5. RESULTS AND DISCUSSION 116 The sugar utilization through the pilot plant is shown in Figure 5-22, using data from the point of maximum ethanol production (t = day 9). This graph illustrates how quickly hexose was fermented to ethanol in the system, and shows a 30% drop in xylose from the SSLin (the SSL entering the settling tank) to fermenter #2. 35.0 SSLin Fermenter #1 Fermenter #2 Figure 5-22 Pilot Plant Trial #1 - Sugar utilization and ethanol production by LNH32 on softwood SSL (t = day 9). The fact that LNH32 metabolized 30% of the xylose in the softwood SSL, with 23% of the total xylose uptake occurring in the presence of hexose in fermenter #1, indicates that expression of the xylose-metabolizing genes by LNH32 was not affected by catabolite repression (which was observed in a bench scale batch fermentation). Approximately 50% of the galactose was utilized by LNH32, confirming earlier studies that indicated galactose fermentation would be incomplete. No arabinose was used by LNH32. On this particular day in the trial, the system CHAPTER 5. RESULTS AND DISCUSSION 117 was operating at a fermentation efficiency of 97% based on the maximum theoretical yield of ethanol from all incoming hexose sugars. Using the ethanol and sugar concentrations for this day, the fermentation efficiency of LNH32 can be calculated a different way, based on a maximum theoretical yield of 0.5 lg ethanol per gram of hexose sugar utilized in the system. On this basis, LNH32 produced 0.52g EtOH / g sugar utilized, for a fermentation efficiency of approximately 102%. This approach suggests that some ethanol must have been produced from pentose sugar metabolism, as it has been suggested in the literature that in practice, the maximum fermentation efficiency based on hexose utilized is closer to 90% (due to cell maintenance, natural by-products, etc.) (Ward, 1989). The viable yeast cell concentration in fermenter #1 was approximately 2.5 g/L on this day, so with an 800 L active fermenter volume, 2 kg of viable LNH32 cells were present in fermenter #1. Using this data, the productivity of LNH32 can be calculated. Ethanol production in the system was 1.8 kg EtOH/hr (14.9 g/L x 2 L/min), so the productivity in the system was 0.9g EtOH / g biomass / hr (1.8 kg EtOH/hr + 2 kg cells*, *all hexose fermentation occurred in fermenter #1). This represents a relatively fast rate of fermentation, which is advantageous in a commercial operation. In addition, this unit productivity may be quite conservative, as hexose fermentation had already reached completion within fermenter #1, so it is reasonable to assume that the system could have processed even more substrate. Figure 5-23 shows the fermentation efficiency based on total hexose, hexose utilized, total sugars, and sugars utilized in the system. CHAPTER 5. RESULTS AND DISCUSSION 118 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Time (Days) Figure 5-23 Pilot Plant Trial #1 - Fermentation efficiency by LNH32 on softwood SSL based on total hexose and sugar available, as well as hexose and sugar utilized. This graph shows that the fermentation efficiency steadily increased over time, and reached a maximum of approximately 100% on hexose utilized, and 97% on total incoming hexoses. The efficiencies given by "total sugars" and "sugars utilized" take into account the concentrations of arabinose and xylose. The following tables provide a sample calculation of the fermentation efficiency using the data from day 13. CHAPTER 5. RESULTS AND DISCUSSION Table 5-2 Sugars utilization on day 13 of the softwood trial. 119 Sugar Influent Concentration (g/L) Effluent Concentration (g/L) Arabinose 1.24 0 Glucose 6.29 0 Xylose 6.01 4.94 Galactose 4.18 3.01 Mannose 16.73 0 Total Hexose 27.2 3.01 Total Pentose 7.25 4.94 Total Sugars 16.73 7.95 * Ethanol production on this day was 11.82 g/L Table 5-3 Fermentation efficiency calculation. Sugars Calculation Fermentation Efficiency (%) Total Hexoses , [ 6 t h a n 0 l ] , x ! 0 0 ((hexose inf)x 0.51) 85.2 Hexoses Utilized [ethanol] ((hexose inf - hexose eff)x 0.51) 95.8 Total Sugars , [ e t h a n ° 1 ] , x l 0 0 ((total sugar inf) x 0.51) 67.3 Sugars Utilized [ 6 t h a n 0 , ] . x l O O ((total sugar inf - total sugar eff) x 0.51) 87.5 CHAPTER 5. RESULTS AND DISCUSSION 120 The difference between 100% fermentation efficiency and the fermentation efficiency achieved based on hexoses utilized represents a loss of ethanol production as a result of sugar metabolism to products other than ethanol (such as biomass, or other fermentation by-products such as glycerol). There was a loss of potential ethanol production due to galactose passing through the system unmetabolized. Based on the average concentration of unmetabolized galactose (approximately 2.5 g/L from Figure 5-19), and assuming that the theoretical maximum ethanol could be produced from this wasted sugar (.51 g/L ethanol per 1 g/L of galactose), the hypothetical ethanol concentration unrealized was 1.25 g/L. If LNH32 were capable of metabolizing all galactose to ethanol, the maximum theoretical fermentation efficiency that could have resulted was approximately 105% (based on hexoses). An interesting result was that high fermentation efficiency (« 85% on total hexoses) was maintained near the end of the trial despite a sharp decline in yeast viability (as illustrated in Figure 5-17). Previous baseline trials had established that it was common to reach a maximum fermentation efficiency early in a trial, followed by a tapering off that was strongly correlated to the declining yeast viability (St. Onge and Cameron, 1997). So while fermentation efficiencies in this trial may not have exceeded the theoretical maximum, the relatively high ethanol production that was sustained could have been due in part to the metabolism of pentose sugars. Using the fermentation efficiency based on hexoses utilized from this recombinant yeast trial, a direct comparison can be made to the fermentation efficiency achieved in a previous baseline trial that used the alcohol plant strain (see Figure 5-24). The composition of the softwood SSL used in each of these trials was comparable, and the crucial operating conditions (pH, feed rate, and temperature) were nearly identical for days 0-9 of the trial. CHAPTER 5. RESULTS AND DISCUSSION 121 LL 20 0 4- - r - - 4 - -+- -4- -A 0 4 8 12 16 20 Time (days) Figure 5-24 Pilot Plant Trial #1 - Fermentation efficiency comparison between LNH32 and the Alcohol Plant Strain from a previous baseline trial on softwood SSL. The fermentation efficiency (based on hexose utilized) of this recombinant strain was comparable to previous baseline results using the yeast strain from the alcohol plant on softwood SSL. The decreased fermentation efficiency on day ten coincided with a doubling in the feed rate to the system, which decreased the retention time from 12.5 to 6.2 hrs on average. Shortly after this adjustment, the fermentation efficiency rebounded, and nearly 100% fermentation efficiency was attained on day 14 of the trial. Also of interest was the difference in yeast viability between the two cases during the time of maximum fermentation efficiency. In the baseline trial, the yeast viability was approximately 75% whereas the viability of LNH32 was 50% when it reached a maximum. This CHAPTER 5. RESULTS AND DISCUSSION 122 emphasizes the possibility that metabolized pentose sugars may have aided in achieving high fermentation efficiency near the end of the trial. 5.2.1.3 Experimental Plan for Pilot Plant Trial #2 with LNH32 The next pilot plant trial attempted to determine the fermentation efficiency and sugar utilization of LNH32 with hardwood spent sulfite liquor acting as the sole substrate. 5.2.2 Pilot Plant Trial #2 - Hardwood Spent Sulfite Liquor This trial was conducted in a similar manner to pilot plant trial #1, with the exception that hardwood SSL was used as the substrate. The utilization of xylose (which is in greater abundance in a hardwood SSL) was monitored, and the fermentation efficiency of LNH32 was compared to the efficiency achieved by the alcohol plant strain in a previous testing. 5.2.2.1 LNH32 Propagation in the 16 L Laboratory Fermenter Prior to inoculating the pilot plant, the recombinant Saccharomyces strain 1400(pLNH32) was grown-up in the 16 L laboratory fermenter using an enriched hardwood SSL medium (see Appendix A for detailed recipe). Batch, fed-batch, and continuous feeding modes were used in growing the yeast to a viable cell concentration of approximately 15 g/L (as illustrated by Figure 5-25). The mass of viable cells increased from lg to approximately 140g over a 110-hr growth period in the laboratory fermenter. CHAPTER 5. RESULTS AND DISCUSSION 123 Time (hours) Figure 5-25 Pilot Plant Trial #2 - LNH32 growth curve on hardwood SSL in the 16 L lab fermenter prior to pilot plant inoculation. The ethanol production during the propagation in the lab fermenter is displayed in Figure 5-26. The feeding process had been refined since Bench scale trial #3, which saw high levels of ethanol, decreasing the amount of sugars converted into new cell biomass. During this yeast propagation, conditions were optimized for cell growth, with the production of ethanol minimized quickly with a low rate of continuous feeding. Aeration was set at 1 vvm (one volume of air per volume of medium per minute, or approximately 6 L/min). CHAPTER 5. RESULTS AND DISCUSSION 124 UJ 0.0' Continuous Feed @ 0.6ml/min 20 40 60 Time (hours) 80 100 120 Figure 5-26 Pilot Plant Trial #2 - Ethanol production by L N H 3 2 on hardwood SSL in the 16 L lab fermenter prior to pilot plant inoculation. The concentrations of the various sugars were monitored to determine which were being consumed by LNH32 during growth phase. These concentrations, along with the ethanol produced, are shown in Figure 5-27. "Ethanol •Arabinose "Glucose Xylose •Galactose •Mannose •Total Hexose 40 60 80 Time (hours) 100 120 Figure 5-27 Pilot Plant Trial #2 - Sugar consumption and ethanol production by LNH32 on hardwood SSL in the 16 L lab fermenter prior to pilot plant inoculation. CHAPTER 5. RESULTS AND DISCUSSION 125 A l l sugars were completely utilized by the end of the propagation. The hexoses in the fermenter were used very quickly, and after the first 18 hrs of the grow-up, they were no longer detectable in the system. While effectively no arabinose was used early in the grow-up, after the concentration of xylose dropped below 2 g/L, it was also utilized quickly and completely. A l l of the xylose supplied was utilized by LNH32 by the end of the propagation. Following the grow-up in the laboratory, the ethanol production from the LNH32 cells grown in the 16 L lab fermenter, and the LNH32 mother culture from which it was derived, were compared in a shake flask experiment. The results of this comparison are presented in Figure 5-28. Despite the sub-par production by both the fermenter culture and the mother culture, they both produced more ethanol than the theoretical maximum from hexose alone (40 g/L EtOH). 3 c o "+•> (0 +•> c CD o c o o o c ro UJ 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 - -• Mother Culture ^ ^ ~ L a b Fermenter 12 24 Time (hours) 36 48 Figure 5-28 Pilot Plant Trial #2 - Ethanol production by L N H 3 2 grown on hardwood SSL in the 16 L lab fermenter, and the mother culture from which it was derived. CHAPTER 5. RESULTS AND DISCUSSION 126 While the fermenter culture plasmid confirmation produced inconclusive results, agar plating on Y P X and Y P D plates indicated that plasmid-bearing cells made up approximately 85% of the yeast population (see Table D-12). These results, combined with the fact that 100% of the xylose supplied was metabolized by LNH32, supported the assessment that the plasmid was retained, thus, this yeast population was subsequently inoculated into the pilot plant. 5.2.2.2 Sugar Utilization and Fermentation Efficiency in Pilot Plant Trial #2 Upon inoculating the pilot plant, a distinctly different yeast strain was noticed in the plant. Dilution plating and sugar utilization tests identified this strain as a wild yeast strain, capable of using hexose sugars for growth, but with no appreciable production of ethanol. In order to eliminate this contaminant, the aeration was minimized, and a constant feed rate of 2 L/min was provided to the plant. These conditions were maintained until 7.5 days into the trial, at which time the contaminant strain was nearly completely eliminated (determined by microscopic examination and agar plating), and the aeration was increased in both fermenters to induce some growth of LNH32. Table 5-4 shows the results of the differential plate count at the beginning, and end of the trial. Table 5-4 Differential plate counts at the beginning and end of the trial. Agar Plate Beginning of Trial, Average CFU/mL End of Trial, Average CFU/mL Y P X 5.8 x 108 6.15 x 107 Y P D 6.8 x 108 7.02 x 107 Lysine 0 5.43 x 10s CHAPTER 5. RESULTS AND DISCUSSION 127 Agar plating on Y P D and Y P X indicated that plasmid-bearing cells made up the majority of the LNH32 yeast population both early and late in the trial at approximately 85% and 88% of the yeast population, respectively. Wild yeast strains were non-existent at the start of the trial (before the contaminant got into the system in an antifoam product) and constituted a maximum of approximately 1 in 100 cells at the end of the trial period. Therefore, LNH32 was the dominant strain in the plant. The concentration of viable cells increased from approximately 0.09 g/L to 0.19 g/L in a period of 36 hrs, and after 60 hrs of cell multiplication, the concentration of viable cells had increased to approximately 0.45 g/L. The response of the yeast population to the changes made early in the trial, and over the course of the trial, is shown in Figure 5-29. 18.0 Time (days) Figure 5-29 Pilot Plant T r i a l #2 - Viable cell concentration of L N H 3 2 in fermenter #1 and #2 on hardwood SSL. CHAPTER 5. RESULTS AND DISCUSSION 128 Figure 5-29 shows that the viable concentration of yeast didn't substantially increase until 8.5 days into the trial. Within 3 days of the normal operating conditions being applied, a maximum concentration of viable cells was achieved at approximately 1.7 g/L in both fermenter #1 and #2. The viable concentration of the population declined for the rest of the trial, to a minimum viable cell concentration of 0.5 g/L. Over the course of the trial, the viability of the yeast population decreased steadily from approximately 75% to 26% (see Figure 5-30). It should be noted that attempts were made to disregard the wild yeast strain (which was distinguishable in the methylene blue dye test) in these viability calculations. The diminishing viability of the population over time was consistent with the findings in previous hardwood baseline trials conducted with the Tembec plant strain (Cameron, 1998). 10 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 Time (days) Figure 5-30 Pilot Plant Trial #2 - A decrease in the viability of LNH32 in fermenter #1 and #2 over time on hardwood SSL. The steady-state usage of xylose in the pilot plant is presented in Figure 5-31. Xylose uptake was at a minimum at the same time as the viable yeast population hit a low point of CHAPTER 5. RESULTS AND DISCUSSION 129 0.1 g/L (see Figure 5-29). Following this, the uptake of xylose continued to increase to the end of the trial. 25.0 a> (A o >> 5.0 0.0 • Effluent • Influent - 2 per. Mov. Avg. 4.0 6.0 8.0 10.0 12.0 Time (days) 14.0 16.0 18.0 Figure 5-31 Pilot Plant Trial #2 - Utilization of xylose by LNH32 on hardwood SSL. Due to the low metabolism of xylose early in the trial, the average usage over the course of the trial was only 16.2%. However, it was encouraging that 35% of the incoming xylose was being metabolized at the end of the trial (14% more than the average usage displayed by LNH32 on softwood SSL). The uptake of galactose, displayed in Figure 5-32, also correlated with the increasing yeast population. After day 9, galactose uptake increased until almost 50% was being used at the end on the trial. The average usage over the entire trial was 27.6%. CHAPTER 5. RESULTS AND DISCUSSION 130 3.0 _ J c 2.5 o 2 2.0 * J c 0) o c 1.5 o o CD 1.0 w O O 0.5 co co o 0.0 - i—4-• • Effluent A Influent - 2 per. Mov. Avg. • • 4.0 6.0 8.0 10.0 12.0 Time (days) 14.0 16.0 18.0 Figure 5-32 Pilot Plant Trial #2 - Utilization of galactose by LNH32 on hardwood SSL. For the most part, the only hexose sugar that escapes fermentation by LNH32 was galactose. Figure 5-33 shows the uptake of hexose in the trial, and illustrates the consequences of diminished viable cell concentrations. This graph shows that when the cell concentration was at a minimum, hexoses other than galactose can also escape fermentation. 20.0 18.0 co •*-> o 5 16.0 o il 14.0 1 12.0 o 8 10.0 o 0) w o X CD 8.0 6.0 4.0 2.0 0.0 - I -Yeast Viability @0.1 g/L m^m , • Effluent • Influent - 2 per. Mov. Avg. w - • 4.0 6.0 8.0 10.0 12.0 14.0 Time (days) 16.0 18.0 Figure 5-33 Pilot Plant Trial #2 - Utilization of hexose by LNH32 on hardwood SSL. CHAPTER 5. RESULTS AND DISCUSSION 131 The concentration of ethanol throughout the pilot plant was monitored daily and is presented in Figure 5-34. No ethanol was detected in the settling tank, indicating that no contaminant strains were producing ethanol prior to the pilot plant. Ethanol production had started to increase when the contaminant was recognized and actions were taken to ki l l it off. This resulted in a regression in yeast production, and a subsequent drop in ethanol production. Once normal operational conditions were re-established, ethanol production rebounded and a maximum concentration of 5.39 g/L was achieved. While this level of production may seem limited, the theoretical maximum was only 7.12 g/L EtOH (based on total hexoses) due to the low concentration of incoming hexoses. UJ Yeast Viability @ 0.1 g/L A Settling Tank • Effluent ^^Theoretical Max. (hexoses utilized) ^"Theoretical Max. (total hexoses) Figure 5-34 Pilot Plant Trial #2 - Comparison between ethanol produced by LNH32 on hardwood SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized. CHAPTER 5. RESULTS AND DISCUSSION 132 At the time that the maximum ethanol production was achieved, the cell viability in the system was only 55%. Cell maintenance and endogenous respiration due to this low cell viability could account for some amount of sugar demand in the process that ultimately takes away from the production of ethanol (Cameron, 1998). It was thought that some of the cells that were considered "dead" (as determined by the methylene blue test) might have still been somewhat active, but just extremely inefficient (Cameron, 1998). If a cell were dying, little or no ethanol would be produced, while sugar and air would still be used as the cell attempted to stave off death (Cameron, 1998). This endogenous metabolism may be responsible in part for the resulting decrease in the fermentation efficiency in systems with low viability (Cameron, 1998). The sugar utilization at the point of maximum ethanol concentration (t = day 11) is provided in Figure 5-35. Virtually all sugars used, were utilized in fermenter #1. SSLin Fermenter #1 Fermenter #2 Figure 5-35 Pilot Plant Trial #2 - Sugar utilization and ethanol production by LNH32 on hardwood SSL (t = day 11). CHAPTER 5. RESULTS AND DISCUSSION 133 On this day, LNH32 metabolized only 13% of the xylose in the hardwood SSL. Virtually no galactose or arabinose was utilized. The fermentation efficiency over the course of the run in the pilot plant is shown in Figure 5-36. These were calculated based on the theoretical maximum yield of ethanol from the sugar source indicated. 1 4 0 0 .0 2 . 0 4 . 0 6.0 8.0 10 .0 12 .0 1 4 . 0 16 .0 18 .0 Time (Days) Figure 5-36 Pilot Plant Trial #2 - Fermentation efficiency by LNH32 on hardwood SSL based on total hexose and sugar available, as well as hexose and sugar utilized. When the viable cell concentration was at a minimum, the fermentation efficiency based on total hexose decreased dramatically as mannose and glucose escaped the system unfermented (there was too much substrate for the concentration of cells present). A maximum of approximately 83% fermentation efficiency was reached based on hexoses utilized, and 76% fermentation efficiency was attained based on total incoming hexoses. The maximum fermentation efficiency came relatively late in the trial (around day 11). Since fermentation CHAPTER 5. RESULTS AND DISCUSSION 134 efficiency is known to decrease over time, it is likely that even higher efficiencies could have been attained earlier in the trial, had there not been problems with contamination. Despite maximum sugar metabolism late in the trial for each type of sugar fermented, the fermentation efficiency continued to decrease to a minimum of approximately 40%. This represents a huge loss of fermentation efficiency as a result of unwanted by-product formation and/or a deadload on the system caused by the endogenous respiration of dying yeast cells (which are hampered by toxicity during extended runs of hardwood SSL) (Cameron, 1998). A comparison between the fermentation efficiency achieved by LNH32 and the alcohol plant strain in a previous baseline trial is shown in Figure 5-37. 100 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 T i m e ( d a y s ) Figure 5-37 Pilot Plant Trial #2 - Fermentation efficiency comparison between LNH32 and the Alcohol Plant Strain from a previous baseline trial on hardwood SSL. The trends in the fermentation efficiency of LNH32 and the yeast strain from the alcohol plant were very similar. Late in the trial, the trend lines for the two cases were CHAPTER 5. RESULTS AND DISCUSSION 135 decreasing at almost the same rate. The maximum fermentation efficiency by LNH32 slightly exceeded production by the plant strain, with 76% fermentation efficiency achieved in the trial. Late in the trial, the fermentation efficiency of the recombinant strain was approximately 5% higher. This was despite the fact that LNH32 only utilized 27.6% of the incoming galactose, had a lower cell viability, and took 9 days to reach a cell concentration at which ethanol could be produced effectively. It should be noted that direct comparisons made between different trial runs on hardwood SSL are not necessarily valid, as certain grades of hardwood may be significantly more toxic to yeast than others (Cameron, 1998). 5.2.2.3 Experimental Plan for Pilot Plant Trial #3 with LNH32 Due to problems early in this trial, the consumption of xylose was delayed, and never reached steady-state (it was still increasing at the end). In order to determine the level of xylose uptake that could be achieved, this trial was extended using the same yeast population. This trial extension was to explore the hypothesis that over time, the yeast population would utilize more xylose, as more of the cells gained a selective advantage through greater expression of the genes on the plasmid. 5.2.3 Pilot Plant Trial #3 - An Extension of Pilot Plant Trial #2 on Blended SSL Table 5-5 summarizes the sugar utilization by LNH32 in pilot plant trial #1 and #2, conducted on softwood and hardwood SSL, respectively. CHAPTER 5. RESULTS AND DISCUSSION Table 5-5 Average sugar uptake in Pilot Plant Trial #1 & #2. Sugar PP Trial #1 - Softwood PP Trial #2 - Hardwood Mannose 100% 90% (25-100%) Glucose 100% 94% (73-100%) Galactose 50% (26-78%) 28% (3-61%) Xylose 21% (1-43%) 16% (3-33%) It should be noted that these are the average sugar uptakes by LNH32 through the entirety of the trials and there were complications in the propagation of the yeast in the hardwood trial. These complications in pilot plant trial #2 resulted in decreased sugar metabolism early in the trial; however, the galactose and xylose concentrations were steadily decreasing at the end of the hardwood run. Since the steady-state uptake rate was not determined for xylose, this hardwood run was extended to investigate long-term trends in the fermentation efficiency of LNH32. 5.2.3.1 Sugar Utilization and Fermentation Efficiency in Pilot Plant Trial #3 In the previous trial, a maximum cell concentration of 1.7 g/L was reached in both fermenter #1 and #2. The viable cell concentration of the population declined for the rest of that trial, to a minimum of 0.5 g/L. Softwood spent sulfite liquor was used at the beginning of this trial to help revive the yeast population. The response of the yeast population to the blended SSL substrate is displayed in Figure 5-38. CHAPTER 5. RESULTS AND DISCUSSION 137 ro o.o4- - + - - | - - 4 - I —|— —i 17.0 19.0 21.0 23.0 25.0 27.0 Time (Days) Figure 5-38 Pilot Plant Trial #3 - An increase in the viable cell concentration of LNH32 in fermenter #1 and #2 on blended SSL. This shows that the concentration of viable yeast in the system rebounded, and steadily increased to a maximum of approximately 1.9 g/L. The population of viable cells averaged 1.6 g/L for the duration of the trial. Results from earlier baseline trials had determined that a minimum viable cell concentration of 1.0 g/L was required in order to achieve substantial ethanol production (Cameron, 1997). The steady-state usage of xylose in the pilot plant is presented in Figure 5-39. The uptake of xylose increased steadily throughout this trial, and eventually all incoming xylose was metabolized. CHAPTER 5. RESULTS AND DISCUSSION 138 18.0 17.0 19.0 21.0 23.0 25.0 27.0 Time (days) Figure 5-39 Pilot Plant Trial #3 - Utilization of xylose by LNH32 on blended SSL. When the xylose was completely used in the system, the average xylose concentration of the incoming blended SSL was 7.5 g/L. The same influent xylose concentration was present in pilot plant trial #1, which involved fermentation by LNH32 on pure softwood SSL. However, in the softwood trial, LNH32 only utilized 21% of the xylose. Differential plate counts determined that, while wild yeast were present in this trial, they were not present in significant numbers, and would therefore not account for the xylose use. By this point in the trial, the population of yeast was much older than the population was in pilot plant trial #1, implying that the proportion of plasmid-bearing yeast, or the expression of the genes on the plasmids, may have increased over time. A substantial drop in pH through the pilot plant indicated that organic acids were likely formed as an unwanted by-product. The production of organic acids can indicate the presence of contaminants, such as the wild yeast that was present early in pilot plant trial #2, however, it was CHAPTER 5. RESULTS AND DISCUSSION 139 confirmed that the concentration of wild yeast throughout this trial was negligible (see Appendix D for detailed differential plate count data). In shake flasks and Durham tube tests, the wild yeast strain produced no ethanol from xylose, and grew at a very slow rate. This enforces the likelihood that LNH32 was the only strain responsible for xylose metabolism in the system. Galactose was again the most substantial sugar to escape fermentation. Uptake had been increasing in the previous trial, but levelled off, and the average usage over the whole trial was only 41.6% (see Figure 5-40). Aside from a couple of excursions, the only hexose sugar that escapes fermentation by LNH32 was galactose. 5.0 3 C 4.0 O re c 3.0 a> o 8 2.0 a> in o tj 1.0 re ro O 0.0 X I * * • Effluent A Influent - 2 per. Mov. Avg. / • 17.0 19.0 21.0 23.0 Time (days) 25.0 27.0 Figure 5-40 Pilot Plant Trial #3 - Utilization of galactose by LNH32 on blended SSL. At the end of the previous trial (on hardwood SSL), the production of ethanol was poor. Only 2 g/L of ethanol was being produced during the transition from the hardwood liquor used last trial, to the blended SSL of this trial. The ethanol profile for the trial is shown in Figure 5-41. CHAPTER 5. RESULTS AND DISCUSSION 140 12.0 17.0 19.0 21.0 23.0 25.0 27.0 Time (days) Figure 5-41 Pilot Plant Trial #3 - Comparison between ethanol produced by LNH32 on blended SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized. The production of ethanol rebounded with the new SSL substrate. A maximum concentration of 7.78 g/L was achieved, and correlated well with the maximum viable cell concentration. At the time of maximum ethanol production, the percentage of viable cells in the population was only 20%. No ethanol was detected in the settling tank, indicating that all fermentation was occurring in the pilot plant. The sugar utilization at the point of maximum ethanol concentration (t = day 24) is provided in Figure 5-42. As was the case in the previous trials, the majority of the fermentation occurred quickly, within the 6 hours spent in fermenter #1. Again, aeration rates were minimized in fermenter #2 to avoid any ethanol consumption in the absence of available sugars. CHAPTER 5. RESULTS AND DISCUSSION 141 25.0 SSLin Fermenter #1 Fermenter #2 Figure 5-42 Pilot Plant Trial #3 - Sugar utilization and ethanol production by LNH32 on blended SSL (t = day 24). On this day, LNH32 metabolized 100% of the xylose in the hardwood SSL. As per usual, mannose and glucose were preferentially metabolized, and were completely fermented in the first stage fermenter. Approximately 50% of the galactose was utilized. Unlike results in previous testing, on this day arabinose was utilized to completion, which correlated well with the complete uptake of xylose. The fermentation efficiency during the transition from hardwood SSL to a blend is shown in Figure 5-43. This was calculated based on the theoretical maximum yield of ethanol from the sugar source indicated. CHAPTER 5. RESULTS AND DISCUSSION 142 17.0 19.0 21.0 23.0 Time (Days) 25.0 27.0 Figure 5-43 Pilot Plant Trial #3 - Fermentation efficiency by LNH32 on blended SSL based on total hexose and sugar available, as well as hexose and sugar utilized. A maximum of approximately 77% was reached based on hexoses utilized, and 71% fermentation efficiency was attained based on the total incoming hexoses. These maximums came quickly once the new substrate was supplied, and the concentration of viable cells had rebounded. Typically fermentation efficiency decreases over time, and has been correlated to the diminishing viability (St. Onge and Cameron, 1997). In this case, the viability was very low to begin with, and only decreased by an additional 4% after the maximum fermentation efficiency was attained, so the fermentation efficiency remained steady for the balance of the trial. 5.2.3.2 Experimental Plan for Pilot Plant Trial #4 with LNH32 The blended SSL used in pilot plant trial #3 consisted primarily of softwood liquor, which helped the yeast population to rebound and increase the fermentation efficiency. In order CHAPTER 5. RESULTS AND DISCUSSION 143 to determine i f the complete metabolism of xylose could be sustained over the long term, this trial was extended a second time, again using the same yeast population. The hypothesis that the LNH32 plasmid-bearing cells had gained a selective advantage in the pilot plant system was tested on hardwood spent sulfite liquor, so as to provide the yeast population plenty of xylose. Finally, an increase in LNH32 viability late in the trial (i.e. when a proportionately higher fraction of plasmid-bearing yeast were present) at a time when xylose utilization was maximized could result in substantially higher fermentation efficiencies. Thus, a yeast purge was also planned for late in pilot plant trial #4. 5.2.4 Pilot Plant Trial #4 - An Extension of Pilot Plant Trial #2 on Blended SSL Table 5-6 summarizes the sugar utilization by LNH32 in pilot plant trial #1 with softwood SSL, and the extensions of pilot plant trial #2, which involved both hardwood and blended substrates. Table 5-6 Average sugar uptake in Pilot Plant Trial #1, #2, and #3. Sugar PP Trial #1 - PP Trial #2 - PP Trial #3 - PP Trial #2 & #3 -SWD HWD Blended Average Mannose 100% 90% (25-100%) 94% (69-100%) 92% (25-100%) Glucose 100% 94% (73-100%) 99% (90-100%) 97% (73-100%) Galactose 50% (26-78%) 28% (3-61%) 42% (27-46%) 35% (3-61%) Xylose 21% (1-43%) 16% (3-33%) 62% (36-100%) 34% (3-100%) CHAPTER 5. RESULTS AND DISCUSSION 144 Despite complications early in the extended trial, galactose and xylose concentrations in the effluent had decreased steadily. By extending the hardwood trial (pilot plant trial #2), the uptake rate of pentose eventually surpassed that achieved in the softwood run. Pilot plant trial #4 was a second extension of pilot plant trial #2, again maintaining the same yeast population. 5.2.4.1 Sugar Utilization and Fermentation Efficiency in Pilot Plant Trial #4 Entering this trial, the cell concentration was 1.6 g/L and had been slowly decreasing near the end of pilot plant trial #3. The response of the yeast population to the new blended SSL substrate is displayed in Figure 5-44. 2.0 -j 1.8 -_J D) 1.6 -C o 1.4 -re C 1.2 0) o c 1.0 o O *-> 0.8 -i/> ra >• 0.6 -Si 0.4 ra > 0.2 -0.0 -27.0 • Fermenter #1 • Fermenter #2 • Ethanol Produced 29.0 31.0 33.0 Time (Days) 35.0 37.0 10.0 3 c o ro L-c Qi O c o o re 2 0 £ 0.0 Figure 5-44 Pilot Plant Trial #4 - Viable cell concentration of LNH32 in fermenter #1 and #2 on blended SSL. Also shown are the ethanol profde, and the point at which a yeast purge was attempted. CHAPTER 5. RESULTS AND DISCUSSION 145 Figure 5-44 shows that the concentration of viable yeast in the system steadily decreased once subjected to the new blend of SSL. This was expected, as the blend was predominately hardwood SSL. The ethanol profde throughout the trial is also shown in Figure 5-44, and a viable cell concentration of 1.1 g/L corresponded with the maximum ethanol production (9.72 g/L EtOH). Following this, the yeast concentration dropped steadily, as did ethanol production. These results are further evidence that at least 1.0 g/L of viable cells are necessary to achieve substantial ethanol production. Also displayed on Figure 5-44 is the point at which a yeast purge was attempted. For a period of 5 hours, all of the yeast recovered by the centrifuge was eliminated from the system. Calculations indicated that this would decrease the yeast population by roughly half. Figure 5-44 shows that the viable yeast population doubled in the days following the purge. In addition, the recovery of the yeast to a concentration greater than 1 g/L correlated well with an increase in ethanol production. The viability of the yeast population entering this trial was approximately 22%. Viability continued to decrease through this trial as it had in the entire extended run to this point. A minimum of approximately 12% occurred around day 35 of this trial (see Figure 5-45). CHAPTER 5. RESULTS AND DISCUSSION 146 O) c '•6 •a 3 CO T J c ra > » + J n ra > tn ra a) > • 50 45 40 35 30 25 20 15 10 5 • Budding, Fermenter #1 • Budding, Fermenter #2 X Viability, Fermenter #1 • Viability, Fermenter #2 A Ethanol Produced Yeast Purged 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 3 c o c <D O c o o o c ra LU 27.0 29.0 31.0 33.0 Time (Days) 35.0 37.0 Figure 5-45 Pilot Plant Trial #4 - Viability and budding of LNH32 in fermenter #1 and #2 on blended SSL. Also shown are the ethanol profile, and the point at which a yeast purge was attempted. F o l l o w i n g the yeas t purge , c e l l v i a b i l i t y d o u b l e d a n d the n u m b e r o f b u d d i n g c e l l s i n the s y s t e m inc r ea sed f r o m 8%, to a h i g h o f 30% at one p o i n t . T h i s i nc rease i n v i a b i l i t y c a n be a t t r ibu ted to an i n c r e a s e d g r o w t h rate and , to a lesser extent , i m p r o v e d sepa ra t ion b y the cen t r i fuge ( S t . O n g e a n d C a m e r o n , 1997) . T h e increased g r o w t h rate a n d b u d d i n g o c c u r s because o f a decrease i n c o m p e t i t i o n b e t w e e n the yeast c e l l s ( C a m e r o n , 1 9 9 8 ) . W i t h the feed rate cons tant , a yeas t pu rge m e a n s that m o r e substrate is r e a d i l y a v a i l a b l e f o r the c e l l s r e m a i n i n g i n the s y s t e m ( C a m e r o n , 1998) . B y r e d u c i n g the a m o u n t o f s l u d g e i n the s y s t e m , the l o a d o n the cen t r i fuge is dec reased , a n d it is p r e s u m e d that this increases the s e p a r a t i o n e f f i c i e n c y ( S t . O n g e a n d C a m e r o n , 1997) . F o r the m o s t part, the m e t a b o l i s m o f x y l o s e w a s cons tan t , a n d n e a r l y c o m p l e t e t h r o u g h o u t th is t r i a l . X y l o s e c o n s u m p t i o n in the p i l o t p l an t is p resen ted i n F i g u r e 5 -46 . CHAPTER 5. RESULTS AND DISCUSSION 147 10.0 9.0 5" 8.0 o> c" 7.0 o 2 6.0 S 5.0 u O 4.0 o S 3.0 o X 2 0 1.0 0.0 j I i -^Effluent A Influent - 2 per. Mov. Avg. — » » 27.0 29.0 31.0 33.0 Time (days) 35.0 37.0 Figure 5-46 Pilot Plant Trial #4 - Utilization of xylose by LNH32 on blended SSL. Overall, 92.5% of the incoming xylose was metabolized by LNH32 in this trial. Differential plate counts on Y P D and Y P X confirmed that the plasmid was present at the end of the trial, and lysine plating established that an insignificant amount of wild yeast was present in the plant. These results clearly indicate that LNH32 cells were capable of long-term sustained xylose uptake. At the end of the previous trial, the production of ethanol was constant at approximately 7.2 g/L. In this trial, although the concentration of incoming hexose increased from the last trial (21.5 g/L to 26 g/L), only 8 g/L of ethanol was produced on average (see Figure 5-47). A maximum ethanol concentration of 9.7 g/L was achieved, and again, no ethanol was found at any point in the settling tank. CHAPTER 5. RESULTS AND DISCUSSION 148 3 c o re i_ *-> c o o c o o o c re LU 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 27.0 •^Set t l ing Tank • Fermenter #2 ^^Theoretical Max. (hexoses utilized) ""Theoret ica l Max. (total hexoses) 29.0 31.0 33.0 Time (days) 35.0 37.0 Figure 5-47 Pilot Plant Trial #4 - Comparison between ethanol produced by LNH32 on blended SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized. The sugar utilization at the point of maximum ethanol concentration (t = day 32) is presented in Figure 5-48. SSLin Fermenter #1 Fermenter #2 Figure 5-48 Pilot Plant Trial #4 - Sugar utilization and ethanol production by LNH32 on blended SSL (t = day 32). CHAPTER 5. RESULTS AND DISCUSSION 149 On this day, LNH32 metabolized 90% of the xylose in the hardwood SSL. Mannose and glucose were preferentially metabolized, and both galactose and arabinose utilization were negligible. The fermentation efficiency of the system was calculated based on the theoretical maximum yield of ethanol, and is shown in Figure 5-49. The fermentation efficiency based on hexose utilized, total hexose, sugar utilized, and total sugar are all indicated. > o c a> o E LU c o '•5 ro 4-1 C Qi E l _ Qi LL 100 90 80 70 60 50 40 30 20 10 0 27.0 A Hexose Utilized • Total Hexose • Sugars Utilized • Total Sugars 29.0 31.0 33.0 Time (Days) 35.0 37.0 Figure 5-49 Pilot Plant Trial #4 - Fermentation efficiency by L N H 3 2 on blended S S L based on total hexose and sugar available, as well as hexose and sugar utilized. A maximum of approximately 81% was reached based on hexoses utilized, and 73% fermentation efficiency was attained based on the total incoming hexoses. These results were comparable to the fermentation efficiency achieved in the first hardwood trial. Similar to that trial, the fermentation efficiency decreased quickly after reaching a maximum when utilizing CHAPTER 5. RESULTS AND DISCUSSION 1 M hardwood as the sole substrate. In this case though, the decline was reversed, with fermentation efficiency rebounding in conjunction with a purge of the yeast population. 5.2.5 Pilot Plant Trials #2 to #4 - Summary of Long Term Fermentation Trends Table 5-7 summarizes the sugar utilization by LNH32 in pilot plant trial #1 with softwood SSL, and the extended pilot plant trials #2 to #4, which involved both hardwood and blended substrates at different stages. Table 5-7 Average sugar uptake in Pilot Plant Trials #1, #2, #3, and #4. Sugar PP Trial #1 -SWD PP Trial #2 -HWD PP Trial #3 -Blended PP Trial #4 -Blended PP Trial #2-#4 -Summary Mannose 100% 90% (25-100%) 94% (69-100%) 100% 95% (25-100%) Glucose 100% 94% (73-100%) 99% (90-100%) 100% 98% (73-100%) Galactose 50% (26-78%) 28% (3-61%) 42% (27-46%) 41% (32-62%) 37% (3-62%) Xylose 21% (1-43%) 16% (3-33%) 62% (36-100%) 93% (83-100%) 44% (3-100%) These are the average sugar uptakes by LNH32 through the entirety of the trials mentioned. The three preceding trials (pilot plant trials #2 to #4) were all presented separately, and the discussion focused on trends discovered during the specific trial period. This section summarizes the entire extended trial, and examines long-term trends in the data. CHAPTER 5. RESULTS AND DISCUSSION 151 5.2.5.1 Sugar Utilization and Fermentation Efficiency in Pilot Plant Trials #2 to #4 A moving average (2 days) of the long-term viable yeast concentration is plotted in Figure 5-50. Time (Days) Figure 5-50 Pilot Plant Trials #2-4 - Viable cell concentration of LNH32 in fermenter #1 and #2 on varied grades of SSL. Also shown is where a yeast purge was attempted. Figure 5-50 shows that the concentration of viable yeast in the system increased to approximately 1.7 g/L, followed by a decrease to a minimum of 0.5 g/L (associated with fermentation of a predominately hardwood SSL). Revival of the yeast population at approximately day 16 was due to a change of substrate, as softwood SSL was introduced to the system. As a result, viable yeast in the system steadily increased to a maximum of approximately 2.0 g/L. Again, at approximately day 24, hardwood SSL made up the majority of the substrate, and a corresponding decrease in yeast viability to 0.7 g/L was observed. At this point in the trial, CHAPTER 5. RESULTS AND DISCUSSION 152 roughly half of the yeast population was removed with a yeast purge, which revitalized the viable yeast population. The viable cell concentration doubled in the days following the purge, and correlated well with an increase in ethanol production. The viability of the yeast population decreased steadily over the course of the extended trial. Upon inoculating the plant, cell viability was as high as 75%, and a minimum viability of approximately 12% was observed around day 35 of the trial (see Figure 5-51). 90 80 cn 70 c '•5 T3 60 3 OQ T3 50 C CO > 40 abi 30 > *-CA 20 coCD >- 10 0 X Hardwood Liquor Grade: T C F • Budding, Fermenter #1 A Budding, Fermenter #2 X Viability, Fermenter #1 • Viability, Fermenter #2 Blended Liquor Grades: T C F (HW) & Solv (SW) Blended Liquor Grades: T C F & Solv 10 15 20 25 Time (days) 40 Figure 5-51 Pilot Plant Trials #2-4 - Viability and budding of LNH32 in fermenter #1 and #2 on varied grades of SSL. Also shown is where a yeast purge was attempted. The only time that a substantial increase was noticed in the relative viability followed the yeast purge as indicated on the graph. This resulted in a doubling of both cell viability and budding in the system. It was thought that this increase in viability can be attributed to 2 main CHAPTER 5. RESULTS AND DISCUSSION 153 factors, an increased growth rate and, to a lesser extent, improved separation by the centrifuge (St. Onge and Cameron, 1997). While a strong correlation was found between viability and fermentation efficiency in baseline trials with the alcohol plant strain (St. Onge and Cameron, 1997), this correlation was not generally present with LNH32. St. Onge and Cameron (1997) concluded that yeast losses in the beer from the centrifuge increased as the volume of sludge to be separated increased through a trial. They believed that the cells that were lost were usually the smaller, freshly budded, and highly viable cells, resulting in the retention of larger and older cells in the system (St. Onge and Cameron, 1997). With the alcohol plant strain, this clearly resulted in a decrease in both the relative viability of the population and the fermentation efficiency with time. It is possible that the same correlation was not observed with LNH32, because in the case of this plasmid-bearing strain, the retention of larger and older cells in the system was in fact, beneficial. It is conceivable that these cells are exactly the cells that are desirable in the system, as they have gained a selective advantage with higher rates of expression of the genes on the plasmid than that attained by new yeast. If this was the case, the inefficiency of the separator was actually inadvertently selecting for cells with higher rates of expression. It may have been expected that the yeast purge from the system would have adversely affected ethanol production as the purge specifically targets the removal of these cells larger and older cells in the system to support the growth of new, smaller, freshly budded, and highly viable cells. In practice, the yeast purge resulted in a positive response in the system with respect to ethanol production. It is possible that this was a short-term effect that was the result of temporarily improved hexose utilization by new cells in the system. It is still unknown what the long term effects of these yeast purges will be on plasmid retention and expression in LNH32. CHAPTER 5. RESULTS AND DISCUSSION 154 The xylose concentration profile over the course of the pilot plant trial is presented in Figure 5-52. The uptake of xylose increased steadily throughout this trial, and eventually, virtually all xylose was metabolized. 25.0 Figure 5-52 Pilot Plant Trials #2-4 - Long-term utilization of xylose by LNH32 on varied grades of SSL. Overall, an average of 44% of the incoming xylose was metabolized by LNH32 during this trial. Xylose uptake continually increased until 100% was being used around day 25 of the trial. At the beginning, throughout, and at the end of the trial there was very little difference between the plate colony counts on Y P D and Y P X confirming that the plasmid was present at the whole time. It has been established that LNH32 cells can achieve sustained xylose metabolism. CHAPTER 5. RESULTS AND DISCUSSION 155 Galactose uptake was varied slightly over the course of the trial. A n average uptake of 37% was achieved (see Figure 5-53). 6.0 5.0 3 | 4.0 n o c o O <D (fl O O ro ro O 3.0 2.0 1.0 0.0 Blended Liquor Grades: T C F (HW) & Solv (SW) Blended Liquor Grades: T C F & Solv • Effluent • Influent - 2 per. Mov. Avg. 0 5 10 15 20 25 30 35 40 Time (days) Figure 5-53 Pilot Plant Trials #2-4 - Long-term utilization of galactose by LNH32 on varied grades of SSL. The results of this extended trial, and pilot plant trial #1 with softwood confirm that LNH32 was incapable of fermenting galactose to completion. This represents an average loss of 1 -3 g/L of potential ethanol from the process. Production of ethanol was dependant on the substrate being used, and thus varied considerably throughout the trial. Early in the trial when a hardwood SSL was being used, production did not exceed 5.4 g/L of ethanol. A maximum of 9.7 g/L was achieved in the trial when the feed was a blended SSL (see Figure 5-54). CHAPTER 5. RESULTS AND DISCUSSION 156 Time (days) Figure 5-54 Pilot Plant Trials #2-4 - Comparison between ethanol produced by LNH32 on varied grades of SSL and the theoretical maximum yield of ethanol based on total hexose available and hexose utilized. The sugar utilization at the point of maximum ethanol concentration (t = 32 days) is presented in Figure 5-55. Figure 5-55 Pilot Plant Trials #2-4 - Sugar utilization and ethanol production by LNH32 on varied grades of SSL. CHAPTER 5. RESULTS AND DISCUSSION 157 The results of the sugar utilization in the plant by LNH32 support the conclusion reached by Ho et al (1998) that xylose is co-metabolized with hexose sugars by LNH32. Since the expression of the genes for utilization of xylose are inducible (and at least partially repressible by the presence of glucose), a stronger expression of the xylose genes in the yeast of fermenter #2 than in yeast from fermenter #1 would be expected. This may have accounted for the ability of LNH32 to complete the metabolism of xylose in fermenter #2. Based on the maximum sugars consumed in the trial, and assuming 10% conversion into biomass and unwanted by-products, a maximum of 14.7 g/L ethanol should have theoretically been produced (as opposed to the 9.7 g/L observed), which would have represented a 123% fermentation efficiency based on hexose utilized (clearly indicating the fermentation of pentose to ethanol). A possible explanation for lower than expected ethanol production could be related to the fast rate of fermentation in fermenter #1, combined with the low level aeration in fermenter #2. Due to a fast rate of fermentation, the majority of the hexose was removed in fermenter #1, leaving little substrate to support the yeast in fermenter #2. In addition, to support proper yeast growth and the oxidation of xylitol to xylulose in the metabolism of xylose, a small amount of air was added in fermenter #2. It is possible that some of the ethanol produced in fermenter #2 was subsequently consumed by the yeast (which can occur under aerobic conditions), resulting in what appeared to be lower fermentation efficiency. This could be controlled with stricter aeration control than that present in the pentose pilot plant. Another factor that may have contributed to lower than expected ethanol production could have been the formation of other by-products such as organic acids and glycerol (Cameron, 1998). Organic acids can be produced when sugars enter the Krebs cycle, but leak out CHAPTER 5. RESULTS AND DISCUSSION 158 early (Cameron, 1998). While no specific measurements were made for organic acids, a slight drop in the pH was observed through the system occasionally, which can indicate their presence. Glycerol formation is another possible reason for low ethanol production (Cameron, 1998). The presence of sulfite during fermentation (which is present in SSL) can stimulate conversion of dihydroxyacetone phosphate (an important intermediate in glycolysis) to glycerol (Ingledew, 1995). Glycerol production reoxidizes accumulated N A D H , and ultimately detracts from ethanol yields (Yang etal, 1982). The fermentation efficiency of the system was calculated based on the theoretical maximum yield of ethanol, and is shown in Figure 5-56. The fermentation efficiency based on hexose utilized, total hexose, sugar utilized, and total sugar are all indicated. 10 15 20 25 Time (days) 30 35 40 Figure 5-56 Pilot Plant Trials #2-4 - Fermentation efficiency by LNH32 on varied grades of SSL based on total hexose and sugar available, as well as hexose and sugar utilized. CHAPTER 5. RESULTS AND DISCUSSION 159 A maximum of approximately 81% was reached late in the trial based on hexoses utilized, and 73% fermentation efficiency was attained based on the total incoming hexoses. In the middle of the trial, a second peak in fermentation efficiency was observed that was associated with a more favourable SSL blended substrate. Shortly after the peak was realized, there was a steady drop in efficiency associated with a predominately hardwood SSL feedstock. Late in the trial the efficiency was improved a second time with a yeast purge, which increased the viability, and subsequently the fermentation efficiency, of the yeast population. Based on experience in both the pilot plant and full-scale alcohol plant, it was expected that it would be difficult if not impossible to grow the yeast on pure hardwood SSL for an extended period of time (Cameron, 1998). In the full-scale plant, the concentration of viable yeast decreases rapidly on transitions from softwood to hardwood SSL. In fact, during production runs lasting more than 5 days on hardwood SSL in the full-scale plant, there have been occasions when the yeast population was lost entirely, with no alcohol production at all (Cameron, 1998). The fact that LNH32 was capable of maintaining reasonable levels of fermentation efficiency throughout the course of this extended trial was a success. Even though LNH32 exhibited fair stability for a plasmid-bearing strain, and it was demonstrated that the plasmid could be retained throughout extended trials with appropriate selective pressure on the system, an ideal transformant would be completely stable and not require the use of selection pressure at any stage of growth or fermentation (Ho et al, 1998). In reality, the use of a plasmid-bearing strain in a continuous fermentation process is not likely to be commercially viable, as there would always be concerns about strain stability. CHAPTER 5. RESULTS AND DISCUSSION 160 5.2.5.2 Experimental Plan for Trials with LNH-ST A new strain, 259A(LNH-ST) or LNH-ST, was provided by Dr. Nancy Ho of Purdue University for research purposes. LNH-ST is a stable strain, meaning that the genes required for xylose metabolism have been inserted into the chromosome of the host (Saccharomyces 259A), and should therefore be present through successive generations, even on non-selective media. LNH-ST will be compared to other strains in shake flask experiments with successive transfers on both softwood and hardwood media. 5.3 Shake Flask Testing with 259A(LNH-ST) While earlier xylose cofermentation results obtained with LNH-ST in bench scale testing were very impressive, with 91% fermentation efficiency achieved on a mixed xylose/glucose solution (see Section 2.5.3.4 for details), these tests were conducted on a well-defined medium using laboratory grade glucose and xylose (Ho, 1999). For any transformant strain to be considered for full-scale use at Tembec, it must be capable of cofermenting the xylose and hexose sugars in SSL under typical alcohol plant conditions. A bench scale comparison of T2 - the Tembec alcohol plant strain, 259A - the parental strain which lacks the genes for xylose utilization, and LNH-ST - the stable transformant strain, was conducted (by assistant Annie Chenier) to test relative levels of ethanol production between the strains on both softwood and hardwood substrates. 5.3.1 Softwood and Hardwood Trials with 259A(LNH-ST) Shake flask experiments were conducted in softwood and hardwood SSL media that were adjusted to a pH of approximately 5.2 (see detailed methodology in Section 4.5) (Chenier, 1999). Ethanol production was monitored, and sugar measurements were made on the CHAPTER 5. RESUL TS AND DISCUSSION • o I supernatant to characterize residual sugars from the fermentation. The results of shake flask testing with L N H - S T , 259A, and the Tembec plant strain, are presented here. Figure 5-57 compares the production of ethanol by T2 - the Tembec plant strain, 259A - the parental strain, and L N H - S T - the stable transformant strain. 20.0 SWD#1 SWD #2 SWD #3 HWD #1 HWD #2 HWD #3 Liquor Type and Transfer Number Figure 5-57 Comparison of shake flask ethanol production by T2, 259A, and LNH-ST (Chenier, 1999). L N H - S T produced approximately 2 g/L more ethanol than the Tembec plant strain (Chenier, 1999). Even the parent strain, 259A, produced more ethanol than the plant strain, by as much as 2 g/L on softwood. The maximum amount of ethanol produced by L N H - S T was approximately 20% higher than the parental strain, and the yeast strain currently used in the alcohol plant, in both hardwood and softwood spent sulfite liquors (Chenier, 1999). Figure 5-58 compares the utilization of arabinose, galactose and xylose by T2 - the Tembec plant strain, 259A - the parental strain, and L N H - S T - the stable transformant strain in CHAPTER 5. RESULTS AND DISCUSSION 162 this shake flask trial. The data presented are residual sugars from transfer #3 (when the yeast was most acclimated, and steady-state results were obtained) for softwood and hardwood SSL tests. 120% ^ 100% § 80% ra £ 60% ro CO 40% 20% 0% • • Arabinose • Galactose • Xylose 1 1 1 T2 259A LNH-ST Softwood SSL T2 259A LNH-ST Hardwood SSL Figure 5-58 Comparison of sugar utilization by T2, 259A, and LNH-ST (reproduced from Chenier, 1999). Both the parental strain 259A and LNH-ST utilized significantly more xylose and galactose than the Tembec plant strain (Chenier, 1999). L N H - S T utilized 24% and 35% of the xylose from hardwood and softwood SSL, respectively. Also, L N H - S T metabolized 31% and 55% of the galactose in the hardwood and softwood SSL, respectively. This represents two to seven times the uptake of xylose, and three to nine times the uptake of galactose by L N H - S T over the Tembec plant strain. Mannose and glucose were completely utilized by all strains tested (not shown), and differences in arabinose uptake were insignificant (Chenier, 1999). The ability of the stable transformant to take up galactose was a very positive result. In previous trials with LNH32, it was determined that the plasmid-bearing strain was incapable of utilizing galactose, and as a result, it is possible that any benefits associated with the uptake of CHAPTER 5. RESULTS AND DISCUSSION 163 xylose were counteracted. The ability of LNH-ST to utilize virtually all available hexose in combination with xylose fermentation was promising. 5.3.2 Experimental Plan for Pilot Plant Trial #5 with LNH-ST The performance of LNH-ST in these laboratory experiments on SSL, justified further optimization and scale-up studies at the bench, and in continuous fermentation trials in the pentose pilot plant. Unfortunately, due to limited time in the study period, detailed bench scale testing was not performed as part of these studies. Instead, in the interest of time, a batch of LNH-ST inoculum was propagated in the 16 L lab fermenter in order to seed the pilot plant for a single run. The goal was to operate the pilot plant first using a hardwood substrate, followed by softwood SSL to assess the alcohol production capabilities of LNH-ST. It was hoped that ethanol would be produced in excess of the maximum that could be achieved by the fermentation of the hexose fraction alone. 5.4 Pilot Scale Testing with LNH-ST A single pilot plant run was conducted with LNH-ST. Pilot plant trial #5 tested the ethanol productivity and sugar consumption by LNH-ST on both hardwood and softwood SSL in a two week trial. Finally, the fermentation efficiency of LNH-ST was compared to the efficiency achieved by the alcohol plant strain in previous research. 5.4.1 Pilot Plant Trial #5 - LNH-ST on Hardwood & Softwood SSL 5.4.1.1 LNH-ST Propagation in the 16 L Laboratory Fermenter Prior to inoculating the pilot plant, the recombinant Saccharomyces strain 259A(LNH-ST) was grown-up in the 16 L laboratory fermenter using an enriched hardwood CHAPTER 5. RESUL TS AND DISCUSSION 164 SSL medium (see Appendix A for detailed recipe). Batch, fed-batch, and continuous feeding modes were used in growing the yeast to a viable cell concentration of approximately 9.5 g/L in 80 hrs. Propagation curves from the inoculum grow up (not shown) were nearly identical to those from pilot plant trial #2 which are detailed in Section 5.2.2.1. 5.4.1.2 Sugar Utilization and Fermentation Efficiency in Pilot Plant Trial #5 Upon inoculating the pilot plant, the viable concentration of yeast increased steadily over the course of the trial. By the end of the trial, the concentration of viable cells was approximately 3.0 g/L in both fermenter #1 and #2 (not shown). Also consistent with the findings in pilot plant trial #2 on hardwood, the relative viability of the yeast population decreased steadily from approximately 80% to 30% (not shown). The sugar utilization at the point of maximum ethanol concentration (t = day 15) is presented in Figure 5-59. 35.0 SSLin Fermenter #1 Fermenter #2 Figure 5-59 Pilot Plant Trial #5 - Sugar utilization and ethanol production by LNH-ST on softwood SSL (t = day 15). CHAPTER 5. RESULTS AND DISCUSSION 165 Figure 5-59 shows that LNH-ST metabolized 25% of the xylose in the softwood SSL (which was the substrate on this particular day), with all xylose uptake occurring in the presence of hexose in fermenter #1. This indicates that the expression of the xylose-metabolizing genes by LNH-ST was not affected by catabolite repression, and also illustrates a very fast rate of fermentation by LNH-ST (All sugars were metabolized in less than 6 hrs in fermenter #1). Approximately 65% of the galactose was utilized by LNH-ST, and utilization was still increasing at the end of the trial, confirming that LNH-ST was capable of fermenting galactose. A small amount of arabinose was used by LNH-ST, and correlated well with the removal of all other fermentable sugars. On this particular day in the trial, the system was operating at a fermentation efficiency of 74% based on the theoretical ethanol yield from all hexose sugars, and 78% based on the theoretical ethanol yield from hexose sugars utilized. The average sugar uptake achieved by LNH-ST in Pilot Plant Trial #5 are presented in Table 5-8. Table 5-8 Average sugar uptake in Pilot Plant Trial #5. Sugar 1 Average Use Over Trial Mannose 93.4% (52-100%) Glucose 99.2% (94-100%) Galactose 53.1% (23-92%) Xylose 7.0% (5-37%) CHAPTER 5. RESULTS AND DISCUSSION 166 A comparison between the fermentation efficiency achieved by LNH-ST and the alcohol plant strain in a previous baseline trial on hardwood SSL is shown in Figure 5 -60 . 120 0 2 4 6 8 10 12 14 16 18 Time (days) Figure 5-60 Pilot Plant Trial #5 - Fermentation efficiency comparison between LNH-ST and the Alcohol Plant Strain from a previous baseline trial on hardwood SSL. The fermentation efficiency of this recombinant strain was somewhat lower than expected, as previous results with the yeast strain from the alcohol plant achieved much more consistent fermentation efficiency. The data for the alcohol plant strain is from a baseline run using hardwood SSL. On day 13 the fermentation efficiency of the stable strain begins to increase in response to a change in substrate to softwood SSL. This explains its perceived improvement over the alcohol plant strain in the late stages of the trial. The erratic fermentation efficiency displayed by the stable recombinant strain in this trial may have been a response to the inadvertent use of a particularly toxic grade of hardwood SSL at the start of the trial (grade: TCF). The presence of toxicity, let alone its effect on fermentation, is very difficult to CHAPTER 5. RESULTS AND DISCUSSION 167 quantify. Nothing in the grow-up of the inoculum for this trial, nor the shake flask experiments conducted with LNH-ST on hardwood SSL indicated that it would have this degree of difficulty on a hardwood substrate. Also, the quick recovery of the system and smoothing out of the fermentation efficiency in response to the change in substrate to softwood SSL offer further support to the theory that the low efficiency achieved in this trial may have been an anomaly. Further testing is certainly required to test the full potential of this stable recombinant strain. CHAPTER 6. CONCLUSIONS 168 6. CONCLUSIONS This research investigated the performance of xylose-fermenting recombinant Saccharomyces yeasts (transformants of robust industrial yeast strains) on xylose-rich spent sulfite liquor from the Tembec sulfite pulp mill in Temiscaming, Quebec. Shake flask, 16 L bench scale, and pilot scale fermentation trials were conducted to determine the ethanol production capabilities of LNH32 (a plasmid-bearing transformant) and LNH-ST (a stable chromosomal transformant). The pilot scale fermentation trials with LNH32 were the primary focus of this research, and were carried out under typical industrial (aseptic) conditions. A total of 4 pilot plant trial runs were conducted with LNH32 during the course of this research, and a single pilot plant run utilizing LNH-ST. The results from these trials, as well as the results of laboratory testing with LNH-ST, have been presented. The major findings of these studies are presented in the following sections. 6.1 Bench & Pilot Scale Trials with LNH32 6.1.1 Plasmid Retention and Strain Growth The stability of the plasmid-bearing strain, LNH32, was verified during both bench and pilot scale trials under industrial conditions. The strain successfully maintained its xylose-fermenting capability when propagated on softwood and hardwood spent sulfite liquors. The presence of the plasmid was verified throughout the duration of extended pilot trials using differential plate counts, gas production with xylose as the sole carbohydrate source, shake flask experiments (using the N.Ho test) and xylose consumption in the pilot plant. The results of these experiments confirm that LNH32 was solely responsible for the disappearance of xylose in these trials. CHAPTER 6. CONCLUSIONS 169 The specific growth rates from bench experiments for LNH32 on softwood and hardwood were calculated as u, = 0.028 hr"1 and 0.020 hr"1, respectively. The doubling times for LNH32 on softwood and hardwood were calculated as tdoubie = 25 hrs and 35 hrs, respectively. As a comparison, the specific growth rate for Saccharomyces cerevisiae and Saccharomyces 1400 are approximately equal to 0.018 hr"1 and 0.014 hr"1, respectively, with doubling times equal to 38 hrs and 50 hrs, respectively (Cameron, 1998). This indicates that LNH32 is capable of exhibiting excellent growth on both softwood and hardwood SSL. 6.1.2 Xylose Metabolism and Fermentation The highest levels of xylose fermentation were achieved during the final two weeks of a five-week long extended pilot plant trial on mixed hardwood and softwood SSL substrates. As much as 100% of the xylose present was consumed by LNH32 during this period. These results established that the plasmid-bearing transformant, LNH32, is capable of maintaining sustained xylose metabolism. However, despite the successful metabolism of xylose by LNH32, no appreciable increase in ethanol yield over the current plant strain was demonstrated. The extended trial began with relatively low levels of xylose metabolism, averaging approximately 16%, and ended with a sustained period of 100% uptake. The ability of the LNH32 yeast population to metabolize a larger proportion of the xylose with time indicates that the expression of the xylose-fermenting genes on the plasmids may increase with time. Finally, a decrease in fermentation retention time near the end of pilot plant trial #1 resulted in reduced xylose metabolism, implying that increased fermentation retention time may also invoke further expression of the genes on the plasmids, resulting in improved xylose metabolism. CHAPTER 6. CONCLUSIONS 170 Finally, pilot plant testing of the sugar utilization in the plant by LNH32 supported the conclusion reached by Ho et al (1998) that xylose is co-metabolized with hexose sugars by LNH32. This has historically been a challenge because: a. glucose is very easy for Saccharomyces to assimilate and is therefore preferentially used over xylose which requires synthesis reactions (it is disassembled, and then reassembled) prior to assimilation, and b. there is evidence that glucose and xylose share the same transporter in yeast (Barnett, 1997). In pilot plant trial #1, LNH32 metabolized 30% of the xylose in the softwood SSL, with the majority (75%) of the total xylose uptake occurring in the presence of hexose in fermenter #1. These results indicate that expression of the xylose-metabolizing genes by LNH32 is not affected by catabolite repression. 6.1.3 Fermentation Efficiency and Ethanol Productivity The fermentation efficiency of LNH32 was comparable to results with the plant yeast strain on both softwood and hardwood SSL, despite the fact that LNH32 utilized only 50% of the galactose (the plant yeast strain is capable of fermenting nearly 100% of galactose), and had a lower cell viability at the time of maximum fermentation efficiency. Incomplete fermentation of galactose by the LNH32 would seem to at least partially explain the lack of increased ethanol production by LNH32. Due to the level of xylose fermentation by LNH32, a 15% increase in fermentation efficiency (based on available hexose sugars) would be realized over the plant yeast strain if LNH32 were capable of fermenting galactose. CHAPTER 6. CONCLUSIONS 171 The highest levels of fermentation efficiency were achieved in pilot plant trial #1 on softwood SSL substrate. A fermentation efficiency of 97% was achieved in this trial based on the maximum theoretical ethanol yield from all hexose sugars available. In the same trial, but calculated on the basis of ethanol yield per gram of hexose sugar utilized in the system, LNH32 produced 0.52g EtOH / g hexose sugar utilized, for a fermentation efficiency of approximately 102%. This approach suggests that some ethanol must have been produced from pentose sugar metabolism, as it has been suggested in the literature (and is regularly observed with T2 in the full-scale alcohol plant) that in practice, the maximum fermentation efficiency based on hexose utilized is closer to 90% (due to cell maintenance, natural by-products, etc.) (Ward, 1989). The specific rate of ethanol productivity for LNH32 was calculated at 0.9g EtOH / g biomass / hr in pilot plant trial #1 on softwood SSL substrate. This unit productivity was possibly conservative, because it was unknown exactly when the fermentation reached completion in the first fermenter, and possible that LNH32 could have processed even more substrate in the 6-hr retention time. Regardless, this represents a relatively fast rate of fermentation, which would be advantageous in a commercial operation. As a comparison, ethanol productivity with Pichia stipitis has been reported as only 0.38g EtOH / g biomass / hr, and productivity by Saccharomyces cerevisiae on well-defined glucose medium is approximately 2g EtOH / g biomass / hr (Johansson, 2001). 6.2 Shake Flask Trial with LNH-ST Shake flask experiments were conducted using T2 - the Tembec plant strain, 259A -the LNH-ST parental strain, and LNH-ST - the stable transformant strain. These tests examined the relative levels of ethanol production and sugar consumption between the strains on both CHAPTER 6. CONCLUSIONS 172 softwood and hardwood substrates. Tests showed that a 20% increase in ethanol yield by LNH-ST over both the parental strain and plant yeast strain in both hardwood and softwood SSL. Also, it was established that the stable transformant is capable of utilizing the majority of the galactose present in both hardwood and softwood SSL. In trials with LNH32, it was determined that the plasmid-bearing strain was incapable of utilizing galactose, and as a result, it is possible that any benefits associated with the uptake of xylose were counteracted by a decreased ethanol yield from hexose. The ability of LNH-ST to utilize virtually all available hexose in combination with xylose fermentation translated into increased fermentation efficiency in laboratory trials. 6.3 Pilot Scale Trial with LNH-ST 6.3.1 Strain Confirmation It can be concluded that LNH-ST was primarily responsible for the disappearance of xylose that was observed over the course of this trial. Testing on wild yeast strains isolated from the plant did not use xylose on agar plates, or in Durham tubes. 6.3.2 Xylose Metabolism and Fermentation During maximum utilization, LNH-ST metabolized 25% of the xylose in softwood SSL with all xylose uptake occurring in the presence of hexose in fermenter #1. This indicated that, similar to results found with LNH32, the expression of the xylose-metabolizing genes by LNH-ST is not affected by catabolite repression. CHAPTER 6. CONCLUSIONS 173 6.3.3 Fermentation Efficiency and Ethanol Productivity In this single pilot plant run with LNH-ST, the fermentation efficiency of the recombinant strain was somewhat lower than expected, under-performing in comparison to results with the plant yeast strain on both softwood and hardwood SSL. A maximum fermentation efficiency of 74% was achieved based on the theoretical ethanol yield from all hexose sugars, and 78% based on the theoretical ethanol yield from hexose sugars utilized. Due to the low fermentation efficiency achieved in this trial, there is no way to confirm that the xylose metabolized in the process was definitively being converted to ethanol. Approximately 65% of the galactose was utilized by LNH-ST, and utilization was still increasing at the end of the trial, confirming that LNH-ST is capable of fermenting galactose. 6.4 Implications for the Tembec Full-scale Operation LNH32 exhibited fair stability for a plasmid-bearing strain, and it was demonstrated that the plasmid could be retained throughout extended trials with appropriate selective pressure on the system. However, in reality, the use of a plasmid-bearing strain in the full-scale Tembec continuous fermentation process is not likely commercially viable, as there would always be concerns about strain stability. The ideal yeast for Tembec will exhibit stable performance under typical alcohol plant conditions, while cofermenting the xylose and hexose sugars in the spent liquor from the sulfite pulp mill. LNH-ST has been shown to outperform the plant yeast strain in laboratory trials, with a 20% increase in ethanol yield. Although LNH-ST did not exhibit the same level of performance as LNH32, this stable recombinant strain still shows the greatest promise for reliable use in the full-scale Tembec operation after the completion of additional testing. CHAPTER 7. RECOMMENDATIONS 174 7. RECOMMENDATIONS 7.1 Extended Pilot Plant Trials with 1400(pLNH32) The positive results of the extended trial with LNH32 warrant further optimization and scale-up studies in the pentose pilot plant. These studies indicated that LNH32 became more successful at metabolizing xylose with time. Over time, it appeared that xylose fermenting yeast cells gained a selective advantage in the system, and subsequently displayed higher rates of expression of the genes. A fermentation efficiency of 100% of theoretical (based on hexose utilized) was achieved in the relatively short pilot plant trial with softwood SSL, and only 30% of the xylose was fermented at the time this was achieved. If xylose metabolism were to eventually reach completion during an extended softwood SSL trial (as observed in the extended hardwood trial), ethanol yields may far surpass the theoretical maximum attainable from hexose alone. Thus, additional trials on varied grades of both softwood and hardwood SSL should be performed, and extended for as long as possible to investigate the long term capability of LNH32 to retain its xylose-fermenting capability, and to test the maximum attainable fermentation efficiencies. It may also be of interest to perform Southern hybridization on isolates from extended runs that are positive for xylose consumption, to check whether the vector carrying XR, XDH, and X K has been integrated into the genome of any cells. If this has occurred, it would result in increased stability of the cloned genes in LNH32. CHAPTER 7. RECOMMENDATIONS 175 7.2 Further Testing with 259A(LNH-ST) The success of L N H - S T in shake flask experiments on SSL, justify further optimization and scale-up studies at the bench, and in continuous fermentation trials in the pentose pilot plant. These studies indicated that the stable chromosomal transformant was capable of enhanced ethanol production through xylose metabolism. In shake flask studies, L N H - S T out competed the parent strain 259A and the Tembec Saccharomyces strain, producing nearly 20% more ethanol than each in both softwood and hardwood SSL. Finally, the results of future studies involving L N H - S T should be monitored closely to see i f there are any notable adverse physiological effects displayed by the cells. It is generally accepted that there can be complications with chromosomal transformants related to the interruption of cell functions as a result of the random insertion of genes (Cameron, 1998). 7.3 Genetic Modification of the Tembec Strain (T2) It has been determined by Ho, (1996) that different Saccharomyces strains respond differently with respect to xylose fermentation even if they contain the same plasmid and the same cloned genes (Ho, 1996). Therefore, consideration should be given to the genetic modification of the existing Tembec strain. The Tembec yeast could be transformed with pLNH32 (to produce a plasmid-bearing transformant strain similar to the strain 1400(pLNH32) used in these studies), and/or could be genetically engineered into a stable xylose-fermenting yeast (to produce a chromosomal transformant strain similar to the strain L N H - S T used in these studies). CHAPTER 7. RECOMMENDATIONS 176 7.4 Yeast Purges with the Tembec Strain (T2) and Recombinant Strains Cell growth during fermentation theoretically lowers ethanol yield, by metabolizing more sugars to CO2 and water to produce the energy and synthesis products required for biomass production. However, observations in both the full-scale alcohol plant, and preliminary testing in the pilot plant has indicated that a young, growing yeast population outperforms an older one (Cameron, 1998). These observations have often been made during the early stages of fermentation when a peak in the ethanol production is realized, followed by a levelling off at a fermentation efficiency 5-10% below the maximum attained. These observations are supported in the literature; Ingledew (1995) stated that the growth of Saccharomyces cerevisiae under anaerobic conditions is tightly coupled to ethanol production, and Kirsop (1982) reported that a growing yeast could produce ethanol at a rate 33 times higher than that observed with resting (non-growing cells). While Kirsop's observation does not necessarily translate to an increase in overall ethanol yield, it does illustrate that the productivity of a growing yeast culture is elevated, which would decrease the fermentation time to attain a maximal ethanol concentration. The theory is that some of the substrate is sacrificed to obtain enough growing yeast cells such that the remaining sugars are catalyzed more efficiently to ethanol (Ingledew, 1995). If regular yeast purges were successfully implemented in the full-scale plant, a 5-10% increase in ethanol production could be realized. More testing is required to determine the ideal timing of batch yeast purges (relative to the grades of pulp being produced). The major concern is that a younger yeast population could be less "robust", and be more susceptible to the effects of toxic compounds in SSL (especially during hardwood SSL fermentation, where sharp declines in the yeast viability are regularly observed). CHAPTER 7. RECOMMENDATIONS 177 If yeast purge testing is conducted during the course of pilot plant trials with LNH32, the effects of the yeast purge on the ability of LNH32 to metabolize xylose should be monitored. As a result of the findings in the extended trial (pilot plant trial #2), it was hypothesized that over time, cells that are able to ferment xylose gain a selective advantage with higher rates of expression of the genes. It was also postulated that these cells are the larger and older cells in the system. A yeast purge specifically targets the removal of these cells to support the growth of new, smaller, freshly budded, and highly viable cells. While testing involving a yeast purge during pilot plant trial #4 with LNH32 indicated a positive response in the system with respect to ethanol production, it is possible that this was a short-term effect that was a result of temporarily improved hexose utilization. It is unknown what the long term effects of these yeast purges will be on plasmid retention and expression in LNH32. Finally, testing should also investigate continuous, rather than batch, yeast purges. 7.5 Monitoring of Xylitol Production with Recombinant Strains Several researchers (Jeffries et al, 1998; Kotter et al, 1993; Eliasson, 2000; Johansson, 2001) in the literature have cited the production of excess xylitol as a major challenge in the fermentation or cofermentation of xylose by recombinant strains of Saccharomyces. Also cited is a diminished consumption of xylose under anaerobic conditions (Jeffries and Himmel, 1998). It is thought that both of these factors could be related to redox imbalances arising from the specific cofactor preferences for xylose reductase and xylitol dehydrogenase (Jeffries and Himmel, 1998). Xylose reductase (from Pichia stipitis) is capable of using either NADH or NADPH for the reduction of xylose to xylitol, with NADPH strongly favoured (Johansson, 2001; Rizzi et al, 1989). Unfortunately, the next enzyme in the pathway, xylitol dehydrogenase CHAPTER 7. RECOMMENDATIONS 178 (native to Pichia), uses N A D H exclusively. Therefore, the excess consumption of N A D P H can lead to elevated levels of xylitol while simultaneously blocking the assimilation of xylitol into the fermentation pathways (Jeffries and Himmel, 1998). Some work has focused on the genetic engineering of xylose reductase to use N A D H preferentially, which may eliminate this imbalance (Jeffries and Himmel, 1998). It is possible that the xylitol production could be monitored in the pilot plant during the course of trials conducted with LNH32 and L N H - S T in order to determine whether the observed consumption of xylose is undergoing complete metabolism, or being stalled due to excess xylitol production. 7.6 Alteration of Gene Expression and Enzyme Ratios in Recombinant Strains Eliasson (2000) found that the ratio of activities for X R / X D H / X K was an important factor as to whether metabolized xylose would accumulate in a xylitol pool or undergo further metabolism to ethanol in experimentation with a recombinant xylose-utilising S. cerevisiae. Simulations and fermentation experiments implied that a 1:10:4 relation was optimal in minimizing xylitol formation (Eliasson, 2000). The plasmid bearing strain utilized in these trials had an X R / X D H / X K ratio of approximately 1:2:0.25 (showing X D H and X K were only one-fifth and one-sixteenth of the levels recommended by Eliasson, respectively). Therefore, it is possible that fermentation efficiency could be improved by changing the ratio (or expression) of the inserted genes that control these enzymes in the transformed host. This of course could only be tested if new recombinant strains were produced with varied gene ratios. CHAPTER 8. BIBLIOGRAPHY 179 8. BIBLIOGRAPHY Amore, R., Kotter, P., Kiister, Ciriacy, M . and Hollenberg, C P . (1991). Cloning and expression in S. cerevisiae of the NAD(P)H-dependent xylose reductase encoding gene (XYL1) from the xylose-assimilating yeast, Pichia stipitis. Gene 109:89-97. Asenjo, J.A., Sun, W.H. and Spencer, J.L. (1991). Optimization of Batch Processes Involving Simultaneous Enzymatic and Microbial Reactions. Biotechnol. Bioengineer. 37:1087-1094. Barnett, J .A. (1997). Sugar Utilization by Saccharomyces cerevisiae. In: Zimmermann, F . K . and Entian, K . (Eds). Yeast Sugar Metabolism. Technonic Publ. U K . Bicho, P.A., Runnals, P.L., Cunningham, J.D. and Lee, H . (1988). Induction of xylose reductase and xylitol dehydrogenase activities in Pachysolen tannophilus and Pichia stipitis on mixed sugars. Appl . Environ. Microbiol. 54:50-54. Bjorling, T. and Lindman, B . (1989). Evaluation of xylose-fermenting yeasts for ethanol production from spent sulfite liquor. Enzyme Microb. Technol. 11:240-246. Cameron, D. (1995). Comparison of various strains on softwood SSL. Tembec Internal Documents. Cameron, D. (1995). Yeast Counts. Tembec Internal Documents. Cameron, D. (1997). A Pilot Plant Optimization of Xylose Fermentation. Tembec Internal Documents. Cameron, D. (1998). Personal communication. Cameron, D. (2000). Pilot Scale Xylose / Hexose Cofermentation by Saccharomyces Yeast in Spent Sulfite Liquor. Biofuels Session. World Biofuels Conference, Seville, Spain. Chenier, A . (1998). LNH32 Plasmid Presence Verification. Tembec Internal Documents. Chenier, A . (1999). Transferts dans le SSLin des levures T2, 259A et 259st. Tembec Internal Documents. Crueger, W. and Crueger A . (1990). Biotechnology: A Textbook of Industrial Microbiology. Science Tech Publishers, 2 n d ed., M A , U S A . Delegenes, J.P., Moletta, R. and Navarro, J . M . (1986). The effect of aeration on D-xylose fermentation by Pachysolen tannophilus, Pichia stipitis, Kluyveromyces marxianus, and Candida shehatae. Biotechnol. Lett. 8:897-900. CHAPTER 8. BIBLIOGRAPHY 180 Dellweg, H . , Rizzi , M . and Klein, C. (1989). Controlled limited aeration and metabolic regulation during the production of ethanol from D-xylose by Pichia stipitis. J. Biotechnol. 12:111-122. Demain, A . L . and Solomon, N . A . (1985). Biology of Industrial Microorganisms. The Benjamin/Cummings Publishing Company, Inc., C A , U S A . Does, A . L . and Bisson, L.F . (1989). Characterization of xylose uptake in the yeasts Pichia heedii and Pichia stipitis. Appl . Environ. Microbiol. 55:159-164. du Preez, J . C , Bosch, M . and Prior, B .A . (1986). Xylose fermentation by Candida shehatae and Pichia stipitis effects of pH, temperature and substrate concentration. Enzyme Microb. Technol. 8:360-364. du Preez, J . C , van Driessel, B . and Prior, B . A . (1988). The relation between redox potential and D-xylose fermentation by Candida shehatae and Pichia stipitis. Biotechnol. Lett. 10:901-906. du Preez, J . C , van Driessel, B . and Prior, B . A . (1989). Effect of aerobiosis on fermentation and key enzyme levels during growth of Pichia stipitis Candida shehatae, and Candida tenuis on D-xylose. Arch. Microbiol. 152:143-147. Eliasson, A . (2000). Ethanol production by recombinant and natural xylose-utilising yeasts. PhD thesis, Lund University, Lund, Sweden. Grootjen, D.R.J., Jansen, M . L . , van der Lans, R .G.J .M. and Luyben, K . C h . A . M . (1991). Reactors in series for the complete conversion of glucose/xylose mixtures by Pichia stipitis and Saccharomyces cerevisiae. Enzyme Microb. Technol. 13:828-833. Grootjen, d.R.J., van der Lans, R .G.J .M. and Luyben, K . C h . A . M . (1990). Effects of the aeration rate on the fermentation of glucose and xylose by Pichia stipitis. Enzyme Microb. Technol. 12:20-23. Guebel, D.V. , Cordenons, A . , Nudel, B .C . and Giulietti, A . M . (1991). Influence of oxygen transfer rate and media composition on fermentation of xylose by Pichia stipitis N R R L Y7124. J. Indus. Microbiol. 7:287-292. Hallborn, J., Gorwa, M.F . , Meinander, N . , Penttila, M . , Keranen, S. and Hahn-Hagerdal, B . (1994). The influence of cosubstrate and aeration on xylitol formation by recombinant Saccharomyces cerevisiae expressing the XYLl gene. Appl . Microbiol. Biotechnol. 42:326-333. Hallborn, J., Walfridsson, M . , Airaksinen, U . , Ojamo, H . , Hahn-Hagerdal, B . , Penttila, M . and Keranen, S. (1991). Xyli tol production by recombinant Saccharomyces cerevisiae. Biotechnology 9:1090-1095. Ho, 1M.W. and Tsao, G.T. (1995). Recombinant yeasts for effective fermentation of glucose and xylose. PCT/US94/12861. CHAPTER 8. BIBLIOGRAPHY 181 Ho, N . W . Y . (1996). Genetic Engineering of New Yeasts for Xylose Fermentation.. Internal Tembec Document. Ho, N . W . Y . (1999). L N H - S T . Internal Tembec Document. Ho, N . W . Y . and Chang, S.F. (1989). Cloning of yeast xylulokinase gene by complementation of E. coli and yeast mutations. Enzyme Microb. Technol. 11:417-421. Ho, N . W . Y . , Chen, Z . and Brainard, A . P . (1998). Genetically Engineered Saccharomyces Yeast Capable of Effective Cofermentation of Glucose and Xylose. Appl . Environ. Microbiol. 64:1852-1859. Ingledew, W . M . (1995). The Biochemistry of Alcohol Production. In: Lyons, T.P., Kelsall, D.R. and Murtagh, J.E. (Eds). The Alcohol Textbook. Nottingham University Press, U K . Jeffries, T.W. and Himmel, M . (1998). Session 2, Applied Biological Research. Appl . Biochem. Biotechnol. 70-72:113-114. Jeffries, T.W. and Kurtzman, C P . (1994). Strain selection, taxonomy and genetics of xylose-fermenting yeasts. Enzyme Microb. Technol. 16:922-932. Jeffries, T.W., Davis, B.P., Dahn, K. , Cho, J., Shi, N . , Lu , P. and Hendrick, J. (1998). Genetic Engineering of Xylose Fermentation in Yeasts. U S D A , Forest Service, Forest Products Laboratory, Madison, WI (found on internet in 1998). Johansson, B. (2001). Metabolic Engineering of the Pentose Phosphate Pathway of Xylose Fermenting Saccharomyces cerevisiae. PhD thesis, Lund University, Lund, Sweden. Kilian, S.G. and van Uden, N . (1988). Transport of xylose and glucose in the xylose-fermenting yeast, Pichia stipitis. Appl . Microbiol. Biotechnol. 27:545-548. Kirsop, B . H . (1982). Developments in Beer Fermentation. Topics in Enzyme and Fermentation Technology. 6:79-131. Kotter, P. and Ciriacy, M . 1993. Xylose fermentation by Saccharomyces cerevisiae. Appl . Microbiol. Biotechnol. 38:776-783. Kotter, P., Amore, R., Hollenberg, C P . and Ciriacy, M . (1990). Isolation and characterization of the Pichia stipitis xylitol dehydrogenase gene, XYL3, and construction of a xylose-utilizing Saccharomyces cerevisiae transformant. Curr. Genet. 18:493-500. Krishnan, M.S . , X i a , Y . Tsao, G.T., Kasthurikishnan, N . , Srinivasan, N . and Cooks, R .G. (1995). Process engineering of high-ethanol-tolerance yeast for the manufacture of ethanol. Appl . Biochem. Biotechnol. 51/52:479-493. Ligthelm, M . , Prior, B . A . and du Preez, J.C. (1988). The oxygen requirements of yeasts for the fermentation of D-xylose and D-glucose to ethanol. Appl . Microbiol. Biotechnol. 28:63-68. CHAPTER 8. BIBLIOGRAPHY 182 Prior, B . A . and Kotter, P. (1997). Pentose Utilization by Yeasts. In: Zimmermann, F .K. and Entian, K . (Eds). Yeast Sugar Metabolism. Technonic Publ. U K . Rizzi , M . , Hartwart, K. , Erlemann, P., Bui-Thanh, N . A . and Dellweg, H . (1989). Purification and properties of the N A D + -xylitol-dehydrogenase from the yeast Pichia stipitis. J. Ferment. Bioeng. 67:20-24. Shuler, M . L . and Kargi, F: (1992). Bioprocess Engineering: Basic Concepts. Prentice Hall PTR, 2 n d ed., NJ , U S A . Skoog, K . and Hahn-Hagerdal, B . (1990). Effect of oxygenation on xylose fermentation by Pichia stipitis. Appl . Environ. Microbiol. 56:3389-3394. Slininger, P.J., Bothast, R.J., Ladisch, M.R . and Okos, M . R . (1990). Optimum pH and temperature conditions for xylose fermentation by Pichia stipitis. Biotechnol. Bioengineer. 35:727-731. Slininger, P.J., Bothast, R.J., Okos, M.R . and Ladisch, M.R . (1985). Comparative evaluation of ethanol production by xylose-fermenting yeasts presented high xylose concentrations. Biotechnol. Lett. 7:431-436. Smook, G . A . (1992). Handbook for Pulp & Paper Technologists. Angus Wilde Publications Inc., Vancouver, Canada. Spencer, J.T.F., Spencer, D . M . and Smith, A . R . W . (1983). Yeast Genetics - Fundamental and Applied Aspects. Springer-Verlag, N Y , U S A . St. Onge, R. (1996). Fermentation Efficiency. Tembec Internal Documents. St. Onge, R. and Cameron, D. (1997). Pentose Pilot Plant (June 1997) - A Second Baseline Run using Yeast from the Alcohol Plant. Tembec Internal Documents. Stanbury, P.F., Whitaker, A . and Hall, S.J. (1995). Principles of Fermentation Technology. Elsevier Science Inc., 2 n d ed., N Y , U S A . Takuma, S., Nakashima, N . , Tantirungkij, M . , Kinoshita, S., Okada, H . , Seki, T. and Yoshida, T. (1991). Isolation of xylose reductase gene of Pichia stipitis and its expression in Saccharomyces cerevisiae. Appl . Biochem. Biotechnol. 28/29:327-340. Tantirungkij, M . , Izuishi, T., Seki, T. and Yoshida, T. (1994). Fed batch fermentation of xylose by a fast-growing mutant of xylose-assimilating recombinant Saccharomyces cerevisiae. Appl . Microbiol. Biotechnol. 41:8-12. Tantirungkij, M . , Nakashima, N . , Seki, T. and Yoshida, T. (1993). Construction of xylose-assimilating Saccharomyces cerevisiae. J. Ferment. Bioengineer. 75:83-88. Toivola, A . , Yarrow, D. , van den Bosch, E. , van Dijken, J.P. and Scheffers, W . A . (1984). Alcoholic fermentation of D-xylose by yeasts. Appl . Environ. Microbiol. 47:1221-1223. CHAPTER 8. BIBLIOGRAPHY 183 Toon, S., Philippidis, G.P., Ho, N . W . Y , ZhengDao, C , Brainard, A . , Lumpkin, R.E. and Riley, C.J. (1997). Enhanced Cofermentation of Glucose and Xylose by Recombinant Saccharomyces Yeast Strains in Batch and Continuous Operating Modes. Appl . Biochem. Biotechnol. 63-65:243-255. Ward, O.P. (1989). Fermentation Biotechnology. Open University Press, G B , United Kingdom. Yang, R.D., Grow, D.A. and Goldstein, W.E. (1982). Pilot Plant Studies of Ethanol Production from Whole Ground Corn, Corn Flour, and Starch. Fuel Alcohol U S A . February: 13-16. Yang, V . W . and Jeffries, T.W. (1990). Purification and properties of xylitol dehydrogenase from the xylose-fermenting yeast Candida shehatae. Appl . Biochem. Biotechnol. 26:197-206. APPENDIX A. RECIPES USED FOR LAB CULTURE CULTIVATION 184 APPENDIX A. RECIPES USED FOR LAB CULTURE CULTIVATION A . l Recipes for Yeast Growth on Softwood SSL Inoculum Batch Recipe: • 7.5g Xylose • 7.5g Molasses • 0.5g Fermaid K • 0.183g(NH4)2SO4 • 0 .183g(NH 4 )2HPO 4 • 0.025g M g S 0 4 » 7 H 2 0 • 0.25g Yeast Extract • 30 mLs S W D SSL • 5 mLs Mother Culture in Y P X These ingredients were made up to 200 mLs with tap water, and sterilized at 15 psi for 45 minutes. The resulting xylose to hexose ratio is 1.7:1. 16 L Fermenter Batch Recipe: • 1 L S W D SSL • 40g Xylose • 1 Og Fermaid K • 1 0 m L s H 3 P O 4 • 0.266g Z n S 0 4 • 0.067g Ca-Pantothenate • 0.106g Thiamine • 4 .16gMgS0 4 » 7 H 2 0 • 20g Molasses The xylose, Fermaid K , and molasses were made up to 200 mLs, 100 mLs, and 100 mLs respectively, and all were sterilized separately. The xylose and Fermaid solutions were sterilized at 15 psi for 30 minutes, and the molasses was sterilized for 1 hour. The rest of the APPENDIX A. RECIPES USED FOR LAB CULTURE CULTIVATION 185 fermenter recipe was made up to 6 L with tap water in the lab fermenter and sterilized at 97°C for 2 hours. The pH was adjusted to 5.0 with concentrated NaOH in all solutions prior to sterilization. The resulting xylose to hexose ratio is 1.2:1. Feed Solution: • 80g Molasses • 60g Xylose • 500 mLs S W D SSL The xylose and molasses were made up to 300 mLs and 200 mLs respectively, and were sterilized separately. The pH was adjusted to 5.0 with concentrated N a O H in all solutions. The SSL and xylose solutions were sterilized at 15 psi for 30 minutes, and the molasses was sterilized for 1 hour. The resulting xylose to hexose ratio is 1.2:1. A.2 Recipes for Yeast Growth on Hardwood SSL Inoculum Batch Recipe: • 7.24g Xylose • 8.14g Molasses • 0.5g Fermaid K • 0 .183g(NH 4 ) 2 SO 4 • 0 .183g(NH 4 ) 2 HPO 4 • 0.025g M g S 0 4 « 7 H 2 0 • 0.25g Yeast Extract • 30 mLs H W D SSL • 5 mLs Mother Culture in Y P X These ingredients were made up to 200 mLs with tap water, and sterilized at 15 psi for 1 hour. The resulting xylose to hexose ratio is 1.7:1. APPENDIX A. RECIPES USED FOR LAB CULTURE CULTIVATION 186 16 L Fermenter Batch Recipe: • I L H W D SSL • 20g Xylose • lOgFermaidK • 1 0 m L s H 3 P O 4 • 0.266g Z n S 0 4 • 0.067g Ca-Pantothenate • 0.106g Thiamine • 4.16gMgSCv7H 20 • 30.6g Molasses The xylose, Fermaid K , and molasses were made up to 200 mLs, 100 mLs, and 100 mLs respectively, and were all sterilized separately. The xylose and Fermaid solutions were sterilized at 15 psi for 30 minutes, and the molasses was sterilized for 1 hour. The rest of the fermenter recipe was made up to 6 L with tap water in the lab fermenter and sterilized at 97°C for 2 hours. The pH was adjusted to 5.0 with concentrated N a O H in all solutions prior to sterilization. The resulting xylose to hexose ratio is 1.2:1. Feed Solution: • lOOg Molasses • 60g Xylose • 500 mLs H W D SSL The xylose and molasses were made up to 300 mLs and 200 mLs respectively, and were sterilized separately. The pH was adjusted to 5.0 with concentrated N a O H in all solutions. The SSL and xylose solutions were sterilized at 15 psi for 30 minutes, and the molasses was sterilized for 1 hour. The resulting xylose to hexose ratio is 1.2:1. APPENDIX A. RECIPES USED FOR LAB CULTURE CULTIVATION 187 A.3 Recipes for YPD, YPX, and Lysine Agar Plates YPD Agar Plates: • 10 g/L Yeast Extract • 20 g/L Peptone • 20 g/L Dextrose • 15 g/L Agar In preparing these agar plates, the sugar was sterilized separately from the balance of the ingredients. After sterilization, all ingredients were mixed, and poured into sterile plastic agar plates. YPX Agar Plates: • 10 g/L Yeast Extract • 20 g/L Peptone • 20 g/L Xylose • 15 g/L Agar In preparing these agar plates, the sugar was sterilized separately from the balance of the ingredients. After sterilization, all ingredients were mixed, and poured into sterile plastic agar plates. Lysine Agar Plates: • 2.4 g/L Yeast Extract • 0.12 g/L Lysine • 10 g/L Agar In preparing these agar plates, ingredients were all sterilized separately, and after sterilization, all ingredients were mixed, and poured into sterile plastic agar plates. APPENDIXB. ADDITIONAL GRAPHS FROM PILOT PLANT TRIALS 1 APPENDIX B. ADDITIONAL GRAPHS F R O M PILOT PLANT TRIALS B.l Pilot Plant Trial #1 - Softwood Spent Sulfite Liquor o cr £ 3 c G J D -C/3 73 O o o C/3 =tfc c J2 CL i£ CQ L. V Q. IA O o x c £ ra I ("1/6} uoiteJiuaouoQ |oui>ma H3k (1/B) uo!)ej)U33uoo iseaA C/3 -J < 5 f -E— Z < -J a. h-O —i o UH on X C L a —i < z o Q Q < CQ X 5 z LU CL C _ < ra a; TJ ^ i P 13 £ s o 5 (1/6) u o j i e j j u a D u o o |Oueqi3 (1/6) uofiejluaouoo |Oueqi3 o 3 CT J 3 c/3 c OJ D. T3 O 0 o 'u. c CQ (n/6) uojjejiuaouoo loueijig o c 3 CO O ft OJ S • — n (1/6) s||ao aiqeiA (%) <l!liq«IA 1183 S E c ti • Total Hexose —*— Hexose Utilized —•-Total Sugars —*— Sugars Utilized -•-Viability • Total Hexose —*— Hexose Utilized —•-Total Sugars —*— Sugars Utilized -•-Viability • Total Hexose —*— Hexose Utilized —•-Total Sugars —*— Sugars Utilized -•-Viability {%) Aouapijig uonejuaiujaj C/j < 2 H H Z < Cu f— O — o u. 0-a - J < z o p S Q < CQ >< Q Z UJ c CL < (n/B) U0!iej)uaouoo |OUBL)13 C c ro a> £ E (%) A o u a p j i i g uoi j t ' iusujja- j APPENDIXB. ADDITIONAL GRAPHS FROM PILOT PLANT TRIALS B.2 Pilot Plant Trial #2 - Hardwood Spent Sulfite Liquor \°\S o 3 C T S 3 C/) i-> c <u D. C/) T 3 O o e3 'C h-+^  c « CL •*-» C N CQ 111 -t-> (1/6) uojieJiussuoQ asouiqejy 1<U SSLin -•-Fermenter #1 —*— Fermenter #2 -•-Surge Tank 16.00 • 12.00 -10.00 -B.OO -6.00 -4.00 -2.00 -0.00 -(1/6) uoiiBiiuaauoQ eso|Ax < H fc < -J CL H O — CL O u. (A X CL i O — i < z g P 5 Q < CQ X 5 z LU CL Cu < CO I L £ CO 1 1 1 a 8 (1/6) uojiBJiuacuoQ jeBng (~|/G) uojiKJiuaauoQ asoan|9 o 3 cr J u E 3 CZ3 4—1 c <o D . on X) o o "2 X CN 13 'C f— c 03 c~ rs cd o o q q 11/6) iioiiKJiuaouoo ssouuu'W (1/6) uoi)ej)U33uoo |ouei)i3 on < Qi f -E— Z < -J CL E— O — c_ S O oi L L on X CL < a —i < z o f-5 Q < CQ >< 5 z UJ CL c_ < (~|/6) uoiej)u»3uo3 esoiseieg ("1/6) uoieiuaauoQ S S G X S H | B J O J . MS o 3 cr LJ 3 C/J 3 U c oo -o o o "2 X I CN c _o (X CN cd ("1/6) U0!iej]u«3U03 [oiieiua (1/6) uoi)BJ)UB3U03 jseex cn -J < 5 f-f-z < o. E— O _ cZ o c^ u. t/3 X CL O -J < z g f— Q Q < cd X 5 § C L CL < S i (1/6) uo|)BJ)ueouo3 |ouei)i3 (1/6) uoiieJiuesuoQ toueuig o 3 cr 3 c u 5, C/3 o o i CN — « .2 E— c _w Cu CN cd (1/6) uoj)ej)U33uoo louegjg C |I cu CP 5 t_ — 1 — 1 )l Produced \ 11 IL U t + t (1/8) S||30 3 |qeiA (%) AiwqeiA lieo O 1 • E CM Ol il ro (%) Aauamiyg uorteuieuue j -J < f— f-Z < -1 Dt, (— o — a. O u. C / ! l Cu i O —i < § P S Q < CQ X 5 z uU Cu Cu < (n/6) uojieJiuaouOQ |oueu,jg o 2; c c (O 111 £ E 1 1 1 1 « ? 1 ^ ^ o ? i 1 b ID £ U_ ^ LU l + t E (%l iCllliq«IA (%) ADuapiyg uo|)ejuauijad APPENDIX B. ADDITIONAL GRAPHS FROM PILOT PLANT TRIALS B.3 Pilot Plant Trial #3 - An Extension of Pilot Plant Trial #2, Blended SSL in -a u 73 C _o CQ CN C .2 ' £ ' - 4 -o c _o 't75 c u X LU c < 13 f— c CL cl CQ ! 1 1 Settling Tank •*• Fef menter #1 •*— Fermenter #2 -•— Surge Tank O Theoretical Max. l / (~l/6) uo|iej)U83uo3 loueuja (~|/B) uo)}ej)ua3uo3 )BB»X C/3 —I < 5 h-H Z < -J n_ i-O — C L O a: L L X C L a -J < z g H 5 a < CQ >< 5 5 O H C L < (0 _0> TJ -Q 0) TO u — 3 a "° 2 o < £ 8 10 "5 P ( T / B | uoijcuueDUOp |oueiu3 (1/6) uo;iej)ue3uo3 |oueu,)3 APPENDIX B. ADDITIONAL GRAPHS FROM PILOT PLANT TRIALS 2 0 6 B.4 Pilot Plant Trial #4 - An Extension of Pilot Plant Trial #2, Blended SSL C/3 -o TJ c CQ f-(1/6) UOI1BJJU33U03 |0ueqi3 •= TJ U C 3 « TJ T -0 * Q. y o 5 c 1 § ^ • 0) 0) J c % cd o CL _o CL <+-o c o 'en c CU *-» X LU c < I 1 -Is 'C f-c CL •4—* CL CC (1/6) s| |83 stqeiA | i (%) /3uap!M3 uo|)e)uatujBj t/3 —1 < CC H f-Z < - J CL t -O — CL O L L CL O —i < z o 5 Q < CC X s z CL O H < ("1/6] uoiiejiuesuoj |Oueu,jg o * £ £ o £ c c ro oi £ E (%) Xou»j3|J43 uout'iueuuej APPENDIX B. ADDITIONAL GRAPHS FROM PILOT PLANT TRIALS B.5 Pilot Plant Trial #5 - LNH-ST on Blended SSL —1 (/) C/3 T3 <u T3 C JJ CO c o H ON CN in c jS c _o c~ IT) cd C/D -1 < 5 t -z < -1 c_ h-O —) c u. X i O < z o H S Q < cd >< 5 z UJ 0-< 3J3> (1/6) uo!iBj|u«3uoo jeBns " o « c » < 3 x a 2 H q 2. E ("1/6) uojiKJiuaauoQ jeBng (1/6) uoj)ej)U83uoo jeBng ( H I D (l/B) uoijejjuaouoo jeBng —1 C/D C/D T3 <U X ! C jy CQ c o H C/D i n .2 "C c JS o_ +u m CQ C/D -J < s H E— Z < Cu 6-O — Cu o uu C/D X QU < o < z g P S Q < CQ X 5 Cu Cu < +11 8 8 (1/6) U O J I B J I U 9 0 U 0 3 O S O U U B L ^ (1/6) uoi]ej)UB3U03 9so}oe|eg 3U-T r T J • Fermenter #1 Fermenter #2 • Fermenter #1 Fermenter #2 i i f ("1/6) uojtejfuaouoo | oue i | i3 s EL o it m £ o to 8 O) 2 T + t < APPENDIX C. RAW DATA FROM BENCH & PILOT PLANT TRIALS 2 APPENDIX C. RAW DATA F R O M BENCH & PILOT PLANT TRIALS C l Bench Trial #1 - Softwood Spent Sulfite Liquor, Continuous Bench Fermentation Trial #1 grow-up with L1400(pLNH32) in the 161 Lab-Scale Fermenter on Softwood SSL. Mode of feed: Batch, Fed-Batch, and Continuous Feed Date May,25/98 May.26/98 May,26/98 May.27/98 May.27/98 May.28/98. May.28/98 Time 5:30pm 8:30am 1:30pm 8:30am 1;30pm 8:30am 1;30pm # of Hours into Trial 6 \5 i6 39 44 6.3. 68 Temperature (°C) 30.9 31.4 31.0 30.4 30.0 30.4 30.1 Aeration (L/miri) 1.5 1.5 1.5 6 6. 6 6: pH 5.07 5.05 5.04 .S.66 i6e •5.6i % Solids (w/w) 4:96. 4.50 4.59 5:i;o 5.46 5.88 7.37 % Sludge Volume 0.12 0.60 0.94 2,30 2.36 3.10 3,20 Yeast Concentration % Viable P/o Budding Viable (g/L) - 67.6 62,5 76:7 77.2 64,3 71:9 - 19 2 5:5 4 5.4 4,3 10.8 - 2.43 1.83 4.47 5.60 6.05 7.58 Ethanol (g/L) 0.40 3.10 2.6.4 . 6.61 6.$6 0.04 i39 Sugar Concentration (g/L) Arabinose Glucose Xylose Galactose Mannose Total Hexose 0.37 0:28 0.37 0.37 ;o.24 0.34 .0.59. 2.03 0.00 O.OQ 0.00 0.00 0.00 ,0,00, 7.35. 6.11 .5.76. 5:68 5.23 3.1.5 9.71 0.90 0.00 0.00 0,00 0.00. 0.00 0,24 3.62 0,00 0.00. 0.00 0.00. 0.00 :o.oo. 6.55 0.00 0.00 0.00 0.00 0.00 P. 24 Confirmation of the N. Ho Strain. Sample LNH 3 2, Lab.FeTmenter(lx) LNH 3 2, Mother Culture Date May.27/98 May.28/98 May.29/98 May.27/98 May.28/98 May 29/98 Time 5:.00pm 5:00pm 5:00pm 5:00pm 5:00pm 5:00pm # of Hours into Trial 0 54 4S 0 24 48 Ethanol (g/L) 0.0 44.0 53.5 0.13 27:0 53.7 APPENDIXC. RAW DATA FROM BENCH & PILOT PLANT TRIALS C.2 Bench Trial #2 - Softwood Spent Sulfite Liquor, Batch Bench Fermentation Trial #2 grow-up with L1400(pLNH32) in the 16L Lab-Scale Fermenter on Softwood SSL. Mode of feed: Batch Date May.28/98 May, 28/98 May.28/98 May:29/98 May 29/98 May.29/98 Time 4:15pm 8:15pm 11:15pm 8:15am 12:15pm 4:15pm # of Hours into Trial 0 4 7 16 20, 24 Temperature (°C) 31.3 30,8 30.1 30.4 30.7 30.8 Aeration (L/min) 1 1 1 1 1 1 PH 5,04 5,05 5.02 5.09 5,03 4,99 % Solids (w/w) 20.34 19.50 18.86 '18.43 18.41 18.06 % Sludge Volume 2.00 1.60 2.00 2.00 1.80 1.90 Yeast Concentration % Viable % Budding Viable (g/L) - - - 17 - 15.3 - - - 5.6 - 7.1 - - - 0.90 - 070 Ethanol (g/L) 0.66 5,51 1.0.02 12.60 12.34 12.09 Sugar Concentration (g/L) Arabihosei Glucose Xylose Galactose Man nose. Total Hexose! 2.04 2.12 2.15 1.93 1.79 1.70 6.72 0.98 0.00 0.00 .0,0.0 0.00 8.03 8.38 8.17 6.32 5.71 6.07 4.95 5.06 4.62 1,44 0.46 0.00 19.28 12.76 2.89 0.00 0.00 0.00 30.95 18,80 7.51 1,44- 0,46 0.00 APPENDIX C. RAW DATA FROM BENCH & PILOT PLANT TRIALS 220 C.3 Bench Trial #3 - Hardwood Spent Sulfite Liquor, Continuous Bench Fermentation Trial #3 grow-up with L1400(pLNH32) in the 16L Lab-Scale Fermenter on Hardwood SSL. Mode of feed: Batch, Fed-Batch, and Continuous Feed Date June.8/98 :June:9/98 .June.9/98 June.10/98 June.10/98 June.1-1/98 June.11/98 June:12/9£ Jun'e:i2/9£ June.13/9£ Jurie.14/9E June.-15/98 Time 4:00pm .8:30am 1:30pm 8:30am 1:30pm 8:30am' 1:30pm 8:30am 1:30pm 11:00am •12:45pm 8:30am # of Hours into Trial 16.5 ~'m 40:5 453 ,W,5 69.5 M 93.5, 1.15; m ml Temperature |°C) 33:5 '32:6 '30.9 31.0 31.4 .31.5 30.9 '31.8 3113 312 :30.6 .31.4 Aeration (Umin) 1.5 1.5 1.5 ->6 6 6 .6 6 6 6: 6' .6: 6. PH 5,10 5.05 5.10 :5:oo 4.93 5:05 '4.93 4.98 . :5.04 5:1.1 5:00 % Solids |vrfw| 4.82 4.47 •4:54 5.29 :5.80 6.35 6.66 :7.51' 7:76 .8:77 9:27 .'9:78 % Sludge Volume 0:12- 0:36 .0:38 0:52 ,0.56 1.10 1.02 1.26 1.38 2:10 3.5^ 0 •.4'.00 Yeast Concentration % Viable % Budding Viabie'tg/L| 14.9; 28 ;31.2 29.1 31.4 38.6 37.1 44:6 .42.4 .49:3 '54.1 •39.1 15.9 12:5 7.7 •17.1 18.6 19.0 .26.1 •26.0 . 19.5 13:2 5.1 11.3 0:09 0,19 0,27 0.26 0.31 0.90 0:79 1.45 1.95 5,36 15:70 ,11,90 Ethanol ( g l ) 0:28 ,2.24 2.23 1.50 2.12 :2.23 1:95 =1.64: 1.54 -1.-71- 0.16 0.07 Arabinose; Glucose, Xylose Galactose Mannose Total Hexose 0.25 0:00 0.18. 0:20 o:oo 0.31 0,28 o.2i: 0:29 0,28 .0:33 0:52 3.45. .0.00 .0.00: •1:54 1:30 0.00 o:oo .0:00 0.00 0.00 0.00 •0:00 7.87 :5.64' 6:78 •8:8.1 11.20, 14.85 15.55 18.06 20:51 24.06 21.52 18:66. 0.46 0.00 0.00 0.00 0.00 .0.00 0.00 0.00 0.00 0.00 .0:00 0.00 1.85 0.00 0.00 0.00 0.00 0:00 0.00 0.00 0:00 0.00 0.00 0.00 5:76; 0.00 .0.00 -1:54 1:30 :o;oo o:oo 0.00 0.00 0:00 0.00 •0:00 Cbnfirmation of the N. Ho Strain. | LKIM3; , Lab Fermenter (Ix) | LKIH32l.Molher Ctflure | Date Time #.of Hours into Trial 5 24 48 0 24 48 6.6 23.2 iii :.6,l3 ilb 53.7 APPENDIX CRAW DA TA FROM BENCH & PILOT PLANT TRIALS 221 C .4 Bench Trial #4 - Hardwood Spent Sulfite Liquor, Batch Bench Fermentation Trial #4 grow-up with LUOOfpLN^) in the 16L Lab-Scale Fermenter on Hardwood SSL Mode of feed: Batch Date June. 16/98 June. 16/98 June. 16/98 June. 16/98 June, 17/98 Time 8:00am T2:00pm 4:00pm . 8:00pm . •8:00am # of Hours into Trial 0 4 8 12. .24 Temperature (9C) 31.0 32.4 32.9 30.8 314 Aeration (L/min) 1 t 1. 1 •1 PH 5.15 5.24 5.05 5.04 4.94 % Solids (w/w) 22.80 22.69 22.56 22.35 ' 22.14 % Sludge Volume 1.74 1.74 1.76 1,80 1.88. Yeast Concentration %Viable 34.7 19.5 T5.1 12: 4.2 % Budding Vjabie(g/L) 26.5 25 19.4 25,9 27.7 3.55 2 1.65 1.35 0.47 Ethanol (g/L) 0.19 0.87 2.71 3.70 4,14 Sugar Concentration (g/L) Arabinose' 1.48 1.48 1.41 1.46 1.32 Glucose 445 3.03 0.00 0.00 0.00 Xylose 21.63 21.75 20.71 19.08 17.69 Galactose 4.35 2.68 2.51 2.57 2.03 Mannose 10.64 9.78 6.76 0.00 0.00 Total Hexose 1.944 15.49 9.27 2.57 2.03. APPENDIXC. RAW DATA FROM BENCH & PILOT PLANT TRIALS 222 C.5 Pilot Plant Trial #1 - Softwood Spent Sulfite Liquor IDate-Tlme (# Hours into Trial 3 into Trial 02-AUP-98 12.30 AM Da^ sjr tO-Aufl-98 04-Auo:'M 05-Au(»:98 06-AUO-98 07-Aug-98 OB-Aug-SO. O^Aug-98 .10-Aus-W 11-AUfl-9B 12-Auff-99 13-Aug-9B 14-AU9-98 l5>Aug-98 16-AUQ-M 17rAufl-98 9.30 AM 9 30 AM 1030AM 10:00 AM 10.00AM 9:45 PM- 1t;15AM 9.15 AM 9.30 AM 9:00 AM "8.30AM 9:15 AM 10.00AM 10:40 AM 9;30AM 9 33 58 01:5 '95.5 1 31 145 167 1 91 214.5 2 38 263 287:5 312 336 1 5 5 6 0 7.0 .30 8.9, 10.0 11.0 11.0 .13 0 ,14.0 1.4 24 3.4 Sef;go/htsl_. |Feerj.Rate:(L7rnin) temperature ('cj |PH O p e r a t i o n * . Storage Tank Levei;(%) lSepV lb ra t lo r ^ n T m / ^ ^ PH . , Sludge, volume (mi/50mi) % Solids (w/w) Sugars. (g/L) Arabinose Glucose xylose-Galactose Mannose iTotal Hexose 2.77 •397 3B6 4.12' 3.91 3.82 4.32 3.W i?40 3.10 3.29; 3.20 346' 0.04. 002 a o r 003 0 03 0.02 003 0.04 002 0.02 0.08 0.06 0.05 21.2 21.46 20.92 21.89 21.91 21.98 2231 2241 22.2 22.37 2219 • 22.2 22.56 1.55 1,64 1.58 1.69 1.76 V53 1.54 1"82 1.55 1.9 '1.32, 1.21 1,18.' &;37 6.9 7; 43 .5.33' -6.98 6.3 6.98 '6.2 597 7.49 7.03 •512 '631 6:84 7.25 7,44' 7.44 '7.91 6.B7 6.99 6.97 6.'69 B.'ir <t'Ar. 54 "5.63 • 47 503 -4'95 5.08 '5.57 4.79 4.89 482 4:8 5.69 -4.59 •3.81 3.9B 17.38 18.35 20)36 19.68 20.75 18.35 19.39 18.22 17.71 23.04 19 14:43 17.68 28 43 3D 28 32.74 29.07 33 3 29 44' •31.06 29.24 • 28.48 36 22 30 82 23 38 27,97. 31.7 28.3 34 336; 353 35 2 35.9 33' 31.1 32 32 8 30.3 '30.1 2.78 396 3 84 4.10. 4.05 3.92 4.22 4 00 350 3.22 3.37' 3,40- 3.65' 0.03 002 0.03 0.02 0.03 0.03 0.04 003 0.02 0 03 0.02 0.04. ,0.03 2042 20.99 21.24 22.8B 21 '.79 21:99 2268' 22:22' ; 2Y4 2' ',22.26 22:47 22.B, 22.85 0024 014 0.14 6,031 0.03 O.OJ 0.02 o!o3 ' 0.031 0.03 004 0.06 .0.161 i:54" 'e.e 1.56 1.54 1.38 1.83 1.52 1.53 1:54 1:62 1I6B ,1.46 1:23 t. 6 57 7.34'. 5.05 6:i 5.34 5.93 7.07 645 7.25 7.04 6 04' '6.19 8.84' 7.08 7.11. .6.06 ;7,66. 6.49 6.93 7.78 7.11 7.98 6,8 59 :5:53. 4,73 4 78 4 89 4 2? 4.82 451 4.86 5.67 5 25 5.62 ' 461 •4.21 3 88. •1716 . 17.8B 19.75 15 6 17,51 17.01 18 45. 20.42 .18:82 221 isV 15.7. 14.67; 28:49 29 23 31.98 24.87 28.43 26.86 29,24 33.16 .30:S2- 34:97 .30.55 25.95 24,74 28.3 27.8. .28 '27,9 291 29 5 30.3 28.3 27.9 289 30.1' .27.7 2B.2; MED '4.98 MED MED MED MED MEO*MIN MIN=»MED MEO MED MED MED MED MED' 509 '4:96 "5.02 5.00 5.07 5.00 4.98 502 502 4.92 '4.94: 4.90 Settling Tank Temperature. (*C) pH Sludge.voiume (mi/50ml) % Solids (w/w) Etriaipl (g/L) Sugars (g/L) ! AraDinose Glucose' Xylose Galactose Mannose* Total Hexose Ftrm«nttr#1 Temperature (°C) 35.0 27.3 27,8 Aeration MAX MAX MAX P H 5.08 4.98 5.00 Ye33t Concentration %'viaoie: - -.40 73.6 %-Budding - . '2.8 e i Viable'(cyL) • 0.036 6.345 Total (C/L) • 0 09 0.47 SlUdge volume (mL/50ml) 003 003 o.i4 % SOIIOS (WAV) ' 3.6? 4,23 EtTianol (g/L) p.057 o.36 2,17 Sugars (g/L) AraDinose o.'43 0.23 0.2 GlUCOSe 2.28 C.87 0 [Xylose ? 33 1.35 1.1 Galactose >;-59 0.93 o iMannose' 5.86 2.95 0 Total Hexose 973 4,75 0 iFermerrtaoon Efficiency (total hex) 037 2.33 14,04 Fermentation Efficiency (hex used) 0.54 276 14 04 'Fermentation Efficiency (all sugar) _Q j g _ _ l g j _ _ l i B l _ 47.4 15.9 0.37. 0.46 0 77 14,58 5.07 3181. 46.02 2543 14.81 9 US' 58.43 58,43 45.29 57.92 57 92 44,89 0.32 20.29 6.45. 64.59 32.40 2.32 0.55 20.17 5 26-.9.56 65.00 80.36 42.63 2.56 3,06 0 70 20.71 9.66 6.63 2.95 0 2.95 62.51 69.25 4B.'45 499 095 21,2 2.79 6,59 72 4B 93 44. 66.71 a 90 .21.22 11.28 72.99. 7B.60 ,56.58 4.35 0.93 3.97 8.26 62.71 8620 2.29 .3.74 21-55 10.5 6 49 67 95 21.6 M.16 3.02 0 • 3.02, Ba'45 90 50 6353 21.5 12.02 90 72' 68 42 |F*m»nt»r*2 Temperature (*C) [Aeration IPH .'east Concentration % Viable' % Budding viaple(cyL) Total (c/U Sludge volume (mi/50mi) it Solids (w/v) Etftaooi (g/L) Sugars (g/l_) Araolndse Glucose Xylose Galactose Mannose-Total Hexose iFeimematJon Efficiency (total hex) Fermentation Efficiency (nexused) Fermentation Efficiency (al sugar) fermentation ETfTciency (sugar usea| 26.1 30.3 X.3 307 30 3 30.9 32 30,4 29.3 30 7 32 7 30.5 30.9 MED MED MED MED MED MED=>MIN MIN MIN MIN MIN MIN MIN. MIN.1 4.79 473 4.76 491 4.92 491 4.84 4.91 4.'eo" 47_5 •4.78 474 4.68 83.8 83.1 81.7 71.6 69 4 73.6 62.9 65.2 45.6 58 2 .47.9' '50"8 40 8.2 7.9 9.3 24.5 181 5.8 16,7 177 2.2 23. 61 7.4 3.5 0.61' 0.7 0.69 1:39 1.27' 1.73 1.65 2.32 a eg 1.84 3.1 3.09: '2.84' 0.73 0.84 084 1.94 1.B3 2.35 2.62 356 1.96 3.16 6.47 B.OB' 7.10. 0.29 030 0.30 0.33 0.55 059 0.47 a 84 078 0.90 0.98 1.06 21.5 1,00' 9.6 12.7 8.18 13.99 1B.34 19.42 2036 20"93 21.131 21.11 21 '.35 •2l':45 21.43 4.5 &79' 9.75 10 2 11.13 14 14,9 11.97 11.5 12 01 1*1.82 1294 0.63 Q 0 86 06 1.42 1.53 1.43 1.06 0.46 021 6.27 0.27 6 0: 0 0. 0 0 0 0 0 6 0 0 0 0. 2!92 151 247 5.11 6.15 5.54 617 5.12 4.39 6.38 6:12 494' 4196 0.65 0 0" 2.08 3.21 2.25 1.59 1.58 1.16 3.39 3.43 3.01 -1.67 0 0 0 0 a 0 0 O 0 0 0 0 0 0.65 0 0 2.08 3 21 2.25 1 59 1.58 1.16 3.39 3 43 301 1.B7 29.12 52.93 56 BB 63.09 66.01 72.02 SO 60 96.42 •77.48- 74.42 7772 8521 93 29 2976 52.93 56 BB 67.74 73.83 77,80 95.61 101,72 eo!54 63.79 8 ™ * 95 82 1000 7" 22.57 41.03 '44.09 48.90 51,16. 55.82 70.22 7473' '60.04 67.68 60 24 67.29 7366 25.29 46.31 47~85 62:72 70.91 73.05 87r78 91,49 70.41 .77 "31 ms 97,47. 91 '.88 SurgiTink PH Sludge volume (mi/50mi) % Solids (w/w) Etiiaiul fij/li) Sugars (g/tl) Araolndse JGIucose pytose IGalacluse Mannose (total Hexose Tneo.Max-EtOH (hex ln:settiingrj<) fne'6.Max:EtOH'(ha( used) 1545 15.12 4.79' 4,69 4.89 4.50 483 4.99 4.79 4.74 .4.77' -472' 4.65 0.42 0.34 6.08 0.17 0.10 017 0.32 6.30 0.30 0.10 0.16 13.16 IB 08 19.07 20 21 20 75 21 09 21.05 21.28 21 4* •21.4 21.42 7.54- 10.27 103 10.98 13.3 14.8 11.68 11;54 11:43 11j7B 1311 •0.52' 1.37 •1,?7 1.63 1.23 04 026 0.23 .0.26 0' 0 0 0 "0 0 0 'O V b 6 b 0: 202 J.94 5.34 6^ 56 5.57 4.41 4! 32 654 6.21. .4 69 Ve5. 0 V.B2 '2.55 2.54 1.72 i.42 1.23 3,42 J.38 2.75 '2*03 6 •6 0 0 b 0 0 0 0 0 0; 0 1.82 •2.55 2.54 1.72 1.42 i:23 3.42 •3,38 ,275 2.03' 15.45 15.46 15,45 15,46 15 45 15 45. 15.45 1545 15:45 15.45 13B7 13.87-15.45 IS 45 14.39 13 32 14,31 14*64 14.65. . 14,66 ,1372 1370 .1234 U92 APPENDIXC. RAW DATA FROM BENCH & PILOT PLANT TRIALS 223 C.6 Pilot Plant Trial #2 - Hardwood Spent Sulfite Liquor Date 22-Sep-9B 23-Sep-98 24-Sep-98 25-Sep-98 26-Sep-98 27-Sep-88 28-Sep-98 39-Sep-98 30-Sep-98 01-03-98 02-Od-9e' 03-Od-98 04-OCJ-96 05-Od-98 06-0(3-98 08J3C1-98 OS-Od-98" Time 900PM 11;3p'AM 10 00 AM 9:00 AM 10 00 AM ittoo AM" 8,30 AM 9:00 AM 900 AM 9:00 AM 9 30 AM 5:30 PM 3QQP~M 1:30 PM 1000 AM '10XAM CHoLTSirto Trial o lis' 37- 60 65' 109 1315 1S6 180' "204 ' '2285 260,5 282 3045 3265 373 3975 # Days into Trial o'.0 06 1.5 1.5 3.5 45' 5.5. 6.5 75' 65 9.5 109 n.a 12.7.' 13.7 15.5 16.6 Sitfftv/rtji; . . . . . . - • - - - • •-Feed Rate (L/min) 2.0 2 0 20 2.0 2.0 3.0 2.0 2.0•» 00' .20 20 20 20 06 1,5 Temperature (*C) 30.0 X O 28.0 28.0 280 28.0 28 0 380 28 0 38 "0 289 4-9 26.0 28.6 38'6. 380 A3 38,0 49 PH 5.6 53 SJ> S.0 5.0 SXI .5.0 5.0 5.0--4.8 4.9 45 49 4.9 49 Qaeretktns . . . ; : ~ Storage'Tank Level (%) 76 76 75 74 68 62 56 51 45 38 .27 26. 13. 14 10; 0. • 0 sehvibratlons'fmm/sec) 2.6 5.5 4.5 4,7 40 5.1 36 3.5 4.5, 4.2 44 4.4 '55' 4.4 Fermentatioht * ... , • • •-— • ••••• --•• PH .3.47 3.08 3,42 3.53 3 45 3";S5 3,71 3,54 3,40 3.14 3.31 •321 300 3.94 Sludge Volume (mfSOrfil) o"6s 0,07 004 0.04 094 0 04. 0 03 ,0,10' 005 0.03 0.05 0.04. % solids '(w/w) 18.77 20 OO 20 35 MOT 21 00 3i'ii 21*07 20^ 96 21.30 21 Jl 2126 21 .54; 2152 21J7-Sugars'(gt.) ,1,01 Arabinose 0 78 0.71 i.01 t.03 .1.06 0.98, 1.05 • 1.18 095 T.03 1132 0.88 0.69 Glucose 3.27 3.55 4.10 4,51 4.36 4.27. 5.18 5.16 3 SO 4,65 4 09 •4:30 3 97 4:32' Xylose 13.69 13.32 1530 •16.16 17.11 .15.16 18.27 1652 -13.97' 17,45 16.43 15.81 15 01 1568' Galactose 1.78 .1.69 206 2,22 2.16 2.06' .331 I.'48 1 91 2.11 2.02 2,00 .1:74 1.98 Mannose S.S9 4 B3 5.49 6.17 8.90 6 42 10. I B 7,70 5.28 6,75 8 48 6.48. 7 65 e.ei is . i i Total Hexose .10.64 10.07 11 67 12.90. 15.42, 12.75 1767 15.34 11.09 .15.51 14.59 14,78 1356 [settling Tank Temperature (*C) PH . Sludge Volume (ml/50rnl) % Solids (w/w) Ethane* (p/L) Sugars (g/L) lArarjinose Glucose Xylose Galactose Mannose Total Hexose 33 9 34 S 37 6 36.6 328 33.3 344 33.4 34,1 34,7 346 34.8. 38.1 32.5 3.58 3,15 343 3,47 3.45 354 369 3.55 3.40 3.19 320 322 326 3.03 0 05 0.04 006 0.04 0.04 0,04' 004 0.05 004 G.04- 0.03 0,03 0 03: 0.04 • 18*52 19.60 2016 •20.79 21.14 21,22 21 30 20.96 21.39' 31.35 21.17 21*51 2158 31'.35 0.10 0 07 0.10 0.04 0.04 0.11 002 003 003 0.03 092 :6.rj6 006 0.04 0,72 079 098 1-03 1 oe 1.06 1 08 hll 096 i:03 109 1.01 0 87 V05 3 20 3.50 386 4.41 4.45 4.31 4.53 •4,88 4,18 445 4 67 3.99 4 37 4.53 1366 13.30 15,10 15.21 15,77 16.87 17.67 18.97 .16,70 17.53 1753 1601 15.94 16.37 1,74 1.72 1.86 2.16 2.19 3.05 2.22 3.23 2.14 2.11 2,14 1:98 199. 1,99 5 20 4,78 8 53 5.92 •578 9.36 10.10 .10.58 851 6.65 9.11 6.32' 833 9.10 .10,14 io.ob 14.27. 12,49 12.42' 15.62 1685 .17 69 1483 .15.21 1592 • -14.29 14.59 15.61. 29 0 268 27.6 38.1. 37.9 286 28.4 37 6 279. 28.0 260 280 28.0 28.1 262' 357: 279 MED MED MED'. ME D-»MIN mu MIN MIN .MIN MIN -»MEP MED .MED MED MED MED MED 'MED. MED •S'IO 502* 5.15 504 S.03 5.08 502 5.00 5.DQ 4.92 5 02 4 93 '-4.3l' 493 . AM 4.94 492 72 :* 78 8 73.5 79.5- 59.4 612 36,2 663 605 609 645 '57.1 '58.5- 585 .50.1 361 35.9 6.5 35.7 8.9 23 8 16 8.5 18.9 17,7 33.1 14*4 22.1 16,4 18.8 185 17.4 > 120 12.4 0.09 0.17 0,19 0.55 0.32 0.20 0,31 0,13 0.10 0.96 1,45 1 87 1.33 1 43 i!21 0.38 0.85 .0.13 6.'21 0.26 069 0.54 033 0.86 02O 6,17 1.57 2.24 3.27 2.27 3.44 .2.4*1 1.04 SIM 0.04 008 0.18 0.15 0.11 0.13 0.17 013 0.13 030 0:38 040 0.33 026 O.X 0.19 0,29 15.80 1578 1555 W62 18.64 1978 20 .'52 31.10 31.19 18.25 17.28 18.31 18.41 17.'58 20'.44 3094 21^ 23 0,07 6.86 290 1 82 2.84 3 33 3.81 320 1.91 3.32 4'.06 4.48 4.05 376 3.42' '29t' 1.88 0,00 0.00 0.00 0.75 0,78 094 1.00 10! 1.00 1.02 0.93 0,75 0:47 0.47 0,40 0.39. 0,39 2.96 161 000 1 06 051 1.10 1,04 1 67 1.62 0.59 0.00 000 0.00 000 000 QOD 0.60 1059 1034 10.14 12,90 13,13 1367 14.31 14,43 15,77 15.12 15.65 13.70 1472 13.77 12.63 io;i_s 1l',70 •1.52 1 49 i;43 000 154 1,54 Vei 1.97 3.C3 1,92 1.93 t.75 1 61 1.43 U t 1.39" 0 85 1.28 5,00 9.49 3.96 000 000 298 2.87 359 6.66 2.33 0.00 000 0.41 000 0.67 000 2.t4 706 1.43 260 2.25 589 5,88 728 1022 4,75 1.75 161 1.84. 1-31 2.06' 08S 4,02 0.98 13.0S 40.73 25.56 39.88 4522 39.46 3090 26 82 46.68 57.03 62 94 5691 53.83 46.03 4087 36.40 3.06 34 43 4597 31,41 47 55 78.22 68.18 64.56 10am 7073 65 20 71'.14 65.55 £623 56.34 '43 52 37.08 0.44 5.44 16.36 1152 17.98 20-38 17,79 13.93 13.09 21 04 25.70 28 37 25.65 23.81 21.65 18'.42 11.90 30.2 29.5 30.9 29.6 283 269 292 28.6 296 292 295 29.5 25.9 38.4 MED ->m 1 MIN ••MED MED MED ->MIN MIN MIN-»MED MED MED MED MED MED MED MED MED 504 4.97 5.00 505 4.99 5.03 -497 4.97' '485 -4.80 4,78 4 90 4.96 4.82 iFarmenW #1 Temperature (*C) lAeratlon Veast Concentration % viable %'Buddthg viaae (g/L) Total (p/L) Sludge Volume (mi/50mi) %'Solid3',(w/w) Ethanol.(a/L)' Sugars (g-t) A/abinose Glucose Xylose Galactose Mannose. Total Hexose Fermentation Efficiency (total rex) Fermentation Efficiency (hex'used) Fermentation Efficiency (all sugar) Ftrmentw#2 Temperature ("C) Aeration P* v'east Concentration % viable % BurJjIng Viable (g/L) . ™ a i . (o/L) Sludge Volume (ml/50ml) % solids (wftv) Emantf ig/L) Sugars, (gl) Arabinose Glucose Xylose Galactose Mannose Tota Hexose [Fermentation Efflclercy (total hex) Fermentation Efficiency (hex used) Fermentation Efficiency (all sugar) fermentation Efficiency (sugar used) 3848 43.99 17*34 34 37 51:54 S7.98 '2323 43.31 097 0.43 14.93 53 38 6364 41 85 72.93 37 03 77.13 9.11 37.64 108.31 •'16 96 107.13' 55 SO 64 05 35.03 56.29 3.30 0.39 17.39 0.96 0.00 15.69 66.38 75.07 29.93 72,96 75.74 83.53 34.14 70.75 0,44 000 12.24 65.01 71.55 29.30 53.32 60 97 66.32 27.48 -53 84 •o.oo 12.42 .1.16 54:49 59.43 5294 3304 33.63 0.30 3U0' 35.95' 39.06 16,21 27.41 Surg* Tank PH Sludge Volume (ml/50ml) %'Solirj3(wM') Ethane* (g/L) Sugars'(gl) Arabinose Glucose' Xylose Galactose Mannose. Total Hexose Theo Max-EtOH (hex in settling tk) TheoMax EtOH (hex used) 5.04 4.98 5.03 492 4.88 462 4,75 477 4,68 4.96. 4.80 0.07 0 07 0.05 0,14 O'IO 0,13 0.12 0,14 6.16 0 04 6,0? 19.96 20.86 21'.'05 14 21 '1209 1535 15.00 17.36 19.79 .20.89 21.04 3,70 305 2.78. 362 4 44 520 4:56 ' 433 3.83 321 251 1,05 106 0.94 099 1.15 0 81 0.44 0,31 0.'40 0.40 0.35 0,42 095 1.14 000 0.00 000 006 0.00 o'oo 000. 6.00 1490 14,79. 15.62 1446 1558 1405 1306 12.70 13.13. 10.75 10.78 1.95 1 95 1.38 1,74 1.60. 1 40 1:34 too 1.21 060 0,92 6.00 2,41 5.77' 000 0.00. 000 OOO 0.00 OOO 000 0,00 3.37 5.31 689 ll74 •1.60 1 40 i'l34 1 00 t.21 080 0.92 7.12 7.12 7.12 7.12 7.13 7.12 7,13 7.13 7.12 7.12 7.13 7.12 7.12 6 33 4.36 5.66 409 2.47- 6.17 6.30 6 46 6,47 655 6.53 688 6.55 APPENDIX C. RAW DATA FROM BENCH & PILOT PLANT TRIALS C.7 Pilot Plant Trial #3 - An Extension of Pilot Plant Trial #2, Blended SSL Date 1 0 : O c t - 9 8 11-OC1T9B 12-OC1 :98 13 -Oc t -98 14 -Oc t -98 1 5 - O C I - 9 8 16^0ct-98 1 7 - O c t - 9 8 1 8 - O C t r 9 8: 19-Oct-9B '.2Q;Oct-98* Time 10:45 'AM 10lOO 'AM 11:30AM 2:15 PM i 0 :00 AM 1 : 3 0 P M 9:30 AM 12:30 PM* 4100 .PM' 10:00 A M IO :6O'AM" #'Hours into Trial 0 23 48 .5 75 .5 •95.5 "l 23 143 1 7 0 . 1 97 .5 215.5 239 :5 # Days'Into Trial bo 1.0 2.0 3.1 4 .0 5:1 6:0 7.1 i9:2- ao- -10 .0 S e t f R o f r i i s _ 1 s . : '. : .—.—: Feed'Rate (L/min) 115- 1.5 1;S = » 2 .0 2 :0 => 5.0 5.0 =*--2.0 2.0 2.0, 2 .0 2.0 2.0. 2:0 Tempera ture(oC) 28 .0 .28.0 28.0 28 .0 2 8 . 0 . 28 .0 28 .0 28:0 28 :0 2 8 . 0 :28.o • bH 4.9 4 9 4 .9=i ' 5 ;0 5.0 5:0 .5.0' 5.0- 5*0 5.0' -5.0 5:0* O b o r a t i o n s ' ' . . . . • : •£ . , i _ — i — u —I Storage'Tank Level .(%) 85 82 78 57 60 ' 5 3 46 41 34 2 S = » - 2 2 = V C 86 (Hwd) 515 Sep Vibrations (mm/sec) 4.0 5:0 4.5 4:5 4.-4 •5.2 4.7. 4 .6 5:8 5:5 Enrrntint a tin n _ _ _ „ _ , • • -- -SSLin 2 . 3 6 ' p H ' 3.21 3.82 3 :89 3 .70 2.63 3 . 8 2 3.62- 3 .72 3 . 7 0 3 7 0 ' SludgeVolume (mV50ml) 0 .05 . 0.1 6 0.06 p:0 4 0.'1 5, 0. i 0 0:06 0 . 0 5 . 0 :05 . 0 .08 .0:05-' %.Solids (w/w) 21.35 20 do 19:38 18 .13 1 8 . 1 2 ie:is> i 8.1 3 18^ 20 i a : i 7 1 7; 98 1 7 .92 Sugars (g'/L) 1-.62' Arabinose K O I .0 .97 •1.34 1:47 1.65. 1.56 t;57 1.57 1 6 6 Glucose 4 . 2 2 ! 3.68 4 :69 4.-37 s.ob 4*; 2 7 4.47 4 . 4 7 - 4": 9 2 5 :p5 , !4."85. Xyjose' 15:30 . 1 1:23 9:68 7 :08 ' 7!41 6 .80 6 .53 6153 7. 26 7:21 '6:6 a; Galactose' 1.99 3 ; 3 2 : 3:33 3 .50 4.11 3 .73 3 .95 3:95 3.97, 4.08 3 . 8 2 ' Mannose' 8.83 9.20 12:45 12.70 14 .04 ,1 2.82 11.01 1.1.01: •1 4.1 4' 14.24'; 13 .85 Total,Hexose 15:64: 1 6.20 20:47. 20 i57 2 3 . 1 5 ^20:82 •19J4 3 -19143 23 .03 23:37' 2 2; 32 Settling Tank. Temperature (bC) . 3 8 . 8 i 34,-1 34.4 33.4 39:1 39:0 3 4 . 7 3 5 : 5 ' 3 4 . 5 . 37.; 0 37,0 PH 3.19 3.68 3.81 3 7 0 3.01 3 .7 4 3 :78 3:73 3*69 '3.'68" 3".69. Sludge; Volume(ml/50rnl) 0:02 .0.03 0.05 0.04 0.04' 0 .04 0:04 rj.04' 0 .04 0 .04 • ,0 :04 . % Solids,(w/w) 21.31 21.21- 19.61 18.39 1 8:1 2 .18.37 * 1 8:1 B 18:1 6 .18:18 '0.00 ' 18.25. 1 9;1 2 Ethanol (g/L) 0 .03 0.02 0.02 0.01 o"02 0 .03 p:02 0.00* "p. 01 :oYj3; Sugars (g/L) 1':68 1:67 .1i68* Arabinose • 1.06 .1.05 1.65 1:65 2 08 1.46 1:60 1.60 Glucose 4 .29 4.i'l 4.'4 3 4'. 6 3 4 : 6 8 ' 4. 'I8 •4'.52 - 4 : 5 2 ' 5:02 5:00 ,4.'93 • Xylose Galactose 15.68 12.63 10.39 7 :86 6 . 6 0 ' 7 : 0 9 7.61 7.61 7 .36 7.56 7 ; 3 0 1.99 •2 .38 3.1.6 3 ' 7 5 3.97 3 .43 3 .96 3 .98 3 . 9 2 4'.01 i;4*5.1 3.96. Mannose a.i 2 ,9 :B6 12*02 I X S ? 12:38 12 .06 . 12.21 12.2 V 14:49 i '4.37 Total Hexose 1 5.40 16 :35 19.61 . 21 .95. 21.03 19,67 20.69 20:69 23:43; . '23 .52. 2 3 : 2 6 Fermenter # 1 27! 7 .27.9 Temperature (oG) 28.1 ' 28 .1 2 7 : 9 28 .4 •.2B:5 2 8 : 0 29.1 27 i9 27 :9 Aeration M E D M E D . M E D r » M A X . M A X - > M E D : M E D , M E D M E D M E D . M E D M E D ,M E D PH"' 4 .99 5*03 4'. 94 5.04 ' 5:0s 5 .03 5 :05 '5.05 5.0V 5:04- 5*00; Y. eastC o n ce'ntrati on 2712 2 2 6 .% Viable 29!6 33.4 22 :4 . 31 .7 35:2 2 8 . 5 22 .5 2 5 . 8 ; 25 .9 . .% Budding 2:5- 6:3 I'41-I 11 .2 19.4 11 .6 11.2 -11.7 "9.6 8:1 9 ^ Viable- (g/L) 0.81 1.27 0 . B 5 1,70 1.91 2 .16 2 .06 1-.79 r .88 6 .9 V 1 7 2 - V j 7 8 : Total (g/L) Sludge Volume' (rnl/50rril) 2.7 4 3.80 3 .79 5.36 5.43 7; 5 8 9; 16 6 9 4 6.64 7.88, 0.'40 '0:3.9 6.41 °- '46. 0:55 ,6.72 0:7 7 '0 .61 ,6';'eo 0:68^  ' .0.67. i '7 ,65 % Solids (w/w) 21.12 20.92 20.34 I 8,81 1 8.25 1 7.84 1 7:63 1 7 , 6 6 '1 7.60' I 7.54 Ethanol (g/L) 1.82 1:97' 2 .36 1.90 3.94 7 .42 7 .19 7 .00 6:45 6.18, 6.61 Sugars'.(g/L) Arabinose 0.40 0!28 0.31 0.B9 0.B2 0 .23 0 : 2 5 ' .0.00 0.*26 0 ;25 .0123, 'O.'OO' Glucose 0.42 6.00" OlOO 1.76 0.96 O.'OO 0 .00 '0.00* ,0.00 0 .00 • Xylose 11.22 B 7 5 6 :47 5;is 5'.33 2 :92 1 .4 1 0.78 '1.52 1.05 0.65 Galactose' 1 i 3 3 1.40 1:94 2:93 3.50 2 .38 2.47 1.28; 2 !89 2":90 2:86 Mannose 1.60 0.00 3:88 7 .99 6.82 0 .00 o-od . 0 . 00 2.; 04 ofrjrj o.'oo Total.Hexose 3.35 1.40 5:82 12.68 11,28 2 .38 2:47 1-.28 4'. 9 3 2,90 ,2:86 Fermentation Efficiency (total hex) 22.66 2 4 . 5 3 •29.39 17.35 35 :98 67.7.7 6 5.' 67 6 3 : 9 3 58.91 76."47 * '56.'44 65:26 60. r 37 6 9 . 6 5 4 2 . 4 0 Fermentation Efficiency (hex used) 28.79 26 .92 46.61 42!39 7 5 . 8 2 76 .22 74.20 67.99 Fermentation Efficiency, (all sugar) 11.71 12J67 15:19 12.19 25:27 47 .60 ' 46:1 2 4 4 :90 •41:38 ' 39.64 Fermenter #2 31:1 ' 3 0 : 9 ' . M I N Temperature (oC) Aeration 29,'1 30 :2 29.6 30 .2 . 30.-7' 30 ,7 31.1 31 .2 32.1 M E D M E D M E D M E D - M E D - - M I N M I N M I N ;MIN M IN • M IN pH 4:78 4.79 4 . 7 0 4 .79 4.75 4,74 4 :70 4.74 4 :74 4:76 •4:78 Yeast Concentration 15.3 2B.0 22 .9 2 2 . 9 . % Viable 36.4 32.3 17:9 3 X 8 36:8 2 8 . 5 21-.6 VoBuddirig 4.7 6.8 610 9:9 16:8 9:4 9:8 ' 11 .'4 i9:7-' 8:8 *.1':'*J'/' Viable-(g/L) 1,'IB •1.U3 U.ti3 1:01 1.7 3 1.81 1.64 1.5H 1.85 V.37 Total (g/L) 3.1 9 3.19 4164 5.36 4.70 6:35 7 .59 10:33 6:61 5.99 .5.98.-Sludge.Volume (rrt/50ml) 0.-32 0.35 0 .42 0.44 ' 0.58 0 .67 0.'7 3 0.57 0 .55 o.;ee 0.66 t 7.54 %':Solids (w/w) 21.04 20.83 20 46 ia:s6 18:12 • 17.71 1 7 t 4 7 1 7 , 4 8 1 7.54' 1 7.42-Ethariol'(g/T_) 2.34 2:41 3.04 2 ,80 6:02 7 . 6 0 7.54 . 7 7 8 6'. 97 ' 7.l '4 7.-54; Sugars (g/L) Arabinose 0.37 0:33 0:36 0 .33 0.00 0 .22 0 .26 0:00 0 .26 0 .25 . 0.Y2 • Glucose 0 .00 0.00 0 .00 0.53 0.00 0.00 0 60 0.00 0 . 0 0 0.66 0.00 Xylose Galactose 1 022 1.12 8:72 0,96 6 .42 1.71 4.34 2.51 '4:05 3.14 1 .Bfi 2.;20 n:oo 2 . 1 3 0 . 0 0 ' 1.59 0 .00 2.47 0:00 2 : 6 0 -'0 .00 ,2.49 M a nnb se o:53 0.00 O.'bO 5.'.1 4 2:20 0.:o6 0 .00 0,00 0:00 0 .00 •0.00-;. Total Hexose 1.65 0.96 1.71 8 ,18 ,5 :34 2.20 2 .13 1.59 • 2.47 2;60 ' 6*5.21 74 :20 45 .80 50,51 2.494 Fermentation-Efficiency.(total hex) RermentatJoh^Efficiency (hex used)* 29,1 4 3 2 . 5 5 30; 01 3 1 . 9 6 37.B5--42.46 26,39 42.64 54'. 98 ' 7 3 . 1 8 69i41 •77:34 68 .86 76 .45 7 1 . 0 6 7 6 : 7 4 6 3 . 6 6 -71.93 7 7 :90 48 .37 53^07. . Fermentation Efficiency:(all sugar) . 1 5;05 15J50 19.56 18.54 3 8 . 6 2 4B .75 48 .37 49:91! 44;71 Fermentation Efficiency (suqar used) 25:15 23 .08 .27.11 3 1 . 9 8 : 55 .74 .56 :69 52.'4 7 5 2 . 6 5 .49.1 0 Surge Tank ,4:76 4.78 PH Sludge-Volume (ml/5pml) % Solids (w/w) 4.74 4.79 4 .68 4,7 6 4 .75 4 .74 4 . 7 0 4 .73 4 .74 o.oa; 1 8.44 '.0:lY 20.73 0 .12 20,44 6 .25 10.14 ' 0 :26 16.80 .0!26 17:57 0 .29 17'. 37 0.21 1 7 . 4 4 ,0*24 1 7;51 6.30 17:34 6 ,30 ' 1 7 , 4 0 7.40 Erhanol'tg/L) 1.96 2:36 2 .82 2:95 5:70 7 .25 7.-31 7.60 6.86 7-. 08 Sugars (g/L)" Arabinose n:-34- 0.39 0.31 0 .28 0.00 0 .24 0 :26 0.00 0 .25 0.24 0,21 Glucose O.OQ . 0 . 0 0 - OiOO 0,45' .•0.00. o.'oo 0 .00 .0:00 o.'oo 0 . 0 0 . 0.00 o.'qp'' Xyjose Galactose 9.58 9.4 7 5:79 3.97 3 . 5 0 ' 2 :15 0.00 ' 0 .00* ,6:00 0:00, 1.04 1.25. 1.55 2:58 2 . 7 5 . 2 .53 2 .05 2.07 2.49 2 .38 2.39 Mannose 0 .00 0.00 0-00 4 .06 .1.-75 0.00 0 .00 0 .00 0 . 0 0 0 .00 0.00 : Total Hexose •r.D4 T.25 1.55 7:09 4:50 2 .53 2 :05 .2 .07 2.'48 2 .38 '.2.39.' 1 0 .95 . 9 ' 6 8 . . Theo.Max EtOH (hex in.settling tk) 8:03 8.03 8 .03 10.95 10.95 10.95 ' 10:95 10:95 10:95 ' 1 0:95 Theo.Max EtOH (hex used) 7-.J9 .7:54. 7.16 6 .78 . 8 . 2 3 : 9:83 .9 .86 1O.'l 4 9:69 9J62 APPENDIXC. RAW DATA FROM BENCH & PILOT PLANT TRIALS 225 C.8 Pilot Plant Trial #4 - An Extension of Pilot Plant Trial #2, Blended SSL Date '20-Oct-9B' •21-Otl-9S' 22-Oct-9 8 23TOC |:9B" 24-00-98 25TOct:98 26rOct:.9B 27-0 et- 98- 2B-Oct:.98 297Oct-98 3 0-.O 9 9 Time # Hours into' Trial ' 1 0:D0 A M 10:00 .AM- 10:30 AWi: :'i0:00 A M io:6o'AM 9:00 AM . 1:30.PM 1 0:OO':'AM 10:00AM: 9:30JAM; 9:30 A M ~ .'24 48:5 7*2 • 96 120 148:5 168- 192 21 5*5" •239:5 # Days.into'Trial o.o 1.0 •2.0 3:0 4.0 -5.0 6.2 7.0 B.0- .9:0 'mo §BtlR6fritB> .. _ , . M — B .—,—. Feed Rate (Umin) •2.0 2.0 2.0 2.0 2.0 .2.0 2.0 2.0 .'2.0. 2.0 2.0: Temperature.(oC) 26.0 2B.o; 28.'0 28.0 28.0 2B:O 28.0 28:o; 28; 0 :28,0. . 28.0 pH •5:0 5.0. :5.Q' 5.0 5:o; 5.0' '5:0 5.0: i5:0? ySIO 'Sfb Oberatfonsi • ' ,. ,, ,, - — — Storage TankUevel. (%) 86 78 72 65 58 53 47 43 38 33 V Sep'Vibrations fmrn/seh) 5:5 5.6 5.5 518 5; 5 5.5 5:7 6.0 .5.8 6.3 6.2-r r v > i n n i a i t r s n : _ " •• * ' -.. SSLin PH 3.70 ,3:57 3.:70 3:ee 3:83 3.99 3.70. ' 3:65 3:59 3.58: ,3:61 Sludge Volume (rrn750rnl) .0.05 0.04 0.08 .0708 0.10 6.08 0.04 "0,03 0.02 .pro 3, "p. 0.3 %.Solids'.(w/w) 17.. 9 2 :20:25 22:20 22:48 •20.44 20:27 20.06 :20,09 18:99' 20:17. 19.84 Sugars (g/ll) "1.25 Arabinose .1.49 1.63 1-.37 •1.46- 1:37- 1.-16 1.36'. ' 1 :.3i - 1-.25 '1.2 6"' Glucose 4:65 6.53 6; 16 '6.49. 5:99 5.07 .6.24 6:02 5,'79 •6:06 sVa*3 Xylose' Galactose .6:68 9.33 8..16 8.90 7.95 7.01 8.37- 8:60 8.'48 •B.20; 8.10. •3.02' 4.67. 4:05 4.40 4:04 3:55 .4.04, 4.14 4:01 '3.08 , 3^ 9 3 Mannose, 1 3.8S 16.44 16:40* 18.14 16.65 14.47. 19.54 1 e;62 -16.87' 16.23 ,16:04 Total Hexose 22.32 2 7.'64 26,61 29:03 26^ 68 23,09, 2B 82 26,78 •26.67- 26 j 1- 25.80 Settling Tank -34.0 33,5 Temperature (oC) 36.6 3.4.6: 34.5 37.0 37.2 35:3 •33.1 ' 34.7 36.'7 pH 3.69 3.60, 3'. 6 8 3.84 '3.81 3.92 3;73 • X66 3/60 •3.58" • 3.59 Sludge, Volume'(ml/50mij 0.04 0.05 0.03 .0.03 0.05 0.03 0.03 0.03 0.02 6.03 0:03' %Solids (w/w) Ethanol'(g/L). 18.1 2 21.32 22.04- 22;47 20.88 •20.48. 20.21 20:13 •20:00' 20:16 20:01 0.03 0.00 0.02 0:02 0.02 o'oe .0.03 0.02 0.01" 0.02* • 0:02 Sugars. (g/L) 1.2B Arabinose 1.68. •1.51- 1,50 V-.37 -f;27. 1:23 1c32' 1,3V 1':15 1:36 Glucose 4:93 6.'49' 6.^ 79 .6:73 5.77 5:51 6.00 • 6:10' 6:06 1 • 6:08 Xylose '7,30 a. 07 8.96 9.00 7.69 7:58 ,8:25' 8.40 8,47 .'8.05' ; B.39, Galactose • 3.98' 4*56 4.36 '.4.46' 3!90 3.87 '3.93 . 4.08 4.03 '3.90' 4.00 Mannose. 14.37. "16,44 '18! 13 18.23 'l6.~0B 15:6V 16:25 16.63 '16.37,. 16:06 ,VB178 Total Hexose 23:26 27,49 29.28 29.-42 25.73 24.99- 26.18 '26:81 26.4'6> 26:07, : 26.34 Farm •nter ft rl: Tempe rature'(oC)' feration H ' ea'st Concentration % Viable % Budding Viable (g/L) Total (g/L) Sludge Volume (rhl/50ml) "Solids'(w/w) Ethanol.(g/L)' Sugars (g/L) [Arabinose . Glucose Xylose'-balactose Mannose-Total Hexose Fermentation Efficiency (total hex) Fermentation Efficiency (hex used) Fermentation-Efficjency^aj^uo^ ,27,9 27.9 27.8 .28.1 27.8 28.'1 28.1 28.0 29!0 ,28:2 284 M E D , M E D M E D M E D M E D M E D M E D . M E D M E D : M E D . :M E D 5.00 5.04' 5.10 5,04 5.20. 5.1 I 5.01 5.02 5.02 •5.01 • 5.01 '22.6 20,4 21.9 is:4 21.2'' 1 7,4 11.0 -. 10,5' 12:2 14',6 •is:6 9:0 7.9. •8.2 3:4 8.5 10.0 •16.9- 15.0 30ll :22:1 • 18.8 1:78 1.26 1.22 1.20 1.29 1:10 • 0.83 0.84' 0,73 0,95 1.28; 7.88 6'.1 8 5,57 6.52 6,08 6.34" '6.97: '8.00 s:99 '6:51 elee" '0:81 0.87 •0,67 0:63 0.70 0.70 0,78 0.75 6.77 0:69 ,0,75. •17.6 5 20.72 20,66 21,87- 20*80 r20!20'. 19:91 '1 9'. 76 ;'l '9". 6 7 19:4 7 •1 9;25 6.61 6:29 6.69 6.55 6.80 8.05 7.65 7,06 5?73 :6:25 6:98 0.23 0.31 0.24 ;0.39 0,34 0.33 0.36- 0.261 0.28 0.27 0:28 .O.OO' 0.00 0.00 '0.00 _ 0:00 0 00 aoq", • 0,00 0.00 o.oq- •poo,'. 0.65 1,40, 1.87 2.92' 2.71 2.05 0.74 "0,00, 1:47 -1:89' : 2.08* 2.86' -3.38 2.92 13i1 S- -3:6o 2.95 2.24 2.40. 2.55 '2:31 1:84 0.00 2.89 2,17 ,2.87 2.80 2.'77 2:13 '1,59 2:93 M'.57 0.00 2.86 6.27 5.09 .6.05 5.80 5:72 4137' 3.99 5.48 3.88: 1:84' 60.37' ,57,45 46:44, 45,47 47.21 80.62 ; 57'.61 ;53I17 43:15 47:07- •5 2756 69:65 .8tVl5 56.65' 57.87 59.40 7,7.69" 69.23 .62,79 .54.65; 65.31 5^6.56 42.-40 •40:35 33.96- 33:25 •34.52 . 44156: 42:34 39.08 31^ 72 34?59 38:63 30.9 30.8 30.9 31'.2 31.0 30.'3 30:3" 31.0 32.'1 30.6 ,30.4 MIN MIN MIN MIN MIN MIN MIN MIN M IN MIN MIN '4.78 4.83 4.91 '4.91 4:98 4:91 •4.92 4.93 4.92 '4.96. " 4 . - 9 8 F«rm •nt»r#2: Temperature .(oC) |Ae ration pH Yeast Concentration' % Viable 22.9 22.6 .% Budding 8:8 8:6 Viable (g/ll) i.3V 1,52 Total (g/L) 5^.98 6,73 Sludge Volume (ml/50ml) 0.68 • 0.65 %:Solids(w/w) ' '7.54 18.83 Ethanol (g/L) 7-54 7,52 Sugars (g/L) lArabinose 0.22 0.31 Glucose. 0-00 O.oo [Xylose. 0-00 0:00 Galactose 2:49 2:9i Mannose p.po o;oo Total Hexose 2.49 2.81 Fermentation Efficiency (total hex) 68.-86 68.68 Fermentation Efficiency'(hex-used) 77:90 79.02 Fermentation Efficiency,(all sugar) 48.37 48:24 Fermentation Efficiency (sugar u'sed)' 53,07, 53,72 Surg« .Tank [pH 4.78 4-.81 Sludge Volume (ml/50ml) 0:30 -0.30 %:sblids(w/w) 1>.40 18,75 Ethanol (g/L) 7.40 7.40 Sugars (g/L) [Arabinose D:2i o:no Glucose 0.00 o.oo' [Xylose 0.00 o.oo Galactose 2.39 2.87 Mannose 000 0,00 Total Hexose 2:39, 2.B7 Theo.Max EtOH (hex in settling tk) 10.95 io.es The6:Max EtOH (hex used) Q.6Q- fl-52 24.8 9.0. l!55 6:25 0,60 20.32 8.00 0,31 0.00 0:00 2.89 OlOO 2.89 ;55.54'-61 .B7' 40.62 44.28. 21 J 7:9 1.26 .5.92 0.75 21,13 8:25 0.31 0.00 1.68 2.83 0.00 7,63 67.27 63:65 4V.88 47.86 6.0 1.1b 5:62 0:7 8 ,20.88 7.92 0.26 0.00 •r.39 2:66 0.00 . 2.66 54.98 60:70 40:21 45.26 .7*1 u.ya 5^ 28 0.72 20.38. 9.72 0.26 6:00 0, 78 2,45 o'.oo 2.45 73:20 80:90' 53.80' '59.68. 1 2.7 1 9.2 0.73 . !5.75 ;0:75 20,01 8:96' 0.32. 0.00 ;0:00 1:B6 0.00 1:86-67.4 7 72:66 49:58 52.85 8:0 10:3 U.50 7.251 - 0:78 1 9.69 .8.03 0.30 0.00 0:00: 1.87 o!oo 1,87 ,60.47 65:15 44.45 .47.35 11.3 15.9 6.69 19.49, d:e9 0.30 0.00 0:00 2.34 0:06 2.'34 ;51.69-:57;01 38:14-41.21 19:1 2410 1:21 '6.34. ;o:75. 19.40 •6.8B. 0.25 .0.00 v.26-:2.28 0:00. 2:28 51,66 56:62 37.97 42:52 -t8:s 19.B' 1.29 6:97-• o:eo, 1 8.99 7;ar. 0.29 0.00. 1.68 1.71 0.00 i-J-i 58.81 62.95 43.23 48.24 0133 •20.07 7.90 o;oo 0,00 6.00 2.86 0.00 2.86 14.40' 12.93. 4189 0:31 21.02 7.44 .0,00 '0:00 1.52~ 2.71 .0,00 .2,71 14:40 1 2.96 4.94 4,90 4.92 f.92 4,91 4.96; ' 4:fl7; '0:44' o':*! 0:38.. ''6.'45 0!'40 :0:52'' •^ 0:58': 20.59 20.29' 19.82 1 9;'57 '19'.44' 19^ 39 19,16 7.85 9.44 8.70 8.20 6.'74 .6.99 . 7.92 0:33 0.2B ;o,33 0:32 0:30 :o:251 0,26 D:OO 0.00 o'.oo. 0.00 o.:bo '6:'o6. 0.00 i.,39 0.84 •O.'OD' 'O.OO oise .1:.23 V".57 2.67 2.49 V.93 1.97 2.36 2.15 1.52 0.00 0,00 0.00 0,00 0.00 .0.00 0.00 2.67 2.49 •'1--B3 1.97. 2:36 '2.15 1;62 14.40 13.28. 13.20 13:28 13.28 13.28 ,13,28 13.05 1 2 03. 12.' 33 1 2.33 ' 12.09- 12:12 1 2.41 APPENDIX C. RAW DATA FROM BENCH & PILOT PLANT TRIALS C . 9 Pilot Plant Trial # 5 - LNH-ST on Blended SSL 226 Bate M-Nw-98 •U-N 0.-90 12-NOT-9B .13-N<w-98 M-Mov-9- 15-NOV-9B 164.QV-99 17-NOV-98 1B-MOV-98 19J4os«-9e . 20-MOV-98 21-NOV-98 22-N_v-98 23-MOY-98 24-NOV-90 -_5-NovflB 2e-Nov-.e Time 1:30 AM 1:30 PM 9 30 AM 1:30 PM 10.30 AM '10:30 AM 1:00PM 1:30 PM 1:30 PM 12:00 PM' 315AM :3:00 PM t;30PM. 3:30 PM •IftOO'AM' 9:00 AM " 900 AM #HbiiSlrto Trial 0 12 s31 60 61 105 131.5 ,156 1BO .202.5 222,75 '253.5 '276 302' •320.V 343.5 3675! # Days intri Trial .0X1 0.5 1.3' 25 .r.4 4*. 55 65 7.5 84 9.3 108 11 "5 126 13.4 '143 ' 163 Sat Aaiiff.c Feed Rate (L/min) 2D 20 •2.0 20 20 20 2.0' 2.0 2 20 20 2 0 30 20 Temperature f°C) 300 30.0 30 0 300 30.0 300 30.0 300 .300 30.0 30. 300 30.0 30.0 30.0 .30.0 30.0 PH ,50 5.0 5.0: 5*0 50 "5.0 5.0 ;50. 5.0 5.0: 5.0 50. '5.0 50 •S.0 50 15.0 Obvations} 1 -". j_ Storacp Tart. Level (%) t5(SwJ) •71SW1J ,B5(Hw.) 60 77 7t 65 sa 51 41 97 (S*d) 75 62 56 52 *6 Sep.vibrations (mm/sec). 53. 60 5 5.2 :5 2 5.1 '5.4 6.0 5.3 S.A 53 !6.0 5.7 •'5.8 Fermentation^ " . _ I I - ' : SSL, pH 3.35 4.11 4.50 '4.57' 4.30 3 76 3.85 3.77 4 03 :4'12 * 0 0 3.69 :3.94 3,86 Sludge Volume (ml/50ml) 030 0.15 0.10 0.05 0.04 .0.04 0.06 •0,05 0 05 0.05 01)6' 1973 OOS 0.02' 003 19.7 % sollas (w/w) 20.23 20 32 20.06 20.3 20.38 20.32 .20.58 20.4 2174 20.37' 1956' 19 97' Sugars {g/L) 1.54 A/aninoae' .1 37 1.47 1 31 1.25 1.28 '1.22 068 ,1,11 159 1.59 131 1.49 1 45 Glucose .434 4,31 V47 4.56 '431 4.37 4.46 4.2 637. 66 563 6 26 635 6.49 xylose 17JS0 .17.10 20.30 20.64 20.38 2021 21.24- 2032 .1079 8.79 956 674 '673, 7.09 Galactose 2.71 2.39 2 36 2.12 212 •2.CB 202 2- 4 24 '4.43 3.94. 4 •4 02' 4.15 Mannose 11.29 9.50 10.14 9.26 e.94 8.92 io.1 8 56' 1643 18 02 14.4 16.83. 16.93 19 08' 1864 16,20 1637 .15.94 1537 15.25 1660 14 76 27 04 29'2£ 2337 27.09 2730 29'.72 Battling Tank' Temperature (*C) 22.3: ,29.0 31 3 33,2 32.2 32,5 .33 7 372 36.8 31 2 29.1 30 26.8 21 .1 Sludge Volume (mi/5Oml) % Solids (w/w) 3.45 4 .X 4.42 4.53 436 3.68' 3 72- 355 4.09 3.97. 3.74 •3 64; 3.86 005 009 0.04 004 004 004 004 •0 03' 0O3 005 004 0 02 0O2_ 002 18.69 19.76 20.25 20.42 .20 25, 20.39 20.6 20.43 .21 65 20 68 '1996 19.64. 19.66 19.56-O'A Etnanoi {gi) 006 0.04 003 0.05 003 '0.06, 0.02 0.03 0.064 0.58 0O6* 0.07 0.07 Sugars'^) 1.45 V.51. Arabinose.- 1.31 1.28 1.31 1.33 1.32 •1.18' 0.7S 1:11 1*414 * :i59 1.49; .1.5 Glucose 4.65 4.38 4 37 •4.54 436 4.17 4.43" 4.18 5.75' 6.72 503 • 6.44 637 • 6.63' xylose Galactose 15.55 1722 19.11 20.28 20.82. 18 06 ,21.96 30.11' 1*1.13 ,9.36 76 7.19 6.99; ?!l 4 2.71 2.32 23 ,2.26' ,'9,23. 2,18 1.9 206' •2.06 359 4.33 4/17 4,01 4 08:- 4.05 Marircse 11 05 9.96. 9.36 9.17 • 5.97 10.57 6.56. •1433 17.51 1531 16*92. 17.19 18.78 Total Hexose 18.41 15.68 16.03 .(6*08 1S.71 12.04 .1 7.06. 14.80 :2367 28.56.' 26.01. 27. ,37 27 64 29.46' F«rrntnttr#1 .343 ' Temperature ("C) 32 0 .25.8 24.7 269 300 30 • 30.1 30 30.1 302 29.9 30 29.9 M.1 383 23.9 Aeration Max Max Mate tow Low LOW Low Med Med Med' Med M M Med 'Mod- Med Low Low PH Yeast Concentration 5.IS '5.05 5.05 .502 5.09' 5.D2 5.02 305 >5.0Q" 5.03 5.00 505 5.05 500 5'48 .5 23 5.24 47.02 .% Viable 78.5 753 77.4 91.9 83.7 783 63.5 56 51.9. 44.5 .38.8 332 36,7 396 43.6 41.8 ,% Budd.no 9.2 12.9 2.2 26.1 5.7 5.93 7' 124 73 10.5 7.12 166 176 16 S.S1 122 11.2 Viable' (g/L) o.oe 0.14 0,22 0.47 0.74 0.91 0S6 1 01 15 1 055 1.63 1.635 2.21 245 305 3 28. 4.02 Tots* (Qrt)' 0.10 0.19 0!29 051 0.68 1.16 1.35 1 60 3.06 3.74 4.36 5.53 6.02 6.19 7.16 '765 8.55 Sludge Volume (ml/50ml) 0^ 05 .007 0.13 0.12 0.21 0.26 .O.K 0.31 0,41 0.42 0.46 0.56 0.55 0.60 0.75 '•0.72i 0.66 % Sollds'fwA.) - |«59 14,57 '14.53 16.36 1755 19.03 20.12 20.23 ; 20.4 3 20.26 208, 20.51 '1955 19,32> 19 03' 9 06: 9.2' Etriaid(c/L) O.OS 0.41 1.25 3.27 6.SO 4 3. 3.53 302 5,04 3.93 1.91 5.2e 6.74 8.71 6.65 Sugars: {g/L) Arabinose 1.95 1.87 2.01 1 22 1.27 1.19 1,3* 1 31 1123 0.31 0.9 0 41 0.37 0.4 03 04 0.41 Glucose 6.75 '5,11 1.49 0.60 0.00 0 086 065 0 0 0 0 0 0 0 0' 6.38: 0 Xylose •20.24 20.28 18.54 12.25 13.25 15.43 18.76 1835 19.1 19.9 20 1561 11.53 8.38 675 586 Galactose 4.24 4.33 4.40 2.57 0.67. t 03 1.71 1.77 1.86 1.7 1.54 1 84 2.43 169 0.83 1.44 .0 1!88 Mannose 21.75' 21.38 •14.52 602 000 0B6 4'. 56; 466 0. 0 0 0 0 0 0 °. . Total Hexose 32*74 3032 ,20.41 9.19 0.67 1.09 7.13 7.50 1.68 1.70 1,54 1 84 2.43 169: 0.83 1 44 1:88 Fermentation Efficiency (total rex) 6.61 •5.01 15,27 39.96 79 43 52.54 •43.14 36.90 61.59 48.53 47.7B 6437 48.49 6266 62.23' 64.18; .65.18 ea.16 -i'ios, 4907' Fernierrtatiori Eiticlercy'friex used) -0.59 £ 4 4 •56,17 93.52' 92.89 5956 ,7753 69.29 6976 • 54.40 5265 73 60 S3 33 66.80 '4 6 '46 (8 32 Fermentation Efficient'(all suqar) 0 27 .2.19 6.S3 17 47 3472 32.97 16.86 16.13-. 2S''.92. .31.26 20.69. 28,10 35 95 F«rm«nt»r #2 Temperature (*C) 263 30 S 30.6 30.4 304 '30 a- 30.7 30.9 303 30.4 303- 384 24 3 23 9 Aeration Low. Low Low Low Med 'Med Med Med Low Low- Low Low Low pH 5.01 5.05 4,99 5.00 5.03 4,99' 4.98" 4,95 4.97 .5.00 4.95 5:08 • 5,16. 6:08 Yeast Concentration 42.53 37.98.' ,%! Viable 83.4 727 74.2 65 52.9 -we ,42' 37.6 33 3 34.9 38.1 42*65-% Budding 17.1 5,4 6*34 5 65 '11'e 97 ' ' : S 165 15.2 335 6,12 :8.67 16.16" Viable '(g/L) Total (gU oas 0.31 OSS 0.71 002 "1.47 ' 1.706. 0.04 1.73 2065 230 2.06 . 3 *!':?3. 091 0 43 1.15 1.09 1.74 269 406 2 87 5 37 •5.92 627 6:94 7.76, 6.03 Sludge Volume (ml/50ml) 0.18 0.21 0 2l" 023 0 31 '0.41 0>3 0,42 0.52 0.56 060, '19.67 0.65 19,2 0.76. 0.65 19.04 % Sollas (w/W) 1S.11 1738 18.95 19 75 1985 2021 20.49 20.19 3068 20 61 1 Etharial'(iyL) V.ia 7.34 5.1 4.22 3.54 58 4.62 3,77 5.18 6.75 854, 8,81 .9 68. 1033' Sugars {g/L) Arabinose ,1.17 1.23 1.2 1.33 M l 1.18 0.54 0.2S 0.46 0,42 027 0.23 0.67' 0.31 Glucose 000 0.00 0 0 04. 0 0 0 0 0 0. 0 . P. 5~99 xylose Galactose 1043 .1268 1S.62 1856 -2069 IB.51 19.52 19.07 15.76 1231. 75 6.2. 654: 1 84 034 098 1.36 1S7 1.77' 1.54 .1,36 1 49 2.18 108 0.31 0 97 ' 1.46 Mannose .1.37 0.00 0.99 3,33 4 3- -.0 0 6- 0 0 2AB 0' 0 0 31 ° v!« Total Hexose 321 0.34 1.97 5.79 667 1:77 1.54 136. 1 49 1 08. 0 97 Fermentation Efficiency (total hex) Fermentation Efficiency (riecused) Fermentation Efficlency^ all sugar) 51.20 S4O0 2238 09 69 91 £3 39 21 62.32 '1 04 27.24 51.57 60.68 22 54 43.26 74 03' 1891 7087 7366 30.98 564 S '62.45 24 SB. 46.07-50.33 70.14 6330 69.79 27 67 48.56 5276 36X10' 61.43 63.97 4555 63.36 64.-1 46.99 69 64 73.31. 51 63 66.41. 74 31 "78.53' '55.1O 37.52 64 09 56.44 75.04 87.45 74.59 -59.96 46.21- 53.46 60.57 5959 57.54 Surgt.Tmk pH Sludge volume (mfSOmi) % Solids (w/w) EtJiaricH (ga.). 5,05 4.98 5 00 503 4,99 4.98 •4.95 4 95 '4,98 433 5.07 S.16. 5.06 0.14 17 31 7.16 0.18 19 5,18 0.14 '19.B6 0 21 201 3.45 0.26 2022 5.63 026 .2038 4:43 0.2S 19.83: 3.63 0 31 20 65 5.34' 0.35 20:49 62 0.40 :t9.7i 337 0.35 1931 ess 051 1909 10 is 080 13.12 987 Sugars, (gfL> Arabinose 1.20 1 26 1,29 1.37 1.19 0.S2 0 27 0.44' 036 026 0 31 -031 031 Glucose 0.00 0 0.36 0 0 0 0 0 0 0 0 'o, 0 Xylose Galactose 12 44 15.3 16.55 20.02' ie:43 ,2036 19.56 16.74 1233 9,17 679 666 5 9 0.45 1.07 "l,53 1.9' 1,69' 1.65 1.33 1.5 2!l5 106 0.33 0.8 ;o: 1.42 Mannose 000 1.31 2*12 448 ,0" 0' 0 D 0 0' 0 0 Tola Hexose 0.45 2.38 4,01 6.38 1*69 1.65 1.33 1.50 2.15 ioe; 0,33 o".9o; '•42 fheo.Max EtOH (nex in settling'tk) Theo.MaxEtOH (hex used) 3.18 8.01 818 7.18 8.1 a' 5.23 e.t'e 4.78 '3.18 7.28 a.i8 7.40 6 13 7:49 6.18 7.42 13.90' 12.79 13.90 1335 13.90;' 13.74 13.90 1341 13S0' 13*16' APPENDIXD. DIFFERENTIAL PLATE COUNT DATA APPENDIX D. DIFFERENTIAL PLATE COUNT DATA D.l Bench Scale Trial #1, LNH32 on SWD SSL Table D-l Yeast Colony Count - May 28 Lab Fermenter Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX 10"6 .5 325 6.5 x 108 YPX IO"6 .1 66 6.6 x 108 YPX 107 .5 43 8.6 x 108 YPX 107 .1 10 1.0 x 109 YPD IO"6 .5 299 5.9 x 108 YPD 10"6 .1 66 6.6 x 108 YPD IO"7 .5 36 7.2 x 108 YPD 107 .1 12 1.2 x 108 SMA 10s .5 Too Many N/A SMA 10s .1 266 2.6 x 108 SMA 10"* .5 307 6.1 x 108 SMA IO"6 .1 76 7.6 x 108 Lysine 10"5 .5 11 2.2 x 106 Lysine 10s .1 2 2.0 x 106 Lysine 10"6 .5 1 2.0 x 106 Lysine 10"6 .1 0 0 Table D-2 Summary of the May 28 Yeast Colony Count Agar Plate Average CFU/mL YPX 7.2 x 108 YPD 6.6 x 108 SMA 5.4 x 108 Lysine 2.2 x 106 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA D.2 Bench Scale Trial #3, LNH32 on HWD SSL Table D-3 Yeast Colony Count - June 11 Lab Fermenter Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX 10s .5 52 1.0 x 107 YPX 10s .1 6 1.2 x 107 YPX 10"6 .5 6 6.0 x 106 YPX 10"6 .1 2 2.0 x 107 YPD 10s .5 56 1.1 x 107 YPD 10s .1 25 2.5 x 107 YPD 10~6 .5 11 2.2 x 107 YPD 10"6 .1 0 0 SMA 103 .5 Too Many N/A SMA 10~3 .1 500 5.0 x 106 SMA IO"4 .5 270 5.4 x 106 SMA IO"4 .1 53 5.3 x 106 Lysine 10"3 .5 0 0 Lysine 10"3 .1 0 0 Lysine IO"4 .5 0 0 | Lysine IO"4 .1 0 0 Table D-4 Summary of the June 11th Yeast Colony Count Agar Plate Average CFU/mL YPX 1.0 x 107 YPD 1.1.x 107 SMA 1 5.2 x 106 1 Lysine 0 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA Table D-5 Yeast Colony Count - June 15 Lab Fermenter Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX IO"6 .5 65 1.3 x 108 YPX 10"6 .1 22 2.2 x 108 YPX 107 .5 9 1.8 x 108 YPX 107 .1 1 1.0 x 108 YPD IO"6 .5 81 1.6 xlO 8 YPD ltr6 .1 18 1.8 x 108 YPD IO"7 .5 7 1.4 x 108 YPD IO"7 .1 2 2.0 x 108 SMA IO"4 .5 Too Many N/A SMA IO"4 .1 Too Many N/A SMA 10s .5 646 1.3 x 108 SMA 10s .1 127 1.3 x 108 Lysine 10^  .5 0 0 Lysine IO"4 .1 0 0 Lysine KT4 .5 0 0 Lysine IO"4 .1 0 0 Table D-6 Summary of the June 15th Yeas t Colony Count Agar Plate Average CFU/mL YPX YPD SMA Lysine 1.3 x 108 1.6 x 108 1.3 x 108 0 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA D.3 Pilot Plant Trial #1, LNH32 on SWD SSL Table D-7 Yeast Colony Count - July 30th Lab Fermenter Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX 10s .5 125 2.5 x 107 YPX 10"5 .1 32 3.2 x 107 YPX 10'6 .5 27 5.4 x 107 YPX ltr6 .1 13 1.3 x 108 YPD IO-5 .5 100 2.0 x 107 YPD 10s .1 48 4.8 x 107 YPD 1 Q - 6 •5 25 5.0 x 107 YPD IO"6 .1 8 8.0 x 107 SMA 103 .5 Too Many N/A SMA 103 .1 Too Many N/A SMA IO"4 .5 Too Many N/A SMA ltr4 .1 302 3.0 x 107 Lysine 103 .5 0 0 Lysine 10"3 .1 0 0 Lysine IO"4 .5 0 0 Lysine IO"4 .1 0 0 Table D-8 Summary of the July 30th Yeas t Colony Count Agar Plate Average CFU/mL YPX YPD SMA Lysine 3.7 x 107 3.9 x 107 3.0 x 107 0 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA Table D-9 Yeast Colony Count - August 17th Pilot Plant Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX IO"5 .5 160 3.2 x 107 YPX 10s .1 46 4.6 x 107 YPX 10"6 .5 38 7.6 x 107 YPX 10"6 .1 9 9.0 x 107 YPD 10s .5 265 5.3 x 107 YPD 10s .1 49 4.9 x 107 YPD 10"6 .5 46 9.2 x 107 YPD IO"6 .1 22 2.2 x 108 SMA 10"5 .5 271 5.4 x 107 SMA 10s .1 33 3.3 x 107 SMA 10"6 .5 37 7.4 x 107 SMA 10"6 .1 10 1.0 x 108 Lysine IO4 .5 44 8.8 x 10s Lysine IO"4 .1 24 2.4 x 106 Lysine 10s .5 13 2.6 x 106 Lysine 10s .1 9 9.0 x 106 Table D-10 Summary of the August 17th Yt ;ast Colony Count I Agar Plate Average CFU/mL YPX 5.1 x 107 YPD 6.5 x 107 SMA 5.4 xlO 7 Lysine 8.8 x 105 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA D.4 Pilot Plant Trial #2, LNH32 on HWD SSL Table D-l l Yeast Colony Count - September 22nd Lab Fermenter Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX IO"7 .5 29 5.8 x 108 YPD IO"7 .5 34 6.8 x 108 SMA IO"7 .5 30 6.0 x 108 Lysine IO"6 .5 0 0 Table D-12 Summary of the September 22nd Yeast Colony Count Agar Plate Average CFU/mL YPX YPD SMA Lysine 5.8 x 108 6.8 x 108 6.0 x 108 0 Table D-13 Yeast Colony Count - September 25* Pilot Plant Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming L YPX 10s .5 34 6.8 x 106 YPX 10s .1 12 1.2 x 107 YPX 10"6 .5 5 1.0 x 107 YPX IO"6 .1 2 2.0 x 107 YPD 105 .5 54 1.1 x 107 YPD 10s .1 9 9.0 x 106 YPD 10"6 .5 9 1.8 x 107 YPD IO-6 .1 1 1.0 x 107 1 SMA 10s .5 48 9.6 x 106 I SMA IO"5 .1 5 5.0 x 106 APPENDIX D. DIFFERENTIAL PLATE COUNT DATA SMA IO"6 .5 10 2.0 x 107 SMA IO"6 .1 0 0 Lysine 10"4 .5 0 0 Lysine 10"4 .1 0 0 Lysine 10"5 .5 0 0 Lysine 10"5 .1 0 0 Table D-14 Summary of the September 25' h Yeast Colony Count 1 Agar Plate Average CFU/mL YPX 6.8 x 106 YPD 1.1 x 107 SMA 9.6 x 106 Lysine 0 Table D-15 Summary of the October 10th Yeast Colony Count Agar Plate Average CFU/mL 1 YPX 6.2 x 107 1 YPD 7.0 x 107 SMA 6.4 x 107 Lysine 5.4 x 10s APPENDIX D. DIFFERENTIAL PLATE COUNT DATA D.5 Pilot Plant Trial #3, LNH32 on Blended SSL Table D-16 Yeast Colony Count - October 20th Pilot Plant Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX 105 .5 259 5.2 x 107 YPX 10s .1 57 5.7 x 107 YPX IO"6 .5 32 6.4 x 107 YPX 10"6 .1 6 6.0 x 107 YPD IO"5 .5 138 2.8 x 107 YPD 105 .1 56 5.6 xlO 7 YPD 10"6 .5 18 3.6 x 107 YPD ltr6 .1 4 4.0 x 107 SMA io - 5 • .5 153 3.1 x 107 SMA 10s .1 58 5.8 x 107 SMA .5 29 5.8 x 107 SMA i o 6 .1 5 5.0 x 107 Lysine 102 .5 Too Many N/A Lysine i o 2 .1 Too Many N/A Lysine IO"3 .5 344 6.9 x 10s Lysine 103 .1 63 6.3 x 10s Table D-17 Summary of the October 20th Yeast Colony Count Agar Plate Average CFU/mL YPX 5.8 x 107 YPD 4.0 x 107 SMA 4.9 x 107 Lysine 6.6 x 10s APPENDIX D. DIFFERENTIAL PLATE COUNT DATA D.6 Pilot Plant Trial #4, LNH32 on Blended SSL Table D-18 Yeast Colony Count - October 30th Pilot Plant Sample Agar Dilution Quantity (mL) # of Colonies Colony Forming Units/mL YPX 10s .5 296 5.9 xlO7 YPX 10s .1 61 6.1 x 107 YPX 1(T6 .5 29 5.8 x 107 YPX IO"6 .1 4 4.0 x 107 YPD 10s .5 320 6.4 x 107 YPD 10s .1 67 6.7 x 107 YPD 10"6 .5 35 7.0 x 107 YPD 10"6 .1 7 7.0 xlO7 SMA IO"5 .5 265 5.3 x 107 SMA 10s .1 53 5.3 x 107 SMA 10"6 .5 24 4.8 x 107 SMA 10"6 .1 3 3.0 x 107 Lysine 10"2 .5 Too Many N/A Lysine io- 2 .1 Too Many N/A Lysine IO"3 .5 296 5.9 x 105 Lysine 103 .1 79 7.9 x 10s Table D-19 Summary of the October 30th \ east Colony Count Agar Plate Average CFU/mL YPX 5.9 x 107 YPD 6.7 x 107 SMA 5.1 x 107 Lysine 6.9 x 10s 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0058762/manifest

Comment

Related Items