Biocatalyzed partial demineralization of acidic metal sulfate solutions by Robert Michael Hunter A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering Montana State University © Copyright by Robert Michael Hunter (1989) Abstract: Oxidation and leaching of pyritic minerals is one of the primary causes of chemical pollution from mining activities. These actions can produce drainage waters that contain high concentrations of sulfuric acid and metals. At present, it is necessary to provide physical-chemical treatment of acid mine drainage. Currently available treatment methods consume significant quantities of chemicals and energy and most produce effluents that contain high concentrations of dissolved minerals. This research investigated the feasibility of the sulfate reduction step in a proposed new process for treatment of acid mine drainage. That process is called biocatalyzed demineralization of acidic metal sulfate solutions and comprises the steps of (1) acid phase anaerobic digestion of biomass to produce volatile acids and a partially stabilized sludge, (2) use of the volatile acids as carbon sources and electron donors for biological sulfate reduction for removal of acidity, metals and sulfate from acid mine drainage, and to produce acetate, (3) use of the acetate solution as feed for methane phase anaerobic digestion to produce methane and to reduce the organic content of the effluent of the process,(4)and use of the methane to satisfy the energy requirements of the process. Single substrate and multiple substrate chemostat studies were conducted to determine if kinetic control (i.e., a relatively short mean cell residence time) could be used to ensure partial oxidation of higher molecular weight volatile acids (propionic and butyric) and production of acetate during the sulfate reduction step. The studies confirmed that propionate produced during acid phase anaerobic digestion can be used as an electron donor for sulfate reduction (during which it is converted to acetate) and that the acetate is available as a substrate for a subsequent methane production step.  BIOCATALYZED PARTIAL DEMINERALIZATION OF ACIDIC METAL SULFATE SOLUTIONS by Robert Michael Hunter A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering MONTANA STATE UNIVERSITY Bozeman, Montana November 1989 COPYRIGHT by Robert Michael Hunter 1989 All Rights Reserved CDSkIS HcUVS ii APPROVAL of a thesis submitted by Robert Michael Hunter This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. ittee Approved for the Major Department Date (f, Hecid, Major Department/ Approved for the College of Graduate Studies Date Graduate Dean ill STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a doctoral degree at Montana State University, I agree that the Library shall make it available to borrowers under the rules of the Library. I further agree that copying of this thesis is allowable only for scholarly purposes, consistent with "fair use" as prescribed in the U.S. Copyright Law. Requests for extensive copying or reproduction of this thesis should be referred to University Microfilms International, 300 North Zeeb Road, Ann Arbor, Michigan 48106, to whom I have granted "the exclusive right to reproduce and distribute copies of the dissertation in and from microfilm and the right to reproduce and distribute by abstract in any format." Signature________ kIov. 2(0 ,IjGdDate iv TABLE OF CONTENTS Page LIST OF TABLES . ................ ■................... vi LIST OF F I G U R E S ...................................... vii ABSTRACT............................................. viii INTRODUCTION ........................................ I Overview ...................................... I Goal and Objectives............................ 2 BACKGROUND . . . . . ................................ 6 The Acid Mine Drainage Problem.................. 6 Formation of Acid Mine Drainage............ 6 The National Problem........................ 7 A Montana Problem (or Opportunity): The Berkeley Pit......... . . , ........ 8 Bioprocessing of Acid Mine Drainage. . ............. 14 A New Biotechnology: Biocatalyzed Partial Demineralization of Acidic Metal Sulfate Solutions......................... 17 Acid Phase Anaerobic Digestion................ 21 Microbial Sulfate Reduction ................ 21 Methane Phase Anaerobic Digestion .......... 24 ,Sulfur Recovery ............................ 26 Modeling of Microbial Growth .................... 27 Model Selection.............................. 2 8 Single Substrate Models .................... 30 Acetate, Propionate and Butyrate Oxidation . . . . 33 Reaction Stoichiometry........................ 3 3 Biomass Yield .............................. 35 Energetics.................................... 3 5 Kinetics...................................... 38 Biochemistry..................' .......... . 41 VTABLE OF CONTENTS-Continued Page EXPERIMENTAL METHODS AND APPARATUS .................. 44 Microorganisms ................................ 44 Media.......................................... 45 Chemostat...................................... 4 7 Analytical Methods . ^ ..................... 4 8 Data Analysis . ...................... .......... 50 Hypothesis Testing........................ 50 Model Development....................... 52 R E S U L T S ............................................ 55 Reaction Stoichiometry. . . . ................... 55 Biomass Yield .................................. 59 Kinetic Parameters.............................. 62 DISCUSSION......................................... 66 Hypothesis Testing.............................. 66 Stoichiometric Evidence.................... 67 Kinetic Evidence.......... 70 Summary.................................... 71 Model Development.............................. 7 2 Process Performance ............ 76 Applications......................... 76 Volatile Acid Turnover .................... 76 Kinetic Control............................ 7 8 Bioprocessing of Acidic Sulfate Solutions from Hydrometallurgical Operations . . . . . . .................. 79 Coal Desulfurization . . . . . . . . . . . . 79 Bioprocessing of Heap Leach Effluents and Acid Mine Drainage.................. 8 0 CONCLUSIONS.................................. 88 NOMENCLATURE ........................................ 90 BIBLIOGRAPHY ........................................ 92 APPENDICES........................................ H 5 Appendix A— Thermodynamic Calculations..........116 Appendix B— The Challenges of Continuous Culture of SRB in Small Reactors......................12 0 Appendix C— Data Tabulations........; ..........125 LIST OF TABLES Table Page 1. Berkeley Pit Water Quality............ .. . . Il 2. Value of Elements in Berkeley Pit . . . . . . 12 3. Annual Value of Metals and Sulfur ........ 14 4. Products of Acid Phase Anaerobic Digestion. . 22 5. Available Data on Biomass Yield............ 3 6 6. Calculated Free Energy Changes ............ 37 7. Available Kinetic Data fdr Growth on Acetate. 39 8. Available Kinetic Data for Growth on Propionate and Butyrate...................... 40 9. Media Composition.............................. 45 10. Flow Rate D ata............ 126 11. Medium Concentration Data. ........ 127 12. Effluent Concentration Data................... 128 13. Calculated Effective Influent Concentrations . 129 14. Acetate Production .......................... 130 15. Sulfate Reduction............................. 131 16. Acid Neutralization............... .......... 132 17. Calculated Hofstee Plot Variables............. 133 18. Calculated Hanes Plot Variables............... 134 19. Biomass Characteristics....................... 135 20. Calculation of Yield Analysis Plot Variables . 136 vii LIST OF FIGURES Figure Page 1. Process Diagram............................ 19 2. Chemostat with pH Control.................. 47 3. Acetate Production Stoichiometry............ 56 4. Sulfate Reduction Stoichiometry . . 57 5. Acid Neutalization Stoichiometry............ 58 6. Analysis of Biomass Yield..................... 61 X i 7. Hofstee Plot of Kinetic Data................ 63 8. Hanes Plot of Kinetic Data.................. 64 9. Plot of Monod Equation...................... 65 10. Process Performance ........................ 77 viii ABSTRACT Oxidation and leaching of pyritic minerals is one of the primary causes of chemical pollution from mining activities. These actions can produce drainage waters that contain high concentrations of sulfuric acid and metals. At present, it is necessary to provide physical-chemical treatment of acid mine drainage. Currently available treatment methods consume significant quantities of chemicals and energy and most produce effluents that contain high concentrations of dissolved minerals. This research investigated the feasibility of the sulfate reduction step in a proposed new process for treatment of acid mine drainage. That process is called biocatalyzed demineralization of acidic metal sulfate solutions and comprises the steps of. (I) acid phase anaerobic digestion of biomass to produce volatile acids and a partially stabilized sludge, (2) use of the volatile acids as carbon sources and electron donors for biological sulfate reduction for removal of acidity, metals and sulfate from acid mine drainage, and to produce acetate, (3) use of the acetate solution as feed for methane phase anaerobic digestion to produce methane and to reduce the organic content of the effluent of the process, (4) and use of the methane to satisfy the energy requirements of the process. Single substrate and multiple substrate chemostat studies were conducted to determine if kinetic control (i.e., a relatively short mean cell residence time) could be used to ensure partial oxidation of higher molecular weight volatile acids (propionic and butyric) and production of acetate during the sulfate reduction step. The studies confirmed that propionate produced during acid phase anaerobic digestion can be used as an electron donor for sulfate reduction (during which it is converted to acetate) and that the acetate is available as a substrate for a subsequent methane production step. IS INTRODUCTION Overview Oxidation and leaching of pyritic minerals is one of the primary causes of chemical pollution from mining activities. These actions can produce drainage waters that contain high concentrations of sulfuric acid and metals. Acid mine drainage can be associated with mining of coal, metal ores (copper, zinc, lead, and uranium), pyritic sulfur and geothermal fluids. An acid mine drainage problem can continue for as long as oxygen and water are in contact with the minerals. The same processes that cause acid mine drainage have also been used in engineered systems for extraction of metal values from low grade ores in dump leaching, heap leaching and tank leaching operations. These processes have been given greater attention in the last few decades as deposits of high grade ores have been depleted and metals prices have dropped (Ehrlich and Holmes, 1.986) . Conventional acid mine drainage treatment comprises lime addition and aeration for ferrous iron oxidation and precipitation. The process is often carried out at a pH above 9 to facilitate precipitation of other dissolved 2metals. Neutralization of the water is sometimes necessary prior to discharge. Other treatment methods, including sulfide addition, reverse osmosis, and ion exchange, have been investigated but have been shown to have higher cost (Peavy and Hunter, 1987). The lime precipitation process is reliable and relatively inexpensive, but it does have some serious disadvantages. One particularly perplexing disadvantage is that lime addition results in supersaturation of treated water with calcium sulfate. This compound, gypsum, usually precipitates after discharge causing cementation of downstream streambed gravels. This phenomenon was one of the most serious environmental effects of discharges of treated wastewater from the Anaconda copper mining opera­ tions in Montana. Even if calcium is removed by soda ash addition in a subsequent treatment step, the resulting effluent can contain such high concentrations of dissolved salts as to be of limited utility. For these reasons, investigation of alternative treatment schemes which include sulfate removal are of interest. Goal and Objectives The goal of this research was to investigate the feasibility of an alternative treatment process, termed biocatalyzed partial demineralization of acidic metal sulfate solutions. The process is comprised of a number of 3steps, most of which utilize basically proven technologies. This reseach focused on a critical step in the process, kinetically controlled microbial sulfate reduction, and assessed its technical feasiblity by testing the following hypothesis: A chemostat containing a mixed culture of sulfate- reducing bacteria (SRB) can be operated at a dilution rate large enough to cause incomplete oxidation of propionic acid and butyric acid and production of acetic acid, if all three acids are present in the reactor feed, but sufficient sulfate is present to allow utilization of only the two higher molecular weight acids. The specific objectives were as follows: 1. to determine an operating pH and temperature likely to be optimum for propionate oxidation while allowing oxidation of higher molecular weight volatile acids, based on literature data, 2. to experimentally determine Monod kinetic parameters (^ max' KS' and Yg) for propionate oxidation under the conditions determined above, 3. to estimate kinetic parameters for acetate oxidation under the conditions selected for optimum propionate oxidation, based on literature data, 44. to experimentally determine kinetic parameters for butyrate oxidation under conditions selected for optimum propionate oxidation, 5. to experimentally determine the stoichiometry of acetate production during oxidation of propionate and buytrate, 6. to use the above parameters to calibrate a multiple substrate growth model, and 7. to confirm that mixed culture sulfate-reducing bacteria use multiple substrates as predicted by the model calibrated using single substrate data. To accomplish these objectives, a number of experiments were conducted using an anaerobic reactor fabricated for this research. Possible applications for results obtained from this research include treatment of acid mine drainage and other acidic metal sulfate solutions, metals recovery and sulfur recovery. The kinetic data developed during this research will contribute to an increased understanding of microbial sulfate reduction. It should be noted that the experiments conducted during this project were not specifically designed to determine.appropriate conditions for essentially complete removal of sulfate. As was noted in the hypothesis this research set out to test, sulfate was believed to be present in the medium in excess of stoichiometric require­ 5ments, often much in excess (i.e., over 1,000 mg/1 in excess). Further experiments with sulfate as the limiting substrate will be necessary to establish appropriate operating conditions for its removal and the kinetic parameters with which to design sulfate removal reactors * 6BACKGROUND ^ The Acid Mine Drainage Problem Formation of Acid Mine Drainage The mechanism of acid water formation when water and air come into contact with iron sulfides associated with coal seams or metal sulfides in ore bodies has been much studied (Goldhaber and Kaplan, 1974). In general, half of the acidity is attributable to sulfide oxidation to sulfate and half to ferrous iron oxidation to ferric iron and its hydrolysis. Acid mine drainage associated with mining of coal forms when FeS2 (pyrite, marcasite, etc.) in coal seams is exposed to moisture and air. Initially, FeS2 is oxidized in the following reaction: 2FeS2 (S) + VO2 + 2H20 — > 2Fe+2 + 4H+ + 4S04-2 (I) Although this reaction can occur abiotically, it is accelerated by Thiobacillus ferrooxidans which also mediates the reaction (Hutchins et al., 1896): 4FeS04 + O2 + 2H2S04 — > 2Fe2 (SO4)2 + 2H20 (2) The oxidation of pyrite is further accelerated by the following reaction: FeS2 + 14Fe+3 + SH2O — > ISFe+2 ■+ 2S04~2 + 16H+ (3) IAlternatively, the ferric iron may precipitate as follows: 4Fe+3 + 12H20 — > 4Fe(OH)3 (J3) + 12H+ (4) Once the pyrite oxidation cycle has started, further oxygen is needed only for the microbially catalyzed oxidation of ferrous iron to ferric iron. Acid mine drainage formed from metal ore deposits results, in part, from reactions similar to those noted above. In addition, chalcopyrite (CuFeS2) is oxidized on exposure to air as follows: 4CuFeS2 + 2H20 + IVO2 — > 4Cu+2 + 4Fe(0H)S04 + 2SO4-2 (5) This reaction is also significantly accelerated by T. ferrooxidans. The role of sulfur bacteria of the genus Thiobacillus in accelerating the above processes has also been much studied (Baker and Wilshire, 1970). These acidophilic, chemolithotrophic bacteria are capable of deriving all of their energy from the oxidation of reduced inorganic sulfur or reduced iron compounds. They consume oxygen and use carbon dioxide as a carbon source. The National Problem It has been estimated that 10,000 miles of streams and 29,000 surface acres of impoundments are seriously affected by mine drainage. About 40 percent of this drainage comes from active mines; the remainder from abandoned surface mines (15 percent) and shaft and drift mines (45 percent) (Goldhaber and Kaplan, 1974). 8The free mineral acid loads associated with coal mine drainage alone in the United States exceed 5,300 tons per day (Goldhaber and Kaplan, 1974.) . Neutralization of this acidity would consume over 1,100,000 tons of lime per year. Manufacture of this amount of lime would consume about 5 trillion BTU per year. A Montana Problem Cor Opportunity); The Berkeley Pit Mining began in the Butte, Montana, area in 1864 when prospectors discovered shallow placer deposits of gold and then silver. By the middle of the twentieth century, over 400 underground mines had operated or were still operating in the area (Camp Dresser & McKee Inc., 1988e). In 1955, Anaconda Copper Mining Company (which merged in 1977 with Atlantic Richfield Company) began an open pit mining operation in Butte known as the Berkeley Pit/ The Pit, which was operated until 1982, encompasses about 767 surface acres and has a volume of approximately 9.7 billion cubic feet from its base in 1982 to its top at elevation 5,520 feet (Anaconda Minerals Company elevation datum). During mine operation, the Pit and associated under­ ground workings in the area were continuously dewatered by pumping the Kelley Shaft and by drainage from the Pit to other underground workings. The water from the Kelley Shaft was used in mining operations and eventually treated and discharged to Silver Bow Creek. Silver Bow Creek is a 9tributary of the Clark Fork River which is a tributary of the Columbia River. In recognition of the potential hazard to human health and the environment, the Silver Bow Creek Superfund site was added to the National Priorities List in 1983 by the U.S. Environmental Protection Agency (EPA) (48 FR 40658, September 8, 1983) pursuant to the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA). The Silver Bow Creek site comprised about 28 stream miles beginning at the Metro Storm Drain in Butte. During work on the Silver Bow Creek site, the importance of Butte area workings as the source of contamination of the site was formally recognized. In 1987, the site was expanded to include the Butte area and its name was changed to the Silver Bow Creek/Butte Area site (52 FR 27627, July 22, 1987, Final Rule). The Butte addition consists of the East Camp area, which comprises the Berkeley Pit and associated underground mine workings, and the West Camp area, which includes the Travona and Emma mines. The two areas are thought to be separated hydraulically by bulkheads installed in the underground workings in the 1960's. The Berkeley Pit and associated flooded mine workings are also identified as the Butte Mine Flooding Operable Unit and plans for a Remedial Investigation/Feasibility Study are currently being formulated consistent with EPA1s National Contingency Plan 10 (NCP) (40 CFR Part. 300), CERCLA and the Superfund Amendments Reauthorization Act (SARA). Preliminary water balance studies indicate that the Pit is currently filling at a rate of about 7.6 million gallons per day (mgd) (Camp Dresser & McKee Inc., 1988c). If this rate does not change, the Pit will overtop (at 5,520 feet) by about the year 2010 (21 years hence). Prior to that time, there is a potential for discharge of contaminated water into the alluvial aquifer (which is exposed on the southeast side of the Pit) by about the year 1996 (7 years hence). The water flooding the Pit is comprised of 4.0 mgd of process water inflows from current mining operations by Montana Resources Inc. and about 3.6 mgd of groundwater inflow. The groundwater inflow includes a component of unknown magnitude resulting from urban stormwater runoff from the City of Butte that flows into the Syndicate Pit and, hence, into an East Camp shaft. Total mine dewatering pumpage from the West Camp area was historically about 2.0 mgd (Camp Dresser & McKee Inc., 1988c). The water level in the West Camp mines has been rising since 1984 at a rate of about 4 feet per month. Water from West Camp mine workings (as measured by flooding of the Travona Shaft) was projected to discharge to Silver Bow Creek alluvial deposits by Spring 1989, however, pumping of the shaft has been initiated with discharge to the City of Butte wastewater treatment plant. It is 11 hypothesized that one reason for the enhanced flooding of the West Camp system is partial failure of the bulkheads that separate the East Camp and West Camp workings because West Camp flooding commenced after dewatering of the Pit ceased (Camp Dresser & McKee Inc., 1988c). Historical data on Berkeley Pit water quality are presented in Table I. The water has been described as an acidic metal sulfate solution (Camp Dresser & McKee Inc., 1988a). If the concentration of metals were to remain at Table I. Berkeley Pit Water Quality Constituent Average concentration, mg/1 unless otherwise noted 1984a 1985a 1986a Z 1987b Cations Aluminum 131 129 178 181 Arsenic 0.13 0.22 0.08 0.70 Cadmium 1.4 1.3 1.6 1.7 Calcium 451 424 450 468 Copper 127 146 192 193 Iron 235 283 812 872 Lead 0.17 - - 0.7 Magnesium 220 205 271 256 Manganese 97 92 133 145 Nickel 0.74 0.65 0.85 - Potassium 4.7 8.4 23 18 Silver 0.024 0.038 0.036 - Sodium 62 63 65 70 Zinc 211 232 409 444 Anions Chloride 12 10 - 19 Sulfate 4,059 4,375 7,060 Other parameters pH, units 2.7 2.4 - 3.0 Silica, as SiO2 86 93 103 a Source: Sonderegger, 1987 b Camp Dresser & McKee Inc., 1988c 12 1987 average levels and the Pit were to fill to the rim (5,520 foot level), the Pit would contain the amount and value of elements indicated in Table 2 based on projected 1990 unit prices (Brower, 1987). Table 2. Value of Elements in Berkeley Pit Element Amount, million pounds Unit price, dollars per pound Value, million dollars Aluminum H O 1.03 113 Copper 118 1.49 176 Magnesium 156 1.48 230 Manganese 88 1.47 130 Zinc 270 0.43 116 Sulfur 1,430 0.04 57 Average concentrations of the following constituents exceed federal ambient water quality criteria for aquatic life (Camp Dresser & McKee Inc., 1988d): Arsenic Cadmium Copper Iron Lead Zinc Furthermore, discharge of water with such high sulfate and iron concentrations and low pH could cause serious acid mine drainage and gypsum precipitation problems downstream. The gypsum precipitation problem caused in treatment of acid mine drainage from the Butte area in the Warm Spring ponds was described by the EPA (1972) as follows: 13 During the course of the field investigations it was noted that from Warm Springs to somewhere between Dempsey and Deer Lodge the bottom materials were 1 cemented1 together.... This condition is due primarily to the high levels of calcium, sulfate, and metallic ions in the discharge from the Anaconda Company settling ponds which precipitates as gypsum and insoluble metallic hydroxides and coat the bottom. The magnitude of the potential gypsum precipitation problem can be indicated by consideration of the following hypothetical scenario: 1. Seven mgd of Pit water is treated with lime to neutralize it and precipitate the metals. 2. Lime addition plus calcium from natural sources results in downstream precipitation of calcium sulfate concentrations in excess of about 2,000 mg/1, its solubility in cold water. Under this scenario, about 400,000 pounds per day of gypsum (2,000 mg/1 as calcium plus 5,000 mg/1 as sulfate) would precipitate on the stream beds of Silver Bow Creek and/or the Clark Fork River. This is equivalent to almost 75,000 tons of gypsum per year. As a four foot by eight foot by 1/2 inch sheet of gypsum sheet rock weighs about 60 pounds, 75,000 tons per year of gypsum is sufficient to "sheet rock" over 500 miles of a hypothetical stream of rectangular cross section, 20 feet wide and 4 feet deep, each year. This potential cementation problem is particularly significant in that the Clark Fork Basin Project (Johnson and Schmidt, 1988) reported that "most of 14 the upper (Clark Fork) river seems unsuitable for trout reproduction due to siltation and other substrate deficiencies." That the problem, of gypsum precipitation has been significant was noted by the EPA (1972) in a study of Clark Fork water quality: The precipitation of gypsum and metallic hydroxides eliminates or greatly reduces the availability of bottom gravels for fish spawning and benthic habitat. An analysis of the annual value of recovery of 100 percent of the copper, zinc and sulfur in Pit water at various pumpage rates is presented in Table 3. The same concentrations and unit prices as were used above were used in this analysis. Table 3. Annual Value of Metals and Sulfur Element Value 2.0 mgd , million dollars per 5.0 mgd year 7.0 mgd Copper 0.9 2.2 3.1 Zinc 1.2 2.9 4.1 Sulfur 0.6 1.4 2.0 Total 2.7 6.5. 9.2 Bioorocessincf of Acid Mine Drainaae A number of investigators have studied the use of sulfate-reducing bacteria (SRB) for treatment of acid mine drainage. To date, however, an economically feasible process has not been developed. 15 Tuttle et al. (1969) described the potential utility of microbial sulfate reduction as an acid mine drainage water pollution abatement procedure. Batch testing of cultures grown in acid mine drainage (pH 3.6) and wood dust at various temperatures indicated that more sulfate was reduced at 37 °C than at 25 °C or 50 °C. Enrichment of wood dust-acid water cultures with sodium lactate resulted in a pH increase from 3.6 to 7.0 in a 10-day period. King et al. (1974) described the use of sulfate- reducing bacteria in accelerating the rate of acid strip mine lake recovery. They indicated that the key to the process is the accrual of enough organic material at the bottom of the lake to allow generation of suitably reduced conditions for sulfate-reducing bacteria to grow. They noted significant differences in both acidity (2,660 mg/1 vs 0 mg/1) and specific conductance (6,400 micromho/cm vs 230 micromho/cm) in lakes and these differences were attributed to the process. Ilyaletdinov and Loginova (1977) reported the findings of batch testing of sulfate-reducing bacteria for removal of copper from the effluent of the Balkhash Mining and Metallurgical Combine. They were able to reduce copper concentrations of 0.08 mg/1 in settling pond effluent (with a sulfate concentration of 1,200 mg/1 and lactate as a carbon source) to 0.02 mg/1 within a few weeks. Batch testing of treatment of effluent from the secondary 16 settling pond with chopped reed and ammonium sulfate additions showed that 1.0 mg/1 concentrations of copper could be reduced to zero in 10 days. The concept was also tested in continuous culture in a six-section reactor with a residence time of 10 days. The bottom of the reactor was covered with mud to provide a habitat for the bacteria. At 25 0Cf the reactor was able to reduce an initial concentration of copper of 1.26 mg/1 to zero. Yagisawa et al. (1977) and Noboro and Yagisawa (1978) presented the results of continuous recovery of metals from a solution prepared by bacterial leaching of copper sulfide ore. Batch culture of sulfate-reducing bacteria in the leaching solution was impossible because of the low pH and high concentration of metals in the solution. The specific growth rate and the rate of removal of metals were found to be strongly influenced by the pH of the culture. The optimum pH for both metal removal and growth was 6. In continuous culture at 30 °C and pH 6, the maximum rate of metals removal occurred at 40 percent of the metal concentration of the original solution. The original solution contained 270.0 mg/1 copper, 102.5 mg/1 zinc and 135.0 mg/1 iron. Sodium lactate and yeast extract were added to the culture. A black precipatate was produced that contained 19.96 percent copper, 6.13 percent zinc and 10.95 percent iron. 17 Olson and McFeters (1.978) described microbial sulfur cycle activity at a western coal strip mine. The sediments of the mine settling pond supported a large and active population of sulfate-reducing bacteria, producing up to 10.5 mg hydrogen sulfide per liter of sediment per day. Metal bound sulfides were found to comprise, at times, over 0.2 percent of the dry weight of pond sediments, leading the investigators to suggest that sulfate-reducing bacteria were precipitating heavy metals in the pond. Grim et al. (1984) described the findings of batch and continuous culture studies of sulfate reduction by an acetate-utilizing strain of Desulfobacter postgatei. Growth was optimized by constant pH control, slow nitrogen purge to prevent inhibition by the sulfide ion, and immobilization of cells on a fixed surface in a continuously stirred tank reactor. A ferric sulfate precipitate adhered to the wall of the reactor apparently allowing cell numbers to increase and facilitating increased sulfate reduction. A New Biotechnology: Biocatalvzed Partial Demineralization of Acidic Metal Sulfate Solutions The shortcomings of other techniques for bioprocessing acid mine drainage required the development of a new process for biocatalyzed partial demineralization of acidic metal sulfate solutions. The process was first described by Hunter, Peavy and Sonderegger (1987). It is Iillustrated generally in Figure I. The process comprises the steps of (I) acid phase anaerobic digestion of biomass to produce volatile acids and a partially stabilized sludge, (2) use of the volatile acids as carbon sources and electron donors for biological sulfate reduction for removal of acidity, metals and sulfate from acid mine drainage, and to produce acetate, (3) use of the acetate solution as feed for methane phase anaerobic digestion to produce methane and to reduce the organic content of the effluent of the process, and (4) use of the methane to satisfy the energy requirements of the process. Key to the feasibility of the process is the use of kinetic control (i.e., a relatively short mean cell residence time) to ensure partial oxidation of higher molecular weight volatile acids (e.g., propionic, butyric, valeric) and production of acetate during the sulfate reduction step. In this way, the higher molecular weight volatile acids produced during acid phase anaerobic digestion can be used BOTH as electron donors for sulfate reduction (during which they are converted to acetate) and as substrates for the subsequent methane production step. If proven technically feasible, the process would have several advantages over prior acid mine drainage bioprocessing schemes. A primary advantage of the technology is that acid phase anaerobic digestion (acido- genesis) of biomass is used to supply electron donors and 18 19 ACID MINE BIOMAM DRAINAGE WASTE METALS VOLATKE ACIDS BIOQAS RECYCLE HEAT METALS ♦ ACETATE RECYCLEELECTRICITY BIOMAM SULFUR DEGASIFICATION SEPARATION SOLIDS DEGASIFICATION STABILIZATION SOUDS SEPARATION SEPARATION SOLDS ACIDOGENESIS DEGASIFICATION AEROBIC TREATMENT CEMENTATION (OPTIONAL) SULFUR RECOVERY METALS PRECIPITATION SULFATE REDUCTION BIOFILM METHANOGENESIS METHANE UTILIZATION CLEAN WATER Figure I. Process Diagram 20 carbon sources to the microbial sulfate reduction step of the process. The presumption here is that volatile acids (e.g., acetic acid, propionic acid, butyric acid, valeric acid, etc.) produced on-site by acidogenesis of biomass will cost less than purchasing preformed electron donors. This presumption is true if a source,of large amounts of relatively easily digested biomass is available near the treatment site. Some other bioprocessing schemes call for use of biomass as a source of electron donors, but they also call for mixing the biomass with the acid mine drainage. This would result in production of a digested biomass that is stabilized but that is also contaminated with heavy metals. The proposed process, on the other hand, would produce a stabilized sludge that is metal-free and that could be used as a soil conditioner in mined land reclamation efforts. Another advantage of the process is that acetate produced during the microbial sulfate reduction step is used as a substrate for a subsequent methane production step. Heat produced from combustion of the methane could be used to provide a portion of the heat required to maintain the contents of the acidogenesis, sulfate reduction and methanogenesis reactors at the elevated temperatures (e.g., 35 °C) necessary for optimum biological action. 21 Considerable work has been done in characterizing acid phase anaerobic digestion, microbial sulfate reduction, and methane phase anaerobic digestion. Important findings are presented below. Acid Phase Anaerobic Digestion The acid formation phase of anaerobic digestion has been thoroughly reviewed (Toerien and Hattingh, 1969; Zehnder, 1978). Its outcome is the conversion of complex organic matter into saturated fatty (volatile) acids, carbon dioxide and ammonia. The volatile acids have been found to be acetic, propionic and butyric acids with lesser amounts of formic, lactic and valeric acids. Acetic acid is the most plentiful followed by propionic acid. Acid phase anaerobic.digestion has been successfully accomplished on a laboratory scale by many investigators (Ghosh et al., 1975; Heijnen, 1984; Pohiand and Ghosh, 1971a and 1971b). The maximum specific growth rate (Mmax) of acidifying bacteria is about 0.3 to 0.5 hr-1. Available data on the products of acid phase anaerobic digestion of wastewater treatment sludge and glucose are presented in Table 4. Microbial Sulfate Reduction There is no known chemical mechanism for the reduction of sulfate by organic matter at normal earth temperatures and pressures (Goldhaber and Kaplan, 1974). Hence, 22 Table 4. Products of Acid Phase Anaerobic Digestion Parameter Hindin and Dustin, 1959 Ghosh et al., 1975 Wise et al., 1985 Substrate Domestic sludge Activated sludge Glucosea Detention time, hours 120 12.7 43.2 pH,units Total volatile solids, mg/1 as 5.15 5.74 5.0 acetic acid Volatile acid components, mg/1 5,519 NRb NR Formic 18 NR NR Acetic 1,882 2,396 846 Propionic 2,717 839 37 Butyric 2,037 890 1,460 Valeric NR 9 00 92 Caproic NR NR 4,450 Lactic 33 NR NR aMethane production suppressed chemically 3^NR = not reported microorganisms are thought to be completely responsible for the process. In dissimilatory sulfate reduction, sulfate is used as the electron acceptor for the oxidation of organic matter. The end product of this oxidation is the excretion of hydrogen sulfide by the microorganism. Until 1980, three genera of sulfate-reducing bacteria were known to exist: Desulfovibrio, Desulfomonas, and Desulfotomaculum. These known organisms can use a limited range of carbon sources, typically fermentation products such as alcohols and organic acids (D'Alessandro et al., 23 1974). The organisms are obligatory anaerobes. Until that time, the known organisms all oxidized organic substrates incompletely, with acetate being the normal end product when lactate, pyruvate or ethanol served as the electron donor and carbon source (Postgate, 1984). In 1980, five new genera, Desulfobacter. Desulfobulbus. Desulfococcus. Desulfosarcina. and Desulfonema. plus new species of the old genera, some of which could oxidize organic compounds completely to carbon dioxide, were reported (Pfennig and Widdel, 1981). As a group, sulfate-reducing bacteria can tolerate a wide variety of environmental extremes, including tempera­ ture, salinity, and pressure (Postgate, 1984). They comprise both mesophilic strains which grow best at temperatures between about 30 °C and 42 °C and thermophilic strains able to grow at temperatures in the 50 °C to 70 °C range. They tolerate pH values ranging from below 5 to 9.5 (Postgate, 1984); Some compounds found in acid mine drainage have been reported to inhibit the growth of sulfate-reducing bacteria. For example, Saleh et al. (1964) reported a minimum inhibitory concentration of 5-50 mg/1 of copper sulfate. Exposure of cell suspensions to 10 millimolar (mM) concentrations of such oxyanions as molybdate, selenate, tungstate, and chromate has caused destruction of adenosine 51-triphosphate (ATP) sulfurylase, the enzyme 24 that catalyzes the first step in sulfate reduction (Taylor and Oremland, 1979). The minimum inhibitory concentration of zinc sulfate, on the other hand, is reported to be 10,000 mg/1 (Saleh et al., 1964). Lead acetate has often been added to solutions used for identification of the organisms. The bacteria have been grown in the presence of carbonates or oxides of lead, zinc, antimony, bismuth, cadmium, cobalt, nickel and copper. Only copper, as the basic carbonate, showed any toxicity (Miller, 1950). Information on the environmental requirements of sulfate-reducing bacteria in pure cultures must be viewed in the light of the ability of the organisms to create microenvironments conducive to their growth. Effective sulfate reduction and iron precipitation has been shown to occur in a pH 2.8 water, for example, when bacteria could not be isolated that grew below pH 5 (Tuttle et al., 1969). A mixed culture of sulfate-reducing bacteria isolated from soil was grown in a chemostat supplied with effluent from a bacterial leaching operation that contained up to 270 mg/1 of copper, 102.5 mg/1 of zinc, and 135 mg/1 of iron (Noboro and Yagisawa, 1978). Maximum growth occurred at a copper concentration of 108 mg/1. Methane Phase Anaerobic Digestion The methane-production phase of anaerobic digestion has been accomplished in conventional chemostats (Ghosh et al., 1975), in fluidized bed reactors, (Heijnen, 1984) and 25 in packed bed reactors (Heijnen, 1984). Several workers have noted that when acetic, propionic and butyric acids are subjected to methane-phase anaerobic digestion, only acetic and butyric acids are metabolized initially. Propionic acid is only degraded when acetic acid concentra­ tions have reached low levels (Pohland and Ghosh, 1971; Heijnen, 1984). When anaerobic digestion is divided into two phases kinetically, two populations of bacteria occur in the methane production phase. The first group are acetogenic bacteria which convert propionic and butyric acids into acetate and hydrogen. The JLtm a x of this population is about 0.01 hr-1 (Boone and Byrant, 1980). The second group is a methanogenic population which converts acetate or hydrogen and carbon dioxide into methane. The JLtm a x of this population is about 0.1 hr-1 with hydrogen plus carbon dioxide as substrate and 0.02-0.03 hr-1 with acetate as substrate (Harper and Pohland, 1986). The biomass yield of both groups is very low. Recent work with an expanded-bed, granular activated carbon anaerobic reactor confirmed that biofilm (attached- growth) reactors can achieve very effective conversion of acetate to methane. Wang (1986) found that such a reactor could produce effluent acetate concentrations of 5 to 6 mg/1 (soluble COD = 15 to 23 mg/1) with an influent acetate 26 concentration of 3,200 mg/1. This performance was achieved at an empty-bed liquid detention time of one day at pH 7.0. Sulfur Recovery Cork and Cusanovich (1978) reported the findings of batch studies of biological conversion of sulfate in solvent extraction raffinates, a waste product of hydrometallurgical copper ore processing, to elemental sulfur. Desulfovibrio desulfuricans was used for sulfate reduction and either of the photosynthetic bacteria Chlorobium thiosulfatophilum or Chromatium vinosum was used for conversion of hydrogen sulfide (H2S) to elemental sulfur. The organisms were cultured separately and a purge system using an inert carrier gas was used to transfer H2S gas from one culture to the other. Lactic acid and yeast extract were used as the carbon source for sulfate reduction. With Chlorobium. an overall process conversion rate of 55 percent was achieved. Cork and Cusanovich (1979) presented the results of pilot scale batch and continuous culture studies of the two-stage conversion process they introduced in 1978. The tests were carried out at an optimum temperature of 30 0C and an optimum pH of 7. At an initial sulfate concentration of 13,400 mg/1, a conversion rate of 91 percent was achieved using lactic acid as the carbon source for sulfate reduction. Utilization of Chlorobium biomass as an alternate carbon source was investigated, but only 10 27 percent of the required carbon could be supplied in that manner. The investigators suggested using other carbon sources such as raw sewage. Uphaus et al. (1983) described another version of purged microbial mutualism using Desulfobacter postcratei. a sulfate-reducing bacterium capable of using acetate as its sole preformed carbon source. This substrate is more economical than lactate and the products of its metabolism include carbon dioxide and hydrogen sulfide which can serve as feed gas for the growth of photosynthetic green sulfur bacteria. Rigid compressed panels of fiberglass taken from commercial ceiling panels were used as a solid support matrix for the cells of the sulfate-reducing bacteria. Addition of concentrated, cell-free Chlorobium culture supernatant to the immobilized Desulfobacter culture increased sulfate reduction rates by 3 to 4 times. Modeling of Microbial Growth A significant part of this research involved develop­ ment and calibration of a model of the sulfate reduction . unit operation of the proposed process. The purpose of the modeling effort was threefold. First, development of the model in the early stages of the project facilitated design of the experimental apparatus and procedures. Second, the model provided a framework for understanding the microbiol­ ogy of the system under study. Finally, a calibrated 28 model is a valuable tool for investigating applications for the basic knowledge gained during the research. The rationale for the model selected for use in the study of the sulfate reduction unit process and an explanation of the model are presented below. Model Selection An appropriate model for the sulfate reduction step of the proposed process should reflect its microbiology. In this situation, two microbiological mechanisms are possible: (I) the sequential utilization of multiple substrates by one or more SRB capable of utilization of acetate, propionate, butyrate, etc. and (2) "natural selection" of one or more microorganisms that are capable of using volatile acids with a higher molecular weight than acetate. The first mechanism is a diauxic (or triauxic) growth pattern and a diauxic (or triauxic) model would be appropriate. With the second mechanism, shifting from one substrate to another does not occur and, at steady state, use of a plurality of single substrate/single microorganism models is appropriate for steady state operation after competition has completed the microorganism selection process. The findings of this research and that of Pfennig and Widdel (1981) conclusively support use of two single substrate/single microorganism models in the instance of utilization of propionate and butyrate. 29 Pfennig and Widdel (1981) described the situation as follows: Enrichment cultures with propionate or butyrate all yielded strains that incompletely oxidized their substrates. Apparently, under conditions of competition such strains have a selective advantage at high concentrations of these sub­ strates. Pure cultures of strains enriched with other substrates were even capable of completely oxidizing propionate and butyrate; however, such strains could not be selectively enriched with these substrates. Thus, single substrate models are used herein because: 1. Acetate, propionate and butyrate are the primary volatile acids present in anaerobic digester supernatent and the only volatile acids considered in this research. 2. The presence of only propionate or butyrate in a medium reportedly select for SRB species that oxidize these substrates incompletely to acetate. 3. SRB species capable of incomplete oxidation of propionate or butyrate are incapable of oxidation of acetate. 4. At low dilution rates, it is likely that three groups of SRB, each of which is capable of oxidizing a different volatile acid, would be present in a chemostat. 5. Under such conditons, the only interaction among the SRB groups would be that, at very low 30 dilution rates, the acetate produced by oxidation of propionate and butyrate would be available for oxidation by acetate-utilizing SRB. Single Substrate Models The continuous stirred tank reactor (CSTR) is one of the simplest reactor configurations used in biochemical operations. These reactors are referred to as chemostats in microbiological research because they provide a constant chemical environment for microbial growth. When that chemical environment contains a single growth-limiting nutrient (substrate), relatively straight-forward mass balances on substrate and cells can be used to develop models for the systems. Mass balances on substrate and cells entering and leaving the reactor are presented below: Accumulation = Mass - Mass + Mass - Mass (6) in out generated consumed . Substrate: (dS/dt)V = QS0 - Q S - rsV (7) Cells (biomass): (dX/dt)V = QX0 - QX + rg/xV - rd/xV (8) where Q = volumetric flow rate t = time V = reactor volume S0 = input substrate concentration 31 S = reactor and output substrate concentration rs = rate of consumption of substrate X0 = input cell concentration X = reactor and output.cell concentration rg/x = rate of generation of cell material rd/x = rate of death and decay of cell material At steady state, when dS/dt = dX/dt = 0, and with sterile input, these equations simplify as follows: Substrate: 0 = (Q/V)(S0 - S) - rs (9) Biomass: 0 = - (Q/V)X + rg/x - rd/x (10) Because bacteria reproduce by dividing in two, the reaction rate for bacterial growth can be expressed as a first order equation: rg/x = (H) where fj, = specific growth rate Similarly, bacterial death and decay can be expressed as rd/x = bX (12) The true growth yield, Yg, is the ratio of the rate of generation of cell material (in the absence of maintenance energy requirements) to. the rate of substrate removal: Xg = rg/x/(-rs) (13) Combining the above equations yields: "rs = (iU/Yg)X (14) 32 Thus the rate of substrate consumption is also first order with respect to the concentration of cells. . The value of the specific growth rate, /x, depends on the concentration of the limiting nutrient available for growth. Unfortunately, a mechanistic equation for the relationship between /x and S has not yet been discovered. The equation that has gained greatest acceptance is the one proposed by Monod (51): M = /%ax*s/(Ks + S) (15) The term Mmax a constant defined as the maximum value possible for /X under a specific set of conditions. Ks is referred to as the half saturation constant and determines how fast M approaches Mmax- Ks the substrate concentration at which m is equal to one half of Mmax- This equation was incorporated into the models developed herein. Taken together, the above equations comprise the following single substrate model: D(S0 - S) = (Mmax/(Ks " S))(X/Yg) (16) X = Yg(So - S)/(I + b/D) (17) where D = dilution rate (Q/V) By substituting equations 11 and 12 into equation 1.0, it is apparent that, at steady state and with X0 =0, n = D + b. That is, the specific growth rate is equal to the dilution rate plus the maintenance coefficient. This fact can be used to develop estimates of Mmax and Ks based on experimental results. 33 Acetate. Propionate and Butyrate Oxidation Since the discovery of SRB species capable of oxidizing acetate, propionate and butyrate in the early 1980's, researchers have investigated the stoichiometry and kinetics of these reactions. Information reported to date is presented below. Reaction Stiochiometrv The stoichiometry of a reaction indicates what changes will occur and to what extent. An understanding of the stiochiometry of a microbial!y mediated reaction allows the yield of the reaction, i.e., the ratio of the mass of products formed to the mass of reactants consumed, to be estimated. A number of investigators have.studied the stoichiometries of SRB mediated oxidations of acetate, propionate and butyrate. Typically, such studies are done using pure cultures of specific SRB. The suggested stiochiometry for acetate oxidation by Desulfotomaculum acetoxidans (Widdel and Pfennig, 1977)t Desulfotomaculum acetoxidans (Schauder et al., 1986), Desulfonema limicola and Desulfonema magnum .(Widdel et al., 1983)' Desulfobacter postgatei (Widdel and Pfennig, 1981)' and SRB in general (Pfennig and Widdel, 1981) is as follows: CH3COO- + SO4-2 — > 2HC03- +.HS- (18) J I 34 From the above, it is apparent that reduction of one mole of sulfate is associated with oxidation of one mole of acetate. The stoichiometry of incomplete propionate oxidation by Desulfobulbus prooionicus was suggested to be as follows by Widdel and Pfennig (1982): 4CH3CH2COO" + 3SO4-2 — > 4CH3COO" + 4HC03" + 3HS" + H+ (19) With this reaction, for every four moles of propionate oxidized, three moles of sulfate are reduced and four moles each of acetate and bicarbonate are produced. Pfennig and Widdel (1981) suggested the following stoichiometry for complete oxidation of fatty acids such as propionate: 4CH3CH2COO" + VSO4=- — > 12HC03~ + 7HS" '+H+ (2.0) The stoichiometry of incomplete oxidation of butyrate by Desulfobacterium autotrophicum was determined to be as follows by Schauder et al. (1986): 2CH3CH2CH2COO" + 3SO4-2 + 6H+ — > 2CH3COO" + 4CO2 + 3H2S + 4H20 (21) Pfennig and Widdel (1981) suggested the following stoichiometry for incomplete oxidation of straight chain fatty acids with even numbers of carbon atoms such as butyrate by SRB: 2CH3CH2CH2COO" + SO4"2 — > 4CH3COO" + HS" + H+ (22) 35 The stoichiometry of complete oxidation of butyrate by Desulfobacterium catecholicum was determined to be as follows by Szewzyk and Pfennig (1987): Pfennig and Widdel (1981) suggested the same stoichiometry for complete oxidation of fatty acids such as butyrate by SRB. Biomass Yield The yield of biomass (growth yield) from a microbially mediated reaction is another aspect of stoichiometry. Biomass yield is generally expressed in terms of dry weight of biomass produced per unit weight (or mole) of substrate consumed. Some published data on observed biomass yields associated with oxidation of acetate, propionate and butyrate by SRB, typically obtained in batch experiments, are presented in Table 5. These relatively low values are typical of other anaerobic processes. Energetics Another aspect of stoichiometry is the energetics or thermodynamics of the oxidation-reduction (redox) reactions mediated by microorganisms. Analysis of the redox reactions involved can, in turn, shed light on the free energy available for microbial growth. Because, in most biological, systems, each step in the oxidation of a substrate involves the transfer of two electrons, free 2CH3CH2CH2COO- + SSO4-2 — > SHCO3- +5HS- + H+ (23) \ 36 Table 5. Available Data on Biomass Yield Substrate/ organism/ reference Biomass Grams of biomass per mole of substrate used yielda Grams of biomass per gram of substrate used Acetate Desulfobacter oostcratei Widdel and Pfennig, 1981 4.4 0.074 Ingvorsen el al., 1984 4.3 0.072 Desulfomaculum acetoxidans Widdel and Pfennig, 1977 5.6 0.095 Mixed culture Middleton and Lawrence, 1977 3.8 0.065 Propionate (incomplete oxidation) Desulfobulbus orooionicus Widdel and Pfennig, 1982 4.7 0.064 Butyrate (incomplete Desulfobacterium autotroohicum Schauder et al., 1986 oxidation) 8.3 0.096 Strain HRM6 Schauder et al., 1986 8.3 0.091 Butyrate (complete oxidation) Desulfobacterium catecholicum Szewzyk and Pfennig, 1987 6.2 0.072 a Dry weight basis energy changes in such reactions are often expressed as free energy change per two electrons (e~) transferred. Table 6 presents free energy changes at unit concentrations of reactants and products and at pH 7.0 (AG0') based on 37 Table 6. Calculated Free Energy Changes Substrate Product(s) Net free energy change, AG"', Kj/2e- Butyrate Acetate -13.92 Propionate Acetate + CO2 -12.63 Butyrate CO2 -12.24 Propionate CO2 -12.10 Acetate CO2 -11.83 calculations presented in Appendix A for the reactions discussed in the previous sections. While free energy changes have often been correlated with biomass yields (Bailey and Ollis, 1986), both Middleton and Lawrence (1977) and Snoeyink and Jenkins (1980) have pointed out that the maximum rate at which / various microorganisms can grow (Atmax) is related to the energy available from the redox reaction that is catalyzed. For example, the maximum specific growth rates associated with the following microbially catalyzed reactions are clearly correlated with a progressively lower amount of energy available to the microorganisms from substrate oxidation: aerobic heterotrophic oxidation, heterotropic denitrification, nitrate oxidation, ammonia oxidation, heterotrophic sulfate reduction, and heterotrophic methane fermentation. For this reason, the calculated values presented in Table 6 suggest that incomplete oxidation of butyrate and propionate might support slightly more rapid 38 growth of SRB than complete oxidation of either these volatile acids or acetate. Kinetics Kinetics describe the rate at which chemical and biological processes occur. In bioprocess engineering, process rates characterize the rates at which microorganisms grow under specified environmental conditions. Two parameters, the maximum specific ,growth rate (Mmax) and the half saturation coefficient (Ks), are used to characterize bioprocess rates. As was noted above, the maximum specific growth rate is the maximum rate of increase of microorganism mass divided by the mass of microorganisms in a system under a specified set of growth conditions unconstrained by the availability of a limiting nutrient. It is related to the minimum microorganism doubling time (tj^min) as follows: td,min = (^n 2)/^max (24) The half saturation coefficient is the limiting substrate concentration at which the specific growth rate is equal to one half of JLtm ax . A number of investigators have characterized SRB capable of utilizing acetate, propionate and butyrate in terms of the above parameters. Typically, the work has been done using pure cultures of a specific SRB. Available data are presented in Tables 7 and 8. Defined (as opposed 39 Table 7. Available Kinetic Data for Growth on Acetate Limiting substrate/ microorganism/ reference Maximum specific growth rate, hr-1 Minumum doubling time, hr Half saturation coefficient, mg/1 Temp. C Acetate Desulfobacter nostcratei Widdel and Pfennig, 1981 0.035a 20 32 Thauer, 1982 - - <12 - Ingvorsen el al. 1984 0.033a 21 30 Schauder et al., 1986 O.025a 28 30 Desulfotomaculum acetoxidans Widdel and Pfennig, 1977 0.058a 12 36 Widdel and Pfennig, 1981 0.023a 30 36 Schauder et al., 1986 O.014a 48 3.0 O.032a 22 - 37 Desulfobacter hvdrocrenoohilus Schauder et al., 1986 O.039a 18 30 Desulfonema so. Widdel et al. , 1983 0.023a'b 30 30 0.0069a 100 - 30 Mixed culture . Middleton and Lawrence, 1977 0.014 250 20 0.019 - 92 25 0.022 5.7 31 Z ^Calculated from minimum doubling time bComplex growth medium to complex) growth media with a pH near neutrality were used unless indicated otherwise. 40 Table 8. Available Kinetic Data for Growth on Propionate and Butyrate Limiting substrate/ microorganism/ reference Maximum specific growth rate, hr-1 Minumum Half doubling . saturation time, coefficient, hr mg/1 Temp. C Propionate Desulfobulbus nronionicus Naninga and Gottschal, 1987 0.110 6.3 35 Widdel and Pfennig, 1982 0.069 10 39 Butyrate Desulfomaculum acetoxidans Widdel and Pfennig, 1981 0.046a 15 36 Desulfovibrio baarsi Schauder et al., 1986 0.017a 42 30 Desulfovibrio saoovorans Nanninga and Gottschal, 1987 0.066a 10.5 35 ^Calculated from minimum doubling time. If formulas developed by Middleton and Lawrence (1977) are extrapolated to 35 °C, an acetate substrate Mmax 0.029 hr-1 and a Ks of 2.7 mg/1 are calculated. In that their studies were done with mixed cultures of SRB, these values probably reflect maximum Mmax values and minimum Ks values for growth of SRB present in their innoculum on that substrate. SRB with lower' Mmax characteristics would be outcompeted by other SRB present in the innoculum. Even if the anomolously high value of Mmax on acetate of Widdel and 41 Pfennig (1977) is considered, review of the data in Tables 7 and 8 suggests that SRB are capable of faster growth on propionate and butyrate than on acetate. Biochemistry Dissimilatory sulfate reduction by SRB begins with activation of sulfate by adenosine triphosphate (ATP) (Gottschalk, 1979). The reaction is catalyzed by the enzyme ATP sulfurylase and forms the products adenosine phosphosulfate (APS) and polyphosphate (PPjJ : SO4-2 + ATP — > APS + PPi (25) The second step in the process is reduction of APS to sulfite by the enzyme APS reductase with the production of adenosine monophosphate (AMP): APS + 2e“ — > SO3-2 + AMP (26) The remaining steps in the pathway are unsettled (Akagi, 1981). With the pathway currently favored by Brock et al., (1984), the third step is the reduction of sulfite to trithionate (with sulfite recycle from the subsequent steps): SO3-2 + 2e“ — > S3O6"2 (27) The fourth is reduction of trithionate to sulfite and thiosulfate: S3O6"2 + 2e" — > S03"2 + S2O3"2 (28) The fifth is reduction of thiosulfate to sulfite and sulfide: 42 P2O3 ^+ 2e~ — > SC>3~2 + (29) Oxidation of acetate by SRB occurs via one of two mechanisms. In Desulfobacter spp., acetate is metabolized to carbon dioxide via a modified citric acid cycle (Thauer, 1982). The normal citric acid cycle is modified in that citrate is synthesized from acetyl-CoA and oxaloacetate is catalyzed by ATP-citrate lyase rather than by a citrate synthase (Holler and Thauer, 1987; Schauder et al., 1987). Oxidation of I mole of acetate to two moles carbon dioxide yields I net mole of ATP by substrate level phosphorylation. In other SRB, acetate oxidation apparently proceeds via a novel pathway termed the oxidative acetyl-CoA/carbon monoxide dehydrogenase pathway (Sporman and Thauer, 1988). The mechanism is in principle a reversal of that of synthesis of one mole of acetyl-CoA from two moles of carbon dioxide which occurs in autotrophic methanogenic bacteria (Fuchs and Stupperich, 1986) and in homoacetic bacteria (Ljungdal, 1986). With at least some SRB that use this pathway, one mole of ATP is required for the activation of acetate to acetyl- phosphate and one mole of ATP is generated by the formation of formate from N10-formyltetrahydrofolate. Thus, this oxidation of acetate does not result in net synthesis of ATP by substrate level phosphorylation. All of the ATP needed for activation of sulfate and SRB growth is 43 generated by electron transport phosphorylation (Spormann and Thauer, 1988). Incomplete oxidation of propionate (to acetate) by the SRB Desulfobulbus propionicus apparently proceeds via the succinate pathway (Stams et al., 1984). Intermediates include propionyl-CoA, methylmalonyl-CoA, succinyl-CoA, succinate, fumarate, malate, oxaloacetate, and pyruvate or phosphoenolpyruvate. This pathway is remarkable in that it is reversible. Incomplete oxidation of butyrate (to acetate) by the SRB Desulfobacterium autotronhicum apparently proceeds by a different pathway (Schauder et al., 1986). Butyrate is activated via CoA transfer from acetyl-CoA which is formed from butyryl-CoA by /3-oxidation. EXPERIMENTAL METHODS AND APPARATUS A number of experiments were conducted to develop the data required to both test the hypothesis presented earlier and facilitate development of a model of the process. In general, the experiments consisted of a series of chemostat runs. During each run, a chemostat was operated at a particular dilution rate until steady state Was reached. Details of experimental methods and apparatus are presented below and in Appendix B. Microorganisms The microorganisms used in the experiment were obtained from three sources to ensure that both acetate­ utilizing and acetate-producing SRB would be present in the innoculum. One source of microorganisms was the American Type Culture Collection. The following SRB were obtained from that source: Desulfobulbus propionicus. ATCC No. 33891 Desulfovibrio sapovorans. ATCC No. 33892 , These cultures were maintained in 20 ml Bellco test tubes with butyl rubber stoppers at 5 0C and used to reinnoculate the reactors when necessary. IMixed cultures of wild SRB and other organisms were obtained from black, anaerobic mud from Silver Bow Creek and from black, anaerobic sludge from the City of Butte, Montana, wastewater treatment plant. Both sources of organisms were known to be impacted by acid mine drainage. These cultures were maintained together at 35 °C in the complete medium described below. 45 Media The growth media were designed to allow growth of any fresh water SRB capable of utilizing acetate, propionate and/or butyrate as an electron donor. The concentrations of constituents in the media containing three acids (acetic, propionic, and butyric) are indicated in Table 9. Table 9. Media Composition Constituent Concentration, mg/1 Monopotassium phosphate 500 Ammonium chloride 1,000 Calcium chloride 60 Sodium sulfate 4,200 Magnesium sulfate + VH2O 2,000 Ferrous sulfate + VH2O 500 Yeast extract 10 Sodium dithionite 30 p-aminobenzoic acid I Biotin I In some of the runs using a medium containing propionate as the sole electron donor, a lower sulfate concentration 46 (3,330 mg/1) was used (See Table 11, Appendix C.)- The above media is essentially Postgate's C medium (Postgate, 1974) with sufficient sulfate to allow incomplete oxidation (to acetate) of the concentrations of propionate and butyrate given below. Growth factors (i.e., biotin and p- aminobenzoic acid) known to be required by some SRB capable of using the electron donors indicated below were also added to the media. For the single acid experiments, single electron donors were added to the media in the following concentrations (mg/1): Concentration, as sodium Electron donor salt as anion Sodium propionate 3,550 2,700 or Sodium butyrate 2,530 2,000 For other experiments, the sodium salts of three volatile acids were added in the following concentrations (mg/1) to simulate acid phase digester supernatent: Concentration, as sodium Electron donor salt as anion Sodium propionate 3,550 2,700 Sodium butyrate 2,530 2,000 Sodium acetate + 3H20 4,380 1,900 The media were autoclaved for 25 minutes at 121 °C and; cooled under an atmosphere of oxygen-free nitrogen. 47 Chemostat The chemostats used in this experiment are depicted schematically in Figure 2. The reactors had a nominal medium volume of 500 ml and were continuously stirred. ANAEROBIC OAS OXYGEN REMOVAL PUMP EFFLUENT ACID REACTOR MEDIUM Figure 2. Chemostat with pH Control The reactors were constructed using a Pyrex glass jar, the top of which was sealed with a black rubber stopper. Black butyl rubber tubing (ID = 3/16 in with a 1/8 in thick wall) and Neoprene tubing were used exclusively to prevent oxygen diffusion into the medium. 48 As little tubing was used as possible to minimize the opportunities for oxygen poisoning. Nitrogen gas was also used to continuously purge the medium reservoirs (to prevent entrance of air as the medium was used), and the reactors (to prevent accumulation of hydrogen sulfide). Oxygen was removed from commercial grade nitrogen by passing it through copper filings maintained at 370 °C that were regularly reduced with hydrogen gas. Gas leaving the reactors was bubbled through beakers containing lead acetate for hydrogen sulfide removal. Continuous pH adjustment was practiced and the acid reservoir was also purged with nitrogen. The pH probe was a double junction type to prevent liquid junction poisoning and was inserted into the medium through a hole in the rubber stopper. Each run using the same electron donor was started by reinnoculating the reactor with the three cultures. Growth was allowed to occur in a batch mode (no fresh medium addition) for one day. The medium input pumps were then turned on and the run begun. Analytical Methods Reactor input and contents were monitored for sulfate, acetic acid, propionic acid, and butyric acid concentrations and for pH. Sulfate concentrations were 49 quantified by measuring sulfur concentrations using inductively-coupled plasma spectrometry after removal of sulfide. Initially, sulfur concentrations were determined on'acidified, diluted, purged (with nitrogen) and filtered samples. Later samples were treated with zinc acetate in alkaline solution, diluted and filtered. Individual volatile acid concentrations were determined on a Varian 3700 gas chromatograph fitted with a flame ionization detector. A 2m by 2mm glass column packed with Supelco 80/120 Carbopack B-DA/4% Carbowax 2OM was used. The column was maintained at 175 °C and a 0.06 M oxalic acid solution was used to condition the column. Biomass concentrations were determined by counting cells in homogenized samples preserved by addition of two percent formaldehyde and refrigeration at 5 °C. The cells were stained with acridine orange to allow direct counting on 0.22 micron black filters by epifluorescence microscopy using an image analyzer. Average cell area, width and length data were also obtained. Biomass weight concentrations were determined by multiplying cell numbers by calculated dry weights assuming the wet cells had a specific gravity of 1.07 and that dry weights were 22 percent of wet weights (Luria, 1960; Robinson et al., 1984) . Average cell volumes were estimated using length and width data with the cells represented as short I I 50 cylinders with a cone on each end, the cones having a height equal to half of the cell width. Data Analysis Data obtained during this research were used for two purposes. The first purpose was to test the hypothesis posed in the Introduction. The second was to calibrate a model of the system studied. Data analysis procedures used to accomplish each purpose are described below. Hypothesis Testing The hypothesis tested during this research is restated as follows: A chemostat containing a mixed culture of sulfate- reducing bacteria can be operated at a dilution rate large enough to cause incomplete oxidation of propionic acid and butyric acid and production of acetic acid, if all three acids are present in the reactor feed, but sufficient sulfate is present to allow utilization of only the two higher molecular weight acids. For this hypothesis to be true, SRB that incompletely oxidize propionate and butyrate (to acetate) must be capable of more rapid growth than those that completely oxidize propionate, butyrate and acetate to carbon dioxide. That is, at steady state in a chemostat, the microbial processes that result in incomplete oxidation must occur I51 more rapidly than the microbial processes that result in complete oxidation. Under such conditions, acetate would be present in the reactor and its effluent. Because the reaction stoichiometries associated with incomplete oxidation are different from those associated with complete oxidation, mass balances of substrate, acetate, and sulfate were also used to test the hypothesis. For example, if, at a certain dilution rate, one mole of acetate is produced and three-quarters mole of sulfate is reduced for each mole of propionate oxidized, then the hypothesis is true to the extent that it concerns propionate oxidation. Analysis of kinetic data was also used in hypothesis testing. For example, if SRB can be grown on propionate at a /I greater than the Atmax of SRB that consume acetate, and incomplete oxidation of propionate occurs, then the hypothesis is true to the extent that it concerns propionate oxidation. If the chemostat were operated at dilution rates above the reported Mmax of acetate-utilizing SRB, these bacteria would not be present in the reactor. Under these conditions, acetate produced by oxidation of propionate or butyrate and not consumed by other types of bacteria occurring in the mixed culture would be present in the reactor and its effluent. Model Development The research plan called for models for propionate oxidation and butyrate oxidation of the form discussed earlier to be formulated by developing estimates of the following parameters: A1Iitax,p = maximum specific growth rate on propionate, hr-1 KSy-p = half saturation coefficient on propionate, mg/1 Ygyp = true growth yield on propionate, mg/mg bp = maintenance coefficient with growth on propionate, hr-1 and AtHiax, b = maximum specific growth rate on butyrate, hr-1 Kgyj3 = half saturation coefficient on butyrate, mg/1 Ygyj3 = true growth yield on butyrate, mg/mg bj-, = maintenance coefficient with growth on butyrate, hr-1 A variety of procedures have been- proposed for analyzing continuous culture (chemostat) data to determine kinetic parameters (Wilkinson, 1961; Dowd and Riggs, 1965; Grady and Lim, 1980; Bailey and Ollis, 1986). It is known that culture history can affect the analysis (Templeton and Grady, 1988). In fact, considerable variability in Mmax 53 and Ks values occurs with mixed cultures and it is considered good practice to quote them as ranges (Grady and Lim, 1980). The procedures used in estimating /Xmax, Ks, Yg, and b herein were to be those recommended by Grady and Lim (1980). Values for Yg and b were to be obtained from plots of a linearized version of the equation: X = Yg (S0 - S)/ (I + b/D) (17) With the equation expressed in the form (S0 - S)/X = (b /Y g ) (1/D) + I/Yg (30) plot is made of (S0 - S)/X as a function of 1/D and the result is a straight line with slope b/Yg and ordinate intercept I/Yg. Values for /Xm a x and Ks were obtained from plots of two different linearized versions of the Monod equation. In addition, a nonlinear regression procedure was used. The double reciprocal or Lineweaver-Burk method of plotting was avoided because it reportedly gives a deceptively good fit even with unreliable data (Grady and Lim, 1980). With the Monod equation expressed in the form of a Hofstee plot (D + b)/S = /xmax/Ks - (1/KS) (D + b) (31) a plot is made of (D + b)/S as a function of (D + b). The result is a straight line with slope 1/KS and ordinate intercept /Xm a x / Ks. A least squares line of best fit may be used with this method.. I I U I t 54 In the second version, the Monod equation is expressed in the form of a Hanes plot S/(D + b) = (I/Zhnax) ^ + ^s/^max (32) and a plot is made of S/(D + b) as a function of S. The result is a straight line with slope !/Atmax and ordinate intercept Ks/Atmax. The least squares technique is not an appropriate means of finding the line of best fit because both axes contain terms which are subject to error (i.e., S) (Grady and Lim, 1980). The line is drawn by eye. The nonlinear regression of the Monod equation data was accomplished on Montana State University's VAX computer using SAS statistical software. The program produced estimates of Atmax and Ks given data pairs of limiting substrate concentration and specific growth rate. The standard error and the 95 percent confidence interval associated with the estimate of each parameter was also determined. 55 RESULTS The results of this research characterize volatile acid oxidation during continuous culture of a mixed population of sulfate-reducing bacteria (SRB) and other anaerobic microorganisms. Tabulated data are presented in Appendix C. The experiments were designed to provide information on reaction stoichiometry, biomass yield and kinetic parameters. Each aspect is described below. Reaction Stoichiometry Under all of the conditions investigated, the only reaction that occurred to any appreciable extent was SRB- mediated oxidation of propionate and.reduction of sulfate. Even at dilution rates significantly lower than the reported Mmax of butyrate-utilizing SRB, very little butyrate oxidation occurred. Similarly, only a small amount of acetate oxidation occurred during the two runs conducted at dilution rates significantly lower than the reported Mmax of acetate-utilizing SRB. Figures 3 through 5'present the results of determinations of the rate of consumption of reactants and creation of products during the chemostat runs. Comparison 56 1 .00 0 .6 0 0 .40 0.20 0.00 0 .800 .600.200.00 Propionate consumption, mM/hr • Propionate i Three Acid ■ High Sulfate Medium Medium Medium Figure 3. Acetate Production Stoichiometry of the rates over the range studied reveals the stoichiometry of the reaction(s). Figure 3 shows that approximately 0.98 mole of acetate was produced for each mole of propionate consumed. This 1:1 relationship was true for runs with a medium containing propionate as essentially the only organic substrate and just enough sulfate to allow its incomplete oxidation by SRB (round data points). It was also true for runs with a medium containing all three acids (triangular data points) and for runs with a medium containing propionate and significantly more sulfate than is required for its incomplete oxidation (square data points). 57 0 .75 0 .60 0 .45 0 .30 0 .15 0.00 0.00 0.20 0 .40 0 .60 0 .80 Propionate consumption, mM/hr • Propionate a Three Acid ■ High Sulfate Medium Medium Medium Figure 4. Sulfate Reduction Stoichiometry The square data points at the two lowest propionate consumption rates were generated at a dilution rate of about 0.018 hr-1 and at reactor acetate concentrations of about 1,200 mg/1. This dilution rate is much lower than the reported Atmax f°r acetate-utilizing SRB (about 0.03 hr-1) and the concentration is much higher than their reported Ks (about 3 mg/1). Even under these conditions, almost all consumed propionate was recovered as acetate on a mole per mole basis. Figure 4 shows that approximately 0.73 mole of sulfate was consumed (reduced) for each mole of propionate consumed. The scatter in the data is probably due to the 58 1.25 0 .75 0 .5 0 0 .25 0.00 0 .800.20 0.40 0 .600.00 Propionate consumption, mM/hr • Propionate a Three Acid ■ High Sulfate Medium Medium Medium Figure 5. Acid Neutralization Stiochiometry fact that total sulfur concentrations were measured on samples from which sulfide had been removed. Moreover, sulfur (sulfate) in the "distilled" medium make-up water may also have varied. (See the discussion of Substrate Concentration in Appendix B). Figure 5 shows that approximately 1.18 mole of monoprotic acid was "neutralized" for each mole of propionate consumed to maintain the reactor contents at pH 7.0 ± 0.2. These protons left the reactor as H2S gas. About 81 percent of the sulfide ion produced by sulfate reduction left the reactor in this way. Because essentially all of the iron in the medium was precipitated Ill as iron sulfide (essentially no iron was present in filtered effluent samples), some of the sulfide left the reactor as iron sulfide. Mass balances indicate that six to seven percent of the sulfide left the reactor in this form. The remaining 12 to 13 percent must have left as dissolved sulfide. Previous investigators of methods for bioprocessing of acid mine drainage have not made it clear that H2S and CO2 must be removed from the treated mine water (e.g., by purging, stripping, etc.) for the processes to reduce the pH of the mine water. This fact has important implications on the economic feasibility of the process which are outlined later in the Discussion section. Biomass Yield Microscopic examination of samples of reactor contents indicated that during all runs the predominant microorganism had the characteristic lemon-shaped morphology of Desulfobulbus propidnicus. Typically, the cells were in pairs or chains. In all runs, however, some other microorganisms were present: typically long, thin vibrioids species and/or larger ellipsoidal species. Some of the vibrioid cells were of a size and morphology similar to that reported for Desulfovibrio baarsii. an SRB capable of growth on propionate and acetate (Krieg and Holf, 1984). The ellipsoidal cells were of a size and morphology 59 60 similar to that reported for Desulfobacter so.. a SRB capable of growth on acetate, and Desulfosarcina so., SRB capable of growth on propionate, acetate, butyrate, and CO2. It should be noted that propionate, acetate and CO2 were available for growth in all runs and other SRB and non-SRB may have been present. The average size of the cells in the reactor varied with the conditions imposed upon the culture. For the purposes of this research, the following cell size averages were used in calculating biomass (weight) concentrations from cell count data tabulated in Appendix C: Condition Average cell area, scf. microns Average cell length, microns Average cell width microns Propionate medium 0.895 1.45 0.919 Three acid medium 0.904 1.43 0.924 Low dilution rate 0.696 1.29 0.821 As indicated, the low dilution rate runs produced a much more heterogenous culture having a much smaller average size. Because a mixed culture was present, significant variability in biomass yield occurred. Figure 6 presents an analysis of biomass yield assuming that propionate was the only substrate (S) being used for growth. It was not feasible to measure yields over a large enough range to allow a maintenance coefficient, "b", to be calculated. 61 Cl E Cl E X CO I O 09 100 80 60 40 20 O O 12 24 36 4 8 60 Space time (1/D), hr • Propionate Medium a Three Acid Medium Low Dilution Rate Figure 6. Analysis of Biomass Yield This was the case because the "window" of growth rates between the Aim a x of propionate-utilizing SRB (0.07 hr-1) and the A%ax of acetate-utilizing SRB (0.03 hr-1) was too small. Average cell volumes used in estimating biomass yields for the three growth conditions investigated were estimated using the length and width data presented earlier with the cells represented as short cylinders with a cone on each end, which was a simplification of the typical cell morphology. The resulting calculated cell areas agreed well with the measured average areas. The yield results were as follows: Condition Propionate medium Three acid medium Low dilution rate Observed yield, mg biomass/mg propionate consumed 0.022 0.030 0.028 An average yield value for a low dilution rate condition. Run 18.2, (which produced Tittle visible FeS) determined by filtration through 0.45 micron filters and drying at 1059C was 0.042 mg cells/mg of propionate consumed. These values are lower than the 0.064 value reported by Widdel and Pfennig (1982) who determined "cell mass" by drying (at 80°C) cells (and, possibly, metal sulfides) which were harvested by centrifugation and washed once with 10 mM sodium phosphate buffer (pH 4.5). These differences are reasonable in that the temperature at which a residue is dried has an important bearing on such results (Taras et al., 1971). The fact that the cell counting technique used typically undercounted cells smaller than the predominant cell type may have also resulted in lower biomass yield values. Kinetic Parameters Kinetic parameters for oxidation of propionate by SRB were estimated using both a Hofstee plot and a Hanes plot of the data tabulated in Appendix C. The calculated kinetic parameters should be considered appropriate for the 63 environmental conditions under which growth occurred (temperature, medium composition and pH, etc.)* They should also be considered approximate because S (and, hence, Mmax anc^ Ks) may depend on S0 when mixed cultures are employed (Grady and Lim, 1980). The "effective" initial substrate concentration (S0) varied among the experiments conducted herein because of the diluting effect of automatic acid additions. The dilution rates used are based on the effluent flow rate which comprise the total of the medium addition rate and the acid addition rate. A Hofstee plot (b = 0) of the results of runs using a propionate medium, a three acid medium, and a high sulfate medium is presented in Figure 7. A Hanes plot of the same 0 .15 0 .09 0 .06 0 .03 0.00 0.00 0 .15 0 .30 0 .45 0 .60 0 .75 (E - I) Dilution rate, h r - 1 • Propionate a Three Acid ■ High Sulfate Medium Medium Medium Figure 7. Hofstee Plot of Kinetic Data 64 data is presented in Figure 8. It should be noted that essentially all of the data points fall on or near a straight line in both plots. Thus, it appears that the Monod equation can be used to effectively characterize mixed culture SRB growth on propionate. Analyses of these plots and the nonlinear regression analysis resulted in the following estimates of kinetic parameters: Nonlinear Parameter Hofstee plot Hanes plot regression JLtm a x , hr-1 0.073 0.067 0.079 Ksyp , mg/1 106 79 187 The nonlinear regression gave a standard error of Mmax of 0.013 hr-1 and a 95% confidence interval of 0.052 to 0.107 hr-1. For Ks, the standard error was 123 mg/1 and the 1000 Substrate concentration, mg/I • Propionate i Three Acid ■ High Sulfate Medium Medium Medium Figure 8. Hanes Plot of Kinetic Data 65 I £ % so I 1.00 0 .80 0 .40 0.20 1 000 Substrate concentration, mg/I e Propionate 4 Three Acid ■ High Sulfate Medium Medium Medium Figure 9. Plot of Monod Equation 95% confidence interval was -78 to 453 mg/1. Thus, a Mmax of approximately 0.07 hr-1 and a Ksyp of approximately 100 mg/1 result. A Monod equation plot using these parameters and the data collected in this research is presented in Figure 9. 66 DISCUSSION The implications of the results presented earlier in terms of hypothesis testing, model development, process performance and process applications are discussed below. Hypothesis Testing The hypothesis tested during this research is as follows: A chemostat containing a mixed culture of SRB can be operated at a dilution rate large enough to cause incomplete oxidation of propionic acid and butyric acid and production of acetic acid, if all three acids are present in the reactor feed, but sufficient sulfate is present to allow utilization of only the two higher molecular weight acids. As was noted earlier, this hypothesis is true if the following are true: I . measurement of reactant and product concentrations (stoichiometry) of the reactions occurring at a selected dilution rate indicates that incomplete oxidation to acetate (as opposed to complete oxidation to carbon dioxide) is occurring, and 67 2. determination of kinetic parameters indicates that, at a selected dilution rate, propionate and butyrate is consumed more rapidly than acetate. Stoichiometric Evidence The stoichiometric results indicate that, with growth on a propionate medium, the following reaction occurs: 4CH3CH2COO" + SSO4-2 — > 4CH3COO- + 4HC03- + 3HS- + H+ (19) A two-tailed test of means using the t distribution of the data presented on Figure 3 (and tabulated in Appendix C, Table 13) indicates that with growth on a propionate medium, the rate (mM/hr) of acetate production is equal to 100 percent of the rate (mM/hr) of propionate oxidized at the 0.05 significance level. The data are also consistent with the conclusion that the same reaction occurs in a three acid medium. A similar two tailed test of those data also revealed that the acetate produced was equal to 100 percent of the propionate oxidized at the 0.05 significance level. Butyrate oxidation, on the other hand, occurred to a very limited extent in the reactors and may not have been accomplished by SRB. Clear reactor effluent stored at 35 °C turned black due to production of H2S indicating that sulfate reduction did eventually occur. 68 The lack of appreciable butyrate oxidation in the chemostat was surprising in that the medium composition, pH and temperature were in the range reported acceptable to butyrate-utilizing SRB. Neither Berav1s Manual (Krieg and Holt, 1984) nor The Prokaryotes (Pfennig, 1981) state that Desulfovibrio sapovorans. the butyrate-utilizing SRB with the largest JLimax, has a specific requirement for vitamins or trace elements. fDesulfomatoculum acetoxidans does have a requirement for biotin which was present in the medium). Some other SRB do have a requirement for specific trace elements; for example, Desulfosarcina variabilis requires molybdate as a trace element for growth on benzoate (and its growth is inhibited completely by even the diffuse light in a laboratory) (Krieg and Holt, 1984). In fact, the medium used herein contained all of the components recommended by Bergy1s Manual for cultivation of Desulfovibrio sapovorans and that reference states that vitamins are not required. The only cultivation of Desulfovibrio sapovorans reported in the English language (Nanninga and Gottschal, 1987) used a basal bicarbonate-buffered (pH 7.0) fresh water medium with vitamins and trace elements. Furthermore, the medium Widdel (1980) used in his discovery of the species apparently contained trace elements as it was a defined medium that did not contain yeast extract. The other butyrate-utilizing SRB capable of growth at the 69 dilution rates studied, Desulfotomaculum acetoxidans. has also only been grown in a defined medium containing trace elements (Widdel and Pfennig, 1981). Thus, butyrate­ utilizing SRB may have an unreported requirement for a trace element for rapid growth. Alternatively, another hypothesis is that those SRB may require a more reduced environment for growth than could be achieved in the small reactors used in this research. (See discussion in Appendix B). Pfennig et al. (1981) recommend addition of sodium dithionite to batch cultures of Desulfovibrio sapovorans and, although it was added to the media used in this research, it may have been oxidized prior to introduction into the reactors. A third hypothesis is that rapid purging of the reactors with an inert gas (N2) reduced the ambient H2 level in the reactors sufficiently to prevent the synthesis a required hydrogenase by Desulfovibrio sapovorans. Tsuji and Yagi (1980) noted that growth of another Desulfovibrio species, vulgaris, was inhibited by rapid sparging with the inert gas argon. They noted that the organisms generated a burst of H2 soon after innoculation in a lactate medium that may have served to induce the synthesis of a low molecular weight hydrogenase required for growth. (It should also be noted that Desulfobulbus propionicus can , utilize H2 if both carbonate and acetate are present). 70 A fourth hypothesis is that a butyrate-utilizing strain of Desulfovibrio sapovorans was not present in the innoculum. The ATCC strain was not provided in a butyrate medium and the high metal content of the environments from which the wild SRB were obtained may have prevented its presence. Thus the hypothesis this research set out to test was not true concerning butyrate under the conditions of this experiment. If this were the case in a large reactor treating acid mine drainage, the butyrate would be present in the reactor effluent and available for conversion to methane in the next step in the process. Butyrate is unlike propionate in that it has been shown to be amenable to rapid conversion to methane in fixed-film reactors (Heijnen, 1984). Kinetic Evidence The results show that propionate-utilizing SRB are capable of much more rapid growth than acetate-utilizing SRB in both a propionate medium and a three acid medium. The JLtm a x of propionate-utilizing SRB is greater than the reported Mmax for acetate-utilizing SRB. It may also be true, however, that propionate- utilizing SRB exhibit a larger Ks than that reported for acetate-utilizing SRB. . Thus, at dilution rates very much lower than the Mmax of acetate-utilizing SRB, a significant amount of acetate consumption could occur. During Runs 71 18.1 and 18.2 (dilution rate = 0.018 hr-1) about 17 percent of the acetate produced by propionate oxidation was itself oxidized or the reactor was populated, in part, by SRB capable of complete oxidation of propionate to CO2. From another perspective, however, a relatively large Ks offers some advantage. Grady and Lim (1980) point out, for example, that systems with a more gradual variation in specific growth rate with substrate concentration, i.e., systems with a relatively large Ks, tend to be more stable in continuous culture. Summary In summary, the results produced by the experiments described herein prove that a chemostat containing a mixed culture of SRB can be operated at a dilution rate large enough to cause incomplete oxidation of propionic acid and production of acetate, if acetic, propionic and butyric acids are present in the reactor feed, and sufficient sulfate is present to allow utilization of all three acids. The mechanism of propionate oxidation was, as expected, oxidized by SRB. Of course, the results do not prove that it is impossible to cause butyric acid oxidation to occur as well because a negative cannot be proven (Platt, 1964). These results, taken with the following findings of others, suggest that biocatalyzed partial demineralization of acidic metal sulfate solutions is technically feasible: 72 1. Ghosh et al. (1975) and Ghosh et al. (1987) showed that acid phase anaerobic digestion (acidogenesis) could be used to produce low molecular weight volatile acids from biomass. 2. Harper and Pohland (1986) showed that hydrogen management can be used to alter the mix of volatile acids produced by acidogenesis. 3. Ilyaletdinov et al. (1977) and Noboro and Yagasawa (1978) showed that SRB could be used to remove copper and other metals from acid mine drainage. 4. Heijnen (1984), Bhadra et al. (1987) and Hamoda and Kennedy, 1987) showed that acetate and butyrate can be efficiently converted to methane in low detention time fixed-film reactors. 5. Lawrence and McCarty (1966) showed that soluble sulfide concentrations up to 200 mg/1 as S exert no significant toxic effects on biological methanogenesis. 6. Cork (1987) showed that photosynthetic sulfur bacteria can be used to recover 95 percent of the sulfur from a H2SZCO2 gas stream. Model Development Substrate utilization models were developed to allow substrate and product concentrations to be estimated given a proposed dilution rate. Use of the models is 73 appropriate if growth conditions are similar to those used in their development (e.g., complex medium, temperature = 35° C, continuous H2S removal, pH = 7.0, etc.). SRB growth on propionate is adequately characterized by the Monod equation: D = H = /%ax,p*Sp/ (Ks/p + Sp) (3 3 ) or Sp = D*Ks/p/ (/hnax,p “ D) (34) Sp = 1 0 0*0/(0 . 0 7 - D) (3 5 ) The biomass concentration can be estimated using the following relationship: %p = Yg/p(so/p sp) (36) Xp = 0 . 0 3 (So/p - Sp) (3 7 ) At dilution rates above about 0.03 hr-1, the following equation can be used to estimate the concentration of acetate in the reactor effluent (Sa): Sa = (So/p - Sp)*(59.04/73.07) (38) where 59.04/73.07 = ratio of molecular weights of acetate and propionate In both of the above equations, S0/p is the effective influent propionate concentration. At dilution rates below about 0.03 hr-1, for the acetate produced by oxidation of propionate to be oxidized to CO2, the series of reactions, kI k2 propionate ---- > acetate ---- > CO2 (39) must occur. Under such conditions, where cultivation of I, acetate-utilizing SRB is possible, the acetate concentration in the feed (S0/aj would be consumed according to the Monod equation. Mass balance considerations suggest that the following equation can be used to estimate the concentration of acetate in the reactor effluent if complete oxidation of propionate to CO2 did not occur (Grady and Lim, 1980). Sa = So/a/(I + k2/D) + (klSp/o/D)/[(l + It1ZD) (I + k2/D) ] (4 0 ) or Sa — So/a/(l + Mmax,a*Sa*Xa/Yg/a (Ks/a + Sa)*D + (Atmax'p*sp*xp/Yg/p (Ks/p + sp) *D) *sp/o /(! + ^max'p*Sp*Xp/Yg/p(Ks/p + sp)*D) * (I + /^max'a*Sa*Xa/Yg/a (Ks/a + Sa) *D) (4 1 ) Under practical operating conditions at low dilution rates, Sa » Ks/a and Sp « Ks/p, and It1 would be first order with respect to Sp and k2 zero order with respect to Sa. The above equation could then be simplified to the following: Sa = Sa/g/(l + Mmax^a*xa/Yg/a*D + (^ max'p*xp/Yg/p*Ks/p)so/p/(D * t1 + (^maxfp*Xp/Yg/p*Ks/p)/D) * (I + (Mmax,a*Xa)/Yg/a*D)) (4 2 ) 74 or 75 if ba = O and bp = O Sa = So/a/(l + 0.03(So/a - Sa)/D + (0.07(So/p - Sp)/90)So/p/(D * (I + 0.07(So/Po " Sp)/90 * D) * (I + 0.03(So/a - Sa)/D)) (43) Thermodynamic evidence presented in Appendix B as well as limited published information on growth rates of SRB capable of complete oxidation of propionate suggest, however, that complete oxidation of propionate is likely to occur more rapidly that sequential oxidation of propionate and then acetate. Desulfococcus multivorans and Desulfosarcina variabilis are reported to grow slower on propionate than Desulfobulbus orooionicus (Kreig and Holt, 1984) but none have suggested that they grow slower on propionate than acetate-utilizing species are capable of growing. In fact, growth of D. variabilis on acetate alone is reported to be "very slow" compared to growth on other volatile acids (Kreig and Holt, 1984). Butyrate-utilizing SRB resisted continuous culture with a medium containing butyrate as the only organic substrate and with a three acid medium during this research. Perhaps the medium lacked a growth factor that has not yet been reported in the literature or the dilution rates used were too great under the environmental conditions imposed on the culture. A minor but measureable 76 amount of butyrate consumption did occur over and above the apparent "consumption" caused by dilution of the effluent by the neutralizing acid added to the reactor. Process Performance The performance of the sulfate-reduction unit operation of the proposed process operating at dilution rates above 0.03 hr--*- is compared to model predictions in Figure 10. Only propionate and acetate components of the medium are considered as the butyrate component appeared to be resistant to change in the process. Applications The knowledge gained during this research has a variety of applications. The most obvious concerns treatment of acid mine drainage. This and other applications are discussed below. Volatile Acid Turnover The rates at which volatile acids are consumed in marine and estuarine sediments are of concern to microbial ecologists (Lovley and Klug, 1982; Dicker and Smith, 1985; Anderson et al., 1987). The findings of this research indicate that, under conditions of high concentrations of propionate, acetate, and sulfate, mixed SRB populations are capable of converting propionate to acetate much faster than acetate can be oxidized by SRB, and, in fact, much 77 3 5 0 0 3 5 0 0 2 8 0 0 2 8 0 0 2100 2100 1 400 1400 0 .30 0 .37 0 .44 0 .58 0 .65 (E - I ) Dilution rate, h r - 1 • Propionate ■ Acetate Dl E o m ceo COO e J. 107 Ollivier, B., Cordruwisch, R., Lombardo, A. & Garcia, J.L. Isolation and characterization of Sporomusa acidovorans sp. nov., a methylotrophic homoacetogenic bacterium. Archives of Microbiology. 1985, 142, 307-310. Olson, G.J., & McFeters, G.A. Microbial sulfur cycle activity at a western coal strip mine. Montana University Joint Water Resources Center, Bozeman, Montana, 1978. Op de Camp, H.J.M., Verhagen, F.J.M., Kivaisi, A.K., de Windt, F.E., Lubberding, H.J., Gijzen, H.J., & Vogels, G. D. Effects of lignin on the anaerobic degradation of (Iigno) cellulosic wastes by rumen microorganisms. Applied Microbiology and Biotechnology. 1988, 29., 408- 412. Pankhania, I.P., Spormann, A. M., Hamilton, W.A., & Thauer, R.K. Lactacte conversion to acetate, CO2 and H2 in cell suspensions of Desulfovibrio vulgaris (Marburg): indications for the involvement of an energy driven reactions. Archives of Microbiology, 1988, 150. 26-31. Parkin, G.F., & Owen, W.F. Fundamentals of anaerobic digestion of wastewater sludges. 867-891. Peavy, H.S., & Hunter, R.M. Treatment of Berkeley pit water. Berkeley pit re-industrialization and mineral recovery project. Helena, Montana: Headwaters Research Institute, 1987. Peppier, H.J. Food yeasts. In A.H. Rose & J .S. Harrison (Eds.), The yeasts. New York: Academic Press, 1970. Pfennig, N. The phototrophic bacteria and their role in the sulfur cycle. Plant and Soil. 1975, 43., 1-16. Pfennig, N., & Widdel, F. Ecology and physiology of some anaerobic bacteria from the microbial sulfur cycle. In H. Bothe & A. Trebst (Eds.), Biology of inorganic nitrogen and sulfur. New York: Springer-Verlag, 1981a. Pfennig, N., Widdel, F., & Truper, H.G. The dissimilatory sulfate-reducing bacteria. In M.P. Starr, H . Stolp, H.G. Truper, A. Balows & H.G. Schlegel (Eds.), The prokaryotes. New York: Springer-Verlag, 1981b. Phelps, T.J., Conrad, R., & Zeikus, J.G. Sulfate-dependent interspecies H2 transfer between Methanosarcina barker! and Desulfovibrio vulgaris during coculture metabolism of acetate or methanol. Applied and Environmental Microbiology. 1985, 50, 589-594. 108 Platt, J.R. Strong inference. Science. 1964, 146. 347- 353. Pohland, F.G., & Ghosh, S. Developments in anaerobic treatment processes. Proceedings of Biotechnology and Bioengineering Symposium 2. 1971a, 85-106. Pohland, F.G., & Ghosh, S. Developments in anaerobic stabilization of organic wastes - The two-phase concept. Environmental Letters. 1971b, I, 255-266. Postgate, J.R. The sulphate-reducing bacteria. (2nd ed.). Cambridge: Cambridge University Press, 1984. Postgate, J.R. Recent advances in the study of the sulfate-reducing bacteria. Bacteriological Reviews. 1965, 29, 425-441. Powledge, T.M. Gene therapy: will it work? Technology Review , 1983, 82-87. Process design manual for sludge treatment and disposal. (U.S. Environmental Protection Agency, EPA625/1-79-011). Center for Environmental Research Information, 1979. Ramkrishna, D., Kompala, D.S., & Tsao, G.T. Are microbes optimal strategists? Biotechnology Progress. 1987, 3_, 121-126. Robinson, J.A., & Tiedje, J.M. Kinetics of hydrogen consumption by rumen fluid, anaerobic digester sludge, and sediment. Applied and Environmental Microbiology. 1982, 44,. 1374-1384. Robinson, J.A., Trulear, M.G., & Characklis, W.G. Cellular reproduction and extracellular polymer formation by Pseudomonas aeruginosa in continuous culture. Biotechnology and Bioengineering. 1984, 23., 1409-1417. Roels, J.A. Biotechnology and Bioengineering. 1980, 22. 33. Rufener, W.H. Jr., Nelson, W.O., Wolin, M.J. Maintenance of the rumen microbial population in continuous culture. Applied Microbiology. 1963, 11, 196-201. Russell, J.B., & Martin, S.A. Effects of various methane inhibitors on the fermentation of amino acids by mixed rumem microorganisms in vitro. Journal of Animal Science, 1984, 59, 1329-1338. 109 Saleh, A., MacPherson, R., & Miller, J.D.A. The effect of inhibitors on sulfate-reducing bacteria: a compilation. Journal of Applied Bacteriology. 1964, 27., 281-293. Sansone, F.J., & Martens, C.S. Volatile fatty acid cycling in organic-rich marine sediments. Geochimica et Cosmochimica Acta. 1982, 46, 1575-1589. Schauder, R., Eikmanns, B., Thauer, R.K., Widdel, F., & Fuchs, G. Acetate oxidation to CO2 in anaerobic bacteria via a novel pathway not involving reactions of the citric acid cycle. Archives of Microbiology. 1986, 145. 162-172. Schauder, R., Preub, A., Jetten, M., & Fuchs, G . Oxidative and reductive acetyl CoA/carbon monoxide dehydrogenae pathway in Desulfobacterium autotrophicum. 2. Demonstration of the enzymes of the pathway and comparison of CO dehydrogenase. Archives of Microbiology. 1989. 151. 84-89, Schauder, R., Widdel, F., & Fuchs, G. Carbon assimilation pathways in sulfate-reducing bacteria. II. Enzymes of a reductive citric acid cycle in the autotropic Desulfobacter hvdrogenophilus. Archives of Microbiology. 1987, 148. 218-225. Schlauch, R.M. Colloid free precipitation of heavy metal sulfides. U.S. Patent No. 4,102,784. July 25, 1978. Schwartz, H.M. & Gilchrist, F.M. Microbial interactions with the diet and the host animal. In I.W. McDonald & A.C. Warner (Eds.), Digestion and metabolism in the ruminant. Sydney, Australia: University of New England, 1975. Selwyn, S.C., & Wakerly, D.C. The production of sulphide from sulphated sewage sludges. Journal of Applied Bacteriology. 1954, 17, xx-xxi. Snoeyink, V.L., & Jenkins, D. Water chemistry. New York: John Wiley & Sons, 1980. Sonderegger, J. Butte mineflooding and the Berkeley pit. Berkeley pit re-industrialization and mineral recovery project. Helena, Montana: Headwaters Research Institute, 1987. Spangrude, G.J., Heimfeld, S., & Weissman, I.L. Purification and characterization of mouse hematopoietic stem cells. Science. 1988, 241. 58-62. H O Spormann, A.M., & Thauer, R.K. Anaerobic acetate oxidation to CO2 by Desulfotomaculum acetoxidans. Demonstration of enzymes required for the operation of an oxidative acetyl-CoA/carbon monoxide dehydrogenase pathway. Archives of Microbiology, 1988, 150. 374-380. Stams, A.J.M., Kremer, D.R., Nicolay, K., Weenk, G.H., & Hansen, T.A. Pathway of propionate formation in Desulfobulbus propionicus. Archives of Microbiology. 1984, 139. 167-173. Strayer, R.F., & Tiedje, J.M. Kinetic parameters of the conversion of methane precursors to methane in a hypereutrophic lake sediment. Applied and Environmental Microbiology. 1978, J36, 330-340. Szewzyk, R. & Pfennig, N. Complete oxidation of catechol by the strictly anaerobic sulfate-reducing Desulfobacterium catecholicum sp. nov. Archives of Microbiology. 1987, 147. 163-168. Taras, M.J., Greenberg, A.E., Hoak, R.D., & Rand, M.C. (Eds.). Standard methods for the examination of water and wastewater (13th ed.). New York: American Public Health Association, 1971. Taylor, B.F., & Oremland, R.S. Depletion of adenosine triphosphate in Desulfovibrio by oxyanions of group VI elements. Current Microbiology. 1979, 3, 101-103. Temple, K.L. Syngenesis of sulfide ores: an evaluation of biochemical aspects. Economic Geology. 59., 271-278. Templeton, L.L., & Grady, C.P.L., Jr. Effect of culture history on the determination of biodegration kinetics by batch and fed-batch techniques. Journal Water Polluton Control Federation. 1988, 60, 651-658. Thauer, R.K. Dissimilatory sulphate reduction with acetate as electron donor. Phil. Transactions of Royal Society of London. 1982, B 298. 467-471. Thauer, R.K., & Badziong, W. Dissimilatory sulfate reduction, energetic aspects. In H. Bothe & A. Trebst (Eds.), Biology of inorganic nitrogen and sulfur. New York: Springer-Verlag, 1981. Thauer, R.K., Jungermann, K., & Decker, K. Energy conservation in chemotrophic anaerobic bacteria. Bacteriological Reviews. 1977, 41, 100-180. Ill Toerien, D.F., & Hattingh, W.H.J. Anaerobic digestion I. The microbiology of anaerobic digestion. Water Research. 1969, 3, 385-416. Toerien, D.F., & Maree, J.P. Reflections on anaerobic process biotechnology and its impact on water utilization in South Africa. Water South Africa. 1987, 13, 137-144. Tomizuka, N., & Yagisawa, M. Optimum conditions for leaching of uranium and oxidation of lead sulfide with Thiobacillus ferrooxidans and recovery of metals from bacterial leaching solution with sulfate-reducing bacteria. Bioextractive Applications and Optimization. 321-344. Traore, A.S., Hatchikian, C.E., Belaich, J., & Le Gall, J. Microcalorimetric studies of the growth of sulfate- reducing bacteria: energetics ' of Desulfovibrio vulgaris growth. Journal of Bacteriology. 1981, 145, 191-199. Traore, A.S., Hatchikian, C.E., Le Gall, J., & Belaich, J. Microcalorimetric studies of the growth of sulfate- reducing bacteria: comparison of the growth parameters of some Desulfovibrio species. Journal of Bacteriology. 1982, 149, 606-611. Tsuji, K. & Yagi, T. Significance of hydrogen burst from growing cultures of Desufovibrio vulgaris, Miyazaki, and the role of hydrogenase and cytochrome C3 in energy production system. Archives of Microbiology. 1980, 125, 35-42. Tuttle, J.H., Dugan, P.R., & Randles, C.I. Microbial sulfate reduction and its potential utility as an acid mine water pollution abatement procedure. Applied Microbiology. 1969, 17, 297-302. Tuttle, J.H., Dugan, P.R. , Macmillan, C.B., Sc Randles, C.I. Microbial dissimilatory sulfur cycle in acid mine water. . Journal of Bacteriology. 1969, 97, 594-602. Tuttle, J.H., Randles, C. I. , Sc Dugan, P.R. Activity of microorganisms in acid mine water. I. Influence of acid water on aerobic heterotrophs of a normal stream. Journal of Bacteriology. 1968, 95, 1495-1503. Uphaus, R.A., Grim, D., Sc Cork, D.J. Gypsum bioconversion to sulfur: A two-step microbiological process. Developments in Industrial Microbiology. 1983, 24. 435-442. 112 Van den Berg, L., & Lentz, C.P. Food processing waste treatment Jby anaerobic digestion. Proceedings of the 32nd Industrial Waste Conference. 1977, 252-258. Van den Heuvel, J.C., Beeftink, H.H., & Verschuren, P.G. Inhibition of the acidogenic dissimilation of glucose in anaerobic continuous cultures by free butyric acid. Applied Microbiology and Biotechnology. 1988, 2_9, 89-94. Vosjan, J.H. Respiration and fermentation of the sulphate- reducing bacterium Desulfovibrio desulfuricans in a continuous culture. Plant and Soil, 1975, 43., 141-152. Wang, Y., Suidan, M.T., and Rittman, B.E. Kinetics of methanogens in an expanded-bed reactor. Journal of Environmental Engineering. 1986, 112, 155-170. Ward, D.M., & Winfrey, M.R. Interactions between methanogenic and sulfate-reducing bacteria in sediments. Advances in Aouatic Microbiology. 1985, 3, 141-179. Whitmore, T.N., Lloyd, D., Jones, G., & Williams, T.N. Hydrogen-dependent control of the continuous anaerobic digestion process. Applied Microbiology and Biotechnology. 1987, 26., 383-388. Wichlacz, P.L., & Unz, R.F. A model of ferrous iron oxidation for attached bacteria. in R.W. Lawrence, R.M.R. Branion, & H.G. Ebner (Eds.), Fundamental and applied biohvdrometallurgy. New York: Elsevier, 1986. Widdel, F. Anaerober Abbau von Fettsauren und Benzoesaure durch neu isolierte Arten sulfatreduzierender Bakterien. Dissertation. Gottingen, Germany: University of Gottingen, 1980. Widdel, F., & Kohring, G., & Mayer, F. Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids. III. Characterization of the filamentous gliding Desulfonema limicola gen. nov. sp. nov., and Desulfonema magnum sp. nov. Archives of Microbiology. 1983, 134, 286-294. Widdel, F., & Pfennig, N. A new anaerobic, sporing, acetate-oxidizing, sulfate-reducing bacterium, Desulfotomaculum (emend.) acetoxidans. Archives of Microbiology. 1977, 112. 119-122. Widdel, F., & Pfennig, N. Speculation and further nutritional characteristics of Desulfotomaculum acetoxidans. Archives of Microbiology. 1981, 129. 401- 402. 113 Widdel, F., & Pfennig, N. Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids. I. Isolation of new sulfate-reducing bacteria enriched with acetate from saline environments. Description of Desulfobacter oostaatei gen. nov., sp. nov. Archives of Microbiology. 1981, 129, 395-400. Widdel, F., & Pfennig, N. ' Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids. II. Incomplete oxidation of propionate by Desulfobulbus propionicus gen. nov., sp. nov. Archives of Microbiology. 1982, 131. 360-365. Wieringa, K.T. The formation of acetic acid from carbon dioxide and hydrogen by anaerobic spore-forming bacteria. Antonie van Leeuwenhoek Journal Microbiology Serology. 1940, 6, 251-262. Wilkinson, G.N. Statistical estimations in enzyme kinetics. Biochemistry Journal. 1961, 80., 324-332. Winfrey, M.R., & Zeikus, J.G. Anaerobic metabolism of immediate methane precursors in Lake Mendota. Applied and Environmental Microbiology. 1979, 37., 244-253. Wise, D.L., Leuschner, A.P., & Levy, P.F. Suppressed methane fermentation of selected industrial wastes: a biologically mediated process for conversion of whey to liquid fuel. In Symposium: bioconversion of waste materials to useful industrial products. 197-207, 1985. Wolfe, R.S., and Higgins, L.J. Microbial biochemistry of methane: A study in contrasts. International Review of Biochemistry. 1979, 2JL, 267-353. Wolin, M.J. The rumen fermentation: a model for microbial interactions in anaerobic ecosystems. In M. Alexander (Ed.), Advances in microbial ecology (Vol. 3). New York: Plenum Press, 1977. Wolin, M.J. Metabolic interactions among intestinal microorganisms. American Journal of Clinical Nutrition, 1974, 27, 1320-1328. Wolin, M.J. Interactions between H2-producing and methane- producing species. In H.G. Schlegel, G. Gottschalk, & N. Pfennig (Eds.), Symposium on microbial production and utilization of gases (H2,CH^,CO). Gottingen, West Germany: Akademie der Wissenschaften zu Gottingen, 1976. 114 Wright, R.M., Asce, A.M., & McCarthy, B.J. Fate and transport of metals in a river system. Environmental Engineering. 313-320. Yagasawa, M., Murakami, Y., Kato, 'Y.,. Tomizuka, N., Yamaguchi, M., & Ooyama, J. Recovery of useful metals from a bacterial leaching solution and abatement of wastewater acidity by sulfate reducing bacteria. Journal of the Mining and Metallurgical Institute of Japan. 1977, 93, 447. Yang, S.T., & Okos, M.R. Kinetic study of mathematical modeling of methanogenesis of acetate using pure cultures of methanogens. Biotechnology and Bioengineering. 1987, 3,0, 661-667. Yang, S.T., Tang, I.C., & Okos, M.R. Kinetics of homoacetic fermentation of lactate by Clostridium formicoaceticum. Applied and Environmental Microbiology. 1987, 53(4), 823-827. Zehnder, A.J.B. Ecology of methane formation. In R. Mitchell (Ed.), Water pollution microbiology. (Vol. II). New York: John Wiley, 1978. Zeikus, J.G. Metabolism of one-carbon compounds by chemotrophic anaerobes. Advances in Microbial Physiology. 1983, 24, 215-289. Zeikus, J.G. Chemical and fuel production by anaerobic bacteria. Annual Reviews of Microbiology. 1980, 34, 423-464. Zellner, G., & Winter, Ji Analysis of a highly efficient methanogenic consortium producing biogas from whey. System Applied Microbiology. 1987, 9, 284-292. Zobell, C.E. Ecology of sulfate-reducing bacteria. Producers Monthly Pennsylvania Oil Producers Association. 1958, 22, 12-29. APPENDICES 116 APPENDIX A THERMODYNAMIC CALCULATIONS The calculated free energy change associated with a biologically catalyzed chemical reaction can be used to estimate the true growth yield of the organism involved. The free energy change is usually also positively correlated with the maximum specific growth rate of the organism. The more negative the calculated free energy change, the higher the yield and growth rate can be expected to be. The following Calculations were used to develop the expected free energy changes under standard conditions for the volatile acid oxidations of concern to this research. Data from Thauer et al (1977) were used to calculate the free energy change per each two electrons transferred at unit concentrations of reactants and products and at pH 7.0. Acetate Oxidation Complete oxidation of acetate to carbon dioxide occurs via the following reaction: CH3COO- + SO4-2 — > 2HC03- + HS- (18) 117 This reaction is the sum of the following two half reactions: CH3COO- + 4H20 — > 2HC03- + 4H2 + H+ (44) G°1 = +104.6 kJ/reaction and SO^ + H"*" + 4H2 — > HS + 4H20 (45) G°1 = -151.9 kJ/reaction Because eight electrons are transferred in each reaction, G°,/2e“ = (104.6 - 151.9)/4 = -11.83 kJ • (46) Propionate Oxidation Incomplete oxidation of propionate to acetate occurs via the following reaction: 4CH3CH2COO- + SSO4-2 — > 4CH3COO- + 4HC03- + 3HS- + H+ (19) This reaction is the sum of the following two half reactions: 4CH3CH2COO- + 12H20 — > 4HC03- + 4CH3COO- + 12H2 + 4H+ (47) G°' = +76.1 x 4 = +304.4 kJ/reaction and 3S04- + 3H+ + 12H2 — > 3HS- + 12H20 (48) G°1 = -151.9 x 4 = -455.7 kJ/reaction Because 24 electrons are transferred in each reaction, G°'/2e- (304.4 - 455.7)/12 -12.61 kJ (49) 118 Complete oxidation of propionate occurs via the following reaction: 4CH3CH2COO- + VSO4- — > 12HC03- + VHS- +H+ (20) This reaction is the sum of the following two half reactions: 4CH3CH2COO- + 28H20 — > 12HC03- + 28H2 + 8H+ (50) G°1 = +181.1 x 4 = +V24.4 kJ/reaction and VSO4- + VH+ + 28H2 — > VHS- + 28H20 (51) G°' = -151.9 x V = -1,063.3 kJ/reaction Because 56 electrons are transferred in each reaction, G°'/2e— = (V24.4 - 1,063.3)/28 = -12.10 kJ (52) Butyrate Oxidation Pfennig and Widdel (1981) suggested the following stoichiometry for incomplete oxidation of straight chain fatty acids with even numbers of carbon atoms such as butyrate by SRB: 2CH3CH2CH2COO- + SO4-2 — > 4CH3COO- + HS- + H+ (23) This reaction is the sum of the following two half reactions: 2CH3CH2CH2COO- + 4H20 — > 4cH3COO- + 4H2 + 2H+ (53) G°1 = +48.1 x 2 = +96.2 kJ/reaction and SO4 + H+ + 4H2 — > HS + 4H20 (54) G°1 = -151.9 kJ/reaction 119 Because eight electrons are transferred in each reaction, G°'/2e- = (96.2 - 151.9)/4 = -13.93 kJ (55) The stoichiometry of complete oxidation of butyrate by Desulfobacterium catecholicum was determined to be as follows by Szewzyk and Pfennig (1987) : 2CH3CH2CH2COO" + SSO4"2 — > SHCO3" +5HS" + H+ (23) This reaction is the sum of the following two half reactions: 2CH3CH3CH2COO" + 2OH2O ■— > SHCO3" + 20H2 + 6H+ (56) G°1 = +257.3 x 2 = +514.6 kJ/reaction and SSO4" + 5H+ + 2OH2 — > 5HS" + 20H20 (57) Got = -151.9 x 5 = -759.5 kJ/reaction Because 40 electrons are transferred in each reaction, G°'/2e“ = (514.6 - 759.5)/20 = -12.25 kJ (58) 120 APPENDIX B THE CHALLENGES OF CONTINUOUS CULTURE OF SRB IN SMALL REACTORS Continuous culture of SRB in small reactors presents a variety of challenges to the researcher. A discussion of some of these challenges and some practical suggestions for meeting them are presented below. Air Leaks Continuous cultures of SRB in small reactors are extremely sensitive to leakage of air into the culture vessel. This research used glass beakers having a working volume of about 500 ml and topped with thick black rubber stoppers and was still intermittently plagued by air leakage. The leakage occurred regardless of semicontinuous purging with oxygen-free nitrogen (N2). Soapy water can be used to detect escaping N2. Tubing for medium and N2 transport should be either butyl rubber (Fisher Scientific) or neoprene (Cole Parmer). Only glass tubing connectors should be used. 121 Wall Growth SRB have an affinity for growing on the reactor walls. As one would expect, the problem is more severe at higher dilution rates. Wall growth can be minimized by intermittent scraping of the reactor walls. The scraping must occur every three to six hours at high dilution rates and less frequently otherwise. A scraper can be constructed by attaching an inverted plastic funnel to a stainless steel rod that perforates the top of the. reactor. The portion of the rod outside the reactor can be enclosed in a length of butyl rubber bicycle inner tube. The bottom of the inner tube is fastened to the top of the reactor and the top to a rubber stopper afixed to the upper end of the rod. pH Control The relatively high concentration of H2S in the reactor tends to degrade the performance of pH electrodes fairly rapidly. A double junction electrode (Markson Scientific) should be used. SRB will also grow on the electrode glass bulb creating a microenvironment of relatively high pH. Intermittent vigorous mixing of the reactor contents using the wall scraper is necessary to prevent excess neutralizing chemical (acid) from being added to the reactor by the pH control system and sterilizing the system. 122 Substrate Concentration Substrate concentrations are effectively decreased by the diluting effect of acid additions (0.1 N HCl) and effectively increased by water losses that occur during medium preparation and storage during a run. A relatively low acid concentration (0.1 N) should be used, however, to prevent large changes in pH during pH control operations. The distilled water used to make up the medium should be boiled briefly under oxygen-free N2 to remove dissolved oxygen. Although liquid heating can be halted at the first sign of boiling, some liquid loss will occur during both boiling and during removal from the autoclave. Furthermore, university "distilled water" will contain varying amounts of sulfate in any case as the following distilled water sulfate concentration data indicate: Date Sulfate concentration, mo/1 10/18/88 10/24/88 12/8/88 12/26/88 12/26/88 12/26/88 12/26/88 2 3 13 9 62 35 2 43 3 3 9 8 1/20/89 2/2/89 2/2/89 3/2/89 3/2/89 123 Temperature Control SRB growth rates are very sensitive to temperature. The reactor should be submerged to the extent feasible in a constant temperature water bath. A temperature controller with an accuracy of ±0.01° C should be used (Markson Scientific). Medium Transfer Small reactors and low growth rates combined require very low medium pumping rates. Master-Flex pumps with No. 13 neoprene tubing (Cole Parmer) can produce the low flow rates (5 ml/hr) required. If a complex medium containing much iron is used, a precipitate will form which will clog small bore tubing unless a glass fiber filter is installed at the influent end of the tubing. Medium Additives Although EDTA addition of SRB media has been suggested to prevent iron limitation, EDTA is reportedly toxic to some SRB (Pfenning et al, 1981). Care must also be taken in choosing an appropriate reducing agent to scavenge oxygen in the medium. Sodium thioglycolate is toxic to some SRB (Pfenning et al, 1981) as is ascorbic acid (Brock et al, 1984). Sodium dithionite at 30 mg/1 was not toxic to propionate-utilizing SRB. It should, however, be added to an essentially oxygen-free medium, as the products of ( I 124 reaction of reducing agents with oxygen may be toxic to some bacteria (Ljungdahl and Wiegel, 1986) . 125 APPENDIX C DATA TABULATIONS o tr pj 126 Table 10. Flow Rate Data Run No. Average flow Medium Acid rate, Suma ml/hr Effluent Dilution rate, hr-1 1.1 10.05 - — — 2.1 6.91 _b - — — 3.1 11.31 _b — — — 3.2 15.23 _b - - — 4.1 12.93 _b — — — 4.2 11.75 _b - — — 5.1 5.93 _b - - — 5.2 5.54 _b - - — 6.1 7.01 _b - - - 6.2 6.32 _b — — — 7.1 12.94 _b - — — 7.2 12.09 _b - - - 8.1 14.96 4.29 19.25 19.25° — 8.2 14.10 7.25 21.35 21.66 0.0442 9.1 15.30 7.73 23.03 23.04 0.0474 9.2 15.56 6.25 21.81 21.66 0.0442 10.1 23.45 7.21 30.66 30.55 0.0629 10.2 24.07 5.27 29.34 29.68 0.0606 11.1 25.15 - - - - 11.2 24.25 4.95 29.20 29.34 0.0599 12.1 21.47 8.35 29.82 29.71 0.0611 12.2 23.23 - - ■ - - 13.1 20.00 8.15 28.15 27.96 0.0575 13.2 17.55 1.15 18.70 18.74 — 14.1 19.83 6.74 26.57 26.26 0.0544 14.2 18.22 1.61 19.83 19.58 - 15.1 13.89 6.12 20.01 19.79 0.0410 15.2 12.37 5.04 17.45 17.44 0.0357 16.1 21.24 8.94 30.18 29.89 0.0613 16.2 19.11 7.08 26.19 26.02 0.0539 17.1 - - - - - 17.2 4.42 2.48 6.90 6.93 0.0143 18.1 5.52 2.98 8.50 8.74 0.0179 18.2 5.38 3.06 8.44 8.63 0.0179 Medium plus acid Flow rate not measured Estimate based on sum 127 Table 11. Medium Concentration Data Run Average concentration, mg/1 No. Acetate Propionate Butyrate Sulfate Targeta 0 2,700 0 3,330 1.1 0 3,360 0 _b 2.1 0 3,240 0 _b 3.1 0 2,850 0 _b 3.2 0 2,850 0 _b 4.1 0 2,980 0 2,900 4.2 0 2,980 0 2,900 5.1 0 2,730 0 3,270 5.2 0 2,730 0 3,270 6.1 0 2,730 0 3,310 6.2 0 2,730 0 3,310 7.1 0 2,710 0. 2,960 7.2 0 2,710 0 2,960 8.1 0 2,750 0 2,960 8.2 0 2,750 0 2,960 9.1 0 2,700 0 3,130 9.2 0 2,700 0 3,130 10.1 0 2,850 0 2,790 10.2 0 2,850 0 2,790 11.1 0 2,700 0 2,710 11.2 0 2,700 0 2,710 Targeta 1,900 2,700 , 2,000 3,330 12.1 1,780 2,760 2,080 2,930 12.2 1,780 2,760 2,080 2,930 13.1 1,900 2,840 2,120 2,910 13.2 1,900 2,840 2,120 2,910 Targeta 1,900 2,700 2 ,000 3,790 14.1 1,880 2,830 2,110 3,390 14.2 1,880 2,830 2 , H O 3,390 15.1 1,890 2,780 2,070 3,650 15.2 1,880 2,830 2,100 3,830 Targeta 0 2,700 0 3,790 16.1 0 3,070 0 4,310° 16.2 0 3,070 0 4,310° 17.1 0 2,730 0 3,800 17.2 0 2,730 0 3,800 18.1 0 2,970 0 _b 18.2 0 2,970 0 _b a Theoretical concentrations for the following runs based on measured chemical weights and water volume. b Sample not analyzed for parameter. c Estimate. 128 Table 12. Effluent Concentration Data Run No. Acetate Average concentration, mg/1 Propionate Butyrate Sulfate 1.1 490 3,200 0 _C 2.1 800 2,420 0 3,970 3.1 1,410 100 0 70 3.2 1,530 70 0 _c 4.1 1,670 80 0 90 4.2 1,950 50 0 90 5.1 1,390 130 0 40 5.2 1,320 160 .0 50 6.1 1,320 90 0 100 6.2 1,310 60 0 60 7.1 1,370 140 0 80 7.2 1,410 140 0 140 8.1 1,200 460 0 . 390 8 .2 1,360 H O 0 20 9.1 1,360 100 0 30 9.2 1,390 220 0 190 10.1 1,210 800 0 940 10.2 1,040 860 0 1,100 11.1 _a _a _a _a 11.2 1,130 840 0 1,230 12.1 2,620 330 ■ 1,270 280 12.2 _a _a _a _a 13.1 2,760 240 1,280 210 13.2 _a _a _a _a 14.1 2,590 590 1,400 1,350 . 14.2 _a _a _a _a 15.1 2,880 220 1,170 1,220 15.2 2,940 120 1,140 1,200 ' 16.1 1,360 370. 0 1,440 16.2 1,060 670 0 1,740 17.1 _b _b _b _b 17.2 _b _b _b 1,070 18.1 1,240 30 0 _c 18.2 1,230 40 0 _c a Data not presented because reactor did not reach steady- state due to mechanical difficulties. k Sample lost due to refrigerator failure. c Sample not analyzed for parameter. 129 Table 13. Calculated Effective Influent Concentrations Run No. Effective influent Acetate Propionate concentration, Butyrate mg/1 Sulfate 8.1 0 2,140 0 2,300 8.2 0 1,790 0 1,930 9.1 0 1,790 0 2,080 9.2 0 1,940 0 2,250 10.1 0 2,190 0 2,140 10.2 0 2,310 0 2,260 11.1 - - - - 11.2 0 2,230 0 2,240 12.1 1,290 1,990 1,500 2,120 12.2 ■ , - - - 13.1 1,360 ■ 2,030 1,520 2,080 13.1 — - 14.1 1,420 2,140 1,590 2,560 14.2 - - - - 15.1 1,330 1,950 1,450 2,560 15.2 1,330 2,010 1,490 2,720 16.1 0 2,180 0 3,060b 16.2 0 2,250 0 3,17 Ob 17.1 - - - - 17.2 0 1,740 0 2,420 18.1 0 1,880 0 - 18.2 0 1,850 0 a Diluted medium concentration based on ratio of medium and effluent flow rates. b Estimate (see Table 11). n 130 Table 14. Acetate Production Run Substrate oxidation rate, mM/hr No. Propionate Butyrate Total Acetate production rate, mM/hr 8.1 0.443 0 0.443 0.391 8.2 0.498 0 0.498 0.499 9.1 0.533 0 0.533 0.531 9.2 0.510 0 0.510 0.510 10.1 0.581 0 0.581 0.626 10.2 0.589 0 0.589 0.523 11.1 - - - - 11.2 0.558 0 0.558 0.562 12.1 0.675 0.078 ■ 0.753 0.669 12.2 - -■ - - 13.1 0.685 0.077 0.762 0.663 13.2 — — — - 14.1 0.557 0.027 0.584 0.560 14.2 - - - - 15.1 0.469 0.064 0.533 0.520 15.2 0.451 0.070 0.521 0.476 16.1 0.740 0 0.740 0.688 16.2 0.563 0 0.563 0.467 17.1 - - - - 17.2 - - ' - . - 18.1 0.221 0 0.221 0.184 18.2 0.214 0 0.214 0.180 131 Table 15. Sulfate Reduction Run Substrate oxidation rate, mM/hr No. Propionate Butyrate Total Sulfate reduction rate, mM/hr 8.1 0.443 0 0.443 0.383 8.2 0.498 0 0.498 0.431 9.1 0.533 0 0.533 0.492 9.2 0.510 ' 0 0.510 0.464 10.1 0.581 0 0.581 0.382 10.2 0.589 0 0.589 0.358 11.1 - - - - 11.2 0.558 0 0.558 0.308 12.1 0.675 0.078 0.753 0.569 12.2 - - - - ■ 13.1 0.685 0.077 0.762 0.544 13.2 — — — - 14.1 0.557 0.027 0.584 0.331 14.2 - - - - 15.1 0.469 0.064 0.533 0.276 15.2 0.451 0.070 0.521 0.272 16.1 0.740 0 0.740 0.504a 16.2 0.563 - 0.563 0.387a 17.1 - - - - 17.2 - - - 0.097 18.1 0.221 0 0.221 - 18.2 0.214 0 0.214 a Estimate (see Table 13) 132 Table 16. Acid Neutralization Run Substrate oxidation rate, mM/hr No. Propionate Butyrate Total Acid neutralization rate, mM/hr 8.1 0.443 0 0.443 0.429 8.2 0.498 0 0.498 0.725 9.1 0.533 0 0.533 0.773 9.2 0.510 0 0.510 0.625 10.1 0.581 0 0.581 0.721 10.2 0.589 0 0.589 0.527 11.1 - - - - 11.2 0.558 0 0.558 0.495 12.1 0.675 0.078 0.753 0.835 12.2 - - - - 13.1 0.685 0.077 0.762 0.815 13.2 — — — — 14.1 0.557 0.027 0.584 0.674 14.2 - - - - 15.1 0.469 0.064 0.533 0.612 15.2 0.451 0.070 0.521 0.504 16.1 0.740 0 0.740 0.894 16.2 0.563 0 0.563 0.708 17.1 - - - - 17.2 - - - 0.248 18.1 0.221 0 0.221 0.298 18.2 0.214 0 0.214 0.306 133 Table 17. Calculated Hofstee Plot Variables Run No. Dilution rate (D), hr-1 Dilution rate/concentration (D/S), 1/mg.hr Propionate Butyrate 8.1 - - - 8.2 0.0442 0.000402 - 9.1 0.0474 0.000474 - 9.2 0.0442 0.000201 - 10.1 0.0629 0.0000786 - 10.2 0.0606 0.0000704 - 11.1 - - - 11.2 0.0599 0.0000713 - 12.1 0.0611 0.000185 0.0000481 12.2 - - - 13.1 0.0575 0.000240 0.0000449 13.2 — — - 14.1 0.0544 0.0000922 0.0000388 14.2 - - _ - 15.1 0.0410 0.000186 0.0000350 15.2 0.0357 0.000298 0.0000313 16.1 0.0613 0.0000915 — 16.2 0.0539 0.0000869 - 17.1 - - - 17.2 0.0143 - - 18.1 0.0179 0.000597 - 18.2 0.0179 0.000448 - 134 Table 18. Calculated Hanes Plot Variables Run No. Substrate concentration (S), mg/1 Propionate Butyrate Concentration/dilution rate (S/D), mg.hr/1 Propionate Butyrate 8.1 - - - - 8.2 H O - 2,490 - 9.1 100 - 2,110 - 9.2 220 - 4,980 - 10.1 800 - 12,700 - 10.2 860 - 14,200 - 11.1 - - - - 11.2 840 — 14,000 - 12.1 330 1,270 5,400 20,800 12.2 - - - - 13.1 240 1,280 4,170 22,300 13.2 — — — - 14.1 590 1,400 10,800 25,700 14.2 - - - - 15.1 220 1,170 5,370 28,500 15.2 120 1,140 3,360 ■ 31,900 16.1 370 — 6,040 — 16.2 670 - 10,900 - 17.1 - - - - 17.2 - - - - 18.1 30 - 1,680 - 18.2 40 2,230 > 135 Table 19. Biomass Characteristics Run No. Cell concentration, IO10 cells/1 Average cell area, sg micron Average cell length, micron Average cell width micron 8.1 - - - - 8.2 21.00 0.670 1.37 0.820 9.1 38.75 1.061 1.56 1.02 9.2 - - - - 10.1 33.52 1.047 1.56 0.971 10.2 18.92 0.870 1.39 0.924 11.1 - - - - 11.2 18.75 0.859 1.39 0.870 12.1 46.96 0.806 1.39 0.866 12.2 - - - - 13.1 55.56 0.808 1.33 0.858 13.2 — — — — 14.1 31.90 1.071 1.55 1.03 14.2 - - - - 15.1 36.32 0.915 1.40 0.931 15.2 32.03 0.920 1.46 0.936 16.1 37.28 0.856 1.42 0.894 16.2 18.64 0.905 . 1.43 0.932 17.1 - - T- - 17.2 - - - - 18.1 51.03 0.729 1.31 0.832 18.2 60.75 0.663 1.27 0.809 136 Table 20. Calculation of Yield Analysis Plot Variables Run Space time S0 - s, Average cell weight, lo-io mg (So - S)/X, No. (1/D) , hr mg/1 mg/mg 8.1 - - - - 8.2 22.6 1,680 1.31 61.1 9.1 21.1 1,690 1.31 33.3 9.2 22.6 1,720 - - 10.1 15.9 1,390 1.31 31.7 10.2 16.5 1,450 1.31 58.5 11.1 - - - - 11.2 16.7 1,390 1.31 56.6 12.1 16.4 1,660 1.29 27.4 12.2 - - - - 13.1 17.4 1,790 1.29 25.0 13.2 — — - - 14.1 18.4 1,550 1.29 37.1 14.2 - - - - 15.1 24.4 1,730 1.29 36.9 15.2 28.2 1,890 1.29 45.7 16.1 16.3 1,810 1.31 37.1 16.2 18.6 1,580 1.31 64.7 17.1 - — - - 17.2 69.9 - - - 18.1 55.9 1,850 0.927 39.1 18.2 55.9 1,810 0.927 32.1 MONTANA STATE UNIVERSITY LIBRARIES 3 762 101 3471 4 (£ ~ -