Loading...
HomeMy WebLinkAboutFuel Cell Power Plants in Rural Alaska Final Report 1983Ai sn FUEL CELL POWER PLANTS IN RURAL ALASKA State of Alaska Department of Transportation and RES Zz Report No. AK-RD-84-02 PROJECT EVALUATION ROUTING SLIP Distributed by Research Section Division of Planning and Programming Department of Transportation and Public Facilities State of Alaska Location: 101 Duckering Bldg. Mailing Address: University of Alaska 2301 Peger Road Fairbanks, Alaska Fairbanks, Alaska 99701 (907) 479-2241 FUEL CELL POWER PLANTS IN RURAL ALASKA FINAL REPORT by Je 3B. Malosh, Ph.D: , P.E. Associate Professor Department of Mechanical Engineering University of Alaska Fairbanks, Alaska 99701 April 1983 Prepared for: Department of Transportation and Public Facilities Division of Planning and Programming Research Section 2301 Peger Road Fairbanks, Alaska 99701 AK-RD-84-2 ABSTRACT On the basis of fuel efficiency alone, the methanol fueled phosphoric acid fuel cell (PAFC) is a very attractive replacement for the diesel electric generator, especially in the bush regions of Alaska. However, because of the transportation costs for liquid fuel to the bush combined with the lower heating value of methanol, the PAFC looses this advantage and produces electricity that in some instances is more costly than the diesel generator. Although the PAFC is at the highest state of development of all fuel cell power plants, it is still not a commercially mature technology. The present cost of a PAFC power plant is on the order of ten times the price of an equivalent diesel electric generator. There is also no large body of published, long term data on fuel cells of any type larger than 1 kw from which an accurate assessment of reliability, maintenance and operating costs can be made. Considering this and the lack of electrical production cost advantage, the evaluation of the methanol fueled PAFC for bush applications should be suspended until more operational data is made public and units are commercially available. ACKNOWLEDGEMENTS I would like to thank Mr. Peter R. Voyentzi of Energy Research Corporation and Mr. Thomas J. Lamb of Engelhard Industries for their cooperation in responding to my questionnaire and providing supplemental information on their products. I would also like to thank Mr. Robert Dempsey of Alaska Interior Resources, Inc., and Prof. Michael Economides of the University of Alaska Fairbanks Petroleum Engineering Department for their assistance in obtaining data on methanol production. Thanks are also in order for Mr. Craig Helmuth of the Fairbanks North Star Borough Community Research Center, Mr. John E. Butts of Envirosphere Company and Mr. H. W. Slater of Battelle Pacific North- west Laboratories for their interest and the information they provided for this project. In addition, I would like to thank Ms. Rheba Dupras of the UAF Library, Ms. Marilynn Griffin of the UAF Engineering Experiment Station and Mr. Lee Leonard of the Alaska Department of Transportation and Public Facilities for their valuable assistance in this project. TABLE OF CONTENTS ABSTRACT ACKNOWLEDGEMENTS TABLE OF CONTENTS Introduction Summary Background Scope of Work Literature Search Consultation with Manufacturers Analysis of Responses Continuing Research Program Engineering Economic Analysis Implementation FIGURES AND TABLES Figure 1: Schematic of an Integrated Phosphoric Acid Fuel Cell Figure 2: Schematic of proposed 2.5KW PAFC test set-up Table 1: Summary of cost estimate for initiation of proposed 2.5 Kw PAFC test program Table 2: Physical quantities measured during PAFC Performance Tests Table 3: Cost Comparison of Electricity Produced by PAFC and Diesel- Electric Generator APPENDIX A Letters from Manufacturers REFERENCES -ii- 10 LS 18 21 26 20 19 20 24 27 34 INTRODUCTION Fuel cells, because of their high energy-conversion efficiency offer an attractive alternative to the diesel-electric generators which are presently used in rural Alaska. The objective of this study was to evaluate the feasibility of using a methanol-fueled phosphoric acid fuel cell (PAFC) as an electric power source to replace the diesel-electric generators now used in many rural Alaskan state build- ings. In pursuance of this goal, this study addresses the safety, operation, reliability, maintenance and cost aspects of the fuel cell power plant that are generic to the Alaskan environment. This task was accomplished with an extensive literature search and consultation with fuel cell manufacturers. A program and cost estimate for a continuing research program is also provided. In addition, based on the best available data, an engineering economic analysis of a PAFC power plant in rural Alaska is presented. SUMMARY The major result of this study is that the methanol fueled PAFC in its present form produces electricity in the bush that is 14% more costly than a diesel electric generator. This result comes out of the cost analysis comparing a similar sized PAFC and diesel unit (100 kw) operating at 100% capacity. At 30% capacity, the PAFC produced electricity that cost 15% less than the diesel generator. The reason lies in the fact that methanol has less than 1/2 the heating value of diesel fuel on a volume basis. The cost per heating unit for methanol at the refinery is about the same as #2 diesel oil but there is a $1 per gallon cost in shipping liquid fuel to the bush communities. This raises the cost of methanol per heating unit 66% higher than diesel fuel. The higher efficiency of the PAFC does not overcome this disad- vantage. To further illustrate this point, if the shipping charge is neglected, the PAFC produces electricity that is more than 20% less expensive than the diesel at all capacities. There are several things that must occur if the cost disadvantage imposed on the PAFC by the methanol fuel is to be overcome. (a) The PAFC must become very competitively priced with diesel generators. (b) The efficiency of the PAFC must increase. (c) The price per heating unit of methanol must become much less than diesel fuel. Several alternative systems are being developed that may avoid this problem. A PAFC plant that uses diesel fuel is being developed but the diesel fuel-hydrogen reformer technology is not at the same level as the methanol systems. The molten carbonate fuel cell has a 25% higher efficiency than the PAFC but is at least five years away from commercialization. Other results of this study are: 1. The PAFC power plant has not reached the commercialization stage yet. It is still in the R&D stage. The unit costs of $500 to $800 per kw are projections that may not be approached for several years. There is presently only one manufacturer willing to sell a PAFC which is a bench top -unit and the price is $25,000 per kw. 2. There is presently no published data oa long term PAFC power system durability. Projections of stack life are based on lab tests. There is no evidence that a semi-commercial PAFC has performed any longer than two years in a power generating capacity. Apparently this data is being generat- ed now in fuel cell development programs sponsored by the Department of Energy (DOE) and EPRI. This is a further indication that commercialization of the PAFC has not been attained. 3. There were no serious safety problems found other than those associated with handling a volatile fuel and phosphoric acid. All PAFC require periodic replenishment of the elec- trolyte. 4. The PAFC power plant will require a microprocessor control system. A logical control is necessary to prevent such problems as_ stack over-temperatures, cell inversion on start-up and also control reformer output under transient conditions. The PAFC responds quickly to transient loads but the reformer does not. A 5. The only agent that will poison the PAFC is sulfur which can be removed during the production of the methanol. 6. A continuing research program of three man-months duration would cost $110,000. This includes a manually controlled PAFC, a data logger and personnel costs. A computer controlled unit would increase the cost to $230,000. Considering the results of the cost analysis and the scarcity of durability data, the evaluation of the methanol fueled PAFC for the bush should be suspended until more operational data is published and the units are commercially available. BACKGROUND Simply described, the fuel cell is a continuously fueled battery that produces direct-current (dc) electricity by a catalyzed chemical reaction. It is like a galvanic combustion engine but has fewer moving parts and is relatively quiet. However, unlike a conventional combustion engine, such as the diesel, the fuel cell has a much higher thermal efficiency. This increased efficiency is a result of the electrochemical nature of the power generation process in the fuel ‘cell. The thermal efficiency of the diesel engine is limited by the ‘ Carnot cycle. Gosia? The fuel cell can be powered by any hydrogen-based fuel. These include fuels derived from both renewable (ethanol, methanol, biogas) and non-renewable (coal, natural gas, fuel oil, gasoline, methanol) resources. A schematic of an integrated PAFC is shown in Figure l. The methanol driven phosphoric acid fuel cell (PAFC) is the most highly developed system presently available. It has equivalent inltlax” Ser. higher energy conversion efficiencies compared to the diesel-electricity generator at both part- and full-load. The PAFC is an attractive replacement for the diesel-electric generator at rural facilities where electric power is generated on-site. The diesel-electric generator is currently the only economically feasible method of providing electric power in rural Alaska. The initial purchase cost of diesel-electric plants varies from $700/kw Fuel Hydrogen rich Natural Gas ees Fuel Processor Fuel Cell Power ~ Phew doc Diesel Fuel (Reformer) Section (Stack) cue) JD-4 Nea ‘ Ethanol Process Heat Useable Waste Heat Figure 1: Schematic of an Integrated Phosphoric Acid Fuel Cell (PAFC) Power Plant. for small units (12 kw) to $100/kw for large units (1000 kw). The projected cost of a PAFC power plant built with available mass produc- tion technology is $500 to $800/kw. A *PAFC power plant cannot presently be purchased for these costs. This means that based on purchase cost, the PAFC could be economically feasible for lower power applications (30 kw), such as cottage industries, rural schools, or other public facilities. The cost of producing electric power is determined by the overall energy conversion efficiency of the power plant. The full-load efficiency of a diesel-electric generator is about 25% under ideal conditions. If water jacket waste heat recovery is included, the efficiency rises to about 60%, again under ideal conditions. The full-load efficiency of a PAFC power plant , also under ideal conditions, is about 40%; with waste heat recovery this rises to 75%. In economic terms this means that theoretically, electricity produced by a diesel-electric generator at a cost of $0.30/kw hr would cost only $0.19/kw hr when produced by a PAFC power plant, assuming an equivalent capital and maintenance cost and cost per fuel heating value for methanol and diesel fuel. On the basis of full-load perfor- mance then, the PAFC power plant appears to be very competitive with the diesel-electric plant. The part-load energy conversion efficiency of a diesel-electric plant is much less than the full-load value. This is because diesel- electric plants must be sized for peak demand and cannot be operated at much less than 60% load without eventually suffering mechanical failures. Since average power consumption in rural facilities aver- ages only 30 to 40% of full load, a dummy load is often required to keep the plant adequately loaded for non-detrimental operation. This significantly reduces the part-load energy conversion efficiency of the diesel-electric plant in some applications. This problem can be surmounted by parallel operation of smaller generating units that come on-line as load demands increase. However, multiple unit systems are more complex and expensive, factors which are to be avoided if Wit waste HEAT RECOVERY EFFICIENCY OF Diese PLANT Ause Goes uP Ay possible in rural applications. The part-load efficiency of the PAFC remains near that of full- load efficiency over a greater portion of the load range than does the diesel-electric plant. The PAFC plant will operate at peak efficiency down to 30% of full-load capacity. Below this point, the efficiency will decrease to 15% to 20%. Comparing these figures to the diesel-electric plant's part-load performance, the PAFC plant is obviously the superior system in terms of part-load economy. SCOPE OF WORK Although the PAFC power plant appears to be an economical replacement for the diesel-electric generator in rural applications, there are a number of potential problem areas where the behavior of the fuel cell, when used as a small power plant, may render the PAFC unsuitable in rural Alaska. The major areas of concern are outlined below. Te Safety ix Explosion potential of the system B. Possibility of catalyst bed melt-down Cc. Toxicity of catalyst chemicals II. Operations A. Low temperature operation B. Long, cold-start time requirements Gs Load swing response time III. Reliability As Catalyst life B. Catalyst poisoning susceptibility _C. Heat exchanger and fuel reformer reliability IV. Maintenance and repair Ae Personnel skill level requirements B. Maintenance costs (labor and parts) Cc. Repair costs (labor and parts) D. Catalyst deterioration during down-time The scope of this work addresses all of the points listed above. The seriousness of these areas is identified along with new problem areas that were uncovered during the study. An answer or means of obtaining an answer is provided for each question in the context of the rural power plant applications. This task was accomplished by the following methods: 1 An extensive literature search covering both military and civilian usage of fuel cell power plants. Consultation via personal phone calls and question- naires with fuel cell component and systems manufac- turers. The on-site visit to a typical rural Alaskan power plant suggested in the proposal document was not made. However, a visit to the fuel cell manufacturers to observe a functioning PAFC power plant was suggested as a method to obtain more information on the PAFC operating characteristics. Both Energy Research Corporation (ERC) and Engelhard can provide a demon- stration given sufficient notice. This will be consid- ered as part of the continuing research phase. In addition, the following items are presented: 1. A program and cost estimate for continuing research on fuel cell power plants. This includes the development of a set of criteria needed to set up a small scale PAFC experimental test facility. An engineering cost analysis of a full scale applica- tion of a PAFC power plant in rural Alaska. LITERATURE SEARCH The literature search was performed on the Lockheed DIALOG system using the keywords "methanol", "phosphoric acid", "fuel cell", "electrochemical" and "power". The Electric Power Data Book, NTIS, Compendex, SCISEARCH and SSIE current research data bases were searched and approximately 800 citations were listed. These abstracts have been analyzed and the pertinent and timely citations are described below. The literature search produced no data on the application of the PAFC in rural Arctic or Antarctic environments, in fact, there were no citations of any long-term applications of the PAFC in any utility application. There were also no citations concerned with the safety, operations or maintenance aspects of the PAFC power plant. However, there were citations pertaining to cost and reliability but they were based on projected laboratory test results . (1), (2), (3) Current citations generally pertain to PAFC stack material dura- bility and life problems, development of new electrolyte and catalyst base materials, cell module and fuel condition development. There are three major participants in the current effort to produce a commercially viable PAFC fuel cell power plant (4). They are United Technologies Corporation (UTC), Westinghouse Electric Cor- poration/Energy Research Company (ERC), and Engelhard Industries. Their major goal is decreasing cost and increasing reliability. The longstanding barrier to the attainment of these goals has been stack materials durability and life (5), (6), (7), (8). All three organiza- tions are improving their cell stack life but ERC presently has the longest lived stack in semi-commercial form (8). This was a small 2.1 kw unit that was satisfactorily tested in the laboratory for two years (18,000 hours). At the present, however, there is no published data on long-term durability tests of PAFC stacks of substantial size (9). It seems that this data is presently being generated. * See references. Much of the current research on electrolytes and catalysts is concerned with improving the durability and decreasing the costs of catalysts, improving PAFC performance with new electrolytes or improv- ing durability of PAFC components*. This abundance of published research on the major PAFC components and the lack of data on life, reliability and cost strongly suggest that the PAFC power plant is not commercially mature. In addition to the the firms mentioned above, there are several other organizations participating in the development of the PAFC. Universal Oil Products (UOP) and Exxon are developing new catalysts solutions. The Department of Energy (DOE), the Electric Power Research Institute (EPRI), UTC and Consolidated Edison are collaborat- ing in a 4.8 mw PAFC demonstration project in New York City. Other major organizations participating in PAFC developments are the Tennessee Valley Authority (TVA), the Institute of Gas Technology (IGT), Argonne National Laboratory (ANL), and Brookhaven National Laboratory (BNL). However, the majority of funds available for this work has come from NASA, DOE, and EPRI. Although the federal govern- ment through DOE and NASA probably spend the most money on fuel cell development, EPRI is by far the best source for practical engineering information on fuel cells. In the course of this study, several other organizations have been helpful in providing information. The Fairbanks North Star Borough Community research Center published excerpts from the Governor's Railbelt Electric Power Alternative Study which contained economic data on fuel cell power plants. This study was performed by Battelle-Northwest. Following this up, it turns out that Electric Bond and Service Company (EBASCO) provided the data to Battelle under a subcontractor arrangement. In conversations with EBASCO, they were very cooperative and will provide a detailed description of the basis for their economic analysis of fuel cell power plants. * See reference (11) to (15) for typical examples. CONSULTATION WITH MANUFACTURERS Contact with the three fuel cell manufacturers were made. These companies are: 1. Energy Research Corporation (ERC) 2. Engelhard Company 3. United Technology Company (UTC) These companies were asked to respond to questions drawn from the problem areas described in the proposal. UTC is building the 40 kw natural gas fired PAFC power plant for the United States Department of Energy, but: they were not interested in supplying components or systems to anyone else at this time and consequently were not asked to respond to the questions. The problem areas, the corresponding questions posed to ERC and Engelhard and their responses are listed below. Copies of the full text of their responses are provided in Appendix A. Ta Safety A. Explosion potential of the system: What is the explosion potential of the system? Can an equipment malfunction produce an explosion or fire in hydro- gen generators or fuel cell stack? ERC: The U.S. Army has required a safety analysis. Safety features have been built into the set and no explosion hazard exists. Hydrogen is utilized as it is produced. There is no appreciable pressure of hydrogen in the system. ENGELHARD : The gases of reaction are always separated on the inlet side of the fuel cell. In case of hydrogen leak, hydrogen detectors are normally utilized and an alarm will sound before the explosive limit is reached. In case of interreaction within the cell, only very small quantities of reactant gases are in the stack at any - 10 - Es one time, and this would not lead to a _ potentially dangerous situation. B. Possibility of catalyst bed melt-down: Will an equipment malfunction produce catalysis melt-down in either the hydrogen generator or fuel cell stack? ERC: Over-temperature protection is also built in; no "meltdown" can occur. ENGELHARD: In an unusual situation a meltdown might occur; however, the microprocessor monitors certain temperatures and an automatic shutdown would occur prior to meltdown. C. Toxicity of catalyst chemicals: What is the toxicity of the catalysis chemicals? ERC: The catalysts are non-toxic and are also not accessible to the user. The fuel cell stack utilizes platinum as a catalyst whereas the reformer utilizes copper and zinc. ENGELHARD: The toxicity of the precious metal catalysts is nil. In the reformer standard catalysts employ copper and zinc which can be toxic when burned. Operations A. Low temperature operation: What problems would you foresee in operating the PAFC power plant at Arctic temperatures (-60°F)? ERC: The MIL Spec requires -65°F operation. Cold temperature operation is a matter of heating the system’ to Sie operating temperature. ENGELHARD: Arctic temperatures could cause the following redesigns from the standard: a) Special coolant liquid b) Redesigned start-up burner c) Auxiliary heaters for standby conditions d) Special start-up or hybrid batteries e) Special insulation None of these appear to be beyond the present state of the art. B. Long, cold-start time requirements: What factors determine the cold start-up time requirements of the PAFC plant? ERC: The unit we presently are producing for the U.S. Air Force will start in 1/2 hour at -25°F and one hour at -60°F. The system is heated up to starting temperature (approximately 250°F) by burning methanol. ENGELHARD: Start-up time depends upon: a) Ambient temperature (-60°F is not beyond reason) b) Availability of auxiliary power c) Use of liquid coolant as warm up medium d) Use of batteries e) Keeping unit in standby Cc. Load swing response time: How does the PAFC respond to transient loads? ERC: It responds automatically. - 12.- III. ENGELHARD: Fuel cells follow very well with no restrictions on load change. Fuel pump to reformer is driven by fuel cell stack voltage. Reliability A. Catalyst life: How long will the catalyst last under normal operation? ERC: Over 10,000 hours for the reformer catalyst and over 30,000 for the PAFC stack catalyst. ENGELHARD: Our catalyst will last a minimum of 40,000 hours with degradation of less than 10%. B. Catalyst poisoning susceptibility: What materials in the fuel will poison the catalyst and will it recover from the poisoning? ERC: Chlorides and sulfur, however, methanol is a very pure chemical in commerce as it is produced with catalysts that would suffer poisoning if the starting chemicals (usually natural gas) were not pure to begin with. ENGELHARD: Aromatics and sulfur bearing compounds will poison the catalyst and thus is irreversible. Higher alcohols will also poison the catalyst, but this is a reversible situation. C. Heat exchanger and fuel reformer reliability: What is the current life expectancy of the heat exchanger and fuel reformer modules? ERC: We estimate approximately five years. aatge= IV. ENGELHARD: The reformer and heat exchangers could expect a nominal 20 year life. Maintenance and repair: A. Personnel skill level requirements: B. Maintenance costs (labor and parts): What level of skill is required for personnel performing maintenance and repairs on the PAFC plant? ERC: U.S. Army levels. There are very few working parts and only three continually moving parts in the Army set, i.e., the air blower for the stack, the small burner blower and the fuel pump. ENGELHARD: Initially it is anticipated that a skilled technician with instrument and electrical background would be required. As manufacturing techniques and design are improved it is anticipated that a lesser qualified level would be required. Cc. Repair costs (labor and parts): What are the current replacement costs for the major parts of the PAFC plant? ERC The parts which may require replacement are the fuel pump and fuel injector which are both automotive components and the burner air fan and air blower motor. ENGELHARD: Replacements are not available at this time. As soon as firm manufacturing costs are available, replacement prices will be furnished. — M4 D. Catalyst deterioration during down-time: Does the catalyst deteriorate during down time? Can this be prevented or controlled? ERC: The catalyst does not appear to deteriorate during down time, however, both the stack and the reformer have to be sealed during down time from the atmosphere either automatically or manually. ENGELHARD: The catalyst does not deteriorate during down time. Both ERC and Engelhard were also asked to quote a price on a 2- 5kw PAFC plant. ERC can supply a 2 kw manually controlled system which incorporates a methanol reformer, control panel and fuel cell stack for $49,500. They can also supply a complete microprocessor controlled power plant for $175,000. Engelhard does not presently have firmly established prices on PAFC systems. As indicated in their letter (Appendix A) they are in a R&D program for the PAFC and consequently are not currently in a position to produce fuel cells. They could provide a 5 kw micro- processor controlled system for an estimated $130,000 if several units were ordered at once. They presently do not have these units for sale but may be able to offer a fuel cell at a more modest cost in the next twelve months. ANALYSIS OF RESPONSES The first result of this analysis of the responses is that some of the questions should have been rephrased and some additional questions posed. The question pertaining to the toxicity of catalyst chemicals should have been expanded to include all chemicals in the PAFC stack. Phosphoric acid is a toxic and dangerous chemical and is present in the stack. One manufacturer, Engelhard, has a method of automatically replenishing it when it is depleted (16), (17), but in other designs this process must be done manually after disassembly of => = the stack. The question concerned with the transient response of the PAFC should have been expanded to include the reformer as well. The PAFC responds very quickly under transient conditions, (35% to rated output in 0.5 seconds) but the reformer apparently does not respond quickly. (10), (18) There is also a serious problem that may occur during the PAFC start-up. If the complete stack is not up to operating temperature when the load is applied, a phenomenon called cell inversion can occur which effectively short circuits part of the stack (19). A question pertaining to this phenomenon should have been included in the requests to both ERC and Engelhard. The most significant result from the responses is that a PAFC power plant will require a logical control system, i.e., a micro- processor to prevent such problems as stack over-temperatures, cell inversion on start-up, and reformer output under transient loads. The requirement for a sophisticated electronic system is a negative factor against the application of the PAFC power plant in the bush, both from a personnel skill level and replacement part aspect. Sulfur is apparently the only significant poisoning agent to the PAFC and it can be removed during the production of methanol. Carbon monoxide will poison the PAFC and can be generated by a malfunction of the reformer. Prevention of this problem could be incorporated in the microprocessor control. The deterioration of the catalyst during down time appears to depend on the design of the PAFC stack. ERC which uses an air cooled stack, recommends that the stack be sealed during down time. Engel- hard, whose stack is liquid cooled, says that their catalyst will not deteriorate during down time. The explosion potential and possibility for catalyst stack melt down are fire safety aspects of the PAFC that are directly related to the adequacy of the logical control system. With a proper control system, the fire hazard of the PAFC is no greater than any device (including a diesel engine) where volatile fuels are in the proximity - 146° of moderately high temperatures. This problem has been dealt with by the New York City Fire Department in approving the 4.8 mw demonstra- tion project for Consolidated Edison in Manhattan (10), (20). According to both ERC and Engelhard, low temperature arctic operation is not an insurmountable problem. However, in case of a power outage, automatic drain systems must be provided in the liquid cooled units to avoid freezing. The question pertaining to catalyst life should have been modified to include fuel cell catalyst, stack life and mechanical durability. As indicated in the literature search, these problems , along with heat exchanger and reformer reliability are the major road blocks to commercialization of the PAFC. The values for catalyst life and heat exchanger and reformer life are projections from mainly bench tests and are not representative of commercial designs. ERC has built several small scale PAFC power plants for the Army so they probably are closer to developing a commercial unit than Engelhard. Maintenance and repair costs and personnel skill level require- ments are difficult to assess at this point because of the lack of commercial maturity of the PAFC power plant. ERC indicated that their unit required only U.S. Army skill levels but this was only for routine maintenance and not major stack tear-downs that are required yearly. The unit in question is also much smaller (2.5 kw) than that required for a typical rural state building (50 kw). Considering the necessity of a microprocessor control system, Engelhard was probably more realistic in anticipating that a skilled technician would be required to work on the PAFC power plant. ~ 17 = CONTINUING RESEARCH PROGRAM The experimental program to continue studying a 2-5 kW methanol fueled PAFC plant is described below. The summary of the cost estimate to initiate the program is shown in Table l. A schematic of the proposed test set-up using an ERC supplied PAFC is shown in Figure 2. Hydrogen can also be used to power this PAFC. Table 2 lists the quantities to be monitored during tests. The number of temperatures to be monitored was taken from information provided by ERC (19). The general test procedure will involve monitoring the fuel input and electrical output of the plant under various environmental and time dependent load conditions. Since there are a large number of data points (Table 2) to monitor, a data logger is necessary for transient tests. A resistive load bank is shown in Figure 2 but it could be replaced by a water conductivity load cell if higher capacity is desired (21). From Table 2, the total cost estimate for initiating the experimental study of a 2.5 kw PAFC is about $110,000. As a pure research tool, this system would be adequate. However, the purpose of this program is to study the more technologically mature PAFC systems. Therefore, the computer controlled unit is probably closer to commercial reality and should be selected. The cost for the program with this unit is $230,000. =e TABLE 1 4. Summary of cost estimate for initiation of proposed 2.5 kW PAFC test program. Item Manually controlled 2.5 kW PAFC with reformer and inverter (including microprocessor control) Hewlett-Packard Model 3054 Data Logger (40 point) or equivalent Load cells and miscellaneous equip- ment supplied by UAF Principal Investigator for three man-months with clerical and student help TOTAL (with computer control) = fo.— Cost $ 55,606 ($177,000) $ 10,000 $ 10,000 $ 35,000 $ 110,606 ( $ 230,000 ) Stack Temperature Air Air Inlet Outlet (0) NOTE: 24V DC and 115 VAC supplementary power must be supplied to PAFC. Data Logger Figure 2: Schematic of proposed 2.5 KW PAFC test set-up. Table 2 Physical Quantities Measured During PAFC Performance Tests Component Quantity Location on Fig.2 Fuel Supply Fuel inlet flow 1 A,B Fuel inlet temperature 2 A,B Fuel inlet pressure 3 A,B Reformer Air inlet temperature 4 Air outlet temperature 5 Combustor temperature 6 Fuel outlet temperature 7 Fuel outlet pressure 8 Fuel Cell Stack Inlet air temperature 9 Outlet air temperature 10 Temperature at a location in stack 1l - 16 Inverter Current 17 Voltage 18 Load Cell Current 19 Voltage 20 Frequency 21 - 20 - ENGINEERING ECONOMIC ANALYSIS The economic analysis is based on the method described in the Handbook of Fuel Cell Technology (22). The benefits of waste heat recovery are not considered. In the analysis, the cost of electricity produced by the plant in dollars/kwh is determined by the following formula. cit + t) K. * 3413 Cc, = F+ ———— + ——— (1) Wat ae Pg where c, = cost of electricity (- $/kwh _ M & O -_* U (M & 0) = annual maintenance operating cost in $/kw U = usage in hours per year. Includes affect of availability Cc. = Total investment cost including installation, in $/kw ea ees ok LIFE YR fe = annual interest, % f = capacity or utilization K. = fuel cost, $/gal. T = Fuel heating value, BIU/gal Pe = efficiency as a function of capacity ae A 100 KW power plant unit size was selected for comparison because of supplemental data available and popularity of the size in bush communities (23). The cost comparison is summarized in Table 3. The investment cost for a 100 kw diesel generator consists of the purchase cost of $235/kw plus $765/kw installation cost to give a total of $1,000/kw. The investment cost for the PAFC plant is $500/kw (3),(18), plus the same installation cost, $765/kw for a total of $1,265/kw. For both the diesel generator and PAFC, the same life, 20 years, same interest rate, 12%, same annual usage, 7,972 hours, and avail- ability, 91%, is assumed. A more realistic availability for the PAFC would be 83%. (18) The annual maintenance and operating cost was assumed to be $43/kw for both plants. As discussed in the literature search, these costs for the PAFC are based on projections from short term laboratory tests and may actually be much greater. The cost for no. 2 diesel fuel, $2.14/gal was determined from current market prices with a $l/gal transportation charge to the bush. (23), (24) However, methanol is not presently produced in Alaska but Alaska Interior Resources Company in Fairbanks is promoting the development of a plant to produce methanol from North Slope natural gas. Their current economic analysis shows that they can sell fuel grade methanol (99% pure) at approximately the same cost per BTU as no. 2 fuel oil. Assuming this to be true, the refinery cost for methanol would be $0.47/gal. With the same $1/gal. transportation charge, the cost of methanol in the bush becomes $1.47/gal. The heating value for no. 2 diesel fuel is 1.352 x 10° BTU/gal and for methanol it is 5.576 x 10% BTU/gal. (25) Two values for operating load range, 100% and 30% were used for comparison purposes. The efficiency of the diesel varies with load capacity. Most diesels cannot be operated continuously at less than about 60% capacity without suffering mechanical damage. For compari- son purposes, the diesel is assumed to have an efficiency of 25% for full capacity down to 60%. Below this demand it is assumed that the diesel generator is maintained at 60% capacity and the excess power is = 29) = a dummy load thereby reducing the overall efficiency of the power plant. The following formulas express this relationship mathematically. For the diesel generator: PO, = 25% for £260% (2a) P¢ = 25(1 + £- 60)% for f£< 60% (2b) At 30% capacity Pe = 18% This is in the range values for generation efficiency for bush regions @3). For the PAFC the efficiency is assumed to be constant at 37% for all capacities. The results of the calculations for Cc, are shown at the bottom of Table 3. These costs are in line with actual charges in bush communities. (23) The most significant result is that at 100% capa- city, the PAFC produces electricity that costs $.04/kwh or 14% more than the diesel generator. However, at 30% capacity, the PAFC pro- duces electricity at $.06/kwh or 15% cheaper than the diesel generator. The reason for this lack of cost advantage of the PAFC over the diesel generator lies in the properties of the methanol fuel. Methanol has less than 1/2 the heating value of diesel fuel on a volume basis. The cost per heating unit for methanol at the refinery is assumed the same as no. 2 diesel oil but there is a $l per gallon cost in shipping liquid fuel to the bush communities. This raises the cost of methanol per heating unit 66% higher than diesel fuel. The higher efficiency of the PAFC does not overcome this disadvantage. To further illustrate this point, if the shipping charge is neglected, the PAFC produces electricity that is more than 20% less expensive than the diesel at all capacities. There are several things that must occur if the cost disadvantage imposed on the PAFC by the methanol fuel is to be overcome: (a) The - 23 - - 97 - TABLE 3 Quantity : Total investiment cost including installation Life: Interest: Annual Usage: U Annual Maintenance & Operating Cost (M&O) Fuel Cost: K. Fuel Heating Value: T Capacity: Efficiency: Electricity Cost: Cy $/kwh 100 Kw Diesel-Electric Generator $1,000/kw $ 235/kw cost $ 765/kw install 20 years 12% 7972 hours (91% reliability) $ 43/kw $2.14/gal (diesel) $1.14/gal refinery cost $1.00/gal transportation 1.352 x 10 BTU/gal. (mo. 2 diesel fuel) 100% 30% 25% 18% $0.24/kwh $0.40/kwh - Cost Comparison of Electricity Produced by PAFC and Diesel-Electric Generator 100 Kw PAFC $1,265/kw $ 500/kw cost $ 765/kw install 20 years 12% 7972 hours (91% reliability) $ 43/kw $1.47/gal (methanol) $0.47/gal refinery cost $1.00/gal transportation 5.576 x 10 BTU/gal (methanol) 100% 30% 37% 37% $0.28/kwh $0.34/kwh PAFC must become very competitively priced with diesel generators; (b) The efficiency of the PAFC must increase; (c) The price per heating unit of methanol must become much less than diesel fuel. Several alternative systems are being developed but the diesel fuel- hydrogen reformer technology is not at the same level as the methanol systems (26). The molten carbonate fuel cell has a 25% higher effi- ciency than the PAFC but is at least five years behind it in commer- cialization (18). Considering the results of the cost analysis and the scarcity of durability data, the evaluation of the methanol fueled PAFC for the bush should be suspended until more operational data is publicized and the units are commercially available. -25- IMPLEMENTATION These investigations have shown that there is a very low proba- bility that fuel cell technology could be integrated into the routine operations of the Department in the near term. Therefore, it is not possible to directly implement the results of this work. The investi- gation has also shown that the fuel cell industry itself is not commercially mature enough to warrant a continuation study by the Research Section at this time. This work does suggest, however, that fuel cell technology has certain inherent advantages over conventional diesel-electric systems for rural power applications. These advantages will only be of benefit to the Department after the fuel cell industry achieves a considerable maturation. This maturation process is, of course, accelerated or decelerated by the market potential foreseen. Unfor- tunately, the portion of a potential market base represented by this Department or Alaska in its entirety, is not expected to be large enough to have much affect. We therefore recommend that no further work in fuel cell technology be scheduled at this time. Instead, the progress of the industry will be monitored through the literature in a nonspecific way. If a significant breakthrough occurs, we may in the the future recommend that investigations be renewed. Leroy E. Leonard, Chief Energy and Buildings Research Alaska DOT/PF ~ 4 = APPENDIX A Letters from Manufacturers = 27M ENERGY RESEARCH CORPORATION RENELASTUR: ~LJOE TWX NO. S10 - 456-0486 6 Fee nia. 969615 . TNENALS CORPORA TON June 23, 1982 UNIVERSITY OF ALASKA College of Mathematics, Physical Sciences and Engineering School of Engineering Mechanical Engineering Department College, ALASKA 99701 ATTN: Dr. James B. Malosh Dear Dr. Malosh: As per our discussion yesterday, we are quoting the 2 kW fuel cell two ways, manual and automatic. ERC sells the 2 kW manually controlled system which incorporates a methanol reformer, control panel and fuel cell stack for $49,500. This system is an excellent teaching and learning tool. Output is controlled manually by changing fuel flow. The alternate system based upon our U.S. Army development is fully automatic and incorporates a microprocessor to initiate and control all system functions. This system is, at this time, expensive and is herein quoted at $175,000. Specifications are attached. We will obtain and supply you with an AC inverter for an additional $6,106 which can be used with either system. The following addresses the questions which were asked in your letter of 9 June 1982 in the order that they were presented. 1. The U.S. Army requirements have required a safety analysis. Safety features have been built into the set and no explosion hazard exists. Hydrogen is utilized as it is produced. There is no appreciable pressure of hydrogen in the system. 2. | Over-temperature protection is also built in - no "meltdown" can occur. 3. The catalysts are non-toxic and are also not accessible to the user. The fuel cell stack utilizes platinum as a catalys whereas the reformer utilizes copper and zinc. = 281+ ENERGY RESEARCH CORPORATION UNIVERSITY OF ALASKA Dr. James B. Malosh June 23, 1982 4, The MIL Spec requires -65°F operation. Cold temperature operation is a matter of heating the system to operating temperature. d. The unit we presently are producing for the U.S. Air Force will start in 1/2 an hour at -25°F and | hour at -60°F. The system is heated up to starting temperature (approximately 250°F) by burning methanol. 6. It responds automatically 7. Over 10,000 hours for the reformer catalyst and over 30,000 for the PFAC stack catalyst. 8. Chlorides and-sulfur, however, methanol is a very pure chemical in commerce as it is produced with catalysts that would suffer poisioning if the starting chemicals (usually natural gas) were not pure to begin with. 9. We estimate approximately five years. 10. U.S. Army levels. There are very few working parts and only three continualy moving parts in the Army set i.e., the air blower for the stack, the small burner blower and the fuel pump. 11. The parts which may require replacement are the fuel pump and fuel injector which are both automotive components and the burner air fan and air blower motor. 12. The catalyst does not appear to deteriorate during down time, however, both the stack and the reformer have to be sealed during down time from the atmosphere either automatically or manually. Pao aes (tht pe —* R: Voyentzie ~ Manager of Marketing PRV/amd Attachment Page No. 2 anor = - 0€ - 3 AND 5 kW METHANOL POWERPLANTS OUTPUT VOLTAGE MODE I/II MODE Iv FUEL FUEL CONSUMPTION NOISE LEVEL STARTUP TIME WEIGHT VOLUME MTBF LIFE 120/290V, 60/400 HZ 28 vbDc (3 kW ONLY) 58% METHANOL - 42% WATER 1.8 LB MEOH/kWHRA INAUDIBLE AT 100 METERS 15 MINUTES 100 LB/kW 4CU FT/kW 750 HOURS 6,000 HOURS / 2000 STARTS Dr. os IS Sea? see Sem NAL OP SEMGESLAARO IMQUSTAIES 2IVi Str December 7, 1982 James B. Malosh Mechanical Engineering Department University of Alaska College, Alaska 99701 Dear Dr. Malosh: It was nice meeting you in Fairbanks last month. Again I apolo- gize for not having answered your inquiry from last July. As I explained to you we are in an R&D program for phosphoric acid fuel cells and consequently are not in a position to produce fuel cells at the moment. We would, however, like to answer your ques- tions and bring you up to date with our program. Many of the questions you have asked have not been addressed di- rectly in our program, but our answers are based on our best ex- perience. 1. The gases of reaction are always separated on the inlet side of the fuel cell. In case of hydrogen leak, hydrogen detec- tors are normally utilized and an alarm will sound before the explosive limit is reached. In case of interreaction within the cell, only very small quantities of reactant gases are in the stack at any one time, and this would not lead to a poten- tially dangerous situation. In an unusual situation a melt down might occur; however, the micro processor monitors certain temperatures and an automatic shutdown would occur prior to melt down. The toxicity of the precious metal catalysts is nil. In the reformer standard catalysts employ copper and zinc which can be toxic when‘ burned. Arctic temperatures could cause the following redesigns from the standard: a) Special coolant liquid. b) Redesigned start-up burner. c) Auxiliary heaters for standby conditions. d) Special start-up or hybrid batteries. e) Special insulation. None of these appear to be beyond the present state of the art . = OF = EME G E 6. Fa SE 52 aD Dr. James B. Malosh University of Alaska December 7, 1982 Page 2 5. Start-up time depends upon: a) Ambient temperature (-60°F is not beyond reason). b) Availability of auxiliary power. c) Use of liquid coolant as warm up medium. d) Use of batteries. e) Keeping unit in standby. 6. Fuel cell follows very well with no restrictions on load change. Fuel pump to reformer is driven by fuel cell stack voltage. 7. Our catalyst will last a minimum of 40,000 hours with degrada- tion of less than 10%. 8. Aromatics and sulfur bearing compounds will poison the cata- lyst and thus is irreversible. Higher alcohols will also poi- son the catalyst, but this is a reversible situation. 9. The reformer and heat exchangers could expect a nominal 20 year life. 10. Initially it is anticipated that a skilled technician with in- strument and electrical background would be required. As manufacturing techniques and design are improved it is antici- pated that a lesser qualified level would be required. ll. Replacements are not available at this time. As soon as firm manufacturing costs are available, replacement prices will be furnished. 12. The catalyst does not deteriorate during down time. Prices of our systems have not yet been firmly established. We understand the position you are in and the need for a price so that a continuation to the next phase of your project may follow. Therefore, we would like to give you an estimated price of $130,000 for a 5KW system including 208V/1P AC inverter, a methanol steam reformer and a system microprocessor. We do not currently have these units for sale and the price is based on our estimated cost to build with our existing facilities and personnel. We may re- quire that we receive several similar orders before we can accept an order should you decide to place one. ae OE | ENGELHAR® Dr. James B. Malosh University of Alaska December 7, 1982 Page 3 Our program is continuing, however, and in the next twelve months we should be in a position to offer you a Fuel cell at a modest price with all the latest up to date technology. Thank you for the opportunity to visit and to explain our fuel cell program to you. We hope in the future that we can be of service to you. Very truly yours, moat Lamb Venture Manager, Fuel Cells TJL:mt - 33 - 10. ll. 12. 13. REFERENCES Christner, L.G., et. al., "Scale-Up of Phosphoric Acid Fuel Cells", Proceedings of the 14th Intersociety Energy Convers. Engineering Conference. , Boston, MA, August 5-10, 1979, V. 1, pp 554-558. Patel, D.N., "Methodology for Predicting Long-Term Fuel Cell Performance From Short-Term Testing", DOE Contract No. 90-6182, Dec. 1980. Anon., "Fuel Cell Power Plants for Dispersed Generation", EPRI Tech. Note TS-1/54321, May 1979. Warshay, M., et.al., "Status of Commercial Phosphoric Acid Fuel Cell System Technology Development", NASA Report No. TM-81641, Jan. 1981. Christener, L., "Technology Development for Phosphoric Acid Fuel Cell power Plant, Phase 2", NASA Tech. Report No. NASA-CR-165426, Dec. 1981. Electric Power Research Institute Report, "Advanced Technology Fuel Cell Program," Final REport No. EPRI EM-1730, RPO 114-2, March 1981. Kaufman, A., Terry, P.L., "Phosphoric Acid Fuel Cell Development," Final TEchnical Report, DOE Contract DAAK70-77-C- 0206, Sept. 1980. : Abens, S., et. al., "Improvement of Phosphoric Acid Fuel Cell Stacks," Final Technical Report, DOE Contract, DAAK70-77-C-0174, July 1980. Fraas, A.P., Engineering Evaluation of Energy Systems, McGraw- Hill Co., New York, NY, 1982. Landgrebe, A.R., Weinstock, I.B., "Application of Electrochemical Technologies to the Utility Industry," ASME Paper 81-WA/AES-4, Nov. 1981. Ross, P.N., "Improvements in the Utilization of PT in Phosphoric- Acid Fuel Cells," J. Electrochem Src., V129, N3, p 115, 1982. Stanchart, P., "Preparation and Evaluation of Advanced Electrocatalysts for Phosphoric Acid Fuel Cells," NASA Reprot Nol CR-165594, Dec. 1981. McAlister, A.J., "Non-Noble Catalysts and Catalyst Supports for Phosphoric Acid Fuel Cells," NASA Report No. CR-165308, March 1981. -3- 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Faroque, M., et. al., "Evaluation of Distributed Gas Cooling of Pressurized PAFC for Utility Power Generation," NASA Report No. CR-165304, May. 1981. Hoover, D.Q., Jr., "Gell Module and Fuel Conditioner Development," NASA Report No. CR-165462, Oct. 1981. Kaufman, A., "Development and Test Fuel Cell Powered On-site Integrated Total Energy System. Phase 3: Full-Scale Power Plant Development ,"" NASA Report No. CR-165328, June 1981. Anon., "Fuel Cell Systems," Published by Engelhard Corporation, Tselin, NJ, 1982. King, J.C., et. al., "Candidate Electric Energy Technologies for Future Application in the Railbelt Region of Alaska," Vol. IV of the Reort to Office of the Governor, State of Alaska, Prepared by Battelle Pacific Northwest laboratories, Oct. 1982. Operating Manual for Phosphoric Acid Fuel Cell Stack, ERC S/N 8116 LA, Energy Research Corp., Donhuen, CT., Nov. 1980. Glosser, K.F., "4.8 mw Fuel Cell Demonstration Program - A Progress Report," presented at the National Fuel Cell Seminar, San Diego, CA, July 14-16, 1980. Lalk, T.R., Willett, K. R., "A Comparison of the Transient Response of Two and Four Stroke Diesel Power Generator Sets," SAE cyte Paper No. 810920, Sept. 1981. Berger, C., Ed., Handbook of Fuel Cell Technology, Prentis-Hall, Inc., Englewood Cliffs, NJ, 1968. Executive Summary, "State of Alaska Long Term Energy Plan," Report No. DEPD-82-530-R3, 1982. "The Energy Report," Vol. III, No. 3, Community Research Center, Fairbanks North Star Borough, Dec. 1982. Adler, U., Baylor, W., Ed., The Bosch Automative Handbook, Robert Bosch GMBH, Stuttgart, Germany, 1978. Haughty, W.E., et. al., "Development of an Adiabatic Reformer to Process No. 2 Fuel Oil and Coal-Derived Liquid Fuels," EPRI Report No. EM-1701, Feb. 1982. - 35 - Stack Temperature Reformer © Compustion Temperaturg Methanol Air Air Inlet Outlet (20) NOTE: 24V DC and 115 VAC supplementary power must be supplied to PAFC. Data Logger Figure 2: Schematic of proposed 2.5 KW PAFC test set-up. Table 2 Physical Quantities Measured During PAFC Performance Tests Component Quantity Location on Fig.2 Fuel Supply Fuel inlet flow 1 A,B Fuel inlet temperature 2 ASB Fuel inlet pressure 3 A,B Reformer Air inlet temperature 4 Air outlet temperature 5 Combustor temperature 6 Fuel outlet temperature a Fuel outlet pressure 8 Fuel Cell Stack Inlet air temperature 9 Outlet air temperature 10 Temperature at a location in stack 11 - 16 Inverter Current 17 Voltage 18 Load Cell Current 19 Voltage 20 Frequency 21