R. Root Woods, John J. Cuttica and Karen A. Trimble
Gas Research Institute, Chicago, Illinois ABSTRACT This paper describes the experiences and results of the major field test of forty-six 40kW onsite fuel cell power plants in the U.S. and Japan through 1985. The field test is a cooperative effort between the Gas Research Institute, gas and electric utility companies, private sector companies, and the U.S. Department of Energy and Department of Defense. The field test is conducted in parallel with tech nology development efforts sponsored by the Gas Re search Institute and the U. S. Department of Energy. Operation of the field test units began in December 1983 with the initial start-up of two units at a racquetball club sponsored by the Southern Califor nia Gas Corporation. As of May 1985, over 300,000 hours of operating experience was achieved on the fleet of 40kW units. This experience has demon strated the onsite fuel cell's superior technical specification, ease of installation, operation and maintenance, and economic benefits over convention al energy service. In addition, this effort iden tified operating deficiencies in the 40kW units' which were either corrected through field retrofits or provided guidance to the parallel technology development efforts. THE ONSITE FUEL CELL COGENERATION SYSTEM ONSITE ELECTRICITY FROM GAS With over 300,000 hours of cumulative operating experience (equivalent to 34 years) on forty-six 40-kW phosphoric acid onsite fuel cell power plants, the gas industry, in cooperation with International Fuel Cells Corporation (formerly the Power Systems Division of United Technologies Corporation), has demonstrated in extensive field tests that onsite fuel cell technology works, that it represents a new business opportunity for the gas industry, and that this high tech hardware can be installed, operated, and maintained by utility personnel. The test results show that onsite fuel cell technology could be used to provide gas industry customers with a complete, economically competitive energy service. The key advantages of fuel cell technology for providing such an energy service would be that it (1) efficiently provides both electricity and heat over a wide demand range and at a net saving to the customer; (2) has potentially high reliability, low maintenance, and automatic, unattended operation requiring minimal support from the gas utility; and (3) emits virtually no environmental pollutants, operates quietly, and conforms to building codes and standards to allow broad application. Because of these advantages, Gas Research Institute (GRI), 612 the U.S. Department of Energy (DOE), and 37 participants (utility companies, private sector corporations, and U.S. government facilities) embarked on a jointly funded program that could lead to the commercial introduction of onsite fuel cell power plants by 1990. Under this program, a field test has been conducted over three years, efficiently providing both elec tricity and heat to a variety of multifamily resid ential, .commercial, and light industrial buildings in the linited States and Japan. The program has included two other efforts complementing the field test: ~ . I o a parallel technology development project to improve reliability, performance, and maintain ability as well as to decrease manufacturing costs, o activities of the Onsite Fuel Cell Users Group, which performs economic studies and business assessments pertinent to the commercialization and application of the technology. ONSITE FUEL CELL CONCEPT The high overall fuel efficiency of 72 to 80 per cent (HHV) of onsite fuel cell plants results from the high electric generation efficiency of 36 per cent (HHV) of the electrochemical operation as well as from the cogeneration of electricity and heat at the point of end use. Installed at multifamily residential, commercial, or industrial buildings, the system can provide electricity and heat econ omically while conserving energy and minimizing the environmental impact (one-tenth the pollution of the cleanest conventional electric power plants). Modular design permits quick, economical factory production and easy expansion of installed capacity. Using natural gas or other fuels (gaseous and liquid fuels derived from coal and biomass and petroleum products), the onsite system produces alternating current electricity at the frequency (60 Hz or 50 Hz) and voltage (120/208 V or 277/480 V) required by the user, maintaining its high fuel-use efficiency over a broad range of electrical output (one-quarter to full-rated power). The system can either be connected to or isolated from the electric utility grid. The heat produced as a byproduct can be transferred to nearby buildings in the form of hot water or steam for space conditioning, water heating, and thermal processing. The system's flexible operational capability allows it to respond instantaneously to changes in the user's demand for energy. ESL-IE-86-06-101 Proceedings from the Eighth Annual Industrial Energy Technology Conference, Houston, TX, June 17-19, 1986 Figure 1. ONSITE FUEL CELL POWER PLANT CONCEPT Fuel Processing Section Power Section Power Conditioning Natural gas is converted to . The hydrogen-rich fuel is Section a hydrogen-rIch fuel by a reacted with oxygen by an The electricity is converted catalyic steam-reforming electrochemical process to by a solid-state Inverter to process. produce DC, electricity and alternating current of the by-product heat in the form frequency and voltage re of hot water or steam. qUired by the customer. All (Oxygen Source) Steam Nat ... ral (or other fuels) .Ii----;tto.-.r--- Hydrogen-Rich Gas Heat Usable Heat External Fuel Cavity Hydrogen Electrical Oxygen Circuit The power plant consists of three integrated sections (Figure 1) designed to process the incoming fuel, generate power, and condition the power to user specifications. The thermal energy is used to maintain the operating temperature of the power plant components and to supply usable heat to the customer. Steam is recovered for use in the fuel-processing section. PARTICIPANTS' ACTIVITIES The gas 37 participants in the field test utilities, 13 combination gas and included electric 13 utilities, 5 electric utilities, 2 private sector corporations, and 4 Department of Defense facil ities. These participants tested forty-six 40-kW power plants at 42 sites in the United States and Japan (Figure 2 and Table 1). Many building types were represented, which demonstrated that the onsite fuel cell technology can meet a wide variety of customer electrical and heating needs. The field test also provided the participants with the experience needed to evaluate and implement an energy service business. Initially, the particip ants instrumented over 100 sites collecting more than 800 site-months of building energy use data. This data base was used to evaluate economics, assess siting considerations, select sites for field test units, and optimize power Electrochemical Process The power secHon of the 4O-kW power plant contains a stack of 270 fuel cells, each made up of an anode, phosphonc acid electrolyte solution, and cathode. The hydrogen in the processed fuel IS passed over the anode, where it dissociates into electrons (carrying a negative charge) and ions (carrying a positive charge). The electrons are coi lected and then flow out through an exter nal electrical circuit, creating a direct current output to the power conditioner. The hydrogen ions move through the electrolyte solution to the cathode, which is bathed with air. As the hydrogen ions contact the cathode, they unite With oxy gen from the air and with the electrons returning from the power conditioner, forming steam that is further processed to yield heat and water. plant interfaces with the existing buildings' systems. Utilities were able to install the field test units without major difficulties. Installation required an average of nine weeks--five weeks to prepare the site and four weeks to complete the installation once the power plant arrived. After completing a two-week training course, service personnel from the participating companies operated and maintained the power plants at their sites. This experience included normal scheduled maintenance, failure diagnosis, component repair and replacement, and return to normal operation, demonstrating that an onsite fuel cell plant can be serviced and maintained by gas utility personnel. OPERATING EXPERIENCE A jointly funded GRI/DOE contract was issued in January of 1982 to manufacture the forty-six 40-kW field test power plants. The first field test unit was completed 18 months later, and the last unit was completed in January 1985. This was the largest manufacturing effort undertaken to date in the production of fuel cell systems. 613 ESL-IE-86-06-101 Proceedings from the Eighth Annual Industrial Energy Technology Conference, Houston, TX, June 17-19, 1986 JAPAN 0 HONOLULU ANCHORAGE, AK 0 Figure 2. LOCATION OF FIELD TEST SITES Market segment R=residential Table 1. FIELD TEST PARTICIPANTS AND TEST SITE DESCRIPTIONS C=commercial I-industrial Location Participant Location S=spec1al ty ~ a . . L Atlanta Gas Light Co. MarIetta, GA R - dormitory 35 Baltimore Gas &Electric Co. Pikesville, MD C - restaurant 30 Brooklyn Union Gas Co. Queens, NY C - airport arrivals 26 Staten Island, NY C - hospital Central Hudson Gas &Electric Corp. Wappinger Falls, NY R - apartments 21 Columbia Gas of Ohio, Inc. Cambridge, OH C - hotel/restaurant 19 Consolidated Edison Co. of New York, Inc. Croton/Hudson, NY C - nursing home 25 Consumers Power Co. Jackson, MI R - apartments 16 Dayton Power &Light Co. Vandalia, OH C - hotel 17 Delmarva Power & Light Co. Hockessin, DE C - nursing home 29 Florida Power Corp. Orlando, FL S - electric generating plant 37 Gasco, Inc. Honolulu, HI C - hotel 3 Georgia Power Co. Gainesville, GA C - hospital 36 International Business Machines Corp. Kingston, NY I - data processing 22 International Fuel Cells Corp. South.Windsor, CT S R&D lab 24 Memphis Light, Gas & Water Div. Memphis, TN I - photo lab 33 Mountain Fuel Resources, Inc. Layton, UT I - photo lab 10 National Fuel Gas Distribution Corp. Cheektowage, NY C - hotel 20 Northeast Utilities Vernon, CT S - telephone switching 23 Northwest Natural Gas Co. Portland, OR C - laundry 5 Osaka Gas Company, Ltd. Osaka, Japan C - restaurant 2 Pacific Gas & Electric Co. San Francisco, CA C - office 6 Panhandle Eastern Pipe Line Co. Louisburg, KS S - gas compressor station 13 Peoples Natural Gas Co. Omaha, NE C - office 12 Philadelphia Electric Co. Philadelphia, PA C - variety club 28 Public Service Electric & Gas Co. Avenel, NJ I - aluminum products 27 Princeton, NJ R - dormitory San Diego Gas & Electric Co. Carlsbad, CA C - nursery 9 Escondido, CA C - health club Southern California Edison Co. City of Industry, CA C - laundry 7 Southern California Gas Co. Fountain Valley, CA C - health club 8 C - restaurant Southern Company Services, Inc. Uni vers1ty, AL S - recreation center 34 Tennessee Gas Pipeline Co. East Bernard, TX S - gas compressor station 15 Tokyo Gas Company, Ltd. Tsukuba, Japan S - gas expo/restaurant 1 Tokyo, Japan C - hotel United Power Association Willmar, MN I - food processing 11 Virginia Natural Gas Norfolk, VA R - dormitory 32 Department of Defense sites Fort Belvoir Ft. Belvoir, VA S - dining hall 31 Elmendorf Air Force Base Anchorage, AK S - base power plant 4 Sheppard Air Force Base Wichita Falls, TX S - dormitory 14 Wright-Patterson Air Force Base Dayton, OR S - USAF museum 18 614 ESL-IE-86-06-101 Proceedings from the Eighth Annual Industrial Energy Technology Conference, Houston, TX, June 17-19, 1986 Figure 3. HOURS OF ONSITE PLANT OPERATION Cumulative Operation ~ Active Power Plants 80 400 70 Operating Goal 60 300 z Vl c: :; ~ 3 0 50 CY :I: ~ '0 / 40 "1J ~ 200 /' c ~ . ;5: Vl '" ~ 30 0 ::> ::> 0 ~ .c: S ... ~ , 20 100 ~ f?;l l!.l !fj I . ; ~ 10 ~ ~ ., !!Hff I 0 o J FMAM J JASONDJ FMAMJ J ASONDJ FMA I 1984 I I 1985 I L1986-.J The experience obtained by International Fuel Cells Corporation (IFC) validated many aspects of the commercial-level quality control techniques used. Over 30,000 fuel cell electrodes were manufactured utilizing only visual and nonoperational quality control techniques. These cells were assembled into stacks of 270 cells, incorporated into the power plants, and initially operated during the power plant's acceptance test at the manufacturer's facility. This technique proved very economical and effective -- only one electrode failed to perform during the test, demonstrating a 99.997% acceptance rate. While the field test units were being manufactured, three pre-field-test experimental 40-kW units were installed and operated for 13,000 hours through 1983. This experience provided verification of the design and uncovered deficiencies that were correc ted in the field test units. The first of the 46 field test units began opera tion in December of 1983 and the last started during the summer of 1985. As of the end of April 1986, over 300,000 hours of operation had been logged on these 40-kW units (Figure 3). While demonstrating the operational feasibility of onsite fuel cell technology, the field test verified and demonstrated every major aspect of the 40-kW specification. This 34 years of cumulative experi ence resulted in approximately 9.3 million kWh of power production. Power plants operated over the full range of a to 40 kW output. In addition, grid-independent units, which are required to handle large motor starts, successfully demonstra ted the 56 kW/80 kVA overload characteristic for 5 seconds without variation in the output voltage. The power plants also verified the heat recovery specification of 100 kBtu/hr low-grade heat up to laOoF and 45 to 55 kBtu/hr of high-grade heat up to 250 0 F. During the operation of each unit, detailed data collection allowed the comparison of measured to predicted efficiency. Using hourly simulations, these analyses indicated that 75 percent of the power plants met or exceeded the predicted electric generation efficiency. On average, low-grade heat recovery reached 95 percent of the predicted values based on actual site flow rates and return temper atures. High-grade heat recovery was also demon strated at specified temperatures and rates. As of the end of April 1986, the power plant with the most operating time had accumulated 11,000 hours of operation. 13 units exceeded 8,000 hOUl"S, 28 passed 6,000 hours, and 39 operated more than 4,000 hours. The average operating time was over 6,000 hours. The fleet's average unadjusted availability (operating time divided by calendar time after initial start-up) is 63 percent at 300,000 hours, well within the range of 55 to 65 percent established as a goal for the field test based on the research nature of the project and the maturity of the equipment. Commercial units are expected to achieve operating availabilities of 95 percent or better to minimize the cost of backup power and service visits. The field test demonstrated the automatic, unatten ded operating characteristics required for energy service. Thirty-seven units achieved a continuous operating run in excess of 1,000 hours (42 days). The longest continuous run exceeded 90 days. Approximately 40 percent of the operating exper ience was obtained during 75 runs greater than 1,000 hours. As another indication of the tech nology's ability to provide energy service with minimal support infrastructure, approximately 25 percent of the units achieved actual time between forced outages in the. range of 90 to 170 days. 615 ESL-IE-86-06-101 Proceedings from the Eighth Annual Industrial Energy Technology Conference, Houston, TX, June 17-19, 1986 Although the high-technology components of the onsite plant (cell stack, reformer, and inverter) operated relatively trouble-free, some problems were encountered involving the durability and reliability of some of the more conventional components. In these cases, the automatic control functions of the power plant safely shut the units down without further equipment damage. Electrical and electronic control system failures--resulting from poor quality control of printed circuit cards and improper specification of control band widths- caused electrical disconnects and plant shut-downs. The power section cooling system has exhibited plugging from corrosion products after about 2,000 hr of operation; for the field tests, this was corrected by flushing the system every 1,500 to 2,000 hr. To eliminate this problem, two field test units were installed in 1985 with modified cooling system manifolds. To date, these units have reached 1,900 and 3,600 hr of operation and have shown no indication of plugging. Shutdowns were also caused by failures in non-fuel-cell ancillary components such as pumps, valves, and leaky pipe fittings. All these problems have been analyzed and should be eliminated. COMMERCIAL SECTOR MARKETS Commercial buildings in the United States represent a sizable potential market for gas-fired cogenera tion systems. A study performed by the Onsite Fuel Cell Users Group characterized this sector as con taining about 4 million buildings, representing approximately 175,000 MW of cumulative 'peak elec trical demand, and expanding at an annual rate of 3 percent. Placement of the 40-kW test units in a variety of commercial building types provided an excellent demonstration of the system's flexibility in meeting the varied energy requirements of the commercial market (Table 2). The market study assumed that the onsite plants would be owned and operated by the gas utility under the regulated portion of its business. The utility would serve as the customer's energy company, meeting his electric and thermal energy demands from the fuel cell while collecting revenues from the customer for gas sold and a leasing or facility fee in return for the use of the fuel cell power plant. At an installed cost of $l,OOO/kW and with the onsite plant operating independent of the electric utility grid, the market for a full catalog of plants ranging from 40 to 500 kW was estimated to be 18,000 MW with potential annual revenues of $11.5 billion. Table 2. THERMAL AND ELECTRICAL INTERFACES OF FIELD TEST UNITS Electrical Interfaces Number of Sites* GC ~ - - ' - - ' = - " - - - - ' - - = - " - - - C ; l - Type of Site -5- [) Multifamily residential Commercial 16 5 Light industrial 4 1 Specialty 8 3 ~ = Connected to utility grid. I = Independent of utility grid. Thermal Interfaces Number of Sites** Type of Site wlS wo/S -1-3- 12 Domestic hot water Space heating o 2 Process heating 6 6 Combination 1 2 ** w/S With thermal energy storage. wo/S = Without thermal energy storage. FUTURE DEVELOPMENTS As a result of the technology development project, IFC has developed the technology basis for an initial commercial onsite fuel cell power plant that will be competitive on a combined cost and performance basis as required by the marketplace. Full-scale components have undergone verification testing. A 200-kW power plant design approach (Figure 4) has been defined that would be suitable for manufacturing in low production rates of 100 units per year and achieve installed costs in the range of $2500 to $3000/kW. IFC is currently in the process of forming a consortium to manufacture and market 200-kW power plants. With initial deliveries expected by 1990, the consortium's efforts over the next few years will be directed at completing the detailed engin eering design, building and testing preproduction prototype units, and providing the facilities to manufacture the initial lot of commercial units. With market acceptance and scaled-up production, the power plant is projected to achieve installed costs in the range of $1,000 to'Sl,500/kW. ()16 ESL-IE-86-06-101 Proceedings from the Eighth Annual Industrial Energy Technology Conference, Houston, TX, June 17-19, 1986 DC Module 8 x 24 x "-1/2 h Fuel processor Cell stack Heat recovery Water recovery Cooling Module 6x12x7ft (Remotely sited) Figure 4. DESIGN APPROACH TO 200 kloJ ONSITE POWER PLANT Electrical Module 8 x 8 x 111/2 h Inverter Transformer Controller Diagnostics 617 ESL-IE-86-06-101 Proceedings from the Eighth Annual Industrial Energy Technology Conference, Houston, TX, June 17-19, 1986