Because of existing utility grid congestion concerns and physical limitations to existing electrical distribution infrastructure, the university concluded that distributed generation posed a practical opportunity to satisfy its power needs. The university's three-fold goal was to (1) develop a university-owned reliable power supply; (2) displace power purchased from the grid; and (3) reduce power congestion along local power supply transmission lines, thereby reducing the risk of possible brownouts affecting certain university operations. Because of its experience in campus energy systems, distributed generation systems, and fuel cell power systems, R.W. Beck Inc. was hired by the university to conduct a feasibility planning assessment of the potential for integrating a fuel cell power plant on its campus. Principal funding for the feasibility assessment was provided by a state renewable energy organization.

Although other fuel cell technologies were commercially available, such as phosphoric acid, proton exchange membrane, and solid oxide technologies, the university chose to focus only on molten carbonate fuel cells, which then represented the only fuel cell technology promising megawatt-level commercial units (see Figure 1). At the time of study, this technology had several commercial applications in output sizes of 250 kilowatts or less.

Figure 1. Schematic of a Molten Carbonate Fuel Cell

The objective of R.W. Beck's assessment was to develop observations and conclusions regarding the technical and economic feasibility of implementing this fuel cell technology in a campus environment as a distributed generation source. The feasibility planning assessment considered the requirements for integrating soon-to-be-commercially available 1-megawatt fuel cell modules, comprising of four 250-kilowatt stacks, which would be grouped in pairs to create a 2-megawatt power unit (see Figure 2). These requirements included (1) identifying the infrastructure necessary to support the fuel cell modules; (2) calculating the fuel cell module's electric and thermal outputs; (3) determining the scope of interconnection necessary to utilize these outputs; (4) identifying installation costs; and (5) identifying the operations and maintenance (O&M) costs to quantify life cycle characteristics of the fuel cell power plant. The assessment also included a review of molten carbonate fuel cell technology and its ability to scale up to a megawatt-level plant. Operating and maintenance issues and long-term equipment replacement and reliability were also addressed.

Figure 2. 1-MW Stack Module (courtesy of OEM)

Figure 3. Typical Campus Thermal Interconnection

Figure 4. Layout of a 2-MW Fuel Cell Unit (courtesy of OEM)

Site Integration

Since the fuel cell's original equipment manufacturer (OEM) designed its fuel cell power plants for outdoor application, a number of outdoor on-campus sites were considered during the process. Certain general characteristics were desired at anticipated campus integration locations, notably (1) the project site could provide for all fuel cell plant needs, including fuel supply, water supply, and wastewater disposal; and (2) the point of integration would accept thermal output and electric output. Since the campus has an extensive district heating and cooling system, it was not difficult finding locations to tie in and accept the thermal output of the fuel cell plant (see Figure 3).

Certain electric characteristics were also desired at anticipated campus integration locations that made site selection more problematic. For instance, as a basis for design feasibility it was decided in advance (for reasons stated elsewhere in this article) that (1) the fuel cell power plant would need to operate parallel to the electric grid; (2) the fuel cell power plant would always need to have backup power from the electric grid; and (3) minimum electric load at the point of interconnect with the fuel cell power plant would need to be greater at all times than the rated output of the fuel cell power plant; otherwise the ability to back-feed into the utility grid would need to exist.

As previously stated, the 1-megawatt fuel cell modules would be packaged in pairs to create a 2-megawatt power unit (see Figure 4). Required resources for operation of a single 2-megawatt fuel cell power unit with a heat-recovery steam/hot-water generator were expected to be approximately 13.65 MMBtu/h of domestic pipeline–quality natural gas at a pressure and temperature consistent with values of typical low-pressure pipeline supplies and approximately 300 gal/hr of municipal-quality potable water. Each 2-megawatt unit was expected to discharge approximately 125 gal/hr of wastewater from the process, which could be piped to a campus sewer or similar municipal system.

Each 2-MW unit required a footprint of approximately 5,700 square feet, which included space for equipment, maintenance access, and a reserved area for stack/module installation and removal. A stub stack measuring a nominal 15 feet would be necessary for the rejection of waste heat from the process.

Performance and Operating Risk

Each 2-megawatt fuel cell unit was proposed to have a design performance basis of 2,000 kilowatts net electric output at a heat rate of 6,824 Btu/kWh (LHV) natural gas input at initial commercial operation, exclusive of output degradation, and a thermal output of 2.8 MMBtu/h as steam. For the purposes of the feasibility assessment, fuel cell power unit availability was assigned to be 90% though the OEM reported greater values. It was further assumed that fuel cell module design output degradation would be 3% of net power output per year on the basis of no more than two cold starts per year.

Each 1-MW fuel cell stack would comprise four 250-kilowatt fuel cell stacks. At the time of the study, the OEM's 2-megawatt fuel cell units and 1-megawatt modules were not operating commercially although their commercial operating fleet consisted of nominal 210-kilowatt/250-kilowatt units that had been in operation for approximately two years. Consequently, there was insufficient information for R.W. Beck to fully review thermodynamic performance risk, operating availability risk, and projected long-term O&M costs for a 2-megawatt unit. Due to this limited experience and the issues that are typically encountered with the introduction of a new technology, the use of project-specific risk mitigation strategies was expected to be both prudent and necessary, and the best way to mitigate risk to the project's technical and financial performance. It was expected that the university would mitigate technical risks through appropriate commercial guarantees, service agreements, guarantees from the OEM, or a combination of these, if it went forward with the project.

Relating to these concerns, the OEM proposed to the university a long-term service agreement (LTSA) for materials and labor to maintain the fuel cell power plant, including major overhaul stack replacement, but excluding routine preventive-maintenance labor and materials, consumables, and warranty repairs covered by any equipment purchase contract. The base period of the proposed LTSA was coincident with the expected lifespan of an individual fuel cell module stack, which would need to be replaced every three to five years. It was expected that routine maintenance would be performed by existing university energy facilities staff.

Environmental Permitting

Air pollutant and wastewater pollutant discharges for fuel cells are relatively negligible, when compared to conventional electric generating equipment. In the proposed campus location, there were no anticipated environmental regulatory air permits required to construct the proposed fuel cell power plant, due to the non-applicability of federal new source performance standards and state air emissions standards. It was determined that the fuel cell power units' reverse-osmosis discharge of minor wastewater streams would likely require ministerial approvals to allow discharge to the municipal sewer system.

Grid Connection

On the assumption that a fuel cell power plant would help both parties, the university desired that the local utility would agree to allow the fuel cell plant to back-feed power generated to the grid to reduce, if not eliminate, demand charge for power, on the assumption that the fuel cell plant would reduce grid congestion and improve grid stability in the area.

Since it was anticipated that the 2-megawatt fuel cell units might have relatively limited capability to rapidly increase or decrease electric output with rapid changes in connected load, there was concern that if connected load to a unit fell below that unit's power output, continuous operation of the unit might be interrupted. Therefore, to mitigate this concern, as well as to avoid possible utility standby charges if this were to indeed occur, it was desired that the minimum electric load at the point of interconnection for each unit needed to be equal to or greater than the rated output for each unit; in other words, the unit would always be displacing load.

Although each 2-megawatt fuel cell unit was expected to have high reliability, under any scheduled or unscheduled outage it would be necessary to rely on power from the grid or an emergency generator as a parallel supply to the fuel cell power plant during the outage as well as for startup, since it could take a relatively long time to start up or re-start a unit after a shutdown or trip depending on the duration of the outage.

One concept proposed by the university to address grid interconnect concerns was development of a dedicated 13.8-kilovolt distribution network for the university's campus. Rather than connect the fuel cell power plant to a lower-voltage load (e.g., 4,160 volts or 480 volts) with many peaks and valleys, connection to a central 13.8-kilovolt system with a flatter load profile might ensure that connected load to each fuel cell unit would always be greater than the unit's output at the point of electrical interconnect thus negating interconnect concerns.

Economic Impact

The economic impact to the university's business operations from power generated by a fuel cell plant was unclear. The university would, it was assumed, benefit from a significant reduction in its electric energy purchase from the utility, and recapture most if not all of the waste thermal energy, which would reduce university fuel consumption for space heating and cooling needs. Having a fuel cell plant as an independent source of generated electricity would also benefit the university environmentally, through reduced consumption of fossil fuels. One key economic issue that was never resolved because of a lack of applicable distributed generation rate structures was the potential for utility standby charges.

Additionally, as an independent source of generated electricity, the fuel cell system also would provide certain other intangible benefits, such as supporting the university's independence and security. By examining the potential of distributing the output of the fuel cell systems to selected campus locations, the university anticipated an increased ability to provide essential services of food and shelter to a considerable number of students, faculty, and staff, in the event of a grid outage. The fuel cell plant, if developed, was also expected to partially mitigate the impact of local grid congestion by displacing existing grid load, thereby reducing the risk of campus-wide economic loss from grid brownouts. While the fuel cell plant would supplement the existing supply of electric power from the utility, however, it would not be a substitute.

Conclusion

R.W. Beck's fuel cell feasibility study concluded that incorporation of fuel cell power generation technology into a campus utility system is technically viable and can provide synergy with a campus system, particularly if that campus has adequate electric and thermal load at the desired point(s) of interconnection. The combined heat and power generation efficiency of the reviewed fuel cell technology was relatively high and desirable for a campus system electric and district heating and cooling system. Because of uncertainties regarding utility backup charges, project economics were never fully resolved. Ultimately, the university chose not to go forward with the project due to certain project-specific issues related to grid interconnect and certain project-specific economic issues.

PAUL D. CLERI, P.E., is a principal engineer with R.W. Beck Inc. in Boston, MA, and can be reached at pcleri@rwbeck.com or 508/935-1600.

DE - September/October 2004