32 kW Fuel Processor

Based on the operating parameters acquired during the laboratory and pilot plant work, HTI began the basic engineering required to construct two 32 kW Brassboard units, one each for WEC and ERC. The basic engineering package included Process Flow Diagrams, P&ID's, Equipment Specifications and Logic for Safety and Control Systems. This information was then given to fabricators for competitive bidding. The successful bidder, Texas Systems and Controls, Inc. (Houston, Texas) began detailed engineering in December 1994.

The brassboard fuel processor was a completely self contained skid. Major equipment on the skid included reactors, air cooled condensers, and control system. A remotely located personal computer provided the interface between the operator and the control system. Heating for the unit, including both process heaters and heat loss compensation heaters, utilized electric heating elements. Feedstock and utility storage were provided off skid. Commercially available catalysts from Haldor Topsoe were used in the FPS.

The WEC skid was shipped to the Southern California Edison Fuel Cell Test Facility in Grand Terrace, California. It was installed and commissioned during the summer of 1995. The skid for ERC was shipped in August, 1995 and installed and commissioned in the fall of 1995.

Unit Operations - WEC

Start up on commercial grade JP-8 began on August 24, 1995. Initially, the unit met operating conditions but was limited to a maximum flowrate of approximately 70% of design capacities, due to electric heater limitations. Additional heater capacity was added, and the unit completed 766 hours delivering jet fuel reformate to the SOFC on October 11,1995. During the run on jet fuel, the feed was desulfurized to less than 1 ppm (wt). Inlet sulftir concentration, present in the purchased JP-8, was 80 ppm (wt).

The unit processed 5.3 kg/hr jet fuel with a hydrogen to fuel ratio of 0.5: 1, and a steam to hydrocarbon ratio of 2.5:1. Reformate composition was nearly identical to the design basis, with a stream of approximately 50% methane (CH4) on a dry basis. See Table 2.

When the jet fuel supply was nearly exhausted, diesel fuel was added and the unit continued operating. From October through February 1996, the fuel processor achieved the requisite 1500 hours delivering fuel to the SOFC. However, during this period; the fuel processor was actually operating 1918 hours, achieving an on line factor of 75%.

Again, the fuel processor met the goals of the diesel portion of the established test program. Exit sulfur levels remained,belowj ppm (wt) during the commercialdiesel (DF-2) run (360 ppm (wt) total sulfur inlet). Reformate composition did not vary between the jet fuel test and the diesel test. Design flowrates of 5.75 kg/hr were achieved as necessary.

Some mechanical difficulties were encountered, consisting mainly of electric heater problems. High operating temperatures and low flowrates combined to require high heat densities in the heaters. This made them susceptible to failure. One control system failure was also recorded. The repair of the failed elements-resulted in an on-line factor of 75%. However, no significant difficulties were encountered in the operation of the unit. The unit achieved the contracted 2000 hour runtime with the SOFC, and was subsequently shutdown for catalyst inspection.

Unit Operations - ERC

The successful startup of the ERC fuel processor began on October 2,1995. Once the fuel processor achieved stabilized operating conditions on October 7, the unit was shut down to install the MCFC. Minor problems occurred in November and December, including a heater failure and flow control problems. These challenges were corrected and the unit was prepared in December for jet fuel and diesel, scheduled for January and February 1996.

Reformate from the fuel processor was delivered to the MCFC on January 23, 1996. The fuel processor continued operation on both jet fuel and diesel into February at which time the testing requirements of 400 hours combined run time (200 hours each on jet fuel and diesel) were accomplished. Total runtime logged with the MCFC was 560 hours. The deviations from design seen in reformate composition were due to slightly different conditions inlet the prereformer.

Table 2 Reformate Composition (Dry Basis)


Table 2 Reformate Composition (Dry Basis)




Jet Fuel


Jet Fuel



0 0

Post a comment