and higher ambient temperatures, the power output is reduced as limited by equipment size. The output is derated by about 8% at -20 F and 2% at 110 F.

The electric and cogeneration efficiencies are plotted as a function of ambient temperature in Figure 4. The efficiencies at -20 F are lower than those at 60 F because the colder air reduces the energy input to the fuel cell unit. The electric efficiency at 110 F is also lower than at 60 F because the additional energy brought in by the hotter air, as limited by equipment size, is not utilized for power generation. In addition, more power is required to compress the hot air. The additional energy in the hot air is transferred to the HRSG flue gas. As a result, the cogeneration efficiency is higher than at 60 F.

Figure 4

Effects of Ambient Temperature

Plot Area Required

The footprint required by the 1 MW market entry unit is compared to a standard tennis court in Figure 5. The 0.75 ft^/kW requirement is very small, and five such units can fit into an area of the size of a tennis court.

Figure 5

Small Plot Area

Figure 5

Small Plot Area

Total BOP Cost

The estimated BOP cost based on the first quarter, 1995 pricing, is shown in Figure 6. The 1 MW unit in the 20-50 units/year production range has a BOP cost of about$l,000/kW, which is the generally accepted cost goal for a successful market entry. A sensitivity study was conducted to determine the economy of scale if the power output were increased to 4 MW. In the 4 MW unit, the fuel cell skid and process skid consist of two parallel trains while the electrical skid has a single train. The increase in capacity from 1 MW to 4 MW reduces BOP costs by approximately 25%.

This study has not incorporated any design optimization. For example, the 3 atm system operating pressure was arbitrarily chosen without rigorous tradeoff analyses. The individual BOP components, such as the HRSG and inverter, have not gone through detailed mechanical design and packaging analysis. Additional design optimization efforts could yield significant benefits including further reduction in BOP costs, increased efficiencies, and an even smaller footprint.

Major BOP Cost Components

The major BOP cost components, as shown in Figure 7, are reformers, fuel cell vessel, turbogenerator,"HRSG, control valves, process skid assembly labor, and power conditioning unit.

The horizontal fuel cell vessel design has significantly cut down the cost of this equipment For the turbogenerator, the key issue is not to reduce the cost but to ensure that the projected cost can be achieved. The turbogenerator developer needs to work very closely with component and parts suppliers to determine the optimum design and manufacturing process. The HRSG design and cost estimate has not been optimized for the small-scale applications for fuel cell units. There should be ample opportunity to further reduce the HRSG cost, physical dimensions, and weight. Due to ever-improving electronics, the power conditioning unit has the highest cost reduction probability.

Figura 7

BOP Cost Distribution

Figura 7

BOP Cost Distribution


Fossil fuels continue to be used at a rate greater than that of their natural formation, and the current byproducts from their use are believed to have a detrimental effect on the environment (e.g. global warming). There is thus a significant impetus to have cleaner, more efficient fuel consumption alternatives. Recent progress has led to renewed vigor in the development of fuel cell technology, which has been shown to be capable of producing high efficiencies with relatively benign exhaust products. The tubular solid oxide fuel cell developed by Westinghouse Electric Corporation has shown significant promise. Modeling efforts have been and are underway to optimize and better understand this fuel cell technology. Thus far, the bulk of modeling efforts has been for operation at atmospheric pressure. There is now interest in developing high-efficiency integrated gas turbine/solid oxide fuel cell systems. Such operation of fuel cells would obviously occur at higher pressures. The fuel cells have been successfully modeled under high pressure operation and further investigated as integrated components of an open loop gas turbine cycle.

Air electrode supported-tubular solid oxide fuel cells, shown in Figure 1, developed by Westinghouse, have considerable design merit. The solid electrolyte eliminates problems typically associated with liquid electrolyte fuel cells. The tubular design allows for an innovative manufacturing technique and safer performance at the high operating temperatures of approximately 1000° C. Finally, the lack of a support tube significantly decreases the concentration polarization at the cathode (air electrode).

Figure 2

Figure 2

Modeling of fuel cell performance was based on the slice-by-slice, axial marching technique employed by Kanamura, (1). As illustrated in Figure 2, the fuel cell is divided into a preset number of subdivisions. Each slice may be viewed as a "minicell", with a given current, i„, flowing from slice anode to cathode. All the slices have the same potential difference, AV, between cathode and anode. The governing electrochemical equation for each slice is :

where £„ is the thermodynamic, equilibrium potential; and AVpoi, „ is the overvoltage due to activation, concentration and ohmic irreversibilities. The equilibrium potential was obtained via the Nernst equation:

£n~-AG°/nF-RJT/nF* In {nfproinv-['ro<1/nW1**} [2]

where "- AG0" is the standard Gibbs free energy decrease, "nF" is the charge transferred per reaction and "f' is fogacity. The fuel stream was an ideal gas mixture of hydrogen and steam; the oxidant stream was air. The overall reaction was hydrogen being oxidized to water.

The Nernst equation [2] simplifies to:

£n=- ACstam fonMtio„/2F - RJV2F* In {PfW Phz* Poix^ } m

The ohmic irreversibility was the dominant irreversibility for these fuel cells operating at a high temperature (minimizing activation polarization) and utilizing no support tube (minimizing concentration polarization). The ohmic losses were routinely greater than 90% of polarization losses.

Figure 3

The ohmic loss calculated in a given fuel cell slice was based on Ohm's Law:

AVn,ohmic=in*Rfl [5]

where R„ is the resistance of the slice to current flow. This calculation of resistance was based on the "transmission line" model (3). The proposed current path in a given slice is shown in Figure 3. Each half of a cell slice behaves as a resistor in parallel with the other half. Excellent correlation was obtained between model predictions and empirical results.

0 0

Post a comment