A fuel cell is an electrochemical device that generates electricity from continuously supplied streams of fuel and oxidant to porous anodes and cathodes. The two streams do not mix or burn but produce electricity by electrochemical reactions similar to a conventional battery. The details of the chemical reactions depend on the type of fuel cell, but in all types, an electrically charged ion is transferred through an electrolyte that physically separates the fuel and oxidant streams. The fuel cell thus provides an elegant means of converting the chemical energy of the fuel directly into electrical energy. Fuel cells represent an exciting power generation technology for the coming decades due to their high efficiency (80% including electrical and thermal), modularity, best environmental characteristics, siting feasibility and potential utilization of fuels such as hydrogen, methanol, natural gas, LPG, and coal gasified fuel. During the last 30 years, five major types of fuel cells have been investigated namely Alkaline (AFC), Phosphoric Acid (PAFC), Molten Carbonate (MCFC), Solid Oxide (SOFC) and Proton Exchange Membrane (PEMFC). For moderate temperature (200°C) and utilization of natural gas as a fuel, approximately 250 stationary PAFC power plants ranging from 40–200 kW are in operation in the USA and Japan.

Possible applications of Phosphoric Acid Fuel Cells using locally produced hydrogen in the Kingdom of Saudi Arabia include dispersed power generation, electrical buses and trucks, and remote power stations. To meet the increasing needs of a developing country like Saudi Arabia and to utilise abundantly available natural resources in the Kingdom, it was thought necessary to begin work on fuel cells as an alternative source of power.

The PAFC R&D activities at KACST’s Energy Research Institute were initiated in 1991 under the framework of HYSOLAR, a Saudi-German R,D&D Program on Solar Hydrogen Production and Utilisation. Within the task of Hydrogen Utilisation, it was envisaged that various components should be developed for half cells, mono cells, and stacks ranging from 100–1000 W. During the last five years, fabrication methods using a rolling technique have been developed at KACST to fabricate PTFE-bonded gas diffusion porous carbon electrodes ranging from 100–1000 cm² in size using platinized carbon catalyst powders (10–25 wt% Pt/C) with PTFE powder (40–45 wt%) as a binder as well as to provide the required hydrophobicity. These porous carbon electrodes were tested in half cells, mono cells, large mono cells, 100 W and 250 W PAFC stacks. The mono cells and stacks were operated continuously for 1000 h using hydrogen as a fuel and air as oxidant at 175°C and 1 bar, and 99% H₃PO₄, with minimum degradation, as reported by Ghouse et al.. Also, the electrodes were characterised using SEM, EDS, XRD, XPS, ICP-AES, and porosimetry techniques, which are reported elsewhere. Based on the R&D experience gained on the 100–250 W stacks, a 1 kW PAFC stack was designed and built with “in-house” developed components, including large size porous electrodes (1000 cm²), graphite bi-polar plates, and external gas manifolds. It was operated continuously at 175–180°C for 250 h without any degradation, using H₂ as a fuel and air as an oxidant at 1 bar. In this paper, test results of the 1 kW PAFC stack are presented.

To prepare the anodes, 10 wt% Pt/C catalyst powder was used, while 25 wt% Pt/C catalyst powder was used to prepare the cathodes using a rolling technique. The procedure adopted for the fabrication of the electrodes in the present investigation is the same as reported elsewhere. The final contents of PTFE in the anodes and cathodes were 40 wt% and 45 wt%, respectively. The anodes were sintered at 340°C, and the cathodes were sintered at 345°C in a nitrogen atmosphere. The matrix layers containing 6% PTFE and 94 wt% SiC with 7 wt% PEO (Polyethylene oxide) solution were applied on to the cathodes using a casting technique and sintered at 335°C in a nitrogen atmosphere. Graphite foils were used to fabricate the bi-polar plates with an impervious layer (Polyether sulphone films-PES) between the graphite plates having 25% porosity, and the graphite strips having 1% porosity. The detailed procedure adopted for the preparation of graphite bi-polar plates used to build the present stack is the same as described by Ghouse et al. for the 0.25 kW PAFC stack. Fig. 1 shows the arrangement of mono cells in a 1 kW PAFC stack. The anodes and cathodes were 30 cm × 20 cm, and their edges were sealed using PTFE tape and kept for 24 h in an oven at 175°C for acid absorption. These acid-soaked electrodes were then assembled in a stack of 30 cells with graphite bi-polar plates. There was a shortage of four cathodes containing 25 wt% Pt/C. as some cathodes with matrices were damaged during stack assembly, so cathodes made with 10 wt% Pt/C were applied after absorbing phosphoric acid for 8 h, replacing cell numbers 5, 15, 27, and 29 in the stack. These were made as substitutes for the cathodes made using 25 wt% Pt/C, based on earlier experience during the assembly of the 0.25 kW PAFC stack having six cells. It would be possible to study the performance of cathodes containing 10 wt% Pt/C in the cells of the 1 kW stack. In addition, it would be possible to compare the results using the cells containing both types of catalyst when the stack is in operation. Cured as well as uncured viton rubber was used at the edges of the external gas manifolds to prevent gas-leakage. Fig. 2 shows the 1 kW PAFC stack assembly. Although an acid management system was present, it was not used during the stack operation. Three cooling plates were used to cool the stack, i.e. for every eight cells one cooling plate was inserted in the stack. The k-type thermocouples were placed using conducting carbon paste on the grooves of graphite bipolar plates i.e. cathode-air side and the lead wires were taken out from the air outlet manifold for connecting to the data acquisition and control (DAC) system. When the stack was in operation, data were recorded using a data acquisition and control (DAC) system, the details of which are given below. The stack was successfully operated for 250 h continuously at a temperature of 175–185°C at the centre of the stack (cell numbers 13–18) with DACS, using H₂ and air at 1 bar, at an average cell voltage of 0.68 V/cell. The total output DC power generated by the stack was 1 kW (20.5 V × 50 A) with a load of thirty 40 W × 24 V lamps, with the H₂ flow-rate at 1050 l/h and air flow-rate at 12,000 l/h. All volumetric flow rates in this paper are given at the temperature of 25°C and 1 bar. Table 3 shows the operating conditions of the 1 kW PAFC stack, and Table 4 shows the specifications of the 1 kW stack. The calculations for number of moles of hydrogen required for a 1 kW stack are given below. The half-reactions for a hydrogen in fuel cell are: Anodic reaction: H₂ ? 2H⁺ + 2e^?, Cathodic reaction: 2e^? + 2H⁺ + 1/2O₂ ? H₂O. Overall reaction is H₂ + 1/2O₂ ? H₂O

Ref.,where N_fu is the Number of moles of hydrogen per sec, A the electrode effective area (400 cm²), J the Current density, mA/cm², n the number of electrons, mol-elec/mol, F the faraday constant 96,500 coul (A/S)/mole-elec, ?_F the Faradic efficiency (assume 100%).

(A) In the case of the current density 125 mA/cm² at 180°C (at the center of the stack) N_{fu=400 cm}²_{×0.125 A/cm}²_{/(2×96500×1)}=2.59×10^?4 g mol/s=0.9326 g mol/h (for one cell) For 30 cells hydrogen gas required=0. 0.9326 × 30 cells g mol/h=27.978 g mol/h Since 1 g mol of Hydrogen occupies at standard conditions, the fuel flow rate in liters per hour is then given by=24.441 l/g mol × 27.978 g mol/h=684 l/h for 30 cells. This means the hydrogen required for the reaction is 684 l/h. Assuming 5% losses are involved, the approximate consumption of hydrogen is 650 l/h. So the H₂ utilization (U_f)=650/1050=62%. Approximate fuel cell stack efficiency (?_FC)=U_f × V_a/V_r=0.62 × 0.68 V/1.23 V=34%.

(B) In the case of high temperature, the cd was 160 mA/cm² 185°C (at the center of the stack), N_fu=400cm²_×0.160A/cm²_{/(2×96500×1)}=3.316×10^?4 g mol/s or 1.1938 g mol/h (one cell). Hydrogen required=1.1938 × 24.441=29.18 l/h for one cell and 875 l/h for 30 cells. Assuming 5% losses are involved, the approximate consumption of hydrogen is 831 l/h. So the H₂ utilization(U_f)=831/1175=71%.

Approximate fuel cell stack efficiency (?_FC)=U_f × V_a/V_r=0.71 × 0.60 V/1.23 V=35%.