Status Of Mcfc Stack Technology At

M.Hosaka, T.Morita, T.Matsuyama and M.Otsubo Ishikawajima-Harima Heavy Industries Co., Ltd.

Tokyo, Japan


The molten carbonate fuel cell (MCFC) is a promising option for highly efficient power generation possible to enlarge. IHI has been studying parallel flow MCFC stacks with internal manifolds that have a large electrode area of lm^. IHI will make two 250 kW stacks for MW plant, and has begun to make cell components for the plant. To improve the stability of stack, soft corrugated plate used in the separator has been developed, and a way of gathering current from stacks has been studied. The DC output potential of the plant being very high, the design of electric insulation will be very important A 20 kW short stack test was conducted in 1995 FY to certificate some of the improvements and components of the MW plant

These activities are presented below.


250 kW stacks will be installed at Kawagoe Power Station in 1998 FY. The 250 kW stack, shown in Fig.i, piles up two 125 kW sub-stacks in series. The sub-stack has 140 cells, that half of the cells, i.e., 70 cells, are making a block on an intermediate holder and the other are under the holder. Each 70 cells' block is compressed by springs instead of air bellow used for past stacks. Two 250 kW stacks and the sub-stacks are electrically connected in series, and generate total of over 500 volts at OCV. Thermal insulation will be set both around the vicinity of stack and outside of the vessel. We plan to monitor the performance every 5 cell instead of every cell as usual, because wiring a great deal of cable in a vessel is difficult and increases the probabilities of causing leakage current The number of cable is still great and should be much fewer in a future.


(1) Structure of separator

IHI has been adopting internally manifolded stack concept. It is important to keep good contact among cell components for both of effective area and wet seal area including internal manifold. Heights of cell components will change during long term operation at high temperature. To follow the thickness change of cell components, we select soft corrugated plate setting in the separator under wet seal. The soft corrugated plate shown in Fig.2 has elastic and plastic characteristics as shown in the figure. Though the corrugated plate shows relaxation for the first time load cycle, the curve does not change after that. It is sufficient for the soft corrugated plate to follow the component's thickness change. The corrugated plate also keeps strength enough under the high temperature operation by creep analysis.

(2) End cell performance

Most of the full sized stack we have tested before shows that the performances of end cell were usually lower than the other cells. This is mainly caused by the way of taking out load current from terminals of end plate. The current distribution of the end cell is affected by the potential distribution of the end plate, that gathers current from the stack. Fig.3 shows the potential distribution in the end plate, in case of the same current among the terminals. The current distribution vertical to the flow direction in a cell will bring fuel utilization's distribution non uniform among channels in a separator. This distribution reduces the performance, and according to circumstances this increases the decay rate of end cell. .

(3) Insulation

Stacks in the plant will produce high electrical potential. Piping, measuring cables, heater plates and holding system connected with stack should be highly insulated not to come about leakage current. Ceramic fiber or glass fiber is usually used as insulating sheet or covering. This material must be taken notice of these lowering electric non-conductance at high temperature. To reduce the probability of occurring leakage current by measuring cables, the cell's potential will be measured by every 5 cell instead of each cell as usual.


The 125 kW sub-stacks for the MW plant will be done pre-treatment and partial generating test at the factory of IHI. Then the stack will be transported by trailer to Kawagoe Power Station where MW plant will be constructed. Therefore the sub-stack should be gotten thermal cycling before operation at Kawagoe.

20 kW short stack was tested to certify cell components that will plan to be used for the 125 kW sub-stack, and to predict the performance of MW plant after thermal cycling. Fig.4 shows the 20 kW stack, which was installed in a pre-treatment vessel at IHI's atmospheric 50 kW stack test stand. The stack had two holding system, one was the current air bellow and the other was spring. To simplify holding system, the air bellow of the stack was removed during the thermal cycling, simulating operation without air bellows at Kawagoe.

(1) Results of test

Stack operating history is shown in Fig.5. Excepting for the hours of the planned thermal cycling, the stack completed a total of 1,793 hot hours with 1,536 hours of on-load operation and. generated 26 MWh electric power. Fig.6 shows performance of each cell before and after thermal cycling. There was no effect on stack performance as a result of the thermal cycling. Gas leakage and short current were almost little there.

(2) Prediction of 250 kW performance

The gas composition, being fed to the cathode side in MW plant, have very low concentration of carbon dioxide and oxygen. We usually take data at higher concentration than MW condition. Recently Morita et al. proposed a correlation of cathode oveipotential [1], which stands up even if at the low concentration. Fig.7 shows the relation to be good for performances of stacks [2]. The stack showed good performance at simulated MW condition, and got over the planned values. The decay rate of the stack under the simulated continuous operation of MW plant also satisfied them.

0 0

Post a comment