Present Status Of Some Technological Activities Supporting The Molcare Project

A. Torazza(°\ G. Rocchini™, M. Scagliotti''' '"'Ansaldo Ricerche S.r.l., Corso Perrone 25,1-16161 Genova (,)ENEL S.p.A. - CRAM, Via Rubattino 54,1-20134 Milano (,)CISE S.p.A., P.O. Box 12081,1-20134 Milano

Introduction

The development of MCFC stack technology is carried out at Ansaldo Ricerche in the framework of the MOLCARE project (1), a cooperation with Spanish companies under a partial UE funding, while a specific research program concerning the physico-chemical characterization of materials is performed jointly by CISE and ENEL (2). The project includes the development, the construction and the testing of a full scale 100 kW prototype, the assessment of stack technology on subscale stacks, the mathematical modelling of the MCFC based plants and the basic researches. The aim of the basic researches, carried out on single* cells, is to improve the effectiveness and durability of both the active and the hardware materials.

The Ansaldo stack technology is based on external manifolding. The full scale 100 kW prototype will be integrated with the sensible heat reformer and other ancillary equipments according to the "Compact Unit (CU)" concept (1). These technical choices stress requirements for manifold gasket configuration, electrolyte migration control, Ap management and porous component compaction.

Stack Technology

The present MCFC technology under development at Ansaldo Ricerche makes use of a green electrolyte matrix, of a nickel cathode, which is oxidised and lithiated in situ during the start up cycle, and of a nickel-chromium anode. A proprietary methodology is used to accomodate proper quantities of lithium and potassium carbonates in each cell.

Subscale stacks having the same active area and an increasing number of cells have been fabricated and operated in the last three years. In particular two stacks of 15 cells with active area of 702 cm2 were tested at atmospheric pressure at CISE for more than 1600 and 1800 hours of hot time respectively (3). In the latter stack the manifold sealings and electrolyte matrix impermeability to process gases were significantly improved.

In 1996 a stack of 50 cells with the same active area was tested for more than 750 hours of hot time using the Ansaldo Ricerche pressurized facility (fig. 1).

At the end of the start tip cycle an average OCV of 1042 ± 5 mV was noticed at approximately 600 °C (fuel: 25%H2-9%C02-11%H20-55%N2; oxidant: 86%air-14%COj). During the operation at atmospheric pressure most of the cells reached the expected performance of 740 mV at a current density j = 158 mA/cm2 with the fuel utilization Ur = 40%. This cell performance should correspond to a stack power output of 4.1 kW. However, the power output of about 4 kW was obtained, but at 206 mA/cm2 with Uf = 50% (fig. 2), due to the anomalous behavior of some bottom cells. The stack power output P(j) measured after 600 hours of hot time is shown in fig. 2 together with the voltage V(j) of the cell n. 30 starting from the bottom, which is representative of the average behavior of the properly operating cells. _ —

Gaschromatographic analyses were performed at regular intervals on the process gases and did not evidence cross-over phenomena through the electrolyte matrices.

Post test analyses are now in progress mainly to get information about the electrolyte sharing among the cells and the porous components, and also about the electrode compaction. In parallel a great effort was devoted to set up the facilities for the fabrication of porous repeat parts (1100 mm width continuous tape casting) and metallic components. A full scale stack assembling area under controlled atmosphere was also set up at Ansaldo Ricerche. The first stack of 20 cells having a useful area of 6760 cm2 is under construction. It will be tested within the end of this year.

Fig. 1 - The 50-celI stack tested at the pressurized Ansaldo Ricerche facility in 1996: (a) at the test facility, (b) after thermal insulation removal.
Current Density (mA/ctr?)

Fig. 2 - Behavior of P© and V© measured after 600 hours of hot time (Uf = 50 % at 206 mA/cm2) on the 50-celI stack tested at Ansaldo Ricerche facility in 1996.

Bench-scale Single Cell Testing

The basic activities on standard bench-scale single cells are currently being carried out at Ansaldo Ricerche and CISE. Tests of a few thousands hours are performed at atmospheric pressure and in pressurized conditions using square cells having an active area of approximately 50 cm1. The aim of this experimental program is to optimize porous electrodes and to test protective coatings on current collectors and housing components.

The cells are fed with fuel and oxidant gas compositions similar to those planned for the next full scale stacks and in particular for the first stack of 20 cells of 6760 cm2.

In the standard conditions of testing (fuel: 35%H2-23%C02-42%H20; oxidant: 86%air-14%C02; UH2=Uca2=60% at 150 mA/cm2) cell voltages of 700-710 mV were obtained at 150 mA/cm2 (fig. 3). When using a 52%H2-13%C02-35%H20 fuel and a 70%air-30%C02 oxidant (U1G = U^ = 40%) cell voltages exceeding 800 mV were obtained at 150 mA/cm2 with the present component technology.

Current Density (mA/cm?)

Fig. 3 - Behavior of cell voltage and power density as a function of the current density measured after 277 hours of hot time on the ARI-X33 cell fed with 35%H2-23%C02-42%H20 fuel and 86%air-14%C02 oxidant (Uf = 57% at 150 mA/cm2).

Materials Characterization

Both the porous electrodes and the protective coatings are characterized before and after the tests in stacks and single cells. This important task is mainly carried out, under the ENEL coordination, at CISE where different diagnostic tools are available including mercury porosimetiy, helium pycnometry, electron microscopy, X-ray diffraction, Auger and X-ray photoelectron spectroscopies. At present, the main effort is focused on the careful evaluation of the morphological properties of porous repeat parts, because of their influence on electrolyte sharing and on cell performance.

The combined use of mercury porosimetry, helium pycnometry and SEM image analyses provides useful data to improve the quality and the homogeneity of fall scale porous components. Also the chromium distribution in Ni-Cr anodes is under study in order to increase their mechanical stability. Satisfactory results were obtained on nickel coated anodic current collectors and ahiminized bipolar plates. The behavior of the shielded slot type anodic current collector made from Ni/AISI310/Ni threelayer was satisfactory both in stack and in single cell runs. Similarly, the effectiveness of the protective coatings in the wet seal area of bipolar plates was demonstrated by the SEM-EDS analyses of the samples of the subscale stacks.

More recently, the experimental methods for the determination of the electrolyte content and the sharing of tHe electrolyte among the different porous components (anode, matrix and cathode) were set up. Preliminary data have been obtained on the 15-cell stack operated at CISE in 1995 and on the 50-cell stack operated at Ansaldo Ricerche in 1996. These results are useful in order to define proper solutions for the electrolyte management in fall scale stacks with external manifolding.

Conclusions

The scaling up approach of the Ansaldo stack technology and the basic research activities on bench-scale cells and material characterization provided satisfactory results.

Some problems faced in the initial phase of cell and stack development were overcome. The new technology is promising because the cell performance is in agreement with literature data. Finally, farther efforts are needed to improve the homogeneity of the cell performance as well as the component stability and durability, even if our data show that cell power densities are already consistent with the final goal of the fall scale 100 kW prototype.

Acknowledgements

The authors thank ENEA, Babcok Wilcox Española, Iberdrola, Fabbricazioni Nucleari and International Fuel Cells for their useful technical contributions.,

References

1. M. Brossa, A. Dufour, E. Hermana, J. F. Jimenez, F. Sanson, "The European MOLCARE program: a 100 kW demonstrative plant and engineering development of the MCFC technology", 1994 Fuel Cell Seminar, San Diego, p. 238.

2. G. Rocchini, M. Scagliotti, A. Torazza, "Technological and basic activities on MCFC", 1994 Fuel Cell Seminar, San Diego, p. 573.

3. P. Araldi, L. Bigoni, A. Colombo, F. Gariboldi, B. Passalacqua, P. Savoldelli, M. Zappaterra, "Assessment of MCFC stack technology in Italy", 1994 Fuel Cell Seminar, San Diego, p. 570.

PERFORMACE OF NEW 10 kW CLASS MCFC USING Li/K AND Li/Na ELECTROLYTE

Yoshihiro Mugikura, Fumiliiko Yoshiba, Yoshiyuki Izaki, Takao Watanabe Central Research Institute of Electric Power Industry 2-6-1 Nagasaka, Yokosuka-slii. Kanagawa-ken 240-01 Japan Koh Takahashi. Sei Takashima, andToshiki Kahara

Hitachi Works, Hitachi. Ltd. 3-1-1 Saiwaicho. Hitachi-shi. Ibaraki-ken, 317 Japan

1. INTRODUCTION

The molten carbonate fuel cell (MCFC) uses generally mixture of lithium carbonate and potassium carbonate (Li/K) as the electrolyte, NiO cathode dissolution is one of serious problems for MCFC life 11The NiO cathode has been found to dissolve into the electrolyte as Ni" ion which is reduced to metallic Ni by H, in the fuel gas and bridges the anode and the cathode. The bridges short circuit and degrade cell performance and shorten cell life. Since solubility of NiO in mixture of lithium carbonate and sodium carbonate (Li/Na) is lower than in Li/K, it takes longer time to take place shorting by NiO cathode dissolution in Li/Na compared with in Li/K. The ionic conductivity of Li/Na is higher than of Li/K, however, oxygen solubility in Li/Na is lower than in Li/K. A new 10 kW class MCFC stack composed of Li/K cells and Li/Na cells, was tested. Basic performance of the Li/K cells and Li/Na cells of the stack was reported.

2. EXPERIMENTAL

The stack consisted of 26 cells. The cell area was 2520 cm2. The electrolyte of six cells, which were located at top of the stack, was Li/NaCOj (Li/Na=53/47 mol %). Other cells used Li/KC03 (Li/K=62/38 mol %). The weight of the separator plate was lighter than of the conventional plates since the separator plate consisted of thin stainless steel plates. Gas follow pattern was cross flow type. In the operation, the maximum temperature in the stack was kept lower than 680 °C. The stack was installed in the test facility of CRIEP1. after the initial performance clieck of the stack with heating up and cooling down processes at Hitachi Works site. The stack was operated at 3 - 7 ata. No gas recycling such as cathode gas recycling was applied.

3. RESULTS

Figure 1 shows effect of pressure on cell performance. The cell performance of Li/Na is lower than of Li/K at 3 ata. However, pressure gain of Li/Na cell is larger than of Li/K cell and is approximately two times larger than of Li/K cell. This pressure gain agrees with the result of small single cell(2). The maximum output voltage of Li/Na cell is higher than of Li/K cell at pressures higher than 5 ata. Each cell performance of the stack was analyzed by the following new method (3). Output voltage (K) is determined by following equation using open circuit voltage (E) and

0 0

Post a comment