Mcfc Stacks

A. TorazzaO, A Dufoui<°>, L. GiorgiW, J. G. MartínO, G. RocchiniC®), F. J. Simón« OAnsaldo Ricerche S.r.l., Corso Perrone 25,1-16161 Genova (°)ENEA, Via Anguillarese 301,1-00060 Roma WlBERDROLA S.A., C/Goya 4, S-28001 Madrid C®)ENEL S.p.A. - CRAM, Via Rubatüno 54,1-20134 Milano (*)Babcock & Wilcox Española, S.A Fabrica Galindo, S-48510 Galindo - Vizcaya


An interesting way of reducing the production costs of the electrical energy by improving efficicncy and, at the same time, having a good integration between environment and power plants is offered by the utilisation of the fuel cells operating at high temperatures. From this point of view, Molten Carbonate Fuel Cells (MCFCs) seem to be one of the most promising technologies because of their environmental friendly operation for various fuels and potential low cost In fact it is well known that the MCFCs overall plant efficiency is typically some 50% and can reach, as a consequence of their high operating temperature, 65% with a bottoming cycle. Moreover MCFCs will be particularly attractive for dispersed power plants of MW size located at user sites. Additional advantages of MCFCs are their good response to base and partial load, short time for plant erection and modularity.

However, the competitiveness of MCFC power plants with respect to the other plant types depends, in a significant way, on the achievement of some basic prerequisites concerning technical, economical and environmental aspccts. An important practical problem regards the process simplicity of the overall system and the proper choice of plant components as concerns their performance, reliability, cost reduction and lifetime.

In particular the MCFC stack, which is the key component, still needs improvements to reach high performance during long term operation as well as a close integration with the balance of plant (BoP), which is essential for the effective management of heat and gas flow. Another important goal concerns an easy maintenance and a low-cost replacement of the stack.

Ansaldo Ricerche, also within a technical co-operation with ENEL and ENEA engaged in the field of MCFCs since the early eighties, Babcock & Wilcox Española, IBERDROLA and their partners are facing the previous problems by taking advantage of their experience concerning the design and construction of a 100 kW MCFC stack which will be tested at the IBERDROLA facility in Guadalix, near Madrid. After this test, the stack will be placed in a 100 kW MCFC prototype (»generation power plant under erection, at the CISE premises, near Milan. These activities are carried out under two JOULE and THERMIE contracts from European Union (EU). As regards the THERMIE project, CISE is the ENEL representative.

In the frame of its co-operation with Ansaldo Ricerche, the role of ENEL is to support the technology development and assessment through the performance of specific activities concerning the physicochemical characterisation of materials and the operation of laboratory monocclls and small size stacks, and to provide basic ideas as concerns the plant type and configuration as well as its operation modalities.

General considerations

At present the main efforts of Ansaldo Ricerche and Babcock & Wilcox Española are addressed to improve MCFC technology, define optimum systems for fuel processing and re-design fuel cell stacks for application to MW class power plants mainly fuelled with natural gas and consistent with the market opportunities concerning dispersed and on-site applications. However some studies concerning the use of coal gas as fuel are also undertaken [1],

To this aim Ansaldo Ricerche and its Spanish partners defined their roles to reach in a short time and in a cost effective way the target of the joint program [2] according to their actual technological background. From this point of view the activities to be developed have been splitted in such a way that each partner, within its specific roles and expertise, may collaborate in the most important tasks of the joint program. Moreover, IBERDROLA, a Spanish electrical utility, will support some technical choices dealing with the BoP and definition of the basic criteria for connection to the grid and for operation modalities.

Some of the previous activities are performed within a new program partially founded by the EU, lasting three years and mainly focuscd to the development and assessment of cost effective stack through the re-design of the stack as well as the proper selection of electrode materials and repeat components by performing suitable tests also using one full area short stack.

Stack design

Many experimental activities aimed at supporting the design and construction of the basic stack module concerned the design, manufacturing and testing of several stacks having cells with an useful area of 100, 702 and 6760 cm2. This program permitted to assess the validity of the present technology, define proper operating conditions and individuate the critical parameters conditioning the integration of the stack in the plant Useful information about the manufacturing of repeat porous and metallic components were also gained so that it was possible to set up a pre-industrial production line, which permitted to define its running criteria and to evaluate properly the relevant costs and specific problems for mass production.

Actually a sound knowledge has been already obtained to review the basic requirements relating to the plant and stack and concerning the cost reduction, modularity improvement, control simplification, easy maintenance and high reliability. Other important aspects to be considered concern assembly procedures, start-up cycles, gas sealing, impurity tolerability and diagnostic tools.

The present technology for the construction of large size stacks is based on elements having a geometrical area of 0.75 m2 with a rated power ranging from 50 to 100 kW. Anodes are based on Ni-Cr alloys with a chromium content ranging from 2 to 5% and are formed using proprietary heat treatments under suitable atmospheres concerning debinding, pre-sintering, sintering, chromium pre-oxidation. After the previous operations anodes are filled with a mixture of Li and K carbonates having a proprietary composition. Cathodes, made of nickel, are oxidised in situ during the conditioning phase of stacks, performed using a suitable atmosphere, to give a lithiated nickel oxide structure. The electrolyte matrix, constituted by Y-LiA102, is placed in the stack in the green state and is typically made of some layers having an overall thickness less than 1 mm. Particular care is devoted to the selection of the y-LiAI02 powders to prevent matrix crack growth and thickness reduction during stack operation. The removal of binders is performed using a proprietary process during stack start-up.

Current collector and gas distributor are constituted by a single piece which is obtained by die forming. This component is carefully controlled to verify the respect of specifications as concerns its height which is very important in order to get a uniform pressure over all the contact region and to reduce significantly the value of the internal resistance. Cathodic current collector and gas distributor are made of AISI310 S stainless steel. Anodic current collector and gas distributors have a more complex structure based on a trilayer sheet made of nickel clad AISI 310 S stainless steel. Springs for cathodic and anodic regions are selected to operate properly at a temperature of about 700 °C. Separator plates are made of AISI 310 S stainless steel and the zones of the wet sealing are aluminised according to a proprietary procedure. Besides the filling of the anodes with Li and K carbonates, a further amount, necessary for a correct stack performance, is added in the anodic current collectors using a proprietary procedure. Its amount is determined-by applying a mathematical model, validated by several experimental observations, which permits to account for electrolyte losses due to corrosion and decomposition processes and migration under the influence of electrical fields.

The present stack design is based on the external manifolding configuration. The mechanical sealing at the operating temperature is obtained using a suitable assembly of a-alumina, zirconia felt and cloth. The design account also for the slide due to hot stack compaction under the axial mechanical load. External manifolds are properly matched to the stack by a system which assures a nearly constant pressure on the seals. The axial mechanical load is applied through a pneumatic bellow which operates with nitrogen and permits to regulate accurately the pressure during the stack start-up and operation. The use of a pneumatic bellow is very effective to reduce the contact resistance and entity of the gas leakage's through wet seals. A mechanical system becomes active when the value of the axial load is less than a prefixed threshold.

The current practice is to equip stacks with suitable sensors in order to collect the values of several physical quantities such as: cell voltage, internal resistance, cell temperature map, inlet and outlet manifold temperature at different regions and stack compaction versus operation time.

The first prototype of the 0.75 m2 technology is a stack having 20 cells which will be operated under a pressure of 3.5 bar. This stack will be fed with anodic and cathodic gas mixtures simulating the operating conditions of the 100 kW stack. In fact the «¡generative 100 kW plant utilises heat sensible reforming system to produce hydrogen rich gas from natural gas. At present the stack is in the assembly phase and its operation is planned at middle October, 1996.

Ansaldo Ricerche experience based on single cells and subscale stacks has shown that some critical aspccts of the present stack technology which includes also conditioning procedures are: aluminization of separator plate in the wet seal region, stack assembly procedures, stack conditioning cycle and non-repeat metallic parts. It is useful to underline that the previous problems are not of technical nature but mainly arise from an economical point of view because at present some of the previous operation are not cost effective.

In order to move towards the pre-commercial application of plants based on MCFC technology, the following intermediate targets have been fixed for the basic stack module:

• Power density 160 mW cm-2 at 750 mV/cell

• Operating pressure 3.5 atm

The previous goals are consistent with the technical and economical requirements of a pre-commercial phase and industrial application because the effective commercialisation requires only the improvement of manufacturing facilities for cost reduction and mass production

To reach the previous targets and improve stack design, a guideline handbook with reference to the stack configuration, mechanical load management, cell number, gas manifolding, choice of materials and/or components, assembly and maintenance* procedures will be defined. The influence of the "size-dependent" parameters will be also examined in order to identify and optimise the size of the commercial units. In addition to these activities, Ansaldo Ricerche is starting a specific technical program aimed at defining a new start-up cycle procedures, which should require less tight controls and envisage the possibility of performing such an operation directly on the plant, and at designing improved fabrication technologies of the non-repeat metallic parts in order to face a mass production with a low manufacturing time and effective costs.

Stack configuration will be defined also by using suitable mathematical modelling to support the choice criteria for single block or submodule arrangement This point is veiy important as concerns the start-up cycle, performance, reparability and maintenance. The mathematical modelling is also useful to define basic criteria and methodology for the design of new stacks and to select the main parameters affecting their sizing.

As regards the development of effective cathodes, the doped lithium cobaltite will be examined and specifications concerning total porosity, mean pore size, surface specific area, electrical conductivity and stability in molten carbonates will be established. Furthermore an experimental activity concerning anodes stabilised whit aluminium has been started and some specimens have been prepared. The characterisation of their performance on monocells is in progress.


Ansaldo Ricerche experience, concerning single cells and subscale stacks, shows that the present technology seems to be valid for the construction of stacks whit a size of about 100 kW even if the single cell performance as well as the anode and cathode stability can be improved.

Mathematical modelling of stack design and performance is an useful tool to reach the designed target and define the expected performance when the state-of-the-art materials are used. Mathematical simulation is also useful for a correct distribution of the process gases.

At last, a new crossed approach "system-to-stack" and "stack-to-system" has been defined, which should overcome the distortion due to the fact of considering separately the stack development and BoP specifications. This new approach permits a comparative analysis of mutual links of stacks, ancillary devices and BoP equipment by accounting for the use of the proprietary Compact Unit concept and comparing the effectiveness of the different fuel processing systems.


Part of the work described in this paper has been performed under the financial support of the European Commission in the framework of the JOULE programmes. The authors thank also International Fuel Cells for its useful technical contributions.


1. A. Dufour, A. Torazza, V. Regis, G. Rocchini, E. Hermana; "MCFC Compact Unit and System Improvements for 100 kW - 2 MW Dual-Fuel Demonstration Plant", Conference Paper Power-Gen Europe, Book II, Vol.5, p. 859, Amsterdam, 1995.

2. M. Brossa, A. Dufour, E. Hermana, J. F. Jimenez, F. Sanson, "The European MOLCARE Programme: a 100 kW Demonstrative Plant and Engineering Development of the MCFC Technology", Proc. Fuel Cells Seminar, p.238, San Diego, 1994.


Suk Woo Nam, Hyung-Joon Choi, Tae Hoon Lim, -Hwan Oh, and Seong-Ahn Hong Korea Institute of Science & Technology Seoul 136-791, Korea


Dissolution of NiO cathode into the electrolyte matrix is an important phenomena limiting the lifetime of molten carbonate fuel cell (MCFC). The dissolved nickel diffuses into the matrix and is reduced by dissolved hydrogen leading to the formation of metallic nickel films in the pores of the matrix. The growth of Ni films in the electrolyte matrix during the continuous cell operation results eventually in shorting between cathode and anode. Various mathematical and empirical models [1-4] have been developed to describe the NiO dissolution and Ni deposition processes, and these models have some success in estimating the lifetime of MCFC by correlating the amount of Ni deposited in the matrix with shorting time.

Since the exact mechanism of Ni deposition was not well understood, deposition reaction was assumed to be very fast in most of the models and the Ni deposition region was limited around a point in the matrix. In fact, formation of Ni films takes place in a rather broad region in the matrix, the location and thickness of the film depending on operating conditions as well as matrix properties. In this study, we assumed simple reaction kinetics for Ni deposition and developed a mathematical model to get the distribution of nickel in the matrix.

Model Development

Ni deposition is known to occur by the following reaction :

Ni2+ dissolved from the cathode and hydrogen dissolved at the anode diffuse opposite to each other and react within the pores of the matrix. The Ni deposition process is similar to the one used for the fabrication of inorganic membranes by chemical vapor deposition in an opposing reactants geometry [5,6]. But, transport mechanism of charged species like Ni2+ in the electrolyte is somewhat different from gas diffusion in the pores. For simplicity, consider the deposition of Ni in a cylindrical pore of the matrix with an initial pore radius r„ at any axial position z, as shown in Fig. 1. As Ni is deposited on the wall of the pore, the pore radius changes. From the deposition model, we can find the pore radius change as a function of pore axial position and time. The progress of Ni deposition can be characterized by the instantaneous radius calculated from the amount of Ni in the deposit layer using molar volume of Ni (pNj) and molecular weight (Mw-Ni) :

0 0

Post a comment