Diagnostic & health management of fuel cell systems: Issues and solutions

Continuous depletion of the crude oil and gradual increase in the oil price have emphasized the need of a suitable alternative to our century-old oil-based economy. A clean and efficient power supply device based on a renewable energy source has to be available to face this issue. Among the different technological alternatives, fuel cell power generation becomes a more and more interesting and promising solution for both automotive industry and stationary power plants. However, many technological hurdles have still to be overcome before seeing the development of industrial and competitive products in these fields. Among them, one of the major issues to be solved is their insufficient reliability and durability for stationary and transport applications. To reach this aim, efficient diagnostic and state-of-health estimation methodologies should be available, able also to operate real-time and with limited number of additional physical sensors. This paper describes the state-of-the-art and the motivations regarding these research issues. It presents also selected recent developments and experimentations in this area. Introduction In the last years, environmental concerns for sustainable development have reached increased attention from the world political, technical and scientific communities. In the general trend towards increasing renewable energies use, hydrogen-energy based technologies and fuel cell systems are more and more seen as key players in the forthcoming next energy mix [1]. Indeed, they are gaining performances and reducing manufacturing costs; moreover, they can be considered both for transportation applications and for stationary power supply infrastructure [2], thanks to the existing duality between two energy vectors: electricity and hydrogen-energy. Regarding transportation applications, fuel cell vehicles present the interests of a high efficiency versus classical internal combustion engine vehicles and a zero (in situ) pollutant emission level. Moreover, their autonomy and refilling time are quite similar to those of classical thermal vehicles. Regarding stationary applications, there is an increasing interest of using hydrogen for the long-term storage of electricity, especially when coupled to intermittent and seasonal power production coming from renewables (windmills, photovoltaic panels). Nevertheless, they are still technical and scientific issues to be solved, before seeing on the market competitive and efficient fuel cell systems, for a broad range of application areas. Among them, the most sensitive subjects are dealing with the global efficiency of the fuel cell system and the limited lifespan of the existing fuel cell systems, especially when considering hard operating constraints. As an example, and according to recent publications from the US department of energy, the maximal durability of a fuel cell systems under transportation operating conditions is today over 2500 hours, where 5000 hours are requested to reach the existing “standard” lifetime of internal combustion engines [3]. Considering global electrical efficiency of a fuel cell system, even if it can already reach about 40% versus lower heating value of the hydrogen fuel, there is still room for an increase of this efficiency [4]. To reach this aim, research on catalytic layers, electrolytic membranes, materials and design of bipolar plates must be pursued, but extending the lifespan and increasing the efficiency must also be seen from the systemic point of view. As a matter of fact, it is possible acting on this lifetime by reducing / mitigating the constraints imposed to the fuel cell stack itself, when still responding to the power demand in a very efficient way. For this purpose, diagnostic and state-of-health management tools have to be developed [5]. Regarding these diagnostic tools, that can relate both on algorithms and also hardware developments, four major research objectives can thus be expressed. First, the durability of the fuel cell stack and of the fuel cell system must obviously be increased. Second, still obvious, the global system efficiency must be increased. Third, the reliability of the fuel cell system should also be increased, or at least not lowered by the possible introduction of additional sensors or actuators. Fourth, in order to respond to varying power cycle profile without soliciting too hardly the fuel cell stack, an electrical hybridization of the fuel cell stack with electrochemical or electrostatic electricity storage devices must be considered. This will also lead to an increase of the dynamic performances of the fuel cell system. Besides these research objectives, as fuel cell systems should be able to enter the market at competitive prices, a constraint can also be added for this research activity: the use of a minimal number of actual sensors. Indeed, this constraint will reduce complexity within the system and costs induced by the new state-of-health diagnostic functionality proposed on the fuel cell system. Moreover, this is also quite conservative regarding reliability level. This paper is organized as follow. Section one will provide some recalls about fuel cell technology and fuel cell systems. Then, a second part will be devoted to the electrical behaviour description of a PEMFC (Polymer Electrolyte Membrane Fuel Cell), classically considered for both transportation and also stationary applications. The description of the different losses reducing the efficiency of the fuel cell system will also be given. Finally, this part will also provide some guidelines regarding the behaviour of a PEM fuel cell stack under faulty (or at least non nominal) operating conditions. The last part of the paper will then provide a flavour on ongoing works relating to diagnostic and health management of fuel cell systems. 1. Fuel cell technology and PEMFC systems A fuel cell is an electrochemical converter which continuously converts the chemical energy from a fuel and an oxidant into electrical energy, heat and other reaction products. The fuel and the oxidant are stored outside of the cell, and are transferred as the reactants are consumed. A cell is a stack of different layers: A porous anode: the gaseous fuel diffuses through the pores of the anode to reach the interface with the electrolyte able to conduct ions, where it is oxidized, electrons are conveyed from the anode to the cathode by an external circuit, A porous cathode: the gaseous oxidant (dioxygen) diffuses through the pores of the cathode to reach the interface with the electrolyte, and is reduced, An electrolyte which conducts the ions from one electrode to the other, The bipolar plates which convey the reactants to the electrodes, evacuates the reactants in excess, the product of the reaction (mostly water), the heat produced by the cell, Silicon seals to avoid gas and cooling fluid leakages. Different phenomena (electrical, electrochemical, fluidic, thermal) occurs that make a fuel cell a highly multi-physics object. Furthermore, these phenomena have very different characteristic time responses. The fastest concerns the double layer capacity effects: at the interface between the electrodes which conducts electrons and the electrolyte which conducts ions, charged particles are not displayed geometrically in a uniform manner. Then comes the transfer of charges due to the electrochemical reactions (oxidation and reduction), the gas diffusion through the electrodes, the water transport, the poisoning of electrodes due to contaminants in the reactants, the thermal changes, and then the degradation and ageing effects. Figure 1 illustrates that for a PEMFC, covering the range from microseconds to years, from 10s to 10s. Furthermore, the electrochemical reactions occur at the surface of the catalytic nanoparticles whereas gas are supplied through centimetric pipes. A fuel cell isn’t only a multi physics but also a timeand spacemultiscale and coupled object. To sum up in one word, a fuel cell is a complex system [6]. The thermodynamical maximum cell voltage is 1.23V. It means that in real life applications, the cells are associated in series in a stack to increase the voltage at a practical value. Different technologies of fuel cells exist. One very commonly used classification is based on the operating temperature. Three classes of fuel cells can thus be defined, operating at low, medium and high temperatures. This classification criterion is interesting because it has a significant impact on the structure of the cell, the balance of plant to operate it and its domain of application. The operating temperature conditions the quality of the heat produced which accompanies the electricity production. The higher the temperature at which the heat flux is produced, the better it can be exploited so as to enhance the overall yield of the system. Also, the higher the operating temperature, the less sensitive the system is to the presence of carbon monoxide in the reactants, which simplifies the processes of conditioning of the gases. Finally, medium and high temperatures enable us to forego the use of noble metals to catalyse the redox reactions, thereby greatly reducing the cost of the electrodes. Table 1 gives few characteristics of the fuel cell technologies, PEMFC: Proton Exchange Membrane Fuel Cell, AFC: Alkaline Fuel Cell, PAFC: Phosphoric Acid Fuel Cell, MCFC: Molten Carbonate Fuel Cell, SOFC: Solid Oxide Fuel Cell. The applications of fuel cell can be divided in three domains that can overlap: portable applications (supply of mobile phone, laptop ..., auxiliary power unit), stationary application (uninterrupted power supply, supply of domestic electricity, domestic CHP, high power CHP ...), transportation (supply of the propulsion or supply of the electric net board, in space, air, sea, terrestrial vehicles). Then, fuel cells can cover a wide range of power and heat generation, from mW to MW [7]. PEMFC and SOFC are currently the more investigated technologies. In this paper, PEMFC technology is focused. Table 1: Classification of fuel cell technologies Type of fuel cell Operating temperature range (°C) Electrolyte Type of ions exchanged Main application area Low temperature PEMFC [50°C-80°C] Polymer Membrane H + Low power portable applications Low power stationary applications Automobile/transport AFC [65°C-200°C] KOH OH Spaceships Medium temperature PAFC [180°C-250°C] H3PO4 H + Domestic heat and electricity co-generation (CHP) High temperature MCFC [600°C-700°C] Carbonate ion salt − 3 CO High power units for CHP, maritime applications SOFC [750°C1000°C] Oxidebased ceramic − 2 O High power units for CHP Fig. 1: Illustration of the multiscale nature of a PEMFC [8]. 2. Behavior and losses of PEMFC 2.1. Characterization and modeling A PEMFC is currently fed by pure hydrogen at the anode and oxygen at the cathode, either pure or coming from ambient air. Hydrogen is oxidized at the anode and then proton H flows from the anode to the cathode, through the polymer membrane, whereas electrons are collected by the bipolar plates and feed the cathode. At the cathode, oxygen, electrons and protons meet to produce water (Figure 2). This process leads to the production of electric power, thermal power and water. The electrolyte is a polymer which conducts protons only if it is well hydrated. Fig. 2: Schematic representation of a PEMFC operation. Reactions occurring at the electrodes are then :