Performance investigation of 2.4 kW PEM fuel cell stack in vehicles

https://doi.org/10.1016/j.ijhydene.2007.05.043Get rights and content

Abstract

The aim of this paper was to investigate the behaviour of a 2.4 kW PEM stack in dynamic conditions typical of vehicle applications, including urban driving cycles taken from European legislation. A management strategy of the main fuel cell system auxiliary components was chosen with the scope of minimizing the energy consumptions associated to humidification, cooling and reactant feeding operations. The results obtained provided the definition of a correlation diagram power–temperature, in particular the application of a statistical indicator of the cell voltage uniformity (coefficient of variation Cv) allowed the specification of a reliable working field in terms of stack durability in real applications. The tests effected on European driving cycles R40 and R47 evidenced that stack power increase rates up to 500 W/s were acceptable for the fuel cell system used in this paper, while for higher dynamics the stack working regularity was limited by low stoichiometric ratio values (R<1). However, during fast transient phases (10 s as order of magnitude) excursions outside the optimal conditions of power, temperature and R resulted acceptable in terms of Cv (mostly lower than 4%).

Introduction

The energy politics of the developed or emergent countries need to deal with two critical issues, i.e. to improve their security in energy supplying, by decreasing their dependence on oil, and to meet the requirements connected to global warming due to green-house gas emissions. In this respect hydrogen fuel cell technology represents a promising solution for the replacement of internal combustion engines in both stationary power plants and road transportation means [1], [2], [3], [4], [5]. However, while today hydrogen is produced from fossil fuels via natural gas reforming as well as the partial oxidation of heavy fuel oil (or diesel) and coal, its utilization in a clean energy production chain involving the fuel cells would imply the exploitation of renewable sources, such as hydro, wind and solar [6], [7], [8].

The use of polymeric electrolyte membrane fuel cells fed by hydrogen is the most suitable for automotive application, because of their high power density, low operative temperature, fast start-up procedure, good dynamic performance [9]. A fuel cell stack has to be inserted in an electric power train in hybrid configuration with an electric energy storage system (batteries or supercapacitors), which are always necessary to guarantee the proper response of the power train to the most rapid power requirements coming from vehicle utilization on the road [10], [11], [12], [13], [14], [15], [16], [17]. However, their contribution should be minimized if size and weight are critical issues in vehicle design. In this case the stack power should provide almost the total of energy requirements coming from the engine.

The above considerations imply that a fundamental task to be faced in realization of a fuel cell power train is the optimization of the fuel cell system in such a way to make it able to follow the dynamic demands of road utilization with minimum fuel consumption and elevated working reliability, taking into account that this dynamic behaviour can be limited by the interaction between the stack and its auxiliary components necessary for fuel, air and water management [18], [19], [20], [21], [22].

The present paper deals with the study of stack management strategies suitable to its utilization in a fuel cell power train in real conditions. In particular, a 2.4 kW polymeric electrolyte membrane fuel cell fed by pure hydrogen was coupled to an 3.5 kW electric drive and a Pb battery pack. The hybridization level between stack and battery was chosen in order to minimize the contribution of the storage system to engine requirements. The study was effected by the experimental analysis of the interaction between stack and purge, humidification and air supply devices during tests designed to investigate stack response to fast load changes. During the tests a control strategy was selected taking into account the criterion of minimizing hydrogen losses, compressor power consumption and external humidification. An optimal operation region in terms of stack power and temperature was detected during steady state tests and verified on European driving cycles R40 and R47. The dynamic response of stack in relation to air compressor control was also checked on the same cycles, during which two different power increase rates could be realized (100 and 500 W/s).

Section snippets

Experimental

The fuel cell power train used for the dynamic experimental tests reported in this paper is shown in the scheme of Fig. 1, with some details about the main auxiliary components for stack management, while Table 1 reports the main technical characteristics of all power train components.

The power train was based on a 2.4 kW PEM stack (Proton Motor GmbH) fuelled with pure hydrogen at low pressure, equipped with all the auxiliary components necessary to the fuel cell operation (air and hydrogen

Results and discussion

The stack management implies the necessity to regulate the stack temperature together with water and reactant control parameters in order to avoid mass transfer limitations and membrane drying-out or flooding. However, the utilization of a fuel cell as power generation system in an electric power train requires the study to be effected with particular regards to the dynamic conditions typical of road demands. Then the interaction between stack and auxiliary components has to be calibrated on

Conclusions

The experimental results reported in the present paper have provided the general lines of a control strategy to be applied to PEM fuel cell system in automotive field.

A management strategy was selected for the main auxiliary components and verified in both steady state and dynamic conditions, allowing to individuate a optimal working region in the power–temperature correlation diagram. The coefficient of variation Cv was adopted as statistical indicator of the cell voltage uniformity. The

Acknowledgement

The authors gratefully acknowledge Mr. Giovanni Cantilena of Istituto Motori for his cooperation in realization of the experimental apparatus and execution of tests.

Cited by (0)

View full text