The air supply system is one of the most critical components in the energy balance of a fuel cell (FC) plant. Particularly, for automotive application the amount of energy required to drive this component can share up to the 25% of the total electrical energy generated by the stack. Such a high energy level is required to achieve air pressure above 2 bar to meet two automotive application strategic targets: high fuel cell efficiency and high power to weight ratio. These two goals involve trade-offs between plant design, control strategies and costs, and finding a proper solution is one of the most troublesome problems faced by both FC makers and automotive companies. The aim of the work is modelling the FC compression system to perform a comparative study of different compressors. This study wish also to address some key-features on the suitability of each compressor with respect to the main system design issues (e.g. energy balance, powertrain performance, system control). In the paper a modelling analysis is presented for four different compressor types to analyze their specific capabilities with respect to the whole system performances and to the energy balance of the plant. The models developed are part of a computational library linked to a dynamic model of a hybrid electric vehicle equipped with a PEM FC and a battery pack. According with the technologies actually considered as being able to the meet the requirements of the FC application, reciprocating, sliding vane, twin-screw and centrifugal compressors have been studied. The modelling approach is based on both the thermodynamic description of the plant and the synthesis of experimental efficiency data. Such a hybrid modelling approach leads to the attainment of opposite goals in term of computational sfficiency and accuracy, and has been preferred with respect to a more complete theoretical one that would have been too complex, computationally expensive and somehow incomplete. Because of its low computational burden the whole model allows performing transient analysis within real-time studies aiming at i) plant optimization; ii) control system design and iii) development of on-board model-based control algorithms. Moreover, the modular computational structure conceived has the indirect benefit of being easily generalizable to different devices belonging to one of the four categories analysed. A normalized dataset derived from performance maps provided by manufacturers together with essential geometrical information are the basic information required to perform simulations. The proposed method allows extending the study to different systems and configuration through interpolations on a compressors database with some physical correction factors. Several analyses have been carried out to test the behaviour of each compressor, as part of a hybrid electrical vehicle powertrain, over a set of standard transient manoeuvres (i.e. load ramp and step) and for a 2 hours real driving cycle. Afterward an energy analysis has been conducted to determine the most efficient system with respect to a performance index defined as function of the achieved vehicle performance targets. From these analyses, it emerges that the best system can be selected after a trade-off between compressor technology, energy saving, powertrain performance and global vehicle control strategy.
|Titolo:||Energy analysis of PEM fuel cell compressors|
|Data di pubblicazione:||2005|
|Appare nelle tipologie:||4.1.2 Proceedings con ISBN|