Modelling of PV-Powered Water Electrolysers

In the near future the hydrogen production by means of advanced water electrolysers powered by renewable hybrid energy systems (Photovoltaic solar/wind) could help to resolve the electricity supply and environmental problems relating to the use of fossil fuels. In the light of this perspective the hydrogen represents an alternative energy carrier, helping to overcome all the problems related to the intermittent nature of solar and wind sources. A non linear dynamic simulator of a photovoltaic-hydrogen energy system has been realised, aiming to provide a useful instrument for the development of innovative strategies for plant control and plant operating guidance. The lumped parameter physical approach has been used, applying the fundamental conservation laws of mass, energy and momentum to every component of the plant. The water electrolyser model has been tailored on the characteristics of an advanced pressurised system, using a Casale Chemicals S.A. advanced cell bipolar design, with alkaline electrolyte (KOH solution), whose mathematical models was described by the authors in previous papers. A first version of this simulator has been improved by introducing a reliable thermal model, able to predict the solar panel temperature profile that affects the PV array performance; the panel model has been modified in order to reproduce precisely the I/V characteristics of any PV module, starting from its nominal data. Thanks to this model improvement, the simulator allowed to be used to maximise the PV power production, evaluating different control strategies: a Maximum Power Point Tracking (M.P.P.T) block has been then introduced in the model to optimise the generated power by the photovoltaic plant. The Joule losses due to the PV field internal wiring and to its feeding connection with the electrolyser have been also considered: it consents to separately compute the energy losses in the different PV-electrolyser coupling configurations, thus evaluating the best panel disposition in order to minimise the electric power dissipation. The simulator proved to be able to robustly predict the performance of the PV-electrolysis system for different configurations, operating conditions and control strategies. A steady-state analysis not appears in fact to be an adequate tool for these purposes.