Efficient Fuel Cell Models for SOFC/GT Hybrid Power Plant System Simulations Designed to be Parameterized with Experimental Results

Kurzfassung

Today, gas steam combined cycle plants are the most efficient power plants converting chemical energy into electrical energy with installed powers of usually several hundred megawatts. They are technologically mature reaching electrical efficiencies around 60 % (based on the lower heating value). Hybrid power plants consisting of solid oxide fuel cells (SOFC) and gas turbines (GT) can reach higher electrical efficiencies and can be built at smaller installed power [1]. For the design and simulation of process systems with fuel cells or electrolyzers as main components it is beneficial to develop electrochemical reactor models, which are fast, accurate and experimentally adaptable. Calculation time is an important factor in complex system simulations. To achieve a good accuracy in a wide operation range a physically meaningful model is required. Furthermore the model should be adaptable to different reactor types by using experimental results to find model parameters. In general the performance of a SOFC is significantly determined by the total area specific resistance (ASR) that accounts for ohmic, activation and concentration polarization losses. Literature provides various models [2, 3, 4]. However, these models are electrochemically complex. Therefore it is necessary to develop a model that can be parameterized through a limited number of experimental results and that provides a tradeoff between calculation time and accuracy. This contribution shows two experimentally validated models. Firstly, a simple ASR correlation is presented that is fitted to experimental test results of a commercially available stack. This model has a high computation speed and ensures a good performance of complex system models. The integration of a simple correlation into a 0D hybrid power plant model is demonstrated and simulation results are shown. Secondly, a flexible ASR model is developed that takes all polarization losses into account separately. This model is therefore applicable to several types of electrochemical reactors and can either be parameterized by material properties, experiments or a combination of both.