SOFCs (solid oxide fuel cells) are a promising technology for electric power generation, because of their high efficiency and fuel flexibility. Relative to more mature technologies, they appear very competitive starting from small-scale applications. The aim of this study is to map SOFC performances varying fuel composition, reproducing gases of technical interest (biomass air and oxy-gasification, water and carbon dioxide co-electrolysis, steam methane reforming) and to provide a forecasting tool to predict performance upon the fuel composition occurring in any application. Therefore, at first commercial SOFC are tested using several fuel gas mixtures produced by means of the chosen technologies (H-2, CO, CH4 molar fractions ranges are respectively 8-50%, 0-42%, 1-8%). Secondly, experimental data are used to build a mathematical tool to predict SOFC characteristic curve from the fuel composition. This result is interesting from the point of view of system-level modelling, because it allows characterizing commercial SOFCs performances in many technical applications. Forecast error is less than 1%. Because of its accuracy, the model can be implemented to assess properly SOFC performance as a consequence of any particular hybridization, whereas fuel composition falls into the range already defined. (C) 2015 Elsevier Ltd. All rights reserved.