Approaching generation variable costs from publicly available data

Generation variable costs are key parameters in economic studies of real power systems focusing on redispatch costs and power flows between regions, such as the ENTSO-E bidding zone review. These costs are commercial sensitive information unlikely to be disclosed, that depend on multiple time-varying factors that are difficult to determinate. This paper addresses the problem of estimating the variable costs of generation at national, regional and unit-scale using publicly available data. Different methods are tested to find the best correlation between spot prices and generation. Results are presented for France and Germany and demonstrate the usefulness of this approach providing a realistic estimation of the merit order between countries and technologies.