Uncertainty estimation of investment planning models under high shares of renewables using reanalysis data

Abstract

Abstract Variable renewable energy, particularly wind and solar, increasingly affects present and future energy systems. Reanalysis products are often used as a meteorological basis for subsequent energy system studies under high shares of variable renewable energy. Since reanalyses contain considerable biases, it is ex-ante unclear how they impact the results of energy system models. For that purpose, this study applies the high resolution reanalysis data set (6 km grid) to a full model chain, from the meteorological basis via power conversion to the optimal electricity system investment and dispatch, and investigates the implications of initial errors. It is shown that power conversion as well as electricity system model results are highly sensitive to these errors associated with the meteorological input data, in particular under high shares of variable renewable energy. The initial errors are observed to propagate through the model chain with at least the same magnitudes, which leads to distinct impacts on the overall configuration of the German electricity system as well as allocation effects of variable renewable energy capacities.