Impact of Different Forecast Horizons in Energy System Simulations

Energy system simulations can be used to optimize energy systems towards minimal investment and operational costs, usually focusing on multiple decades. A perfect forecast for the whole timeframe is usually assumed. In recent years, myopic optimizations have been increasingly introduced into energy system models, to reduce the computational complexity of energy models or to allow a more accurate depiction of the energy system. This paper aims to compare different forecast horizons and their impact on the results of energy system simulations. The following three fundamentally different approaches were considered: perfect foresight, myopic optimization, and a custom simulation in which an energy system is newly built from scratch in every simulated year. Results show that in this studied case, the total costs of an energy transition did not differ much with varying forecast horizons. Very similar final energy systems were achieved with different technology-building pathways. The time period used in this paper is from 2020 to 2060. As a benchmark model, a simplified version of the German energy system based on hourly values was used.