Considering Reactive Power in Generation Expansion Planning for Energy System Design

Abstract

Planning models for studying future energy system designs predominantly use the "DC" power flow approximation, a linearization of the AC power flow equations which ignores reactive power. This can lead to results that are physically infeasible ("AC-infeasible"). We investigate whether the use of an approximation which incorporates reactive power can improve the designed system with respect to AC-feasibility. As a first experiment, we solve a generation expansion problem using aggregated data representing the European energy system. The results show that the AC-feasibility improves significantly with the alternative approximation, leading to 97% AC-feasible snapshots, whereas the DC approximation only leads to ~73%. Since we observe an increase of one billion € in capacity investments and 239 million € in operating costs, this can likely be attributed to the improved visibility of grid bottlenecks. Hence, in our experiment the DC approximation leads to an underestimation of the total system costs by 1.239 billion €.