Applying Spatial Decomposition in Energy System Models

The European energy transition requires the expansion of renewable generators and consequently, the expansion of grid infrastructure and flexibility options. When modelling complex energy systems, the application of decomposition techniques is reasonable to keep models computationally tractable. Our paper focuses on the efficient optimization of energy systems covering large geographic areas. To accelerate the solving process, we apply a variant of Benders decomposition, which does not only apply temporal but also spatial decomposition. We show that by applying our decomposition approach, the solving time decreases up to 75% compared to the runtime of the equivalent monolithic model solved by a commercial solver and the barrier algorithm. Furthermore, we show that additionally applying spatial decomposition results in time savings compared to an only temporally decomposed approach.