Aggregation Methods for modelling hydropower and its implications for a highly decarbonised energy system in Europe

Given the pursuit of long-term decarbonisation targets, future power systems face the task of integrating the renewable power and providing flexible backup production capacity. Due to its general ability to be dispatched, hydropower offers unique features and a backup production option not to be neglected, especially when taking the flexibility potential of multireservoir systems into account. Adequate hydropower representations are a necessity when analysing future power markets and aggregation methods are crucial for overcoming computational challenges. However, a major issue is that the aggregation must not be a too flexible representation. In a first step, a novel equivalent hydro system model implementation including a possibility to integrate pumping capacity and appropriate handling of multiple water paths (hydraulic coupling) by making use of an ex-ante optimisation is proposed. In a second step, a clustered equivalent hydro system model implementation employing k-means clustering is presented. A comparison of both aggregation approaches against the detailed reference system shows that both aggregated model variants yield significant reductions in computation time while keeping an adequate level of accuracy for a highly decarbonised power system scenario in Europe. The aggregation methods can easily be applied in different model types and may also be helpful in the stochastic case.