Hydropower systems consist of complex river systems making hydropower modelling computational expensive. The power output of hydropower stations are depending on the efficiency curve of a turbine, the head-height and the discharge through a turbine. The cascade ordering of reservoirs in river system results in further interdependencies and thus long simulations. Especially when simulating bigger energy system, e.g. the European system, a simplification is required. In earlier research, an 'equivalent model' is calculated by using a bi-level optimization problem with a particle swarm optimization (PSO) approach. The principle is to intelligently aggregate plants in one area into an equivalent model that mimics the behavior of the detailed model. Due to the simplifications, the risk of losing interdependencies is high. The equivalent model underestimates production in high price periods, and overestimates in low price periods respectively. This paper combines recently ideas and introduces two new adjustments that improves the equivalent performance.