Abstract
Due to a higher share of power production from renewable sources with high short-term variation, hydro systems must more often operate closer to their components' physical limits. To simulate system behaviour, a hydropower system simulator must therefore include most physical details. We present a simulator for hydropower investment analysis that combines a medium-term production planning model based on stochastic dual dynamic programming principles with a detailed and deterministic short-term hydro scheduling model. To reduce computation times, the system description for the short-term model may include only a snipped subset of the plants and reservoirs without deteriorating the results. The simulator is verified in a case study where an investment decision has been analysed for a Norwegian hydropower producer. The combination of medium-term optimization and short-term, detailed simulation is a useful decision support tool and provides both economic results and detailed physical information about the system behaviour.