Hydropower reservoir management using multi-factor price model and correlation between price and local inflow

Abstract

Hydropower producers with reservoir capacity have a special challenge when it comes to weighing the short-term profit from selling power in the day-ahead spot market against waiting for better electricity prices. In this paper, we propose a medium-term scheduling model for a price-taking hydropower producer, using a horizon of two years. We use the price of forward contracts to forecast future spot prices, and use multiple factors to describe movements in price. Further, we include a short-term correlation between movements in electricity price and local inflow. Our main contribution is a comparison of the performance of our scheduling model to a model in which price and local inflow are assumed to be independent and a model in which price movements are described using only one factor. We quantify the loss in expected revenues of using the latter two models compared to the case where price movements are in fact driven by multiple factors and correlated with local inflow. In both situations, we find the loss to be approximately 2-3 %. We have based our study on a Norwegian hydropower plant.