Distributed Balancing of Wind Power Forecast Deviations by Intraday Trading and Internal Ex-ante Self-Balancing -- A Modelling Approach

Abstract

Wind power generation does on the one hand contribute to a less polluting and more sustainable electric power generation mix. On the other hand, its power output is variable and subject to forecast errors. In real-time, deviations from wind power forecast are handled by the system operator. But expected deviations can already me minimised by power generating companies before real-time. Ways to decrease their expected deviations are intraday trading and/or re-scheduling of own power plants. Both can be regarded as forms of self-balancing before the period of delivery (ex-ante), intraday trading as a form of external ex-ante self-balancing and re-scheduling as internal ex-ante self-balancing. Both can decrease the need for real-time balancing by the system operator. As existing intraday markets are often plagued by low liquidity, it is important to model such not-perfectly liquid intraday markets and simulate different trading and scheduling strategies. This paper presents an approach to model the choice between purely internal self-balancing and internal self-balancing combined with intraday trading on a not fully liquid intraday market. Results from the model runs indicate that intraday trading on a not-perfectly liquid market can be beneficial from the producers as well as from the system's perspective. However, it can result in increased costs if the possibilities to trade on the intraday market are very limited. This is important to consider when investigating the question whether it would be beneficial to distribute a larger share of the balancing responsibility among the power generating companies in order to relieve some pressure from the system operator.