Day-ahead bidding strategies for hybrid power plants using probabilistic generation forecasts

Hybrid Power Plants (HPPs) face increasing market price volatility and an appropriate design for their Energy Management System (EMS) could turn it into an advantage. Traditionally, control approaches are supported by deterministic forecasts and would benefit from a probabilistic approach to better manage market risks. This study investigates day-ahead bidding strategies to enhance HPP’s profitability through EMS thanks to the development of probabilistic generation forecasts. The case study consists of a wind-PV-battery hybrid power plant located at Risø, Denmark. Day-ahead (DA) and imbalance price forecasts are deterministic while quantile-regression forecasts are generated for wind and PV production. The generated quantiles are then employed to simulate realistic Monte Carlo scenarios by preserving autocorrelation at lag 1. The EMS is modeled through the HyDesign-HiFiEMS tool with two approaches: (1) An EMS based on deterministic DA forecasts only. (2) An EMS with a stochastic optimization. Results show that the stochastic optimization approach leads to a 4.7% increase in profitability coming mainly from better spot market revenues and reduced battery degradation.