Scheduling in Concentrating Solar Power Plants Based on Model Predictive Control and Binary-Regularization Methods

Abstract

This paper proposes to use model predictive control with binary-regularization $(\mathbf{MPC}^{BR})$ to manage the generation problem in concentrating solar power (CSP) plants with thermal energy storage (TES). The main advantage of the proposed method is to allow rescheduling the generation at regular periods taking advantage of the most recent energy prices and weather forecast, and of the current plant's state. A second benefit of the $\mathbf{MPC} ^{BR}$ formulation is the inclusion of a turbine protection method based on a binary-regularization term that penalizes for changes of the turbine output and binary constraints that limit the number of daily turbine startups. Binary variables are used to avoid penalization of the turbine start-up and shutdown. These extensions of the classical MPC approach aim to preserve the turbine lifetime. An interesting question is if this protection mechanism affect the economic results of the CSP plant. An included economic study shows that the proposed scheduling method provides a good trade-off between the economic profits obtained from energy sales and the protection of the generation block. The study is based on a realistic simulation of a 50 MW parabolic trough collector-based CSP with TES under the assumption of participation in the Spanish day-ahead energy market scenario.