VELARDE RUEDA, PABLO ANIBAL, FERNANDEZ CAMACHO, EDUARDO
No
Renew. Energ. Focus
Article
Científica
1
2
25/07/2026
001836067900001
2-s2.0-105045595145
The economic dispatch of concentrating solar power plants with thermal energy storage is strongly affected by uncertainty in irradiance forecasts. This paper proposes a risk-aware stochastic model predictive control (RA-MPC) strategy that combines expected operating cost with the Conditional Value-at-Risk (CVaR) of a composite operational index accounting for economic loss, dispatch deficit, low-storage exposure, and solar curtailment. Using linear auxiliary inequalities and the Rockafellar–Uryasev reformulation, the resulting receding-horizon problem is formulated as a convex quadratic program. Under the affine plant model, removing the CVaR term makes the optimal control sequence scenario-independent; the risk term thus breaks this degeneracy and enables forecast uncertainty to influence dispatch. The controller is evaluated on a 50 MW parabolic-trough plant with 300 MWh of thermal storage and benchmarked against deterministic MPC, scenario-based stochastic MPC, rule-based control, and robust minimax MPC. On six representative days, RA-MPC reduces the CVaR index by 27.4% and the dispatch deficit by 28.3% relative to the non-risk-aware stochastic MPC baseline. Over a 365-day campaign, it increases mean daily revenue by 32.6% (38.01 versus 28.67 kEUR/day), reduces dispatch deficit by 23.9%, and achieves 100% hierarchical solver feasibility. In a continuous 30-day stress test, RA-MPC limits the number of days ending below the 60 MWh terminal-reserve threshold to 3, compared with 11–21 days for the rule-based and minimax baselines. These results demonstrate that explicit tail-risk management can improve both economic performance and operational reliability in renewable generation with storage. © 2026 Elsevier Ltd
Electric load dispatching; Heat storage; Losses; Optimal control systems; Risk assessment; Risk management; Risk perception; Solar energy; Solar irradiance; Stochastic control systems; Stochastic systems; Concentrating solar; Conditional Value-at-Risk; Model-predictive control; Power; Risk aware; Risk-aware control; Stochastic model predictive controls; Stochastic optimizations; Stochastics; Thermal energy storage; Stochastic models