Open Access

ARTICLE

Fuzzy Chance-Constrained Bi-Level Dispatch of CSP–PV Systems with Source–Load Interaction and CCER-Based Deep Peak Regulation Incentives

Menghao Zhou¹, Jie Chen^2,*, Zhuang Zhao³, Liming Huang³
1 School of Electrical and Energy Engineering, Shanghai Dianji University, Shanghai, China
2 Intelligent Manufacturing Modern Industrial College, Xinjiang University, Urumqi, China
3 Ultra-High Voltage Branch, State Grid Xinjiang Electric Power Co., Ltd., Urumqi, China
* Corresponding Author: Jie Chen. Email:

Energy Engineering https://doi.org/10.32604/ee.2026.087256

Received 13 June 2026; Accepted 16 July 2026; Published online 22 July 2026

Abstract

High photovoltaic (PV) penetration increases net-load fluctuations and weakens the incentive adequacy of conventional deep peak regulation compensation for concentrated solar power (CSP) plants. This paper proposes a fuzzy chance-constrained bi-level dispatch framework for a CSP–PV system that coordinates source–load interaction, CSP deep peak regulation compensation, and PV-side Chinese Certified Emission Reduction (CCER) revenue sharing. The upper level reshapes the load profile through price-based demand response (PBDR), while the lower level coordinates thermal units, CSP with thermal energy storage (TES), and PV generation under network, reserve, carbon, and settlement constraints. Simulations on a modified IEEE 30-bus system show that PBDR reduces the load peak-to-valley difference by 60 MW and the net-load peak-to-valley difference by 94.5 MW. Relative to the Scenario 5 benchmark comparison dispatch (Scenario 5-BCD), the controlled comparison shows that CSP deep peak regulation under the complete mechanism releases 185 MWh of additional PV accommodation space and eliminates the remaining PV curtailment. Compared with robust optimization, the proposed fuzzy chance-constrained programming (FCCP) model reduces operating cost by CNY 56,700 and avoids 42 MWh of PV curtailment. Compared with a conventional single-level mixed-integer linear programming (MILP) model, the proposed bi-level model reduces operating cost from CNY 1.2806 million to CNY 1.2359 million and cuts thermal carbon emissions by 214 tCO₂. The results show that PBDR, deep peak regulation compensation, and CCER revenue sharing provide complementary incentives for low-carbon flexibility.

---

Keywords

CSP–PV system; deep peak regulation; CCER revenue sharing; price-based demand response; fuzzy chance-constrained programming

---