Multi-Market Joint Optimization and Cooperative Benefit Allocation for Wind–PV–Thermal–Storage Aggregators Considering Battery Degradation Cost
Abstract
High penetrations of wind and photovoltaic (PV) generation increase renewable-output uncertainty, balancing-market exposure, and the demand for flexible ancillary services. This paper proposes a multi-market stochastic optimization and cooperative benefit-allocation framework for a wind–PV–thermal–storage aggregator considering BESS degradation cost. A price-taking three-stage offering model is formulated for participation in the energy, spinning-reserve, frequency-regulation, peak-shaving, and balancing markets, with non-anticipative offer-curve construction, thermal-unit commitment, BESS intertemporal constraints, imbalance settlement, and equivalent-throughput degradation cost. A cooperative game is then constructed from the maximum expected net profits of all 15 non-empty resource coalitions. The strict Shapley value is retained as the benchmark, while a Shapley-based flexibility-adjusted allocation uses explicit leave-one-participant-out contribution indices and an individual-rationality-constrained adjustment parameter. For the representative test system, expected net profit increases from CNY 1.6433 million in Case 1 to CNY 1.7340 million in Case 5, a cumulative increase of 5.52%. BESS integration accounts for a 1.35% increase relative to Case 1, while subsequent expansion from the BESS-inclusive Case 2 to Case 5 provides a further 4.12% increase relative to Case 2. Scenario-number, ancillary-market-size, and degradation-cost sensitivity analyses show limited variation in the main economic results over the tested ranges. The strict Shapley allocation slightly violates BESS individual rationality; the final minimum adjustment, α = 1.22 × 10−4, restores individual rationality for all participants. The ordinary core is empty because of a small CNY 367.97 non-superadditivity gap, and the final allocation attains the individual-rationality-constrained least-core bound.