Under the global imperative for climate action and sustainable development, accelerating the transition towards high-penetration renewable energy systems remains a universal priority, central to achieving the United Nations Sustainable Development Goals. However, the inherent uncertainty and volatility of renewables such as wind and solar PV pose fundamental challenges to power system stability and flexibility worldwide. These challenges, if unaddressed, could significantly hinder the reliable and sustainable integration of clean energy on a global scale. While pumped storage hydropower (PSH) represents a mature, large-scale solution for enhancing system regulation capabilities, existing planning methodologies frequently suffer from critical limitations. These included oversimplified scenario representations—particularly the inadequate consideration of escalating extreme weather events under climate change—and computational inefficiencies in solving large-scale stochastic optimization models. These shortcomings ultimately constrained the practical value of such approaches for advancing sustainable energy planning and building climate-resilient power infrastructures globally. To address these issues, this paper proposed a bi-level stochastic planning method integrating scenario optimization and improved Benders decomposition. Specifically, an integrated framework combining affinity propagation clustering and isolation forest algorithms was developed to generate a comprehensive scenario set that covered both typical and anomalous operating days, thereby capturing a wider range of system uncertainties. A two-layer stochastic optimization model was established, aiming to minimize total investment and operational costs while ensuring system reliability and renewable integration. The upper layer determined PSH capacity, while the lower layer simulated multi-scenario system operations. To efficiently solve the model, the Benders decomposition algorithm was enhanced through the introduction of a heuristic feasible cut generation mechanism, which strengthened subproblem feasibility and accelerated convergence. Simulation results demonstrated that the proposed method achieved a 96.7% annual renewable energy integration rate and completely avoided load shedding events with minimal investment cost, verifying its effectiveness, economic efficiency, and enhanced adaptability to diverse operational scenarios.

In day‐ahead electricity markets with high renewable penetration, price prediction errors are prevalent. These errors significantly increase the downside risk of energy storage arbitrage, potentially diminishing profits or even causing sustained losses. To address the lack of effective downside protection for energy storage systems operating in highly uncertain environments, this paper proposes a reinforcement learning–based battery‐dispatch method. The method is enhanced by three mechanisms to improve policy robustness and risk management capabilities. Residual injection disturbs predictive inputs to simulate various bias scenarios, guiding agents toward more conservative decision‐making. Action hard projection maps outputs in real time onto feasible regions, ensuring physical feasibility and training stability. Teacher model behaviour cloning incorporates low‐risk demonstrations based on actual prices, accelerating convergence and avoiding high‐risk actions. The approach underwent long‐term empirical validation using highly volatile data from the Germany–Luxembourg market for 2020–2024. Results indicate that, although the proposed method yields slightly lower average returns compared to the traditional prediction‐and‐optimization baseline, it significantly reduces maximum drawdowns, loss probability and profit volatility, thereby demonstrating robust downside‐risk protection. This study validates reinforcement learning’s capacity for effective risk control in energy storage dispatch and provides a viable pathway for robust asset management in highly volatile electricity markets.

In the global transition toward low-carbon power systems with high renewable energy penetration, pumped storage has emerged as a strategic cornerstone for modern power grids. However, the collaborative planning of pumped storage and backbone-grids faces critical challenges, including the lack of explicit quantification of the resilience value of pumped storage and the coarse treatment of N-1 connectivity constraints. This paper proposes a bi-objective resilient backbone-grid planning approach that integrates the pumped-storage hub effect, aiming to minimize total life-cycle costs and the system resilience mismatch index. The proposed framework incorporates network connectivity, N-1 connectivity (edge connectivity ≥ 2), and dual-scenario power flow security as rigid constraints. Furthermore, a three-stage constrained evolutionary algorithm TER-NSGA-II is developed. During the N-1 connectivity reinforcement phase, the max-flow min-cut theorem is employed to achieve precise validation and guidance for edge-connectivity enhancement. Case studies on the IEEE 118-bus system, together with extended validation on the IEEE 300-bus system, show that the proposed method can explicitly quantify the resilience value of pumped storage, obtain Pareto solutions that balance economy and resilience under strict edge-connectivity constraints, and demonstrate competitive overall performance in terms of solution-set quality, feasible-domain search stability, and scalability compared with NSGA-II and the more recent NSGA-III/NG benchmark.

In power systems with a high penetration level of wind power, wind turbines are required to temporarily extract stored rotor kinetic energy to provide primary frequency support, thereby compensating for the limited frequency regulation (FR) capability of the overall system. However, wind turbine participation may lead to underresponse (insufficient frequency support) or overresponse, potentially causing a secondary frequency dip (SFD). Electrolytic aluminum load (EAL), as an industrial load with FR potential, can rapidly adjust its active power input by controlling the electrolytic cell voltage, equivalently increasing the system’s FR capacity and thereby enhancing the load disturbance resistance of power systems with high wind power penetration. This paper first analyzes the causes and mechanisms of the SFD induced by wind turbine overresponse based on a Unified transfer function structure (UTFS) model and introduces the concept of a frequency stability region. Within this region, the virtual droop and virtual inertia coefficients for wind turbines are tuned to prevent SFD during FR. Simultaneously, by involving EAL in system FR, the analysis reveals that its participation essentially equivalently expands the system’s frequency stability region. Building on this analysis and considering spatiotemporal influencing factors, a coordinated wind-aluminum FR control strategy across multiple timescales is proposed to avoid SFD. Finally, a 39-bus simulation system built in DIgSILENT is used for validation. The simulation results indicate that the proposed control strategy effectively suppresses SFD under high wind power penetration conditions, and that the incorporation of EAL significantly expands the frequency stability region of the power system.

The increasing penetration of renewable generation and the rapid growth of data-center demand require coordinated use of heterogeneous flexibility while respecting network operating limits. This study proposes a three-objective electricity–computing–carbon dispatch method that minimizes physical system cost, direct carbon dioxide emissions, and unutilized renewable energy. Data centers are represented through task arrivals, deadlines, server capacity, and facility efficiency, whereas pumped-storage plants retain independent power limits, reservoir states, terminal conditions, and connection locations. A diversity-archive disturbance–multi-objective particle swarm optimizer (DAD-MOPSO) couples archive diversity and stagnation feedback with resource-group temporal-block disturbance. In 30 paired runs, DAD-MOPSO produced feasible final solutions in all runs and reduced mean spacing by 51.8% relative to conventional MOPSO, with a Holm-adjusted p-value of 0.0071. Repeated evaluation of 12 workload-resource cases showed that pumped storage provided the dominant improvement in renewable-energy utilization, while workload rescheduling provided a smaller marginal contribution. Factorial analysis indicated partially overlapping rather than universally superadditive flexibility. Fixed-total-capacity tests further revealed pronounced siting dependence: the same 1200 MW/7200 MWh storage capacity was feasible in 30/30 runs at Bus16 but only 16/30 runs at Bus27 because of undervoltage. The proposed framework therefore provides a unified basis for coordinating heterogeneous flexibility and distinguishing marginal, interaction, siting, and network-limited effects.