Cost-effective and reliable power system transitions under deep uncertainty: A multi-scenario multi-objective robust decision making framework

• A MORDM framework is proposed for robust planning under deep uncertainty. • Distinguishes constraint-related and objective-related uncertainties in planning. • Incorporates hundreds of deep uncertainty scenarios and accelerates solutions. • Considers distinct planning approaches in the search phase for candidate solutions. Power system transitions toward high renewable penetration face deep uncertainties in economic, technological, and climate dimensions. The classical agree-on-assumption paradigm, which optimizes solutions under prespecified scenarios, may lead to costly or unreliable outcomes when alternative futures materialize. Several agree-on-decision paradigms have been proposed for decision making under deep uncertainty, yet their application in power system transition remains limited. We propose a multi-scenario multi-objective robust decision-making (MORDM) framework that explores diverse candidate solutions and identifies those that are cost-effective, reliable, and robust across multiple scenarios. The framework distinguishes between constraint-related uncertainties (such as climate projections) and objective-related uncertainties (such as techno-economic costs) and integrates four planning approaches and solution techniques: single-scenario based robust decision making (RDM), multi-scenario stochastic programming (SP), min–max cost robust optimization (RO), and min–max cost regret. Case studies on China’s power system transition toward carbon neutrality by 2060 demonstrate that RDM and RO are Pareto-dominated by robust regret in terms of cost, cost regret, and energy not served (ENS). SP achieves slightly lower total costs but results in higher ENS and cost regret than robust regret.