Multi-agent collaborative diffusion of CCUS in China’s thermal power industry under green financial policies: A stochastic evolutionary game model

As the largest carbon emitter, China's thermal power industry faces critical challenges in achieving dual carbon goals. This study constructs a tripartite stochastic evolutionary game model integrating government, financial institutions, and thermal power enterprises to analyze the collaborative diffusion of Carbon Capture, Utilization, and Storage (CCUS) technology under green financial policies. By incorporating Gaussian white noise to simulate real-world uncertainties and employing the Milstein method in MATLAB for numerical solutions, the key findings are as follows: (1) Gaussian white noise modeling accurately simulates CCUS diffusion dynamics across stakeholders, revealing uncertainty-driven interactions. (2) In China's tripartite decarbonization phase, thermal power enterprises show acute policy-market sensitivity, balanced by regulatory-financial institutional pillars. (3) Carbon trading has an asymmetric driving effect. A 20 % carbon quota cut shortens the enterprise decision-stabilization period by 20 % more than a 20 % price hike, yet it’s less effective than price regulation in stabilizing financial institutions’ strategies. (4) Green credit stability collapses below the 2.86 % rate threshold, while a 3 % rate optimizes system evolution. (5) Policy synergy has two optimal ranges. System evolution is best when technological investment subsidies hit 0.16 times the cost. Within the 3.15–4 times carbon-price penalty range, an optimal penalty multiplier can curb fluctuations, boost technology adoption, and promote strategy choices for all parties. This research provides theoretical and policy insights for CCUS diffusion in China, and based on the conclusions, suggests relevant policy adjustments to aid the low-carbon transition of China’s thermal power industry under its coal-dominated energy structure.