Attribution analysis to Co-planning renewable energy and storage capacity based on Shapley Additive Explanation

In global power decarbonization, planning decisions for diverse resources must account for the complex interplay of technical and policy requirements. The impact mechanisms of heterogeneous factors, such as resource endowments, and technology costs, on planning schemes remain unclear. It is essential not only to rely on optimal planning schemes, but also to elucidate the internal optimization logic for decision-makers, quantifying the influence of various factors on planning decisions. Using Shapley Additive Explanation, an interpretability framework for co-planning schemes is designed, taking the collaborative planning of renewable energy and storage as an example, with the internal optimization logic of the planning revealed from both local and global interpretability dimensions. This framework quantitatively analyses the impact of different factors on co-planning for Northwest power resources during decarbonization. When the proportion of renewable energy is relatively low, technology costs drive the co-planning in Northwest China. However, as low-carbon reforms deepen, carbon costs and storage capacity costs become the primary drivers, with carbon costs ideally set at 23–32$/ton to enhance planning benefits. Moreover, the interactive effects of different factors impact planning nonlinearly. Wind power costs below 210$/kW and short-term storage capacity costs above 147$/kWh could drive an approximately 20 % expansion in long-duration storage capacity.