Uncovering spatial clustering and heterogeneity of energy consumption drivers of wastewater treatment plants in China: A shap-lisa integrated approach

Under China's "dual carbon" targets and the forthcoming 15th Five-Year Plan, improving energy efficiency in wastewater treatment plants (WWTPs) is a pressing environmental management challenge, yet the spatial heterogeneity of energy consumption remains poorly understood. To address this, we develop a policy-oriented analytical framework combining decomposition analysis, machine learning based prediction, and spatial interpretation to identify the drivers of WWTP energy consumption using a nationwide dataset. This approach is specifically designed to disentangle the multidimensional and spatially varying factors that influence energy consumption driver, offering a more nuanced understanding than conventional uniform assessments. Our results reveal pronounced regional disparities in energy use patterns. Treatment capacity and influent water quality are the dominant drivers, but their impacts vary significantly across regions. Notably, while pollutant removal requirements are site-specific, the broader socioeconomic and climatic determinants of energy consumption nevertheless show strong regional patterns. These findings provide a quantitative basis for replacing uniform national benchmarks with localized efficiency standards tailored to regional realities. By providing a spatially explicit and interpretable framework, this study offers actionable insights for aligning wastewater sector management with China's broader sustainable development goals.