Carbon emissions in cold-climate railway halls: automated batch simulation and sensitivity analysis based on spatial design variable combinations

High-speed railway stations in cold regions typically feature large spatial volumes and long operating hours, resulting in substantial heating energy demand and associated carbon emissions. This study examines the impact of spatial design parameters on the carbon emission performance of waiting halls in cold-climate railway stations, utilizing Harbin West Station in China as a case. To address the limitations of conventional single-case simulations, a fully automated parametric simulation framework was developed by integrating EnergyPlus with Python scripting, enabling efficient batch evaluation of multiple spatial design scenarios. Based on this framework, 125 spatial configurations were generated by systematically varying the window-to-wall ratio, skylight-to-roof ratio, and ceiling height, and their carbon emission intensities were calculated. Cluster analysis and multiple linear regression were applied to classify spatial configurations by carbon performance and to quantify the sensitivity of emissions to individual variables and their interaction effects. The results indicate that the skylight-to-roof ratio and ceiling height are the dominant parameters influencing carbon emission intensity, exhibiting pronounced nonlinear interaction effects, whereas the window-to-wall ratio shows a limited marginal impact under heating-dominated conditions. The proposed framework provides a scalable and data-driven methodological basis for low-carbon spatial design and retrofitting of large-scale public buildings in cold regions.