A study on the global sensitivity of early-stage design parameters on energy consumption in classroom buildings of Beijing’s universities

To support early-stage energy-efficient design in university classroom buildings, this research surveyed 40 classroom buildings to establish an energy consumption characteristic parameters database in Beijing. Four typical archetypes—“Rectangle”, “L”, “U” and “Courtyard”—were identified. Using the Sobol global sensitivity analysis method, the study systematically revealed the influence mechanisms and sensitivity rankings of nine key design parameters on heating, cooling, and lighting energy consumption. Parameters with high, moderate, and low sensitivity levels affecting different categories of energy consumption were identified. Findings show that the sensitivity rankings are similar among the “Rectangle,” “L”, and “U” archetypes but differ in the “Courtyard” archetype for the same type of energy consumption. However, rankings vary significantly across different energy consumption types. Heating energy consumption is primarily influenced by the heat transfer coefficient of exterior windows (U window ) and the solar heat gain coefficient (SHGC). For cooling, SHGC is the dominant factor, while lighting is most affected by south and north window-to-wall ratios (WWR south , WWR north ). This study provides a reproducible methodological framework for sensitivity analysis of architectural design parameters and offers the data foundation for engineering optimization decision and energy efficiency standards.