A Refined Framework for Energy Disaggregation Using Load-Temperature Plots in Cooling-Dominated Academic Buildings

Abstract

This study introduces a refined hourly-based Load-Temperature Plot (LTP) framework for disaggregating building energy use into base load (), dynamic operational load (), and cooling-sensitive load. By leveraging a temperature-insensitive zone and defining interpretable metrics— (base load) and (entire operational load)—the framework increases the granularity of energy-use disaggregation. Validated through hourly submeter data, the LTP-based baseline achieved an error rate of 0.72% for total energy use and 3.41% for cooling-sensitive loads. Application to 14 academic buildings revealed strong correlations among , , and Energy Use Intensity (EUI), confirming their reliability for benchmarking energy performance and suggesting scalability to other building types. Moreover, an experimental study focusing on nighttime air-conditioning optimization illustrates the framework’s capacity to guide base load reduction, yielding 13.97% energy savings across daily operations and a 43.1% drop in nighttime base load, all while maintaining essential operations. Overall, the refined LTP framework presents a scalable, actionable tool for energy management, enhancing sustainability and energy efficiency in cooling-dominated contexts.