Data-driven building energy benchmark modeling for bank branches under different climate conditions

Abstract

Energy benchmarking in buildings is a significant analytical method to determine and classify the energy performance of buildings. The objective of the present paper is to establish a compre-hensive data-driven energy benchmark model for banking buildings. In this context, 587 bank branches are selected to cover every climatic region in Turkey, with more than 210,000 data obtained from the installed energy analyzers and building energy audits. A mathematical model is developed and verified to define the EUI of the existing bank branches. Calculated energy scores and benchmark study implied an energy efficiency potential across the sector. Investigating the results of underperforming buildings, the building shell and construction date are found to be significant. As results indicated an efficiency potential, a scenario-based energy benchmark model is proposed based on the new insulation regulations in Turkey to see the possible energy con-sumption and carbon gas emission reductions. The developed model is adapted to the scenario -based model by revising U-values concerning the regulation and climate zones. The scenario -based strategy has led to a 2136.7 MWh/year reduction in energy consumption. The total EUI was found to decrease from 159.44 to 124.35 kWh/m2year, also resulting in a 20.5% improve-ment equal to 2057.3 equivalent tons of CO2/year in annual GHG emissions with the application of new regulations.