Multi-step building energy model calibration process based on measured data

Building energy models are a key element in regulatory compliance calculations. These performance calculations often do not accurately reflect actual operating conditions. Therefore, evaluation of comparing use building with the outcome dynamic simulation can be misleading, this difference is also known as gap. The reduction gap an important task aimed to provide confidence for efficiency. This paper focused on reducing technical issues (e.g. poorly adjusted thermal parameters envelope, inefficient boiler operator or lack adjustment heat pumps, baseboard radiators air handling units) which one main causes application multi-step, optimization-based, calibration methodology performed white-box environment (EnergyPlus) using three months ten minute time-step data adjust HVAC parameter values genetic algorithm software (Jeplus) validated real test site. Resulting BEM that fits building’s hourly benchmark into international standards levels: indoor temperature by Thermal Zone (TZ), production and electric consumption from comprise all components model. A batch 1500 h heating operation, obtained management system, has been used calibrate results complied requirements American Society Heating, Refrigerating Air-Conditioning Engineers (ASHRAE) Guideline 14–2002 at interval, NMBE ?±10%,Cv(RMSE) ?30% R2 ?75% International Performance Measurement Verification Protocol (EVO) Cv(RMSE) ?20% aforementioned levels, considered step forward area calibrating white box models. In addition, prove strength robustness results, model checked long testing independent period 2.500 operations same level compliance. demonstrator library school located Denmark. system composed four air–water pumps deliver whole compound backup support gas boiler. heated unit ventilation purposes.