Journal of Korean Institute of Architectural Sustainable Environment and Building SystemsISSN:1976-6483(Print) 2586-0666(Online)한국건축친환경설비학회

This paper demonstrates a calibration methodology for adjusting model parameters and factors using assistant tools ― python and GenOpt. In this study, a building simulation model is developed for a demonstration building for a low-energy solar house. The model is composed of a 3D model describing building architecture and HVAC system model using TRNSYS simulation program. In this study, model calibration methodology consists of two phases in order to improve the accuracy of building energy model and reduce errors of simulation results: Parameter adjustment phase and factor optimization phase. Trial and error approach is used for the first phase and GenOpt tool is employed for the second phase to improve the model performance. Based on American Society of Heating, Refrigerating and Air-conditioning Engineers (ASHRAE) Guideline 14 indices, the performance of the calibrated building simulation model is evaluated to obtain Normalized Mean Bias Error (NMBE)_hourly, values within ±10% and Coefficient Variation of Root Mean Square Error (CV(RMSE))_hourly, values within 30%. The performance indices are calculated with simulated and measured temperatures of each indoor zone and buffer tank. Daily data is used for calibrating heat pump (HP) electrical energy consumption because it is hard to fit the operating time of HP which can be different to time that heating load occurs. As a result of the calibration process, it is found that the values of NMBE_hourly and CV(RMSE)_hourly achieves within followings: Average indoor temperature 一 NMBE_hourly is 1.2% and CV(RMSE)_hourly is 2.2%. Buffer tank temperature NMBE_hourly is 1.5% and CV(RMSE)_hourly is 8.2%. And the value of NMBE_daily and CV(RMSE)_daily are used for HP electrical use as following: NMBE_daily is 9.5%, CV(RMSE) is 23.4%.