Defining the calibration process for building thermal performance simulation during the warm season: a case study of a single-family log house
Matic Možina, Luka Pajek, Pravin Diliban Nadarajah, Manoj Kumar Singh, Mitja Košir
Abstract
The study aimed to develop a holistic, evidence-based calibration strategy and define a hybrid step-by-step methodology for building thermal simulations during the warm season. The methodology was demonstrated using the thermal performance of a naturally ventilated single-family log house in a temperate climate. The calibration process was supported by on-site measurements of indoor thermal conditions and occupant behavioural reports between April and July. Hybrid calibration methodology combines statistical indicators, such as NMBE, CV(RMSE), and R2, and a graphical evaluation of the deviation between the simulated and measured values. After 28 proposed calibration steps, a final calibrated model predicted the actual thermal response of the log house with an accuracy of ±1°C at 71.6% or ±2°C at 98.4% of the considered time. The results showed that occupant-related information, material properties and infiltration rates are crucial to improving the reliability of building thermal models. Moreover, in the calibration process of log houses, appropriate determination of building envelope materials’ thermal conductivity, wood's solar absorptivity, and airtightness showed high importance. The study highlights the importance of calibration for reliable building thermal models, specifically in overheating prevention simulations.