Proposal for improving the linear regression method and uncertainty calculation in building airtightness tests

Improving the energy efficiency of buildings and the quality of indoor air requires accurate assessments of airtightness. The conventional regression method, Ordinary Least Squares (OLS) regression—as shown in ISO 9972—encounters challenges in the occurrence of fluctuating wind conditions, affecting the reliability of air permeability measurements. This study explores the potential of alternative regression techniques, specifically Weighted Least Squares (WLS) and Weighted Line of Organic Correlation (WLOC), to enhance the precision and reliability of building airtightness tests across varied environmental settings. Through the analysis of a comprehensive dataset derived from over 6,000 on-site blower door tests conducted on a multitude of house configurations, this research assesses the relative accuracy of these methods. Additionally, it introduces a new approach to global uncertainty calculation for the WLOC method. 
Findings indicate that while all methods exhibit similar in predicting the airflow at 50 Pa, WLS and WLOC_2 reduce prediction error by up to 6 percentage points at 4 Pa under wind speeds exceeding 4 m/s compared to other methods. At 50 Pa, the OLS 95% confidence interval covers the reference airtightness value for only 25% of the data, compared to WLOC_2 with 42% and WLS with 91%. At 4 Pa, while OLS interval covers only 21% of measurements, WLS overestimates uncertainty with 100% coverage, and WLOC_2 achieves 82% coverage. These results support incorporating weighted regression methods in airtightness testing standards.