Infrared thermography for quantitative thermal performance assessment of wood-framed building envelopes in Canada

Abstract

Since many buildings in Canada were built prior to the advent of national and provincial energy codes and standards, quantifying building envelope thermal performance is an important step in identifying retrofit opportunities in existing buildings. This study aimed to use external quantitative infrared thermography (IRT) to estimate effective U-value of opaque building envelopes (considering the effect of thermal bridging sources) of a conditioned at-scale structure comprised of four wood-framed wall assemblies commonly used in Canada. Furthermore, the effect of vignetting artefacts on effective U-value measurements was assessed, followed by a practical approach to correcting for it to improve accuracy of U-value estimation and calibration of energy models. Additionally, a comprehensive uncertainty analysis was performed to evaluate the impact of input variables on the accuracy and uncertainty of results. Finally, apart from qualitative and quantitative thermal assessment of the building envelope, a novel relative quantitative infrared index (IRI) methodology was proposed as a means to facilitate rapid evaluation and subsequent ranking of building envelope thermal performance. The results indicated that vignetting effect has an adverse impact on the accuracy of results, in particular for well-insulated walls where deviations of −42.31% to −83.33% were observed. However, when the proposed practical approach was implemented, substantial improvements in accuracy of walls’ U-value were obtained, ranging from −2.33% to −12.50% after correction versus −13.95% to −58.33% without correction. Moreover, the results indicated that the energy model was substantially more accurate when the effect of thermal bridges were accounted for, and the adverse effect of vignetting was addressed in the estimation of U-value. In this case, ASHRAE Guideline 14 criteria were satisfied: Normalized Mean Bias Error (NMBE) < 5%, and Coefficient of Variation of the Root Mean Square Error (CVRMSE) < 15%. The findings of the uncertainty budget demonstrated that the influence of parameters on U-value depends on the type of wall assembly. Ultimately, wall thermal performance rankings based on IRI were consistent with their U-value rankings, implying that IRI can be a reliable metric for relative quantitative comparison of building envelope thermal performance, regardless of boundary conditions.