Comparative assessment of quantitative infrared thermography approaches for experimental thermal transmittance determination using UAVs

Abstract

Rapid and non-contact evaluation of building thermal efficiency remains a major challenge in urban energy management. The increasing use of Unmanned Aerial Vehicles (UAVs) equipped with thermal cameras has expanded the potential of aerial thermography as a diagnostic tool for assessing building energy performance. This study compares five Quantitative Infrared Thermography (QIRT) methodologies (Methods A-E) for estimating the U-value of building envelopes, benchmarking them against theoretical reference values derived from in-situ wall characterization. The research analyzes the sensitivity of these methods to boundary condition uncertainties through two winter experimental campaigns in southwestern Spain. Results reveal that no single experimental method is universally robust; accuracy is critically dependent on the algebraic formulation of each method and specific environmental variables. Methods A and B exhibited extreme sensitivity to wind speed, showing deviations of up to 52 % with minor perturbations (±1 m/s), rendering them unsuitable without precise anemometry. Method E showed deviations of ≈40 % linked to internal surface temperature variations, while Method C proved highly volatile regarding external surface temperatures. Conversely, Method D, which incorporates both convective and radiative terms, emerged as the most balanced approach, showing intermediate stability (≈20– 27 % deviation) across all variables. Consequently, this study identifies the optimal application scenarios for each method based on field measurement capabilities, providing a decision-making framework to ensure the reliability of UAV-based thermographic assessments.