Studying the impacts of test condition and nonoptimal positioning of the sensors on the accuracy of the in-situ U-value measurement.

The non-destructive thermal characterization of building envelopes relies significantly on various factors such as climate conditions, monitoring devices used, indoor environment, and conditioning systems. In the case of both the temperature-based method (TBM) and heat flux meter (HFM) approaches, U-value is determined considering the ideal condition of steady state. However, it is challenging to accurately define the true thermal condition of buildings when monitoring is affected by inherent uncertainties of the chosen approach and inadequate instrumentation of building envelopes. This paper presents the outcomes of an experimental campaign, that aimed to evaluate the impact of incorrectly positioned exterior sensors, on the precision of U-value measurements. This study simultaneously employed the TBM and HFM approaches. To enhance the accuracy of the results, rigorous outlier detection and statistical analysis were employed on the data collected from three autonomous monitoring systems. The findings of this study revealed that the applied data analysis yielded more satisfactory results for the TBM approach compared to HFM. However, regardless of the approach used, the effectiveness of outlier detection relied heavily on the accuracy of the monitoring systems. When removing an individual outlier, the monitoring systems characterized with higher accuracies provided U-values that were closer to the theoretical values, than less accurate ones.