Evaluation of the precision and accuracy of the QUB/e method for assessing the as-built thermal performance of a low-energy detached house in UK

Abstract

In this paper, the findings from a field study of the thermal performance of a low-energy dwelling are presented and discussed. The study aimed at evaluating the precision and accuracy of the QUB/e method in the field under UK climatic conditions. A series of in situ measurements were carried out in a low-energy detached house located in the University Park campus, University of Nottingham to determine both the whole building Heat Transfer Coefficient (HTC) and the thermal transmittances (U-values) of the external walls and the glazings. The values determined with well-established quasi-static measurement methods (coheating test, ISO 9869-1 average method) were used as reference (benchmark) values to assess the accuracy of the QUB/e method. In total 18 consecutive QUB/e tests were conducted.The QUB/e method was able to provide precise and accurate estimates of the whole building HTC: 94% of the results (i.e. 17 out of 18) were found to be within ± 15% from the mean and 78% of the HTC estimates (i.e., 14 out of 18) obtained with a QUB/e test could not be considered statistically different from the HTC determined with the coheating test (benchmark). It should be noted that the reported confidence intervals associated with the HTC estimates obtained with the QUB/e tests were slightly larger than the one reported for the coheating test (i.e., 11.6% in average and 6.1%, respectively). The Mean Bias Error (MBE) and the Root Mean Square Error (RMSE) associated with the HTC estimates obtained with the QUB/e tests were approximately 11%.The QUB/e method was also able to provide precise and accurate estimates of the thermal transmittances of the external walls and the glazings. We observed that all but seven (i.e., 119 out of 126 or 94%) U-values estimates of the external walls obtained with a QUB/e test could not be considered statistically different from the U-values determined with the ISO 9869-1 average method (benchmark). With regard to the glazings, this was observed for all U-values estimates (i.e., 162 out of 162 or 100%). It should be noted that the reported confidence intervals associated with the U-value estimates obtained with both test methods were close (i.e., approximately 10% in average) for all elements except for the ICF wall.