Analysis of monitoring systems through the study of measurement uncertainty to improve the estimation of the Key Performance Indicators of the energy behaviour of the building's thermal envelope

The Energy Performance of Building Directive 2010/31/EU notes the importance of implementing an MCS to achieve quality assurance ((10288/17, 2017), (Brajterman et al., 2018)) and the importance of comparing EPC issued before and after the renewal of buildings. The Heat Transfer Coefficient (HTC) is a potential Key Performance Indicator (KPI) to be used [4] in the future Energy Performance Certification (EPC) of buildings, making it necessary to improve the technologies of Monitoring and Control Systems (MCS) integrated into Automated Building (AB), as well as having experimental tests for reliable HTC estimation.

To estimate reliable HTC, it is necessary to assess its uncertainty through the measurement uncertainty propagation of the physical variables involved in the HTC estimation. Where decreasing HTC uncertainty involves collecting reliable MCS data to estimate measurement uncertainties as low as possible. In turn, knowing the best technology allows the MCS to be optimized, providing the best sensor layout and technological accuracy, among others aspects.

Considering all of the above, in order to estimate a more reliable HTC for use in future EPC and to advance to an optimized MCS, this research study performed a 3D MCS in a tertiary building to identify the best location, precision and accuracy of the sensors through the uncertainty analysis of indoor air temperature. The results show the importance of sensor accuracy, in which there is a discrepancy between the manufacturer`s accuracy and the sensors’ experimental accuracy, showing how it can impact the HTC uncertainty. The main conclusion of this study is that the overall Temperature Uncertainty (𝑼𝑼𝑻𝑻) of the monitored thermal zones is 2.4 to 10.7 times bigger than the manufacturer's sensor accuracy. Thus, using the manufacturer's accuracy as the overall temperature uncertainty value could greatly underestimate the measurement uncertainty. Underestimating the measurement uncertainties of the HTC variables implies underestimating the HTC uncertainty, resulting in unreliable HTC estimations.