Uncertainty in Building Energy Performance Characterization: Impact of Gas Consumption Decomposition on Estimated Heat Loss Coefficient

Characterization of building energy performance indicators such as the Heat Loss Coefficient (HLC) based on in-situ measurement data calls for thorough building physical insight, a well- designed measurement set-up to collect sufficient, qualitative data and adequate data analysis methods. On-board monitoring may be an alternative for dedicated experiments to perform the data collection task. This paper analyses the sensitivity of the end-result of the characterization, the HLC estimate, to flaws in the monitoring data set. More specifically, the impact of not installing submeters to disentangle the gas consumption for space heating and the production of domestic hot water is evaluated. Hereto, multiple gas decomposition methods are applied on a case study monitoring data set, after which the HLC is assessed. The results show deviations up to 33% for the mean estimate. Nevertheless, the 95% confidence intervals largely overlap.