Quantification of uncertainty in zero-flow pressure approximation

Abstract

Listen

Translate article

Multiple authors stated that, when performing fan pressurization test, Ordinary Least Square (OLS) method should not be used as a regression technique anymore. However, alternative methods require first to quantify components of uncertainty in pressure and air flow rate measurements. This work aims at quantifying the uncertainty in zero-flow pressure approximation, which is mainly due to short-term fluctuation of wind speed and direction. This has been done by statistically analysing the uncertainty indicator of 40 zero-flow pressure tests performed on 30 different units on eight different sites in Brussels. First, the analysis showed that this uncertainty could be reduced by increasing the period of measurement used to compute zero-flow pressure approximation. Second, it shows that the standard deviation of zero-flow pressure measurements was the variable with the most significant impact on the quality of the zero-flow pressure approximation. Third, it provides three different linear models to predict uncertainty as a function of different variables. This study experienced some limitations due to the available sample of tested units. These limitations lead to important further work: the validation of the model on another sample of buildings and its adaptation if needed. Further work should also focus on integrating these results on the uncertainty in envelope pressure measurements and on the uncertainty in airtightness estimation of the building.