Dynamic measurement of the airflow rate in a two-zones dwelling, from the CO2 tracer gas-decay method using the Kalman filter

Abstract

The measurement of natural airflows is challenging, because of low pressures involved in natural ventilation, and because of their variability. The tracer gas decay method is the most used method to assess natural airflows. It is, however, not adapted to multi-zone dwellings. A protocol based on the decay method is tested here involving the Kalman filter, and a correction procedure of concentrations. Both intend to allow the measurement of the global airflow-rate of a multi-zone passive-stack ventilated dwelling. The correction procedure of concentrations inhibits infiltrations that would have an impact on the accuracy of the tracer decay method implemented in a multi-zone dwelling. The Kalman filter allows measuring a dynamic airflow-rate, while the background tracer gas concentration varies which would bias the conventional decay method. It is a key issue when implementing the CO2 decay method in a multi-zone dwelling as, depending on its sources, the CO2 concentration in any room of the dwelling is likely to vary. The robustness of the proposed protocol was tested through a parametrical analysis in laboratory twincells and it showed a significantly lower sensitivity compared to the conventional 2-points decay method. Maximum deviations on the airflow-rate were systematically lower from 4 to 11 points. The concentration correction procedure allowed increasing the accuracy up to 190 points. The ability of the Kalman filter to assess the dynamic airflow-rate was questioned. It was strongly influenced by the process noise variance associated to state parameters of the Kalman filter, which are defined by the experimenter.