Identification of indoor airborne contaminant sources with probability-based inverse modeling methods

Indoor environment quality exacerbation calls for effective control and improvement measures. Accurate and prompt identification of contaminant sources ensures that the contaminant sources can be quickly removed and contaminated spaces can be isolated and cleaned. This study discusses the use of inverse modeling method to identify potential indoor pollutant sources with limited pollutant sensor data.;The study reviews various inverse modeling methods for advection-dispersion problems and summarizes the methods into three major categories: forward, backward, and probability inverse modeling methods. The adjoint probability inverse modeling method is indicated as an appropriate model for indoor air pollutant tracking because it can quickly find source location, strength and release time without prior information. The principles of the adjoint probability method are then introduced. Corresponding adjoint equations for both multi-zone airflow models and computational fluid dynamics (CFD) models have been established. The developed inverse algorithm successfully and quickly identifies the source location with known release time in two CFD and two multi-zone cases.;Subsequent efforts are focused on tailoring the method for even complex source identification tasks, including release time identification with known source location, simultaneous time and location identification, location identification of constant and dynamic point source, and location identification of multi-point or area source. Afterwards, the study proposes a two-staged inverse modeling approach integrating both multi-zone and CFD models, which can provide a rapid estimate of indoor pollution status and history for a whole building.;In order to ensure a smooth application of the algorithm in the real world source identification, a thorough sensitivity analysis has been carried out. Variation in air velocities, measurement errors and mass range and the resulted changes in calculated location probabilities have been quantified and studied, which illustrates the robustness of the inverse algorithm. A CO2 experiment performed in an apartment is then performed to test the effectiveness and accuracy of the method. The predictions from the method are successfully verified against the actual experiment, indicating good capability of the algorithm in finding indoor pollutant sources in reality. The method finally finds another important application in the contaminant sensor network design. A novel approach that revolutionizes the sensor system design has been proposed and verified to be able to significantly facilitate the fast and accurate indoor contaminant source identification.