An innovative three-dimensional computational fluid dynamics-iterative ensemble Kalman filter model for the prediction of heavy gas leakage and dispersion in enclosed workplaces: Case study for hydrogen sulfide leakage

Abstract

Toxic heavy gas leakage in enclosed workplaces can result in severe safety accidents. Reasonable estimations of toxic heavy gas leakage and its subsequent concentration distribution are crucial for disaster assessment and emergency response. This study combines a three-dimensional (3D) gas dispersion model developed based on computational fluid dynamics (CFD) with the iterative ensemble Kalman filter (IEnKF) algorithm for accurate estimation of heavy gas leakage source parameters and detailed prediction of concentration distributions, with a case study of hydrogen sulfide (H2S) leakage. The inversions of two uncertain parameters are considered: H2S leakage velocity and air supply velocity. The results show that the 3D CFD-based gas dispersion model can work well with the robust IEnKF algorithm to predict the spatiotemporal concentration distribution of heavy gas with high confidence. Moreover, the multiple correction procedures performed in the IEnKF module can greatly improve the efficiency and accuracy of parameter estimations compared to the single correction of the EnKF algorithm. When the maximum number of iterations is set to ten, the assimilation time steps required to achieve satisfactory estimations of uncertain parameters are reduced by 86.67%; the accuracy of H2S leakage velocity and air supply velocity estimation is improved by 10.59% and 46.25%, respectively. In addition, after 13 assimilation time steps, the estimated H2S leakage velocity (17.2 m/s) is almost equal to the assumed true value (17 m/s). This study provides a novel approach to assess the impact of poisonous dense gas leakage in enclosed workplaces.