A coupled Computational Fluid Dynamics and Wells-Riley model to predict COVID-19 infection probability for passengers on long-distance trains

Abstract

Coupled Wells-Riley (WR) and Computational Fluid Dynamics (CFD) modelling (WR-CFD) facilitates a detailed analysis of COVID-19 infection probability (IP). This approach overcomes issues associated with the WR ‘well-mixed’ assumption. The WR-CFD model, which makes uses of a scalar approach to simulate quanta dispersal, is applied to Chinese long-distance trains (G-train). Predicted IPs, at multiple locations, are validated using statistically derived (SD) IPs from reported infections on G-trains. This is the first known attempt to validate a coupled WR-CFD approach using reported COVID-19 infections derived from the rail environment. There is reasonable agreement between trends in predicted and SD IPs, with the maximum SD IP being 10.3% while maximum predicted IP was 14.8%. Additionally, predicted locations of highest and lowest IP, agree with those identified in the statistical analysis. Furthermore, the study demonstrates that the distribution of infectious aerosols is non-uniform and dependent on the nature of the ventilation. This suggests that modelling techniques neglecting these differences are inappropriate for assessing mitigation measures such as physical distancing. A range of mitigation strategies were analysed; the most effective being the majority (90%) of passengers correctly wearing high efficiency masks (e.g. N95). Compared to the base case (40% of passengers wearing low efficiency masks) there was a 95% reduction in average IP. Surprisingly, HEPA filtration was only effective for passengers distant from an index patient, having almost no effect for those in close proximity. Finally, as the approach is based on CFD it can be applied to a range of other indoor environments.