Virus aerosol transmission, dispersion, and infection probability simulation: A case study in subway carriages

Abstract

BackgroundSubways are typical congregate settings and may facilitate aerosol transmission of viruses. However, quantified transmission probability estimates are lacking. PurposeTo model spread and diffusion of respiratory aerosols in subways by simulation and calculation of infection probabilities. MethodsThe internal environment of carriages of Shanghai Metro Line 10 was used to establish a study scene. The movement of tiny particles was simulated using the turbulent model. Trend analysis of infection probabilities and viral quantum doses was conducted in a closed subway carriage scene by a quantum emission-infection probability model. ResultsUnder a typical twelve-vent air conditioning configuration, respiratory droplet aerosols within a subway carriage dispersed rapidly throughout various regions due to airflow, with limited short-term diffusion to other carriages. Concurrently, owing to the uncertainty of airflow patterns, the airflow might circulate and converge within carriages, causing delayed outward dispersion or hindered dispersion of droplet aerosols upon entry into these zones. Passengers boarding the carriage could exacerbate the formation of these zones. When the air conditioning system functioned adequately (air exchange rate=23.21 h−1), the probability of a virus carrier transmitting the virus to other passengers within the same carriage via aerosol transmission was approximately 3.8%. However, in the event of air conditioning system failure (air exchange rate=0.5 h−1), this probability escalated dramatically to 30%. Furthermore, a super-spreader (with virus spreading exceeding 90% of the average) elevated the infection probability to 14.9%. Additionally, due to the complexity of turbulence within the carriage, if local diffusion occurred in 1/2 zones of a carriage, the anticipated infection probability would increase to 8.9%, or during the morning or evening rush hours leading to elevated aerosol concentrations, the infection probability would rise to 4.7%. The subway transmission probability for common coronaviruses diminished to as low as 0.9%. ConclusionCombined computational fluid dynamics and infection probability analysis reveals that in the prevalent twelve-vent air conditioning configurations, despite being a major transportation hub with substantial spatial-temporal overlap, the internal space of subway carriages exhibits a certain level of resistance to virus aerosol transmission owing to built-in ventilation capabilities. However, turbulence and passenger positioning may lead to localized hovering of droplet aerosols, thereby increase the risk of virus transmission. Furthermore, super-spreaders, poor operational status of built-in air conditioning system, and high passenger volume at morning or evening peak hours exert profound effects on virus transmission and infection probability.

Keywords

• computational fluid dynamics
• droplet aerosols
• trajectory simulation
• virus aerosol transmission
• infection probability simulation