Modelling the transmission of COVID-19 in indoor spaces

Thursday, June 17 at 02:15am (PDT)
Thursday, June 17 at 10:15am (BST)
Thursday, June 17 06:15pm (KST)

SMB2021 SMB2021 Follow Wednesday (Thursday) during the "MS17" time block.
Share this


Raquel González Fariña (Cardiff University, United Kingdom), Katerina Kaouri (Cardiff University, United Kingdom)


It will take time to achieve herd immunity against SARS-CoV-2 through vaccination. In the meantime, we need to find ways to safely lift restrictions and resume economic and social activities while containing the virus. Therefore, indoors transmission of COVID-19 has been intensively researched since the beginning of the pandemic. There is an emphasis on quantifying viral transmission in areas with a high density of people. These high-risk spaces include healthcare clinics, schools, public transport, nursing homes and supermarkets. In this mini-symposium, we present several mathematical approaches for modelling the dynamics of transmission of COVID-19 in diverse indoor locations. All three main routes of transmission – airborne (through droplets and aerosols) and through surfaces (fomite) will be discussed and the effects of ventilation, antiviral technologies, and other non-pharmaceutical interventions will be presented. Deterministic and stochastic mathematical models are considered and are solved using a combination of analytical and numerical techniques. In many cases, Computational Fluid Dynamics simulations are employed to model air flows and transport of SARS-CoV-2 in detail. Many of these results have been used to inform policies related to the COVID-19 pandemic in the UK and elsewhere and the speakers comprise an interdisciplinary group from academic and government organisations.

Christian Kähler

(Universität der Bundeswehr München, Germany)
"From droplets to pandemic – how to prevent SARS-CoV-2 infections via droplets and aerosols"
The SARS-CoV-2 pandemic is currently presenting humanity with major challenges. Containing the spread of the virus requires enormous financial, technical and social efforts, and it is impossible to predict how well humanity will cope with the problem. Since the infectious disease not only has an acute course, but can also cause long-lasting systemic damage to infected individuals, prevention of infection is most important. It is generally accepted that the transmission of viruses is largely via droplets and aerosol particles. Therefore, the question of how these aerosol particles are generated and released and how they spread through the room and cause infection is particularly important to answer. Next, there is the question of how to best protect against infection. The answer to this question depends on the areas for which protection is to be established, because different protective measures have to be taken in a pedestrian zone than in buses and trains or in offices, schools and restaurants. To address these two problems, the first part of the talk will present the formation of aerosol particles in the body, their ejection by breathing, speaking, singing and coughing, and their dispersion in space. In the second part, the effectiveness of different protective measures is analyzed experimentally using laser based measurement data. In particular, the effectiveness of different masks for individual protection, as well as the usefulness of room air cleaners and protective walls, is demonstrated quantitatively. A deeper understanding of the spread processes and the protection options is imperative to effectively limit the spread of the pandemic and thus the costs for the state, the economy and society. Whether society is finally ready to protect itself effectively depends on the insight of the population, but also on the way the measures are implemented politically. This will also be discussed during the lecture, because this pandemic can only be contained if science, technology, politics and the population pull together.

Chenfeng Li

(Swansea University, United Kingdom)
"CFD simulation of airborne virus transmission aided by a machine learning surrogate model"
In this study, the airborne virus transmission is investigated using computational fluid dynamics simulation. The study is carried in three steps. First, a standard boxroom scenario is considered, and different conditions in relation to the door, window and mechanical ventilation are investigated using OpenFOAM, a well-established CFD simulator. The resulting data are organized a series of relation curves to reveal the sensitivity of virus transmission with respect to the change of ventilation conditions. Next, a machine-learning based surrogate model is constructed from the simulation data obtained from the first step. The experiment shows at an acceptable level of accuracy, the surrogate model can quickly predict the flow field and the associated airborne virus transmission for the boxroom scenario at different environmental and ventilation conditions. In the last step, the study focuses on the impact of having people in the room. To achieve this, a coupled CFD-DEM approach is adopted, where the air flow is captured by the CFD solver, and the moving objects are captured by the DEM (discrete element method) solver. The two solvers are fully coupled, representing accurately the influence of people on the air flow, thereby the airborne virus transmission. In all these studies, we assume the virus particles are sufficiently small, such that they do not have a significant impact on the air flow and merely get transported by the air. The information obtained this investigation quantify the relative risks of virus transmission with respect to changing environmental and ventilation conditions, as well as the impact from human activities.

Simon Parker

(Defence Science and Technology Laboratory, United Kingdom)
"Transmission of virus in carriages - development of a multiroute viral exposure model for public transport"
Understanding the potential exposure of passengers during the use of public transport is of particular interest for transport planning. The Transmission of Virus on Carriages model has been developed to address this need. This stochastic model estimates passenger exposure to SARS-CoV-2 on an underground train carriage via three routes: close range droplet, small aerosol and surface contact. Passenger input data reflects realistic ridership at different stations on the journey. In addition to the effect of ventilation details, the effect of airborne and surface viral decay and surface deposition are accounted for. Touching of high-touch surfaces within the carriage is included and is used to study the transfer of fomite contamination between surfaces and passengers. Variation in passenger density throughout a journey contributes to different close range exposures. Stochastic representation of infectious passenger boarding and other events allow the model to be used to explore key parameters such as disease prevalence in the travelling population, carriage loading and adherence to mask wearing rules. The results from the model clearly show the relative importance of different routes of exposure and the value of different mitigation measures. Future work is focused on improving the representation of multiple respiratory activities and on extending the model to other types of public transport.

Raquel González Fariña

(Cardiff University, United Kingdom)
"Predicting the spatiotemporal risk of airborne infection in indoor spaces using an advection-diffusion-reaction equation"
Probabilistic models for indoors airborne transmission of viruses, such as the Wells-Riley model (Riley et al., 1978) and its extensions applied to COVID-19, for example (Buonanno et al., 2020), assume that the concentration of infectious particles in the room is uniform in space. We have developed a spatially dependent generalisation to such models to determine the infection risk in indoor spaces. We model the concentration of airborne infectious particles using an advection-diffusion-reaction equation where the particles are emitted by an infected person, advected by airflow, diffused due to turbulence, and removed due to the room ventilation, biological inactivation of the virus and gravitational settling. The model is quasi-three-dimensional and incorporates a recirculating flow due to air-conditioning. We are able to obtain a semi-analytic solution which allows for very fast simulations. An important aspect of our work is that we include realistic particle size distributions. We consider that the particle emission rate and the gravitational settling rate depend on the size of the particles. Most airborne transmission models only account for a single particle size and, thus, assume a constant settling rate. We find that this simplifying assumption may significantly alter the predicted infection risk in a room. We also quantify the effect of several ventilation settings and different activities such as breathing, talking and coughing, on the particle concentration and the infection risk. Finally, we determine the time to probable infection (TTPI), at any location in a room, paving the way for formulating recommendations. Good agreement with CFD models and existing data is obtained.

Hosted by SMB2021 Follow
Virtual conference of the Society for Mathematical Biology, 2021.