It is well known that in research the most effective outcomes and solutions often result from asking the right questions. But what if the question comes from an unexpected place? In February 2020, faced with a myriad of government guidelines on the safe use of indoor spaces during the emerging coronavirus pandemic, the Bursar turned to the Fitz Engineers for some answers. The resulting project led to both a paper published by the Royal Society, and a free online tool to safely assess the aerosol distribution of COVID-19 in an indoor space. Here we learn more about this unusual collaboration, the researchers and staff involved, and the impact the findings have had for the College and the wider community.
Fitz Fellow, Epaminondas Mastorakos, is a Professor in the Department of Engineering, and he is a specialist in applied thermodynamics. As he explained, “Sometimes in science you start with a solution and are looking for a problem, but this time the problem, the practical question, was the starting point. The discussion in College began before the start of the first UK lockdown, in January and February, when advice was beginning to circulate about the need for face-coverings, physical-distancing and caution around the use of indoor spaces.”
During what would be the last ‘in person’ College Committee, the Bursar, Andrew Powell, outlined the emerging government advice for workplaces, with its clear insistence on the use of face-coverings and 2-metre distancing. While these restrictions are now grimly familiar, in February 2020, there was significant concern both that measures might be unnecessarily restrictive or, perhaps more worrying, insufficient to prevent the spread of infection. As is often the case with a diverse group of thinkers, discussion started around how solid the science was, how precise the Government’s understanding of risk could be and, most practically, would we be able to meet in Wilson Court in future and would a face covering make us any safer?
For the Bursar, these discussions must have been initially frustrating. The speed with which the College was compelled to make practical decisions and the often-overwhelming flow of information meant that academic debate felt like a luxury we could ill-afford. But in a demonstration of the resourcefulness of the College community, we did not simply debate: the Bursar threw down the gauntlet to the engineers present, and challenged them to model the risks of aerosol transmission and return with some advice. It was agreed, as Professor Mastorakos explained succinctly, “This was not a medical problem, it was a question of fluid mechanics”.
The science behind the question of indoor ventilation brought together two areas of expertise: the study of jet engines, more specifically how a combustor mixes liquid fuel and air to make vapour; and the study of thermal fluids in understanding the distribution of this vapour within an enclosed space. It was the combination of this knowledge, along with a solid understanding of stochasticity, which generated the possibility of predicting relative risk and the development of the online calculator. Professor Masorakos explains:
“Stochasticity is important: with two people in a room, at 2m distance, a few times out of ten a cough might not transmit sufficient aerosol, but a few times it will. This variability is ‘bread and butter’ in our work on turbulent fluid mechanics. It isn’t a special quality of the coronavirus, or any virus, it is about droplets moving through the air, over time.”
To build the model, Professor Mastorakos’ team (Leo Mesquita, one of our Fitz post-doc Research Associates, his post-doc Pedro de Oliveira from Magdalene, his PhD student Savvas Gkantonas from Churchill, and another one of our own Fitz engineering supervisors and Bye-Fellow, who is now at Imperial, Dr Andrea Giusti) took measurements from a number of rooms around the College and the Department of Engineering. The practical understanding of the systems was crucial, and with the support of the Fitz Maintenance Manager, Julian Eddy, the ventilation systems in the Auditorium and Upper Hall were reviewed and their impact upon aerosol transmission assessed.
“We looked at all sides of aerosol and droplet transmission to understand, for example, the fluid mechanics involved in coughing and speaking,” said Professor Mastorakos. “The role of turbulence and how it affects which droplets settle by gravity and which remain afloat in the air is, in particular, not well understood. We hope these and other new results will be implemented as safety factors in the app as we continue to investigate.”
The team’s findings were published in January 2021, in the Proceedings of the Royal Society A and were covered extensively by the media. The implications of the results were significant for health and safety advice: standing 2m opposite an infected speaker is not safe without the use of a protective mask; an unprotected cough is not safe at any reasonable distance; the settling of the aerosol by gravity was found to be small compared to the viral decay; the presence of air currents strongly affects the suspended viral dose; and an infected person speaking for one hour in a room may lead to infection risk levels of 10-20% with inadequate ventilation, but the risk can be reduced by at least a factor of three, if ten air changes per hour are employed through ventilation.
Based on their models, the researchers have also built Airborne.cam, a free, open-source tool which can be used by those managing indoor public spaces, such as shops, workplaces and classrooms, in order to determine whether ventilation is adequate. The tool is already in use in several academic departments at the University of Cambridge. The tool is now a requirement for any higher-risk spaces at the University, enabling departments to easily identify hazards and control-measure changes needed to ensure aerosols are not allowed to become a risk to health.
For Fitz, the research has been an important factor in discussions of health and safety throughout the pandemic. In September 2020, with the return of the student population, the ongoing work gave an important perspective on the safety of teaching supervisions in-person, and the adaptations that were necessary to continue the practice such as room size, face-coverings and adequate ventilation. In October, it helped us to decide in which rooms, and with what capacity we might meet. And now, as we plan for the next emergence from lockdown, it will once again help the College – and all of its members – make informed choices about the facilities we can use, and how.
'Evolution of spray and aerosol from respiratory releases: theoretical estimates for insight on viral transmission', Proceedings of the Royal Society A, P. M. de Oliveira, L. C. C. Mesquita, S. Gkantonas, A. Giusti and E. Mastorakos. Published:20 January 2021https://doi.org/10.1098/rspa.2020.0584