Introduction: During the second wave of the SARS-CoV-2 pandemic, the numbers of patients admitted into hospital rose steadily across the UK from a weekly average of 122 (September 1st 2020) up to 2,037 (17th December) eventually reaching a peak of 4,232 cases per week on 9th January 2021. During this time, the NHS continued to deliver with other non-COVID admissions and elective surgery.
The increase in hospital admissions led our ITU teams to consider expanding to ward areas to manage SARS-CoV-2 critically ill patients. They requested an urgent review of the feasibility, efficacy and safety of UVC air sterilisation devices to reduce the length of the fallow period between patient procedures required for the room ventilation system to clear air potentially contaminated by viral-loaded aerosol generated particles (AGPs).
Computational Fluid Dynamics (CFD) offers insights into ventilation efficiency and contamination spread not previously available in the original design of the existing treatment spaces.
Objective: to review existing data on UVC air sterilisation
Main Questions
- What effect does UVC radiation (250 nm) have on bacteria and viruses particularly SARS-CoV-2
- How could we predict UVC efficacy in a real world setting through the use of computational modelling and/or environmental air sampling?
Assessment: Quantifying the rate of far-UVC viral inactivation within a general room is complex and multi-physics in nature. The judicious application of CFD modelling to the understanding of the complex anisotropic air flows associated with fixed and moving objects (including humans) can suggest effective optimisations of ACH strategies achieved by combining air scrubbing with pre-existing HVAC’s [7].
Our Computational Fluid Dynamics (CFD) models provide information on the dispersion of airborne particles in healthcare environments where AGPs present a significant risk. CFD is a useful tool to understand the dynamics of infectious particles through the air. It has been used successfully to study the effect of different ventilation regimes, and layouts within clinical areas [8,9].
A CFD expert group was established to model the flow dynamics within a dental treatment room in Birmingham Women’s and Children NHS Foundation Trust. (Prof Tony Fisher, Royal Liverpool University Hospital, Prof Paul A White, Cambridge University Hospitals NHS Foundation Trust, Fred Mendonça and Pawan Ghildiyal, Open CFD Ltd, Peter Bill, Birmingham Women’s and Children NHS Foundation Trust, Claire Greaves, Nottingham University Hospitals and Prof Chris Hopkins, Hywel Dda). This high-risk AGP environment was taken as a surrogate of the endoscopy suite which is the focus of this study (vide supra).
Computational Fluid Dynamics (CFD) modelling: CFD in the Engineering Sciences of buildings is well-established [10]. Solving the Navier-Stokes equations governing continuum air movement, combined with the discrete particulate transport, gives us a compete view of complex ensemble of turbulence, buoyancy, aerosol dispersion, evaporation and wall interaction across the full range of particle sizes of interest, notionally [0.1 .. 100µm3].
In this study of indoor ventilation, sponsored by UK Research and Innovation [11], using fully ISO9001:2015 QA’ed open-source CFD [12], the efficacy of several ventilation strategies was assessed. The strategies include mechanical (controlled by the building air-management system), natural (opening window), augmented ventilation (from UV air-cleaning devices) and several roof-diffuser vent designs.
CFD tracks the age of the air (AoA) from a fresh source to anywhere within the room volume, therefore identifying which parts are well ventilated, and conversely, locations of dead air or recirculating bubbles. Mean-AoA then becomes a meaningful measure of clean air circulation in the enclosure.
A CFD model of Birmingham Children’s Hospital’s 44.7m3 fully equipped dental treatment room comprises three occupants; dentist, patient and nurse. Roof ventilation slightly offset from directly above the treatment chair was supplied in the centre of the room at 5 ACH with balanced extraction to one side.
This was carried out by installing a mobile UVC unit into the room which provided 7 ACH of recirculated air leading to a total ‘equivalent’ of 12 ACH, i.e. 5 ACH of outdoor air supplied by the mechanical HVAC system and 7 ACH of recirculated and clean/sterilised air supplied by the UV mobile unit.
When the UVC steriliser was introduced into the room, there was a reduction in fallow time by up to 75%. On standard settings (180m3.h-1), the age of air within the room reduced to 6 minutes and on the purge or boost (360m3.h-1) setting, this was further reduced to 4 minutes.
Similar studies in well ventilated rooms of different sizes and ventilation strategies suggests a close link between the ventilation rate (ACH) and mean-AoA as an absolute measure in minutes. Positioning of UVCs or direction of airflow gives only +/-15% variation in the clean air mixing.
Conclusion: CFD modelling suggests a significant improvement in ‘clean air’ after introduction of the UVC sterilisers and early work with the PCR detection method for SARS-CoV-2 seems to confirm this. The improvement is seen in both age of air and ‘air mixing’ when placed in an optimum position, with no significant issues raised from real-world testing. There is ongoing work to develop guidelines for using portable air cleaning devices in a range of clinical settings.