Development of Computational Models to Design Upper-Room Ultraviolet Germicidal Irradiation Air Disinfection Systems in Hospital Environments

Lead Research Organisation: University of Leeds
Department Name: Civil Engineering


Ultraviolet Germicidal Irradiation (UVGI) irradiation has been known for many years to have a lethal effect on microorganisms. It is routinely used in water disinfection and can be a viable method of reducing airborne pathogens in indoor environments to decrease the risk of cross transmission of infection. Despite recommendations for use in high risk healthcare environments such as tuberculosis (TB) wards, the application of UVGI remains limited due to lack of evidence gained directly in clinical settings. However two major clinical studies that are about to be published will prove the effectiveness of UVGI devices against TB transmission and thus a significant increase in the future demand for UVGI air disinfection is anticipated. To translate these clinical based research findings into successful control strategies tailored to the needs of a particular healthcare environment, robust guidance on designing safe and effective UVGI air disinfection systems is now needed. This project aims to address this need by developing a design tool for simulating the behaviour of upper room air disinfection devices in realistic hospital environments, and guidance documents to enable hospital managers, architects and engineers to (a) determine if UVGI disinfection is suitable for a particular environment and (b) to ensure any UVGI installations are both effective and safe. The proposed study will use computational fluid dynamics (CFD) simulations to carry out a parametric study quantifying the factors that influence the performance of a UV device to produce an empirical model of UVGI disinfection within a ventilation design model. The model will quantify the mean effectiveness, as well as stochastic variations, and provide an output in terms of UV device performance and relative risk of infection for ward occupants. This will enable better design and specification of UVGI systems without the use of resource intensive CFD models. Parametric studies using CFD models and the new design tool will then be used to draft three design guidance documents; suitability of upper-room UVGI systems in healthcare environments, safe installation and operation of UVGI systems, and optimising UVGI system design to minimise airborne infection risk.


10 25 50
Description Research focused on development and application of computational models to simulate performance of upper-room UV air disinfection devices in hospital environments. The study included CFD simulation of airflow in naturally ventilated wards, CFD simulation and experimental assessment of UV systems and development of zonal models to assess performance at ward and room level.

The study showed that simulating UV dose in a CFD model provides valuable insight into interaction between UV devices and airflow patterns. The most significant variable was specification of the UV field. Deriving a field from experimental measurement is the most feasible method. A 2-D field definition was adequate for spatial assessment of dose but a 1-D field resulted in over-prediction. Assessment of surface reflections demonstrated no impact on disinfection performance for all but the smallest rooms. Simulation of inactivation was shown to yield realistic results compared to preliminary bioaerosol chamber experiments, however further experimental data is needed before this can be used with confidence.

Simulation of real scenarios focused on large naturally ventilated multi-bed wards as these spaces are hard to characterise and may present a high transmission risk due to multi-occupancy design. Experimental study (conducted with EP/G029768) showed cross-ventilation provides high dilution, but infection risk was substantially higher when windows were closed. CFD simulations were able to capture the contaminant dispersion seen in experiments and showed airpaths were predominantly determined by window angle and ventilation rate. UV lamp placement was not intuitive, with lamps located on leeward side of the ward leading to higher doses in the ward. Results demonstrate critical need to consider airflow and UV field interaction. Combined experimental and simulation study of enclosed UV devices showed CFD models compare well to bioaerosol experiments and device effectiveness depends on positioning, room size, occupancy and ventilation.

Multi-zonal infection risk models combined with 2-zone UV models enabled exploration of the influence of a UV installation in single and multi-room environments. Models indicated that where the infectious source is known, device placement close to the source is most effective solution. However, where sources are not known and resources are limited, installation in communal areas is more effective at reducing total infection risk than installation in individual ward spaces. Zonal models developed to approximate UV dose compare well to CFD simulations and experimental data published in the US. Paper detailing these outputs presented at ASHRAE IAQ 2013 conference and subsequently invited for publication in Science and Technology for Built Environment Journal (ASHRAE published 2015).

Invited paper published in Photobiology and Photochemistry details CFD modelling approach and key considerations. A more accessible paper summarising UVGI was published in European Medical Hygiene in 2013. Invited presentation at ASHRAE IAQ 2010 conference led to invitation to join an ISO standards committee to develop international guidance on in-duct UV systems. Collaboration with Dr Escombe to install UV lamps in a London TB Clinic and invited paper at the Fogarty UVGI conference at Harvard University led to international networking and engagement with plans to develop international guidance involving researchers, clinicians, policy makers and device manufacturers worldwide.
Exploitation Route Research has non-academic applications both in system design/assessment and in the development of future guidance.

In terms of design, dose and risk models offer a straightforward way of exploring the benefits of upper-room UV installation and the numbers and room level locations of units needed to achieve an appropriate level of protection. These models are spreadsheet based and are feasible for both industry and clinical teams to use. While the models currently have a field data for a limited number of UV fixtures within them, they can be used to input generic UV field intensities to explore typical performance within a ward space. CFD modelling approaches, while not appropriate for most clinical teams, are potentially feasible for manufacturers/architectural engineering organisations to implement where there is a need to understand airflow/UV device interaction.

In the case of developing guidance, it was concluded at an early stage in the project that developing standalone guidance would have little impact. It was for this reason that the focus of guidance development (objective 5) shifted to modelling specific outputs and longer term strategies to engage with international researchers to develop guidance with impact. Project outcomes led to consensus at the Fogarty UVGI conference that modelling UV field and dose distributions using CFD simulations is an appropriate method for quantifying UV performance. Ongoing work that developed directly from this conference is exploring application of the CFD and dose models to determine appropriate guidance values for UV output and hence the number of fixtures recommended in a room. Current guidance (NIOSH, 2010) specifies an average intensity of 0.3 to 0.5 W/m2 in a plane through the centre of UV devices, based on a single set of experimental data. CFD models can be used to explore how the airflow affects this specification, whether corrections should be made in certain environments and whether the current parameter specification is appropriate for all devices and rooms.
CFD models: Research demonstrates feasibility of using CFD simulation in the design and assessment of upper-room UVGI installations and sets out the important parameters for consideration in the model. These models can be used in system design to optimise lamp placement and UV dose distribution as well as to conduct performance/risk assessment for different pathogens. The models have already been exploited to aid the design and installation of the first UK upper-room UVGI installation at a TB clinic in London, in collaboration with Dr R Escombe, Imperial College. Field models were used to position lamps to ensure the best coverage and consider the likely airflow paths. Further ongoing exploitation providing insight for experimental studies. Full-scale bioaerosol room experiments are currently underway - comparison with CFD simulations enables understanding of mechanisms and validation of inactivation models. They also potentially enable quantification of the DNA repair seen in multi-pass upper room systems compared to single-pass testing .

Ventilation: Simulation of naturally ventilated ward spaces has already informed experimental studies in Noakes Challenging Engineering grant (EP/G029768). CFD methods are now being used within DeDeRHECC project (EP/G061327) to explore redesign of large open wards in the UK NHS. Future exploitation could help optimise layout and ventilation of naturally ventilated ward spaces, particularly in developing countries where airborne disease risk (eg tuberculosis) is much greater. Further ongoing work in Noakes Challenging Engineering grant is exploring how optimisation techniques that have been successfully applied to ventilation flows can be used in conjunction with CFD models to numerically optimise placement of UV lamps in a space. This is particularly beneficial for large spaces with multiple UV fixtures.

Dose/Inactivation models: Research builds on a growing body of work modelling airborne infection risk in the built environment led by our group and key researchers in Hong Kong, Taiwan and the US. Models developed here demonstrate that air disinfection devices can be built in to such models. Future development is likely to include coupling with more accurate disease data as well as applying cost benefit approaches developed in project EP/G029768.
Sectors Construction,Environment

Description Authored section on UVGI disinfection for new edition of CIBSE guide A (Chapter 8) published in 2015. This sets out improved guidance for UK practitioners on application of UVGI disinfection in buildings.
First Year Of Impact 2015
Sector Construction,Energy,Environment
Impact Types Policy & public services

Description CIBSE Guide A chapter
Geographic Reach National 
Policy Influence Type Citation in other policy documents
Description PhD studentship in UVGI
Amount £90,000 (GBP)
Organisation National Council on Science and Technology (CONACYT) 
Sector Public
Country Mexico, United Mexican States
Start 02/2011 
End 02/2014
Title CFD simulation of UV performance 
Description Computational Fluid Dynamics model to simulate dose and microorganism inactivation due to upper room UV devices. 
Type Of Material Computer model/algorithm 
Year Produced 2012 
Provided To Others? No  
Title Multi-zone UV model 
Description Spreadsheet based model to simulate UV dose and inactivation in single room or series of connected rooms with air exchange between them. 
Type Of Material Computer model/algorithm 
Year Produced 2012 
Provided To Others? No  
Description Collaboration with Dr R Escombe 
Organisation Imperial College London
Country United Kingdom of Great Britain & Northern Ireland (UK) 
Sector Academic/University 
PI Contribution Involvement in ongoing project led by Dr R Escombe, Imperial College to install upper-room UVGI lamps in a London TB Clinic. Contribution included airflow assessement, modelling lamp fields and potential benefits considering likely airflow and UV field monitoring.
Start Year 2009
Description SET for Britain poster 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? Yes
Geographic Reach National
Primary Audience Policymakers/parliamentarians
Results and Impact Poster on "The Influence of Infection Control on the Design of Hospital Wards" presented at House of Commons. Poster in top 5 out of more than 60.

Year(s) Of Engagement Activity 2011
Description TechWorld 2011 poster 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Poster presentation - University Excellence Competition, TechWorld 2011 (United Kingdom Trade and Investment), Excel Centre " Approaches to Evaluating Airborne Infection Risk in Large Naturally Ventilated Hospital Wards".

Year(s) Of Engagement Activity 2011