Evidence on the effectiveness and safety of ultraviolet germicidal irradiation technologies in reducing SARS-CoV-2 in the air of occupied rooms

Buchan (2020)^{Footnote 5}

Simulation study

UK

Nov 2020

Researchers simulated the use of a far UV-C lamp (new technology that emits a narrow bandwidth of 207-222 nm, which are safe for humans) located in the upper corner of a 3 meter by 3 meter air conditioned room projecting down into the room (whole room UVGI), occupied by a single person. There were two vents located in the upper corners of the room, and tests were conducted at two different velocities (0.1 ^ms-1 / 8.0 air changes per hour (ACH) and 0.01 ^ms-1 / 0.8 ACH). This was to determine the efficacy of far UV-C in inactivating SARS-CoV-2 when different velocities of ventilation were used alone, or in combination with far UV-C.

To represent far UV-C inactivation values of SARS-CoV-2, the inactivation value of other human coronaviruses was used. The viral load of SARS-CoV-2 was released into the room using two second pulses and two second pauses to represent breathing.

• When the far UV-C lamp was used with high ventilation, the SARS-CoV-2 viral count was reduced by 90% in six minutes and 99% in 11.5 minutes. This viral count reduction was performed in less than half the time it took for high ventilation of 8.0 ACH alone to reduce viral count.
• The use of a far UV-C lamp in combination with ACH ventilation at 0.8 and 8.0 velocities resulted in quicker SARS-CoV-2 inactivation at all distances, compared to using 0.8 or 8.0 ACH ventilation alone.
• When the viral load of SARS-CoV-2 was released using two second pulses and two second pauses to represent breathing, the use of the far UV-C lamp in combination with 0.8 or 8.0 ACH ventilation, resulted in a ~20% or 57% further reduction in viral concentration, respectively, compared to 0.8 or 8.0 ACH ventilation used alone.

Volchenkov (2021)^{Footnote 6}

Field investigation

Russia

2003-2020

Researchers examined the effectiveness and safety of upper and whole room UVGI in reducing SARS-CoV-2 and TB transmission among employees and patients within a hospital building.

The UVGI source consisted of 240 wall-mounted UV-C fixtures (one fixture per 18 m²). Each UV-C fixture contained two low-pressure mercury lamps (T8 30 W, wavelength = 254 nm) for upper and whole-room UVGI, respectively.

The upper-room UVGI lamp was used 24 hours per day, 7 days per week, when people were present and absent from the hospital rooms, while the whole-room UVGI lamp was only used when people were not present in the room.

• There was no SARS-CoV-2 transmission reported among TB and HIV patients located in hospital rooms with the UV-C fixtures, while community acquired COVID-19 cases were reported among staff from April-June 2020.
• Safety results found in Table 5.

Li (2021) ^{Footnote 10}

Simulation study

China

Jul 2021

This study aimed to evaluate the SARS-CoV-2 infection risk in different indoor locations at a university, and the efficacy of engineering control measures (including upper room UVGI) in different exposure scenarios. The Wells-Riley equation was used to model SARS-CoV-2 infection risk. The model assumed that the inactivation rate for upper room UVGI was 12± 1.3 h^-1 (based on prior research using mycobacteria). Wavelength and power of the UVGI was not specified. General ventilation looked at increasing air change rates from 0.5 to 4 per hour. Masks included a range of risk reduction estimates from surgical, dental, homemade and N95s.

The five exposure scenarios included: sleeping or talking in a dormitory, studying or talking in a classroom, playing basketball in a gym, studying or whispering in a library, and eating in a dining hall.

For the scenarios in a classroom, gym, library, and dining hall:

• The average SARS-CoV-2 infection risk was lower when UVGI was used in addition to masking and general ventilation, compared to only masking and general ventilation.
• The infection risk was approximately the same when general ventilation was used with HEPA vs. with UVGI.
• The lowest infection risk was found when a combination of general ventilation, masking, UVGI, and HEPA was used.

For the scenario in a classroom:

• The SARS-CoV-2 infection risk was 35% with general ventilation and masking vs. 28% with general ventilation and UVGI (exposure time was 24 hours).
• Note: The model did not take into account the type of UVGI technology or its installation location.

Hill (2021)^{Footnote 14}

Simulation study

USA

Jul 2021

This study aimed to examine the optimization of a UVGI disinfection system on the survival fraction of SARS-CoV-2 virions that are within host particles (shielding them from UVGI) in the air or on surfaces.

For the purposes of this review, only the upper room UGVI applications were considered.

In the simulations, virions in a group of particles were exposed to UV light and the average survival fraction of the virions was calculated under varying conditions regarding the number of light beams and the distance between the angles of light. UV wavelengths of 260 nm and 302 nm were studied.

• When exposed to UVGI, the survival fraction of SARS-CoV-2 virions in host particles increased as the particle size increased.
• Increasing the number of UV light beams hitting the particle resulted in a lower virion survival fraction, even though the total UV energy emitted remained the same.
• Increasing the separation between the angles of UV light beams hitting the particle resulted in a lower virion survival fraction (compared to light beams coming from a similar direction).

Swanson (2021) ^{Footnote 9} preprint

Simulation study

USA

Apr-May 2021

This simulation characterized the probabilities of SARS-CoV-2 infection, hospitalization, and death associated with aerosol exposure from in-person classes and the impacts of masking and UV-C ceiling fans. The UV-C ceiling fans have disinfecting ultraviolet lights built into the base of the fan that are aimed up at the ceiling, thus an upper room UVGI application.

A semester of courses in a real college with approximately 11,000 students embedded within a larger university was modelled. The schedule input for the model included 11,968 students and 342 faculty in 1,025 courses. Immunity rates from 60-95% were used in the simulation to determine the impacts of masking and UV-C fan ceiling interventions.

• Upper room UV-C ceiling fans in every classroom reduces the risk of infection (>40%) more than universal masking alone.
• Compared to no intervention, a combination of masking and UV-C fans show the greatest reduction SARS-CoV-2 infection risk.
• Under a low SARS-CoV-2 transmissibility scenario with 60% immunity and using UV-C ceiling fans, the probability of exceeding 50, 100, 250, and 500 student infections was >0.999, 0.997, <0.001, and <0.001 respectively. The probability of exceeding 1, 2, 10, and 20 faculty infections was 0.936, 0.156, 0.002, and <0.001, respectively. At 90% immunity probabilities drop to <0.001 for the above thresholds in students and staff.
• Under a high SARS-CoV-2 transmissibility scenario with 60% immunity and using UV-C ceiling fans, the probably of exceeding 50, 100, 250, and 500 student and 1, 2, 10, and 20 faculty infections was >0.999, and at 90% immunity was 0.814, 0.034, <0.001, and <0.001 for students and 0.652, 0.008, 0.002, and <0.001 for staff, respectively. Adding masking decreased the probability of exceeding 500 student and 20 faculty infections at 60% immunity to 0.554 and 0.005, respectively.
• Scenarios for 70%, 80%, and 95% immunity were also provided.
• Similar trends were shown for hospitalizations and death.

D'Alessandro (2021) ^{Footnote 8}

Simulation study

Italy

Mar 2021

An Eulerian–Lagrangian model was developed to examine the effect of UV-C irradiation on inactivation of airborne virus/bacteria particles in a cloud of saliva droplets. Clouds produced from one, two, and three cough ejections were modelled. The UV-C source was a lamp at a wavelength of 254 nm, with a power of 25 watts (W) or 55 W.

In the model, the radiation dose sufficient to inactivate SARS-CoV-2 was used as the "susceptibility constant" for the virus/bacteria (8.5281 x 10^-2 m²/J).

• UV-C irradiation was shown to effectively inactivate the majority of SARS-CoV-2 particles in a cloud of saliva droplets after 4 seconds.
• The UV-C lamp with a power of 55 W was more effective at inactivating SARS-CoV-2 over a period of 10 seconds compared to 25 W.
• Note: visualizations were provided in the study that show SARS-CoV-2 inactivation in a cloud of droplets produced from one, two, and three cough ejections.

Hou (2021) ^{Footnote 12}

Simulation study

USA

Mar 2021

Researchers used ray-tracing to simulate the impact of different room design parameters on the safety and effectiveness of UV-C irradiation of SARS-CoV-2. The simulation involved the use of an occupied test room (24 x 30 feet, floor area = 15.80 m²), which had one wall-mounted UVGI lamp (254 nm, 25.47 W).

Three room configurations were examined: configuration 1 (square, 3.97 m length x 3.97 m width, UVGI lamp located on one wall), configuration 2 (rectangle, 4.57 m x 3.44 m, UVGI lamp located on long wall), configuration 3 (rectangle, 4.57 m x 3.44 m, UVGI lamp located on short wall). Four ceiling height / UVGI device mounting heights (C/M height) were examined: C/M height 1 (2.44 m / 2.13 m), C/M height 2 (2.74 m / 2.13 m), C/M height 3 (2.74 m / 2.29 m), and C/M height 4 (3.05 m / 2.44 m).

A simulated case study involving an occupied hospital room (7.32 m length x 4.57 m width, default ceiling height = 2.74 m, floor area = 27.87 m²) with two upper zone UVGI lamps (Atlantic Ultraviolet Corporation Hygeaire model LIND24-EVO, lamp power = 25.6 W each, 254 nm) was also performed to determine the impact of different room design parameters on the safety and effectiveness of UV-C irradiation of SARS-CoV-2.

Four scenarios for UVGI fixture location were examined: scenario 1 (UVGI fixtures located above the bed), scenario 2 (UVGI fixtures located opposite of the bed), scenario 3 (UVGI fixtures located on the left and right side walls from the bed), and scenario 4 (UVGI fixtures located above and beside the bed). Three patient room layouts were examined: layout 1 (L-shaped, 7.32 m length x 4.57 m width, default), layout 2 (rectangular, 6.01 m x 4.57 m), layout 3 (rectangular, 7.01 m x 3.96 m). Three room surface UV-C reflectance coefficients were examined (0.05 (default), 0.1, and 0.2). Three scenarios for ceiling height / UVGI fixture mounting height were examined: height 1 (2.74 m / 2.13 m), height 2 (2.74 m / 2.29 m; default), and height 3 (3.05 m / 2.44 m).

SARS-CoV-2 disinfection effectiveness was measured using average fluence rate (µW/cm²).

Simulation study:

• Configuration 3 resulted in the highest SARS-CoV-2 disinfection effectiveness of all three room configurations (average fluence rate = 41.94 µW/cm²). However, all three room configurations resulted in the effective deactivation of SARS-CoV-2 within 19 seconds, at a fluence rate less than 48 µW/cm².
• C/M height 1 has the most effective SARS-CoV-2 upper zone disinfection effectiveness (average fluence rate = 56.56 µW/cm²), while all other C/M heights had an average fluence rate of less than 48 µW/cm², which is the threshold for average fluence rate.
• Accounting for both SARS-CoV-2 disinfection effectiveness and safety, C/M height 3 is the optimal option.

Case simulation study:

• Scenarios 1 and 4 had the most effective SARS-CoV-2 upper zone disinfection coverages of 18.85% (average upper zone fluence rate = 48.19 µW/cm²) and 14.83% (average upper zone fluence rate = 48.61 µW/cm²), respectively.
• All three room layouts resulted in effective SARS-CoV-2 disinfection (maximum irradiance at the 1.83 m to 1.98 m range: layout 1 = 0.31 µW/cm² for a maximum of 5.2 hours; layout 2 = 0.28 µW/cm² for a maximum of 5.7 hours; layout 3 = 0.33 µW/cm² for a maximum of 4.9 hours).
• Safety results found in Table 5.

Beggs (2020) ^{Footnote 7}

Simulation study

UK

Oct 2020

In this study, researchers simulated the best and worst case scenarios for using upper-room UVGI (~254 nm) to determine its efficacy in decreasing SARS-CoV-2 transmission in occupied buildings.

The upper room UV-C susceptibility constant for SARS-CoV-2 was assumed to be 0.377 m²/J (best case) and 0.0377 m²/J (worst case), and the amount of UV irradiation (UV flux) required to inactivate 50%, 70%, and 90% of the SARS-CoV-2 virus in a 1 to 8 ACH ventilated room (dimensions = 4.2 m x 4.2 m x 2.5 m) with an upper room UVGI lamp (height = 2.1m above floor) was determined. The UV-C lamp (30 W) used was assumed to have an average upper-room flux of 50 µW/cm².

• Best-case scenario: When the UV-C susceptibility constant for SARS-CoV-2 is 0.377 m²/J, at the highest ventilation rate of 8 ACH, the average irradiation needed for 50%, 70%, and 90% SARS-CoV-2 inactivation is 2.6 µW/cm², 4.4 µW/cm², and 8.5 µW/cm², respectively.
• Worst-case scenario: When the UV-C susceptibility constant for SARS-CoV-2 is 0.0377 m²/J, at the highest ventilation rate of 8 ACH, the average irradiance needed for 50%, 70%, and 90% SARS-CoV-2 inactivation is 25.5 µW/cm², 44.4 µW/cm², and 84.8 µW/cm², respectively, which is a ~10 factor increase compared to the best-case scenario.
• Even in the worst-case scenario (0.0377 m²/J), SARS-CoV-2 disinfection rates >90% can effectively occur in a 2.5 m high room with ventilation rates between one to six ACH, and one UV lamp (30 W) located every 18.58 m². This will result in an average UV flux of 50 µW/cm².

Feng (2020) ^{Footnote 11}

Simulation study

USA

Jan 2021

This study aimed to evaluate the effectiveness of a novel portable UV air cleaner in reducing airborne droplets with SARS-CoV-2 in a patient's room (4.8 m length x 4.3 m width x 2.4 m height). Simulations were conducted using a computational fluid-particle dynamics model. In these simulations, a patient emitted droplets with SARS-CoV-2, and the effectiveness of the portable UV air cleaner was assessed under different flow rates and ventilation conditions. Wavelength and power of the UV air cleaner was not specified.

Effectiveness was measured by the reduction in concentration of droplets with SARS-CoV-2 suspended in the room and in the main ventilation system.

• A portable UV air cleaner could filter up to 82% of airborne droplets with SARS-CoV-2.
• Increasing the flow rate of the UV air cleaner could improve its efficiency in filtering droplets with SARS-CoV-2, which meant a higher number of droplets filtered in a unit time.
• However, increasing the flow rate could also increase convection and airflow recirculation in the room, resulting in a wider distribution of airborne droplets with SARS-CoV-2 in the room.
• Note: Simulated visualizations are provided, which show the SARS-CoV-2 droplet deposition patterns in the room under different UV air cleaner flow rates and ventilation conditions.

Volchenkov (2021) ^{Footnote 6}

Field investigation

Russia

2003-2020

Researchers examined the effectiveness and safety of upper and whole room UVGI in reducing COVID-19 and TB transmission among employees and patients within a hospital building.

The UVGI source consisted of 240 wall-mounted UV-C fixtures (one fixture per 18 m²). Each UV-C fixture contained two low-pressure mercury lamps (T8 30 W, wavelength = 254 nm) for upper and whole-room UVGI, respectively.

The upper-room UVGI lamp was used 24 hours per day, 7 days per week, when people were present and absent from the hospital rooms, while the whole-room UVGI lamp was only used when people were not present in the room for sterilization purposes.

• In 17 years of using the UV-C fixtures in the hospital, no cases of overexposure have been reported due to upper-room UVGI, while some cases of overexposure were reported due to whole-room UVGI (people were not supposed to be present).
• Effectiveness results found in Table 1.

Hou (2021) ^{Footnote 12}

Simulation study

USA

Mar 2021

Researchers used ray-tracing to simulate the impact of different room design parameters on the safety and effectiveness of UV-C irradiation of SARS-CoV-2. The simulation involved the use of an occupied test room (24 x 30 feet, floor area = 15.80 m²), which had one wall-mounted UVGI lamp (254 nm, 25.47 W).

Three room configurations were examined: configuration 1 (square, 3.97 m length x 3.97 m width, UVGI lamp located on one wall), configuration 2 (rectangle, 4.57 m x 3.44 m, UVGI lamp located on long wall), configuration 3 (rectangle, 4.57 m x 3.44 m, UVGI lamp located on short wall). Four ceiling height / UVGI device mounting heights (C/M height) were examined: C/M height 1 (2.44 m / 2.13 m), C/M height 2 (2.74 m / 2.13 m), C/M height 3 (2.74 m / 2.29 m), and C/M height 4 (3.05 m / 2.44 m).

A simulated case study involving an occupied hospital room (7.32 m length x 4.57 m width, default ceiling height = 2.74 m, floor area = 27.87 m²) with two upper zone UVGI lamps (Atlantic Ultraviolet Corporation Hygeaire model LIND24-EVO, lamp power = 25.6 W each, 254 nm) was also performed to determine the impact of different room design parameters on the safety and effectiveness of UV-C irradiation of SARS-CoV-2.

Four scenarios for UVGI fixture location were examined: scenario 1 (UVGI fixtures located above the bed), scenario 2 (UVGI fixtures located opposite of the bed), scenario 3 (UVGI fixtures located on the left and right side walls from the bed), and scenario 4 (UVGI fixtures located above and beside the bed). Three patient room layouts were examined: layout 1 (L-shaped, 7.32 m length x 4.57 m width, default), layout 2 (rectangular, 6.01 m x 4.57 m), layout 3 (rectangular, 7.01 m x 3.96 m). Three room surface UV-C reflectance coefficients were examined (0.05 (default), 0.1, and 0.2). Three scenarios for ceiling height / UVGI fixture mounting height were examined: height 1 (2.74 m / 2.13 m), height 2 (2.74 m / 2.29 m; default), and height 3 (3.05 m / 2.44 m).

SARS-CoV-2 disinfection effectiveness was measured using average fluence rate (µW/cm²).

Simulation study:

• Configuration 3 is the safest room configuration for individuals occupying the lower zone, as it has the optimal distance between the UVGI device and the opposite wall, which allows less UV light to be reflected off the wall and onto those occupying the lower zone.
• C/M height 4 is the safest option with the lowest likelihood of UV overexposure to occupants in the room, because it has the largest difference between the ceiling and the UVGI device mounting height, causing lower amounts of UV light to be reflected from the ceiling to the lower zone.
• Accounting for both SARS-CoV-2 disinfection effectiveness and safety, C/M height 3 is the optimal option.

Case simulation:

• The highest risk of UV-C overexposure for humans was in Scenario 3, while Scenario 1 had the lowest risk of UV-C overexposure.
• As the UVGI fixture mounting height increased, there was a decrease in room areas with UV-C irradiance above the safety threshold (0.2 µW/cm²).
• Height 2 had a reduced level of lower zone UV-C irradiance compared to Height 1, indicating that the higher the UVGI lamp is located on the wall, the lower the risk of over-exposure.
• Increasing the surface reflectance coefficients from 0.05 (default) to 0.1 and 0.2, increased the upper zone average fluence rate, and caused lower zone UV-C irradiance greater than the safety threshold (0.2 µW/cm²), when the wall height was below 1.83 m. The greatest lower zone UV-C irradiance was observed at a reflectance coefficient of 0.2.
• Effectiveness results found in Table 1.