Air and Surfaces UVGI Disinfection for Medical and IVF Clinics

Author: Normand Brais P.Eng., M.A.Sc., Ph.D

Normand Brais holds a Mechanical Engineering degree, a Master in Mechanical Engineering and a PhD in Nuclear Engineering from Ecole Polytechnique de Montreal. He was appointed Professor at the Energy Engineering Institute after he graduated. He has founded several technological companies in various fields such as air pollutants from combustion equipment, biomass combustion, photonics, and air or surface ultraviolet disinfection. In 1995 he founded Sanuvox Technologies, which is now a worldwide technological leader in UV disinfection of air and surfaces for hospitals, labs, and commercial buildings


Germicidal ultraviolet systems have been widely used in the healthcare industry for the disinfection of equipment, surfaces, and air supply systems. Applications include medical equipment disinfection, whole room decontamination, walls and floors disinfection, cooling coils disinfection, and operating room disinfection. UV is a highly effective and predictable technology to insure sterilization of almost any types of surfaces and to eliminate microbial growth. Public health agencies such as the CDC (Center for Disease Control) recommend the use of UV as an effective technology to disrupt the transmission of pathogens in building ventilation systems.The current rarity of UV disinfection stems from the fact that it has been erroneously assumed that air filters are sufficient to provide sterilized air. The last 25 years of data has shown that this is far from reality. When dealing with sub-micron bio-contaminants in the size range of 0.1 to 0.4 micron, even the best filtration technologies fail to stop them all. HEPA filter challenged with a concentration of 1 million viable particles per cubic meter at a flow rate of 1,000 m3 / hr, could allow as much as 500,000 particles every hour to go through. During the course of a single day, a total of 12 million bio-viable particles will penetrate the filter and contaminate the aseptic zone.Those uncaptured bio-contaminants particulates can be rendered innocuous by a proper use of Ultraviolet Germicidal Irradiation (UVGI) technology. This chapter presents an overview of how germicidal ultraviolet light can be engineered to sterilize a wide spectrum of microorganisms such as viruses, bacteria,

1 Introduction

Evidences have accumulated over the years that following the standard guidelines and codes for designing health-care facility ventilation systems is far from being sufficient to ensure a sterile environment (1) (2) (3) (4) . Sterility is generally defined as 99.9999% reduction of a population of microorganisms. This means that as little as one microorganism in a million is expected to survive after disinfection. Standard traditional air filtration with HEPA (High Efficiency Particulate Air) filters or ULPA (Ultra Low Penetration Air) filters in hospital, labs, and clinics ventilation systems to control airborne pathogens have been widely adopted. However, multiple studies have demonstrated that despite the use of such filters, viral and bacterial airborne contamination are still ubiquitous in these ventilation systems(5) (6) (7) . The most common explanation for filter ineffectiveness is often pointing at the filter rack seal joints bypass, filter puncture leakage, and poor installation or maintenance. Although all these points remain valid and can always be improved, the physical cause is rooted in the fact that all filters show a significant drop in their capture efficiency for a certain range of particulates sizes. In this critical size range, the particles are either too small to be captured by interception/impaction or too large or to be removed by diffusion/electrostatic. This is just a straight forward consequence of the fundamental principles of filtration physics(8). HEPA filters are no different and also display a weakness at a critical particle size between 0.1 and 0.4 microns as shown in fig.1 and Fig.2. The HEPA filter efficiency drops to a minimum value of 99.95% at a critical point called MPP (Most Penetrating Particle) which is around 0.2 microns.

If a HEPA filter is challenged with a concentration of one million particles per cubic meter falling within its vulnerable size range, simple calculations show that for each cubic meter of air as much as 500 particles every hour will pass through the HEPA filter. During the course of a single day, a 1,000 m3 /hr “fresh air” ventilation system will allow12 million viable particles to contaminate the aseptic zone. It is also worth noting that in order to perform according to specifications, the air velocity facing the filters must be below a specified value. For HEPA filters it is generally recommended not to exceed an incoming velocity of more than 1.3 m/sec. Since many air handling units are often designed to operate at much higher velocities, the actual performances of HEPA filters end up being substandard.

Within the filter vulnerable particle size extended range of 0.02 to 0.7 micron, the following microorganisms also called “viable particulates” are typically found (Table 1). Medical specialists will easily recognize that many of the listed bio-contaminants in this critical size range are highly undesirable inside a medical environment.

Table 1 shows that when challenged by one million particles, some viable microorganisms can penetrate through the filter for certain particular sizes around 0.2 microns. Considering that sterility is defined as less than one survivor on a million microorganism population, it is quite clear that the air disinfection process initiated by HEPA filtration is not sufficient for IVF laboratories and as such requires a finishing step. Unlike filtration, ultraviolet germicidal irradiation (UVGI) does not capture or retain microorganisms, it sterilizes them by damaging their DNA/RNA strands as they pass-by an intense germicidal UV light zone. Contrary to a filter that accumulates particulates until the pressure drop increases to a point where it needs replacement, UV disinfection systems have a negligible pressure drop and require comparatively very low maintenance.

2 Fundamentals of Ultraviolet Disinfection Process

2.1 UV light spectrum

Being outside our visible wavelength range, the UV light spectrum that extends from 100 to 400 nm is not visible to the human eye. The UV spectrum has been arbitrarily subdivided into four bands:

• UV-A band (400–315 nm) — the most abundant in sunlight reaching the Earth’s surface

• UV-B band (315–280 nm) — primarily responsible for skin reddening

• UV-C band (280-200 nm) — the most effective for germicidal effect

• Far or Vacuum UV (200 – 30nm) – Ozone producing and ionizing radiation

2.2 UV disruption of DNA and RNA

The discovery of microbial disinfection by UV light dates back to 1877 (9) . Then later, in 1928 F.L. Gates(10) identified the specific wavelength of UV light that was responsible for the observed germicidal effect. The detailed fundamental physical mechanisms explaining the interaction of specific wavelengths of light with specific DNA/RNA molecular bonds were finally revealed by quantum mechanics, developed during the first half of the 20th century. More recent biochemical research has shown that the most effective germicidal wavelengths of 265 nm coincides with the peak absorption spectra of nucleic acids (11) as shown in fig.4. On the basis of this correlation, the majority of the damages inflicted to sterilized microbes were found in their genetic material. The primary recognized mechanism in UV disinfection is now confirmed to be cumulative molecular damages to DNA and RNA strands. The disruption of nucleic acids by UV light has the ability to affect the complete spectrum of microorganisms, making them all sterile given a sufficient dosage and consequently, making them unable to infect a host. Within the limits of experimental accuracy, the lethal action of germicidal UV appears to be independent of the nature of the organism and, unlike antibiotics, there has been no signs of adaptive resistance after almost a century of wide usage for drinking water disinfection. Most if not all commercially available germicidal light sources are based on fluorescent tube technology and emit between 30 and 35% of their input power at 253.7 nanometers, a wavelength very close to the peak germicidal wavelength of 265 nm as shown in Fig.4.

UVGI sterilization of microorganisms is therefore achieved in practice with the low cost and widely available wavelength of 253.7 nm. The quantum energy carried by UV-C photons is high enough to dissociate most single chemical bonds between carbon, hydrogen, oxygen, and nitrogen atoms. The molecular disruptions caused by these energetic photons result in irreversibly damaging the nucleic acids of a microorganism until it is no longer viable. Amongst various UV radiation damages to DNA, the formation of cyclobutane pyrimidine dimers (CPDs) and pyrimidine-pyrimidone 6-4 photoproducts (6-4 PP’s)(12) , CPDs are caused by covalent bonding between two adjacent pyrimidines. However, as shown in figure 5, UV-C usually generates thymine dimers in the greatest quantity, cytosine dimers in low quantity, and mixed dimers at an intermediate level (13). In UV irradiated RNA viruses, the nucleotide uracil forms pyrimidine photoproducts. At irradiation dose of magnitude high enough to overwhelm the nucleic acid repair mechanisms, damages result in irreversible alterations, impairment of replication and genetic transcription, and eventual death of the organism. For a thorough assessment of the photochemistry ofUVinduced damages to nucleic acids, and on repair mechanisms, review the in-depth description by Kowalski (13).

2.3 UV Dose – Response calculation

To acquire a mathematical understanding of the UVGI disinfection process predictability, it is helpful to picture it as the analogue of a bombardment of photons hitting a microbe. Each photon carries an amount of energy called a quantum Eλ , of a value linked to the light wavelength according to the Planck-Einstein relation : 𝑬𝝀 = 𝒉 𝒄/𝝀 Eq.(1) Where h = Planck’s constant, 6.626 x 10-34 Joule.sec c = Speed of light in vacuum, 2.998 x 108 m/sec λ = wavelength, m Using the Planck-Einstein relation, the energy conveyed by each UV-C photon at a wavelength of 253.7 nm is equal to 7.83 x 10-19 Joule. Therefore the number of photons per Joule is the inverse i.e. 1.28 x 1018 photons per Joule.

Remembering that one watt of power is defined as a rate of one joule of energy per second, then a UV intensity of 100 Watt/m2 provides a flow of 1.28 x 1020 photons per second per square meter. Now, considering that a virus of 0.2 micron diameter represents a target area of only 3.14 x 10-14m2 , despite its small size, this virus will be bombarded by as much as 4 million photons every second! Given a sufficient exposure time to this photonic assault, photochemical damages will accumulate enough to render the organism biologically dysfunctional. In reality, regardless of the tremendous number of photons shooting at this virus, only a very small number hit their target successfully to initiate a photochemical reactions. The real effective inactivation cross sectional area of a target microbe is a function of many parameters, among them, the quantum chemical yield, the outside capsid protective layers, and the particular distribution of its DNA sequence. A promising and useful predictive method based on the above described photon bombardment concept and successful hit probability has been published to predict the UV susceptibility of microorganism as a function of their genome without using classical experimental bio-lab test procedure. (14)

Based on the above described UV bombardment analogy, a mathematical relation can be written to express the UV dose response for a population of microorganisms. It is statistically fair to infer that the rate of decay of a microbial population will vary proportionally to the number of successful hits over a period of time. This rate of successful hits can be described as the product of the UV power per unit area I, the number of bio-organism N, the bioorganism effective UV inactivation cross section k, also called the bio-organism UV susceptibility constant, and the exposure time t as follow: Hit rate = 𝒅𝑵/𝒅𝒕 = 𝒌 𝑵 I t Eq. (2) Integration of equation (2) yields: 𝑵(𝒕) = 𝑵𝟎𝒆 −𝒌𝑰𝒕 Eq (3) Where N0 = initial number of microorganisms, Nt = number of microorganisms surviving after any time t, k = a microorganism-dependent UV susceptibility constant, in m2 /Joule, I = the irradiance UV intensity received by the microorganism, in Watt/m2 t = exposure time, in seconds The fraction of microorganisms initially present, which survive at any given time, is called the survival ratio S and can be expressed as: 𝑺 = 𝑵𝒕 𝑵𝟎 Eq(4) The sterilized fraction is what is called the disinfection rate, is simply 1 minus the survival ratio. 𝑫𝒊𝒔𝒊𝒏𝒇𝒆𝒄𝒕𝒊𝒐𝒏 = 𝟏 − 𝑺 = 𝟏 − 𝒆 −𝒌𝑰𝒕 Eq(5) As explained above, we can define the germicidal UV dose by the total number of UV photons emitted per unit area during a time interval, which can be written as: 𝑼𝑽 𝑫𝒐𝒔𝒆 = 𝑰 × 𝒕 in Joule/m2 Eq (6) By substituting eq.(6) in eq.(5), we finally get the well verified germicidal UV Dose-Response relation: 𝑫𝒊𝒔𝒊𝒏𝒇𝒆𝒄𝒕𝒊𝒐𝒏 = 𝟏 − 𝒆 −𝒌 𝐔𝐕 Dose Eq (7) What equation (7) reveals is that a given dose produces a given disinfection rate, whether the dose consists of low intensity for a long exposure time, or a high intensity for a short time. A key difference between surface decontamination and airborne disinfection is the available exposure time. Whereas the exposure time for any induct disinfection will be of the order of a few seconds or a fractions of a second depending on airflow velocities, the exposure could be minutes or even hours for stationary surfaces such as wall, floors or air cooling or heating coils. Therefore, the UV intensities for airborne microorganisms are required to be orders of magnitude higher than that typically used for stationary surface disinfection. The graph of Fig.6 illustrate this exponential time decay relation for some microorganisms of concern under a constant UVGI intensity of 10 mW/cm2 . It shows that there are significant differences in the exposure time required for the same level of disinfection between the most and least UV susceptible microorganisms of concern. The one that requires the highest dosage will be the one governing the sizing of the UV system.

2.4 Susceptibility of microorganisms to UV energy

Organisms differ in their susceptibility to UV inactivation. A few examples of familiar pathogenic organisms are included in each group for reference. It is important to note that it is impossible to list all the organisms of interest in each group. Depending upon the application a public health or medical professional, microbiologist or other individual with knowledge of the microbial threat or organisms of concern should be consulted. In general, vegetative bacteria are the most susceptible to UV, followed by mycobacteria, then the bacterial spores and finally the fungal spores which are the most resistant to UV energy. Within each group, an individual species may be significantly more resistant, so care should be taken using this ranking only as a guideline. It should be noted that the spore forming bacteria and fungi, also have vegetative forms which are markedly more susceptible to inactivation than the spore forms. Viruses are particularly problematic to categorize as their susceptibility to inactivation is even broader than that of bacteria or fungi. 

Based on Eq. (5), it is clear that smaller values of k represent organisms that require a higher UV dose for disinfection. Units of k are m2 /Joule which is the inverse of the units used in UV dose. For example, the value of the UV susceptibility of Influenza-A virus has been measured experimentally by Jensen in 1964 and was found to be 0.0119 m2 /J. Based on this value, one can easily calculate the required UV dose to be applied to reach 90% disinfection of a population of influenza-A virus using the following formula: 𝑫𝟗𝟎 = 𝒍𝒏(𝟏𝟎) 𝒌 = 𝟐.𝟑𝟎𝟑 𝒌 in J/m2 Eq.(8) The D90 value for Influenza-A virus is therefore equal to 19.3 J/m2 . The D90 value has a high practical interest as it allows the UV system designer to quickly evaluate the required UV dosage to obtain a desired disinfection level. For example, providing a UV dose of twice the D90 will result in a disinfection level of 99%. Delivering three times the D90 dose will result in 99.9% disinfection rate, and so on. It can be easily demonstrated mathematically that the number of 9s, also called the disinfection LOG value, is simply equal to the delivered UV dose divided by the D90 value. To reach sterility, a condition that we have previously defined as a disinfection level of 6 LOG or 99.9999%, at least 6 times the D90 value of the most resistant microorganism must be delivered. Extensive compilations of published UV susceptibility values “k” can be found in several references in the available literature (13) .

3 UVGI Dosage required for adequate Air Disinfection of IVF Clinics

Given the nature of the sensitive procedures performed within IVF clinics where extremely fragile embryos are being manipulated, the target air disinfection level should be as close as possible from total sterility. To determine the required UV dose for IVF clinics, we should first examine the list of the microorganisms of concern listed in table 2 that fall within the vulnerable size range of HEPA filters and compare their UV susceptibility k to find out the most resilient species.

By examining the D90 values shown in table 2, it is obvious that the most resilient microorganism is by far Francisella Tularensis which requires a UV dose of 25.59 mJ/cm2 for 90% disinfection. In order to reach 6 LOG of overall disinfection after filtration, the UV system must therefore be designed to at least sterilize 499 of the 500 remaining bacteria, i.e. a disinfection rate of 499/500 = 99.8% which is just a little short of 3 LOG (99.9%). Consequently, the UV system sizing criteria consist in delivering a UVGI dose of a little less than 3 times 25.59 mJ/cm2 . Exact calculation shows that a dose of 75 mJ/cm2 must be delivered to the air stream before entering the aseptic space. The disinfection level obtained when a UV dose of 75 mJ/cm2 is delivered after HEPA filtration is shown in table 3.

This UV dosage insures an overall disinfection of at least 6 log for all of the microorganisms of concern. When as described here, filtration is used in conjunction with UVGI disinfection, a combined disinfection efficiency can be calculated using the following formula: 𝑫𝒊𝒔𝒊𝒏𝒇𝒆𝒄𝒕𝒊𝒐𝒏𝒐𝒗𝒆𝒓𝒂𝒍𝒍 = 𝟏 − (𝟏 − 𝑭𝒊𝒍𝒕𝒆𝒓𝒆𝒇𝒇)(𝟏 − 𝑼𝑽𝒆𝒇𝒇) Eq.(9) Therefore, to attain an overall disinfection of 99.9999% i.e. 6 LOG of sterility equivalent, the following UV disinfection efficiency is required for a given filtration efficiency: 𝑼𝑽𝒆𝒇𝒇 = 𝟏 − 𝟏𝟎−𝟔 𝟏−𝑭𝒊𝒍𝒕𝒆𝒓𝒆𝒇𝒇 Eq.(10).

According to Eq.(10), if the HEPA filter MPP efficiency is 99.95% for the most penetrating particle size, then the UV disinfection efficiency must be designed to be superior to 99.8% so that an overall disinfection above 99.9999% or 6 LOG is achieved. Using eq.(7) with the controlling UV susceptibility value of Francisella Tularensis, the minimum required UV dose to reach 99.8% disinfection is computed to be 75 mJ/cm2 . Repeating the above calculation but with a higher performing ULPA filter, where the MPP efficiency is equal or greater than 99.99%, then the UV dose needed to complete the disinfection and insure air sterility drops to 50 mJ/cm2 .

3.1 UVGI air disinfection system design guidelines for IVF

The question that every designer wants to be answered here is how much UV power does it take for a given air flow to deliver the target UV dose that will insure sterility. Before getting into an example of such a design, it is important to note that equation (7) does not give any indication of the distribution of UV energy as a function of x,y,and z coordinate given by set of UV lamps positioned inside an air duct . In an air duct, the physics is complicated by the movement of the target microorganisms in a turbulent stream or even sometimes stratified air stream. Whereas the turbulent case approached the well mixed condition, the worst case scenario remains the stratified air stream often miscalled laminar. That coupled with the fact that the light intensity decays as an inverse square law with the distance from the source, and that the physical geometry of the duct, number of UV lamps, and UV lamps positions in the duct, all affect the final delivered UV dose. Only a computerized program using a numerical integration to sum the contribution of all the UV elemental sources can properly calculate the UV irradiation field inside a duct. Such calculation also incorporate the important contribution of UV reflective properties of the duct wall surfaces on the performance enhancement of UV disinfection.

The first design guidelines for UVGI airstream disinfection systems were developed in the 1940s (15) . These guidelines propose rudimentary charts and tables to size lamps and reflective surfaces so as to obtain a desired average UVGI dose. These sizing methods, though sometimes admirably detailed for that time, suffer from several fundamental deficiencies, the most important being:

− The real three dimensional intensity field is not defined, it is instead simply evaluated based on the lamp power rating or else relying on basic photometric data taken at lamp midpoints.

− Lamps are specified without regard to lamp positioning.

− The correction factors for reflectivity ignore duct dimensions and lengths.

Still today, too many UV systems are unfortunately sized using crude rules of thumb, such as filling the available cross-section of ductwork with a row of lamps. Such misuse has invariably ended up with poor performances and deceived some UV system users. The available computational power of modern day computers allows for adequate custom sizing of any induct UV disinfection system. Proper calculation for predicting the applied UV dose must take into account the relevant input parameters describing a rectangular UVGI system in terms of its geometry, lamp characteristics, lamp placement, lamp orientation, and surface reflectivity. The program requires the following input parameters for each computation:

− Airflow rate

− Height, Width, and Length of duct.

− % Reflectivity of inner surfaces

− Lamp UV output power (W), lamp length and diameter

− 3-D positioning coordinates of each UV lamp (xi, yi, zi)

− Target microorganism susceptibility constant k (m 2 /J)

Figure 7 shows an example of the typical output of such software for a set of five lamps of 50 Watt each of net UV output installed at the center of a 500 mm square duct lined with reflective polished aluminum. The air flow is 1000 m3 /hr at a temperature of 22 C. The numbered iso-contours represent the cumulative UV dose distribution in mJ/cm2 received by the air stream at the exit of the duct. The UVGI dose iso-contours expressed in mJ/cm2 show that even in the worst case where the air flow is not well mixed and totally stratified, the lowest irradiated zones are still subjected to a UVGI dose above the 50 mJ/cm2 required level.

Taking all these critical sizing variables into account, Kowalski (17) has proposed a dimensional analysis to assess the sensitivity of all the above design parameters on the disinfection performance of a UV system. The outcome is a useful simplified general scaling correlation equating the delivered UV dose as a function of air flow, UV output power, and duct length. The formula is as follows: 𝑼𝑽 𝒅𝒐𝒔𝒆 ~ 𝑷 x 𝑳 / 𝑸 Eq. (11) Where P = Power output of UV source in Watt Q = Air flow in m3 / sec L = UV exposure length For upscaling or downscaling purposes, eq. (11) tells us that if the flow rate is doubled in the same duct size, then the UV power or number of lamps must be doubled as well to maintain the same disinfection performance. The same can be said about the duct UV exposure length L, if it is reduced by half to make the system more compact, then the UV output power will have to be doubled to compensate. Applying this correlation to the previous example where 250 Watt of UV output over 2 meter exposure length provided a given disinfection level to an air flow of 1,000 m3 / hr, we can easily answer the question of how much UV output would be require to insure the same level of disinfection for an air flow of 2,000 m3 /hr over the same exposure length, the answer is simply twice the UV output i.e. 500 Watt. By observing in eq.11 that the air flow Q is the product of the duct cross section A in by the air velocity V and that the UV exposure time tis simply the ratio of the duct length L to the air velocity V, we can rewrite it as follows: 𝑼𝑽 𝒅𝒐𝒔𝒆 ~ 𝑷 × 𝒕⁄𝑨 Eq(12) This scaling relation concisely expresses the fact that the delivered UV dose is the product of UV lamps output power in watts with the exposure time in seconds divided by the duct cross section area. Fig.7 shows an actual picture of the UVGI disinfection system described and calculated in this example. 

It is worth mentioning that the inner walls reflectivity provide a significant contribution to the total UV field. These reflections that echo between surfaces are called inter-reflections. The resulting intensity due to the inter-reflections will achieve steady state at the speed of light, converging to a finite value that depends on duct geometry and inner surface reflective properties. The physical process of inter-reflections is also taken into account by computer models. Neglecting to use highly reflective duct lining surfaces such as polished aluminum severely impairs UV system performance. It should be noted that the reflective properties for UVC wavelength is very different compared to visible light reflectivity. Despite its reflective features for visible light, stainless steel has a low UVC reflectivity of only 20%. duct to improve UV disinfection performance. UVC reflectivity data is published and must be included in a proper calculation to maximize the energy efficiency of the UV disinfection system.

3.2 Effect of air velocity, temperature, and lamp aging on UV system output

Air temperature and velocity may vary over a wide range within a ventilation system, causing significant variations in UV lamp output and must be adequately accounted for. UV lamps are designed in such a way that the lamp surface temperature must be between 38 °C and 50 °C to reach the maximum UV output ( Fig.8). In moving air the temperature of the UV lamps could become too low and the UV output will fall as seen in fig.9. and fig 10. To minimize the chilling effect and allow higher operating efficiency under cold air flow condition, it is preferable to install the lamps parallel to the flow instead of perpendicular cross-flow.

UV lamp output decreases over time due to lamp aging. UV lamps are rated in effective hours of UV emission and not in end of electrical life hours. Most UV lamps emit intensity levels at the end of their useful life ie 20,000 hours, that are 80% or more than those measured at 100 hours of operation. UVGI systems should be designed for UV output at the end of effective life. Modern day UV lamps will continue to emit blue light long after they have passed their useful germicidal lifespan. Variations in UV lamp output due to lamp type and ambient conditions can be of equal or greater magnitude than depreciation due to aging. These changes are cumulative and can reduce lamp output by as much as 50% over a range of typical conditions found in ventilation systems (18). Consequently, Consequently, good engineering practice must take these sources of output variation into consideration.

4 Ultraviolet surface disinfection

Environmental surfaces play an important role in the transmission of healthcare associated pathogens. Several studies have demonstrated that cleaning surfaces in healthcare facilities is often suboptimal (18). There are two types of UVGI surface disinfection systems, they can be mobile or fixed.

4.1 Mobile UVGI surface disinfection units

Mobile UVGI units are momentarily placed in contaminated areas to disinfect whole room surfaces. The unit shown in Fig.11 has sufficient UVGI power to provide a 6 LOG disinfection for Clostridium Difficile spores in a square room of 5 m by 5 m on all exposed surfaces within 15 minutes. Such mobile unit is equipped with multiple motion sensors that will cause it to shut down is someone enters the room during the sterilization cycle. It is also equipped with a data logger that will keep a time and location record of every disinfection cycle performed during a given period.

4.2 Fixed automated UVGI disinfection units

In critical areas such as operating rooms and labs or in notoriously contaminated places like bathrooms and equipment storage rooms, permanent automatic UVGI units can be used (See Fig. 13). Those units are activated automatically when the rooms is unoccupied after each entry or use. They include a programmable logic controller with a timer, redundant motion detectors and door switch for personnel safety. With a properly engineered UV output relative to the size of the room that ensures a minimum UVGI intensity of 30 microwatt/cm2 on the target surfaces, a disinfection cycle time of 5 to 10 minutes has demonstrated up to 6 LOG disinfection of the most commonly found pathogens. See typical expected disinfection results in Fig.12.

4.3 Air conditioning cooling coils disinfection

Heating Ventilating Air Conditioning (HVAC) systems provide a suitable environment for the growth of a wide range of fungal spores and bacteria. The presence of moisture or high relative humidity is an effective catalyst for the germination and growth of fungal spores. Dust provides a nutrient base on which fungi can grow. Environmental bacteria can grow biofilms and thereby provide fungal spores a nutrient base. Consequently, the following molds and bacteria are ubiquitous inside HVAC: aspergillus, penicillium, mucor, cladosporium, fusarium, alternaria, pseudomonas aeruginosa, e.coli, salmonella, and legionella.

Nearly every HVAC technician has had the experience of opening a unit to find the drain pan and coil covered with a slimy residue of mold biofilm. Not only these conditions can be unhealthy and occasionally deliver unpleasant smell for building occupants, but it also ruins the heat transfer capacity of the system and consequently increases the energy operating cost.

Various coil-cleaning methods have been used to try to control this problem. Many of those techniques involve the use of detergents or even solvents, which can pose safety issues – health and flammability, for example – and high pressure washing that diminishes the life of the coil, because sometimes acids are involved. Often coil cleaning isn’t done with regularity and even when it is done on schedule, the mold growth can return in a very short time, usually less than a month.

Air conditioning cooling coil fins constitute a fertile wet surface area at constant temperature that ends up being a major bacteria and mold incubator and reservoir. The removal of fungal growth inside cooling coils HVAC system is a common well-known application of fixed UVGI systems. Elimination of the air conditioning reservoir of microorganisms significantly reduces airborne infections. Because biofilm coated coils also impairs their heat transfer performance, the energy consumption is reduced with substantial energy saving paybacks

Since the UV light can be operated 24 hours a day every day, the disinfection of the air handling units requires very little power. Fig.13 shows a typical installation to maintain cooling coils biologically clean along with the engineering sizing software calculation to insure adequate UV dosage across the coil surface and between the fins.

To maintain a coil free of bio-contaminant, a constant minimum UV intensity of 0.25 mW/cm2 is required on its surface. Simple petri dish contact tests performed over the last 20 years have shown that this intensity is more than sufficient to ensure that 99% of the most resilient mold Aspergillus Niger will be rendered sterile after 3600 seconds of exposure. It follows that the required average UV output power per square meter of coil is only 2.5 watt and, considering the standard UV lamp efficiency of 33% and an overall uniformity compensation multiplier of 1.33, a total input power consumption of 10 watt per square meter of coil is all it takes to keep it sterile and clean at all times.

5 UVGI system maintenance guidelines

5.1 Lamp Replacement

UV lamps should be replaced at the end of their useful life based on recommendation of the equipment manufacturer. Although lamps can operate for as much as two consecutive years, it may be good maintenance to simply change the lamps annually (8760 hours under continuous use) and therefore insure that adequate UV dosage is always provided. Lamps can operate long after their useful life but will have reduced germicidal UV output. Switching lamps ON and OFF too often may lead to premature lamp failure. Consult the manufacturer for any specific information regarding expected lamp life in hours and the impact of frequent on-off switching on lifespan.

5.2 Lamp Disposal

UV lamps should be disposed of in the same way as other mercury-containing devices such as conventional commercial fluorescent bulbs. Most lamps must be treated as hazardous waste and cannot be discarded with regular waste. Low mercury bulbs can generally be discarded as regular waste however some state and local jurisdictions classify these lamps as hazardous waste. The US EPA has promulgated “Universal Waste” regulations to several types of hazardous waste including mercury bulbs. These regulations allow users to treat mercury lamps as regular waste for the purpose of transporting to a recycling facility. This simplified process was created to promote recycling

5.3 Inspection

UVGI systems must include a feedback component to alert maintenance personnel of UVC lamp failure. Any failed lamp should be replaced immediately. In the event the lamp becomes dirty or soiled due to inadequate pre-filtration or airborne bio-aerosols it should be cleaned. Any UV lamp may be cleaned with a lint free cloth and commercial glass cleaner or isopropyl alcohol.

5.4 Safety Design Guidance

In-duct UV systems shall be fully enclosed to prevent leakage of UV light to unprotected persons or materials outside of the HVAC equipment. All access panels or doors with access to the lamp chamber where UV may penetrate either directly or through reflectance shall be affixed with warning labels in appropriate languages. The labels shall be affixed to the exterior of each panel or door in a prominent location visible to persons accessing the system (19) . Lamp chambers shall be equipped with an electrical disconnect device. Positive disconnection devices are preferred over switches. Disconnection devices shall be capable of being locked or tagged out. Disconnection devices shall be located outside of the lamp chamber and adjacent to the primary access panel or door to the lamp chamber. Switches shall be wired in series such that the opening of any access will shut-down the UV system. On/off switches for UV lamps must not be located in the same location as general room lighting. Switches must be positioned in such a location that only authorized persons have access to them and should be locked to ensure that they are not accidentally turned on or off.

6 Conclusion

When properly engineered, germicidal ultraviolet systems can be extremely efficient for the disinfection of equipment, surfaces, and air supply systems in the medical field. Applications include medical equipment disinfection, whole room decontamination, walls and floors disinfection, cooling coils disinfection, and operating room disinfection. UVGI is a predictable, low cost, and mature technology to eliminate microbial growth. Public health agencies such as the CDC (Center for Disease Control) recommend the use of UVGI to disrupt the transmission of pathogens in building ventilation systems. 

The current uncommonness of UV disinfection is essentially due to the false perception that HEPA air filters are sufficient to provide sterilized air. Years of cumulated field experiences have shown that filters are certainly necessary but not sufficient. When dealing with sub-micron bio-contaminants in the size range of 0.1 to 0.4 micron, even the best filtration technologies fail to stop them all. Unlike filters, UVGI technology does not capture the bio-contaminants but it can effectively sterilize them when a proper dosage of ultraviolet is applied. The UVGI technology has the ability to sterilize a wide spectrum of microorganisms such as viruses, bacteria, and mold spores in air stream as well as on contaminated wall and objects.


1. American Institute of Architects. Guidelines for desing and construction of hospital and healthcare facilities. Washington, DC : AIA, 2001.

2. American Society of Heating, Refrigerating, and Air-Conditioning Engineers. HVAC Design Manual for Hospital and Clinics. Atlanta : ASHRAE, 2003.

3. —. Handbook – HVAC Applications. Atlanta : s.n., 2003.

4. CDC. Guidelines for Environmental Infection Control in Health-care Facilities. Atlanta : U.S. Department of Health and Human Services Centers for Disease Control and Prevention., 2003.

5. Shinsuke Kato, Minki Sung, -University of Tokyo. Using UVGI to counter contaminant dispersion. IFHE Digest. 2011.

6. Menzies D, Popa J, Hanley JA, Rand T, Milton DK. Effect of ultraviolet germicidal ligths installed in office ventilation systems on workers’ health and well being: double-blind multiple cross-over trial. s.l. : The Lancet Medical Journal, 2003. pp. 1785-91. Vol. 362.

7. Efficacy of environmental measures in reducing potentially infectious bioaerosols during sputum induction. Menzies, D., Adhikari, N., Arietta, M., Loo, V. 7, s.l. : Infection Control and Hospital Epidemiology, 2003, Vol. 24.

8. Kowalski, W.J., Bahnfleth, W. “MERV Filter Models for Aerobiological Applications. Air Media. Summer 13-17, 2002.

9. Research on the effect of light upon bacteria and other organisms. Downes, A., Blunt, T. London : Proceedings of the Royal Society, 1877. 26:488.

10. Gates, F.L. On nuclear derivatives and the lethal action of UV light. 1928.

11. Davidson, J.N. The biochemistry of nucleic acids. London : Metheun, 5th edition, 1965.

12. Setlow, R.B. Cyclobutane-type pyrimidine dimmers in polynucleotudes. Science. 1966, Vol. 153: 379-386.

13. Kowalski, W. Ultraviolet Germicidal Irradiation Handbook. Berlin : Springer-Verlag, 2009.

14. Kowalski, W. J., Bahnfleth, W. P., Hernandez, M. T. A genomic model for the prediction of Ultraviolet inactivation rate constants for RNA and DNA viruses.

15. Luckiesh, M., and Holladay, L. L. Designing installations of germicidal lamps for occupied rooms. . s.l. : General Electric Review, 1942. Vol. 45(6), 343-349.

16. Kowalski, W. J. Design and optiization of UVGI air disinfection systems. A thesis in Architechtural Engineering. s.l. : Penn State, 2001.

17. Bruno Lee, William Bahnfleth, Katja Auer. Life cycle cost simulation of in-duct Ultraviolet germicidal irradiation system.

18. Kowalski, W. Hospital Airborne Infection Control. Boca Raton : Taylor & Francis, 2012. 978-1-4398-2196-1.

19. Engineers, American Society of Heating Refrigerating and Air Conditioning. ASHRAE Handbook- Chapter 16. 2008.