The CDC says UV light kills "viral, bacterial, and fungal species" (i.e., germs): cdc.gov/niosh/nioshtic-2/20034387.html

@Yehuda_NYC It takes a relatively long exposure to kill anything. In labs the working enclosures are exposed to UV for half an hour to overnight to kill bugs on the surfaces. In an HVAC system, air flows by the UV source, limiting exposure for airborne particles. Also, be sure never to look at or even have skin exposure to light from such a UV source.

for such a small hole I would reccommend offset aviation snips not regular.

@Armand This EPA publication says that the effectiveness of an UV light in an HVAC duct varies from 9% to 99% depending on the organism (see the executive summary and the table on page 11). The 9% was for spores.

@Yehuda_NYC It depends on the light intensity, wavelength and "dwell time" of exposure. For airborne agents, to get something at least moderately effective, you need some combination of bright UV sources along a long stretch of ductwork or slow movement of air through the irradiation area. I would be dubious about an untested DIY setup actually having a significant effect. Hope you find something effective!

@Armand The US EPA says UV Light in HVAC is effective it killing biological warfare organisms, as per the link I provided. What support do you have of your contrary view?

@Yehuda_NYC It's not a contrary view; I'm just saying that dosage matters. If you just slap a relatively low intensity UV source in a duct with air whizzing by, it won't do much. I have 25 years experience working with microorganisms in biology labs and have found as a practical matter that microfiltration (e.g. 0.2 micron filters) and autoclaving (pressure cooking) work well. I realize those are not practical for your application, so I urge you to make sure what you use has been tested as a system (as built) to be effective.

@Yehuda_NYC William P. Bahnfleth noted "Most important, one must define ... what dose needs to be delivered over what period of time for the design level of microbial susceptibility and under what ambient conditions—temperature and air speed. This defines the average fluence (the germicidal energy per unit area incident on a point in space from all directions) ... that is needed and can be the basis for selecting an appropriate number and configuration of lamps. System design should be supported by calculations to confirm the required UV-C dose is delivered by the system."

@Yehuda_NYC Further technical details at: 2019 ASHRAE Handbook—HVAC Applications, Chap. 62, “Ultraviolet Air and Surface Treatment.” and 2020 ASHRAE Handbook—HVAC Systems and Equipment, Chap. 17, “Ultraviolet Lamp Systems.”

@Armand It might be ideal to calculate effectiveness for each application, but the EPA publication indicates we are in the ballpark. Surely dismissing UV light in HVAC ducts as ineffective is going too far (as you wrote "It takes a relatively long exposure to kill anything.")

@Yehuda_NYC I can't actually access the 2008 article on any of my browsers - they just hang on it - so I can't see the irradiance levels and airflow they recommend. Seems like you are aware of the potential issues, though, so good luck with your project!

If you're constantly pumping enough UV into air that floating germs die, you should consider the resulting ozone levels that will be a by-product. Ozone is carcinogenic; do you want lung cancer or a cold?

@dandavis An HVAC application for UV would use bulbs rated not to produce any ozone (ones that don't produce the shorter 1xx nm UV that generates ozone, just the ~265nm germicidal UV-C). Ozone also directly hurts lung function, so has bad effects even with short exposures.

@Armand This seems to be the same report in a different format: cfpub.epa.gov/si/…

@Yehuda_NYC Yes, the data in the table in the link look quite reasonable. In particular, I see the one system tested that did a good job on bacterial spores slowed the airflow by a factor of about 6.5x compared to the others, and systems 2-4 with much lower power didn't do a very good job.