Crossposting this essay by my friend Gavriel Kleinwaks, who is approximately the world expert in Far UV deployment.
[Edit: Austin is very kind--I am not close to being the world expert in far UV deployment; there are people who run/used to run companies trying to do that, and researchers who work with them very closely, who know more about far-UV deployment, and I'm largely consolidating information from them. -Gavriel]
Disclaimer: In the interest of avoiding the problem where a post languishes in drafts forever, I am not attempting to be particularly rigorous, but I’m happy to answer questions to fill in details about how I’ve come to the conclusions I’m writing about. This post is informed by my work experience at 1Day Sooner, but only reflects my personal takes, not the official stance of my organization.
I work on indoor air quality to reduce disease burden and pandemic risk, and far-UV is an extremely promising intervention. One of the questions I get most often is, “If this is so promising, why isn’t it everywhere?” The answer requires a lot of breakdown and I get a lot of follow-up questions as soon as I start to answer, so this post is an attempt to run through the entire conversation at once. I’d never actually get around to posting this if I also included a deep dive on far-UV or tried to prove its value, so if you don’t know yet why it’s promising, you can check out this report.
Top line takeaway: Far-UV isn’t widely used because no individual organization has the incentive or bandwidth to install it, and the industry is stuck in a vicious cycle, making companies unreliable. Filtration is already widely used, but more user-friendly filters would greatly improve implementation.
Q: What are the major roadblocks to widespread use of high-quality filters?
A: The market is hard for consumers to navigate; it’s difficult for consumers to be well-informed enough to assess which filters are high-quality. For example, ionizers (ozone producing and, according to some researchers, basically ineffective1) and portable HEPA filters (no harmful byproducts, universally agreed to be useful) are both sold as “air purifiers.”
Q: What are the major roadblocks to widespread use of far-UV?
A:
- Far-UV is not intuitive to most people (since other forms of UV are directly harmful to interact with), and people responsible for improving air quality in a given public space may not have the bandwidth to pursue hard-to-understand options over easy-to-understand options like filters.
- Far-UV does not have an immediately visible or easily measured impact on a given space, whereas filters can be immediately shown to remove particulate matter from the air using a standard particle counter. Since far-UV only addresses pathogens and not other airborne pollutants, its effect isn’t immediately visible.
- Reduction in disease transmission is very difficult to measure in one specific setting, because people in that setting can get sick elsewhere. This difficulty also means that the benefits of air cleaning are diffuse, but the costs are concentrated. In other words, setting-specific disease reduction has positive externalities, and its impact is most clear in a closed environment or an area where air cleaning tech has been installed in lots of individual buildings simultaneously.
- Businesses want a cost-benefit analysis for the use of far-UV, but a good cost-benefit analysis is impossible to produce without better evidence of real-world disease reduction.
- (EDIT 12/13/2025: Aerolamp is a new option on the market! Thanks to Jeff Kaufman for posting about it.) There are very few providers, and hardly any of them sell an off-the-shelf product. You usually can’t just buy a lamp to try it out—you have to call the company, get a consultation, and often have someone from the company come install the lamp. It’s a lot of overhead for an expensive product that most people have never heard of.
- It could be that even after far-UV is installed, it is only a good idea to have the lamps on at certain times and not others—for example, far-UV might have a harmful photochemical effect on a particular cleaning product, which would warrant the lamps being turned off during cleaning times, until the product dissipates.2
Q: Okay, sounds like more real-world research is needed on far-UV. What are the major roadblocks to that research?
A:
- Research is very expensive!
- People don’t want to feel like part of an experiment. When you ask people3 to run an experiment in an environment that they occupy, in which you’d install far-UV for the experiment, they get worried about it and ethical approval is more difficult to achieve.4 Hypothetically you can circumvent that problem by finding organizations that want far-UV anyway, and ask to measure the outcome of the installation,5 but then we get back to the problem of organizations wanting proof that the far-UV is worth their time/money/effort.
- Far-UV companies are historically unreliable or short-lived, which is a disaster for study planning. Far-UV companies that we have encountered are typically on the brink of failure, so by the time ethical approval and funding for a study has been successfully arranged, your intended supplier might have gone out of business. Some suppliers also produce low-quality lamps, which may not actually output their purported power, and are therefore less effective than anticipated.
- All of the above means that studies are difficult to power—it’s hard to get enough people into any one study with high-quality lamps at the right time.
Q: So far-UV is facing a vicious cycle. What differences in the state of the world would break the cycle—i.e. What’s your wish list?
A:
- A reliable source of off-the-shelf, low-cost, high-quality lamps with an attached ozone scrubber6: the ozone produced by high-power lamps is probably the greatest safety concern of far-UV, so having the attached ozone scrubber would eliminate some installation complexity, and a reliable company producing easy-to-install lamps would reduce overhead and confusion on the part of people responsible for ordering the lamps.
- Such a company would probably have to run as a nonprofit for a long time. There is very little technical difficulty to producing this type of lamp, and there is a nonprofit organization in the process of building it,7 but they are extremely bandwidth-constrained.
- Large research investments in clinical trials for far-UV efficacy in high-control settings: high-power clinical trials would probably inspire enough investor confidence8 to support a more robust far-UV market (although it would not necessarily solve the externality problem).
- Large investment in far-UV installations and data collection across many public buildings in a concentrated area, e.g. all non-residential buildings in a college town: this type of installation would address the externality problem and deliver real-world data.
- Large research investments in far-UV safety/photochemistry: concern about photochemical byproducts of far-UV is probably the main reason that far-UV does not have the full support of the research community. We already know that far-UV installations should happen only in environments with good ventilation, but for ease of communication, we want to be able to provide better-informed guidance on the ventilation rate.
- A high-profile installation: an installation at a large well-known company, or funded by such a company, could set trends for other organizations. (“If Google gets far-UV, we want it too!”)
- To support such installations, it would be useful to have a highly visible certification for “healthy air” in a way that references pathogen concentration, analogous to LEED certification, possibly an expansion of the WELL standard.
- One problem is that the ideal flagship adopters are the exact companies that have already invested heavily in high-quality air, and are reluctant to invest more without a great cost-benefit proposition.
Q: If you need big research funders or a high-profile installation, why not try to start a government research program? This seems like exactly the sort of thing they do.
A: The American government is institutionally conservative/risk-averse and we have heard from government offices that they want to see more excitement from the private market before initiating their own investments9 in research. However, we really hope that governments get excited about far-UV eventually. One very reasonable outcome would be for us to find, through careful research and statistical analysis, that far-UV is almost never worth it for individual private-sector actors to use in peacetime, but that far-UV is invaluable during a pandemic. If that is the case, it would be important for the government to stockpile far-UV lamps and pay for their widespread installation, to only be turned on in the event of a pandemic.
Q: Are there any good standard-setters outside of the government?
A: America’s authoritative voice on indoor air quality is the American Society of Heating, Refrigeration, and Air-Conditioning Engineers (ASHRAE). They have no regulatory authority, but their standards are used and referenced by many other organizations, and some building codes require compliance with ASHRAE standards. As a professional association of HVAC engineers, the general body of ASHRAE doesn’t have much epidemiological expertise and isn’t traditionally focused on disease reduction. The COVID pandemic brought indoor air quality to the forefront of epidemiology, so ASHRAE has had more interest in disease reduction, but their standards still focus on HVAC interventions and do not provide guidance on the use of germicidal UV.
Q: If ionizers produce ozone and have mixed evidence for their efficacy, why aren’t they facing the same market issues as far-UV?
A: This question keep me up. Far-UV manufacturers will say that it’s because ionizer companies are better at sales, have bigger backers, and are more willing to use dishonest sales tactics, but that only kicks the question back to why far-UV companies are worse at sales and lack these same large backers as ionizer companies. I have no direct experience of ionizer companies’ sales tactics, but my experience with far-UV sales is that, in general, salespeople communicate very differently from researchers in ways that can read as dishonest, and that the field of far-UV is not uniquely virtuous about communication style when business is on the line. I believe that the success of ionizers has more to do with their purported use for eliminating other pollutants (not just pathogens) from the air, so consumers view them as an all-purpose air cleaner. Ionizers also started being marketed for home use in the mid-1900s, before the health concerns of ozone inhalation became widely known, so it’s possible that ionizers were just a well-timed product, and that now their use might slowly die off.
Q: Do you have a wish list for filtration?
A: Filtration doesn’t face the same hurdles as far-UV in terms of the vicious cycle relationship between research, market adoption, and government investment, and because it has benefits aside from pathogen reduction. We only spend so much time talking about far-UV because it takes longer to explain as a technology and because the adoption problem is so complicated. The problems with filtration are already being addressed by other parties, who have much greater comparative advantage in this area, so filtration doesn’t “need” 1Day Sooner. My wish list for filtration, which is already well underway, is:
- Quieter filters: fan noise is one of the largest problems for filter use, because people get annoyed and turn them off. CleanAirKits already makes a great one, and I hope they conquer the air filter world.10
- Cheaper “smart” filters: high-quality filters that include useful features, like automatically adjusting fan speed in response to air quality conditions, are expensive. Smart features reduce the chance that someone will turn off the filter and forget to turn it back on. There are clear market incentives for smart filters to hit a price floor, and we already know a couple of people working on it.
- Better public understanding of HEPA filters’ benefit over ionizers: this is hard to tackle because it’s such a diffuse problem, but experts talk about it as much and as widely as they can, and government materials, informed by these researchers, already endorse HEPA filters over ionizers. Another route, which California may take soon, is banning ionizers outright.
I may add to/edit this list as more questions occur to me or as I learn more about the far-UV market. Thumbnail image from this biodefense meeting because I thought it was funny.
Personal communication with multiple researchers. I have not personally reviewed the literature—a quick glance shows research that seems to both prove and disprove the usefulness of ionizers, but generally I trust the researchers who I spoke with. As a history note, a coalition from the ionizer industry once sued a large group of air quality researchers over publishing research with negative outcomes.
Chamber studies can help address this type of uncertainty, but imperfectly—it’s hard to set up a chamber that imitates all the factors you’d want to consider, such as molecules’ diffusion or deposition. These factors might make far-UV look riskier than it really is, or they might mean that important considerations are missed.
Personal communication with a researcher who tried this avenue.
A study in nursing homes in Nova Scotia has been approved and is currently underway; the entire field is breathlessly awaiting results.
Yeah, this feels a bit like a distinction without a difference to me.
Corollary question that I get because of living in San Francisco amid the starry-eyed tech adventurers and entrepreneurs-by-temperament: If I think that having a reliable far-UV lamp is necessary, why don’t I go found a startup about it?
On the personal fit side, I’m just not that driven to found a startup. It’s not my comparative advantage and I don’t think I’d enjoy it that much—but who knows. As for the field itself, far-UV hasn’t made anyone any money yet; hence the existing companies being constantly on the brink of going out of business. Anyone getting into this field would have to just commit to eating a very large loss that I am not willing to contend with. This type of coordination problem definitionally can’t be solved by the market. (I definitely hope that this line of business becomes profitable; just describing the current state of affairs.) Plus, engineering in the physical world has very high overhead—if I wanted to work on the mechanics of far-UV, I’d just go work for the nonprofit working on that and increase their bandwidth and help them work faster, rather than start from scratch to figure out manufacturing best practices and whatnot. (Of course, this very much relies on my belief that the nonprofit in question is good at what they do—people in other industries may very well see many more opportunities to fix things that other people in the field are doing wrong.)
That all being said, I very much believe that the entire far-UV field shouldn’t rest on the shoulders of one or two manufacturers, so if you are a mechanical engineer digging around for new challenges and you’re not put off by everyone else’s failure to turn a profit, I’d be enthusiastic about your building a lamp and would do my best to help you get in touch with people you could learn from.
Personal communication with the group.
Personal communication with an industry sales rep. We were discussing a large but low-control study that some researchers were preparing, and he was pretty scared that it would show no effect. I was surprised to hear that he considered scientific literature so paramount to his business, because I figured that potential buyers never read the scientific literature anyway, but he told me that the decision-makers, like facilities managers, work with engineers who do read the literature, and that investors and research grantors do too, and null results make them disinclined to invest or fund higher-control research.
Why run low-control studies anyway? They can be much less expensive (depends on the setting) and they’re more reflective of real use!
Ironically, government research and a government standard would probably make everyone else in the private sector feel a lot better about far-UV! While navigating this field, it has felt like each group is waiting for the other to move first.
Corollary: it might be that one good use of philanthropic money would be to heavily subsidize and distribute ClearAirKits filters, or take out a massive ad campaign for them. My guess is that such a project wouldn’t meet an EA funder bar, but would be good for some major public health funder that likes local/immediate community work, like Chan Zuckerberg. I have not researched this carefully.
I would really recommend to look into pre-schools in the Nordics. They have high sickness rates and importantly: The government pays parents to stay home with sick kids. Even a 5% reduction in absence is worth millions and the government explicitly asks for solutions to this.
But there is more, anyone can set up a nursery, and the authorities track absence rates across pre-schools (I know, because kids who are immunocompromised get preference in pre-schools with the lowest absence rates). Setting up one's own pre-school is paid for by the state - they are called cooperatives. So one can literally possibly set up a pre-school running far-UV at almost no cost.
There is a challenge in the ethics of this: Is it ethical to do this? I have not checked but would be happy to either myself, or find someone even better positioned to figure out if this is actually possible. I know from talking to many parents that parents are extremely motivated to have their kids be sick less often. Another thing about the Nordics: It is quite evidence based and the freedom of choice for individuals has weight in government decisions - if parents are fully informed and still opt in, it might not be a problem.
This has changed! You can buy an Aerolamp for $500. I have one for my own use, and my dance organization uses four.
[EDIT: expanded this into a post]
@Gavriel Kleinwaks, do you back these?
I do! (Not in the financial sense, tbc.) But just want to flag that my endorsement is confounded. Basically, Aerolamp uses the design of the nonprofit referenced in my post, OSLUV, and most of my technical info about far-UV comes from a) Aerolamp cofounder Viv Belenky and b) OSLUV. I've been working with Viv and OSLUV for a couple of years, long before the founding of Aerolamp, and trust their information, but you should know that my professional opinion is highly correlated with theirs—1Day Sooner doesn't have the equipment to do independent testing.
I think it's the ideal outcome that a bunch of excellent researchers took a look at the state of the field and made their own product. So I'm not too worried about relying on this team's info, but you should just have that context.
Fwiw, Mox (moxsf.com), run by Austin Chen, has installed a couple of Aerolamps and they were easy to set up and are running smoothly.
Thanks so much for flagging this update!
Could/should big EA-ish coworking spaces like Constellation pay to have far-UV installed? (either on their floors specifically or for the whole building)
I think they should, and have spoken directly to Constellation about this, but it would probably help if people who actually use the space also express enthusiasm about the idea. My suggestion is to install far-UV in rooms that host gatherings where lots of people are regularly talking in relatively close quarters, like communal dining areas and rooms used for talks/presentations.
(I think it's a good idea not only because of direct disease reduction, but also to increase social proof/normalcy.)
Earlier this year year, I sent a Works in Progress piece about Far UV to a friend who is a material science engineer and happens to work in UV applications (he once described his job as "literally watching paint dry," e.g. checking how long it takes for a UV lamp to dry a coat of paint on your car). I asked
Friend replied:
You can/should take this with a grain of salt. On the other hand, I generally believe that some EAs tend to be very credulous towards solutions that promise a high amount of efficacy on core cause areas, and operate with a mental model that the missing ingredient for scale is more often money than culture/mechanics: that everything is like bed nets. By contrast, I believe that if some solution looks very promising to outsiders but remains in limited use -- e.g. kangaroo care for preemies in Nigeria or removing lead from turmeric in Bangladesh -- there is likely a deep reason for that that's only legible to insiders, i.e. people with a lot of local context. Here, I suspect that many of us don't have a strong understanding of the costs, limitations, true efficacy, and logistical difficulties of UV light.
That's my epistemology. But if someone wants to fund and run an RCT testing the effects of, say, a cluster of aerolamps on covid cases at a big public event, I'd be happy to consult on design, measurement strategy, IRB approval, etc. (Gotta put that university affiliation to use on something!)
I think this is true for many options, but not the Aerolamp:
Hi Jeff, I think we're talking about the same lifespan, my friend was talking about 1 year of continuous use (he works in industrial applications).
Thanks for clarifying! I do think in a context like this one, where people are thinking about why offices etc don't install far-UVC, your friend's phrasing is likely to confuse people. For example, if I recommended someone not buy a car because it only had a one-year lifespan, I think they'd be grumpy if they later learned I meant it would only last one year of 24/7 operation. When we talk about "lifespan" we're normally bringing in assumptions about expected usage.
Hey Seth, thanks for the comment. Very reasonable point on the importance of considering insider legibility/local context. I think far-UV doesn't fit the pattern you're describing, though, because the people writing about it are "insiders" and it's a very small field where everyone is talking and sharing information--there isn't an equivalent to local context that someone working in the field wouldn't quickly discover. Richard Williamson (who wrote the article you linked) works for Blueprint Biosecurity and is a leading expert in this field.
In the case of far-UV, I don't think there's a super deep reason for the lack of scale, and field insiders still debate the best path forward to scale up. I think what's going on is that it's at the "early adopter" tech stage and in the case of any early-commercial-stage tech, it's hard to say exactly what would speed up adoption and exactly why people aren't buying it faster. The most basic principle at play is inertia: it's always easier to not buy expensive, relatively unproven new tech than to buy it. Probably RCTs/a NIST standard/lower cost would help, but these things still wouldn't guarantee adoption.
ASHRAE has long had standards and working groups on UVC, and recently published standard 241 on Control of Infectious Aerosols. The goal is to reduce transmission risk, not to support any one particular technology. Filtration is usually cheaper than Far-UVC and easier to maintain for the same level of infection control. Far-UVC/UVC is better in some niches, particularly in healthcare settings that require high air flow rates.
I suggest getting involved in ASHRAE and the research community that has been working on and developing standards for infection control for over a century.
From an EA perspective, I think it is more effective to promote adoption of ASHRAE Std 241 than the adoption of Far-UVC specifically.
-perspective of a Ph.D. mechanical engineer and ASHRAE member with experience in IAQ
Do you think the general superiority of filtration over Far-UVC is likely inherent to the technologies involved, or would the balance be reasonably likely to change given further development of Far-UVC technologies? In other words, is it something like solar, which used to be rather expensive for the amount of output but improved dramatically with investment, economies of scale, and technological progress?
(Of course, we could improve filter technology as well, although it strikes my uninformed eyes as having less potential room to improve.)
Far-UVC produces ozone. That's inherent to the technology. That can be managed with ventilation, so places with already high ventilation rates where you don't want the added static pressure in the air supply from greater filtration are a good fit for UVC. In other places, in-room air cleaners tend to be cheaper to operate and maintain. The "best" technology depends on space constraints, ventilation rates, first cost, maintenance, etc. If far-UVC gets cheaper, I expect it will become more widely used. But I don't think it will fully dominate the space.
My position is the same as last year - I think there is enough technology at great enough maturity where it makes more sense to push for ASHRAE 241 adoption. Why doesn't your local grocery store have airborne infection control technology? Is it because lack of sufficient technology development or cost, or more likely, because they don't even know that's a thing they should do?
Thanks! I may be thinking about it too much from the consumer perspective of owning a condo in a 100-year-old building, where the noise of filtration is a major drawback and the costs of a broader modernization of HVAC systems would be considerable.
I haven't polled grocery store owners, but an owner would bear all the costs of improving air quality yet may capture few of the economic benefits. Although customers would care a lot in a pandemic, they probably wouldn't otherwise care in a way that increases profits -- and managers are incentivized toward short-term results. Cynically, most of their employees may not have paid sick time, so the owner may not even realize most of the benefit from reduced employee illness. (Of course, regulators could require compliance -- but that's not an awareness problem. So maybe the candidate intervention is lobbying?)
Just get a room air filter for your condo. There are different models and they are usually quite quiet unless on the highest setting. I can't hear mine on the two lowest settings (up to 100 CFM). UVC doesn't remove small particulates, which is the most serious air pollutant of concern from a health perspective for most homes.
Commercial buildings have to comply with locally required codes and standards. Code authorities could adopt some or all of ASHRAE Standard 241 in their jurisdiction just like they do with ASHRAE Standard 62.1 (ventilation) and 90.1 (energy).
I have colleagues in other organizations who are involved in ASHRAE; one contributed to the development of 241! 241 is great, and to be very clear, my job is not to promote far-UV, it is to promote indoor air cleaning. Far-UV gets a lot of attention in my social circles because it is exciting and new, and people have a lot of questions about that specifically--this post is meant for the very narrow case of answering questions I've directly gotten about far-UV.
If this describes you, I'd also love to help (eg with funding) -- reach out to me at austin@manifund.org!