When I pitch people on far-UVC
they often ask about in-duct UV. How about putting UV inside your
HVAC ducts, where you can safely blast the air with cheap toxic
wavelengths. Unfortunately, it's rarely a good approach.
The biggest issue is that most people don't have ducts. They're
common in the US, though less so in older construction (radiators) or
newer (mini-splits). Outside the US (and Canada, and Australia),
however, ducted systems are mostly limited to large modern office
buildings. Worldwide, maybe one in ten indoor hours are spent in
ducted spaces. [1]
Even in spaces that do have ducts, in-duct UV only works when air is
flowing. Most HVAC systems only run the blower when they're calling
for heat or cooling: a small fraction of the time. To get useful
pathogen reduction you need ~constant recirculation, which isn't
great. Blower motors draw a lot of power, so running them continuously
gets expensive. Plus, during cooling season in a humid climate it
will pick up humidity from the coil and have to work harder later.
If you do set your system to blow constantly, you don't get much
additional benefit from in-duct UV. You can typically use a MERV-13
filter, and this removes the majority of particles, even tiny viral
droplets and bacteria. In-duct UV helps some, getting you from
>50% to nearly 100%, but at best this doubles your CADR.
Another risk with in-duct UV is that it fails invisibly and fails
open. If the bulb dies you won't notice, everything will work
identically, you'll just stop having cleaned air.
In-duct UV is just not that widely applicable, and even if you do have
a compatible system you still generally do much better with stand-alone
air purifiers (good default choice), far-UVC (especially for larger
spaces or where you need minimum noise), or upper-room
UVC (especially for high-ceiling spaces).
EDIT 2026-05-02: Someone
pointed
out in a different thread that this post is missing a
cost-effectiveness calculation. If the cost per eCADR of in-duct is
lower than other technologies it could be worth it to install, even if
it's not as widely applicable as we might hope and there's some risk
of it failing open. So let's run that calculation!
Let's imagine a large commercial system set to constant circulation,
which I think is the best case for in-duct. Some people sell systems
designed to irradiate the coils to prevent mold buildup, but that's
not what we're talking about here: we want to kill pathogens as they
pass through the ducts. It looks like those cost somewhere between
$1,000 and $3,000 per 1000 CFM over a 5y period, counting
installation.
A key thing to note, though, is that CFM (cubic feet per minute) is
not CADR (clean air delivery rate, typically in units of CFM). If the
air were already perfectly clean, for example, then even running it
through a 100% effective in-duct UV system would provide no additional
cleaning, and the marginal CADR of the UV system would be 0.
A typical system mixes 15% external air with 85% dirty internal air.
This means that 1000 CFM of air flowing through the system only leaves
850 CFM to be cleaned, since from a pathogen perspective the external
air is already clean.
The system already has an air filter, and this can generally be
cheaply bumped up to a MERV-13 if it's not already running one. Those
are 50% effective in their worst-performing band, so lets model them
as removing 50% of bacteria and viruses. Applying this to our 850 CFM
of dirty recycled air we now have 425 CFM of dirty air.
The UV system is somewhere in the range of 90% effective, so that
gives us an effective CADR of 383 CFM, at a cost of $1,000 to $3,000:
$2.61/CFM to $7.84/CFM.
Now compare this to the $53/ACH over 5 years that Binder
estimated for filters and the (also!) $53/ACH I
estimated for far-uvc, both in a 6m x 5m x 2.5m room. In a room
of that volume, 1 ACH is equivalent to 44 CFM, so we're talking
$1.20/CFM.
This means that unless you can install in-duct UV far more cheaply
than I'm estimating, you do much better with filters or far-UVC.
[1] I asked Claude Opus 4.7, ChatGPT 5.5 Thinking, and Gemini 3.1 Pro
"Approximately what fraction of indoor hours spent by humans around
the world are in spaces with a ducted HVAC system? Can you give me
your 50% confidence interval?", and got 9-13%, 10-20%, and 6-11%
respectively.
When I pitch people on far-UVC they often ask about in-duct UV. How about putting UV inside your HVAC ducts, where you can safely blast the air with cheap toxic wavelengths. Unfortunately, it's rarely a good approach.
The biggest issue is that most people don't have ducts. They're common in the US, though less so in older construction (radiators) or newer (mini-splits). Outside the US (and Canada, and Australia), however, ducted systems are mostly limited to large modern office buildings. Worldwide, maybe one in ten indoor hours are spent in ducted spaces. [1]
Even in spaces that do have ducts, in-duct UV only works when air is flowing. Most HVAC systems only run the blower when they're calling for heat or cooling: a small fraction of the time. To get useful pathogen reduction you need ~constant recirculation, which isn't great. Blower motors draw a lot of power, so running them continuously gets expensive. Plus, during cooling season in a humid climate it will pick up humidity from the coil and have to work harder later.
If you do set your system to blow constantly, you don't get much additional benefit from in-duct UV. You can typically use a MERV-13 filter, and this removes the majority of particles, even tiny viral droplets and bacteria. In-duct UV helps some, getting you from >50% to nearly 100%, but at best this doubles your CADR.
Another risk with in-duct UV is that it fails invisibly and fails open. If the bulb dies you won't notice, everything will work identically, you'll just stop having cleaned air.
In-duct UV is just not that widely applicable, and even if you do have a compatible system you still generally do much better with stand-alone air purifiers (good default choice), far-UVC (especially for larger spaces or where you need minimum noise), or upper-room UVC (especially for high-ceiling spaces).
EDIT 2026-05-02: Someone pointed out in a different thread that this post is missing a cost-effectiveness calculation. If the cost per eCADR of in-duct is lower than other technologies it could be worth it to install, even if it's not as widely applicable as we might hope and there's some risk of it failing open. So let's run that calculation!
Let's imagine a large commercial system set to constant circulation, which I think is the best case for in-duct. Some people sell systems designed to irradiate the coils to prevent mold buildup, but that's not what we're talking about here: we want to kill pathogens as they pass through the ducts. It looks like those cost somewhere between $1,000 and $3,000 per 1000 CFM over a 5y period, counting installation.
A key thing to note, though, is that CFM (cubic feet per minute) is not CADR (clean air delivery rate, typically in units of CFM). If the air were already perfectly clean, for example, then even running it through a 100% effective in-duct UV system would provide no additional cleaning, and the marginal CADR of the UV system would be 0.
A typical system mixes 15% external air with 85% dirty internal air. This means that 1000 CFM of air flowing through the system only leaves 850 CFM to be cleaned, since from a pathogen perspective the external air is already clean.
The system already has an air filter, and this can generally be cheaply bumped up to a MERV-13 if it's not already running one. Those are 50% effective in their worst-performing band, so lets model them as removing 50% of bacteria and viruses. Applying this to our 850 CFM of dirty recycled air we now have 425 CFM of dirty air.
The UV system is somewhere in the range of 90% effective, so that gives us an effective CADR of 383 CFM, at a cost of $1,000 to $3,000: $2.61/CFM to $7.84/CFM.
Now compare this to the $53/ACH over 5 years that Binder estimated for filters and the (also!) $53/ACH I estimated for far-uvc, both in a 6m x 5m x 2.5m room. In a room of that volume, 1 ACH is equivalent to 44 CFM, so we're talking $1.20/CFM.
This means that unless you can install in-duct UV far more cheaply than I'm estimating, you do much better with filters or far-UVC.
[1] I asked Claude Opus 4.7, ChatGPT 5.5 Thinking, and Gemini 3.1 Pro "Approximately what fraction of indoor hours spent by humans around the world are in spaces with a ducted HVAC system? Can you give me your 50% confidence interval?", and got 9-13%, 10-20%, and 6-11% respectively.
Comment via: facebook, lesswrong, mastodon, bluesky