UV-C germicidal irradiation (UVGI) is effective in HVAC systems, but only when the dose is engineered rather than assumed. Inactivating 90% of common airborne bacteria and viruses requires roughly 2,000–10,000 µJ/cm² at 254 nm, and air travelling at 2.5 m/s spends only about 0.4 s inside a 1 m irradiated duct section. The lamp bank therefore has to deliver several thousand µW/cm² of average irradiance to matter. Coil irradiation is the easier win, because biofilm on a wet cooling coil receives continuous exposure for thousands of hours. Safety is non-negotiable: 254 nm causes photokeratitis and erythema within minutes of direct exposure.
Dose is the only variable that counts
UVGI reduces to two equations. Dose equals irradiance multiplied by exposure time, D = E × t, and survival follows S = exp(−k × D), where k is the organism-specific susceptibility constant in cm²/µJ. Typical k values for vegetative bacteria and enveloped viruses in air fall between 0.0003 and 0.001 cm²/µJ, so the dose for one log reduction (D90 = 2.3/k) lands between about 2,300 and 7,700 µJ/cm². Fungal spores are 5–10 times more resistant, and bacterial spores more resistant still.
- Enveloped viruses in aerosol, including influenza A and coronaviruses: D90 typically 1,500–4,000 µJ/cm².
- Common vegetative bacteria such as E. coli and Serratia: D90 typically 3,000–7,000 µJ/cm².
- Mycobacterium tuberculosis in aerosol: D90 typically 5,000–10,000 µJ/cm².
- Mould spores such as Aspergillus and Penicillium: D90 typically 20,000–60,000 µJ/cm².
- Bacterial spores such as Bacillus subtilis: D90 typically 20,000–80,000 µJ/cm².
Work the arithmetic backwards to size a system. To achieve 4,000 µJ/cm² in 0.4 s of transit, the average irradiance across the duct must be about 10,000 µW/cm², which is a substantial lamp bank in a duct of any size. Halve the velocity to 1.25 m/s, or double the irradiated length to 2 m, and the required irradiance halves. That trade between duct length, velocity and installed lamp watts is the entire design exercise.
Three ways to apply UV-C in an air system
| Application | Coil and drain pan irradiation | In-duct airstream disinfection | Upper-room air irradiation |
|---|---|---|---|
| Primary target | Surface biofilm and mould | Airborne organisms in one pass | Airborne organisms in occupied room |
| Exposure time | Continuous, thousands of hours | 0.1–0.5 s per pass | Minutes, via convection mixing |
| Typical installed power | 25–60 W per m² of coil face | 1.5–4 W per 100 m³/h of airflow | 15–30 W per 10 m² of floor area |
| Realistic performance | Coil pressure drop restored, 10–25% latent capacity recovery | Typically 60–90% single-pass inactivation | Equivalent to 10–20 extra air changes per hour |
| Main constraint | Lamp output drops in cold, fast air | Duct length and fan energy | Strict eye-level shielding required |
| Cost driver | Lamp count | Lamp count and duct modification | Fixture optics and commissioning survey |
When to specify UV-C and when to spend the budget elsewhere
- Specify coil irradiation when coils foul repeatedly, drain pans grow slime, or a chilled-water coil operating at 6–12 °C stays wet for most of the year.
- Specify coil irradiation when restoring coil cleanliness measurably cuts fan power; a fouled coil can add 40–120 Pa of avoidable resistance.
- Specify in-duct airstream UVGI when the application involves recirculated air among vulnerable occupants – healthcare waiting areas, aged care, transit hubs – and there is at least 0.9–1.5 m of straight duct available.
- Specify upper-room UVGI when the risk is person-to-person transmission in a high-occupancy room and outdoor air cannot be increased without unacceptable cooling load.
- Spend on filtration instead when the target is particulate, dust or PM2.5, because UV-C removes nothing; it only inactivates biological material.
- Spend on ventilation instead when outdoor airflow per person is below code minimums, since dilution addresses CO&sub2;, odour and bioaerosols together.
- Avoid in-duct UVGI when duct velocity exceeds about 4 m/s with no room to extend the irradiated length, because the achievable dose becomes negligible.
- Avoid UV-C entirely when maintenance staff cannot be trained on lockout procedure and annual lamp replacement.
Installation details that decide whether it works
- Design for end-of-life output. Low-pressure mercury lamps last 9,000–17,000 operating hours but depreciate 15–30% in the first year, so size on maintained output and replace annually in continuous-duty systems.
- Account for air temperature and velocity. Standard lamps peak near 40 °C wall temperature; in a 10–13 °C airstream at 2.5 m/s, output can fall 30–40%. Use high-output or amalgam lamps in cold, fast air.
- Line the irradiated section with specular aluminium or polished stainless steel. Reflective surfaces commonly raise average irradiance by 30–50% for no additional power.
- Specify ozone-free lamps with doped quartz envelopes that block the 185 nm line, unless ozone generation is deliberately intended in an unoccupied process.
- Check material compatibility. Direct UV-C degrades EPDM, PVC, polyurethane and nylon within months; use PTFE-jacketed wiring, silicone gaskets, aluminium tape and stainless fixings inside the irradiated zone.
- Provide a UV-blocking viewport, door interlock switches on every access panel, and permanent warning labels at each access point.
Keruite builds air handling units and ventilation equipment at an 80,000 m² plant in Dezhou, China, with an in-house AMCA-210-style test laboratory, and can factory-fit UV-C coil irradiation sections with interlocked access panels so the safety chain is verified before shipment.
Occupational safety requirements
The recognised occupational exposure limit for 254 nm radiation is approximately 6 mJ/cm² over an eight-hour period, which a person standing in a direct beam can accumulate in well under a minute. Practical control therefore relies on enclosure and interlocks rather than administrative rules. Fit positive-break interlock switches that de-energise the lamps when any access door opens, provide UV-safe polycarbonate or glass viewports for observation, and give technicians UV-rated face shields and gloves for lamp changes. Remember that lamps contain 5–15 mg of mercury each and must be handled as regulated waste. Finally, record the lamp installation date on the unit and log annual replacements, because an aged lamp that still glows visibly may be delivering only half its rated germicidal output.
Commissioning and verification
Verify the installation with a calibrated 254 nm radiometer rather than trusting the catalogue figure. Take irradiance readings at three or four positions across the duct cross-section, note the airflow and air temperature at the time of measurement, and record the values as the maintained baseline. If the measured average is more than 20% below the design figure, the usual causes are cold-air derating, a missing reflective liner, or lamps installed across the airflow instead of parallel to it. Repeat the survey after every lamp change so drift is caught early.
FAQ
Does UV-C in a duct actually kill viruses at normal air speeds?
It inactivates a meaningful fraction, typically 60–90% per pass, when the bank is sized for the dose. At 2.5 m/s the air is exposed for only about 0.4 s, so achieving 4,000 µJ/cm² needs roughly 10,000 µW/cm² average irradiance, reflective duct lining and adequate installed watts.
How often should UV-C lamps be replaced?
Annually for systems running continuously, even though rated life is 9,000–17,000 hours. Germicidal output declines 15–30% within the first year while visible glow persists, so lamps look functional long after the dose has dropped below design. Log installation dates on the unit.
Is UV-C better than a MERV 13 filter?
They solve different problems. Filters remove particulate including PM2.5 and dust; UV-C only inactivates biological material and removes no mass. For bioaerosol risk they complement each other, and a common arrangement is MERV 13 filtration plus coil irradiation to control fouling.
Can UV-C damage HVAC components?
Yes. Direct exposure embrittles EPDM gaskets, PVC drain lines, polyurethane insulation facings and nylon cable ties, sometimes within 6–12 months. Inside the irradiated zone, use PTFE-jacketed wiring, silicone or metal gaskets, aluminium foil tape and stainless steel fasteners.