UV dose: why flow rate matters more than watt rating

The science of UV inactivation is built around one central number: dose, measured in millijoules per square centimetre (mJ/cm²). Every target organism — algae, bacteria, protozoan cysts, viruses — has a known dose requirement for a given level of inactivation. Get below that number and UV barely touches it. Exceed it and you achieve reliable kill rates. Wattage determines the intensity of the lamp; dose is what the water actually receives.

What UV dose actually means

UV dose — also called fluence — is the product of two things: the intensity of the UV light passing through the water, and the amount of time the water spends in the UV chamber. Written as a simple relationship:

Dose (mJ/cm²) = UV intensity × exposure time

UV intensity is fixed by the lamp — a 9W low-pressure UV lamp emits a characteristic amount of 254 nm radiation. You cannot easily change it. But exposure time is entirely controlled by your flow rate: the slower water moves through the UV chamber, the longer each litre spends in the UV field, and the higher the dose it receives.

This means flow rate is the most powerful variable in the whole system. A flow rate that is twice as fast delivers half the dose. A flow rate that is half as fast delivers twice the dose. The lamp's wattage is the same in both cases.

How much dose do different targets actually need?

The peer-reviewed literature — particularly the comprehensive review by Hijnen et al. (2006) in Water Research — provides dose requirements for a wide range of organisms. Translated into aquarium terms:

Target	Dose for 1-log (90%)	Dose for 3-log (99.9%)
Green water algae (Chlorella)	~10 mJ/cm²	~30 mJ/cm²
Cryptosporidium oocysts	<7 mJ/cm²	<20 mJ/cm²
Giardia cysts	<7 mJ/cm²	<20 mJ/cm²
Typical aquarium bacteria (E. coli equivalent)	3–9 mJ/cm²	6–9 mJ/cm²
Ich theronts (free-swimming) †	No peer-reviewed dose-response data †	No peer-reviewed dose-response data †
Adenovirus	>50 mJ/cm²	>150 mJ/cm²

Bacteria, protozoa, and virus rows: data from Hijnen, Beerendonk & Medema (2006), Water Research 40(1):3–22, the primary peer-reviewed synthesis of UV dose-response data across 120+ studies covering human and animal waterborne pathogens. Green water algae dose figures are based on hobbyist and aquaculture data rather than the Hijnen pathogen review, which does not cover algae species. † Ichthyophthirius multifiliis theronts are not covered by the Hijnen review, which focuses on human and animal water treatment pathogens. No controlled collimated-beam dose-response study for Ich theronts has been published in the primary peer-reviewed literature. The commonly cited applied aquaculture recommendation of 100 mJ/cm² originates from extension literature, not a peer-reviewed dose-response trial. Gratzek et al. (1983, Journal of Fish Diseases) demonstrated UV control of Ich in a closed recirculating system at the system level — showing UV reduced Ich spread — but did not measure individual theront inactivation doses.

The table shows why UV works so well for green water but gives mixed results for general disease prevention. Clearing green water only requires 20–30 mJ/cm² — achievable by most units at moderate flow. Achieving 3-log inactivation of common bacteria requires around 6–9 mJ/cm², while protozoan parasites and harder targets may need up to 20–50 mJ/cm² — achievable only at lower flow rates. And viruses like Adenovirus are practically outside the range of any hobbyist UV unit.

How to think about your unit's actual dose

You cannot directly measure the dose your UV unit delivers without specialist equipment. But you can reason about it from the basic relationship between flow rate and exposure time.

Most hobbyist in-line UV units (the type plumbed into your filter return) have a UV chamber volume of roughly 0.2–0.5 litres. At a flow rate of 500 L/h, water spends about 1.4–3.6 seconds in the chamber. At 200 L/h, that rises to 3.6–9 seconds. A lamp that delivers a given intensity has roughly 2.5 times more opportunity to irradiate each litre at the lower flow rate.

A practical rule of thumb that follows from the science:

• For green water only: the manufacturer's recommended maximum flow is probably fine.
• For reducing bacterial pathogens: run at 50–60% of the recommended maximum flow.
• For protozoan control (Ich, Velvet): run at 30–50% of the recommended maximum flow, accepting reduced turnover rate.
• For reliable virus inactivation: not achievable with typical hobbyist equipment regardless of flow rate.

The turbidity problem — green water defeats itself

UV radiation at 254 nm is attenuated by anything dissolved or suspended in water. Tannins, dissolved organic compounds, suspended particles, and — crucially — the algae cells of a green water bloom all absorb or scatter UV light before it reaches other organisms.

This creates a self-limiting problem: the denser a green water bloom becomes, the harder it is for UV to penetrate far enough to inactivate all the cells. The cells near the lamp are inactivated quickly; the cells at the far side of the chamber may receive a much lower dose. A severe green water bloom will clear more slowly than a mild one, not because the UV is less effective against the algae, but because the algae itself is blocking the UV from reaching everything.

The same attenuation applies in tanks with high tannin content (heavily botanicals-dosed setups), very high dissolved organic loads, or any form of cloudy water. UV is most effective in clear, low-tannin water.

Does higher turnover compensate for lower dose per pass?

A natural follow-on question: if a single pass through the UV chamber doesn't deliver enough dose to reliably inactivate your target, can you compensate by running the pump faster — pushing 4 or 8 passes per hour through the unit rather than 1 or 2?

The answer requires understanding a trade-off you cannot avoid.

The constraint: flow controls both variables simultaneously

Flow rate determines both how many passes the water makes per hour and how long each litre spends in the UV chamber. These work in opposite directions. Pushing water through faster gives you more passes per hour — but it also reduces the dose per pass, because each litre has less exposure time in the UV field:

• Double the flow rate → twice the passes per hour, half the dose per pass
• Halve the flow rate → half the passes per hour, twice the dose per pass

This is not an equipment limitation. It is a direct consequence of the physics: Dose = UV intensity × exposure time. For a fixed lamp in a fixed chamber, those two variables are coupled to flow rate.

What the mathematics of a recirculating system actually predicts

For organisms that inactivate following first-order kinetics — where each increment of dose produces a proportional log reduction, without a minimum threshold — the mathematics of a well-mixed recirculating system produces a specific prediction. This analysis applies the CSTR (continuous stirred-tank reactor) model to a UV side-loop, following the approach published by Zhu, Saucier, Chen & Durfey (2002) in the International Journal of Recirculating Aquaculture, which developed and validated this model for UV disinfection in recirculating aquaculture systems.

The practical implication: for organisms following first-order kinetics, simply running the UV faster does not accelerate clearance unless you also increase lamp power. You are redistributing the same UV energy across more shorter exposures, and the net effect on the water column is similar.

Where more passes genuinely help

Higher turnover does make a real difference in two specific situations.

1. Continuous sources during an active outbreak. When Ich theronts or bacteria are being continuously released into the tank — from infected fish, from substrate — more passes per hour means each newly released organism encounters the UV unit sooner after entering the water column. The Gratzek et al. (1983) recirculating system study demonstrated this practical benefit: continuous UV exposure reduced Ich transmission even in a system where flow rate, not optimal per-pass dose, was the operating parameter. In an active outbreak, reducing the average time-to-UV-encounter for a freshly released theront has genuine value, even if the per-pass dose is modest. [This framing is our analytical interpretation of the Gratzek et al. findings, not a direct statement from that paper.]

2. Protozoa: no dose threshold, so every pass counts. The peer-reviewed literature on UV kinetics consistently shows that protozoa inactivate without a minimum threshold dose — they follow first-order kinetics from the very first mJ/cm². This means even a low dose per pass genuinely inactivates some fraction of theronts rather than contributing nothing. Multiple low-dose passes do accumulate meaningful effect for this class of organism. The same cannot be said for organisms like Adenovirus, which require a very high dose before significant inactivation begins — no realistic number of passes at hobbyist UV levels will reach it.

Photoreactivation: a complication specific to bacteria

For bacteria, there is an additional consideration. Liltved & Landfald (1996, Water Research) demonstrated that fish-pathogenic bacteria — including Aeromonas salmonicida — can repair UV-induced DNA damage through photoreactivation (triggered by visible and near-UV light) after leaving the UV chamber. Their research found effective inactivation doses needed to be 3–4 times higher than single-pass laboratory figures to account for this repair. Multiple low-dose passes followed by bright-light exposure in a planted tank may see more repair between passes than a single high-dose pass would. For bacteria, higher dose per pass is preferable to more lower-dose passes — another argument for running at lower flow than the manufacturer maximum when bacterial reduction is the goal.

The professional approach: separate the circuits

Recirculating Aquaculture Systems — where UV sterilisation is extensively engineered and studied — resolve this trade-off by running the UV unit on a dedicated pump circuit at a low, carefully chosen flow rate, entirely separate from the main recirculating flow that handles biological filtration and tank circulation. This gives maximum dose per pass on the UV circuit while maintaining full turnover on the biological side.

For a home aquarium, this approach is worth considering if you are running UV specifically for biosecurity — for instance in a quarantine tank where per-pass dose matters most. A small powerhead running at 100–200 L/h through the UV steriliser, while your main filter return runs at its optimal biological flow rate, gives you higher dose per pass and adequate passes per hour without compromising the filter. [This is our recommendation based on the engineering principles above, not a hobbyist-specific recommendation from the peer-reviewed literature.]

For most everyday planted tank use cases — clearing green water, managing an Ich outbreak — the single-circuit approach with appropriate flow reduction (see the rules above) remains practical and effective.

Your lamp's output is falling

All UV lamps — both low-pressure (the most common type in hobbyist units) and medium-pressure lamps — degrade in output over time. The UV-C output of a standard low-pressure mercury lamp typically falls to about 65% of its initial value after 8,000–9,000 hours of operation. That is roughly 11–12 months of continuous use.

A lamp running at 65% output delivers 65% of the dose it did when new — all else being equal. A unit that was marginal at the right flow rate when the lamp was new may deliver a meaningfully sub-threshold dose by the end of its first year. Most manufacturers recommend annual lamp replacement regardless of whether the lamp is still producing visible light (it will be — lamps lose UV output long before they stop producing visible glow).

If you are using a UV unit specifically to manage a recurring problem (green water, elevated pathogen risk), note when the lamp was installed and replace it at the 12-month mark.

A practical sizing approach

Rather than matching UV wattage to tank volume (the common but misleading advice), think about it this way:

1. Decide what you're trying to achieve — green water only, or general pathogen reduction?
2. Choose a flow rate that delivers the dose you need — this will be lower than the manufacturer maximum for anything beyond green water.
3. Check the turnover rate that results — your tank still needs adequate overall flow for gas exchange and biological filtration. See the guide to flow in a planted tank for what adequate coverage actually looks like. If running the UV at a low enough flow rate reduces overall tank turnover too much, either run a separate UV pump circuit at low flow (independent of your main filter return), or accept that UV is best suited to targeted use cases (clearing a bloom, managing an outbreak) rather than continuous 24/7 operation.
4. Replace the lamp annually regardless of visible output.

For the science of what UV actually inactivates and why some things pass through unaffected, see the companion article on UV sterilisers: what they actually kill. The bacteria in your biological filter — which UV deliberately does not reach — are covered in the guide to comammox and the nitrogen cycle.