UV or Chlorine? How to Choose the Right Disinfection Barrier for Drinking WaterDose, UV transmittance, contact time, by-products and residual protection - practical selection notes from the HONCO engineering team

2026-11-16

Left: HONCO UV reactor. Right: packaged dosing skid. Both are product photographs from the HONCO catalogues, not renders.

Disinfection looks like a simple purchase: add a lamp chamber or a dosing pump and the water is safe. In practice, most disinfection failures we see in the field are not equipment failures. They are design failures - the barrier was placed where it cannot work, or it was sized without the numbers that decide whether it still works after a year of operation.

Two questions decide the design. First, what has to be killed, and how tolerant is it to the disinfectant used? Second, does the water have to stay protected after it leaves the plant - in a storage tank, a rising main, or a building's internal network? Ultraviolet light and chlorine answer those questions in very different ways, and the answer that suits a household under-sink unit is rarely the answer for a municipal network.

1. Key takeaways

Two jobs, not one. Primary disinfection inactivates pathogens at the point of treatment; residual protection stops them regrowing downstream. UV does the first job only. Chlorine does both, but does the first one less well.

UV is the stronger barrier against chlorine-tolerant protozoa. Cryptosporidium and Giardia survive chlorination even at high CT, while UV inactivates them at a comparatively low dose.

UV leaves no taste and no by-products, but also no protection. A UV-only system feeding a storage tank or a long network is unprotected the moment the water leaves the lamp chamber.

Chlorine is the only common disinfectant that travels. It is also the one that forms trihalomethanes and haloacetic acids on contact with natural organic matter, so its dose has to be controlled rather than maximised.

Most municipal and building projects need both. UV as the primary barrier and low-dose chlorine for residual. Household point-of-use systems usually need UV alone - provided turbidity and organics have already been removed upstream.

2. What the two barriers actually do

UV light at 254 nm damages the nucleic acid of bacteria, viruses and protozoan oocysts so that they can no longer replicate. It is a physical, photochemical process: nothing is added to the water, nothing is removed from it, and nothing remains afterwards. A low-pressure mercury lamp puts roughly 90 percent of its radiant output at 254 nm, close to the absorption peak of DNA. Dose is the product of intensity and contact time, expressed in millijoules per square centimetre (mJ/cm2).

Because UV works through light, anything that absorbs or scatters 254 nm light reduces the dose the water actually receives. Turbidity, iron, manganese, humic matter and dissolved organics all compete with pathogens for photons. That is why UV transmittance (UVT) at 254 nm is a more important number in a UV design than the flow rate printed on the lamp label.

Chlorine works by oxidation. Hypochlorous acid (HOCl) is the effective species; it dissociates to hypochlorite ion (OCl-) with a pKa near 7.5, so at pH 7.5 roughly half of the free chlorine is in the weaker form, and at pH 8.5 most of it is. Chlorine performance is therefore described by CT - concentration multiplied by contact time - a product that is highly sensitive to pH and temperature. Its great advantage is persistence: a measurable residual can travel kilometres of main and still be present at the tap.

3. Side-by-side comparison

Item UV disinfection Chlorination
Primary mechanism Photochemical damage to nucleic acid Oxidation of cell membranes and enzymes
Cryptosporidium / Giardia Effective at low dose Poor; not a reliable barrier
Residual protection None Yes, measurable at the tap
Disinfection by-products Not formed Trihalomethanes and haloacetic acids with NOM
Feedwater sensitivity Turbidity, UVT, iron and manganese pH, temperature, ammonia (chloramine formation)
Contact time needed Seconds Typically 20-30 minutes for primary credit
Taste and odour impact None Chlorinous taste above about 0.5 mg/L free chlorine
Operating cost profile Power, lamp replacement, sleeve cleaning Chemical, dosing control, by-product monitoring
Typical failure mode Silent: aged lamps or fouled sleeves Visible: over-dosing or under-dosing
Best fit Point of use, direct drinking, secondary supply Municipal networks, storage, long distribution

This table is not a ranking. A UV system with a fouled quartz sleeve can deliver water that looks perfect and is not disinfected; a chlorinated network can meet its residual target and still carry Cryptosporidium. Both failures are invisible without measurement.

4. Sizing a UV barrier: three factors you cannot skip

A UV unit is usually selected on flow rate, and that is where most under-sized installations begin. Flow rate tells you how much water passes; it does not tell you how much light reaches the water. Three correction factors sit between the nominal lamp rating and the dose the water actually receives.

Transmittance. Ask for a measured UVT at 254 nm on a real sample, not a laboratory assumption. Drinking water is often above 90 percent UVT; filtered surface water, rainwater and reclaimed water can sit far lower, and a ten-point drop in UVT can mean a large increase in required lamp power.

Lamp ageing. Low-pressure lamps are typically rated for 8,000 to 12,000 hours. Output falls gradually over that life, so the reactor must be sized at end-of-lamp-life, not on day one. Designing to 60 to 80 percent of initial output is normal practice.

Sleeve fouling. The quartz sleeve that separates lamp from water accumulates scale and iron deposits. A fouling factor of 0.8 is a common allowance, 0.7 for harder or iron-bearing water. The design dose is the target dose divided by the product of all three factors - which is why a reactor quoted at 40 mJ/cm2 under nominal conditions may deliver far less in its second year.

5. Chlorination: CT, residuals and by-products

Chlorine design starts with CT. A worked example: a target CT of 15 mg-min/L can be met with 0.5 mg/L free chlorine and 30 minutes of contact, or with 1.0 mg/L and 15 minutes, provided pH and temperature are favourable and short-circuiting is controlled. Baffling matters as much as tank volume, because a poorly baffled tank delivers only a fraction of its nominal contact time.

Ammonia changes the picture completely. Free chlorine reacts with ammonia to form chloramines, which are much weaker disinfectants but far longer-lived. Many networks use that deliberately to maintain a residual; it is a poor choice for primary disinfection.

By-product control is a dosing problem, not a treatment problem. Trihalomethanes and haloacetic acids form from the reaction between chlorine and natural organic matter, so the effective controls are upstream precursor removal - coagulation, filtration, activated carbon - and keeping the dose no higher than the residual target actually requires.

Residual management closes the loop. Many utilities aim for 0.2 to 0.5 mg/L free chlorine in distribution: enough to protect, low enough to avoid taste complaints. Above roughly 0.5 mg/L, chlorinous taste becomes noticeable to most consumers.

6. The hybrid design most projects should use

For municipal supply, building water systems and any scheme with storage or a long network, the most robust arrangement is a sequence rather than a choice: pretreat to bring turbidity and organics down, disinfect with UV for the pathogens chlorine cannot handle, then add a small chlorine dose for residual.

The order matters. UV placed after filtration sees lower turbidity and higher UVT, so it needs less power. Chlorine placed after UV is not competing with organics for the same photons. And the residual dose can be small precisely because it no longer carries the whole disinfection burden.

For household under-sink systems, direct drinking water units and bottling lines, UV alone is usually right: the water has already passed through pre-filters and a membrane, it is consumed within hours, and there is no distribution network to protect. Chlorine at that scale buys no extra safety and costs taste.

7. Maintenance that keeps the barrier real

A disinfection barrier is only as good as its monitoring. The following items decide whether the system is still doing its job twelve months after commissioning.

  • Lamp hours and output. Log run hours, replace at the manufacturer's end-of-life point, and keep a spare set on site. Lamp-on is not the same as lamp-working.
  • Quartz sleeve cleaning. Inspect on a fixed interval - quarterly in hard or iron-bearing water - and clean before the UV sensor reading drifts, not after.
  • UV intensity sensor calibration. The sensor is the only evidence that the dose is being delivered; an uncalibrated sensor reports comfort, not reality.
  • Free chlorine and pH testing. Daily at the plant, weekly at the furthest tap. Chlorine demand changes with season and raw water quality.
  • By-product monitoring. Quarterly trihalomethane and haloacetic acid testing wherever surface water or high-organic sources are used.
  • Turbidity ahead of the UV chamber. A sudden rise means the UVT assumption behind the design no longer holds.

8. Common misconceptions

UV removes scale and heavy metals. It does not. UV changes nothing chemically; if hardness, iron or nitrate is the problem, the answer is softening, oxidation-filtration or reverse osmosis, not a stronger lamp.

A lit lamp means disinfected water. Not necessarily. A lamp at the end of its life behind a fouled sleeve still produces visible light and an inadequate dose.

A strong chlorine smell means good disinfection. Usually the opposite: it means the dose is high relative to the organic load, or that chloramines are forming in the way that produces the familiar swimming-pool odour.

UV is expensive to run. For clean water it is among the cheapest barriers per cubic metre; the cost rises sharply only when UVT is low and the reactor has to be oversized.

One disinfectant covers every project. Almost never. The right answer depends on what is in the water, where the water goes after treatment, and how it is measured.

HONCO supplies UV reactors, dosing systems and the pretreatment that makes both work - filtration, activated carbon, softening and reverse osmosis - for household, commercial and municipal projects.

Send your flow rate, water analysis and target standard, and we will return a disinfection scheme with the dose calculation shown.

Get the latest price? We will reply as soon as possible (within 12 hours)