---
title: "Ozone for Legionella and water disinfection: how it works, where it fits and where it doesn't"
source_url: https://legionella.io/articles/ozone-water-treatment-legionella/
canonical_url: https://legionella.io/articles/ozone-water-treatment-legionella/
pillar: "Best Practice & Future of Legionella Control"
summary: "Ozone kills Legionella and biofilm on contact but leaves no residual to reach your taps. Here's its niche - cooling-tower side-streams - and where it fails."
primary_keyword: "ozone water treatment Legionella"
date_published: 2026-06-22
date_reviewed: 2026-06-26
author: "Legionella.io editorial team (REMOTE TECH LTD)"
reviewed_against: "HSE L8 and HSG274 guidance"
region: "United Kingdom"
license: "(c) REMOTE TECH LTD. Quote freely with attribution and a link to source_url."
---

# Ozone for Legionella and water disinfection: how it works, where it fits and where it doesn't

Ozone is the strongest common water disinfectant - an unstable gas, generated on the spot, that oxidises Legionella, biofilm and organic load on contact. It also decays back to oxygen within minutes, leaving no lasting residual. That single fact decides where it belongs: ozone can scour water as it passes a contactor, but it cannot follow that water out to a tap and keep it safe. Whole-building hot and cold control is not its job.

Its honest place in building water is narrow - a side-stream on a cooling tower, or a specialist process-water duty - rather than the system-wide residual that HSG274 expects for distal control [1]. It belongs in the same conversation as the other alternative oxidants and physical methods catalogued in [Emerging treatments: UV, copper-silver ionisation and more](https://legionella.io/articles/emerging-treatments-uv-copper-silver-ionisation-and-more/). Treat what follows as orientation for a risk-assessment-led decision, not a design or a dose.

## How ozone actually disinfects

Ozone (O3) is three oxygen atoms in an unstable arrangement that wants to fall apart. That instability is the disinfectant: as ozone breaks down it strips electrons from cell membranes, proteins and the biofilm matrix, killing Legionella and degrading the organic film it shelters in. As an oxidiser it sits above chlorine, chlorine dioxide and monochloramine in raw strength, so the case for ozone disinfection against Legionella is its sheer oxidising power.

You cannot buy ozone in a drum. It will not survive storage or transport, so every system makes it where it is used. On-site ozone generation pulls dried air or concentrated oxygen through a high-voltage corona-discharge cell - a UV cell in smaller units - splitting and recombining oxygen into ozone, which is then dissolved into the water through a venturi injector or a bubble contactor. Generate it, dissolve it, use it, and within minutes it has reverted to ordinary oxygen.

That speed cuts both ways, and it is the entire story of ozone water treatment Legionella decisions.

## The no-residual problem

Here is why ozone is never the answer for a building's hot and cold system. Distal control - keeping the far outlets safe - depends on either temperature reaching every tap, or a disinfectant residual that survives the journey to every tap. HSG274 frames continuous chemical dosing as a secondary measure that carries a persistent residual out to the outlets where Legionella amplifies, with temperature kept as the primary control under ACoP L8 [1][2].

Ozone gives you neither. By the time water reaches the first branch, the ozone has gone. There is no ozone residual at the basin tap, let alone the top-floor shower. You could ozonate water entering a building and still grow Legionella in the very biofilm the ozone never reached, because it decayed in the riser. This is the same limitation that makes UV a point-treatment rather than a system-wide control - set out in [UV disinfection for Legionella: where it works and where it falls short](https://legionella.io/articles/uv-disinfection-for-legionella-where-it-works-and-where-it-falls-short/) - and it is exactly the gap that residual oxidants exist to fill.

That contrast is the cleanest way to read ozone vs chlorine dioxide. Chlorine dioxide is a weaker oxidiser than ozone, but it persists; it holds a measurable residual across a wide pH range out to the far ends of a system, which is the whole point of dosing it. Monochloramine pushes persistence further still, trading oxidant strength for a residual that reaches the index outlet and works into biofilm - a trade examined in [Monochloramine for Legionella control: how it works and where it fits](https://legionella.io/articles/monochloramine-legionella/). Ozone occupies the opposite corner: the strongest kill, the shortest reach. For the full comparison of residual options, the side-by-side is [Secondary disinfection for Legionella compared: chlorine dioxide vs copper-silver vs UV vs monochloramine](https://legionella.io/articles/secondary-disinfection-for-legionella-compared-chlorine-dioxide-vs-copper-silver-vs-uv-vs-monochloramine/).

## Where ozone earns its place: the cooling-tower side-stream

Cooling towers are the application where ozone's short life is not a fault. A tower is a recirculating loop with a large, accessible body of water in the basin and no requirement to deliver a residual to a distant tap - you are trying to keep the circulating water and the wetted surfaces clean, not protect a shower head three floors up. Cooling towers are among the highest-risk systems precisely because they throw out breathable aerosol [1][3], so their treatment is judged on the loop, not on distal outlets.

### Picture the side-stream rig

An ozone cooling tower side stream is worth sketching, because the layout is the design.

Imagine the tower basin at the bottom. A pump draws a slip-stream - a fraction of the circulating water - out of the basin and pushes it through a treatment skid rather than treating the whole flow at once. On that skid sits the ozone generator, fed by an oxygen concentrator or a dried-air supply and wired to a high-voltage corona cell. The ozone it produces is drawn into the side-stream water through a venturi injector, then held briefly in a contact vessel so it can oxidise organisms and organic load before it decays. Treated water returns to the basin, where it mixes and lifts the oxidising condition of the whole loop a little at a time.

Two safety devices are not optional on that picture. At the top of the contact vessel, an off-gas destruct unit catches ozone that comes out of solution and converts it back to oxygen before it can reach the plant-room air. On the return line, dissolved-ozone or ORP monitoring tells the controller how much to make, while a separate ambient ozone gas detector watches the room itself. Lose the destruct unit or the gas detector and you have built a hazard, not a treatment.

Beyond cooling, the same logic puts ozone into specialist process water - bottling and beverage lines, ultrapure and food-contact duties - where treating water at a point of use and then letting the ozone disappear before the water is consumed is a feature, not a flaw. None of that is whole-building plumbing.

## Safety and materials you have to settle first

Ozone is a toxic gas with a low workplace exposure limit, so off-gassing is the dominant hazard. Any ozone that leaves solution has to be captured and destroyed, and the plant room needs ambient gas detection and a COSHH assessment that treats the gas - not just the water - as the risk. The exact exposure limit and the monitoring regime that goes with it sit in the workplace-exposure guidance, not in a figure here.

Materials matter as much. Ozone is aggressive enough to attack many elastomers, seals and gaskets and to oxidise certain metals, so wetted components and any pipework exposed to ozonated water have to be specified for the duty. Confirm what the system is actually built from before anyone sizes a generator around it. Dosed concentrations, contact times and destruct-unit sizing are competent-design figures for your specific application, not numbers to assume from an article.

## General guidance, not a specification

This piece explains how ozone behaves and where it tends to belong; it is not a design, a dosing schedule, or legal advice. Ozone treatment is itself a control measure with its own gas hazard, which means a competent water-treatment specialist has to risk-assess it on your real system, confirm it against the approvals that apply to your water use, and keep both the water side and the gas side under monitoring for as long as it runs. Ozone-system and biocidal approvals vary by application and are not settled by anything implied here.

## FAQ

### Can ozone control Legionella in a building's hot and cold water?

No - not as a system-wide control. Ozone destroys Legionella on contact but leaves no residual, so it cannot protect distal outlets the way temperature or a residual oxidant does. For hot and cold systems, HSG274 points to temperature as the primary control with a persistent secondary residual where it is needed [1][2]; ozone provides neither at the tap.

### Is ozone better than chlorine dioxide?

For raw oxidising power, yes; for protecting a plumbing system, no. Chlorine dioxide is the weaker oxidiser but holds a residual to the far outlets, which is what distal control needs. Ozone's strength is spent at the point of contact and gone within minutes. They are not interchangeable - ozone suits a treated loop such as a tower side-stream, chlorine dioxide suits a system that needs a residual everywhere.

### What is the main hazard with an ozone system?

The gas. Ozone off-gassing in an enclosed plant room is the controlling risk, which is why a compliant system always pairs a contactor with an off-gas destruct unit and ambient ozone monitoring, under a COSHH assessment. The water-treatment benefit is real, but the safety case stands or falls on managing the gas, not the water.

## What to do next

If ozone is on your shortlist, first be honest about the duty. If the problem is distal Legionella in hot and cold water, take ozone off the list now and look at temperature, flushing and a residual option instead. If the duty is a cooling tower or a specialist process loop, the next step is a competent-specialist feasibility review that prices the off-gas destruct, the ambient gas detection and the materials specification alongside the generator itself - and that logs dissolved-ozone or ORP readings, gas-detector checks and Legionella samples in one auditable record rather than across separate clipboards.

## Related reading

- [Secondary disinfection for Legionella compared: chlorine dioxide vs copper-silver vs UV vs monochloramine](https://legionella.io/articles/secondary-disinfection-for-legionella-compared-chlorine-dioxide-vs-copper-silver-vs-uv-vs-monochloramine/)
- [UV disinfection for Legionella: where it works and where it falls short](https://legionella.io/articles/uv-disinfection-for-legionella-where-it-works-and-where-it-falls-short/)
- [Monochloramine for Legionella control: how it works and where it fits](https://legionella.io/articles/monochloramine-legionella/)
- [Emerging treatments: UV, copper-silver ionisation and more](https://legionella.io/articles/emerging-treatments-uv-copper-silver-ionisation-and-more/)

## Sources

[1] HSE, "Legionnaires' disease: Technical guidance (HSG274)". https://www.hse.gov.uk/pubns/books/hsg274.htm
[2] HSE, "Legionnaires' disease. The control of legionella bacteria in water systems - Approved Code of Practice and guidance (L8)". https://www.hse.gov.uk/pubns/books/l8.htm
[3] HSE, "Systems most likely to create legionella risk". https://www.hse.gov.uk/legionnaires/risk-systems.htm