---
title: "Cold water temperature guidelines to prevent Legionella"
source_url: https://legionella.io/articles/cold-water-temperature-guidelines-to-prevent-legionella/
canonical_url: https://legionella.io/articles/cold-water-temperature-guidelines-to-prevent-legionella/
pillar: "Water Temperature Control"
summary: "A single below-20°C reading can hide a cold water system that's warming up. How UK duty holders measure cold water temperature that proves Legionella control."
primary_keyword: "cold water temperature"
date_published: 2025-05-27
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."
---

# Cold water temperature guidelines to prevent Legionella

Cold water has one job in Legionella control: stay cold. The working benchmark in UK guidance is to keep it below 20°C, because that is roughly where the bacteria stop sitting dormant and start to multiply [1]. The catch is that a single reading below 20°C proves almost nothing on its own.

A cold tap can read 14°C the moment a contractor runs it on a cool March morning, and the same outlet can sit at 23°C on a still afternoon in July. Cold water temperature is not a number you capture once. It is a property of the whole system, and the readings only count if they catch that system on a bad day rather than a convenient one.

## The office that passed every check and still warmed up

Picture a three-storey office — nothing unusual. Mains cold water feeds a storage tank in the roof void, and the tank feeds outlets down through the building. The water safety file looks healthy: sentinel temperatures logged every month, and every cold reading sitting comfortably below 20°C. On paper, that is control. (This is an illustrative composite, not a real site, but every detail in it is ordinary.)

Two problems hid in the gaps. First, the readings were always taken on the contractor's regular mid-morning visit, after the tap had run for a couple of minutes, and only at the two sentinel outlets — the nearest and the furthest from the tank. Both happened to be busy outlets that flushed themselves through the working day. Second, the cold water storage tank, sitting in an uninsulated roof void, was never measured at all.

Across a hot fortnight in July the void baked. The stored water crept up toward room temperature, and a little-used kitchenette tap on the top floor — fed off a short branch that drew about one mug of water a day — sat in the low twenties for hours each afternoon. Nobody recorded it, because nobody measured the thing that was actually warming.

The signal had been in the file the whole time. The incoming cold reading had drifted month on month: 13, then 15, then 17. Each figure was logged as a pass because it stayed under the line, and a clear trend was treated as a string of unrelated single points. That is the quiet chain a responsible person learns to read earlier — not one alarming number, but a direction of travel nobody plotted.

## What "cold enough" actually means

Strip it back to the guidance and cold water control rests on one figure and a handful of places to check it. HSE's hot and cold water guidance describes the conventional benchmark — cold water kept below 20°C where practicable — because below that the organism stays dormant rather than growing [1]. Hot water is the mirror image, stored hot and delivered hot; that side is covered in [Hot water temperature guidelines](https://legionella.io/articles/hot-water-temperature-guidelines-to-prevent-legionella/). Treat the 20°C figure as guidance to confirm and apply through your own assessment, not a fixed legal line stamped on every building.

Cold water rarely fails at the mains. It fails on the journey. The usual culprits:

- **Storage tanks in warm spaces** — roof voids and plant rooms that heat up in summer, especially when the tank is oversized and turnover is slow, so the same water sits warm for days.
- **Cold pipes shadowing hot ones** — runs through warm risers and ceiling voids with no insulation or separation, picking up heat all the way to the outlet.
- **Low-use branches and dead legs** — short spurs where water stands long enough to drift up to room temperature. This cold water stagnation is the exact condition the bacteria want.
- **A rising incoming temperature** — mains water that arrives a few degrees warmer in August than in February, quietly eating the margin you had in winter.

A particular blind spot sits downstream of thermostatic mixing valves, where blended water is neither properly hot nor properly cold; if your building has them, [Thermostatic mixing valves and Legionella risk](https://legionella.io/articles/thermostatic-mixing-valves-and-legionella-risk/) is worth a read alongside this.

## Lessons to carry to your own site

The point of the office above is not the building. It is the four habits that would have caught the problem before July.

**Measure the thing that warms, not just the thing that's convenient.** The stored water in the tank, the outlets most exposed to ambient heat, the branches that barely run — these tell you far more than a busy sentinel that flushes itself clean. Sentinel outlets are the start of the picture, not the whole of it.

**Take the reading at the worst moment.** A cold outlet should still be cold after the water has been standing, in warm weather, at the end of a long run. Sample when the system is under pressure, because that is when people are actually exposed. It is worth checking your own measurement method — how long you run the tap, where you place the probe — against HSE's description of cold sentinel monitoring [1].

**Plot the trend, not the pass or fail.** A column of "below 20°C" ticks can still be sliding the wrong way. A reading that climbs three degrees over three months is telling you something a single threshold never will, and HSE's technical guidance leans on the risk assessment, not a fixed calendar, to set how often you look [2].

**Write down the decision behind the task, not only the result.** "This kitchenette outlet is checked monthly and flushed weekly because it is low-use and fed off a warm riser; a reading above 20°C escalates to the responsible person and triggers an insulation review." That one sentence turns a tick-box into a managed control — and it is what an auditor, or your own successor, needs to see.

## Before you rely on a number

None of the figures here replace your own assessment. The 20°C benchmark, how often you measure, which outlets count as sentinels, and what you do when a reading climbs are all decisions for a competent, site-specific risk assessment carried out under the L8 framework [3]. A slow-filling tank in a cool basement and a small tank baking in a loft are not the same problem, and the same reading can mean different things in each. If a temperature is drifting upward and the cause is not obvious, get a competent person to look at the system — the tank, the routing, the insulation — and not just the outlet in front of you.

## FAQ

### Why 20°C, and is it a hard limit?
It is the practical benchmark because it sits at the bottom of the range where Legionella begins to grow; hold cold water below it and you keep the bacteria dormant rather than multiplying [1]. Treat it as the target you design and monitor toward, not a single statutory threshold — your risk assessment decides exactly how it applies to your system [3].

### Should I run the tap before measuring cold water temperature, or measure it standing?
The two readings answer different questions. A reading taken straight away shows what has been sitting in that branch; a reading after the water has run shows what the supply or tank is delivering. For finding Legionella risk the standing reading is often the more revealing, because stagnant warm water in a low-use spur is exactly what you are hunting for. Check both against HSE's guidance on how cold sentinel temperatures are taken [1].

### Our cold water storage tank is in the loft — is that a problem?
It can be, particularly in summer. A tank in an unheated, uninsulated roof void can warm toward room temperature on hot days, and if it is oversized for the building the water turns over slowly and stays warm for longer. Measuring the stored water itself — not only the outlets fed from it — is how you find out, and insulation, relocation or right-sizing the tank are the usual fixes.

## Related reading

- [Hot water temperature guidelines to prevent Legionella](https://legionella.io/articles/hot-water-temperature-guidelines-to-prevent-legionella/)
- [Thermostatic mixing valves and Legionella risk](https://legionella.io/articles/thermostatic-mixing-valves-and-legionella-risk/)
- [Key components of a Legionella risk assessment](https://legionella.io/articles/key-components-of-a-legionella-risk-assessment/)
- [Legionella sampling 101: how and why to test your water](https://legionella.io/articles/legionella-sampling-101-how-and-why-to-test-your-water/)

## Sources

[1] HSE, "Hot and cold water systems". https://www.hse.gov.uk/legionnaires/hot-and-cold.htm
[2] HSE, "Legionnaires' disease: Technical guidance (HSG274)". https://www.hse.gov.uk/pubns/books/hsg274.htm
[3] 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