Temperature is the one Legionella control you can read straight off a dial, which is exactly why it gets misjudged. A hot tap that hits 60°C and a cold tank sitting at 15°C feel like proof the water is safe. They are not. They are two readings, at two points, at one moment — and Legionella multiplies in the gaps between your check points, not on top of them.

The relationship between temperature and Legionella growth is easy to state and easy to break. Get the principle straight and you can stop staring at individual readings and ask the question that actually matters: can the whole system hold these conditions, week in and week out, in the parts you never routinely measure?

The relationship, in numbers

Legionella is a warm-water organism. Below about 20°C it survives but multiplies slowly; across the roughly 20–45°C band it grows, fastest somewhere in the middle; and above about 50°C it begins to die off, faster the hotter the water gets [1]. That growth band is the whole reason for the conventional UK approach to hot and cold water temperatures: keep cold water below 20°C where practicable, and keep hot water genuinely hot — typically stored at around 60°C and delivered so it reaches about 50°C at the outlet within a minute, or around 55°C in healthcare premises [2]. Your risk assessment sets the working figures; these are the benchmarks it usually starts from.

Two features of that relationship trip people up. It is a curve, not a switch — water does not flip from dangerous to safe the instant it crosses a line, so brief, marginal compliance does far less than a comfortable margin held steadily. And temperature acts on bacteria that are rarely floating free. They shelter in biofilm and sediment, so a quick brush past 50°C at a tap does not sterilise a colony living in scale at the base of a calorifier or inside a little-used shower hose [3]. That is why temperature and cleanliness are the same job, and why the biofilm hiding in your pipework blunts a reading that looks fine on paper.

Why one reading in spec isn’t control

Picture a real system rather than a thermometer. Hot water may leave a calorifier at 60°C, but calorifiers stratify — the bottom can sit far cooler than the top, and that cooler layer is where sediment collects. Out on a circulating loop, water loses heat with every metre, so the return can arrive well below where it left; a far outlet on a long run may struggle to reach 50°C within a minute even when the plant room looks perfect.

Cold water has the mirror problem. A tank in a warm roof void, or a cold pipe run clipped alongside a hot one, drifts up into the growth band — most of all in a hot summer, when mains water arrives warmer to begin with.

Then there are the places no routine check ever reaches: the capped spur left behind a removed basin, the room shut for refurbishment, the outlet behind a store cupboard. These dead legs and stagnant branches sit close to room temperature, undisturbed, which is close to ideal for the bacteria. A clean set of sentinel readings tells you nothing about them. This is the heart of the Legionella temperature question: control is a property of the whole system, and a reading only describes the point and the moment you took it.

How to check the whole system, not just the taps

Verifying temperature control is less about collecting numbers and more about collecting the right numbers in the right places — and knowing what to do when one falls out of range. A sequence that holds up:

  1. Map where heat is won and lost. Walk the system from source to furthest outlet. Note the calorifier or heater, the cold tank, the loops, and every point where hot and cold run close enough to trade heat. Done when you have a simple thermal picture, including the longest and least-used branches.
  2. Choose temperature monitoring points that represent the system, not the convenient ones. Sentinel outlets — the nearest and furthest on each loop — plus a rotating sample of the rest, weighted towards low-use and far-flung fittings. Done when your list includes the outlets most likely to fail, not just the ones beside the plant room.
  3. Fix the limit and the response before you measure. Write the target, the acceptable result, and the exact action if a reading misses. Done when a technician can act on an out-of-range reading without phoning for instructions.
  4. Measure how the system actually runs. Cold after a short run, hot at the outlet within a minute, and — importantly — sometimes after a realistic period of low use rather than a convenient flush. Done when the readings reflect service conditions, not best-case ones.
  5. Treat every exception as a question. An out-of-range reading is the start of an investigation: why here, why now, is it the loop, the lagging, or the use pattern? Done when each miss has a recorded cause and a closed action, not just a red cell in a spreadsheet.
  6. Re-walk the picture after any change. New tenants, a refit, a season, a decommissioned wing — each can move where water sits warm. Done when the written scheme and the monitoring points are updated to match the building as it is now.

Where the relationship quietly breaks

A few spots defeat good intentions more often than the rest.

Thermostatic mixing valves. A thermostatic mixing valve blends hot and cold down to a safe delivery temperature, often near body temperature, to stop people scalding. That is sound for safety and a small gift to Legionella, because the blended water and the short length of pipe past the valve can sit squarely in the growth band. The answer is to keep that section short, keep it used, and maintain the valve — not to abandon the TMV.

Seasonal cold water. The cold side that comfortably held below 20°C in February can drift over it in a warm August roof space. Same system, same monitoring regime, different risk. Cold control is half the relationship, not an afterthought to the hot side.

The outlets you never use. Low-use and no-use outlets undo more temperature regimes than any plant fault. They cool to room temperature and stagnate, and no thermometer is checking them. Either bring them back into genuine use, flush them on a defined schedule, or remove the pipework — leaving it capped and forgotten is the one option that fails.

A word on the numbers

Those temperatures are conventional UK benchmarks, not laws of nature and not a substitute for judgement. A competent person and your risk assessment set the figures, the monitoring points and the actions for your specific building — a hospital, a leisure centre and a small office will not land in the same place. And hold the obvious tension in mind: water hot enough to control Legionella is hot enough to scald, which is why mixing valves exist and why hot-water safety has to be designed in, not bolted on after a complaint.

Where to start this week

Pick one circulating loop — ideally one you suspect, not your best — and walk it end to end with a calibrated thermometer. Take the source, the storage, the nearest outlet, the furthest, and then one outlet you know barely gets used. Five readings on one loop will tell you more about whether the relationship holds than a hundred tidy entries from the taps nobody worries about. If the far or low-use point disappoints, you have found your first real action — and a reason to widen the check across the rest of the system.

FAQ

Is there a single temperature that kills Legionella instantly?

No. Heat kills Legionella faster the hotter the water, so it is a relationship between temperature and time, not a single magic number. Water around 60°C kills the bacteria quickly; the same outlet briefly touching 50°C does much less, especially to colonies sheltered inside biofilm and scale [1].

If every outlet passes its temperature check, do I still need to worry about the cold side?

Yes. Cold water that creeps above about 20°C — in a warm roof space, a long run beside hot pipes, or across a hot summer — moves into the growth band even when the hot side looks fine [2]. Cold control is half the relationship, not a box ticked once.

Does hitting the right temperatures mean I can skip sampling?

Not automatically. Temperature is a control; sampling is verification, and HSE guidance is clear that whether and how often you sample follows the system and the risk assessment rather than a fixed calendar [4]. Strong temperature records can reduce the need for routine sampling, but they do not overrule the risk assessment’s call.

Sources

[1] 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 [2] HSE, “Hot and cold water systems”. https://www.hse.gov.uk/legionnaires/hot-and-cold.htm [3] CDC, “How Legionella Spreads”. https://www.cdc.gov/legionella/causes/index.html [4] HSE, “Testing and monitoring your water system for legionella”. https://www.hse.gov.uk/legionnaires/testing-monitoring-water-system.htm