A sentinel outlet that should run hot reads 44°C. The easy move is to write it in the logbook, tick the exception box and wait for next month’s check. The useful move is to ask why the heat stopped arriving, because a missed reading is almost never the actual problem. It is the symptom of one sitting somewhere upstream.
Most plumbing temperature issues come down to a short list of faults wearing different costumes: heat that never reaches the outlet, cold water that quietly warms up, or a mixing valve undoing both. Find the fault and the reading corrects itself. Re-tick the box without finding it and the same exception turns up again on the next round.
First, decide whether the reading is real
Before you condemn a length of pipe, rule out the cheapest explanation: a bad reading. A probe that is out of calibration, a measurement taken at the wrong point, or someone drawing off a basin two minutes before you arrived can all produce a number that looks like a fault and isn’t.
So confirm the basics. Use a calibrated thermometer, measure at the sentinel and TMV points your scheme names, and let the outlet run for the time the scheme specifies before you record hot or cold. A single out-of-range value on an outlet whose neighbours pass, and which itself passes on a clean re-test, is usually a measurement artefact. A reading that repeats is a fault. That distinction tells you whether you are chasing pipework or chasing your own method.
Where hot water loses its heat
When hot water is not reaching temperature, work the causes in order of likelihood rather than tearing into the calorifier first.
Start with demand and run time. Low-use outlets sit full of cooled water; the reading climbs slowly because you are emptying a cold slug before the hot arrives. If it gets there, the outlet is fine and the use pattern is the issue.
Next, check the set point. “Energy-saving” tweaks to storage or circulation temperature are a frequent and quiet culprit. Someone turns the calorifier down a few degrees over the summer, nobody reassesses the risk, and suddenly several outlets drift under their target. The numbers changed; the reasoning didn’t.
Then look at circulation and balance. If a whole branch runs cool, the secondary return pump may be weak or failed, or the system may be out of balance so the far legs are starved while the near ones run hot. A return that comes back to the calorifier cold is the tell.
Only after that do you go hunting for dead legs and long runs. A length of pipe between the circulating loop and the tap holds stored heat away from the outlet, so the water arrives tepid no matter how hot the cylinder is. Stratification inside the calorifier, a cool layer drawn off the bottom while the sensor reads a hot top, produces the same result first thing in the morning. Where a system genuinely cannot hold temperature, thermal disinfection may form part of the remedial response, but it treats the consequence; the fault still needs fixing.
When cold water won’t stay cold
Cold-side faults are less dramatic and more often missed, because a tap that runs lukewarm rarely generates a complaint. The mechanism is almost always heat gain: a cold pipe routed alongside an uninsulated hot one, a cold tank baking in a warm roof space, or a long cold dead leg sitting at room temperature.
Measure along the run, not just at the tap. If the water is cold at the tank and warm at the outlet, the gain is in the distribution; if the tank itself is warm, the storage is the problem. A cold-water tank sitting in a hot plant room or loft is a recurring offender, and one a proper water tank inspection usually flags before a temperature check does. Insulation, separation from hot services, shading or relocating the tank, and clearing stagnant legs are the levers here.
A route from symptom to fix
Use this to get from the reading in front of you to a confirmed cause and an action, rather than guessing. Work the most likely cause first, prove it with the check, then act.
| Symptom | Most likely cause | The check that confirms it | The action |
|---|---|---|---|
| Hot outlet runs tepid (~40-45°C) though stored water is hot | Long run or dead leg between the circulating loop and the outlet | Time how long the outlet takes to climb; compare temperature at the loop vs at the tap | Bring circulation closer, remove the dead leg or re-route, then re-test |
| A whole branch of hot outlets runs cool | Weak or failed secondary return pump, or system out of balance | Confirm the pump runs; measure the return temperature reaching the calorifier | Service or replace the pump, re-balance the legs, verify a warm return |
| Hot water cool only first thing or after quiet periods | Set point lowered, or stratification drawing off a cool layer | Check the calorifier set point and where the draw-off and sensor sit; log overnight temperatures | Restore the set point, correct draw-off/destratify, re-verify |
| TMV-blended outlet stays lukewarm and won’t shift | Failing or scaled thermostatic mixing valve, or a blended dead leg downstream | Isolate and test the hot and cold feeds to the valve separately; check its service history | Service or replace the TMV, shorten the blended section, schedule maintenance |
| Cold outlet creeping above ~20°C | Heat gain from a warm space or adjacent hot pipe, or stagnation | Measure along the run and at the tank; check insulation, separation and use pattern | Insulate or separate, relocate/shade the tank, flush or remove the stagnant leg |
| One reading out, neighbours fine, passes on re-test | Measurement artefact, not a system fault | Re-test with a calibrated probe at the named point after the correct run time | If it repeats, treat as real and trace it; if not, note and keep monitoring |
The thermostatic mixing valve row deserves a flag of its own. A TMV protects against scalding, which is non-negotiable, but place one badly and everything downstream of it becomes a length of permanently blended, lukewarm water that no amount of storage temperature can rescue. When a blended outlet won’t come right, test the hot and cold feeds into the valve before you blame the valve, and keep the run beyond it short.
Knowing when to stop and escalate
Some faults are yours to fix on the spot; others are signals about the whole system, and the pattern is the deciding factor. One cold outlet on one branch is a defect. The same weakness across several outlets, several months, or both hot and cold sides at once is a systemic failure, and the right response is to feed it back into the risk assessment rather than keep patching outlets. Documented cases where unaddressed temperature drift led to Legionella growth tend to read as a string of ignored single readings, not one dramatic break, which is the theme of worked example.
Escalate when you suspect cross-flow between hot and cold through a faulty valve, when the calorifier or secondary pump is failing, when a TMV cannot hold its blend safely, or when you cannot restore control with the access and competence you have on site. UK control rests on the framework in the ACoP L8 and HSG274: a site-specific risk assessment, a written scheme, and a competent person who owns the response [1][2]. A fault you cannot close belongs in that chain, in writing, with the action you took and the result you got, not just a tick to say someone walked past with a thermometer.
Before you act on any number
The benchmark figures here are guidance, not site law. HSE’s hot and cold water guidance describes a familiar set: cold water kept below about 20°C where practicable, hot water stored at around 60°C or above, and hot water reaching roughly 50°C at the outlet within about a minute of running, with 55°C used in healthcare premises [3]. Your own targets, dwell times and the action you take when a reading is out come from your site’s risk assessment and a competent person, not a figure copied off a web page. None of this replaces a registered plumber or engineer doing the actual repair. Where you use sampling to verify or investigate, the frequency follows the system and the risk assessment rather than a fixed calendar [4].
Common questions
Why does one outlet always fail temperature when the rest of the branch passes?
Almost always because it sits at the end of a dead leg or long run, or is barely used, so its water has cooled or warmed before you draw it. Time how long it takes to reach temperature: if it gets there slowly, the fault is the pipe geometry or the use pattern, not the heat source. Persistent failure usually means removing the redundant length or bringing circulation closer.
The water feels hot at the tap, so why does the sentinel still fail?
“Feels hot” is not a measurement. The check is what counts: a calibrated probe, at the named point, after the run time your scheme sets. Pass it properly on re-test and you had a method or timing problem. Fail it after a clean re-test and the outlet really is under target and needs tracing, however warm it felt to the hand.
Can turning the storage temperature down to save energy cause a temperature problem?
Yes, and it is one of the more common hidden causes. Lowering calorifier storage or circulation temperature can pull outlets below target and weaken Legionella temperature control without anyone noticing, because the change stays invisible until the next round of checks. Run any energy-saving adjustment to water temperatures back through the risk assessment before you make it, not afterwards in the readings.
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, “Legionnaires’ disease: Technical guidance (HSG274)”. https://www.hse.gov.uk/pubns/books/hsg274.htm [3] HSE, “Hot and cold water systems”. https://www.hse.gov.uk/legionnaires/hot-and-cold.htm [4] HSE, “Testing and monitoring your water system for legionella”. https://www.hse.gov.uk/legionnaires/testing-monitoring-water-system.htm