A sentinel outlet is a tap or shower you check on purpose, on a set rhythm, because it sits at an extreme of the system: the first outlet off the calorifier and the last one on the hot loop, the nearest cold tap to the storage tank and the furthest one from it. Watch those few points and you are reading the temperature behaviour of the whole pipework through a small, representative sample.

That is both the appeal and the trap. Sentinels are an early-warning sample, not a clean bill of health for every fitting between them. Used well, they tell you cheaply and monthly whether your hot water stays hot and your cold water stays cold across the worst-case stretches of the building. Read lazily, they let a tidy logbook stand in for a controlled system - which is exactly how sites get caught out.

What actually makes an outlet a sentinel

The word does the work: a sentinel stands watch at the boundary. On a hot water service, the outlet closest to the calorifier or cylinder should be the easiest place in the building to get hot water. If that one runs tepid, suspect the source - the calorifier setting, a failing thermostat, short cycling. The outlet furthest along the circuit is the hardest to keep hot, because the water has travelled and shed heat on the way. If the far outlet still arrives hot, everything upstream of it on a continuous run almost certainly is too.

Cold water mirrors the logic. The tap nearest the cold water storage tank is where cold should be coldest; the furthest tap is where the water has had the most time, and the most warm plant-room exposure, to drift upward. Hold both ends and you have bracketed the system.

Most buildings need more than one pair. A site with several risers, separate loops or distinct circuits will have a set of sentinels for each. Your risk assessment and written scheme define exactly which outlets count, and L8 and HSG274 are the references that frame how that monitoring is set up [1][2]. The aim is not to check everything every month; it is to check the nearest and furthest outlets that fail first.

Where the sentinels sit: a sketch you can draw

Imagine the system as two trees. Start with the hot water tree: a calorifier at the trunk, a flow pipe branching out to every outlet, and - on a circulated system - a return pipe carrying water back to be reheated. Mark two points in red: the first branch off the trunk (the nearest sentinel) and the tip of the longest branch (the furthest sentinel). Add a third mark on the calorifier itself, where you record flow and return temperatures.

Now draw the cold tree beside it: the storage tank at the top, feeding outlets below. Mark the branch nearest the tank and the tip of the longest branch.

Under each red mark, write four things: the temperature you expect, the task (a monthly check), the person who owns it, and what happens if the reading is wrong. Any mark missing one of those four is a gap in control, not just a gap on a drawing. Drawn this way, it quickly shows whether a site understands its sentinels or has simply inherited a list of outlets from a previous contractor.

How to monitor them without fooling yourself

The routine is simple; the discipline is in doing it honestly. Sentinel temperatures are commonly checked monthly as part of routine temperature monitoring, with calorifier flow and return readings on the same cycle - though the frequency your scheme sets follows the system and the risk assessment rather than a fixed national number [2][4].

When you run a hot sentinel, let it flow and time it. General HSG274 guidance is that hot water should reach roughly 50°C within about a minute (higher, near 55°C, in healthcare settings) [2]. For a cold sentinel, the expectation is that the temperature stays below about 20°C after running for a couple of minutes [2][3]. Treat those as targets to confirm against current guidance, not gospel - and record the actual figure you measured, not a tick in a box. A logged number lets you see drift; a tick hides it.

One mistake quietly invalidates a lot of monitoring: measuring on the wrong side of a thermostatic mixing valve. A TMV blends hot and cold down to a safe delivery temperature, so a blended outlet reads tepid by design and tells you nothing about whether the hot supply is hot. Measure the supply temperature upstream of the TMV, at the point the scheme specifies. If you only ever sample the mixed water at the spout, you can pass every month while a hot circuit slowly fails behind the valve.

If a sentinel reads out of range, the reading is the start of the work, not the end of it. Investigate, act, and re-check - and write down the decision, not only the fault. A clear record of what you found, what you did and when you confirmed the fix is what turns a temperature log into evidence of control; on keeping flushing and sampling records covers what that paper trail should contain.

What sentinel readings can’t see

Here is the part generic guides skip. Two compliant sentinels do not certify everything between them. A short dead leg, a capped spur nobody uses, or a single low-use shower on a corridor can sit stagnant and warm while both ends of the run read perfectly. The sentinels bracket the extremes; they do not patrol the middle.

That is why sentinel monitoring is a system-health check, not an outlet-by-outlet guarantee. Sound schemes pair the monthly sentinel checks with a rotating programme that brings every other outlet into the temperature record over a defined period, plus separate attention to low-use and dead-leg points. It is also why temperature control and sampling answer different questions. A clear Legionella sample describes one outlet at one moment and follows site-selection rules of its own under BS 7592 [5]; it never substitutes for keeping the water hot, cold and moving. A run of clean sentinel readings followed by a sudden failure is exactly the pattern behind many temperature-related incidents - worked example shows how that drift builds unseen.

Treat all of this as general guidance, not advice for your particular building. Which outlets are your sentinels, the temperatures you hold them to, and how you respond to a failure are decisions for a competent person working from your own risk assessment and written scheme - the standards set the frame, not the site-specific answer.

Set yours up this month

Pull your written scheme and find the outlets named as sentinels. Walk the building and confirm they still exist, are still in use, and still sit at the system’s real extremes - extensions, new wings and decommissioned outlets all move the goalposts. For each one, check that you measure on the correct side of any TMV, that the expected figure is written down, and that someone owns the monthly reading. If any of that is missing or out of date, fix it before the next round of checks.

FAQ

How many sentinel outlets should a building have?

There is no single number. At a minimum, a simple system has a nearest and a furthest sentinel on both the hot and the cold service; larger buildings add a pair per loop, riser or circuit. Your risk assessment and written scheme set the list, and it should be revisited whenever the pipework or the use of the building changes.

Should I take the temperature before or after the mixing valve?

For a sentinel hot-water reading, measure the supply upstream of the thermostatic mixing valve - that is the temperature that tells you whether the circuit is hot. The blended outlet downstream reads cool by design, so checking only there can mask a failing hot supply. Where outlets are TMV-served, the scheme should name the exact measurement point.

Do sentinel checks replace monitoring every other outlet?

No. Sentinels are a representative early warning for the system’s extremes, not proof that every tap between them is fine. They work alongside a rotating check of other outlets and specific attention to low-use points and dead legs, so the whole system is covered over time rather than just its endpoints.

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 [5] BSI, “BS 7592:2022 - Sampling for Legionella bacteria in water systems. Code of practice”. https://knowledge.bsigroup.com/products/bs-7592-sampling-for-i-legionella-i-bacteria-in-water-systems-code-of-practice-1