What an engineer actually looks at inside your cold water tank

An annual look inside a cold water storage tank takes a torch, a mirror or a phone camera, and a trained eye, and most of it is a list you could learn in an afternoon. The lid should be close-fitting and sound, with no light getting in. The insect and vermin screens on the overflow, warning and vent pipes need to be intact, the insulation continuous. Inside, you want no sediment, scale, debris, biofilm or algae, water that runs clear and without odour, corrosion held in check. And underneath all of it sits the question that decides the rest: is the water actually turning over, or standing still in a dead zone (the standard inspection checklist)? You inspect at least annually under HSG274 Part 2 and BS 8558:2015, and six-monthly is the sensible interval for commercial and multi-residential tanks, partly to line up with quarterly sampling.

The harder part is not the checklist. It is the decision the checklist feeds: does this tank need a clean and disinfection, or does it pass for another year. That call is risk-assessed, not automatic. It gets triggered by contamination found on inspection, by remedial works on the system, by an extended period out of use, by an actual contamination event, or by the risk assessment calling for it. There is no fixed sediment depth that forces a clean the moment you exceed it. Anyone quoting you a magic number in millimetres is inventing a threshold the guidance doesn't set.

The temperature that frames the whole thing

Everything in the tank is being judged against one number. Stored cold water should sit at or below 20°C, and above that the growth risk climbs sharply, which is why HSE draws the line there. The cold outlets fed from the tank should run to below 20°C within two minutes. A tank that is warm, or one whose insulation has gaps that let it gain heat, is drifting into the band where legionella multiplies, and that changes how you read everything else you find inside it.

The checklist, and what each item is really checking

Each line on the list maps to a way water goes bad in storage.

• The lid, close-fitting and sound, with no light getting in. A poor lid lets in light that feeds algae, along with dust, insects and worse.
• The screens on the overflow, warning and vent pipes, intact against insects and vermin. A torn screen is a straight path in for a dead pigeon or a colony of flies.
• The insulation, continuous and without gaps. It is not there for frost; it holds the stored water below 20°C by keeping plant-room heat out.
• The interior, free of sediment, scale, debris, biofilm, algae and corrosion. Sediment is a nutrient bed, and biofilm is where legionella lives and hides from whatever you dose the water with.
• The water itself, clear and free of odour. Cloudy, or smelling of anything, is a finding rather than a quirk of the site.
• The geometry, with inlet and outlet placed so the water turns over and no dead zone where a pocket sits stagnant. A tank can be spotless and still be a risk if half its volume never moves.

Any single finding can be minor on its own. Read together, they tell you whether the tank is being kept, or is slowly turning into a culture vessel.

When an inspection becomes a clean

This is the judgement the whole visit is building towards, so it is worth being blunt about how it gets made. You are not measuring the tank against a pass mark. You are asking whether what you have found, in the round, means the water can no longer be trusted until the tank is drained, cleaned and disinfected. Heavy sediment, visible biofilm, an ingress point that has clearly been open, water that is discoloured: any of those tips a routine inspection into remedial work. A light dusting of sediment in an otherwise sound, cold, well-sealed tank may not. That is a competent person's call, recorded with the reasoning, not a number lookup.

The reason to labour this is that "there is no fixed threshold" gets heard as "it is vague, so anything goes". The opposite is true. Because there is no number to hide behind, the inspector has to write down what they saw and why they decided as they did, and that written reasoning is what an auditor reads.

What the disinfection actually involves

When the call goes the other way, the disinfection has a defined shape under BS EN 806-5 and BS 8558. Drain the tank. Physically clean it, because you cannot chlorinate your way through a layer of sediment and biofilm. Shock-chlorinate to 50 mg/l of free chlorine and hold it for at least an hour, with the residual still at 50 mg/l or above when the contact time ends, which is how you know the dose was not simply eaten by the dirt you left behind. Then flush the whole system through until the chlorine is back below 0.5 mg/l, and confirm the result with microbiological sampling around 48 hours later before the tank goes back into service.

That last step is the one people rush. The sample two days later is what turns "we cleaned it" into "we cleaned it and proved it worked", and returning a tank to service before that result is back is putting a signature on an outcome you have not seen.

The tank has a history, and it matters

One tank inspected once tells you about today. The same tank inspected every year tells you a story: the sediment that keeps coming back points at something upstream, the insulation gap noted last year is still there, the screen gets torn on the same overflow every winter. That year-on-year record is where an annual inspection earns more than a same-day pass or fail.

This is the part BlueWave is built to hold. Each tank is an asset on its site with its own service history, so this year's inspection opens with last year's findings attached, and a recurring pattern shows up as a pattern instead of five disconnected visits. Opening the lid and reading what is in the tank stays with the engineer. Holding the history, so this year's findings sit against last year's, is the part worth handing to software.

For where the tank inspection sits in the wider calendar, see the task frequency table. For what to do when a cold outlet fed by that tank reads above 20°C, the failed-reading workflow is the exception path. And the inspection record itself has to earn its place in the logbook, which carries its own requirements.

Open the lid and the first question is the simplest one. Is this water still cold. A tank that has drifted above 20°C is a problem however clean it looks, because keeping it covered and keeping it moving were always, in the end, about keeping it cold.