Hot Water Tank Not Heating: How to Find the Fault

What is actually meant to be heating your tank?

Stand at the cylinder and look at it properly. If there is a round plastic cap on the side or on the top, with a heat resisting flex running into it from a switch on the wall, that is an immersion heater. Many cylinders have two of them, one high up and one near the bottom. If there are two thick pipes entering the side of the cylinder, usually with a pump, a motorised valve and a bundle of wiring nearby, the boiler is heating it through an internal coil.

Plenty of London cylinders have both. The boiler coil does the everyday work and the immersion sits there as a backup, which is exactly why so many households do not notice for months that one of the two has failed. If your water has been fine on the immersion and you have simply never switched the boiler side on, or the other way round, you already know which half is broken.

The other thing worth identifying is whether the cylinder is vented or unvented. A vented cylinder is fed by a cold tank in the loft and usually has a copper look with a vent pipe rising from the top. An unvented cylinder is a sealed white or grey metal case fed straight off the mains, with a pressure reducing valve and a tundish somewhere in the pipework. The diagnosis for a cold tank is much the same either way, but unvented units are sealed pressure systems and anything involving their safety devices is not a job for the householder. A third possibility in a great many purpose built flats is not a cylinder at all but an electric thermal store, and Pulsacoil repair in London is a different discipline again, because in those units the water you wash in is heated on its way through rather than stored in the vessel.

• Plastic cap and a flex from a wall switch: electric immersion heater
• Two pipes into the side plus a pump and valve: boiler fed through an internal coil
• Both present: the boiler is primary, the immersion is backup, and either can be dead
• Loft tank above and a vent pipe: vented cylinder
• Sealed metal case, mains fed, tundish nearby: unvented cylinder

Cylinder not heating up on the immersion heater

Switch the immersion on at its wall switch and leave it. A 3kW element in a typical domestic cylinder takes somewhere between one and two hours to bring a full tank up to temperature, so judging it after fifteen minutes tells you nothing, and for the first half hour a tank that is simply still warming up looks exactly like a tank that is never going to warm up at all. What does tell you something is the pipe leaving the top of the cylinder. Put a hand on it after twenty minutes. Warmth there means heat is going in.

If there is nothing at all after a couple of hours, the usual culprits are the element, the rod thermostat buried in it, or the thermal cut-out next to the terminals. The element is a resistance wire sealed inside a metal sheath and it fails in one of two ways. It goes open circuit, in which case nothing happens and nothing trips, or the insulation inside the sheath breaks down and it starts leaking to earth, in which case the breaker or RCD goes the instant you switch it on. Neither can be repaired. A sealed element is a replacement part.

If the element is sound, the fault is in the head. The rod thermostat can stick open so it never switches the element on, and the thermal cut-out can trip and lock out after an overheat. A tripped cut-out is a symptom rather than a cause: something let the water climb past roughly 85 degrees, and until that is found, resetting it just buys you a few more hours. Both parts come out of the head without draining the cylinder, which is why an engineer tests before condemning anything. The page on an immersion heater not working takes the electrical side further, including how the element is proved dead on a meter rather than by guesswork. An immersion heater repair is a head and an element rather than a cylinder job, which is why it is normally the shortest route back to hot water when the vessel itself is sound and only the heat source has given up.

One more thing people miss. On an Economy 7 setup the lower element is usually wired to a night rate supply and only has power for a few hours in the early morning. If you are testing at 2pm and getting nothing, that may be the tariff working as designed, not a fault.

Never switch an immersion on when the cylinder has been drained or is only part full. An element run dry destroys itself in seconds and can distort the boss it screws into. If anyone has drained the tank, leave the immersion switch off until it is refilled.

Hot water cylinder not heating from the boiler

When the boiler heats the cylinder, several things have to happen in sequence, and a failure in any one of them leaves you with a cold tank and a boiler that looks perfectly healthy. The programmer has to be calling for hot water. The cylinder thermostat strapped to the side of the tank has to agree that the water is below its setting. That signal has to open the motorised valve. The valve has to reach the end of its travel and close the microswitch inside it. Only then does the boiler fire and the pump push hot water round the coil.

Start at the programmer, because it is free to check and it is wrong more often than anyone admits. Put hot water into constant rather than timed and see what happens over the next twenty minutes. If the boiler fires and the tank warms, your schedule had drifted or a battery backed clock had lost the time. That is the end of the job.

If the boiler will not fire for hot water but happily fires for heating, the split tells you a lot. It means the boiler, the gas and the pump are fine, and the fault is in the hot water leg only: the cylinder thermostat, the hot water side of the motorised valve, or the wiring between them. On a common Y plan system a single three port valve serves both heating and hot water, and the most frequent failure is the synchronous motor or the microswitch inside the actuator head, which unclips from the valve body without touching any water. That exact pattern, no hot water but the heating works, has its own page going through the valve, the cylinder thermostat and the wiring between them in sequence.

A less obvious one is a seized valve. The valve paddle can stick in the mid position or in the heating position after years of sitting still, and the motor is not strong enough to shift it. The give away is a faint buzz from the valve with no movement, or an actuator head that gets warm while the tank stays cold. The head can often be swapped on its own, and there is a manual lever on the side of most valves that lets an engineer prove whether the body is free before ordering anything.

• Boiler fires for heating but not hot water: cylinder stat, valve or wiring in the hot water leg
• Boiler never fires at all: boiler side fault, not a cylinder fault
• Valve buzzing, tank cold: seized valve body or a failed actuator motor
• Pump silent when hot water is calling: pump, or the valve microswitch never made
• Everything runs, coil pipes stay cold: coil blocked or sludged, or the pump is passing nothing

The tank is warm at the top but stone cold at the bottom

This is normal, right up until it is not. Hot water is less dense than cold, so a cylinder stratifies: the hot sits in a layer at the top and the cold stays underneath. When you draw off hot from the top, cold mains or cold from the loft tank comes in at the bottom to replace it, and the boundary between them slowly rises. A tank that is hot at the top and cold at the bottom is simply a tank that is part way through heating up or part way through being used.

It becomes a fault when the hot layer stops growing. If the top eighteen inches are hot and the rest never moves after several hours of heating, the heat is only entering high up. On a two element cylinder that means the lower element has failed and you are living on the top boost, which is a third of the tank at best. On a boiler fed cylinder it means the coil is not transferring heat down the length of the tank, usually because it is coated in scale or because flow through it has dropped to a trickle. The practical result in most households is no hot water in the shower after the first few minutes, while a basin still runs warm for much longer simply because it draws so much less.

The quick physical check is to run a hand slowly down the side of the cylinder from top to bottom and find where it goes cold. Do it after a long heating period rather than after a shower. If the cold line sits in the same place every time no matter how long the heat has been on, you have found the limit of what is actually working.

The tank heats but the water is never properly hot

Lukewarm water that never reaches a decent temperature is a different fault from no heat at all, and it usually points at a thermostat rather than a heat source. Both the rod thermostat inside an immersion head and the strap on cylinder thermostat used with a boiler drift downwards with age. A stat marked 60 can be switching off at 45 after enough years, and 45 degrees at the cylinder is a short, disappointing bath. Water not getting hot enough to be useful, rather than not getting hot at all, is a separate complaint with a separate short list behind it, and it almost always ends at a control rather than at a heat source.

Before anyone changes a part, check the setting itself. Cylinder thermostats have a dial that people knock while reaching past them, and immersion thermostats have an adjuster under the cap. The right number for a domestic cylinder is around 60 degrees. That is hot enough to keep stored water safe and low enough that the element or coil is not being scaled to death every cycle. Turning it higher to compensate for a tired system is not the fix, it just makes the scaling worse and raises the scald risk at the taps. A stat that fails the other way, stuck closed rather than drifting low, takes the system into hot water cylinder overheating territory, and that is a safety fault rather than a comfort one.

The other cause of permanently lukewarm water on a boiler fed cylinder is flow temperature. If the boiler is running at a low flow temperature to suit the heating, it may simply not have enough temperature difference left to drive heat across the coil into the tank. Most systems are set to raise flow temperature when hot water is calling, and when that setting has been lost or the boiler has been swapped without it being restored, you get exactly this: heating that works beautifully and hot water that never quite arrives.

Why hard water makes a tank slower and slower to heat

Most of London sits on hard water, and hardness is the single biggest reason cylinders and elements fail here earlier than the same kit does elsewhere. When water is heated above roughly 60 degrees, dissolved calcium comes out of solution and deposits onto the hottest surface it can find. Inside a cylinder that is the immersion element sheath or the outside of the boiler coil.

Scale is an insulator. A coil or element with a hard crust on it still gets hot, it just cannot pass that heat into the water around it. So the symptom develops gradually: the tank takes longer and longer to come up, then it stops reaching temperature at all, and the element runs hotter and hotter behind its own insulation until the sheath fails or the cut-out starts tripping. A scaled element that is tripping its cut-out is not a faulty cut-out. It is a cut-out doing its job.

You can sometimes hear it before you measure it. A scaled element crackles, ticks or makes a kettling noise as steam pockets form under the deposit and collapse. That sound from a cylinder is worth acting on, because an element at that stage is close to the end and the debris shedding off it ends up in the bottom of the tank.

The practical consequence for diagnosis is that in a hard water area, a slow cylinder is a real fault rather than a quirk. A tank that used to be ready in ninety minutes and now needs four hours is telling you something has built up on the heat transfer surface, and that trajectory only goes one way.

If the cylinder is making a rumbling, kettling or crackling noise while it heats, and the water is also taking much longer than it used to, get it looked at rather than waiting for it to stop working. Heavy scale on an element can damage the boss and turn a simple element change into a cylinder replacement.

What you can safely check yourself before calling anyone

This is the honest part. A reasonable share of cold cylinders are fixed in five minutes by the householder, and there is no sense paying someone to press a button you could have pressed. None of the checks below involve opening anything, touching water under pressure, or removing a cover.

Work through them in order and stop when the tank starts warming. If you get to the end of the list and it is still cold, you have not wasted the time: you have given an engineer a much shorter list to work from, which shortens the visit. If you are not yet certain whether the cylinder or the boiler is the problem, the no hot water checklist before you call works the same ground from the other end and takes the whole system rather than the tank as its starting point.

• Check the programmer is calling for hot water, then put it to constant for twenty minutes and watch
• Check the immersion switch on the wall is actually on and its neon is lit
• Check the breaker or fuse for the immersion circuit at the consumer unit has not tripped
• Check the cylinder thermostat dial has not been knocked down to a low number
• Check nobody has left the system on holiday mode or an away setting
• Feel the pipe leaving the top of the cylinder twenty minutes after switching on, and feel your way down the tank side to find where it turns cold
• On a boiler fed system, turn the heating on separately and see whether the boiler fires for that

Do not reset a tripped breaker repeatedly to see if it holds. If a breaker or RCD trips the moment the immersion is switched on, the element is almost certainly leaking to earth. Leave the immersion switched off and the circuit isolated, and have it tested.

What an engineer checks, and in what order

A competent visit follows the heat, not the guesswork. The first question is whether the heat source is running, and the second is whether the heat it produces is reaching the water. Everything else is detail.

On the electric side that means a voltage check at the immersion head to prove power is arriving, then a continuity test across the element terminals. A healthy 3kW element on a 230 volt supply reads somewhere near 18 ohms, because resistance is voltage squared divided by wattage. Nothing at all means an open circuit element. Then an insulation resistance test between a terminal and the element body, which is a separate instrument and a separate test, because an element can read perfectly on continuity and still be the thing tripping your electrics. Both tests get done before anyone says the word replacement.

On the boiler side it means proving the call for heat electrically. Is there a live at the cylinder thermostat, and does the stat switch when the dial is turned up? Is that switched live arriving at the valve? Does the valve drive fully open, and does its microswitch make and send a signal to the boiler? Each of those is a measurement at a terminal rather than an opinion, and the one that breaks the chain is the fault. Where the appliance itself is locking out rather than the hot water leg failing, that is boiler diagnosis rather than cylinder work, and a lockout such as Vaillant F28 points at ignition rather than at anything in the airing cupboard.

If the electrics prove sound and the boiler is firing with the pump running, the last suspect is heat transfer itself. Flow and return on the coil get checked for temperature difference. A coil taking hot water in and sending it back almost as hot is not giving up its heat to the tank, which means scale on the coil, sludge in the primaries, or a cylinder that has reached the end of its working life. That is also the point where the conversation turns from repair to replacement, and it should be based on those readings rather than on the age of the tank.

When the tank needs replacing rather than repairing

Sometimes the heat source is fine and the cylinder is the problem. The giveaway is a tank that heats, but holds almost nothing: hot for one shower and then cold, with no change in how anyone uses it. That is usually volume lost to scale and debris settled in the bottom, or a coil so encrusted that the effective heat transfer area has collapsed.

The other end of life signal is water where there should not be any. A weep from the boss around an immersion head can be a perished gasket and is repairable. Damp that appears from the seam of the cylinder itself, or rust coloured staining running down the case, means the vessel has failed, and a failed vessel is not something anyone patches. Unvented cylinders in particular are pressure vessels, and a corroded one gets replaced rather than repaired.

Age alone is not a condemnation. A well fed cylinder in a soft water area can run for decades, while a heavily scaled one in London can be finished far sooner. The deciding factors are recovery time, how much usable hot water you actually get, whether the coil still transfers heat, and whether the vessel is sound. Four measurements, not a birthday.

Where a replacement is the answer, the useful conversation is about sizing and heat source rather than brand. A household that has grown, or a bathroom that has gained a decent shower, often needs more stored volume than the original installer allowed for, and getting that wrong is the reason so many new cylinders still run out. Weighing a repair against a replacement comes down to recovery time, usable stored volume, whether the coil still gives up its heat and whether the vessel is sound, which are four readings an engineer can take in a single visit rather than four opinions about the age on the label.