Immersion Heater Not Working

Before anything else: is the immersion actually getting power?

An immersion heater is a 3kW electric element screwed into the side or top of your hot water cylinder, sitting in the water. It draws around 13 amps, so it is almost always fed from its own circuit in the consumer unit, through a double pole switch near the airing cupboard, with a heat resisting flex running from that switch down to the element head. Every one of those links can fail and give you exactly the same cold tank that a dead element does.

Start at the consumer unit. Look for a breaker or an RCBO that has dropped, and for any fuse carrier on an older board. An immersion circuit is often labelled, but on a board that has been added to over the years the label may belong to something else entirely, so check every device rather than the one that looks right. If a breaker has tripped, that is information in itself: something drew a fault current, and simply resetting it without knowing what is not a plan.

Then find the immersion switch. It is a large white double pole switch, usually with a neon, mounted on the wall by the cylinder cupboard. Some properties have two, marked bath and sink, or boost and normal. Make sure the one you need is actually on.

Finally, look for a fused spur. Some installations feed the immersion through a spur with a cartridge fuse, and a blown fuse there kills the element while leaving the breaker happily in place. A spur with a neon that is dark, on a circuit whose breaker is up, is a strong hint.

• Breaker or RCBO at the consumer unit: is it up, and has it been tripping repeatedly?
• Double pole immersion switch: on, and the correct one if there are two
• Fused spur, if fitted: fuse intact, neon lit
• Timer or programmer, if the cylinder is also heated by the boiler
• Flex from switch to element head: discoloured, hardened or hot to the touch

If the immersion has tripped the breaker more than once, leave it switched off and isolated rather than resetting it again. Repeated resets push fault current through the earth path each time, and on an older installation that is not something to keep doing while you wait for someone to look at it.

Immersion heater stopped working but the switch neon is lit

A lit neon tells you less than people assume. It confirms that voltage has arrived at the switch, and on most switches it confirms that the switch is in the on position. It does not confirm that both poles have made contact, that current is reaching the element, or that anything downstream is intact.

Double pole immersion switches fail in a very particular way. They carry 13 amps continuously for hours at a time, year after year, and the contacts slowly pit and carbonise. Eventually one side stops making a reliable connection. The neon, which is usually wired across the incoming side, carries on glowing while the element gets nothing. The giveaway is a switch faceplate that feels warm, a faint smell of hot plastic near the cupboard, or a brown bloom around the rocker.

None of this can be settled by looking. Proving whether voltage reaches the element head means removing a cover on a circuit that has to be isolated and tested dead first, and that is the point at which the checks stop being homeowner checks.

Immersion heater on but no hot water: timers and off-peak

If the property is on Economy 7 or a similar off-peak tariff, the main immersion element is often wired through a time switch or a separate off-peak meter supply, and it is only live during the cheap night window. Water heated at two in the morning and used through the day will run cool by late afternoon, and switching the immersion on at that point may do nothing at all, because the circuit is simply not energised until the window opens again. Properties on this arrangement usually have a second, shorter boost element in the top of the cylinder on the normal rate, precisely so you can get a tank top up during the day.

That also explains one of the most common complaints, which is that the immersion works but there is hardly any hot water. A short boost element in the top boss only heats the water above it, which is roughly the top third of the cylinder, and that is what it was designed to do. If you are getting a small amount of genuinely hot water rather than a lukewarm tank, the boost element is alive and the lower element is the one that has failed.

Where the cylinder is also heated by the boiler, check the programmer as well. A hot water channel that was turned off for the summer, or that lost its schedule during a power cut and reverted to an all day off setting, produces a cold tank with nothing wrong anywhere. Old mechanical time switches with a rotating dial also drift, and a dial that has slipped several hours means the tank heats at the wrong time of day. If the radiators are hot while every tap in the house is cold, the immersion is not where this starts, and the page on no hot water but the heating works sorts a boiler side fault from a cylinder side one before anyone takes a cover off.

• Off-peak circuit dead during the day is normal, not a fault
• Boost element heats the top third of the tank only, by design
• Programmer hot water channel off, or reset to a default after a power cut
• Mechanical timer dial slipped so the tank heats at the wrong hours
• Holiday or summer setting left on from months ago

Is it the immersion thermostat or the element?

Under the round plastic cap on the element head there is a rod thermostat, pushed down a dry pocket in the centre of the element. The rod is made of two metals that expand at different rates, and as the water heats the rod grows and opens a set of contacts at whatever temperature the dial is set to. When the water cools, the contacts close again and the element switches back on. That is the whole mechanism.

A thermostat that has stuck open never lets the element switch on. You get no heat at all, no noise from the cylinder, no trip at the consumer unit, and an element that tests perfectly good. This is the outcome worth hoping for, because the stat is a separate part that slides out of its pocket and can be changed without draining a drop of water out of the cylinder.

A thermostat that has drifted low is the cause of most lukewarm tanks. The dial says 60 but the contacts open at 45, the water tops out short of properly hot, and every bath feels like it has been rationed. People assume the element is weak. Elements do not get weak; a 3kW element either delivers 3kW or it delivers nothing. Water that is warm but not hot is a control fault, not an element fault, and that one distinction saves a lot of unnecessary draining. If the water is getting warm rather than nothing at all, the page on water not getting hot works through the other causes of a lukewarm supply, including the ones that have nothing to do with the immersion.

A thermostat that has stuck closed is the one to take seriously. The element never switches off, the water keeps climbing, and the only thing standing between you and a tank of scalding water is the thermal cut-out. Symptoms are water far too hot at the taps, a cylinder that never seems to stop heating, and on an unvented system a tundish that starts to discharge. If your hot water has gone scalding, turn the immersion off at the wall and leave it off.

Setting the thermostat around 60 degrees is the sensible compromise. Hot enough that the stored water is safe, cool enough that limescale forms on the element more slowly. Turning it to maximum does not give you meaningfully more usable hot water and in London water it will shorten the life of the element noticeably.

The immersion heater reset button has tripped: what that means

Alongside the thermostat in the same head sits a thermal cut-out, a one-shot safety device that disconnects the element if the water passes roughly 85 degrees. It has a small reset button, usually red, and it stays tripped until somebody presses it.

A tripped cut-out gives you a completely dead immersion. No heat, no noise, breaker up, switch on, element in good order. It looks identical to a stuck thermostat from the outside, and the two are told apart by testing rather than guessing.

The important point is that a cut-out does not trip for no reason. It trips because the water got too hot, which usually means the thermostat failed closed, or because the element was energised with the cylinder low on water, or because a control fault let the element run continuously. Pressing the reset and walking away puts the same set of circumstances straight back into service, and the next time it may be the tank that has to deal with it.

Resetting a cut-out means taking the cap off the element head. Inside that cap are terminals live at mains voltage, within easy reach of a finger, and often with a flex whose insulation has been baking next to a hot cylinder for a decade. That cover comes off after the circuit is isolated at the consumer unit and proved dead with a tester, not after the wall switch has been flicked off. This is the point where the sensible answer is to stop and get someone with a meter.

Do not press the cut-out reset on a live circuit, and do not keep resetting one that trips again. A cut-out that trips repeatedly is reporting a real overheat, and the fault causing it has to be found and fixed rather than overridden.

Immersion heater not heating water: how the element is proved dead

If the supply is good, the switch is good and the controls are good, what remains is the element, and it gets condemned on two measurements rather than on an opinion.

The first is continuity across the two element terminals. A healthy 3kW element on a 230 volt supply reads somewhere close to 18 ohms, because resistance is voltage squared divided by wattage. A reading in the high teens means the resistance wire inside the sheath is intact end to end. If the meter reads open circuit, the wire has burnt through, and there is nothing to be done about it. The element is a sealed sheathed unit packed with magnesium oxide, crimped at the factory, and no part of it can be opened, cleaned or reconnected.

The second is insulation resistance, measured between a terminal and the metal body of the element with an insulation tester rather than an ordinary multimeter. A sound element reads effectively infinite. One that reads in the low megohms has water in the packing and is leaking current to earth, and that is the element that has been tripping your RCD. An element can pass continuity and fail insulation, which is exactly why both tests are done and why an element is never condemned on one.

• Around 18 ohms across the terminals: element intact
• Open circuit across the terminals: element gone, replacement only
• Infinite insulation to the body: element electrically sound
• Low insulation to the body: element leaking to earth, replacement only
• Element good, no heat: fault is the thermostat, the cut-out or the wiring

Immersion heater tripping the electrics when it is switched on

A breaker or RCD that drops the instant the immersion is switched on is the classic signature of an element that has failed to earth. Water has worked its way into the magnesium oxide inside the sheath, usually through a pinhole caused by corrosion or by a hot spot under heavy scale, and the leakage to the metal body is exactly what an RCD exists to detect. Once that has happened the element is finished. Drying it out does not bring it back.

It is worth knowing that a trip is not always the element. Water tracking down inside the terminal cap, from a weeping head gasket or from a drip landing on the head from somewhere higher up, will trip an RCD just as convincingly. So will a flex that has been pinched against the hot boss and worn through to the conductor. Both of those are repairable without a new element, which is another reason for the insulation test before anything is thrown away.

If switching the immersion on trips the whole consumer unit rather than just its own device, leave it isolated and mention that when you call. That behaviour points at the earthing or RCD arrangement itself as much as at the immersion, and it wants looking at properly rather than being worked around.

Why immersion elements fail early in hard water

Most of London draws its water from chalk aquifers, and that hard water is the single biggest reason immersion elements here do not last as long as the same part in a soft water area. Calcium stays dissolved while the water is cold and comes out of solution as limescale the moment it touches a surface above roughly 60 degrees. The hottest surface in your cylinder, by a wide margin, is the element sheath.

Scale on an element is an insulating blanket. The element was designed to dump its heat straight into the water, and a furred sheath stops it doing that. The wire inside carries on producing the same 3kW, the heat cannot escape at the same rate, the core temperature climbs, and eventually a hot spot destroys the packing. That is when the element either goes open circuit or starts leaking to earth. It is the whole life story of a London element.

You can usually hear it coming before it fails. A scaling element crackles and kettles as it heats, which is water flashing to steam in the pockets under the scale and collapsing again. A cylinder in the airing cupboard that has started to sound like a kettle is telling you the element is furred and living on borrowed time. Scale also flakes off and settles in the base of the tank, which is why draining an old cylinder in a hard water area often produces a startling amount of grit. Our longer piece on hard water and hot water cylinders in London covers what that scaling does to a cylinder over its whole life, not just to the element in it.

Hot water tank not heating when the boiler is meant to heat it

Plenty of people searching for a dead immersion actually have a cylinder that should be heated by the boiler, with the immersion sitting there as the backup. If the radiators are working and the tank is cold, the immersion is not the problem and never was.

On a conventional system the boiler sends hot water through a coil inside the cylinder. That circuit is controlled by the programmer calling for hot water, a cylinder thermostat strapped to the side of the tank deciding whether it is satisfied, and a motorised valve actually opening the path. Any one of those breaking gives you a cold tank and warm radiators, and they are told apart by checking which of them is calling and which is responding. Two port valves fail in a recognisable way: the small motor burns out, the valve stays shut and the boiler never sees a call for hot water, and on the ones with a manual lever, moving it by hand suddenly produces hot water.

If your heating works and only the hot water has gone, that narrows things considerably and is worth saying on the phone, because it changes what the engineer brings. We have separate pages for a hot water tank not heating and for no hot water when the heating works, and both go through the valve and thermostat checks in more detail than belongs here.

Which of these checks are safe to do yourself, and which are not

The line here is not where most people assume it is. The safe checks all involve looking, switching and listening. The unsafe ones all involve taking a cover off something that may still be live.

Everything on the supply side is fair game: the breaker, the switch, the spur fuse if it is a plug in cartridge you can inspect, the programmer, the timer, and which element is meant to be live at this hour. So is listening to the cylinder while it should be heating, and feeling whether the tank is warm at the top and cold at the bottom, which tells you a top element is working and a bottom one is not. So is turning the immersion off and leaving it off when something is clearly wrong.

What is not a homeowner job is the element head. The cap over the terminals has mains voltage behind it, and the circuit needs isolating at the consumer unit and proving dead with a tester before it comes off. That rules out pressing the cut-out reset, checking the thermostat setting, tightening a terminal or inspecting the flex where it enters the head. It also rules out testing the element, since insulation resistance needs an instrument that puts hundreds of volts through the circuit.

• Safe: check the breaker, the switch, the spur, the timer and the programmer
• Safe: feel the cylinder top and bottom to see what is heating
• Safe: switch the immersion off and leave it off if it trips or runs scalding
• Not safe: removing the element head cap for any reason
• Not safe: pressing the thermal cut-out reset on a live circuit
• Not safe: unscrewing an element from a cylinder boss

Never switch an immersion on if the cylinder has been drained or may be part empty. An element run dry destroys itself in seconds and can distort the boss it is screwed into. If anyone has drained the tank, the immersion stays off until it is refilled and vented.

What an immersion heater repair visit involves

The first job is deciding whether this is a supply fault, a control fault or an element fault, because the answer determines everything afterwards. That means isolating the circuit, proving it dead, opening the head and taking readings rather than forming an opinion from the outside.

The tests take a few minutes. Continuity and insulation resistance on the element, continuity across the thermostat contacts hot and cold, the state of the thermal cut-out, and a look at the terminals and the flex, because burnt connections at an element head are a fire risk rather than an inconvenience.

Whatever was changed, the last thing that happens is a full heating cycle with the engineer still in the property: switch on, confirm the element is drawing current, confirm the tank heats, confirm the thermostat cuts out at temperature, confirm nothing trips and nothing weeps at the boss. Leaving before that is how a job fails the same evening. Our engineers are qualified for the work described and will say plainly if what they find on the day changes the picture. What an immersion heater repair covers, from an element swap to a burnt switch and flex, is set out on the service page.

The rate is fixed before we attend and the total depends on how long the job takes, so a thermostat swap and a seized element on a fifteen year old cylinder are not the same visit and are never presented as if they were.