5 boiler leaks not caused by
a defective new gasket
Sometimes the gasket was never the problem. Water travels before it drips — the puddle on the floor beneath the boiler rarely marks the source. Here are five leaks that often get misdiagnosed as gasket failure, and what is actually happening in each case.
Safety first: in the UK, work on gas appliances and any work inside a boiler casing must only be carried out by a Gas Safe registered engineer or otherwise competent person where the law and manufacturer instructions allow it. This article explains diagnostic patterns. It is not an instruction to open or work on a live boiler.
Before assuming gasket failure: a competent person should confirm where moisture first appears under safe, controlled conditions. Water can travel along pipes, fittings and casing surfaces before it drips, so the visible puddle is not always the source.
This article is not a boiler repair guide. Its purpose is to stop gasket joints being blamed when the leak source is elsewhere, so the correct gasket is only replaced when the joint is actually the failure point.
The union is dripping. The joint is opened, the gasket is replaced and the assembly is put back into service. It drips again within days or weeks. The gasket gets blamed — wrong size, wrong material, wrong brand. So another one is tried.
The gasket is not the problem. The sealing face is.
Years of hard water cycling through a heating system deposits calcium and magnesium scale on every surface that hot water contacts. On a flat union face, scale builds up in an uneven ring — sometimes visibly, more often as a thin hard layer that looks like bare metal until you scrape it. The gasket compresses against this ring rather than against a flat surface. One section of the gasket is over-compressed, the rest is not compressed enough. The seal fails at the gap.
A new gasket on a scaled face is a temporary repair at best. The face must be clean, flat and free of contamination before the gasket goes in. Sealant should not be relied on to compensate for a contaminated face.
How to confirm it
Remove the old gasket and examine the union face under good light. Run a fingernail across the surface. Any roughness, raised ridge or white residue is scale. A clean face feels uniformly smooth.
The fix
The sealing face should be cleaned and inspected by a competent person before a new gasket is fitted. The face should be free of scale, residue and raised contamination so the gasket seats on the metal face, not on deposits.
Water appears around the pump body — not at the union connections, but from the pump itself. It is not a lot. It comes and goes with system temperature. It looks like an internal seal or O-ring has gone.
Sometimes it is. The possible source depends on pump construction. A wet-rotor circulator may have no conventional mechanical shaft seal; check the union faces, vent, head-to-body seal and housing. On pumps that do use a shaft seal, contaminated water can contribute to wear.
Black magnetite sludge — iron oxide from corroding radiators and pipework — circulates in unprotected heating systems. It deposits on pump impellers, in bearing surfaces and around shaft seals. A shaft running in abrasive sludge wears the seal surface. The seal starts to weep. You can replace the seal, but if the system is not flushed and protected with inhibitor, the replacement seal is likely to fail for the same reason.
Replacing the pump entirely and ignoring the sludge problem is an expensive short-term fix. The new pump runs in the same contaminated water and is likely to fail for the same reason on a similar timescale.
How to confirm it
Drain a small amount of system water into a clear container. Dark or black water can indicate magnetite contamination, but water colour alone does not fully prove system condition — interpret it together with filter condition, sludge history and system water testing.
The fix
The system water quality should be assessed before the pump seal or pump is replaced. Depending on condition, the corrective work may include system cleaning or flushing, inhibitor replacement and fitting or servicing magnetic filtration.
The PRV is dripping — or occasionally discharging a short burst of water down the overflow pipe. The obvious diagnosis is that the valve seat or gasket has failed. The PRV is replaced. It drips again within weeks.
A PRV that drips regularly is usually telling you that system pressure is periodically exceeding its set point. The valve is doing exactly what it is supposed to do — it opens at overpressure and closes when pressure drops back. It is not failing. It is working.
The most common cause is an expansion vessel that has lost its air pre-charge. The vessel is supposed to absorb the pressure increase as the system heats up and water expands. When the vessel diaphragm fails or the pre-charge bleeds down, there is nothing to absorb that expansion. Pressure spikes on every heat cycle. The PRV opens. The valve gets blamed.
Check the manufacturer's specified cold-fill pressure and then observe the rise toward the PRV set point during heat-up. A cold reading of 1.5 bar is not, by itself, proof of an expansion-vessel fault. Expansion vessel pressure and condition should be checked by a competent person using the boiler manufacturer's procedure. Do not assume that replacing the PRV will solve repeated discharge if system pressure rises sharply during heat-up.
How to confirm it
Note the cold fill pressure. Note the pressure when the system reaches working temperature. If pressure rises significantly above 1.5–2 bar on heating, the expansion vessel is not absorbing correctly. In a typical domestic system, a correctly functioning vessel shows relatively little pressure rise between cold and hot — compared to a faulty vessel where pressure climbs significantly on each heat cycle.
The fix
The expansion vessel should be recharged or replaced where testing confirms loss of pre-charge or diaphragm failure. The correct pre-charge must come from the boiler or vessel documentation, not a generic value. Then assess whether the PRV seat itself needs replacement after the root cause is resolved.
A component is removed for service — a pump, a heat exchanger, a valve. When it goes back together, the joint leaks. The diagnosis: the gasket must have been damaged during disassembly, or the wrong size was fitted.
The more likely cause: the original gasket was put back.
A gasket that has been in service under compressive load for any significant time has permanently deformed to the shape of that joint. The material has compressed and not recovered. When you remove it, it looks intact. It is flat, in one piece, no visible cracks. So it goes back in.
It will not seal reliably. The geometry of the compressed gasket no longer matches the face geometry. It seals unevenly, holds temporarily under hand-tight assembly and starts to weep within the first heat cycle. The joint leaks. The gasket gets blamed. The gasket was not the wrong part — it had already done its job and should not have been reused.
Removed intact does not mean reusable. In normal service reassembly, a compressed flat gasket should be replaced unless the equipment manufacturer and governing procedure explicitly allow reuse. A new gasket costs less than the labour to open the joint again.
How to confirm it
Look at the recovered gasket. A gasket that has been in service will be noticeably thinner at the contact ring than at the outer edge — compression set is visible as a step or groove where it bore against the face. That gasket is finished.
The fix
In normal service reassembly, use a new flat gasket unless the equipment manufacturer and governing procedure explicitly allow reuse. Service stock should include the common gasket sizes needed for planned reassembly work. The cost of carrying spares is negligible against the cost of a second call-out to replace a gasket that should have been replaced the first time.
Water appears under a condensing boiler — on the floor, at the base of the casing, pooling beneath the unit. The instinct is to open the boiler and look for a leaking joint or connection inside. Nothing obvious appears.
On condensing boilers, the flue gas condenses as it cools and produces acidic condensate water that drains from the heat exchanger through a condensate trap. This trap and its associated drain pipework sits at or near the base of the boiler. If the condensate drain is blocked, kinked, frozen in winter, or incorrectly sloped, the water backs up and overflows — onto the boiler floor, down the outside of the casing, and onto the floor beneath.
It looks exactly like an internal water leak. It is not water from the heating circuit at all — it is flue condensate. A pH strip can help: condensate is usually acidic, often around pH 3–5, while treated heating water is typically closer to neutral or mildly alkaline, depending on inhibitor and system condition. A pH strip is a clue, not the only diagnostic test.
A frozen condensate pipe is one of the most common causes of boiler lockout in cold weather — and the puddle it produces when it thaws gets blamed on everything except the condensate drain.
How to confirm it
A pH strip can help identify condensate — usually acidic, often around pH 3–5, while treated heating water is typically closer to neutral or mildly alkaline. Treat pH as a clue alongside other evidence. Condensate also has a faint acidic odour; heating water smells of inhibitor or is odourless.
The fix
The condensate drain should be inspected for blockage, freezing, incorrect fall or poor external protection by a competent person. Corrective work should address the drain route, slope and freeze protection where relevant. Condensate drain issues are a boiler service item — the root cause is drain management, not the heating circuit.
The diagnostic rule that saves time for a competent technician
A competent technician usually starts with observation before replacing parts — not with assumptions, and not with part numbers.
- Confirm the area is dry. Every surface, every joint, every pipe run — checked with tissue or cloth, not just a look.
- Observe at cold working pressure where safe and permitted. The first point of moisture is usually the source.
- Trace moisture upward from the visible drip. A drip on the floor came from above. A drip on a pipe came from a fitting above that point.
- Use liquid identity as a clue. Condensate is usually acidic; treated heating water is closer to neutral. A pH strip costs pennies and saves diagnostic time.
- Check system pressure behaviour before opening joints. If the leak appears or worsens when the system heats up, check the expansion vessel and system pressure first.
The most expensive diagnostic mistake is replacing a component before confirming the root cause. A new gasket on a scaled face leaks again. A new pump in contaminated water fails again. A new PRV on a system with a failed or uncharged expansion vessel may drip again. The component was not the problem. The condition was.
If a joint still leaks after gasket replacement, the root cause has not yet been demonstrated. Recheck leak origin, dimensions, material, face condition and assembly before replacing another part.
Scale on the face. Reused gasket. Shaft seal in contaminated water. Overpressure from a failed or uncharged expansion vessel. Condensate backing up from a blocked drain. Five different causes. Five different fixes. One thing in common: none of them are solved by fitting another gasket and hoping.