Banging and Hammering Pipes: Water Hammer, Expansion Noise and the Leaks They Cause Later

A pipe that bangs is telling you something measurable. Water hammer is a pressure surge, expansion noise is thermal movement, and a rattle is a missing clip. Only one of the three will eventually split a joint.
What water hammer physically is
Water hammer is not a metaphor. It is a pressure transient produced when a moving column of water is stopped abruptly, and it obeys arithmetic that has been understood since the nineteenth century.
Water is effectively incompressible. When a valve closes quickly, the water immediately upstream cannot slow down gradually; its momentum has to go somewhere, and it converts into a pressure spike that travels back along the pipe as a wave at close to the speed of sound in water, reflects off the next change of section, and returns. The audible bang is that wave arriving at a fixed point, usually a bend, a tee or a clip, and moving the pipe against the structure. The wave then oscillates and decays over a fraction of a second, which is why the noise often has a short ringing tail after the initial crack.
The magnitude of the spike depends on the velocity of the flow before closure and on how fast the closure happens, not on the static pressure in the system. That is the counterintuitive part and it explains why a house at modest mains pressure can still hammer badly: a long straight run feeding a fast solenoid valve produces high velocity and near-instant closure, and both terms are in the calculation.
Why a fast-closing valve is almost always the trigger
Traditional taps close on a screw thread over a second or more, which gives the column time to decelerate. Modern fittings do not. A washing machine or dishwasher solenoid closes in milliseconds. A quarter-turn ceramic disc tap closes in a quarter of a wrist movement. A single lever mixer slammed shut does the same. A thermostatic shower valve shutting on a temperature excursion can do it without anyone touching it.
This is why the complaint is almost always tied to a specific appliance or a specific tap, and why the first diagnostic question is not what the noise sounds like but what was being used at the moment it happened.
The four noises, and how to separate them
Four distinct phenomena get reported as banging pipes. They have different timings, different triggers and different remedies, and confusing them is the reason so many households end up with an arrestor fitted to a problem an arrestor cannot fix.
| Noise | When it happens | Cause | Remedy family |
|---|---|---|---|
| Single sharp bang, sometimes with a short ring | The instant an outlet or appliance valve closes | Water hammer, a genuine pressure surge | Arrestor, pressure reduction, slower closure |
| Repeated knocking or drumming while water runs | During filling, not at the end | Valve chatter, usually a worn or oversized float valve | Replace or adjust the fill valve |
| Ticking, creaking or a slow series of clicks | Minutes after heating starts or stops | Thermal expansion and contraction against a fixed point | Free the pipe at the constriction, sleeve where it passes through |
| Rattle or buzz that follows the flow | Whenever a particular outlet is open | Unclipped pipe vibrating against a joist or a board | Clip properly, at correct centres |
Expansion and contraction ticking
Copper expands by roughly the same order as most building metals, which over a long heating run amounts to several millimetres between cold and full flow temperature. Where a pipe passes through a tight notch in a joist, through a hole drilled fractionally undersized, or under a clip done up hard, that movement happens in jerks rather than smoothly, and each jerk is a tick. The signature is timing: it starts after the heating comes on, not at the moment a valve moves, and it fades as the system reaches steady temperature. It is a nuisance rather than a risk, and the fix is to relieve the constriction and sleeve the pipe where it passes through timber.
Loose clipping and pipes touching structure
An unclipped horizontal run will move under the reaction force of flow starting and stopping, and will buzz continuously while water passes. In floor voids the noise is amplified by the boards above. This is the easiest of the four to confirm: it correlates with flow rather than with valve closure, it is continuous rather than impulsive, and pressing on the floor above usually changes it. Schedule 2 of the Water Supply (Water Fittings) Regulations addresses the underlying requirement, and Water Regs UK guidance on installing pipework puts it directly: pipework should be appropriately clipped to avoid water hammer and other stresses that would affect the integrity of the installation.
Failing valves and fill mechanisms
A worn valve produces its own family of noises that are neither hammer nor expansion, and it is worth ruling these out early because the parts are cheap.
Float valves and appliance solenoids
A diaphragm float valve in a WC cistern or a cold water storage tank that is worn, or that is a poor match for the supply pressure, will chatter as it approaches closure: the diaphragm partially closes, pressure recovers, it reopens, and it oscillates audibly for several seconds. The sound is a rapid knocking during filling rather than a single crack at the end, and it usually appears alongside a slow or noisy fill. Replacing the valve, or fitting one with a pressure rating matched to the supply, ends it. Our guide to fixing a leaking toilet cistern covers the same components from the leak side.
Appliance solenoids are the opposite problem. They work correctly and close so fast that they create hammer by design. Nothing is wrong with the valve; the remedy belongs on the supply to it, not in it. This is the single most common origin of hammer in a modern flat, and it is also the easiest to treat because the point of application is known and accessible.
Why it matters: the damage hammer does
Left alone, water hammer is not merely irritating. Each surge loads the system above its working pressure, and the loading is cyclic, which is the regime that produces fatigue rather than immediate failure. The consequence is rarely a burst on the day. It is a joint that weeps eighteen months later in a location nobody connects to the noise.
- Compression joints loosen incrementally as the olive is worked against the fitting, which is why hammer damage shows up first at the accessible joints under sinks and behind appliances.
- Soldered joints on runs with a hard fixed point can develop fatigue cracks at the fillet, particularly where the pipe is also restrained from moving.
- Clips and brackets loosen, which increases movement, which increases the load on the joints: the fault accelerates itself.
- Flexible connectors and appliance hoses are stressed at the crimp, and these fail suddenly and at full bore rather than as a weep.
- Float valves, ballvalves and cistern components wear faster under repeated surge, which produces overflow and the drip that follows.
- Meters and pressure-sensitive devices can be affected, and any check valve on the supply is cycled every time.
| Component | How surge damages it | How it shows up later |
|---|---|---|
| Compression joints | Cyclic loading works the olive loose | A slow weep at a joint that was sound for years |
| Soldered joints | Fatigue at the fillet where the pipe is restrained | A pinhole or hairline crack, often in a void |
| Push-fit fittings | Grab ring bedding, O-ring displacement | Intermittent seepage under pressure only |
| Appliance hoses | Stress at the crimped ferrule | Sudden full bore failure, usually unattended |
| Cistern and float valves | Accelerated diaphragm and seat wear | Overflow, continuous fill, wasted water |
| Clips and brackets | Fasteners loosen under repeated impulse | Increasing noise, then greater joint movement |
Because the eventual failure is displaced in time and place from the noise, hammer damage is frequently recorded as a random leak. Where a property has hammered for a long period and has now developed a weep, the two are worth treating as related, and the diagnostic question becomes which joints on the affected run are accessible. Our note on pinhole leaks in copper pipe covers the failure mode this most resembles.
Diagnosing it properly before spending anything
Almost all of this can be established in twenty minutes with no tools beyond a pressure gauge that screws onto an outside tap or a washing machine connection.
- Establish the trigger. Operate each outlet and each appliance in turn and note which one produces the noise and at what point in its cycle. A bang at closure is hammer; a knock during fill is a valve; a tick after the heating fires is expansion.
- Measure the static pressure with everything closed, and again with an outlet running. A large gap between the two indicates high velocity in the supply, which is one of the two terms that sets the size of the surge.
- Check whether the property already has a pressure reducing valve, and whether it is doing anything. Many flats have one fitted at handover and never checked since.
- Check for a check valve or a meter with a non-return function on the incoming supply. These create a closed system, which removes the mains as a place for the surge to dissipate and makes hammer markedly worse.
- Look at the run itself where it is accessible: clip spacing, contact with joists, long unsupported horizontal spans and unrestrained changes of direction.
- Note whether the noise is new. Hammer that has started recently on an unchanged system usually means something else changed, most often a new appliance, a replaced tap or a new valve on the supply.
For context on the pressure figures, the minimum service standard water companies work to is expressed as a static head at the boundary rather than as a comfortable working pressure, so a property can be well inside the standard and still have velocity high enough to hammer. Our guide to high water pressure, signs and fixes sets out what the numbers mean in practice, and low water pressure causes covers the opposite complaint, which occasionally has the same root.
The remedies that actually work
There are only four real interventions, and the order matters. Fitting an arrestor to a system with an unchecked pressure problem treats the symptom and leaves the loading in place.
| Remedy | What it fixes | What it does not fix | Reference standard |
|---|---|---|---|
| Pressure reducing valve on the incoming supply | Reduces the working pressure and the flow velocity behind every outlet | Poor clipping, expansion noise, worn valves | BS EN 1567 for water pressure reducing valves |
| Anti-water-hammer arrestor near the offending valve | Absorbs the surge at source with a sealed air or spring chamber | Surges originating elsewhere on the system | BS EN 14451 for devices preventing water hammer |
| Re-clipping and sleeving the run | Rattle, buzz and expansion ticking | The pressure surge itself | Schedule 2, Water Supply (Water Fittings) Regulations 1999 |
| Replacing or adjusting the offending valve | Chatter, and closures fast enough to cause the surge | System-wide high pressure | Manufacturer pressure rating for the fitting |
A word on arrestors, because they are sold as a universal answer. A modern arrestor is a sealed chamber with a piston or bellows separating water from a gas cushion, and it works by giving the pressure wave something compressible to push against. It must be fitted close to the valve causing the surge, on the same branch, to do its job; fitted at the other end of the property it absorbs very little. Older installations sometimes relied on a plain vertical capped pipe as an air pocket, which works initially and then stops as the air dissolves into the water. If a property has one of those and hammer has returned, the pipe can often be recharged by draining down, but a proper sealed device is the durable answer.
Pressure reduction is the intervention with the broadest benefit, because it lowers the velocity term for every outlet at once and reduces the standing load on every joint, fitting and appliance in the building. Where a pressure reducing valve is fitted, it should be a product conforming to BS EN 1567 and it should be approved for contact with wholesome water; the WRAS approvals directory is where that is checked. The same applies to arrestors.
What the Water Regulations actually require
Noise is not merely a comfort question in this context; it is evidence of a mechanical loading that the Regulations address directly.
The Water Supply (Water Fittings) Regulations 1999 apply to anyone installing, altering, disconnecting or using a water fitting connected to a supply from a water undertaker, and to anyone who causes or permits it. Regulation 3 prohibits doing so in a way that causes or is likely to cause waste, misuse, undue consumption or contamination, and prohibits leaving a damaged, worn or faulty fitting in use where it creates those risks.
Schedule 2 sets the physical requirements that bear on surge. Paragraph 3(b) requires every water fitting to be constructed of materials of such strength and thickness as to resist damage from any external load, vibration, stress or settlement, pressure surges, or temperature fluctuation to which it is likely to be subjected. Paragraph 5 requires every water fitting to withstand an internal pressure of not less than one and a half times the maximum pressure to which it is designed to be subjected in operation, and paragraph 12 applies the same one and a half times requirement to the water system as a whole. Paragraph 11 requires an adequate number of servicing valves and drain taps so that parts of the system can be isolated and drained, which is what makes any of these remedies practical to install.
Two practical conclusions follow. The first is that persistent hammer in a system is an indication that the installation is not meeting the resistance requirement in paragraph 3(b), and should be treated as a defect rather than a quirk. The second is that any work done to cure it is itself work on water fittings, and should be carried out by someone competent to certify it; Thames Water and the other undertakers direct customers to WaterSafe approved plumbers and ask for a certificate of water regulations compliance on completion. Where the noise has already been followed by a weep somewhere in the building, the sequence set out in finding a leak behind a wall is the right next step.
How we help with this
If the article describes a problem you actually have, these are the visits that deal with it.
- acoustic leak detection
Ground microphones and correlators on a pressurised pipe.
- leak detection in London
Finding a hidden leak without opening the property up first.
- central heating leak detection
For a boiler losing pressure or a circuit that will not hold.
Frequently asked questions
What is water hammer, in physical terms?
A pressure transient caused by stopping a moving column of water abruptly. Water is effectively incompressible, so when a valve shuts quickly the momentum of the flow converts into a pressure spike that travels back along the pipe at close to the speed of sound in water, reflects and returns. The bang you hear is that wave moving the pipe against a fixed point. The size of the spike depends on the flow velocity before closure and on how fast the closure happens, not on the static pressure alone.
How do I tell water hammer from expansion noise?
By timing. Water hammer is a single sharp bang at the exact instant an outlet or appliance valve closes, sometimes with a short ringing tail. Expansion noise is a tick, creak or slow series of clicks that begins a few minutes after the heating starts or stops and fades as the system reaches steady temperature, caused by a pipe moving in jerks against a tight notch or an over-tightened clip. A rattle that runs continuously while water flows is neither; it is an unclipped pipe.
Does water hammer actually cause leaks?
Yes, but rarely on the day. Each surge loads the system above working pressure, and the loading is cyclic, which produces fatigue rather than immediate failure. Compression joints work loose as the olive is repeatedly stressed, soldered joints can crack at the fillet where the pipe is restrained, push-fit O-rings displace, and appliance hoses are stressed at the crimp. The result appears months later as a weep or a pinhole somewhere nobody connects to the noise, which is why the two are often recorded separately.
Will a water hammer arrestor fix the problem?
Only if it is fitted in the right place and the pressure is not the underlying issue. An arrestor is a sealed chamber with a piston or bellows separating water from a gas cushion, and it absorbs the surge by giving the pressure wave something compressible to push against. It has to be fitted close to the valve causing the surge, on the same branch; installed at the far end of the property it does very little. Products should conform to BS EN 14451 and be approved for use on wholesome water.
Is reducing the pressure better than fitting an arrestor?
Usually, because it treats the cause rather than one symptom. A pressure reducing valve on the incoming supply lowers the flow velocity behind every outlet at once and reduces the standing load on every joint, fitting and appliance in the building, whereas an arrestor addresses a single branch. Pressure reducing valves conforming to BS EN 1567 and approved for wholesome water are the relevant product. In many properties the right answer is both: reduce the pressure, then arrest the one fast-closing appliance that still surges.
Do the Water Regulations say anything about banging pipes?
Yes, indirectly but clearly. Schedule 2 of the Water Supply (Water Fittings) Regulations 1999, paragraph 3(b), requires water fittings to be constructed of materials of such strength and thickness as to resist damage from external load, vibration, stress or settlement, pressure surges or temperature fluctuation. Paragraphs 5 and 12 require fittings and the system to withstand one and a half times the maximum working pressure. Persistent hammer indicates the installation is not meeting that resistance requirement, and is a defect rather than a quirk.