Precision Underground Diagnostics

Acoustic Leak Detection London

Acoustic work is the quickest way to put a point on a floor when the conditions suit it, and those conditions are narrow enough to be worth stating plainly. The run has to be under pressure, the pipe has to carry sound, and the building has to be quiet enough for that sound to survive the journey to a sensor. Copper, steel, iron and lead carry it well. Plastic largely does not, and no amount of listening changes that.

Book an engineer See pricing

Quick answer

Acoustic detection reads the noise a pressurised leak makes and places it along a run by comparing arrival times at two sensors. It works well on copper, steel, iron and lead, poorly on plastic, and not at all on a run that is not under pressure. The survey fee is agreed before travel and covers detection and the written record; repair is quoted separately once a point exists.

What builds the fee for an acoustic survey

Stage of the survey	What sets the fee	What lengthens it
Confirming the run is an acoustic candidate	Pipe material and pressure state, which decide whether the method applies at all	Undocumented pipework that has to be established on site
Isolating other draw offs and recording the ambient floor	How many outlets, pumps and services have to be closed or noted	A building that cannot be quietened during trading hours
Ground microphone sweep to bracket the run	Length of run and how much of it is reachable from the surface	Thick surface build up, carpet, screed and external hard standing
Correlation between two fixed sensor points	How many accessible fittings exist to place sensors on	Long spans with no intermediate fitting to work from
Reproducing the result from a second geometry	Whether a backup method must be carried in case the signal is muted	Two runs that disagree and require a different method entirely

We do not publish a price list, because the figure is built from the property rather than picked from a menu. The detection fee is confirmed in writing before an engineer travels, and repair is quoted separately once there is a located point to quote against.

So a survey here starts with the circuit rather than the microphone. The suspect run is brought to normal working pressure and everything else on the system is valved off, because a second open outlet is a second noise source and a correlator has no way of telling them apart. Once the run is alone we take contact readings at whatever fittings can be reached, then set two sensors on known points and let the arithmetic place the fault somewhere along that span.

What property professionals get out of this is more useful than the mark on the floor. An acoustic result is repeatable: the same two sensor positions on the same pressurised run give the same distance again tomorrow. We write down where the sensors sat, the span we measured between them, the propagation figure we used for the material and the offset the instrument returned, so anyone asked to act on the finding can check the geometry rather than take it on trust.

The acoustic survey fee is agreed in writing before an engineer travels and covers the detection visit and the written record of it. What sets the figure is the length of run to be covered, how many sensor positions can actually be reached, and whether a backup method has to be carried in case the pipe material mutes the signal. Repair is quoted separately once a point exists to quote against, and is never folded into the detection fee.

What you get

Electronic Correlation: Our correlators analyse the time delay of a leak's sound between two sensors to calculate the precise distance to the fault, critical for long service runs under concrete slabs.
Ground Microphones: High-sensitivity contact microphones amplify the specific frequency range of a pressurised water leak, distinguishing it from structural or electrical noise in complex buildings.
Non-Destructive: The investigation requires no exploratory demolition. It is the optimal first step for leaks under original floorboards, parquet, or tiled commercial kitchens.
Targeted for Pressure Systems: Acoustic methods are definitive for leaks in mains water, central heating, and chilled water lines, where escaping fluid generates consistent high-frequency noise.
Data-Led Diagnosis: We provide a report detailing the likely leak location, depth, and recommended access point, forming a technical basis for any subsequent restorative work.
London Property Suitability: Highly effective in Victorian and Edwardian homes with suspended timber floors, and in post-war commercial builds with cast-iron or copper pipes in screed.

How it works

A method, not a guess

Confirm the run is a candidate before any listening starts

Pipe material, pressure state and reachable sensor points are established first. A depressurised run, a gravity feed or a plastic supply are not acoustic problems, and we say so before a fee is earned for sweeping a floor that was never going to produce a signal.

Take the noise out, then record what noise is left

Other draw-offs, pumps, cistern valves and neighbouring services are closed or noted. We record the ambient floor the building sits at, because a signal is only a signal relative to that. Where a building cannot be quietened during trading hours we schedule the listening for a window when it can.

Contact sweep to bracket, correlation to place

A ground microphone sweep narrows the run to a stretch. Two sensors then go on fixed accessible points, the measured span and the pipe material go into the instrument, and the time difference between the two arrivals returns an offset from one sensor. The inputs matter as much as the output, so all of them are written down.

Reproduce the answer from a different geometry

No point is marked on one correlation. We move a sensor, re-measure the span and run it again. Two independent geometries that agree inside a tight tolerance is a finding. Two that disagree is a signal we do not trust, and the honest response is a different method rather than an average of the two.

Before you book anyone

Six questions to put to an acoustic detection firm before booking

Ask what they will do before the microphone comes out

The answer decides most of the job. Acoustic equipment reads a pressurised run, so before anything is placed on a floor somebody has to establish that a pressurised run is what is failing. That is pressure decay testing, circuit by circuit, with the appliance valved off. A firm that arrives and starts sweeping is performing stage two without stage one, and the sweep has no target. If the answer to this question is a description of equipment rather than a description of a sequence, you have learned something useful for the price of a phone call.

Tell them what the pipework is made of and listen to the response

Acoustic detection is excellent on copper, steel, iron and lead and close to useless on plastic, which is standard in most London properties plumbed or replumbed since the 1990s. The leak noise is absorbed into the pipe wall within a short distance of the breach instead of travelling along the water column, so a sweep can pass directly over a substantial leak and register nothing at all. The right response to being told the run is plastic is that acoustic will be used where metal sections allow and something else will carry the rest. The wrong response is that it makes no difference.

Ask whether they carry a correlator or only a listening stick

These are different instruments doing different jobs. A listening stick or ground microphone amplifies what is underneath it, so it narrows a run to a stretch by comparing loudness, which is affected by surface build up as much as by distance to the leak. A correlator places two sensors on known points and calculates position from the difference in arrival time between them. Only the second produces a number somebody else can check. Both have their place, and a firm quoting for a pinpointing survey with only the first is quoting for something narrower than you think.

Ask what their definition of located actually is

The word found does a lot of unexamined work in this industry. Some published terms treat a leak as located once it has been narrowed to a five metre square, which in most London flats is a whole room, and a room is not a thing a repair can be cut to. Ask for the definition in writing before you book, ask it of everyone you are comparing, and note that a firm committing to a dimensioned point is describing a different service from one committing to an area. The difference between those two shows up later as the size of the opening.

Ask what happens to the quiet in the building during the survey

A leak signal is only a signal relative to the noise around it. Other open outlets, circulating pumps, cistern float valves, plant and even traffic through a suspended floor all add to what the sensor hears, and none of them announce themselves. A survey that does not isolate other draw offs and record the ambient level is producing impressions. In an office or a trading premises this also has a practical consequence: the listening may need to happen outside occupied hours, and that is a scheduling conversation worth having before the day rather than on it.

Ask what the report will contain besides the answer

For a domestic owner the mark on the floor may be enough. For a leaseholder, a freeholder, a managing agent or anyone who will be asked to pay for a consequence, the working matters more than the conclusion, because the person who has to act on the finding was almost never in the room when the reading was taken. Sensor positions described against fixed features, the measured span, the material assumed, the offset each correlation returned, what was isolated and held, and what could not be reached. If a firm cannot say in advance what will be in the document, assume the document is a paragraph.

Compare like for like

Acoustic detection is a narrow instrument used well or badly

Acoustic work is fast and precise inside its operating envelope and near useless outside it, which makes the interesting comparison not who owns a correlator but who is honest about when it applies. Here is how the same job differs between a listen and sweep visit, a firm whose income depends on the opening that follows, and the method used here.

What to check	A visit that starts by scanning	A firm that finds and opens on one price	London Leak Detection 247
Whether the method suits the pipe	Rarely established first. A plastic supply absorbs the frequencies the method depends on, and a sweep can pass over a substantial leak registering nothing at all.	Not a limiting factor, because an inconclusive listen still supports a recommendation to open and look.	Pipe material and pressure state are confirmed before any listening. Where the run is plastic or unpressurised we say acoustic is the wrong instrument before a fee is earned for sweeping it.
Controlling the noise floor	Other outlets stay open and plant keeps running, so the sweep is competing with the building.	Often skipped for time, since the goal is a plausible area rather than a defensible point.	Every other draw off is closed, plant is noted and where possible stopped, and the ambient level is recorded so that a candidate signal can be judged against it rather than against an impression.
How the location is calculated	By ear and by where the noise seems loudest, which is a function of surface build up as much as of distance.	Loosely, because the output only has to be good enough to justify the next stage.	Two sensors on fixed accessible points, the true span between them measured rather than paced, the material propagation figure entered deliberately, and the offset taken from the time difference between arrivals.
Whether the reading is reproduced	One pass, one conclusion. There is nothing to compare it against.	One pass is sufficient for the purpose it serves.	A second correlation is run from different sensor positions with the span re-measured. Two geometries that agree is a finding. Two that disagree is a signal we do not trust, and the honest answer is a different method rather than an average.
What is recorded about the inputs	Nothing. The result is an assertion with no way to check it.	The conclusion only, since the working is not what is being sold.	Sensor positions described against fixed features, the measured span, the material and propagation figure used, and the offset each run returned, so that anybody can check the geometry later without us.
Access needed in a block	Whatever door opens on the day, which is usually only the flat reporting the problem.	The same, because a single sided result is still enough to propose work.	We say in advance which parts of the building we need to reach, including the flat above and the riser cupboard, because a survey that can only reach one side of a party structure is limited from the outset and that is better known before attendance than after.

Recurring questions

The arguments that come up again and again

Public discussion of acoustic detection circles the same handful of disagreements, and almost all of them turn out to be arguments about method rather than about price. Below are the patterns that recur most often in UK property and DIY discussion, and what we think each one is actually about.

People whose acoustic survey found nothing and who were charged anyway

This is the most common grievance and it splits cleanly in two. Where the pipework was plastic, the method was never going to carry and the failure was in accepting the booking rather than in the survey. Where the pipework was metal, a blank sweep usually means the noise floor was never controlled or the run was not under pressure at the time. Our take: both are avoidable before attendance. Asking what the pipe is made of and what will be isolated is a two minute conversation that decides whether the visit can succeed at all.

Confusion between a listening stick and a correlator

Threads comparing quotes frequently reveal that two firms quoted for visibly different work under the same heading, and neither said so. The distinction that matters is whether position is inferred from loudness or calculated from arrival time at two sensors. Our take: ask which instrument is coming and what it will be placed on. A correlator needs two accessible points on the same run, so if the property has no reachable fittings between the suspected ends, that constraint exists whoever you book and is better discussed beforehand.

Arguments about whether acoustic work is affected by pipe depth

It comes up constantly and the answers given are usually too absolute in both directions. Depth and surface build up attenuate the signal and blur the apparent position, but they do not produce a single accuracy figure that applies to every job. Our take: any firm quoting one accuracy number for all properties is quoting a brochure. The useful thing to ask for is the tolerance claimed for your specific run and the reason for it, because that is the figure a contractor needs to price an opening.

Leaseholders trying to establish that the leak is not theirs

Water arriving from above produces a familiar stalemate: the flat below wants an investigation, the flat above denies a problem, and nobody instructs. The advice that recurs is to notify the managing agent in writing early, because the lease and the block policy often decide who pays before anyone argues. Our take: the fastest way out is evidence rather than argument. Isolating and pressure testing the demised pipework in each flat and showing that one side holds while loss continues is a result worth commissioning in its own right, and a negative finding should be written up as carefully as a positive one.

People wondering whether to hire equipment and do it themselves

Consumer grade listening devices are cheap enough that this comes up regularly, and the honest answer in the threads is usually that the equipment is the easy part. Our take: that is right, and the difficulty is not hearing something but knowing what to discount. Without a controlled noise floor, a known pipe material, a measured span and a second reading from a different geometry, a loud patch is just a loud patch. The instrument does not supply the sequence.

Questions

Asked before every booking

Our pipework is plastic. Is acoustic detection still worth booking?

On its own, usually not, and that is a thing worth being told before a visit rather than after one. Plastic absorbs the high frequency content the method depends on, so the signal falls away within a short distance of the breach and a sweep can pass directly over it without registering anything. On a plastic run the honest sequence is elimination by pressure decay first, then tracer gas or thermal depending on what the floor is made of, with acoustic kept in reserve for the metal sections at either end.

What does a correlator actually measure?

The difference in arrival time of the same noise at two sensors, and nothing else. Turning that into a distance needs two inputs from us: the true pipe span between the sensors and the speed sound travels in that material and diameter. Get either wrong and the instrument returns a confident number in the wrong place. That is why we record the span we measured, the material we assumed and the figure we used, and why a second run from different positions is not optional.

How do you know the noise is a leak and not a pump or the flat next door?

By removing the alternatives and then repeating. Everything else on the circuit is isolated, mechanical plant is noted and where possible stopped, and the ambient level of the building is recorded before we start hunting. A genuine pressurised leak is continuous and holds its frequency character. Plant noise starts and stops, and a neighbouring service does not move when we move a sensor along our run. If a candidate signal cannot be reproduced, it does not go in the report as a location.

Can this be done in an occupied office or a tenanted block?

Yes, with the access and the timing agreed in advance. What we need is reach: the riser cupboard, any valve chambers, and enough fittings on the suspect run to place sensors on. In a block that frequently means the flat above or below as well as the one reporting the problem, because a single sided survey of a party structure is limited from the outset and we would rather say that before attending than write it into the limitations afterwards.

What is in the written record afterwards?

Sensor positions described against fixed features of the building, the measured span, the material and propagation figure used, the offset returned by each correlation run, and the marked point dimensioned from two walls so it can be found again without us. Alongside that, what was isolated and held, what we could not reach, and the conditions in the building at the time. The parts that were ruled out carry as much weight as the part that was found.

How much does acoustic leak detection cost in London?

We do not publish a figure, because the fee is a function of the property rather than a menu item: the length of run to be covered, how many accessible fittings exist to place sensors on, and whether a second method has to be carried in case the pipe material mutes the signal. As a market reference, Checkatrade's published UK figures put leak detection at an average of £162, a typical range of £115 to £234 and a full spread of £80 to £1,600, with hourly diagnostic rates of £95 to £145. That spread reflects the difference between a single method visit to a small flat and a full elimination on a house with several wet systems. Our survey fee is confirmed in writing before an engineer travels.

Does acoustic leak detection work on plastic pipe?

Poorly, and it is worth being told that before a visit rather than after one. Plastic absorbs the higher frequency content the method relies on, so the signal falls away within a short distance of the breach. On a plastic run the workable sequence is elimination by pressure decay first, then tracer gas or thermal depending on what sits above the pipe, with acoustic reserved for any metal sections at either end. A firm that accepts an acoustic booking on a known plastic supply without mentioning this is not doing you a favour.

How long does an acoustic survey take?

Most domestic surveys run to a few hours, and the variable is rarely the listening itself. Isolating other outlets, establishing the ambient noise level, finding two accessible sensor points on the same run and then reproducing the reading from a second geometry all take longer than the sweep does. Buildings that cannot be quietened during occupied hours take longer still, because the useful listening window may not be during the working day.

Can you survey without turning the water off?

The run being surveyed has to stay under pressure, since a depressurised pipe makes no noise. Everything else, though, gets closed: other outlets, branches that can be valved off, and where possible circulating pumps and plant. So the answer is that we turn things off selectively rather than generally, and we record what was isolated, because what held is part of the evidence.

What if two correlation runs disagree?

Then neither is reported as a location. Disagreement between two geometries usually means one input was wrong, most often the assumed span or the material, or that the signal being tracked is not the leak. The correct response is to re-measure and re-run, and if the disagreement persists, to change method rather than to average the two results into a confident looking number that is wrong in a new place.

Will you mark the leak on the floor?

Yes, physically, and dimensioned from two fixed features of the building so it can be found again once furniture has moved and we are no longer there. The report carries the same dimensions along with photographs keyed to the marks. That matters because the person who eventually opens the floor is frequently a different trade attending on a different day.

Water going somewhere it shouldn’t?

Tell us the symptoms and your postcode. We’ll confirm the visit, the fixed detection fee and the arrival window before you commit to anything.

Book a detection visit