How Non-Invasive Leak Detection Works on an Insurance Claim
Non-invasive is a claims term as much as a technical one. Which methods genuinely read through the fabric, which ones need a hole first, why the distinction decides how much of a claim is spent on searching rather than repairing, and how the method chosen shapes the evidence.
Non-invasive is used loosely in this trade, and on an insurance claim the looseness costs money. An insurer paying for a search under a trace and access clause is paying twice over when a contractor opens a floor to look for something a meter could have found through it: once for the search, and again for the making good. A method that reads through the fabric keeps that spend down, and it preserves the physical evidence the claim will later be tested against.
This guide sets out which methods are genuinely non-invasive, which are not, and how the choice affects the claim itself. For the sequence a survey follows from instruction to report, see our guide to what a leak detection survey involves.
Why Non-Invasive Matters to an Insurer
Trace and access is a monetary cap. It pays the reasonable cost of finding a hidden leak and making good the damage caused by that search, and it is stated on the policy schedule. Every square metre of floor lifted speculatively comes out of the same limit as the survey that would have avoided lifting it, and once the limit is exhausted the policyholder pays the balance.
There is a second reason, and it matters more on a disputed claim. The Financial Ombudsman Service guidance on gradual damage sets out when it will normally require an insurer to pay despite a gradual damage exclusion: the damage was caused by an insured event, the policyholder could not reasonably have been aware of it because it was concealed, and they acted as soon as they reasonably could have. The concealment limb is a physical fact about the building, recorded in moisture distribution and pipe route. Somebody who has already cut the floor open has destroyed the record of it.
What the insurer is buying, and in what order
A named failed component comes first, because that fixes whether the peril is insured at all. What the physical evidence says about the age of the failure comes second. Minimum access comes third, and only once the first two are settled. The written record comes fourth. ICOBS 8.1.1R in the Financial Conduct Authority Handbook requires claims to be handled promptly and fairly and not unreasonably rejected, and a handler cannot do either of those things from a report that says there is water under a floor.
What Non-Invasive Actually Means
The working definition is simple: nothing is cut, drilled, lifted or excavated. In practice a non-invasive survey still involves a fair amount of handling, all of it reversible:
- Furniture and floor coverings moved, and appliances pulled out from under worktops.
- Bath panels and boxing removed where they were designed to come off.
- Existing inspection hatches, manifold covers and meter chambers opened.
- Stop taps, isolating valves and service valves operated.
- Instruments placed on finished surfaces and read through them.
Everything else stays where it is until a source has been located or bounded.
The line between survey and access
That line is where most billing disputes sit. The non-invasive survey is the diagnostic stage. Access is the controlled opening-up that follows once a source has been located or bounded, and it is the line on an invoice that maps directly onto the trace and access clause. A contractor who opens up in order to look, rather than in order to confirm, has collapsed the two stages into one and produced an invoice no adjuster can apportion.
The middle category nobody names
Several standard techniques are neither. Pressure testing requires the system to be isolated and drained, which means valves closed and often a joint broken. A borescope inspection needs an aperture, even if it is a small one. Both are minimally invasive rather than non-invasive, and a report that describes them as non-invasive is storing up an argument for later.
The Genuinely Non-Invasive Methods
Four techniques read through the building fabric. Each one answers a different question, and none of them answers all of them.
| Method | What it detects | Where it works best | Where it fails |
|---|---|---|---|
| Acoustic listening and ground microphones | Sound of water escaping a pressurised pipe | Mains and supply runs, buried services, solid floors | Low-pressure and gravity systems, plastic pipe, noisy environments |
| Leak noise correlation | Time difference of the same leak sound at two sensors | Long buried runs between two contact points | Where only one contact point is accessible, or the pipe material changes mid-run |
| Thermal imaging | Surface temperature differential | Heating leaks, hot supply, evaporative cooling, mapping pipe routes under screed | Dried-out leaks, cold water on a cold floor, anything behind insulation |
| Non-invasive moisture meters | Relative moisture behind a surface, to a shallow depth | Mapping the extent of wetting and its gradient across a floor or ceiling | Absolute values, deep build-ups, materials with conductive content |
Acoustic detection and correlation
Water escaping a pressurised pipe makes a noise, and the noise travels along the pipe and through the ground. Ground microphones pick it up at the surface, and a correlator placed at two contact points calculates the position of the source from the time difference between them. It is the first choice on incoming supply and mains, and it is the method most affected by the material the pipe is made of: plastic damps the signal far more than metal, which is why the material is stated in the report rather than assumed.
Thermal imaging and moisture mapping
A thermal camera reads surface temperature, not moisture, and it cannot see through anything. What it sees is the effect of water on the surface above it: warmth from a heating leak, or the cooling that evaporation produces. Paired with a non-invasive moisture meter it produces a map of where water has reached and which way it moved, which is the evidence the concealment argument later rests on. Both are indicative instruments, and their readings are reported separately from quantitative pin meter readings because the two are not comparable. See our thermal imaging page for what the equipment can and cannot resolve.
Tracer gas, and why it sits on the boundary
An inert hydrogen and nitrogen mix introduced into an isolated section escapes through a breach far smaller than water will pass, rises, and is picked up at the surface with a sniffer probe. It is the definitive method under screed, slab and underfloor heating, where acoustic and thermal both struggle. It is non-invasive at the surface, but it requires the section to be drained and a connection made, so on a report it belongs in the minimally invasive column with the connection point recorded. Our tracer gas detection page describes the procedure.
What Counts as Invasive, and When It Is Justified
Invasive work is not a failure of method. It is the stage where a located source gets confirmed, and every claim that ends in a repair passes through it. What matters to the claim is that it happened second, in a defined area, with a photographic record either side.
| Technique | Classification | What it requires | Claim consequence |
|---|---|---|---|
| Moisture mapping, thermal, acoustic | Non-invasive | Access to the surface only | Survey line on the invoice, no making good |
| Pressure and flow testing | Minimally invasive | System isolated and drained, connection made | Survey line, with any disturbance stated |
| Tracer gas | Minimally invasive | Drained section and an injection point | Survey line, connection point recorded |
| Borescope inspection | Minimally invasive | A small drilled aperture | Access line, aperture location and size recorded |
| Lifting floors, cutting ceilings, removing boxing | Invasive | Prior agreement on the area to be opened | Access line, with reinstatement, inside the trace and access limit |
| Excavation of external pipework | Invasive | Located source, and a position on making good agreed first | Access line, and see the ombudsman guidance below |
Excavation is a special case
The Financial Ombudsman Service publishes technical guidance on damage to underground pipes making two points that change how external work is scoped. Damage is not confined to physical damage and extends to loss of function, and an insurer may need to make good excavation damage caused during an investigation even where the claim is ultimately declined. Both belong in the conversation before the ground is opened. Whether the pipe is the policyholder's responsibility at all is a separate question, and Ofwat sets out the boundary between the supply pipe and the communication pipe.
How the Method Chosen Affects the Claim
Method selection is not a technical preference exercised in private. It changes what the file contains and what the policyholder pays.
| Approach | Effect on trace and access spend | Effect on the evidence |
|---|---|---|
| Non-invasive survey first, then defined access | Search cost is a survey fee plus one agreed aperture | Moisture gradient intact, concealment demonstrable, source photographed in situ |
| Exploratory opening-up with no survey | Open-ended, and drawn from the same limit as the making good | Gradient destroyed, cause inferred rather than shown, dates unsupported |
| Single method used in isolation | Often a second attendance, charged again | Findings that cannot be corroborated if challenged |
| Methods combined and each result recorded | One attendance, one apportionable invoice | Systems eliminated on the record, which survives a desktop review |
A negative result is still a result
On a private job a homeowner wants a leak found. On an insured loss, a report stating that the heating circuit held pressure for a recorded duration, the supply was static at the meter and the waste stack was sound has moved the claim, because it redirects the file towards rainwater ingress, condensation or a fabric defect. That is why consumer no find, no fee arrangements do not translate to insurer work: the investigation is itself the deliverable.
The Order the Methods Are Used In
Sequence is what makes a survey proportionate rather than exhaustive, and proportionality is what an adjuster looks for when deciding whether the search was reasonable.
- Establish whether the incoming supply is losing water at all, with every outlet closed and the meter watched.
- Isolate and pressure test system by system, recording the pressure applied, the hold duration and the observed drop rather than the word failed.
- Map moisture across the affected area to establish extent and direction before forming any view on the source.
- Apply thermal imaging to the narrowed area, ruling out the known false positives on site.
- Confirm with acoustic equipment on pressurised runs, or tracer gas where the build-up defeats it.
- Only then agree the aperture, open up, and photograph the exposed component before anything is repaired.
What Non-Invasive Detection Cannot Do
Every method has a failure mode, and a report that does not state them is less useful than one that does. Overstating what a camera can see is the quickest way to have a finding set aside on review.
The false positives that get ruled out on site
Thermal imaging produces apparent anomalies from underfloor heating loops running normally, solar gain through a window, thermal bridging at a junction, recently cleaned or mopped floors, and voids behind plasterboard. Each is excluded by a control observation rather than by judgement, and the exclusion is recorded. Non-invasive moisture meters respond to conductive material as well as water, so a foil-backed board or a metal mesh reads high when it is perfectly dry.
What no method will do
Nothing reads through a substantial insulated floor build-up with certainty. Nothing recovers a leak that has already dried. Nothing distinguishes condensation from a plumbing leak on a single reading, which is why the two are separated by pattern and by sequence rather than by instrument. Where the evidence does not support a conclusion, the honest report says so, and it is worth more to the file than one that guesses.
What This Means for Claims Support
Detection supports a claim in three places. It names the failed component, which decides whether the peril responds. It records the physical characteristics that speak to when the failure began, which decides the gradual damage argument. And it keeps the search spend inside a clause that has a ceiling on it, which decides whether the policyholder ends up funding the difference. The method-by-method comparison is set out further in our guide to acoustic, thermal and tracer gas detection.
Anything introduced into the wholesome water supply during testing falls under the Water Supply (Water Fittings) Regulations 1999, which prohibit materials likely to cause contamination, and approval under the Water Regulations Approval Scheme is how that is evidenced. The Association of British Insurers publishes quarterly property claims statistics for the current market position. For scope, instruction and reporting on an insured loss, see our insurance leak claims page and the trace and access page.
How we help with this
If the article describes a problem you actually have, these are the visits that deal with it.
- our full guide to leak detection
How each method works and when it is the wrong tool for the job.
- insurance leak claims
What escape of water cover pays for, and what it does not.
- leak detection in London
Finding a hidden leak without opening the property up first.
Frequently asked questions
Is leak detection really non-invasive?
The survey stage is. Acoustic listening, correlation, thermal imaging and non-invasive moisture mapping all read through the building fabric, so nothing is cut, drilled or lifted. Pressure testing and tracer gas need a system isolated and a connection made, which makes them minimally invasive rather than non-invasive. Confirming the failed component almost always needs a defined aperture, but that happens second, in an agreed area, and it is invoiced as access rather than as survey.
Why does my insurer care whether the method was non-invasive?
Because trace and access is a monetary cap covering both the search and the making good of search damage, and speculative opening-up draws on the same limit that the survey would have protected. A non-invasive survey also preserves the moisture distribution and the evidence that the pipework was concealed, which is what the ombudsman's gradual damage test turns on. Once a floor has been opened by guesswork, neither the cost nor the evidence can be recovered.
Which method is best for finding a hidden leak?
None of them individually. Acoustic equipment works on pressurised supply pipework and degrades on gravity and plastic systems. Thermal imaging locates heating leaks and evaporative cooling but cannot see through a floor and fails on a leak that has dried. Tracer gas reaches breaches under screed and slab that nothing else finds, but needs the section drained. The method is chosen for the system under suspicion and the construction, and the results are read together.
Can thermal imaging see through walls and floors?
No. A thermal camera reads surface temperature, so what it shows is the effect of water on the surface above or in front of it, not the water itself. That makes it good at heating leaks, hot supply and evaporative cooling, and poor at cold water on a cold floor or anything behind insulation. It also produces false positives from underfloor heating loops, solar gain, thermal bridging and recently mopped floors, each of which is excluded on site.
What happens if the survey rules everything out?
That is a usable finding on an insured loss rather than a failed job. A report stating that the heating held pressure for a recorded duration, the supply was static at the meter and the waste stack was sound redirects the file towards rainwater ingress, condensation or a defect in the building fabric. Its value depends on the report recording exactly what was tested, at what pressure and for how long, because a bare statement that nothing was found proves nothing.
Does non-invasive detection cost less than opening the floor up?
Usually, and the gap widens the more concealed the pipework is. A survey is a defined attendance with a written output. Exploratory opening-up is open-ended, needs reinstating whether or not it found anything, and both halves are drawn from the trace and access limit. The bigger risk is that search damage carried out before the insurer authorised it may fall outside the clause entirely, because the clause pays for a search that was reasonable and proportionate.