Your roof is leaking. The question is not where the water is appearing — that is obvious. The question is where it is actually getting in. Those two locations are rarely the same. Water travels through roofing build-ups, sometimes for several metres, before you see it. The wrong starting point costs you time, money, and often a needlessly large repair.
This article is a comprehensive reference to all nine methods used in professional flat roof leak detection in the UK — from low-voltage electronic integrity testing and high-voltage dry testing, through moisture mapping, vacuum box testing, tracer gas, infrared thermography, and buried electrical field mapping. For each method we explain the principle, when it is appropriate, what the standards say, and where it falls short. It is the kind of briefing we give clients before we go on site — written by specialists who have co-authored CIRIA C817 and BS 8102:2022.
Flat roof leak detection methods: at a glance
| Method | What it detects | Best used when | Works on a buried roof? (green / blue / ballasted) | Conductive substrate required? | Locates the breach? | Key standard / guidance |
|---|---|---|---|---|---|---|
| Low-voltage wet testing (Buckleys Wet Roof Pro) | Membrane breaches via electrical field disruption | Accessible membrane, wet or wettable surface, conductive substrate below | ✗ Membrane must be accessible | Yes | Yes — to within a small area | LRWA GN18:2025 |
| High-voltage dry testing (Buckleys Dry Roof Pro) | Membrane breaches via spark discharge | Dry non-conductive membrane on conductive substrate; not for adhered or bituminous systems | ✗ Membrane must be dry and exposed | Yes | Yes — precise spark location | LRWA GN18:2025 |
| PIC testing | Weld continuity on single ply membranes | Single ply systems only; QA during or after installation | ✗ Pre-covering QA only | Yes | Yes — weld location | Manufacturer QA specifications; SPRA |
| Moisture mapping (Tramex capacitance) | Moisture distribution within build-up | Non-conductive substrates; screening large areas | Partial — scans through some overburden but limited depth; not through ballast or growing media | No | No — maps wet zones, not breach point | ASTM D7954; Tramex |
| MTN Roof Scan | Moisture distribution via steel ball capacitance | Alternative to Tramex on certain roof types | ✗ Surface access required | No | No — moisture mapping only | Manufacturer guidance |
| Vacuum box testing | Weld and seam integrity at details | Single ply laps, penetrations, upstands; QA not field-sheet investigation | ✗ Pre-covering QA only | No | Yes — visible bubble at defect | CIRIA C817; SPRA guidance |
| Infrared thermography | Surface temperature differential indicating wet insulation below | Large warm roof areas; correct weather conditions only | ✗ Does not work on green roofs, ballasted, or inverted roofs | No | No — indicates wet insulation zones only | ASTM C1153 |
| Flood testing | Active water passage through membrane | Post-repair verification only; requires structural engineer sign-off | ✗ Overburden must be removed first; structural loading risk | No | No — confirms leak exists, not location | BS 6229:2025; CIRIA C817 |
| Tracer gas | Membrane breaches via hydrogen escape | Non-conductive systems where EIT is ruled out; inaccessible build-ups | Partial — works where gas can be introduced to the build-up void; wind affects accuracy | No | Yes — sniffer locates escape point | RAWTA member guidance |
| Buried leak location (electrical field mapping) | Membrane breach under overburden | Green roofs, blue roofs, hard landscaping, ballast — specifically designed for this | The correct method for buried waterproofing — no lifting required | Yes (below membrane) | Yes — to a highly localised area | CIRIA C817; GRO Code |
Why visual inspection usually is not enough
Start with what most people try first. A competent roofer looks at the obvious suspects: terminations, penetrations, upstands, flashings. That is always the right first step. Many leaks are directional — driven by wind and rain rather than field sheet failure — and sometimes a careful visual is all you need.
But when the visual does not resolve it, you need equipment. And the right equipment depends on what kind of roof you are dealing with.
Method 1: Low-voltage wet testing (the Buckleys Wet Roof Pro)
The most widely used method for accessible flat roofs in the UK. The Buckleys Wet Roof Pro applies a low-voltage electrical field across a wet roof surface. A guard ring is set around the perimeter of the test area, connected to a generator unit. That generator creates an electrical field within the water on the roof surface.
Any breach in the waterproofing membrane creates a pathway to earth — and that disrupts the electrical field. The operator walks the roof methodically with detector poles, reading where the disruption is occurring. Done correctly, you get a reliable location to within a small area. The requirement: the roof needs to be wet, with sufficient water to form a continuous conductive film. In the UK that is rarely a problem.
More information on the Buckleys Wet Roof Pro: buckleysinternational.com. The definitive UK guidance on EIT methodology for liquid-applied systems is LRWA Guidance Note 18 (2025), produced with RAWTA — free to download.
Method 2: High-voltage dry testing (the Buckleys Dry Roof Pro)
When the roof is dry — or needs to be dry for the test to work — the Dry Roof Pro is the tool. This uses high voltage, passed through either a copper brush or a roller, which is swept across the membrane surface.
The principle: the membrane must be non-conductive, sitting on a conductive substrate. If the membrane is intact, the high voltage finds nowhere to go. If there is a breach, the voltage ionises the air through the gap and you get a spark — a visible, audible signal of exactly where the membrane has failed.
The Dry Roof Pro can test up to 1,000m² per hour in good conditions. One important constraint: this does not suit adhered systems or bituminous systems where you are not looking for a homogeneous weld. On those systems, testing pressure can damage adhesion bonds. Membrane type determines method.
Dry Roof Pro specification: buckleysinternational.com. For guidance on when high-voltage dry testing is appropriate versus wet testing, see LRWA GN18 and the LRWA British Standards guidance.
Method 3: PIC testing (pulse impedance continuity)
PIC testing is used primarily for workmanship quality assurance on single ply membranes. It checks weld continuity rather than gross membrane failure. On a single ply system, PIC testing makes sense. On an adhered system, it is a different matter: an adhered system does not have a homogeneous weld to check. Applying PIC without that understanding risks damaging the very thing you are trying to inspect. A competent investigator chooses the test to the system.
Method 4: Moisture mapping (Tramex capacitance scanning)
When the roof is not conductive — ruling out the wet electrode methods — moisture mapping becomes the viable route for identifying where ingress has occurred. The Tramex Dec Scanner and Tramex Roof and Wall Scanner (RWS) use capacitance to find variations in moisture level through the roof build-up without surface penetration.
The Dec Scanner moves across the surface in a grid pattern generating a continuous impedance reading. The rubber electrode mat transmits a low-frequency signal; moisture conducts it back more strongly than dry material. You build up a moisture map of the whole roof — compliant with ASTM D7954 — to a depth of up to 150mm. It will not tell you where the membrane has failed, but it tells you where moisture is sitting in the system — and that tells you where to look.
Tramex moisture mapping equipment: tramexmeters.com
Method 5: MTN Roof Scan (steel ball capacitance)
The MTN Roof Scan uses steel balls rolled across the membrane surface to map moisture variations using the same underlying capacitance principle. It is an alternative to the Tramex approach on certain roof types.
Buried leak location: a different category entirely
When the waterproofing is buried — under a green roof, blue roof, hard landscaping, or ballast — and you need to find the precise point of failure without lifting the overburden, a different technology is required. Vector partners with Texplor Group for buried leak location surveys using electrical field mapping. This locates the membrane breach to a highly localised area without any destructive investigation. Without it, buried membrane investigations typically involve lifting extensive areas of landscaping at considerable cost.
Learn more: Buried Leak Location. The requirement for permanent leak detection on buried waterproofing is set out in CIRIA C817 (blue-green roofs) and recognised as best practice in the GRO Green Roof Code. NHBC Standards Chapter 7.1 requires electronic integrity testing before green roof layers are placed on NHBC-registered developments.
Method 6: Vacuum box testing
Vacuum box testing is a targeted QA method used to verify the integrity of individual laps, seams, and details on single ply membranes — particularly at penetrations, upstands, and other junctions where a continuous weld cannot be guaranteed by visual inspection alone.
The principle is simple. A rigid transparent box with a foam-sealed base is placed over the suspect seam. Soapy water is applied to the surface and a vacuum is drawn inside the box — typically to around 20–35 kPa depending on the membrane specification. If there is a void or breach in the weld, the pressure differential draws air through it and the soapy water bubbles visibly. The transparent lid means the inspector can see exactly where the failure is without moving the box.
It is not a field-sheet method — it cannot practically cover large areas — but for detail verification it is fast, unambiguous, and requires no conductive substrate. On a single ply system, vacuum box testing at all laps and critical junctions immediately after installation is the most reliable way to confirm workmanship quality before the membrane is covered or loaded.
Vacuum testing is referenced in CIRIA C817 alongside EIT as an accepted QA method. The Single Ply Roofing Association (SPRA) publishes guidance on its application on single ply systems, and most single ply system manufacturers specify vacuum testing as part of their approved installation and QA procedures.
Method 7: Infrared thermography
Infrared thermography is sometimes proposed as a roof survey method. It is worth understanding what it does — and what it does not do — before specifying it.
An infrared camera measures surface temperature, nothing more. The principle applied to roofing is that wet insulation absorbs heat during the day and releases it more slowly after sunset, creating a thermal differential that shows up on a post-dusk scan. In theory, warmer patches in the cooling roof indicate wet insulation below.
The limitations are significant. The camera is reading surface temperature — it cannot see through the membrane into the build-up. Results depend entirely on the right conditions: a clear sunny day providing sufficient solar loading, followed by a cool clear evening with no wind, no recent rain, and no dew point issues. On inverted roofs, ballasted roofs, green roofs, white or reflective membranes, and roofs with certain closed-cell insulations, infrared is largely useless regardless of conditions.
Critically, even when it works, infrared identifies where moisture has accumulated in the insulation — not where the membrane has failed. The two locations are often different. It is a moisture mapping tool, not a breach location tool, and on most UK roof types it is less reliable and more weather-dependent than capacitance scanning.
There are circumstances where it adds value — screening a large accessible warm roof build-up on the right evening, for example — but it should never be relied on as a primary investigation method, and its limitations should be understood before commissioning it.
ASTM C1153 sets out procedures for infrared thermographic inspection of roofs. For UK membrane-type guidance on when IR is and is not appropriate, see LRWA British Standards guidance and RAWTA.
Method 8: Flood testing
Flood testing is one of the oldest roof leak detection methods and is still referenced in guidance including BS 6229:2025 and CIRIA C817. The principle is simple: dam a section of the roof, fill it with water to a set depth, and observe over a test period whether water appears internally.
In practice, it has serious drawbacks that limit when it should be used.
The first is structural. A flat roof is not designed to hold standing water across its entire surface simultaneously. A 150mm head of water adds approximately 150kg per square metre of load. On a roof with any deflection characteristics — lightweight metal deck, timber, or a structure that has any movement — that additional loading can cause slight deflection, which can itself open up laps and details that were previously satisfactory. Flood testing can create the defects it is trying to find.
The second is time. Water takes a long time to find its way through a roofing system and manifest internally, particularly through a build-up with multiple layers. A test period of 24 or even 48 hours may not be sufficient to detect a slow leak. By the time you have a result, you may have saturated the insulation without identifying where the breach is.
The third is that flood testing, even when it works, only confirms that a leak exists — it does not locate the breach. You still need a separate investigation to find the source.
Where flood testing has a legitimate role is in post-repair verification on accessible, structurally appropriate roofs — confirming that a specific repair to a specific area is now watertight. In that confined context, with a structural engineer’s sign-off on loading, it can be useful. As a primary diagnostic tool on a complex roof, it is rarely the right choice.
Method 9: Tracer gas
Tracer gas is a genuinely useful method for situations where other approaches cannot reach. It is most commonly used when the membrane is inaccessible, non-conductive, and the conditions rule out capacitance mapping.
The method works by introducing a safe, inert gas mixture — typically 5% hydrogen in nitrogen — into the build-up beneath the membrane via a small access point. Hydrogen is the lightest molecule in the atmosphere. Where there is a breach in the membrane, gas escapes and rises to the surface. A sensitive detector probe is then swept across the roof surface, reading hydrogen concentration in parts per million. The highest readings indicate where the breach is.
The advantages: it works on non-conductive systems, it is non-destructive, and it can locate breaches that are too small to be found visually. The gas mixture is non-toxic, non-flammable, and safe to use in occupied buildings.
The limitations: it requires access to the build-up to introduce the gas, which is not always possible without some minor intervention. Wind can disperse the escaping gas and affect readings, so calm conditions are preferred. On very large or complex roofs with multiple build-up layers, gas diffusion can make precise location harder.
In the right situation — a non-conductive membrane, inaccessible build-up, failed EIT attempts — tracer gas can find breaches that nothing else will. It belongs in the toolkit of a specialist investigator, not as a first resort but as a capable option when standard methods are ruled out.
Choosing the right method
There is no single correct approach. Method selection depends on membrane type, build-up, surface conditions, and what the investigation needs to answer. Applying the wrong method wastes time. Applying pressure-based tests to adhered systems can damage the roof you are trying to protect. If you have had an investigation that did not find the source, or you are trying to understand which approach is right for your project, we are happy to talk it through before anyone gets on site.
Frequently asked questions
Do I need a conductive substrate for electronic integrity testing?
Yes — both low-voltage wet testing and high-voltage dry testing require a conductive substrate directly below the waterproofing membrane (typically a concrete or metal deck). On timber decks or where insulation sits between the membrane and the deck, a conductive medium such as a wire mesh or conductive glass felt can be installed during construction to make EIT possible. This is why specifying EIT at design stage matters — it is difficult to retrofit. See LRWA GN18:2025 for full substrate requirements.
Can you electronically test a green roof or buried waterproofing?
Not with standard EIT once the overburden is in place — you cannot pass the electrode across a surface covered with growing media, ballast, or paving. For buried waterproofing, the correct method is electrical field mapping (EFM) — a buried leak location technique that locates membrane breaches to a highly localised area without lifting the overburden. CIRIA C817 and the GRO Green Roof Code both recognise EIT as a pre-covering QA method and EFM as the post-burial investigation method.
What is the difference between moisture mapping and leak detection?
Moisture mapping — using Tramex capacitance scanning or infrared thermography — tells you where moisture has accumulated within the roof build-up. It does not tell you where the membrane has failed. Leak detection methods such as low-voltage EIT or electrical field mapping locate the actual breach in the membrane. The two are complementary: moisture mapping identifies which areas of the roof need closer investigation; breach-location methods then find the source. Treating a moisture map as a leak location survey is a common and costly mistake.
Is flood testing safe for flat roofs?
Flood testing carries real risks that are often underestimated. A 150mm head of water adds approximately 150kg per square metre of loading — structural sign-off is essential before proceeding, particularly on lightweight metal deck or timber structures. Additionally, water can take many hours or days to manifest internally through a multi-layer build-up, making short test periods unreliable. Flood testing can also cause deflection in the structure which opens up laps and details that were previously satisfactory — potentially creating the defects it is trying to find. It has a legitimate role in post-repair verification of specific, structurally appropriate areas, but should not be used as a primary diagnostic method on complex roofs.
What qualifications should a roof leak detection specialist hold?
For electronic integrity testing and roof waterproofing inspection, look for: NVQ Level 3 in waterproofing testing and inspection; membership of RAWTA (the Roofing and Waterproofing Test Association) — the primary UK body for independent on-site test houses; and a demonstrable CV in flat roofing and waterproofing systems. The LRWA and SPRA also maintain guidance on competence requirements for their respective membrane types. Under the Building Safety Act 2022, appointing an incompetent inspector carries legal liability for duty holders — competence is a legal requirement, not a preference.
Which leak detection method works on a single ply membrane?
Single ply membranes (TPO, PVC, EPDM) suit several methods depending on the question being asked. For QA of weld continuity at laps and details, vacuum box testing is the most reliable approach. For field-sheet investigation of a roof with a conductive substrate, low-voltage wet EIT is the standard method. High-voltage dry testing also works on non-conductive single ply membranes on conductive substrates. Note that black EPDM membranes containing carbon black are generally not suitable for EIT due to their conductivity — white or grey EPDM can be tested. PIC testing applies to weld verification on single ply systems and should not be used on adhered membranes.
How does permanent sensor monitoring differ from leak detection surveys?
Leak detection surveys are point-in-time investigations — they answer the question “where is the breach right now?” Permanent sensor monitoring is continuous infrastructure embedded during construction that answers the question “is the building envelope still performing?” on an ongoing basis. Sensors embedded in the roof build-up monitor moisture and relative humidity in real time, triggering automated alerts the moment conditions change. This is specified in CIRIA C817 for blue-green roofs and referenced in BS 6229:2025 as a method for detecting leaks between inspections. The two are complementary: sensors provide early warning; a survey then locates the precise breach.
Does electronic leak detection work in the rain?
Low-voltage wet testing actually requires a wet surface — rain on the roof is not a problem and in the UK is rarely one. High-voltage dry testing requires a dry membrane surface and cannot be performed in wet conditions. Moisture mapping with Tramex capacitance equipment can be used in most weather conditions. Infrared thermography requires specific dry, sunny conditions and is ruled out by recent rain. Tracer gas is best performed in calm dry conditions as wind can disperse escaping gas and affect readings.