Hydrogen 5/95 · Sewerin Variotec · Buried & sub-slab specialist

Tracer Gas Leak Detection London

The specialist detection method for buried MDPE mains, screed-covered UFH and sub-slab supply pipes acoustic cannot pinpoint. Inert hydrogen-nitrogen 5/95, semiconductor surface sensors, ±150mm typical accuracy on hard surfaces.

Book a Tracer Gas Survey

Get a Survey Quote

Your browser may ask to share your location — this helps us find the nearest available plumber faster. You can decline and still send the form.

Tracer gas — the specialist method for leaks acoustic and thermal miss

Tracer gas leak detection is not the first-pass method for a residential leak survey. Acoustic listening is faster, requires no drain-down, and pinpoints most pressurised leaks on accessible pipework with ±100mm accuracy. Thermal imaging locates hot-circuit leaks visible at the wall or floor surface. Tracer gas is the method we reach for when both of the faster techniques have been tried, the pressure test still confirms a real leak, and the leak is either too small, too plastic, or too deeply buried for the other methods to resolve.

The physics is why it works where the others fail. Hydrogen is the smallest molecule in nature — an atomic diameter of about 0.7 angstroms, compared to helium at 0.98 angstroms and CO₂ at 3.3 angstroms. That size means hydrogen migrates through leak points that liquid water can flow through but larger tracer molecules cannot — a hairline crack in a copper pipe, a partially-corroded compression olive, a pinhole in a PEX-A loop. Injected at working pressure into a drained pipe, the hydrogen escapes through the leak, migrates through the surrounding structure (soil, screed, plasterboard) to the surface, and is detected by a semiconductor sensor with ppm-level selectivity for hydrogen. The peak surface reading is the pinpoint.

The 5/95 forming gas mixture is chosen for safety. Pure hydrogen has a lower explosive limit of 4.1% in air. Our tracer mixture is 5% hydrogen in nitrogen — meaning at any dilution with room air, the hydrogen concentration falls immediately below the LEL and cannot support combustion. This is why forming gas is the industry-standard tracer for leak detection: enough hydrogen to migrate reliably and be detected at ppm levels, safely below combustion thresholds by design. The nitrogen carrier is inert. Every survey is conducted with the room ventilated as belt-and-braces good practice, but there is no fire risk from the tracer gas itself.

Every tracer gas survey we run is delivered by a Water Regulations 1999 competent engineer (UK Certification Ltd certificate 136356 issued 8 September 2025, expiry 18 August 2030) with G3 unvented hot water certification (certificate 136359, same date range). Water Regulations competency is required to work on any pressurised drinking-water supply, including the temporary disconnection during a tracer gas survey. Public liability £5,000,000 via SiriusPoint through Eaton Gate MGU, policy BE26ACTT000000018221, period 07/05/2026 to 06/05/2027.

The five-stage tracer gas survey — drain-down to pinpoint

Every tracer gas survey follows the same disciplined sequence. Skipping the drain-down produces false-negative results because residual water absorbs the tracer gas.

01

System isolation and full drain-down

Before any gas is introduced the affected circuit is fully drained. On a buried supply pipe survey the internal stop-tap is closed, all downstream draw-off points opened until the residual pressure falls to zero, and the incoming service run at the boundary stop is disconnected at the internal stop connection. On a heating or UFH circuit the loop is drained via the lowest filling loop. A full drain-down is critical — residual water in the pipe absorbs the tracer gas and produces false-negative or delayed surface readings.

02

Tracer gas injection at working pressure

The circuit is charged with hydrogen-nitrogen 5/95 forming gas — a safety-graded mixture of 5% hydrogen (below the 4.1% lower explosive limit at any dilution with room air) and 95% nitrogen (inert carrier). The gas is injected at typical mains pressure (2–4 bar) via a calibrated regulator. Hydrogen is chosen because it is the smallest molecule in nature — its atomic diameter (0.7 angstroms) means it migrates through leak points that liquid water can flow through but larger tracer molecules (helium, sulfur hexafluoride) cannot.

03

Migration wait — hydrogen through structure to surface

After injection the survey pauses for 10–20 minutes to allow the hydrogen to migrate from the leak point through the surrounding structure (soil, screed, plasterboard) to the surface. Hydrogen migrates at approximately 0.5–1 metre per hour through compacted soil, 3–4 metres per hour through screed, and reaches the surface of a solid stone floor within 15 minutes of injection at a typical leak depth. Migration time is a function of leak depth and structure permeability — the engineer times this based on the specific survey environment.

04

Semiconductor surface sensor sweep

Using a Sewerin Variotec 460 or Inficon Sensistor semiconductor gas sensor with a hydrogen-selective detector head, the engineer sweeps the surface at 10–20cm intervals. The detector responds specifically to hydrogen (ppm-level sensitivity) and ignores common atmospheric gases (methane, CO₂, propane). A rising ppm reading identifies the leak point at surface — the peak reading is confirmed by re-sweeping from opposite directions to rule out sensor drift.

05

Confirm and mark — bring the leak to a workable margin

The peak surface reading is marked on tape with x-y offsets from a fixed reference (a wall corner, a doorframe, a boundary manhole). A second confirmation reading is taken from a right-angle direction. Pinpoint accuracy on hard surfaces is ±150mm; on soft ground ±300mm; on multi-layer floor build-ups (screed under tile under underlay) ±250mm. Every reading is timestamped and the survey ends with a written record of the pinpoint plus the ppm peak value.

When tracer gas is the right method

Four specific London scenarios where tracer gas is the correct detection method — usually after acoustic and thermal have already been tried:

Buried MDPE mains supply — acoustic gave no signal

Thames Water leak notice with 400–800 L/day loss, but a first acoustic survey at the property returned no signal — either the leak is too small for acoustic threshold or the MDPE plastic is absorbing the leak-noise. Tracer gas injection into the drained supply pipe at 3 bar, 20-minute migration wait, semiconductor sweep along the pipe route in the front garden pinpoints the leak within 200mm. Keyhole excavation at the marked point, not a full driveway rip-out.

Underfloor heating in screed under a heritage stone floor

UFH loop losing pressure but the finished floor is 300mm of reclaimed York stone slabs which acoustic cannot get useful readings through. The heating circuit is drained, charged with hydrogen 5/95, and the surface is scanned after 15 minutes migration. Peak reading identifies the leak at a specific point on a specific loop — one stone slab lifted, single push-fit repair, slab replaced. The alternative would have been lifting the entire kitchen floor.

Cold mains supply run under a bathroom floor slab

Water bill spike traced to the internal cold-water supply between the boundary stop and the kitchen — the run passes under a bathroom floor slab. Acoustic listening in the bathroom was inconclusive (bathroom fan and cistern refill noise). Tracer gas identified the leak below the WC pan where a compression joint was buried in the slab from a 1985 rewiring — the joint had been sitting under the pan for 40 years and finally corroded through.

Communal riser suspected leak — no accessible pipework

Mansion block reporting cold water main pressure fluctuations. The rising main is buried inside a plasterboard-boxed riser shaft with no access panels. Tracer gas injection at the plant room isolation point, migration wait, then a floor-by-floor sweep of the riser exterior identified the leak point behind a third-floor landing wall — the point where the riser passes through the intermediate concrete floor. Freeholder-authorised access hatch cut on the third floor, repair completed.

The tracer gas kit — sensors, gas, and calibration

The equipment specification matters more here than the survey time. A calibrated hydrogen-selective semiconductor sensor discriminates a real leak signal from atmospheric noise. Uncalibrated general-purpose gas sensors return false positives on methane, propane, or common cleaning products. Our tracer kit:

Sewerin Variotec 460 Tracergas

Handheld hydrogen-selective semiconductor gas detector. 0.5 ppm hydrogen sensitivity. Bluetooth data logging for report integration. The workhorse tracer sensor on every gas survey.

Inficon Sensistor ISH2000

High-sensitivity hydrogen leak detector for large sites and communal risers. Sub-ppm detection, faster response time than the Variotec on rapidly-migrating surface signatures.

Hydrogen 5/95 Forming Gas Cylinder

BOC-supplied hydrogen-nitrogen 5/95 (5% hydrogen, 95% nitrogen inert carrier) — the standard tracer mixture for leak location. Below the LEL of hydrogen at any air dilution, safe for domestic use with ventilation.

Calibrated Regulator + Injection Manifold

Pressure regulator with dial gauge and safety relief for injecting tracer gas at controlled pressure (typically 2–4 bar). Custom injection manifold with isolation valve for clean disconnect from the tested circuit.

Sewerin SNOOP+ Confirmation Fluid

Foaming leak-confirmation liquid applied at the pinpoint to visually confirm gas escape — belt-and-braces final check before the customer commits to structural opening.

Tracer gas vs the other detection methods

Every leak survey we run starts with acoustic. Tracer gas is deployed as a specialist follow-up when acoustic and thermal have already returned no clear result. The full method-suite comparison:

MethodWhen to useWhy tracer gas is different
Acoustic (first-pass)Any pressurised leak on accessible pipework — copper heating, hot water, hidden hot pipes. Typical ±100mm accuracy.Fast, non-invasive, no drain-down needed. Try acoustic first.
Thermal imaging (second-pass)Hot water/heating leaks in walls, floors, ceilings where pipe location is unknown. FLIR radiometric survey.Non-invasive, visualises leak effect on surface, ideal for hidden pipe routes.
Tracer gas (specialist)Small leaks below acoustic threshold, buried plastic MDPE where acoustic gives no signal, screed-covered UFH, sub-slab supply pipes.When acoustic and thermal both return no result but the pressure test confirms a real leak, tracer gas is the correct next method.
Moisture mappingConfirming leak spread pattern once located — supports trace-and-access insurance documentation.Not a pinpoint method itself — confirms wetness, not the leak source.
CCTVDepressurised leaks in wastes and drains — bath waste crack, dislodged shower trap, root ingress on below-ground.Only for drainage, not pressurised circuits.

Cost — tracer gas leak detection surveys

Every tracer gas survey is quoted in writing before the engineer travels. Prices below are indicative; exact cost depends on circuit complexity and drain-down access. The no find, no fee guarantee applies on every scope.

Survey scopePrice (inc. VAT)Includes
Buried supply pipe tracer gas survey (single-family home, up to 15m run)£400–£550System drain-down, tracer gas 5/95 injection at working pressure, migration wait, semiconductor surface sweep along pipe route, marked plan, written report within 24 hours
Long-run buried supply survey (over 15m or complex route)£550–£750Extended migration wait, dual-sensor confirmation, freeholder-format report
Underfloor heating tracer gas survey (single UFH loop)£450–£600UFH loop drain-down, gas injection, screed migration wait, floor sweep, pinpoint on loop diagram
Multi-loop UFH tracer survey (large residential or light commercial)£600–£850Loop-by-loop isolation, separate gas charges, individual pinpoints on manifold diagram
Sub-slab supply pipe tracer survey (leak under concrete floor)£500–£700Concrete floor sweep, extended migration allowance, cross-check with second sensor head
Communal riser tracer survey (mansion block, per riser)£550–£800Multi-storey sweep, plant-room injection, freeholder-format report with per-floor readings

Real London tracer gas surveys

Three recent tracer gas surveys, anonymised. Kit, method, pinpoint and outcome.

Highgate detached — buried MDPE mains under front garden

Thames Water leak notice 550 L/day. Prior acoustic survey by another firm returned no signal — plastic MDPE at low mains pressure produced sub-threshold hiss. Tracer gas 5/95 injected at 3 bar into the drained supply, 25-minute migration wait, Variotec 460 sweep along the 6.5m pipe route identified a 42 ppm peak at 4.2m from the boundary stop. Keyhole excavation exposed a fractured MDPE-to-copper coupling at 60cm depth with poor original clamping. Repair by fusion coupling and thrust plate, £820 including reinstatement.

Notting Hill mews conversion — UFH leak in engineered oak/screed

Ground-floor UFH loop losing 0.2 bar per day. Screed depth 65mm under 18mm engineered oak. Acoustic returned no signal through the multi-layer floor. UFH loop drained, hydrogen 5/95 injected at 2.5 bar, 18-minute migration wait, floor sweep pinpointed leak at 1.6m from the north wall. Single oak plank lifted, screed opened 200x200mm, pinhole in PEX-A loop repaired with push-fit coupling, plank re-laid. £780 total including engineered oak reinstatement.

Belgravia mansion block — communal cold rising main

Freeholder reporting unexplained cold-water top-up on the tank in the roof, no visible leaks in any flat. Rising main runs buried inside a plasterboard riser shaft between the plant room and the tank room, no access panels except at floor level. Tracer gas injected at plant room isolation, 20-minute migration wait, floor-by-floor sweep identified a 65 ppm peak at fourth-floor landing height. Freeholder-authorised access hatch cut, corroded pinhole leak on 42mm copper repaired by bronze compression sleeve. Full survey plus report £750, repair separately.

Tracer gas leak detection across every London borough

Same tracer method and semiconductor sensor kit across all 32 London boroughs. Click a borough for a page tailored to local plumbing patterns.

Frequently asked questions about tracer gas leak detection

Is hydrogen tracer gas safe to use inside a domestic property?
Yes — we use a 5/95 forming gas mixture (5% hydrogen, 95% nitrogen). The 5% hydrogen concentration is below the lower explosive limit (LEL) of hydrogen at any dilution with room air — the LEL is 4.1% in pure air, and diluting our 5% mixture with any amount of room air drops the hydrogen concentration below LEL immediately. There is no fire or explosion risk in normal domestic use. The nitrogen carrier is inert and displaces no oxygen in the volumes used. Every survey is conducted with the room ventilated (window open, extractor fan on) as belt-and-braces good practice.
Why hydrogen rather than helium or another tracer gas?
Hydrogen is the smallest molecule in nature — atomic diameter around 0.7 angstroms compared to helium at 0.98 angstroms and CO₂ at 3.3 angstroms. That size difference matters at the leak point. A hairline crack in a copper pipe or a partially-corroded compression olive will leak liquid water and hydrogen but will not leak helium or CO₂ reliably. Hydrogen tracer gas therefore finds leaks that helium tracer methods miss. Hydrogen is also cheaper than helium and readily available through BOC as forming gas.
When is tracer gas the right method rather than acoustic or thermal?
Tracer gas is the specialist method for four specific scenarios: (1) buried plastic MDPE supply pipes where acoustic returns no signal because plastic absorbs the leak-noise; (2) very small leaks below the acoustic threshold — typically below 5 ml/min flow rate; (3) screed-covered underfloor heating where both acoustic and thermal are limited by the floor build-up; (4) leaks under solid concrete floor slabs where surface access for other methods is impractical. On any leak where acoustic and thermal produce clear results, tracer gas is unnecessary and we do not deploy it.
Does the circuit need to be drained before tracer gas is injected?
Yes — residual water in the pipe absorbs hydrogen tracer gas and produces false-negative or delayed surface readings. Every tracer gas survey begins with a full drain-down of the circuit under investigation. On a buried supply pipe survey the internal stop is closed and downstream taps opened until pressure drops to zero. On a heating or UFH circuit the loop is drained via the filling loop. The drain-down takes 15–30 minutes typically and is included in the survey timing.
How long does a tracer gas survey take on site?
Buried supply pipe surveys 90–150 minutes including drain-down, gas injection, migration wait, and semiconductor sweep. UFH tracer surveys 120–180 minutes because the migration wait through screed is longer. Communal riser surveys 150–210 minutes per riser. All timings include preparation, injection, migration, sweep, and confirmation. Written report emailed within 24 hours of the survey visit.
What accuracy can I expect from the pinpoint?
Pinpoint accuracy depends on the surface type and structure between the leak and the sensor. On hard surfaces (concrete, tarmac) ±150mm. On soft ground (planted garden, block paving) ±300mm. On multi-layer floors (screed under tile under underlay) ±250mm. On communal risers with plasterboard build-ups ±200mm at the correct floor level, ±0 for identifying which floor level. Every pinpoint is confirmed by cross-sweep from a different direction before we mark the plan.
Do you carry the hydrogen forming gas cylinder yourself?
Yes — every tracer gas survey includes a fresh BOC-supplied hydrogen 5/95 forming gas cylinder brought by the engineer to site. There is no cost passed to the customer for the gas itself, and no cylinder rental appears on the invoice — it is included in the survey fee. The gas quantity used on a typical survey is a small fraction of the cylinder capacity, which allows us to service multiple properties from the same cylinder.
Will the tracer gas damage plumbing or fittings?
No — hydrogen 5/95 is entirely inert with respect to copper, brass, plastic, PEX, MDPE, rubber olive seals, and all standard plumbing materials. Nitrogen is inert by definition. Hydrogen at the 5% concentration and typical mains pressures used in surveys does not react with pipework or fittings and produces no residue. The gas is vented from the system after the survey via the same drain-down points used for filling.
Do you carry public liability insurance?
Yes — £5,000,000 public liability via SiriusPoint International Insurance Corporation (UK Branch) acting through Eaton Gate MGU Ltd, policy number BE26ACTT000000018221, current period 07/05/2026 to 06/05/2027. The certificate is issued with every survey quote — managing agents and freeholders routinely require the certificate before authorising access, particularly on communal riser surveys.
What certifications do your engineers hold?
Our senior leak detection engineer holds Water Regulations 1999 competency (WaterSafe registration, UK Certification Ltd certificate 136356 issued 8 September 2025, expiry 18 August 2030) and HWSS G3 unvented hot water certification (certificate 136359, same date range). Water Regulations competency is required to work on any pressurised drinking-water supply — including the temporary disconnection during a tracer gas survey. Both certificates are supplied to the loss adjuster on any insurance-backed survey.

Related plumbing services

Book a tracer gas leak survey

Written no-find-no-fee quote before the engineer travels. Semiconductor pinpoint report within 24 hours.

Request a Survey Quote
Call 020 7101 0629