A hydrogen test bench is tested for leaks again after several tube fittings have been replaced. During a pressure holding test, the measured pressure decreases slightly. Is a connection actually leaking, has the gas temperature simply changed, or is the cause located in another component of the test setup? A portable H₂ leak detector subsequently indicates an increased hydrogen concentration at one of the fittings. But is this result sufficient to demonstrate that the complete assembly meets its permissible leak rate?
When detecting hydrogen leaks, different test methods must be clearly distinguished. Pressure decay measurement can provide information about the overall leak tightness of a sealed test volume. Sniffer testing, on the other hand, is primarily used to detect escaping gas at a particular location and identify the source of a leak. The use of a suitable test gas allows detection sensitivity and safety-related testing conditions to be adapted further.
Hydrogen places particular demands on these methods. The gas has high diffusivity, can permeate through certain sealing materials and may adversely affect the mechanical properties of metallic components under unfavourable conditions. Further considerations include its wide flammability range in air, low ignition energy and the often high operating pressures found in storage systems, electrolysis plants, fuel cell systems and test benches.
This technical article explains how pressure decay testing, H₂ sniffer testing and tracer gas methods can be combined correctly. It focuses on defining a permissible leak rate, temperature-related pressure changes, hydrogen-compatible materials and fittings, and selecting suitable pressure transducers and gas leak detectors. A calculation example demonstrates why even a small temperature change can cause a substantial apparent pressure loss.
Table of Contents
- Define the Test Objective, Scope and Permissible Leak Rate
- Consider Hydrogen-Specific Properties During Leak Detection
- Distinguish Pressure Decay, Sniffer Testing and Tracer Gas Methods
- Define Test Volume, Fittings and System Boundaries
- Check the Sealing Principle and Mechanical Condition of the Fitting
- Evaluate Material Compatibility, Permeation and Embrittlement
- Correctly Evaluate a Pressure Decay Test
- Account for Temperature Changes and Pressure Stabilisation
- Calculation Example: Pressure Decay with Temperature Correction
- Distinguish Pressure Loss, Leak Rate and Acceptance Limit
- Select Hydrogen-Compatible Pressure Transducers and Measuring Ranges
- Locate Leaks Using H₂ Sniffer Testing
- Evaluate Sniffing Distance, Background Concentration and Response Time
- Use Forming Gas Containing Hydrogen as a Tracer Gas
- Consider Helium as an Alternative Test Gas
- Combine the Three Test Methods Effectively
- Check Operating Pressure, Temperature Cycling and Dynamic Loads
- Consider Explosion Protection, Test Gas Handling and Safe Pressure Testing
- Perform a Systematic H₂ Leak Test Procedure
- Evaluate Measurement Uncertainty, Calibration and Detection Limits
- Diagnose Typical Errors and Misinterpretations
- Suitable Hydrogen Measurement and Leak Detection Technology from ICS Schneider
- Conclusion: Verify Overall Leak Tightness and Local Leak Detection Separately
- Frequently Asked Questions About Hydrogen Leak Detection at Fittings
1. Define the Test Objective, Scope and Permissible Leak Rate
Before beginning a leak test, it must be clearly established what the test is intended to demonstrate. Is the objective to verify that an entire hydrogen line is sufficiently leak-tight? Is it necessary to locate a suspected leak? Or must compliance with a defined maximum leak rate be demonstrated for an individual fitting?
These test objectives require different measurement methods and acceptance criteria.
During a pressure holding test, for example, the pressure change in an enclosed test volume is evaluated over a defined period. In a sniffer test, the objective is to determine whether escaping test gas can be detected at specific locations.
A negative sniffer test result initially means only that no signal above the relevant detection or evaluation threshold was identified under the test conditions applied. It does not constitute general proof of absolute leak tightness.
For a quantitative leak tightness requirement, a permissible leak rate must be specified. This can be expressed in mbar·l/s, Pa·m³/s or an appropriate mass flow or reference volumetric flow unit, for example.
The requirements depend on the application. A hydrogen line in a test bench is subject to different conditions from a storage system experiencing numerous pressure cycles or a connection within a fuel cell installation.
The safety consequences of a leak must also be considered. Even a small leak may be relevant, particularly in enclosed housings, poorly ventilated areas or locations with potential ignition sources.
Before selecting the equipment, the test object, test medium, test conditions, required sensitivity and permissible leak rate should therefore be documented. Without this information, it is not possible to determine reliably whether a simple pressure decay test is sufficient or whether a more sensitive tracer gas method is required.
2. Consider Hydrogen-Specific Properties During Leak Detection
Hydrogen differs significantly from many other technical gases in its physical properties. It is colourless and odourless and has a very low molar mass.
Under atmospheric conditions, hydrogen is considerably lighter than air. Depending on airflow and environmental conditions, escaping gas can therefore rise rapidly. However, hazardous concentrations can still accumulate in enclosed rooms or beneath structures.
Another important characteristic is its high diffusivity. Hydrogen can permeate through certain materials and sealing compounds that appear comparatively leak-tight when used with other gases.
Hydrogen can also affect metallic materials under particular combinations of material condition, temperature, pressure and mechanical stress. Hydrogen-assisted damage and embrittlement mechanisms are especially relevant.
For leak testing, this means that a system showing a low leak rate with another gas does not necessarily have the same leak rate when exposed to hydrogen.
Flammability is another particularly important consideration. Under typical atmospheric conditions, the flammability range of hydrogen in air is approximately 4 to 75 vol.%. Actual flammability is also influenced by temperature, pressure, oxygen concentration and other conditions.
Hydrogen leak detection must therefore always be accompanied by a suitable risk assessment. Selecting a sensitive measuring instrument does not replace safe test conditions, appropriate ventilation or monitoring of the surrounding environment.
Even with very small leaks, a clear distinction must be made between a technically permissible leak rate and a release that is acceptable from a safety perspective.
3. Distinguish Pressure Decay, Sniffer Testing and Tracer Gas Methods
Different testing methods are available for examining hydrogen fittings. They complement each other because they assess different aspects of leak tightness.
| Test Method | Measured Quantity / Result | Main Advantage | Important Limitation |
|---|---|---|---|
| Pressure decay testing | Pressure change in an enclosed test volume over time | Assessment of overall leak tightness under defined conditions | No direct leak location; sensitive to temperature and test volume |
| Local H₂ sniffer testing | Detection of an increased H₂ concentration at the test probe | Locating leaks at fittings and connections | Reading depends on gas distribution, measurement distance and instrument characteristics |
| Hydrogen tracer gas testing | Detection of a defined H₂ concentration in a suitable test gas mixture | Targeted leak detection using a defined test medium | Quantitative leak rate only with a suitable, appropriately calibrated method |
| Helium tracer gas testing | Helium-specific detection using a suitable leak detector | Sensitive leak testing for appropriate applications | Requires a separate test medium, suitable instrumentation and adjusted evaluation |
A pressure decay test can reveal that the overall leak tightness of an isolated volume is questionable. However, it normally provides no information about which fitting is leaking.
Sniffer testing, on the other hand, allows the leak to be localised. It is particularly useful when several connections are positioned close together and a specific fitting needs to be examined.
Special tracer gas leak detectors may be required for quantitative verification of small leak rates. Such instruments are designed for a defined test gas and the corresponding measuring range.
A portable gas leak detector displaying an H₂ concentration in ppm is therefore not automatically a calibrated leak rate measuring instrument with a reading in mbar·l/s.
Selection depends on the test objective. For reliable acceptance testing, combining several methods may be more useful than relying on a single instrument.
4. Define Test Volume, Fittings and System Boundaries
A pressure decay test evaluates the gas volume that has actually been isolated from the rest of the system. Before testing, it is therefore necessary to establish which pipes, valves, sensors and connecting components belong to the test volume.
The volume of test adapters, measurement lines and connected pressure transducers may also be relevant.
For the same small leak rate, a large test volume produces a slower pressure change than a small one. The size of the test volume therefore directly influences the sensitivity of pressure decay measurement.
The isolation valves used must be suitable for the test. Internal leakage through an isolation valve, for example, can cause a pressure decrease even though the fittings under examination are leak-tight.
During the actual holding period, the pressure supply must also be safely and clearly isolated from the test volume according to the selected test method.
Test connections and temporary adapters require particular attention because they can introduce additional potential leak points.
For a systematic investigation, clearly defined test boundaries should therefore be documented. These include the locations of isolation valves, all components contained within the test volume and the assumed or determined internal volume.
In extensive installations, it may be useful to test individual sections separately. This can make it easier to associate an unusual pressure change with a particular part of the installation.
However, such subdivision must follow an approved test concept. In particular, safety-related pressure limits and the function of protective valves must not be changed through improvised isolation measures.
5. Check the Sealing Principle and Mechanical Condition of the Fitting
A tube fitting is not necessarily leak-tight simply because its thread has been tightened firmly. The decisive factor is where the sealing function is intended to take place within its design.
In a compression or cutting-ring fitting, sealing is achieved through the intended interaction between the tube, fitting body and compression or sealing element.
In other connections, sealing may be achieved using a metallic cone, a defined sealing surface, a suitable O-ring or another manufacturer-specific design.
The thread often serves primarily to provide mechanical clamping force. It is not automatically the gas-sealing surface.
For hydrogen leak detection, the connection should therefore first be identified unambiguously. A tapered thread seal requires different conditions from a parallel-threaded connection with a separate seal.
Correct tube preparation is also important. Unsuitable tube dimensions, damaged surfaces, incorrect insertion depths or improper initial assembly can impair the sealing function.
Excessive tightening torque is likewise not a universal solution for eliminating leaks. It can damage sealing surfaces, overload threads or permanently deform a compression fitting.
A suspect fitting must therefore not be retightened under pressure without corresponding manufacturer instructions.
A more detailed assessment of thread types, adapters and sealing principles is provided in the ICS technical article Checking Hydrogen Measurement Points for Leaks: Correctly Evaluating Threads, Adapters and Leak Rates.
6. Evaluate Material Compatibility, Permeation and Embrittlement
For hydrogen applications, all wetted components must be suitable in terms of material, pressure, temperature and operating conditions.
This applies not only to the piping. Fitting bodies, seals, valve seats, pressure sensor diaphragms and test connections are also part of the wetted measurement and testing system.
Depending on the material and operating conditions, hydrogen can permeate through elastomer seals. This process must be distinguished from a localised leak at a damaged sealing surface.
Permeation can produce a measurable gas release during sensitive leak testing even when no individual mechanical defect exists.
Different effects are relevant to metallic materials. Under certain conditions, hydrogen can affect their mechanical properties and increase the risk of hydrogen-assisted cracking.
Not every stainless steel grade is equally suitable for all hydrogen operating conditions. Even within one material group, strength, microstructure, surface condition and mechanical loading can influence suitability.
A general specification such as “stainless steel” or “FKM seal” is therefore insufficient for a demanding high-pressure hydrogen application.
Repeated pressure cycling and unfavourable combinations of high mechanical stress with conditions promoting hydrogen-induced damage are particularly critical.
The test gas used also influences the assessment. A leak rate determined using nitrogen cannot automatically be treated as an identical value for hydrogen.
The suitability of wetted materials must be demonstrated using the specific manufacturer’s approval and the actual operating conditions.
7. Correctly Evaluate a Pressure Decay Test
During pressure decay testing, a defined test volume is pressurised with a suitable test medium under approved conditions. The pressure change is then monitored over a specified period.
For evaluation, the pressure difference between the beginning and end of the holding period is often calculated first:
Δp = p1 − p2
However, the pressure change alone is not a complete leak rate. It must be evaluated in relation to the test volume, measurement period and thermodynamic conditions.
Under simplified isothermal conditions with a constant test volume, an equivalent pressure-volume rate can be estimated:
QpV ≈ V · Δp / Δt
Here, V is the test volume and Δt is the time interval considered.
If pressure is entered in mbar, volume in litres and time in seconds, the resulting value is expressed in mbar·l/s.
This relationship represents a simplified evaluation under the stated assumptions. The value is not automatically a directly transferable reference volumetric flow rate or hydrogen mass leak rate under different operating conditions.
More demanding tests must also consider changes in temperature, ambient pressure and, where applicable, test volume.
Data acquisition also plays an important role. Short pressure fluctuations, unsuitable averaging or an insufficiently stabilised test system can lead to incorrect conclusions.
Pressure decay testing is therefore particularly informative when the complete test volume and relevant operating conditions are known, and the pressure measurement equipment has sufficient resolution and sufficiently low uncertainty.
8. Account for Temperature Changes and Pressure Stabilisation
Gas pressure changes not only because of leakage but also because of temperature variations. In an enclosed volume, even slight cooling can produce a measurable pressure decrease.
For a simplified model using an ideal gas, constant volume and unchanged gas quantity:
p · V / T = constant
At constant volume, this gives:
p2 / p1 = T2 / T1
Absolute pressures and absolute temperatures in kelvin must be used.
Temperature changes are particularly relevant when the test object has recently been filled with gas. During filling, temperature differences may occur between the supplied gas, piping and surrounding environment.
Subsequent thermal equalisation can cause a pressure change even when no leak is present.
Solar radiation, air movement and changing room temperatures can also influence longer pressure holding tests.
Measuring ambient temperature alone is not always sufficient. Accurate correction requires the relevant gas temperature or another suitable, representative thermal description of the test volume.
In a large test object, different temperature zones may develop. In this case, a single temperature measurement point may be insufficient.
For a simple, approximately homogeneous condition, the measured final pressure can be referenced to the initial temperature:
p2,T1 = p2 · T1 / T2
The temperature-corrected pressure difference is then:
ΔpT-korr = p1 − p2,T1
This correction assumes that the simplified gas model and the temperature representation are sufficiently suitable for the test setup.
At high pressures, deviations from ideal gas behaviour may become relevant. In such cases, a real-gas model may be required.
9. Calculation Example: Pressure Decay with Temperature Correction
A simplified calculation example considers an enclosed test volume of 10 litres. At the beginning of the observation period, the absolute pressure is 6.000 bar at a gas temperature of 20.0 °C.
After 20 minutes, an absolute pressure of 5.990 bar and a temperature of 19.6 °C are measured.
The directly observed pressure decrease is therefore 10 mbar. Without considering temperature, the entire decrease could incorrectly be attributed to leakage.
| Parameter | Initial Value | Final Value |
|---|---|---|
| Test volume | 10.0 l | 10.0 l |
| Absolute pressure | 6.000 bar | 5.990 bar |
| Gas temperature | 20.0 °C | 19.6 °C |
| Absolute temperature | 293.15 K | 292.75 K |
| Observation time | 0 min | 20 min |
| Observed pressure decrease | 10.0 mbar | |
First, the final pressure corresponding to the measured final condition is calculated at the initial temperature of 20.0 °C:
p2,T1 = 5.990 bar · 293.15 K / 292.75 K
Rounded, this gives:
p2,T1 ≈ 5.9982 bar
The temperature-corrected pressure difference is therefore only approximately:
ΔpT-korr = 6.0000 bar − 5.9982 bar ≈ 0.0018 bar
This corresponds to approximately 1.8 mbar.
Under the simplified assumption that this remaining pressure decrease is actually caused by leakage, the following equivalent pressure-volume rate is obtained:
QpV ≈ 10 l · 1.8 mbar / 1,200 s
QpV ≈ 0.015 mbar·l/s
Without temperature correction, the same approximate calculation would give:
QpV,unkorr = 10 l · 10 mbar / 1,200 s
QpV,unkorr ≈ 0.083 mbar·l/s
The difference is substantial. In this example, a large proportion of the observed pressure decrease can already be explained by the small temperature drop.
However, the result still does not prove that a leak is present. The remaining pressure difference may also result from measurement uncertainty, an inadequate temperature model or other influences.
The figures are intended solely to illustrate the measuring principle. They do not specify a permissible test pressure or a universally applicable leak rate acceptance limit.
For an actual hydrogen leak test, the test volume, pressure, temperature distribution, gas type and measurement uncertainty must be considered according to the intended application.
10. Distinguish Pressure Loss, Leak Rate and Acceptance Limit
The terms pressure loss and leak rate are frequently used interchangeably in everyday industrial practice. However, they describe different quantities from a measurement perspective.
Pressure loss describes a change in pressure. Leak rate, on the other hand, describes the quantity of gas passing through a leak per unit of time under defined conditions.
For gases, a pressure-volume leak rate is frequently expressed in mbar·l/s or Pa·m³/s.
The conversion between these units is:
1 mbar·l/s = 0.1 Pa·m³/s
Leak rates may also be expressed as mass per unit time or as volumetric flow under specified reference conditions.
A volumetric flow expressed in ml/min can only be compared unambiguously with a leak rate if pressure, temperature and any normal or reference conditions are known.
Dependence on gas type is particularly important. A defined leak path can produce different gas flow rates with hydrogen, nitrogen or helium.
The pressure difference across the leak also affects the actual flow rate. At high pressure ratios, different flow conditions may occur compared with testing at low differential pressure.
A leak rate determined using nitrogen or forming gas must therefore not automatically be reported as an identical hydrogen leak rate at operating pressure.
The permissible acceptance limit must refer to the actual specified test conditions and the method used.
Quantitative verification also requires suitable reference leaks, calibration methods or other designated testing equipment.
11. Select Hydrogen-Compatible Pressure Transducers and Measuring Ranges
The pressure measurement equipment used for a hydrogen leak test must be suitable for both the test medium and the pressure and temperature conditions present.
Resolution, absolute measurement uncertainty, long-term stability and temperature-related signal changes are particularly relevant during pressure decay measurement.
A pressure transducer with a very large measuring range may be suitable for monitoring high operating pressures. However, its absolute resolution or uncertainty may be insufficient for detecting an extremely small pressure change over a longer period.
Pressure measurement should therefore be designed so that the expected pressure loss can be distinguished with sufficient confidence from measurement noise, zero-point drift and thermal influences.
The distinction between gauge pressure and absolute pressure is also important.
A gauge pressure transducer measures relative to ambient pressure. If atmospheric pressure changes during a longer test, this can influence the indicated gauge pressure.
However, absolute gas pressure is required for thermodynamic calculations. When gauge pressure measurement is used, the corresponding ambient pressure must therefore be properly accounted for.
The Druck UNIK 5000H offered by ICS is specifically designed for demanding hydrogen applications. It features hydrogen-compatible wetted materials and is available with different pressure ranges and output signals.
The available measuring ranges extend from 700 mbar to 700 bar. Depending on the configuration, accuracy options of up to ±0.04 % of full scale according to BSL are specified.
These characteristics may be useful for pressure monitoring in hydrogen test benches. However, whether the particular sensor is sufficiently sensitive for a specific leak rate and holding time must be evaluated separately using the measurement uncertainty and test volume.
A suitable hydrogen pressure transducer is therefore an important component of the test setup but does not, by itself, constitute a complete leak rate measurement system.
12. Locate Leaks Using H₂ Sniffer Testing
Sniffer testing is used to detect escaping gas at specific locations on an assembly. It is particularly useful for tube fittings, flanges, valve seats and test connections.
The instrument has a sensor or probe used to examine the area around a potential leak.
With an active sniffer probe, the gas sample is transported to the sensor using a small pump. Other gas leak detectors use a designated direct gas detection method in the vicinity of the probe tip.
A locally increased H₂ signal may indicate escaping hydrogen. Its interpretation depends on the gas type, background signal and suitability of the detector.
For practical leak localisation, the probe is moved along the relevant connections in accordance with the manufacturer’s instructions. The measurement point must be sufficiently accessible without exposing personnel to hazardous high-pressure gas jets or potential ignition sources.
Where connections are closely spaced, escaping gas may also be detected at neighbouring locations because of air movement. The local assignment of a signal may therefore need to be confirmed through repeated, controlled measurement.
The testo 316-2 listed by ICS has an integrated pump and a flexible probe. It is designed to detect methane, propane and hydrogen.
For hydrogen, the ICS product description specifies a measuring range from 10 ppm to 4.0 vol.% and an instrument-specific response time of t₉₀ less than 2 seconds.
However, a local concentration reading does not automatically allow direct determination of the quantity of gas escaping per unit of time. A suitable quantitative leak testing method and defined test conditions are required for this purpose.
13. Evaluate Sniffing Distance, Background Concentration and Response Time
The result of a sniffer test depends on more than the size of the leak.
One important factor is the position of the probe. As the distance from the leak increases, the gas can become more diluted by the surrounding air.
Air movement also influences gas distribution. Near a fan or open door, the local gas concentration may differ from that in an area with little air movement.
For an active probe, the aspirated volumetric flow rate is another factor. A blocked probe filter or damaged sampling line can impair detection sensitivity.
The speed at which the probe is moved must match the detector response time. If the probe passes too quickly over a connection, a short-duration local gas concentration peak may not be detected adequately.
The background signal must also be considered. In an environment where low hydrogen concentrations are already present, distinguishing the general background from a small local leak can become more difficult.
For sensitive hydrogen sniffer testing, the manufacturer INFICON points out, among other things, that exhaled breath can influence the H₂ background concentration.
Zero adjustment, background assessment and the test method specified by the manufacturer must therefore be coordinated.
Quantitative sniffer testing additionally requires calibration with suitable reference leaks, defined probe movement and appropriate evaluation.
High sensitivity stated in a data sheet alone does not demonstrate that every leak can be reliably detected under all environmental conditions.
14. Use Forming Gas Containing Hydrogen as a Tracer Gas
For certain leak testing applications, hydrogen is not used as a pure gas but as a component of a defined test gas mixture.
A common example is forming gas containing 5 vol.% H₂ and 95 vol.% nitrogen.
This gas mixture is used in suitable leak testing systems because the hydrogen component can be detected with appropriately sensitive detectors.
The manufacturer INFICON describes corresponding applications for leak testing hydrogen components and containers.
However, the use of such a gas mixture must be matched to the specific test installation and approved testing procedure. The flammability classification and safe use of a defined mixture depend on its composition, relevant safety data and actual operating conditions.
The term forming gas alone does not constitute universal safety approval.
During tracer gas testing, it is also essential that the detector used is suitable for the hydrogen concentration and intended measurement method.
A detector generally responds to the detectable tracer gas component. Without suitable calibration and conversion, its reading must not be interpreted as the leak rate of pure hydrogen at the subsequent operating pressure.
The difference relates both to the concentration of the test gas and to its flow characteristics.
For quantitative testing, it must therefore be documented which gas mixture was used, its concentration, how the measuring system was calibrated and which leak rate or reference conditions apply to the acceptance criterion.
Forming gas can complement local leak detection. Whether it is sufficient for the required final leak tightness verification of the particular hydrogen installation must be determined separately.
15. Consider Helium as an Alternative Test Gas
Helium is frequently used as a tracer gas for sensitive leak testing methods.
Special helium leak detectors are available for this purpose and are designed for the test gas and respective testing method.
This method is particularly useful when small leak rates must be detected reliably or when hydrogen testing under the intended conditions is not the appropriate first test stage.
The European Industrial Gases Association (EIGA) explicitly describes the possibility of helium leak testing for gaseous hydrogen installations, for example using a suitable nitrogen-helium mixture.
The reason is that passing a nitrogen leak test alone is not always sufficient to demonstrate the required hydrogen leak tightness.
A helium leak test has its own test parameters, including helium concentration, pressure conditions, detector type and required sensitivity.
The leak rate determined using helium is not automatically identical to that of hydrogen under subsequent operating conditions.
Particularly for demanding quantitative tests, conversion or evaluation must therefore be performed using a suitable model or the corresponding test standard.
ISO 20485:2017 describes methods for non-destructive leak testing using tracer gas and an appropriate gas-specific leak detector.
The standard can therefore provide a methodological basis for suitably defined test procedures. Its applicability depends on the test object and specific technical requirements.
For equipment selection, it must consequently be established whether qualitative leak localisation is sufficient or whether quantitative tracer gas testing with a defined detection limit is required.
16. Combine the Three Test Methods Effectively
Combining pressure decay testing with local sniffer testing can significantly improve the reliability of a leak tightness assessment.
Pressure decay measurement initially evaluates the behaviour of an isolated volume. If an unusual pressure change is detected, suitable local gas leak testing can help narrow down the source spatially.
A tracer gas method can be used as an additional procedure, particularly where the required detection sensitivity cannot be reliably achieved using a simple pressure decay test.
A possible testing strategy therefore consists of several technically distinct stages:
Define the test object and acceptance criteria → Perform a suitable preliminary test using an approved test medium → Carry out quantitative overall leak tightness testing where required → Locate leaks using a suitable tracer gas → Repair → Repeat acceptance testing
This sequence is a general concept and not a universally approved pressure testing instruction.
In particular, combining different testing methods must not be interpreted as a requirement to pressurise every installation with hydrogen at every stage.
The selection and sequence of test media depend on the safety-related design and approved testing procedure.
One important advantage of combining methods is that a suspicious calculated pressure change is not prematurely attributed to a particular fitting.
A positive sniffer signal must also be interpreted through an appropriate assessment of the test object. Detecting a local leak does not, by itself, establish whether the complete assembly meets its quantitative leak tightness criterion.
After a repair, it is therefore not sufficient to examine only the connection that was worked on. The overall leak tightness required by the test scope must also be verified again in accordance with the test plan.
17. Check Operating Pressure, Temperature Cycling and Dynamic Loads
A fitting may show no problems under static test conditions but behave differently during actual operation.
Important influencing factors include differential pressure, temperature, mechanical movement, vibration and the number of load cycles.
Hydrogen test benches and fuel cell systems may, for example, experience repeated pressure cycles. These loads affect the mechanical stress on the fitting and potentially the condition of its sealing elements.
Temperature changes can also alter the contact pressure. Metallic and non-metallic components have different thermal expansion characteristics.
A connection that performs satisfactorily at room temperature does not necessarily exhibit the same leak rate under all intended high- or low-temperature conditions.
At high operating pressures, the pressure dependence of the gas flow through an existing leak path must also be considered.
Leakage behaviour determined at low test pressure cannot always be extrapolated linearly to substantially higher operating pressures.
The properties of the seal may also change under pressure. For elastomers, relevant factors include compression, swelling, permeation and ageing.
Additional testing or evaluation under representative operating conditions may therefore be necessary for dynamically loaded installations.
The test concept must always remain within the approved load limits of all components.
Leak testing also does not automatically replace verification of mechanical strength or any required pressure test under the applicable technical regulations.
18. Consider Explosion Protection, Test Gas Handling and Safe Pressure Testing
Hydrogen leak detection presents particular safety risks. In addition to the potential ignition of escaping hydrogen, the stored energy of a pressurised gas system must be considered.
Testing pressurised hydrogen systems must therefore only be carried out according to a suitable, approved procedure and by appropriately qualified personnel.
Before testing, the system boundaries, permissible test pressure, gases used, pressure relief arrangements and required protective measures must be defined.
During pneumatic testing in particular, sudden component failure can present a serious hazard because of the release of stored pressure energy.
Safe discharge of test gas must also form part of the test plan. Hydrogen must not be released without control into enclosed areas or near potential ignition sources.
Work on existing hydrogen installations also requires suitable procedures for isolation, depressurisation, inerting and recommissioning.
EIGA guideline DOC 15/21 describes requirements for pressure testing, leak testing and purging before introducing hydrogen into gaseous hydrogen installations, among other topics.
In particular, the guideline points out that nitrogen testing alone may be insufficient for final verification of hydrogen leak tightness.
When locating leaks, the suitability of the instrument for the relevant potentially explosive atmosphere must also be checked.
Any Ex marking must correspond to the actual equipment category, gas group, temperature class and intended operating conditions.
A gas leak detector is also not automatically a complete system for monitoring and safeguarding the surrounding workplace.
Necessary protective functions may require additional suitable fixed or portable hydrogen gas detectors, ventilation equipment and other protective measures.
Requirements for stationary hydrogen detectors are addressed, among other documents, in ISO 26142:2010. Selection of the complete safety system must also be based on the specific installation.
19. Perform a Systematic H₂ Leak Test Procedure
A reproducible leak testing procedure must consider both the measurement task and the relevant safety requirements.
The following steps describe a general test structure. Specific pressures, gases, holding times, connection methods and safety measures must be taken from a procedure approved for the particular test object.
- Identify the test object: Record the assembly, fittings, materials, sealing principles and permissible operating conditions.
- Define the test objective: Determine whether overall leak tightness verification, local leak detection or quantitative leak rate measurement is required.
- Establish acceptance criteria: Document the permissible leak rate, required sensitivity and applicable testing requirements.
- Perform a risk assessment: Evaluate stored pressure energy, hydrogen release, ignition sources, ventilation and necessary protective functions.
- Approve the testing procedure: Define the test medium, pressure stages, test durations and safe implementation according to applicable requirements.
- Document the test boundaries: Record all included lines, valves, sensors, adapters and the relevant test volume.
- Inspect connections: Check fittings, tube geometry, sealing surfaces and available assembly information.
- Prepare measuring equipment: Verify the suitability and calibration status of pressure transducers, temperature measurement equipment and any leak detectors.
- Perform a suitable preliminary test: Carry out the initial test specified for the installation under the approved conditions.
- Allow for stabilisation: Observe the required thermal and measurement-related equalisation periods.
- Record pressure and temperature profiles: Document the relevant measurements with sufficient time resolution.
- Evaluate pressure decay: Assess the measured pressure change while considering temperature, test volume and uncertainty.
- Prepare local leak detection: Where required, use a suitable tracer gas method according to the approved test plan.
- Examine fittings: Use the leak detector according to the manufacturer’s procedure and clearly identify suspect connections.
- Check the plausibility of findings: Consider the background signal, possible permeation and nearby gas sources.
- Arrange necessary repairs: Work on defective connections only after safe depressurisation and in accordance with the manufacturer’s instructions.
- Repeat testing: Perform the prescribed leak tightness verification again after repairs.
- Complete the test: Safely discharge the test gas, restore the intended system condition and document results and approvals.
The exact sequence may vary depending on the type of installation and applicable test specification.
For high-pressure hydrogen systems, the steps described must not be interpreted as independent authorisation to introduce pressurised gas. The safe testing method must be established by the responsible technical personnel.
It is particularly important that the test is not regarded as completed simply because a local sniffer signal disappears.
Full acceptance must demonstrate compliance with the previously defined criteria.
20. Evaluate Measurement Uncertainty, Calibration and Detection Limits
The reliability of a leak test depends significantly on the measuring instruments used and their suitability for the particular testing method.
During pressure decay measurement, pressure sensor uncertainty, zero-point drift, temperature measurement, test volume and observation time can all contribute to the overall uncertainty.
Changes in ambient pressure and differences between the measured component temperature and actual gas temperature may also contribute.
If the expected pressure decrease is smaller than or comparable to the relevant uncertainty contributions, the pressure curve may not provide reliable evidence of compliance with the required leak rate.
A more sensitive leak testing method may be necessary for such applications.
During sniffer testing, detector sensitivity, calibration, background gas, test gas concentration, probe position and measurement duration are important influencing factors.
A portable H₂ detector with a ppm display may be very suitable for locating leaks. However, obtaining a quantitative leak rate in mbar·l/s requires a designated calibration and evaluation procedure.
Suitable reference leaks or testing procedures specified by the manufacturer are used to calibrate quantitative tracer gas leak detectors.
The resulting measurement uncertainty must be appropriate for the required acceptance limit.
Test conditions must also be comparable. A leak rate determined at low test pressure must not be compared with a limit for substantially different operating pressures without suitable evaluation.
Complete documentation therefore includes not only the numerical value but also the gas type, concentration, differential pressure, temperature and relevant measurement or calibration scope.
For qualitative leak localisation, the detection limit or decision criterion associated with the method must instead be clearly stated.
The statement “no leak detected” has limited meaning unless the test sensitivity and conditions are described.
21. Diagnose Typical Errors and Misinterpretations
Unusual results during a hydrogen leak test may originate from actual leaks, the measurement arrangement or the interpretation of the testing method.
A systematic diagnosis therefore distinguishes between the mechanical connection, test medium used and measuring technology.
| Observation | Possible Cause | Suitable Check |
|---|---|---|
| Pressure decreases slightly during the holding test | Temperature change, actual leak or measurement drift | Evaluate absolute pressure, relevant gas temperature and test conditions together |
| Pressure decreases only immediately after filling | Thermal equalisation or insufficient stabilisation | Assess stabilisation behaviour according to the intended test plan |
| Pressure decay without a detectable local H₂ signal | Leak outside the examined area, unsuitable sniffing conditions or another pressure-related influence | Check test boundaries, complete gas path and detection sensitivity |
| H₂ sniffer indicates a signal at several adjacent fittings | Gas distribution caused by air movement or multiple leaks | Check background concentration and local signal assignment using a suitable procedure |
| Leak signal increases when the probe remains at one location for longer | Gas accumulation, detector response time or local release | Evaluate instrument-specific probe movement, background signal and measurement duration |
| Nitrogen test passed, but hydrogen leak test indicates a problem | Gas-dependent differences in leakage or permeation characteristics | Check dependence on test gas and hydrogen compatibility of the components |
| Pressure signal fluctuates despite unchanged system conditions | Temperature drift, electrical noise or unsuitable pressure measuring range | Check the pressure transducer and measurement chain using a suitable reference |
| Leak signal at an elastomer seal without visible damage | Possible permeation or local leakage | Evaluate the material, test medium, operating duration and sealing concept |
| Leak reappears after several pressure cycles | Mechanical stress, settling behaviour or unsuitable connection | Investigate loading history, assembly specifications and seal condition |
| Fitting continues to show leakage after retightening | Damaged sealing surface, incorrect assembly or overloaded sealing element | Professionally assess the connection after safe depressurisation |
| Quantitative leak rate differs significantly from the ppm sniffer reading | Different measured quantities and testing methods | Evaluate leak rate and local gas concentration separately |
| Local leak test passes after repair, but overall pressure decay test still indicates a problem | Another leak, leaking test valve or other pressure-related influences | Examine the complete test volume and associated test connections |
The table contains possible causes and suitable diagnostic approaches. It does not replace complete diagnosis of the actual test object.
It is particularly important to recognise that a small pressure change does not automatically demonstrate a mechanical defect.
Likewise, a locally elevated H₂ signal must not be interpreted directly as a quantitative leak rate without suitable calibration.
For technical evaluation, the results must be related to the specified testing procedure, permissible leak rate and actual operating conditions.
22. Suitable Hydrogen Measurement and Leak Detection Technology from ICS Schneider
22.1 testo 316-2: Gas Leak Detector with Integrated Pump
The testo 316-2 is a portable gas leak detector designed to locate leaks in suitable gas lines, fittings and valves.
It features an integrated pump, a flexible probe and a bar graph display to assist with local leak detection.
The device can detect methane, propane and hydrogen. For H₂, the ICS product page specifies a measuring range of 10 ppm to 4.0 vol.%.
According to the documented instrument specifications, the response time is less than 2 seconds (t₉₀).
The testo 316-2 is therefore particularly useful when a locally increased hydrogen concentration needs to be detected at an accessible fitting.
When selecting the instrument, it must be recognised that the testo 316-2 provides local concentration measurement or leak localisation functionality and does not automatically replace quantitative leak rate measurement in mbar·l/s.
Any approvals required for use in potentially explosive atmospheres must be checked separately. General hydrogen detection capability does not automatically mean that the device is approved for every hazardous area zone.
22.2 testo 316-EX: Gas Leak Detector for Suitable Hazardous Areas
The testo 316-EX is also designed to locate leaks in gas lines and technical installations.
The ICS product description specifies the detection of methane, propane and hydrogen, as well as a version with ATEX approval.
The device features a flexible probe and displays gas values in the ppm or volume percentage range.
For the documented configuration, ICS specifies a response time of approximately 14 seconds (t₉₀).
Its response characteristics therefore differ from those of the testo 316-2. The probe movement procedure must be appropriate for the particular instrument’s response time.
When working in potentially explosive atmospheres, the exact Ex marking of the instrument must be checked against the requirements of the intended location. The earlier ATEX directive reference on the product page does not replace verification of valid markings and approval for use.
The testo 316-EX is also intended for local gas detection and must not be regarded as a complete quantitative leak rate measurement system without further verification.
22.3 Druck UNIK 5000H: Precision Pressure Measurement in Hydrogen Applications
The Druck UNIK 5000H is a pressure transducer specifically optimised for hydrogen applications.
It features suitable wetted materials and is available with different pressure ranges and electrical output signals.
The available measuring ranges extend from 700 mbar to 700 bar. Depending on the configuration, documented accuracy options reach ±0.04 % FS according to BSL.
The sensor can be used in suitable test benches, storage systems and other hydrogen-carrying installations.
The combination of hydrogen compatibility, a suitable measuring span and stable signal processing is particularly useful for pressure holding tests.
However, suitability for a particular leak rate must be evaluated separately. High accuracy in normal pressure measurement does not automatically mean sufficiently low measurement uncertainty for very small pressure changes over long holding periods.
The specific process connection, sealing principle and available approvals must also match the application.
22.4 HT-H2 Series: Hydrogen-Compatible Pressure Transducers for Different Pressure Ranges
The HT-H2 series includes pressure transducers for hydrogen applications such as storage, refuelling, production and fuel cell technology.
The measuring ranges documented by ICS include versions up to 1,000 bar.
Depending on the variant, different electrical signals and mechanical process connections are available. Specific approval options are also offered for corresponding configurations.
The sensors may be suitable for pressure monitoring in hydrogen testing systems when the measuring range, material configuration and electrical integration meet the requirements.
Here too, general process pressure monitoring must be distinguished from sensitive pressure decay measurement.
For quantitative leak rate verification, the pressure transducer, temperature measurement, test volume and evaluation procedure must be considered together.
22.5 MINIMESS 1215 DVGW: Test and Service Connections for Suitable Gas Pressure Systems
The Original MINIMESS 1215 gas filling valve with DVGW approval is designed as a system access point for suitable gas pressure installations.
Depending on the approved configuration, it allows suitable testing and monitoring instruments to be connected, gas to be sampled and other service functions to be performed.
The gas filling valve version documented by ICS is specified for operating pressures up to 250 bar. ICS also provides manufacturer documentation for applications involving gaseous hydrogen.
For hydrogen test benches, a suitable test connection can help provide a defined and reproducible pressure measurement point.
However, the specific order variant, approved gas service, sealing materials and permissible pressure and temperature conditions are decisive.
The general MINIMESS designation or DVGW approval alone must not be interpreted as blanket approval for every hydrogen application.
Connecting, disconnecting and depressurising testing components must also follow the safe procedure specified for the particular coupling and complete system.
Further suitable sensors and components are available from ICS Schneider in the Hydrogen Applications and Leak Detection / Leak Testing Instruments categories.
23. Conclusion: Verify Overall Leak Tightness and Local Leak Detection Separately
Reliable hydrogen leak detection at fittings requires a clear distinction between verifying overall leak tightness and locating individual leaks.
Pressure decay testing evaluates the behaviour of a defined test volume. It can provide indications of leakage but is sensitive to temperature changes, test volume and the characteristics of the pressure measurement equipment.
Sniffer testing allows escaping hydrogen to be localised. However, an increased H₂ concentration at the test probe is not automatically equivalent to a quantitatively determined leak rate.
Tracer gases such as hydrogen-nitrogen mixtures or helium can provide useful additional testing methods when the test medium, detector and detection sensitivity are appropriate for the task.
Gas-specific behaviour is particularly important. Passing a nitrogen test does not always guarantee the required leak tightness during subsequent hydrogen operation.
Materials, sealing principles, temperature and dynamic pressure loads must also be considered together. A fitting can only be evaluated reliably when its actual operating conditions are known.
For quantitative verification, the permissible leak rate, testing method and associated measurement uncertainty are ultimately decisive.
Define the test objective and acceptance limit → Check hydrogen compatibility of components → Define test volume and gas paths → Select a suitable test medium → Evaluate pressure and temperature profiles → Locate leaks using appropriate sniffer testing → Check the plausibility of findings → Perform necessary repairs → Verify overall leak tightness again → Document the results
The most important practical principle is therefore: A pressure decay test initially answers the question of how the entire tested assembly behaves. Sniffer testing helps locate a specific leak. Only a suitable testing concept matched to the application combines both results into a reliable leak tightness verification.
24. Frequently Asked Questions About Hydrogen Leak Detection at Fittings
24.1 Which Methods Are Suitable for Detecting Hydrogen Leaks at Fittings?
Depending on the test objective, suitable methods include pressure decay testing, hydrogen sniffer testing and specialised tracer gas methods. Pressure decay testing evaluates an enclosed test volume. Sniffer testing helps locate leaks. Very small quantitative leak rates may require more sensitive, appropriately calibrated tracer gas procedures.
24.2 Can I Check a Hydrogen Line for Leaks Using Only a Pressure Decay Test?
Under suitable conditions, pressure decay testing can verify compliance with a defined overall leak tightness requirement. Its suitability depends on test volume, test gas, temperature stability, measurement uncertainty and permissible leak rate. Additional methods may be necessary for particularly small leak rates or local fault identification.
24.3 Why Does Pressure Sometimes Decrease Even When There Is No Leak?
Cooling of the gas can produce a measurable pressure decrease in a constant test volume. Measurement drift, ambient pressure changes or insufficient thermal stabilisation can also influence the result. These causes must be investigated before the pressure decrease is attributed to leakage.
24.4 How Is Pressure Decay Temperature-Corrected During Hydrogen Leak Testing?
Under simplified conditions, pressure is evaluated using the relationship between absolute pressure and absolute temperature. For example, a measured final pressure can be referenced to the initial temperature. Absolute pressures and temperatures in kelvin must be used. More advanced models may be necessary at high pressures or with complex temperature distributions.
24.5 How Can a Leak Rate Be Calculated from Pressure Decay?
For a constant test volume under simplified isothermal conditions, the equivalent pressure-volume rate can be calculated from the test volume, pressure change and time interval. This quantity can be expressed in mbar·l/s, for example. Actual thermodynamic conditions and measurement uncertainty must be considered for reliable quantitative verification.
24.6 What Is the Difference Between ppm and mbar·l/s in Leak Detection?
ppm describes the gas concentration at the measurement location. mbar·l/s is a unit of pressure-volume leak rate. A local ppm reading cannot be converted directly into a leak rate without an appropriate calibration and measurement procedure.
24.7 Can I Locate Hydrogen Leaks with a Portable Sniffer Detector?
Yes, provided that the instrument is expressly suitable for detecting hydrogen and is approved for use under the intended environmental conditions. ICS offers the testo 316-2 and testo 316-EX, for example. Both are intended for local gas leak detection but differ in equipment features and operating conditions.
24.8 Is the testo 316-2 Suitable for Hydrogen?
Yes. The ICS product information lists hydrogen as a detectable gas. For H₂, a measuring range of 10 ppm to 4.0 vol.% is specified. The device features an integrated pump and flexible probe. Suitability for the specific leak detection task must be checked against the required detection sensitivity and operating environment.
24.9 What Advantage Does the testo 316-EX Offer?
The testo 316-EX is designed to detect hydrogen, methane and propane and is available in a documented Ex configuration. For use in potentially explosive atmospheres, the specific marking must be checked against the applicable hazardous area zone and operating conditions.
24.10 Can Nitrogen Testing Be Used to Verify Hydrogen Leak Tightness?
Nitrogen leak testing can form part of an appropriate test concept. However, it does not always provide sufficient evidence of leak tightness during subsequent hydrogen operation. Depending on the testing concept, EIGA recommends additional helium testing or controlled hydrogen testing after an initial inert gas test for relevant applications.
24.11 What Is Forming Gas in Hydrogen Leak Detection?
In leak testing, forming gas commonly refers to a defined mixture of 5 vol.% hydrogen and 95 vol.% nitrogen. The hydrogen component serves as a tracer gas and can be detected using a suitable H₂ detector. Safe application and interpretation of the results depend on the specific gas composition and approved testing procedure.
24.12 Why Is Helium Frequently Used for Very Small Leak Rates?
Helium can be detected with very high sensitivity using suitable tracer gas leak detectors. It is therefore appropriate for certain quantitative leak tests. However, the measured helium leak rate must be evaluated according to the defined test conditions and intended hydrogen application.
24.13 Can Permeation Through a Seal Appear Similar to a Leak?
Yes. Depending on the material and operating conditions, hydrogen can permeate through certain seals. A sensitive measuring method can detect gas escaping without an individual mechanical defect necessarily being present. Whether this is permissible depends on the specified leak tightness requirement.
24.14 Can I Retighten a Leaking Fitting While It Is Pressurised?
Retightening under pressure must not be treated as a general repair method. It can cause mechanical damage or hazardous gas releases. The procedure must be based exclusively on specific manufacturer approval and the authorised safety procedure. As a general rule, the connection must be safely depressurised before repair work.
24.15 How Do Temperature Changes and Pressure Cycling Affect Hydrogen Fittings?
Temperature changes can influence contact pressure and the behaviour of sealing elements. Repeated pressure cycling creates additional mechanical stresses. Consequently, a connection may exhibit different leakage behaviour under operating conditions compared with a single static test.
24.16 Which Pressure Transducer Is Suitable for Hydrogen Pressure Decay Testing?
The pressure transducer must be suitable for hydrogen, the intended pressure range and actual installation conditions. Detecting small pressure changes also requires consideration of resolution, drift, uncertainty and temperature evaluation. ICS offers the Druck UNIK 5000H and HT-H2 series, among others.
24.17 Which Standards and Technical Regulations Are Relevant to Hydrogen Leak Testing?
ISO 20485:2017 is relevant to tracer gas methods, among other standards. EIGA guideline DOC 15/21 addresses testing of gaseous hydrogen installations. ISO 19880-1:2020 may be relevant to hydrogen refuelling stations. The specific requirements applicable depend on the installation type, area of application and agreed testing procedure.
24.18 What Information Does ICS Schneider Need to Select Suitable Testing Equipment?
The required information includes the type of hydrogen installation, component to be tested, number and configuration of fittings, and intended test medium. Operating and test pressures, temperature range, test volume, permissible leak rate and required detection sensitivity are particularly important. Additional information is needed about existing pressure transducers, sealing materials, test connections, Ex requirements, available gases, test duration and documentation requirements. For quantitative leak testing, it should also be established whether acceptance testing will use pressure decay, hydrogen tracer gas or helium and which reference conditions apply to the permissible leak rate.