Helium tracer-gas testing is widely used to locate and quantify leaks in vacuum systems. A mass-spectrometer leak detector responds selectively to helium and reports a gas flow or throughput, provided the detector is calibrated and the test method is defined. ISO 20485 describes tracer-gas techniques and ISO 20486 addresses reference-leak calibration. This article explains vacuum, sniffer and integral methods, the meaning of leak-rate units, virtual leaks and practical test planning.
Leak-tightness is a system property
A chamber reaches equilibrium when effective pumping speed balances the total gas load from outgassing, permeation and real leaks. Excess leakage limits the achievable pressure and admits air and moisture that can condense on cold surfaces and interfere with contamination-sensitive tests. Any stated working pressure therefore assumes a compatible total gas load and verified integrity of the vessel, seals, feedthroughs and connected circuits.
There is also a safety dimension: LN₂ and GN₂ circuits are pressure-bearing systems inside or around the chamber. A leaking cryogenic line releases nitrogen into the chamber or the room – enrichment in an enclosed space is an asphyxiation hazard, which is why cryogenic circuits are leak-tested in their own right.
How helium mass spectrometer leak detection works
Helium is suitable as a tracer because it is inert, does not readily condense at normal leak-test conditions and has a low background concentration in air. A mass-spectrometer leak detector ionises the incoming gas, separates mass 4 and compares the signal with a calibrated reference. The usable detection limit and response time depend on the detector, test volume, pumping path, background and selected method.
- Calibrate the detector with a reference leak whose value and uncertainty are traceable for the intended range.
- Allow the measured volume and pumping path to determine response and clearing time before moving to the next joint.
- Record background, test pressure, tracer application and detector settings together with the reported leak rate.
- Use a method-specific acceptance value derived from the total allowable gas load and system safety requirements.
Vacuum (spray) method vs sniffer method
In the vacuum method, the chamber is evacuated with the detector connected to the pumping line, and helium is sprayed from outside onto welds, flanges and feedthroughs one by one. It is the most sensitive approach and localizes leaks precisely; enclosing the whole chamber or single joints in a helium-filled envelope turns it into an integral test with a single overall number.
The sniffer method reverses the direction: a positively pressurised circuit is filled with tracer gas and sampled externally. It is appropriate for operating circuits such as nitrogen or cooling-water lines, but its achievable sensitivity depends strongly on ambient helium background, probe handling and gas distribution. It should not be assigned one universal detection limit.
- Vacuum method: high sensitivity and localised testing of the chamber shell, flanges and feedthroughs.
- Sniffer method: testing of positively pressurised circuits under controlled tracer concentration and ventilation.
- Integral enclosure method: one overall result for a component or chamber, with less localisation information.
- Use helium sparingly and control room background so previous applications do not mask later measurements.
What leak rates in mbar·l/s actually mean
A leak rate in mbar·l/s is a pressure-volume throughput referenced to stated conditions. The numerical value must be interpreted together with gas species, temperature, test pressure and calibration convention. Conversion to molar flow can avoid ambiguity when results from different laboratories or unit systems are compared.
Acceptance is system-specific. The allowable integral leak rate follows from the target pressure, installed effective pumping speed, contamination budget, hold-time requirement and any separate limits for pressurised or cryogenic circuits. A generic joint-level number should not replace that gas-load calculation.
Virtual leaks vs real leaks
Not every leak-like symptom is a path to atmosphere. Virtual leaks are trapped volumes inside the vacuum – blind tapped holes under screws, double weld seams, unvented O-ring grooves – that release gas slowly and mimic a real leak in the pressure curve. A helium spray test shows nothing, because there is no path from outside.
Telling them apart
- Rate-of-rise signature: real leaks produce a linear pressure rise, trapped volumes and outgassing flatten over time.
- Residual gas analysis: an air-like nitrogen–oxygen signature points to a real leak, dominant water vapour to outgassing.
- Helium response: a real leak answers to spray within the response time; a virtual leak stays silent.
- Prevention is design: vented screws, relief bores, continuous inside welds and vented seal grooves avoid trapped volumes.
When to test – and designing for testability
Leak testing is not a one-off acceptance step but a lifecycle instrument. It belongs after manufacturing and first assembly, after every intervention that opens seals or exchanges feedthroughs, at periodic maintenance intervals, and after transport or relocation, where settling and handling loads work on flanges. A recorded baseline for each joint makes later comparisons meaningful.
Testability is a design property: accessible flange positions, a dedicated test port for the detector, helium access to double-seal interspaces and enough clearance to bag individual joints all decide whether a leak check takes an hour or a day. Deepvac offers helium leak testing as a factory option and as part of maintenance and repair service.
Takeaway
Leak testing is most useful when acceptance limits, calibration, response time and test method are defined for the specific system. Helium testing finds real paths to atmosphere; rate-of-rise analysis and residual-gas data help separate those paths from outgassing and trapped volumes.
Frequently Asked Questions
There is no universal acceptable value. The limit should be derived from target pressure, effective pumping speed, permitted gas load, hold time, cleanliness requirements and the applicable test method. Pressurised nitrogen, water or cryogenic circuits need separate safety-related acceptance criteria.
A mass spectrometer tuned to helium (mass 4) is connected to the evacuated chamber, and helium is sprayed onto welds, flanges and feedthroughs from outside. Wherever a real leak path exists, helium enters and is reported quantitatively as a leak rate in mbar·l/s after the method-specific response time. For pressurized circuits the direction reverses: the circuit is filled with helium and a sniffer probe samples the outside.
A virtual leak is a trapped volume inside the vacuum – a blind tapped hole, a double weld seam, an unvented seal groove – that releases gas slowly and imitates a real leak in the pressure curve. Helium spray shows nothing because no path to atmosphere exists. Rate-of-rise behaviour and residual gas analysis distinguish the two; vented screws and relief bores prevent the problem by design.
At defined events rather than a fixed calendar alone: after initial assembly, after any seal change or feedthrough work, after transport or relocation, and as part of periodic maintenance – often annually for chambers in regular campaign use. Keeping a documented baseline per joint turns each later check into a quick comparison instead of a full investigation.
Technical references
Primary standards, agency material and current Deepvac product pages used to verify the technical statements in this article.
- ISO 20485:2017: Non-destructive testing, Leak testing, Tracer gas methodInternational Organization for Standardization (ISO)
- ISO 20486:2017: Calibration of reference leaks for gasesInternational Organization for Standardization (ISO)
- Leak Artifacts and Vacuum Leak Primary StandardNational Institute of Standards and Technology (NIST)
- CERN Accelerator School: Vacuum TechnologyCERN