The most frequent cause: temperature
In a closed vessel containing gas, the pressure rises and falls with temperature. When pressurizing with a gas, compression heats it; as it cools toward ambient temperature, the pressure drops even if there is no leak. That drop can be mistaken for a real loss or, conversely, a small leak can be masked by warming of the surroundings.
Other sources of inconsistency
- Insufficient stabilization time: the system is still settling thermally and mechanically when the measurement begins.
- System volume: with a large volume, a small leak produces a small pressure change; with a small volume, the same orifice is more noticeable.
- Instrument: limited resolution, transmitter drift or a gauge that cannot distinguish the change being sought.
- Leaks in the instrumentation itself: connections, hoses, test valves and plug seals.
- Seating of seals and gaskets: they settle over time, and the leak changes between repetitions.
- Test fluid: a nearly incompressible liquid behaves differently from a gas, and air trapped in a hydrostatic test mimics a leak.
What it looks like in practice
Two repetitions of the same procedure give different drops; the test passes in the morning and fails in the afternoon; the result changes with shade or sun on the equipment. If the drop decreases over time until it stabilizes, it is usually thermal; if it remains constant, a real leak is more likely.
How to reduce the variability
- Allow the temperature to stabilize and record it together with the pressure.
- Use a duration and an acceptance criterion defined in advance.
- First verify that the test setup itself does not leak, by isolating it with a blind plug.
- Use a method suited to the size of the leak: pressure decay, bubble testing, tracer gas detection or flow.
- Correct or compensate for temperature whenever possible.
For equipment that contains pressure, the test must be performed by qualified personnel following the applicable procedures and codes.