Leak-Down vs Pressure-Hold Test: Which to Use
Why this matters
Both tests look for leaks, but they answer different questions. A pressure-hold test asks "does the whole boundary stay tight." A leak-down test asks "how fast does this volume bleed off, and through which path." Picking the wrong one wastes time or hides the fault. Knowing which to reach for is the difference between a clean diagnosis and a second trip.
What each test does
Pressure-hold (static hold). Bring the isolated system to a target pressure, close the source, and watch whether the pressure stays put over a set time. It is a pass/fail on the entire sealed boundary at once. Simple, requires only a gauge and an isolation point.
Leak-down. Apply a controlled, continuous supply at a known pressure and measure the percentage of pressure lost across a restriction or the rate at which the volume bleeds down. Instead of a yes/no, you get a number: how much is escaping, and often a clue about where, because you can listen at suspected exit points while pressure is held against them.
The key difference: pressure-hold is a closed snapshot, leak-down keeps feeding the system so you can localize while it is under steady pressure.
Side-by-side
| Factor | Pressure-Hold | Leak-Down |
|---|---|---|
| Question answered | Is the boundary tight, yes or no | How fast is it leaking, and roughly where |
| Source during test | Closed off after reaching target | Kept connected, supplying continuously |
| Output | Pressure drop over time | Percent loss or bleed rate |
| Localizing the leak | Separate step after a fail | Often during the test, by listening |
| Setup complexity | Low (gauge + isolation) | Higher (regulated supply, often two gauges) |
| Best for | Confirming a section holds | Quantifying severity and finding the path |
When to pick which
Use a pressure-hold when:
- You need a quick pass/fail before returning a system to service.
- You are verifying a repair you just made (did the new seal hold).
- You only have a gauge and an isolation valve.
- The system is small enough that a leak shows up clearly in the hold window.
Use a leak-down when:
- The hold already failed and you need to know how bad it is.
- You need to compare against a spec stated as a percentage of loss.
- You want to localize the leak while it is under pressure (listen, feel, watch bubbles at each exit).
- A slow leak is borderline and you need a rate to judge severity instead of a vague drift.
Reading the results
Pressure-hold: any clear, steady drop beyond the spec is a leak. A drop that stops after a few seconds is usually temperature or trapped air settling, not a fault. Always re-run a borderline hold to separate settling from a real leak.
Leak-down: the percentage or rate tells severity. A small, single-digit loss is often acceptable in many systems; a large loss means a significant path. Where you hear or feel the escape tells you the location. Hold pressure against each candidate (a valve, a seal, a joint) and check which one passes the medium.
Don't confuse either with a load test
Both of these are static-boundary tests. Neither tells you the system performs correctly while running under real load. A system can pass a pressure-hold and still fail when it is working hard, because heat, vibration, and flow open paths that a still test never sees. After a clean pressure test, a load test is often still warranted.
Common pitfalls
- Calling a leak-down "failed" when the supply regulator itself is undersized and cannot keep up.
- Running a pressure-hold too short to catch a slow leak.
- Forgetting that temperature change during either test fakes both a leak and a fix.
- Localizing during pressure-hold without re-supplying, so the pressure is already gone by the time you reach the suspect joint.
References
- Manufacturer documentation for acceptable leak-down percentages and hold-test windows.
- Trade-standard practice for static and dynamic leak testing.
- OSHA general guidance for working with pressurized systems.
- See related: The Pressure Test Method (Generic); Load Test vs No-Load Test.