Dry gas instead of liquid for leak test in pump w/ mech seal.

[moon161]

Does anyone here have experience with using dry gas / cavity decay test instead of presurized water to leak test pumps with rotating mechanical seals?

Thanks,

 

[BEMPE16524]

This method, also called "pressure decay testing" or "cavity decay testing," works by pressurizing a sealed cavity with a dry gas, typically nitrogen or air, and monitoring for pressure drops that would indicate a leak. References:

1. API 682: Pumps - Shaft Sealing Systems for Centrifugal and Rotary Pumps
2. API 610: Centrifugal Pumps for Petroleum, Petrochemical, and Natural Gas Industries
3. ASME B73.1: Specification for Horizontal End Suction Centrifugal Pumps for Chemical Process
4. ASME BPVC Section V: Nondestructive Examination (NDE)
5. ANSI/HI 1.3: Rotodynamic (Centrifugal) Pumps for Design and Application
6. ANSI/HI 9.6.6: Pump Piping for Rotodynamic Pumps
7. ASME PCC-2: Repair of Pressure Equipment and Piping

 

[moon161]

Thanks! Any idea what the gas leakage through a static seal is compared to the liquid leakage?

 

[BEMPE16524]

the pump cavity is pressurized to a specific level, and the rate of pressure decay is measured over time. A stable pressure indicates a good seal, while a decay in pressure suggests a potential leak.

In API Standard 682, Section 10.3.3 describes the "Air Test" or "Integrity Test" for leak testing mechanical seals. This test, performed after final assembly, uses pressurized air to ensure the assembly is free from leaks. It specifically tests for proper assembly, ensuring no assembly defects like damaged O-rings or improperly seated components, which could lead to operational leakage. Typically, this is done with 50 psig air for a short period to verify that the pressure holds without significant loss. Although not expected to be completely leak-free, any detected leakage should meet the acceptance criteria defined in the standard.

 

[A8yssUK]

Thanks! Any idea what the gas leakage through a static seal is compared to the liquid leakage?

If your doing it purely via pressure drop of the system.. gases are harder to seal than liquids so your drop in pressure will be quicker with a gas vs a liquid. However there isn't any rule of thumb for how much quicker as it depends on a lot of things.

I should also add here, liquid pressure testing is much safer than testing with gases. Compressed gasses when uncompressed take up a lot of volume while liquids generally do not compress much at all so are less energetic when released. Especially important when testing larger volumes.