My personal practice for an air test is to be at least one zip-code away from the site.
LMAO @ joe
I don't know of any specific distance requirements, though I'd be interested in seeing that NASA document eliebel mentions. Regardless, there will always have to be someone within that area to do the test.
Note that the piping code requires a leak test, not a proof test. If you're looking to do a proof test (ie: 1.5 times MAWP), then what is typically done is to disassemble the piping system and do a hydrostatic pressure test, and then reinstall the system and do the leak test. The proof test isn't required by code.
ASME B31.3 requires a "leak test" at 1.1 times design pressure. They also require a relief valve set at 10% above that pressure. We always put this relief device at the source of the test pressure, and ensure by calculation that the relief device is large enough to handle the flow from the test source.
When pressurizing the piping system, you should also verify you are only pressurizing those portions of the piping system that are rated for the pressure. Creating a test plan is imperative. Color code a P&ID or something along those lines. Make sure you know what lines you're pressurizing, and make sure the test plan incorporates valve open/close requirements, blanked off piping sections, etc...
I'd also suggest bringing up pressure slowly - in 10% increments. Hold for 5 minutes at each level, and do a soap bubble leak test at each level.
Leak testing requires someone to do the test, so they will always be near the piping during the test. One primary reason for keeping others back isn't just to keep them safe from potentially dangerous pressure, it is to maintain some control over piping and valve configuration. You need to keep people away from valves or other controls that need to be kept in a specific position during test, especially because you're generally in a construction mode and people may be less aware of lockout/tagout procedures.
)
