Note that there must be no leakage from piping joints. Even if you don't see a drop in pressure during the test, you must still inspect every joint (welded, threaded or otherwise) for leakage. A properly done hydrotest, one where the air has been scrupulously removed first (i.e. by applying a vacuum prior to filling) will be very sensitive, such that even a very small amount of leakage will result in a measurable drop in pressure at the test gauge- but a test where there's a substantial gas pad will not be anywhere nearly as sensitive- even a substantial leak may not show up as a pressure drop unless you wait a long time. A test where leakage through something which is not a piping joint, i.e. a valve packing, does not automatically constitute a failure of the test.
The hydrotest is called a "leakage test", but in reality it's a pressure integrity test, carried out at a multiple of the MAWP which itself is a multiple of the maximum operating pressure. A subsequent leakage test at maximum operating pressure using another medium (i.e. a gas for lines which will be handling gases) may be required by the end user depending on the service, after all assembly is complete, but it is not a requirement of B31.3.
Assuming piping is not underground, 'watching the gauge' is for pathetic inspectors/engineers too lazy to actually perform their job and check the welds for leaks, weeps, and seeps.
Every inch of every weld is to be inspected for leaks. The hydro pump can run continuously, to maintain the test pressure. Valve packing leaks are acceptable, and to be expected -- packing is 'set' for normal operating pressure, not hydro pressure. So packing will leak or weep. Acceptable. Flanges with temporary gaskets may leak. Acceptable, but to be noted for rework prior to commissioning the system. Valves themselves may leak past the seat; acceptable, but if a leak-tight valve is needed, note this one for repair or replacement.
Starting to see why it is mandatory to get up off your butt and feel each weld joint for leakage? A successful hydro has nothing to do with 'watching the gauge' or disconnecting the pump and keeping pressure. Real world, there are many sources of acceptable leakage. Point of fact, I was part of the crew that hydro'd the turbine building Main Steam piping at a nuke, and due to 4 angled Stop Valves, we had 300 gpm of leakage past those valves. But that was our only leaks, and the test was acceptable. At 300 gallons per minute.
Duwe6: a star to you for injecting some commonsense into this. I've had more than one colleague let alone client who confused a hydrotest with a leakage test, and hadn't actually read and understood the testing requirements in B31.3.
It is one key reason why B 31.3 is not suited to underground piping / pipelines as it is clearly designed in instances like this for above ground pipes / spools where it is feasible and practical to examine each weld and joint.
The plus side is that you can do a hydrotest /leak test quite quickly as you don't need to allow time for a small leak to show up as a pressure drop....
Totally agree with duwe6 - great post
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
