Pop test of relief valves 1

[mthakur]

Hi all,
We have problems with the performance of relief valves on site, the valves when taken to test bench in " as received" conition for pop test , they generally do not lift on the cold diff set pressure (within the tolerance) which then reduces the next run campaign of the relief valves. AS per contractor spec for pop test of valves , the max pressure applied to the valves are 10% in excess of the CDSP or cold diff set pressure, if the valve doesnt found to be lifted, no pressure is applied considering it unsafe ( as you are going over the design pressure) this is considered as valve sitting on "high" . API 576 does allow to increase the pop pressure to 1.5 times if the valve doesnt not lift at first instance. This is done to ensure that valve seat gets released of found stuck.

My question is around inceasing the test pressure to 1.5 times the CDSP, how do you justify increasing the pressure to 1.5 times the design pressure (or CDSP) Secondly I would like to have the comments from other uers if they have had similar experience with pop test.

 

[Duwe6]

You are confusing Vessel *design* pressure with PRV *set* pressure. A PRV set at 250 psig could be within its working limits at 900 psig. A knowledgable PRV shop will know what the limits are. If the guy running your test bench does not know; then, and only then, do you have a problem.

 

[mthakur]

Thanks Duwe6, Thanks for the answer, In most cases ( or in typical Ammonia plant environment) the design pressure of the system is the set pressure of pressure relief valve. If we go as per API576 and pop test the PSV at 1.5 times the CDSP and consider the PSV acceptable, this also implies that we are accepting the fact that system will actually get pressurised to 1.5 times the design pressure before the relief valve opens up. it may not be appropriate for the system. Your thoughts on this

 

[Duwe6]

Your PRV *have* to relieve at the MAWP / Design Pressure of the system. Unless the system is very hot, the Cold Differential Set/Test Pressure should equal the 'Pop' pressure. In other words, no compensation for the PRV spring getting hot and a little 'limp' and relieving at a lower 'Hot' pressure. Per CROSBY, the differential test pressure for 600-800°F is only 2%, so a CDTP for a 1000# PRV would be 1020#.

To have to go to 1.5x set pressure to get your PRV to 'Pop' means that the PRV is waaay out of calibration, jammed with deposits, or the spring is broken. And I agree that this is not acceptable. Your PRV's are being used too long between Test & Reset's. You might try a 2-port Full Port ball valve under 2 PRV's. That makes one 'live' and one valved-out at all times. Allows you to remove one PRV and clean & reset it mid-campaign. We do this with our ammonia-tank PRVs. Each PRV goes to Test & Reset every other year [send one PRV out each year], but the tanks stay in service for a decade.

 

[mthakur]

Agreed, 1.5 times (max) the set pressure to achive pop test as recommended by API RP 576 us way to high and imagine the case with high pressure valve. I still wonder how this value has been given in API.

At the same time it is also true that 1.1 times the set pressure is also too low, normally we have plant run campaign of three years and so are the freqency of re-test with relief valves. In relief valves there are close tolerances within sleeve and stem which can get jammed due to rust, deposition of foreign particles, rust,, dust etc (more in cae of valves with no bellows), seat may get stuck with nozzle, only 10% overpressure may not be sufficient to lift the valve. The outcome of this exercise will not provide enopugh information for assessing the PSV condition required to determine its overhaul and re-test frequency.

 

[don1980]

mthakur - one possible reason for a PRV not opening within the allowable tolerance is sticking. The point made by API 576 in saying to go as high as 1.5 x set pressure is for you to gauge the severity of the sticking. Understand that this is a data gathering step that is performed during a bench test. No one is saying that it's acceptable for an installed valve to not open until 1.5 x set-P. API is just saying that if the valve doesn't open by 1.5, then there's little value in continuing this bench test at higher pressure - the valve is severely stuck and needs a major overhaul or possibly replacement.

This is a perfectly reasonable bench testing procedure.

 

[Duwe6]

Yup. don1980 is exactly right. Your valves are delivering test data that is telling you that your interval is waaay too long. And have the bench guy stop at 1.5 * setpressure. That PRV is 'Stuck', as opposed to 'Fail-High'.

That PRV needs more TLC - tender, loving care, and a shorter run interval.

 

[Thevalveman]

If the valve does not open within +/- 3% of set pressure then it is considered to have failed the Pre-test! ( out of set pressure tolerance as per ASME)

Now, some customers that we deal with use 10% overpressure! this is their own guide lines set by their inspectors which take into account over pressure for the piping and vessels.

API 576, is a guide line or recommended practice, at 1.5 times the set pressure this is to take into count the inlet flange rating of the valve as to not exceed the allowable working pressure of the PRD eg: 600# inlet flange set at 1440 psi at 1.5 = 2160 psi, the maximum test or hydro test pressure for the valves inlet, or primary pressure zone for the valve!

In the event that you want to find out what the valve is really doing, you can us this for investigation purposes, like, to find out if the valve has been set to incorrectly from a service company or if it is indeed a process problem!