Pressure Relief Valve Change-out Frequency

[gel2]

With respect to pressure piping (and not pressure vessels), what is the required replacement frequency of PRV's. I have heard 3 years but, can not find it in a code (i.e. B31.1/3/or 9). Thanks.

 

[metengr]

For in-service pressure retaining items you need to default to either API or the NBIC (NB-23), which are in-service inspection/repair codes not construction codes. Since I am most familiar with the NBIC, there is particular reference in Part RB for testing (not replacing) PRV's and intervals. Review this information in the 2004 Edition/2005 addendum of the NBIC in Part RB-8410.

 

[dtn6770]

No Code requirement for replacing. A fully fledged PSM program should have PRV inspection and testing requirements well established in its Mechanical Integrity Program!

Many follow API's lead on the subject...

API 510 (8th ed., 4th ad., 8/03), 6.6 - "Pressure relief valves shall be tested at intervals that are frequent enough to verify that the valve perform reliably." "Pressure-reliving devices should be tested and maintained in accordance with API Recommended Practice 576." "The intervals between pressure-relieving-device testing or inspection should be determined by the performance of the devices in the particular service concerned." "Test or inspection intervals on pressure-relieving devices in typical process services should not exceed 5-years, unless documented experience or a risk assessment indicates that a longer interval is acceptable." "For clean (non-fouling), noncorrosive services, maximum intervals may be increased to 10 years." And so on...

Utilities such as air, water, and nitrogen (friendly and non-threatening services) can usually have infrequent inspections/tests.

"Process" services are 5-year intervals, max., 2-3 years usually.

 

[dcasto]

We set up a "gradeing" program for inspect periods on PSV's (also piping and vessels). We used the 2 dimentional array used in most hazard assesment programs. On one axis is Probability of a failure, the other consequences. We evaluated the things that would increase the probability for a piece of equipmet to fail, things like corrosion, vibration, high temperature, frequency of actual use, material selection could even decrease this measure, ect... Then the consequences were looked at, things like chemical properties, high pressure (dispersion), location, line and vessel sizes, ect.
Then we looked up the values on each axis to determain frequency. We did a "common sence engineering" check and saw that our clean, dry low pressure natural gas lines had a 5 year interval and our rich amine psv's and lines had a 1 year inspection level.

 

[gel2]

Thanks. Very helpful!

 

[MJCronin]

See this thread..

http://www.eng-tips.com/viewthread.cfm?qid=15220

-MJC