You've all been very coincidents with your answers but...

[edanyel]

I'm sorry for insist in the matter...
Now please, think about this... overpressure event, PSV lifts...there is some chattering or some hammering and after a while...pressure drops and PSV closes without leaks...

1)If you don't inspect the PSV, let´s say by a Prepopping test, how do you detect the possible damage in the moving parts of PSV internals?

2)In other words how could you assure that the PSV internal conditions haven´t changed if you don't perform a pop test?

3) So I ask again; as far as you know isn't there any standard or code that address this matter??

I'm sorry for bothering with my doubts, thanks again.

 

[JimCasey]

If it closes and does not leak, that is a good indication that it is not damaged.

The annual inspection should be adequate to monitor the condition of the valve otherwise.

 

[JLSeagull]

A certified company like Furmanite tests the pressure safety relief valves every year on a test stand. This can be done at site.

 

[Ashereng]

As far as I am aware, PSV inspection is a site dependent procedure. Some sites require that you inspect right away. Some sites do not.

On those sites that do not require immediate inspection before starting up again, you are effectively starting back up without knowledge of whether your PSV has any damage.

In some case, even if the PSV is damaged, the consequence is minimal. For example, a PSV on the discharge of the pump line back to the tank. If the PSV is leaking, it is only going back to the tank.

I guess more information on why you are looking for a "code" reference may help. Are you arguing in favor of inspecting immediately and as a condition for starting up in all cases? Or are you arguing in favor of a case by case approach?

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[hydtools]

Maybe this book has the information you want.

http://www.rigzone.com/store/product.asp?p_id=1529

Ted

 

[gerhardl]

... and again: what is your real problem?

Is your real problem that a certain PSV (or several PSVs?) opens and release with the manner you describe, and stands closed only for shorter process runs between each release?

Eg, opens once each week, month or so, and each time this interfers with your regular process, creats incertainity or cost mony?

If this is the case you should look into the process properly and evalute some different aspects than only a regular check of the setpoint and thightness of the PSV.

You either have to raise the setpoint (if this can be allowed) or get hold of a PSV which allows less distance between actual process pressure and relief pressure (Pilot operated ? Costly but accurate!)

...or you have to look at the process in general to obtain better process control.

 

[dcasto]

ASME VIII states every 5 years as a maximum for inspecting relief valves.

 

[zerosum]

When a PSV demand occurs, good practice is to perform and record an analysis of the cause, and takes steps where possible to eliminate future similiar events. That analysis should consider whether the demand scenario was sufficient to damage or degrade the device and then a decision made whether to shut down the process for relieving device inspection (assuming non-redundant PSV's).

All these actions should be suitably documented.

 

[dcasto]

edanyel,

How do you know the the relief valve is damaged the day after it was installed? The whole process of testing and re-testing is risk management. If you believe their is a risk of failure after a PSV opens then retest the valve. If you keep track for about a 5 years of the test results, you'll document that either A, you assumption was right, or B your assumption was wrong, or C you'll do a design of experiments and find resonable point at which to test, like say after every 10th time the valve opened we found enough damage to warrent a test or rebuild.

Zerosum is correct too, find a root cause and stop the need for testing except on a 5 year schedule per ASME.