Check Valves in Relief Flow Path

[don1980]

What is your opinion on allowing a check valve between the protected vessel and the PSV, or between the PSV and the vessel to which the fluid is routed? I didn't see this addressed in API-521 or any other code or standard. If you're opposed to allowing check valves in the flow path, is this a blanket opinion or one that might vary depending on the corrosiveness or fouling tendency of the fluid?

 

[lizking]

I don't think check valves are "outlawed" by any code or standard. But, if used, the consequences of their failure should be considered as it relates to compromising the relief system capability. Also, if used, they should be on a regular inspection schedule: at least as often as the relief valve itself.

 

[JLSeagull]

Any obstruction such as a check valve should be discussed during the process hazard analysis. I would object strongly but it might take a while to prepare a code basis.

 

[dcasto]

Even a simple flpper check can fail and the flapper can lodge in a 90 and cause a huge drop. Also the check valve will have an inheient pressure drop that would have to be taken into account as part of the PSV set pressure.


Overall I can't see why the check valve could not be after the relief valve in the line unless this is a retrofit or the original design overlooked the need for a relief valve. If either of two is correct, then pay the money to correct the problem.

 

[Ashereng]

My opinion is that check valves have no place in a relief system. I have not come across an application where I was convinced that a check valve is needed.

A check valve between the tank and PSV dangerous if the check valve fails close for any reason - you now no longer have a PSV.

A check valve between the PSV and vessel to which you are relieving to is also dangerous, if the check valve fails close for any reason - you now no longer have a PSV.

In any case, I do not rely on a check valve to isolate product - I always use the basis that a check valve leads. If I need positive isolation, I used something else.

If the process fluid is corrosive, you can use a rupture disc, or match the PSV material to suit. There are usually alternatives.

As a general blanket statement, I don't put check valves into my relief system.

"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?

 

[tkdhwjd]

If there are no specific guideline regarding a use of check valve on the relief path, you should thoroughly review the need with hazop or equivalent review. Risk is potential loss of containment in the event the check valve fails for some reason (and I have seen many failed check valves). Although highly unlikely, what if the check valve is installed backward. Backward installation is possible, but it gets normally caught during prestart up safety review/walkthrough. But then againm you may have a legitimate reason for requiring a check valve. Are you allowed to share why one would be required?

 

[MortenA]

What should the purpose of the check valve be? API-521 is for pressure vessel (or the ASME code is). And the PSV wount open when there is a negative pressure - even if this should happen (i dont know your system).

Dont just throw check valves in on all line "just in case". They are not that reliable and they will require inspection and maintenance.

Best regards

Morten

 

[don1980]

OK, let me explain this a little more clearly. In a chemical plant you often have one PSV that protects multiple pressure vessels. For example, a distillation column typically has a single PSV that protects the column, condenser, reflux drum,and the process side of the reboiler. Rarely is there a check valve in a distillation system but in other systems they're common. Here's a specific example: a large centrifugal compressor with interstage cooling and a knock-out drum at each interstage. A check valve is installed in the vapor line from each drum, and fire is the only scenario. If it were not for the check valves we could install a single PSV on the final drum. There are many other examples. Take for example a case where thermal expansion is the only scenario. Suppose you have three vessels in series but with a check valves between each one. A single thermal expansion PSV could be installed on the downstream drum to protect all three. If the process stream had any tendency to foul then I wouldn't consider such an option, but for a fluid that presented little risk of fouling it might make sense to just install a single PSV.

 

[dcasto]

I'd install check valves on each stage or where there is a check valve. 4 little relief valves are not that much more than 1 giant economy size, especially the risk of a check valve failing.

 

[JAlton]

There was an attempt some years back to permit check valves underneath (up stream) of Pressure Relief Valves. However, ASME Sec. VIII did not accept the proposed Code Case. The concern was the same as expressed by Ashereng above.

I voted against the item myself, because I agree with Ashereng as well.

JAC