Failure on the bellows of PSV's 1

[leont]

THE POINT TO DISCUSS IS, FAILURE OF THE BELLOWS ON SAFETY RELIEVE VALVES
I have not a huge experience with this kind of failures on safety valves.
This has been a recurrent issue and never a failure analysis has been completed. In the past the fuequency of failure was between 4 and 6 months. Original investigation showed stress corrosion cracking initiated from the outside of the bellows. The bellows were upgraded to Inconel 625 however the failure shows up over and over. Last failure happend two days after new PSV installation.
How big could be the back pressure on the bellows to cause the collapse of the bellows?
Do we really need a big back pressure or just small one will be enough for a big damage?

I will be talking to vendor shortly. However, how can I calculate what this back pressure value is?

Thank you in advance for your support

leont

 

[Latexman]

What is the process fluid?

The bellows must be strong enough to withstand the back pressure without collapsing. It must be flexible enough not to affect the valve performance. The bellows must also be designed to resist flutter during a relieving cycle caused by an inappropriate blowdown setting, or be shielded from it. Such instability can cause premature failure due to metal fatigue. PSV chatter will utterly destroy the bellows (and probably the PSV). The bellows must also be corrosion resistant. Because of the thin metal used in a bellows, pin holes due to corrosion can easily occur in a material considered suitable for the same service in a thicker gauge. The back pressure limits listed for bellows in PSV manufacturers' catalogs do not contain much safety factor!

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[LittleInch]

How big could the back pressure be to cause failure - something bigger than what you put on the data sheet...

As latexman says, the real killer is fluctuations/fatigue, either of the relief valve itself or the back pressure. Try sticking a PT on the discharge line and then monitoring it for presusre pulses or multiple change sin pressure over a day / week. I bet no one looks at or records the pressure fluctiations in that line, but you need to to eliminate or start considering other types of relief valve.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Ripz]

If your Inconel is Failing, it's likely not SCC. Most likely it's Chattering.

 

[don1980]

I agree with the previous responders. I'm guessing that your operating pressure is too close to the set pressure, and the valve stem is moving (maybe just simmering rather than popping open), causing enough movement to fatigue the bellows. If so, consider using a POSV rather than a balanced bellows.