Fire Case PSV Chatter Risk 1

[ChemEngSquirrel]

I have a new vessel with a range of relief cases. I'm proposing to install a single conventional PSV.

The governing case is gas blow by. For the fire case, the relief load is less than 25% PSV capacity so i believe chatter may be a risk for the fire case (as per guidance in API 521). However, this seems to be a common problem for fire relief.

Do i need to alter my PSV design just to avoid chatter in the fire relief case?

Thanks.

 

[MortenA]

As far as i know chatter is primarily about avoiding damage to the seat? If I'm rift i don't that will be a major concern after a fire that lead to a PSV release.

Best regards, Morten

 

[ChemEngSquirrel]

From what Morten says, you don't need to worry about chatter and the 3% rule for PSV inlet lines if fire is the only sizing case. It would help if anyone can provide a reference to support this?

Thanks.

 

[Duwe6]

Chatter isn't a Code condition, its operational. Per Code the PRV is there to keep the vessel from going BOOM. Chatter is something that needs to be engineered around; oversize disks like to chatter - use the smallest disk that works. But worrying about chatter during a Fire Case is akin to worrying about the remaining coal tonnage on the Titanic - ordinarly a significant problem, but rendered trivial by the current disaster conditions.

After a significant fire, ALL PRV's will be pulled and rebuilt or replaced - springs may have lost temper, disks may be chattered, etc. Disaster has a very different set of rules than Normal Operations.

 

[Latexman]

I assume inlet dP < 3% for the gas blow by case.

If so, is inlet dP > 3% for fire case, or is inlet dP < 3% for fire case and the PSV will cycle from set pressure to re-seat pressure due to the dynamics of the system/fire? I don't consider the latter, chattering, but one does need to consider the frequency of the chattering.

Good luck,
Latexman

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

 

[zdas04]

The essence of the advice above is that relief cases less than the controlling case just don't matter to the code.

If I have two credible scenarios that have widely different required rates, then I put two PSV on the vessel--a small one sized for the lower case set around 95% of MAWP, and another one sized for the difference between the required rate for the small case and the required rate for the bigger-flow case set at something like 102% of MAWP. This staged approach results in much smaller valves (less potential for chatter) and better control.

Read Duwe6's post carefully. After a fire you need to be a lot more concerned with assessing what components actually survived than worrying about chatter.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[Latexman]

. . . but one does need to consider the frequency of the cycling. duh!

Good luck,
Latexman

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

 

[don1980]

Side Note: Valve chatter is a safety issue - it has little to nothing to do with preserving the investment in the valve. A chattering PSV won't deliver the expected capacity. When a valve chatters, its flow capacity is a fraction of the value for which it was sized. Also, a chattering valve can sieze up at any position between 0 - 100%, or it can just come apart (especially when in liquid service, due to the high impact forces). The risk of having to repair/replace the PSV is a trivial matter.