PSV Sizing - Fire Case Only 1

[Tweec]

A late stage water filtration vessel operates fluid-packed. When sizing the PSV for FIRE condition only, should a liquid, steam, or vapor service valve be used? The contents are mostly water with some oil. One school of thought says that the vessel will first relieve liquid (b/c it's fluid-packed), then vapor, so liquid is the determining factor. Another feels that the valve should be rated for the vapor, in particular, steam, b/c the primary fluid will boil in the event of a fire. While the third says that although the contents are primarily liquid water, that the valve only needs to be rated for vapor, not steam, service. There is a significant cost difference between vapor and steam service valves. The valve is for fire only, so in the event it relieves, it will be replaced, so the most basic PSV is desired.

 

[dcasto]

see thread1203-180139

 

[Tweec]

I read that thread. It doesn't answer the questions completely. I was hoping for some new thoughts.

 

[juancito]

PSV`s for fire design depends on what is inside the vessel, vessel dimentions and operation and design conditions.
From my understding its selection is more delicate than a std PSV.
I recomend you to transfer your query to a PSV supplier,

 

[doct9960]

Tweec,

You would probably get more answers if you reference this forum question in the Chemical or safety relief valve forum groups. Here are the forum links...
forum124
forum798
forum1203

Here's my insight. If your vessel does not generate steam or is not subjected to steam during normal operating conditions, then it is not a "steam service" vessel.

If your vessel & relief valve is contructed to ASME VIII Div. 1 check:
1. the requirements your Jurisdiction
2. ASME VIII-1 UG-125 thru UG-137
3. ASME VIII-1 Nonmandatory Appendix M, M-13
4. API RP 520
5. valve manufacturer

Here are some websites that may be of interest to you...

http://www.pdhcenter.com/courses/m112/Selection and Sizing of Pressure Relief Valves.pdf

http://www.tycoflowcontrol-na.com/ld/CROMC-0296-US.pdf

 

[doct9960]

Tweec,

I forgot to mention API RP 521 and DIERS (Design Institute for Emergency Relief Systems). API 520 & 521 contains conventional or traditional methodologies of relief systems, while DIERS is the new design approach.

BTW, you mentioned your vessel operates fluid-packed, which I interpret as a "liquid full" system. Your vessel could be considered as a two-phase relief. Your vessel should be analyzed by your own or designated relief specialist. If you are the Owner, you are liable for the relief design because of a lot of uncertainties (e.g. operational factors, fluid data, reaction data, disengagement behaviour, etc.)

I am not into relief system design myself. I just happened to attend one of the training seminars of our process group years ago. I learned that relief systems would not be effective if your vessel is not structurally designed for a fire case. You also need to consider fireproofing and fire-resistant insulation covered with stainless steel cladding. If your bare vessel is engulfed in fire, the rise in temperature will reduce the strength of your vessel material, which could eventually lead to failure or bursting. The vessel might even burst first before the PSV opens. If your vessel is in a fire hazard area, this may lead to a domino effect and set the surrounding areas and equipment on fire.

 

[Tweec]

doct9960, thanks for the info. I've posted another thread regarding API 521, which I've read several times (as well as API 520). The topic of fire relief is covered specifically in 521, but the version I have is unclear and not thorough. I am in the process of obtaining the latest version, which I hope will help answer my other questions.

juancito, I've contacted several mfrs, but their response is "you size it, we verify our valve can handle it." I recently located one that has expressed an interest in discussing it further. Perhaps that will be fruitful. BTW, API 521's orifice sizing equation does not take into account what is inside the vessel, other than the calculation of F'. And even then, F' a suggested range is given that should be used when the data is unknown for calculation. AND, there is no calculation of the heat absorbed b/c latent heat decreases over duration of the fire to zero.

I thank both of you for taking the time to respond.

 

[vesselguy]

Tweec,

Unless your vessel is normally filled all the way up to near the vent nozzle, otherwise, how can it be reasoned that it would be liquid be vented first? Water is not an expansible fluid unless it is heated to boiling point and becomes steam.

Like the other guys said, read throughly API 520/521. Search the net, when I was sizing PSV, I found a free PSV vendor software that I used to help check my sizing. Now, only I can recall what the hell it was. Just search the major PSV vendors on the net and I'm sure you'll find it.

Good luck.