3 replies to this topic

[jprocess]

Dear All,
We should do fire case relief calculation for a slug catcher.As you all know slug catcher is a huge equipment in comparison with other typical vessels.
Slug catcher is finger type pipe array, 8 fingers of 40'' nominal diameter, 200 m long.The liquids collection header is 40'' nominal diameter and estimated at 15m long.
As you see the exposed wetted area is a huge value results a huge total absorbed heat and consequently a huge relief load.This may warrent using more than one psv.
I want to know that should we consider all the exposed wetted area for this special case(specially its 200 m length!)?
Another concern is that the liquid inside slug catcher will reach its critical conditions befor psv set pressure! so there is a problem with latent heat calculation! API 520&521 have no discussion in this field!
I found this article about supercritical relief: " Rigorously Size Relief Valves for Supercritical Fluids " but as you know with papers we can not talk to clients!
Also should we consider unwetted portion of slug catcher that contain gas?
Thanks in advance.
Cheers.

[gvdlans]

I would consider only those parts where a pool fire is likely to occur, which is at locations where flanged connections are provided. Probably you will have a concrete slab below the inlet header and below the liquid outlet header. You can assume a pool fire on one of these slabs as your fire case. A pool fire below a welded pipe is not likely to occur.

I do not understand you when you write "but as you know with papers we can not talk to clients!". Can you talk to clients based on a posting in this forum?

[jprocess]

Dear Sir,
Thanks a lot for your post.
We can not talk with clients with papers and also based on comments.This is only a forum that engineers can share their experiences.In some cases these experiences can help us to solve our problems.
Best of Lucks.
Cheers.

[JoeWong]

To admin,
Suggest to relocate this post to Relief Devices Forum as it is relief related...

jprocess,

I would consider only those parts where a pool fire is likely to occur, which is at locations where flanged connections are provided. Probably you will have a concrete slab below the inlet header and below the liquid outlet header. You can assume a pool fire on one of these slabs as your fire case. A pool fire below a welded pipe is not likely to occur.

I do not understand you when you write "but as you know with papers we can not talk to clients!". Can you talk to clients based on a posting in this forum?

I do agree with gvdlans to certain extent. The basis needs to be fully clarified and documented. One may say that a truck carrying LPG passing by this area, accident and causing serious leak under the slugcatcher bottles....any one may answer how often a truck full of LPG pass by this area, one may say the truck loading area is normally at the plant boundary and slugcatcher is located in remote area. Above event almost not possible......one may propose restrict LPG truck pass through this area. Layout to minimise LPG trukc come into this area...Ultimately all this requirements / actions shall be properly documented and cascade down to all disciplines...

A slopping slugcatcher with 200m lenght may result the gas end high and potentially exceed 7.63m requirements. You may have the chance not to consider the gas end. Just check...

As for liquid end, not much different than a vessel expose to fire case. The only different is the pool fire extent may be limited by the concrete slab below.

Another concern is that the liquid inside slug catcher will reach its critical conditions befor psv set pressure! so there is a problem with latent heat calculation! API 520&521 have no discussion in this field!

This always occur at low operating pressure and temperature with reasonable high design pressure system...other than using the gas expansion method...not sure what other "practical" method can be used.

As for << Rigorously Size Relief Valves for Supercritical Fluids >>, below is the extraction of procedure.
 Extraction.bmp   556.69KB   167 downloads.

I did not go very detail but just spotted the heat input equation is the API equation 21000A^0.82. Wonder if this equation is applicable for "supercritical" fluid. Heat transfer flame->wall->supercritical fluid may not so appreciable if supercritical fluid behave "like gas"...if A used is entire vessel, then of course it is conservative...

Secondly it calculate mass flux considering isentropic expansion when it across the PSV by considering idea nozzle, this quite inline with the recent findings by Dr. Darby from Texas Uni and Mr. Shackelford from Berwanger...

I am just thinking (if you want), you may consider to simulate the system with step change (dt of 3 degC) in temperature (w/o changing relieving pressure), consider entire vessel is "wetted" for conservatism for heat input in order to calculate the step relief load. Calculate the relief load with HNDI method proposed by Dr. Darby. For sure it is very ACADEMIC and TIME consuming...but at least a method for you...

JoeWong