8 replies to this topic

[Danybony]

Hello,

I'm come across a quite mangy PSV sizing. I didn't find any similar argument across the forum, so I've started a new topic. I have a PSV that is mounted on a pig trap (receiver) that must be sized for Fire Scenario but that has a Set Pressure of 211.2 barg (relief at 255.6 according to API): I've tried to simulate the crude oil conditions at that pressure but obviously HYSYS fails in doing that (supercritical ?). How would you proceed? Any hint or help is very very appreciated.Thanks!!

 

Daniele

[fallah]

Daniele,

 

Please specify if fire scenario is only credible scenario for mentioned PSV. If so, the PSV mostly cannot protect the pig receiver from failure due to the fire case, then the sizing wouldn't be so critical and can be done with a proper assumption for latent heat of the inventory during PSV pop up...

[Danybony]

Hello Fallah,

thanks for the answer. Yes, fire scenario is actually the only credible one: I first tried to size the psv for thermal expansion of the trapped liquid crude oil but considering the thermal flux that you get applying the API 521. I don't know if this approach is correct (in my opinion, it's not!!). But on the other side, I don't know how to calculate the latent heat of vapourization of the crude oil, unless you decide to apply the standard value proposed by API 521 (115 kJ/kg, if I remember well).

[fallah]

Daniele,

 

Please upload a detail sketch of the system along with the relevant needed info such as:

 

- Is the crude inside the pig receiver being trapped by isolation valves during a fire case?

 

- Is the trapped crude in liquid state?

 

- How much is the volume/weight of the trapped liquid?

 

And any related info you have in hand at the moment...

[Danybony]

Hello Fallah,

I've attached an extract of the pig trap from the P&ID. The pig trap is essentially a piece of 16" pipe with a total length of 9000 mm, so the volume of trapped crude is not much. The crude inside is in liquid state (dead oil) and is possibly trapped inside in case of emergency.

 

Daniele

Attached Files

•  PIG TRAP.pdf   24.68KB   121 downloads

[fallah]

Daniele,

 

For the piping system you described: can be isolated and can be exposed to fire; a normal and good enginering practice says it needs to PSV for fire case just if the weight of the trapped liquid would be higher than around 2 tons. Then if a PSV isn't required, a TSV should be installed for thermal expansion case...

[Jiten_process]

Downboy,

 

It is difficult to do PSV sizing for supercritical phase condition. I am still not sure whether the data and pressure which you are going is correct. however, I am will not spy on it and i assume you are firm with your crude data and set pressure and fire scenario. 

 

What I can tell you is. For such scenario where fluid goes to supercritical condition on set pressure, there are two approach i have come across. First is, reduce the set pressure of PSV so as to get the fluid on below critical condition, however this is costly approach as PSV size increases unnecessary. Second approach is, if you refer API-521, (i dont remember exact clause number at the moment but) for such instances, API has recommended latent heat value to be considered. Alternatively sometimes, you can consider hexane property also however client/consultant concurrance is strictly required for that. 

 

Lets c what other experts have to say on this. 

 

Good luck...

Edited by Jiten_process, 13 January 2014 - 07:10 AM.

[chemsac2]

Danybony,

 

That is the typical situation for PSVs on pig traps. Crude oil pipeline design pressure would be high due to pump shut-off. Typical crude oil critical pressure would be much lower at 50-60 barg.  

 

API 521 suggests sizing PSVs for heavy liquids, supercritical fluids similar to non-wetted fire case i.e. gas expansion.

 

I would be surprised if PSV size would be any bigger than D or E orifice.

 

In fact, Shell DEP asks only for TSVs on pig traps.

 

I have seen few projects without PSV/TSV on pig traps as pigging is manual operation and hence pool fire is unlikely as traps would be emptied after pigging. Also, hardly any source of spark exists in pig trap areas as they are away from processing areas. Also, pig traps are designed for pipeline code (ASME B31.4/31.8) and not ASME section VIII which asks for PSVs by default for all stamped vessels.

 

Sachin

[Danybony]

Sachin,

the problem is that sizing such valve for non-wetted fire case is impossible: at 255.6 barg the relief temperature would exceed the maximum allowable metal temperature (591.3 °C, if I remember well).

What about sizing it for thermal relief of liquid but considering the thermal flux calculated as per fire scenario-boiling liquid, even if it's a kind of misuse of API 521 ?.

 

Daniele