PSV discharge into bigger Expansion Bottle

[Chap123]

Hello Guys,

My client design a skid in which they discharge PSV into another pipe that already installed with two PSV.

I dont think its a good engineering to vent relief pressure into the piping system.

What you guys think?

Attach is the snap shot

 

[georgeverghese]

What is this new PSV 1002 designed to do - blocked in liquid expansion thermal relief ? Is this PSV a balanced bellows type ?

 

[LittleInch]

It's not great and could have complex permutations of pressure during a HAZOP.

Cascading pressure reliefs are sometimes used where volumes are low and running relief pipework or vents is difficult, but this doesn't look like it to me.

Also in effect the 240 psig set point is actually 250 because the final relief to atmosphere is controlled by the dual port relief (two PSVs in One body??)

PSV 1003 below it doesn't have any vent connection??

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Chap123]

Thanks guys.

PSV-1002 is meant to release vapour fluid when the system between ESDV-1000 and P-1000 is blocked and thermal expansion (or fire) occurs and the expansion bottle is empty and hence the PSV will release into an empty "vessel". Since thermal expansion is a low volume release the expansion bottle (if empty) is sufficient to contain the release and avoid release into the environment. In the event of a fire, and the expansion bottle full, PSV-1002 will release into a full fluid and hence cause PSV-1000 and 1001 to release. According to API 521 it is not recommended to have restrictions downstream of a PSV and it is not all.


PSV 1003 is just venting in the air.

 

[LittleInch]

I can't see P 1000, but even then PSV 1003 is connected to the same pipe as P 1002 via a manual valve. OK it's set 10 psi lower, but that's nothing so it will be merrily venting to air (should be shown on the P&ID).

Are the ESD valves held open reverse check valves? Never seen that before.

so yes it could be done, but seems a bit strange to me.

I would set the PSV a lot lower than 240 to avoid interaction with the other two valves

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Chap123]

Thanks ,

Do you know where it shown in API521 that corresponds to my situation?

 

[Latexman]

Its hard to know with a partial P&ID preventing one to know what is going on upstream and downstream. I doubt API 521 has something like this.

If 3" 300# FP is closed, why would the pressure on the right side not thermally expand like the left side? What does FP mean? If the right side does not thermally expand like the left side, I can see the logic, but 240 psi is too close to 250 psi. If the right side does thermally expand like the left side, I do not see the logic, and the answer to George's balanced bellows question gets critical.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[georgeverghese]

This perception that this expansion bottle will be empty is completely wrong. It would be full of fluid at the max normal operating pressure of the main line and hence this PSV 1002 will have to operate to relieve at that imposed backpressure.

Balanced bellows type PSVs' can tolerate some backpressure while conventional type PSVs' have a zero tolerance to backpressure at the exit. This is wrt the integrity in setpoint maintenance.

A balanced type PSV will, in this case, allow a max built up backpressure of up to 0.8x240psig = 190psig, with no setpoint drift, while a conventional type PSV is essentially a differential pressure relief device, and is completely useless in this application.

So, to allow for some room for increasing backpressure as PSV1002 operates, I would say you cannot have a main line operating pressure exceeding 150-160psig, and that PSV1002 should be balanced bellows type. The sizing of the RV should also take into account this max backpressure on this RV. This is a sloppy design, but given this is very small capacity solar thermal relief ( firecase relief is not applicable to piping, only to equipment designed to pressure vessel codes), it may be tolerated if these conditions can be met. You have a high pressure trip PAHH 1009 set at 150psig, so one of these preconditions appears to be met.

 

[Chap123]

Thanks for the explanation.

Our oeprating pressuure is 120psig.

I will have a look if vendor used balanced Bellows type valve.

Latex i have attached bigger picture for your consideration.

 

[georgeverghese]

Would suggest the following additional modifications, which I believe will futher reduce process safety risk here:
a)Reduce setpoint at PSV 1000/1001 to 185psig, which is safely less than the 192psig backpressure upper limit for a balanced bellows type PSV1002.
b)Reduce setpoint at PAHH 1009 to 0.9x185, round off to 165psig (max permissible inlet pressure limit for conventional PSV1000/1001 to avoid relief simmer)

 

[Latexman]

It's a trans-loading skid.

Please let us know whether PSV-1002 is balanced bellows or not. It's critical.

Since they are not trying to maintain an inert vapor pocket in the "expansion bottle" (unless they are doing it with a manual procedure), and assuming NGL vapors are not regulated too heavily (I may be wrong here. Not my area of experience), I think I would get a conventional PSV and pipe the outlet of PSV-1002 to atmosphere or to the outlet of PSV-1000. So if PSV-1002 is a conventional PSV, you now know a fix, because I don't think a conventional PSV will work as drawn.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[Chap123]

Thanks Latexman,

Its a conventional PSV valve.

If we changed it into balanced as per what george said, design will work ad such ?

 

[Latexman]

Do you have any vapor pressure vs. temperature data for the typical (or worst case/highest vapor pressure) LNG the facility will handle? Is everything at ambient temperature? I see no insulation. Is the pump and piping inside or outside (shaded or exposed to sun)? What's the temperature of your hottest day?

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.