Is it a good idea to relieve in a atmospheric tank? 1

[alpha0125]

Hello,

PSV set at 313 psig

Is it a good idea to relief this PSV into a atmospheric closed water tank? I heard that presure will convert into velocity and turbulenace when it exit from the tailpipe of PSV into a 180 barrel water tank.

Thank you!

 

[zdas04]

It is done all the time, it is successful when the tank vent is adequate for the flow. It is unsuccessful (and downright dangerous) when the tank vent is inadequate.

I often see this configuration on sites where there was zero engineering in the exhaust piping (and no one considered the dP in the pipe to the tank changing the setpoint on an conventional PSV) or tank vent.

I have done the engineering on a couple of these and the clients were dumbfounded that I specified a pilot operated PSV (added cost), piping that was generally 2 pipe sizes bigger than they expected, and changes in the tank vent.

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[EmmanuelTop]

Please post a sketch of the system including all relevant operating/design parameters. Like David says, almost anything can be connected to almost anything, but I would be reluctant to connect the relief system of an equipment designed for 1,000 or 350 psi(g) to an atmospheric vessel, for many reasons.

Dejan IVANOVIC
Process Engineer, MSChE

 

[LittleInch]

Is it a good idea?

Depends whether you are relieving liquid or gas, and, as ZDAS04 says, how much / volumetric flow rate compared to the vent size of the tank and volume of the tank

If liquid then it might create turbulence, vapour and impinge on the side of the tank, but with the details you've provided (virtually none), it is very difficult to comment further.

Get it right - can be a good way of collecting whatever is coming out of the PSV.
Get it wrong - can be sudden and catastrophic failure of your water tank (think shrapnel and explosion)

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

 

[SNORGY]

alpha0125,

Further to the correct points raised above, I am also in the habit of adding 1.5 velocity heads to the vent pipe head loss calculated at the PSV capacity. You might find that a ~1 meter 8"-D gooseneck suddenly becomes 12"-D, or worse.

Pop tanks are common, but the vents are almost always too small for the service people try to put them into. I have often pulled off the thief hatch and added a second vent, where the service is not sour.

 

[SNORGY]

This is perhaps not a designated "pop tank" but the rationale is the same.

 

[alpha0125]

I have attached my little sketch.

I am also thinking about what if all the liquid are sent to tanks and pipeline is flowing dry. now you have gas going into the tanl? what will happen in this case?

 

[LittleInch]

Nothing attached

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

 

[don1980]

Please refer to the posting guidelines for this site - it's not for school work.

 

[alpha0125]

Sorry,

Here is the attachment

http://files.engineering.com/getfil...2db8-4454-9426-9306d13f6fab&file=question.png

 

[tickle]

Looking at the sketch, is there a reason why the pipework up to the isolation valves on the tanks is not rated to 1,000 psig? I think this would be cheaper and would remove a lot of the issues that I can foresee that will come up during the HAZOP. Note that the tank vents should be sized for the gas break-through case (without suitable controls / shutdowns on the separator).

 

[alpha0125]

tickle,

I am not sure why they did not install valves that good for 1000 psig. I think because pressure won't get to 100 psig unless you do not have choke nipple.
however there is a choke nipple which should reduce our pressure from 1000 psig to very less. and the dump valves are not 100% in reality. All the flanges downstream of choke nipple are 150 psi series for some reason. I think it was a mistake. however, we do not want to replace these flanges to 300 # or 600 # since it takes long time to replace all the flanges ( there are probably 14 to 16 of them on each line).

However my question is about does it matter if i send my everything to tank? i am thinking i am inserting all the pressure again to atmospheric tank even though there is hatch to relief excess pressure which i am sure if it is pratical to do it.

 

[LittleInch]

This is clearly some sort of manual / simple control system here? Your "choke nipple" will be at choked flow.

The thing you miss out is what is the venting arrangement on the tanks? given you're dropping 1000 psi across a choke, what you get in the middle / the other side must be an awful lot of gas and not much liquid? Be warned that the PSV sizing for multiphase flow is very difficult as I can't see how you won't have a lot of gas in both the oil and water streams.

However the key point is if the operator does as you say, then how much time / volume has tank 1 got left before it overflows?

You might want to consider a three way valve on the line instead of two tank valves so that if it changes over, then it automatically opens the other tank (lock open the inlet valve).

given that the flow and pressure will be the same whether you flow into the tank via the tank valve or via the PRV, then I would say this is the next best thing to do, but how you guard against overflow is the next problem on your list.

Of course only having one tank connected to the PRV is dangerous if you ever shut that tank off for maintenance as it could either jet liquid in or if it is shutoff then it won't work.

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