5 replies to this topic

[stu]

Dear sirs,

I have come across a Chemical Injection Skid P&ID where there is a PSV at the Pump ( Plunger Pump) discharge.

The PSV outlet is routed to the Source tank. The source tank is designed for atmospheric pressure operation. Is it wise for us to route the high pressure PSV discharge into a tank that is designed for atmospheric pressure operation?

Please advice.

Regards,
Stu

[Art Montemayor]

Stu:

I don't think you are thinking about this application correctly. You are dwelling on taking a very high pressure and discharging it into an atmospheric vessel. If you stop and think about the specific operation - the injection of a high pressure LIQUID into an atmospheric vessel - then it should be crystal-clear that there is nothing wrong with the proposal because: there is no expansion energy released by the HP liquid. It would be a different story if we were dealing with a HP gas.

I am presently applying the same, identical operation in taking the PSV discharge from a triplex reciprocating pump and discharging it at the source tank. I did this in order to avoid doing it in the suction piping - something I always try to avoid - something that we old & experienced engineers always try to do to avoid gas binding a PD pump or contributing to a liquid hammer in the system. It is always preferable to recycle back to the source - if it is safe and convenient.

PS. I edited you post so it made sense in English.

[fallah]

In addition to valuable explanations of Art:

In the case of flashing or two phase flow concerns,at first the amount of relief would be very low with respect to tank volume and on the other hand the tank certainly has its own protection against overpressure.

[Jack Bauer]

With reragrd to concern address by Stu, I think I may have similar issue and may need you opinion on this as well.

I'm reviewing a relief system for a tank.
The scenario is as follow;

Crude inlet to a crude tank. Tank design pressure is Water Full, operating at atmospheric.
There is a PSV at the inlet pipe,(PIPE A), governing scenario is Blocked Outlet, with set P is 15barg.
Downstream of the PSV, there is a shutdown valve, SDV. Max Upstream P is high enough to cause overpressure in the event of blocked outlet.

So i assumed the Blocked Outlet is caused by the SDV closure.
The PSV outlet relief piping is routed to the same pipe (PIPE A), after the SDV, then the relief flowrate will eventually enter the crude tank.

The question is; does the arrangement looks weird since the relief pressure from the PSV (set @ 15barg) is fed to the pipe and eventually to the atmospheric tank.
Is this configuration will ovepressure the tank, considering the relief pressure is 15barg?

Or does this config is OK, considering the the ovepressure source is liquid (crude) not gas? and liquid expansion is insignificant.

Tank size : ID=65meter, Ht=16m.
Pipe size: Main line is 24in, PSV size 8in x 10in, T orifice.

See attached skecth for better understanding.

TQ for your response.

Attached Files

•  Tank 1.pdf   46.53KB   72 downloads

[fallah]

If two phase flow wouldn't be created (considering low set pressure of PSV and superimposed back pressure due to liquid static head in tank seems there wouldn't be) in PSV discharge line during PSV relief,there is no concern about tank overpressure due to PSV relief.

But if there would be two phase flow in PSV discharge line,PVRV on the top of the tank should be reviwed to handle the load due to PSV relief without tank overpressure.

[paulhorth]

Jack,

Putting non-flashing liquid into an atmospheric tank will not overpressure it, if the tank vent is large enough. But, it will fill it up, and if the flow continues after the tank is full, then you will be spilling crude out of the vent, and all around the tank, creating a fire hazard and a pollution problem. It is also possible that the tank will be overpressured once it has become full, because the back pressure due to liquid flowing out of the vent will be several hundred times greater than for the same volume flow of vapour.
DP is proportional to (rho) (v)^2, and whhile v will be the same for the same volume flow, rho will be the liquid density in place of the vapour density.

So you need to have in place something that tells the operators that the tank is filling, (a high level alarm), then some means to shut off the flow at the source.

Paul