6 replies to this topic

[J_Leo]

Dear Fellow Engineers,

 

Please refer to the relief system sketch attached. This is only the simplified system to shown the scenario I am talking about.

 

I am sizing the PSV1 in the sketch. If the control valve to the drum fails open, more vapor flows to the drum.  Because the downstream PSV2 on the distillation column is set at lower pressure (150psig compared to 170psig), I expect PSV2 will relief first. The concern is there is a block valve (not CSO / LO) in the pass between PSV1 and 2. Should I consider the Control Valve Fail Open as a credible relief scenario for PSV1 sizing?

 

Thank you.

 

Regards,

Leo

Attached Files

•  Relief system sketch.pdf   164.1KB   34 downloads

Edited by J_Leo, 03 June 2017 - 08:36 PM.

[Bobby Strain]

Sketch?

Bobby

[J_Leo]

Bobby,

 

The sketch is added. I found the topic was posted twice so I tried to delete one. I deleted the sketch instead of the extra post by mistake.

[Pilesar]

You must consider all credible scenarios and control valve failure is credible. Do not ignore it. You should evaluate the system to determine the overpressure risk for the scenario. Perform the calculations. Perhaps the flow will pass through enough so that you do not have an overpressure event. There will be some pressure restriction in the downstream piping even if the downstream block valve is open. You must calculate that. There is a separate scenario for 'blocked outlet'. If the control valve on the inlet line is for flow control, then blocking the outlet will result in the inlet valve full open and that would be another credible scenario.

[J_Leo]

You must consider all credible scenarios and control valve failure is credible. Do not ignore it. You should evaluate the system to determine the overpressure risk for the scenario. Perform the calculations. Perhaps the flow will pass through enough so that you do not have an overpressure event. There will be some pressure restriction in the downstream piping even if the downstream block valve is open. You must calculate that. There is a separate scenario for 'blocked outlet'. If the control valve on the inlet line is for flow control, then blocking the outlet will result in the inlet valve full open and that would be another credible scenario.

 

Hi Pilesar,

 

Thank you for your advice. I did the calculation already. Since we can only take half of the normal OVHD flow from the drum for credit, relief is required if this is a credible scenario. Based on your advice, relief is required through PSV1.

For the blocked outlet case, I thought it is a double jeopardy if we consider the control valve fail open at the same time, no matter what kind of control valve it is. Per your statement, it is no longer a double jeopardy if it is a flow control valve. Is this from any standards such as API? 

 

Thank you.

Regards,

Leo

[Bobby Strain]

If you are the owner, it is up to you to decide. However, if you are doing engineering for a client, be sure you don't assume any unnecessary risk to your company. It's not uncommon for designers to base relief loads not only due to control valve failure, but to include normal flow, too, through the control valve bypass.

 

Bobby

[J_Leo]

Bobby,

 

Thank you for the reminder. I considered 100% open for the bypass when I did the relief load calculation.

 

Regards,

Leo