12 replies to this topic

[Ryush806]

I have a new vessel that will be installed that will be liquid full and I need some help with a relief valve.

 

We are not considering fire case for the relief valve as the fluid has a high boiling point and a relief valve will provide little protection. The cases we identified were blocked outlet and thermal expansion. Blocked outlet load far exceeds thermal expansion as can be expected.

 

The tank will be far away from any existing knock out pots and flare header so it was suggested that the relief valve could relieve to the downstream side of the block valves that could be closed creating the relief scenario. A simplified drawing of this idea is attached. I know you would have to run calculations to make sure back pressure would not be too high and then potentially use a bellows or pilot valve. From what I can read in API 520 and API 521 this seems to be an ok thing to do, but it rubs me the wrong way. Maybe I've just never seen it done before. Does anyone else have any opinions on the matter? If you believe this may be an acceptable solution, is there anything I need to make sure and check? If you don't like this idea, please tell me why.

 

Thanks so much for your help.

 

 

Attached Files

•  Capture.JPG   14.6KB   6 downloads

[AlertO]

Hi Ruyush806

 

 

Normally, PSV discharges to another system we ensure the low pressure is there. From your figure, your PSV cann't protect your system if you have any additional valves downstream which can be closed because the PSV back pressure is already at the relieving pressure.

 

have a good day.

[fallah]

 it was suggested that the relief valve could relieve to the downstream side of the block valves that could be closed creating the relief scenario.

 

Hi,

 

Please specify "downstream side" to which you referred...Is it closed drain?...Is it safe location?...Is the liquid a hydrocarbon or flammable fluid?...

[shan]

Yes.  It works.

[Bobby Strain]

You don't need a PSV to protect for a closed manual valve.

 

Bobby

[Ryush806]

 

 it was suggested that the relief valve could relieve to the downstream side of the block valves that could be closed creating the relief scenario.

 

Hi,

 

Please specify "downstream side" to which you referred...Is it closed drain?...Is it safe location?...Is the liquid a hydrocarbon or flammable fluid?...

 

 

By downstream I mean toward the right side of the block valve in the illustration I provided. Keep in mind that the illustration is simplified and I mean for it to show that the PSV would discharge past all block valves in the line even though I only show one.

 

You don't need a PSV to protect for a closed manual valve.

 

Bobby

 

 

Can you explain why I wouldn't need a PSV there? If my centrifugal pump can overpressure the tank, then it seems that someone inadvertently blocking the outlet of the tank would require a PSV.

 

Thanks!

[fallah]

Ryush806,

 

Please upload a simplified sketch showing all equipment (pump, tank,..) and also PSV downstream facilities...

[Fr3dd]

It doesn't seem right. Just remember that the psv outlet must discharge to the final disposal system (atmosphere, LP/HP Flare, etc) with a free path; that means no valves that could closed by mistake or restrictions. It's hard to believe in a WHOLE downstream system without a single valve.

 

 

Can you explain why I wouldn't need a PSV there? If my centrifugal pump can overpressure the tank, then it seems that someone inadvertently blocking the outlet of the tank would require a PSV.

 

If your pump is centrifugal, it will run until the pressure of the system reachs its shutoff pressure; if your system is designed properly, the drum should be able to withstand this pressure. That's why you don't see PSV at centrifugal pump discharges (only low flow control) but you see it in other types of pump.

[Ryush806]

Ryush806,

 

Please upload a simplified sketch showing all equipment (pump, tank,..) and also PSV downstream facilities...

 

Done. There's really not much to this system.

 

It doesn't seem right. Just remember that the psv outlet must discharge to the final disposal system (atmosphere, LP/HP Flare, etc) with a free path; that means no valves that could closed by mistake or restrictions. It's hard to believe in a WHOLE downstream system without a single valve.

 

 

Can you explain why I wouldn't need a PSV there? If my centrifugal pump can overpressure the tank, then it seems that someone inadvertently blocking the outlet of the tank would require a PSV.

 

If your pump is centrifugal, it will run until the pressure of the system reachs its shutoff pressure; if your system is designed properly, the drum should be able to withstand this pressure. That's why you don't see PSV at centrifugal pump discharges (only low flow control) but you see it in other types of pump.

 

Changed the drawing to show the PSV discharge going straight to storage tank just to make sure and not confuse the issue that the discharge will go past the last valve most definitely.

 

This new vessel is an add on to an existing process so the company does not want to consider changing pumps. Also the new vessel will be quite large and increasing the MAWP will result in a large price increase as well. Normally, as you said, we could design the whole system so that this would be a non-issue.
 

Of course I realize that changing the pump might be cheaper in the long run...I don't get to make these decisions though unfortunately.

 

Thanks again for everyone's help!

Attached Files

•  Capture.JPG   26.63KB   5 downloads

[shan]

You may consider using a Buckling Pin Relief Device instead of the conventional PSV.

Edited by shan, 30 July 2015 - 09:45 AM.

[Fr3dd]

@Ryush806

 

So basically you're relying on and atmospheric tank to do the work relieving overpressure of what it seems to be a pressure vessel. This configuration is not right as pressure in atmospheric tanks is usually low (below 15psig).

 

I would strongly recommend you to take a closer look at API 521, although it says that a single PRV can discharge to a system with a lower pressure, it also states that this can only be considered if the relief load of the PRV does not result in overpressuring the lower-pressure euipment. I think you won't comply with this under this configuration.

 

I would also advice not to look at the tank as another pressure vessel; it is constructed differently, set for a different purpose and protected against overpressure (or vacuum) in different ways that pressure vessels.

 

Please note that all this recommendations are based on the assumption (since you have not provided any data regarding operating pressure of the system) that it is a pressure vessel, operating at a relatively high pressure (more than 15 psig). Please, have your supervisors take a closer look to this, since safety in operations may be compromised.

[Ryush806]

@Ryush806

 

So basically you're relying on and atmospheric tank to do the work relieving overpressure of what it seems to be a pressure vessel. This configuration is not right as pressure in atmospheric tanks is usually low (below 15psig).

 

I would strongly recommend you to take a closer look at API 521, although it says that a single PRV can discharge to a system with a lower pressure, it also states that this can only be considered if the relief load of the PRV does not result in overpressuring the lower-pressure euipment. I think you won't comply with this under this configuration.

 

I would also advice not to look at the tank as another pressure vessel; it is constructed differently, set for a different purpose and protected against overpressure (or vacuum) in different ways that pressure vessels.

 

Please note that all this recommendations are based on the assumption (since you have not provided any data regarding operating pressure of the system) that it is a pressure vessel, operating at a relatively high pressure (more than 15 psig). Please, have your supervisors take a closer look to this, since safety in operations may be compromised.

 

Thank you very much for your answer! You confirmed my suspicions that this isn't a good arrangement. 

[fallah]

Ryush806,

 

Why didn't you think about the possibility of conducting the PSV outlet to the surge tank at pump's suction side?