8 replies to this topic

[KenCummingsPEng]

We're being asked to install PSV's on a lean amine system where it has been discovered (following an incident) that the booster pumps are capable of producing a discharge pressure higher than the pipe rating of the large distribution network.  The spillback system (which, fortunately, goes back to the large atmospheric surge tank) is being upgraded but local code does not allow us to take full credit for this instrumentation and we're required to use the PSV to protect against the PV fail closed scenario.

 

We have API 520 Pt. 2 as a required code and considered the word "shall" in the first sentence of section 11.1 as making free draining as a REQUIREMENT.  This led us to doing the initial design with a riser pipe from the pump discharge, in order to have the PSV above the top of the surge tank thereby ensuring that the discharge line from the PSV is self draining.

 

Our customer is, rightly so, looking to minimize cost and is suggesting that we should be able to just operate in a liquid filled system, and have a minimal piping system with the PSV discharge line relieving into the spillback line, thereby saving substantially on installed cost.  They used the example of thermal PSVs, often installed 'around' block valves (for instance, on an exchanger that can be isolated) as an example that PSVs can be accepted in cases where it is liquid service on both sides.

 

I'm struggling with this contradiction and would appreciate your comments/suggestions, especially if there are specific API, ASME, .... codes that can be referenced (we are bound to both sections of 520, ASME Section VIII, B-31.3 and CSA B51 by local regulation).

 

Thank-you in advance for any guidance you can provide.

 

Ken

[KenCummingsPEng]

Please note - I believe that this is the same question as that posted by Alfreedo recently.  If you have answered there, don't bother reiterating it here, I'll follow both posts.  Sorry about the second similar question.

[Zauberberg]

What is it, that you are proposing? How does relative elevation look like, between the pump, the proposed PSV, and the tank? A sketch would help tremendously in looking at this problem.

 

Having the PSV at any elevation below the (maximum) tank level will cause the PSV discharge line to be liquid-filled, and this liquid will also impose certain backpressure on the PSV.

[fallah]

We're being asked to install PSV's on a lean amine system where it has been discovered (following an incident) that the booster pumps are capable of producing a discharge pressure higher than the pipe rating of the large distribution network.  The spillback system (which, fortunately, goes back to the large atmospheric surge tank) is being upgraded but local code does not allow us to take full credit for this instrumentation and we're required to use the PSV to protect against the PV fail closed scenario.

 

We have API 520 Pt. 2 as a required code and considered the word "shall" in the first sentence of section 11.1 as making free draining as a REQUIREMENT.  This led us to doing the initial design with a riser pipe from the pump discharge, in order to have the PSV above the top of the surge tank thereby ensuring that the discharge line from the PSV is self draining.

 

Our customer is, rightly so, looking to minimize cost and is suggesting that we should be able to just operate in a liquid filled system, and have a minimal piping system with the PSV discharge line relieving into the spillback line, thereby saving substantially on installed cost.  They used the example of thermal PSVs, often installed 'around' block valves (for instance, on an exchanger that can be isolated) as an example that PSVs can be accepted in cases where it is liquid service on both sides.

 

I'm struggling with this contradiction and would appreciate your comments/suggestions, especially if there are specific API, ASME, .... codes that can be referenced (we are bound to both sections of 520, ASME Section VIII, B-31.3 and CSA B51 by local regulation).

 

Thank-you in advance for any guidance you can provide.

 

Ken

 

Ken,

 

The local code rightly don't allow you to take credit for spill back line in the case of pump blocked discharge; and if it can be supposed when the PSV is activated, the PV at spill back line is stuck closed and the PSV discharge line is connected to spill back line at PV downstream; then the part of PSV discharge line which is common with spill back line and will be terminated to the surge tank can be without the lioquid inside hence appears there would be no problem in this regard just the back pressure at the moment of PSV relief should be checked and it's better using a modulating pilot operated PSV in order to not only avoiding PSV chattering in liquid service but also to cope with the probable high back pressure in PSV downstream at the moment of PSV relieving...
 

[KenCummingsPEng]

Sketch Attached.

 

Thank-you for the interest/comments so far.

Attached Files

•  Amine PSV Sketch.JPG   43.17KB   17 downloads

[Bobby Strain]

This is not a thermal relief! I would opt for the free draining installation. Someone already made a poor judgement in rating the discharge piping. Don't compound the problem.

 

Bobby

[fallah]

Ken,

 

Because the valve at incoming line to the surge tank is CSO; then the main issues for alternate design would be the capability of spill back line to handle the relief load in line size standpoint and checking if the maximum total back pressure at the moment of PSV relieving can decrease the PSV rated capacity...

[KenCummingsPEng]

Thank you for your additional responses.

 

a) Bobby - sorry I was unclear.  When I mentioned 'thermal' it was to indicate the existence of 100% liquid PSVs, contrary to our 520 Pt2 section 11.1 interpretation. 

 

Naser - we agree.  Fortunately, the high pump discharge pressure (and PSV setpoint of ~ 1,700 psig), combined with an early decision to use a regulating pilot operated PSV, make our life a bit easier in this area.  Additionally, we don't have to account for much spillback flow in the line when the PSV operates (just the intersection of the pump curve and the PSV set pressure) as any higher flow in the spillback would push the pump out on its curve and negate the requirement for the PSV to operate.

[KenCummingsPEng]

sorry about the emoticon - it was supposed to be the letter 'b', followed by a bracket .  This emoticon is on purpose: