PSV Discharge Piping

[MickMc]

Our Contractor on this FEED stage of project has taken this approach to configuration of PSV discharge piping going to closed system in this case to flare with the purpose being to save money on isolation valve side. (see attachment)

I have not seen this approach before and we have concerns that although they have stated that back pressure is within 30% margin for Balanced PSV and Velocity is below our limit of 0.7 Mach however their calculations are preliminary at this stage with approximate lengths. Our Process guys seem relaxed about this but in Piping our concern is that their approximation does not include the short length before expander and possible vibration at isolation valve if exceeding 0.7 Mach.

We are pushing them for calculations but would welcome any input especially from any members who have seen this approach and any experience when PSV lifted. Note we have this configuration in existing plants designed by same Contractor but seemingly no PSV lifted/popped as yet.

Regards.

Mick.

 

[Latexman]

Not much information to go on. A 4x6 PSV is assumed, and could be L, M, N, or P nozzle size. Which one? P_SET = ? Fluid = ? or Fluid MW = ? Isolation valve is what type? Full bore? 6" pipe looks 5.6D long, about 3' or 1 m. Flare header superimposed backpressure (inerting pressure) = ? 14" pipe length to sub header? Sub header pipe size = ? Sub header pipe length to flare header? Flare header size = ? Flare header length to flare? Flare D = ?


Good Luck,
Latexman
Pats' Pub's Proprietor

 

[LittleInch]

I assume what they're doing is saving on the valve size? SO just ask them to minimise the pipe spool length.

TBH I can't see any glaring issues here.

I would be worried if they reduced the pipe size from the relief valve flange size, but it doesn't look like that.

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

 

[Latexman]

How strict is this 0.7 Mach limit? That seems to be the key question to me. Unfortunately, to get that, one must start at the exit to atmosphere and work backwards to the PSV exit flange. Not an easy task to do in detail, with all the iso's (or 3D model), bits, and pieces, but not so hard for a first pass with pipe diameters and guesstimated lengths.

Good Luck,
Latexman
Pats' Pub's Proprietor

 

[georgeverghese]

Running up at 0.7Mach seems pushing it to the limit - at these velocities, dp/unit length can be steep and unpredictable. Safer to stay within 0.5Mach at the 14inch sub lateral. Reduce the length of the 6inch segment to the minimum. Ask the relevant engineers to do a coarse AIV screening calc on this relief load at this location.

 

[MickMc]

Guys,

Thanks for response.

In the case in attachment it is 4x6 PSV and isolation in this case is gate valve and all valves are mandated full bore at PSV and API 520 guidelines are being met regarding flange sizes etc. This is just one instance in this process plant area with all PSV taking same approach where expander is required due to either velocity or back pressure which presently is every PSV to flare closed system and is to save on valve size. The 0.7 Mach limit is not to be exceeded as per company standard and for FEED stage they are not required to do Flarenet calculation deferring responsibility to chosen EPC contractor and their rationale is that they have used same design in existing plants for same client but I am a bit uncomfortable with this copy and paste approach, note we have no feedback from exiting plants on any PSV lifting with this arrangement.

We have other process unit with different FEED contractor where they are adopting more conventional to me arrangement of isolation valve after expander which is at PSV outlet flange and in EPC stage we will be looking for consistent approach which will be driven by cost. I was hoping to get some feed back on any practical experiences with this arrangement whether positive or negative.

We will be having further meeting with contractor to address our concerns and I will feed back this.

Regards.

Mick.

 

[georgeverghese]

Oops, I missed that trick from your wily process engineers. Indeed, the full bore gate valve should be on the 14inch side.

 

[mk3223]

Imo, may consider to minimize the discharge by moving the gate valve next to the PSV outlet with a drip ring for the 3/4" instrument connection.
Need to orientate the gate valve stem in the horizontal position.