We have a PSV that has been popping

[Bill3752]

We have a PSV that has been popping. It is just downstream of a centrifugal compressor. I wonder if installing a dampener ahead of the relief valve might eliminate some of the pressure spiking. I have used these on reciprocating compressors routinely, but never thought of using after a centrifugal compressor.
Any thoughts?

 

[mk3223]

IMO, it sounds that the pressure spike from the centrifugal compressor was resulted from the operation of the gas distribution system. An air receiver or storage tank may be needed to compensates for peak demand of the system and help balance the supply of the compressor.

 

[georgeverghese]

Discharge pressure pulsations occur on recip compressors and not on centrifugals. If a PSV on a centrifugal is popping, a dampener here wont help. Reasons for this PSV popping may be one or more of the following

a) The max operating discharge pressure is too close to the setpoint of the PSV
b) Some mechanical defect on the PSV or poor calibration of the PSV
c) The setting of the recycle PIC setpoint on this compressor has been altered recently or this PIC loop is too slow to handle the discharge pressure swings.

 

[The Obturator]

Can be any of the situations already mentioned. Is it possible to post the specification sheet of the PSV? This will allow more review.

Per ISO, only the term Safety Valve is used regardless of application or design.

 

[georgeverghese]

The current design probably expects too much from the response behaviour of the capacity controls PIC. If there are large swings in gas demand downstream, this PIC, not matter how well it is tuned, simply may not be able to cope with the flow - pressure swings. In this case, a surge drum on the compressor discharge will help to slow down the pressure swings and give the PIC more time to respond. Talk to a senior process engineer who can assist with deriving a suitable size for this drum. Perhaps this is what you meant by "pulsation dampener" on the compressor discharge.

 

[dcasto]

use a pilot operated set 10% higher than the desired MAOP if you cant tune the compressor surge controls better.

 

[The Obturator]

A Pilot Operated PRV may well be the best solution. I know of one major PRV manufacturer that has an accessory offering of a "Pressure Spike Snubber" developed specifically for compressible fluid compressor service. You should be able to use this having some 5% differential between MAOP and set pressure. Since the POPRV has a sensing line, in may also be worth your while consulting the supplier where the best pick up point should be. I believe you have a small conventional spring operated PRV that is not designed for this service. There are other spring operated PRV's available that may help here too, those with very small op/set differentials with low blowdown.

Per ISO, only the term Safety Valve is used regardless of application or design.

 

[Bill3752]

Thanks for the input. Here are some responses:

mk3223 - This RV is on the second stage of a syngas compressor system. There is no intermediate separator. I don't think a full sized tank is in the cards; that is why I asked about the dampener vessel.

George - This is indeed a centrifugal compressor, 2nd stage. The operating pressure runs about 90% of the RV set point. The RV is a POSRV assumed to be able to handle 95% operating ratio. We believe the relief valve is mechanically sound (actually we have rotated a spare). Once running, the system doesn't experience large operating extremes, but your question on the PIC control is valid, since the recycle is done from the final stage. Yes, the dampener was meant to include consideration of a (smaller) tank.

avalveman and dcasto - The relief valve is an Anderson Greenwood 25312 POSRV. I am checking to see if a snubber is installed; if not, this is worth pursuing.

 

[pennpiper]

Bill3752,
Have you given any thought to adding a Rupture Disc under the PSV Inlet?

Sometimes its possible to do all the right things and still get bad results

 

[LittleInch]

I think you're trying to solve the symptom not the cause.

If this is the 2nd stage out of 3 then where is this pressure spike coming from that doesn't affect the other stages?

Can you put some numbers down here
Design pressure
Set pressure of the RV
Operating pressure of the system
Level of the pressure spike (if known)
Duration of event,
number per day?
Design code of the system

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

 

[georgeverghese]

If you believe the control engineers have tuned this 3rd stage (final) recycle PIC as best as they can, then one way to dampen the pressure spikes would be to add more control volume within the compressor system ie. increase 2nd and 3rd stage discharge scrubber volumes. Also find ways to smoothen out flow demands at the users downstream. Also check if the recycle PCV is sized to handle the high recycle flows expected when flow demand drops quickly at the downstream end. Another obvious thing to check would be whether you have sufficently stable suction pressure as compressor suction flows fluctuate.

 

[georgeverghese]

On the main discharge line going to the users, there should be a check valve just downstream of the recycle line tee off (tee branch which leads to the recycle PCV). If this main line check valve doesnt hold well during the recycle event, some of the pressurised downstream inventory would reverse flow into the recycle line and cause this discharge PSV to blow also.