Centrifugal compressor recycle cooler sizing 1

[Justice100]

Hello all,

I have a 3 stage variable speed centrifugal compressor, compressing natural gas. The first stage has a recycle cooler and discharge cooler and the 2nd stage has a discharge cooler. The compressor discharges into a manifold fed from another set of compression trains.

I am trying to come up with a basis for sizing the recycle cooler. My thoughts were to simulate the compressor (with compressor curves included) at full recycle with the discharge pressure at the operating manifold pressure to replicate a start-up, where the speed is varied to achieve the discharge pressure. The remaining variable is then the recycle flowrate that I am now unsure how to choose. A recycle flow rate for the 1st stage, using the design flowrate would appear conservative as it should not be operated at design flow in recycle. Alternatively, a lower flowrate could be used with a good margin above the surge point.

Does anyone have an opinion on this or point me to a good reference?

 

[rotw]

I would say that the flow at control line should be - in principle - the basis for sizing on a single stage machine. Three process stages are considered here, those are connected thermodynamically and mechanically (in terms of single shaft speed) - as I suppose. I think you would probably need to check that there is no additional flow requirement (i.e. beyond the control line flow) with the 1 stage in full recycle, in order to satisfy some interstage pressure or a certain process requirement otherwise. That requirement would be imposed considering how all the stages behave. An example of interstage pressure constraint, consider a stream injection/extraction at compressor stage discharge; this is a case where a maximum/minimum pressure need to be maintained in order to make any injection/extraction (respectively) physically feasible.
So in short, control line as baseline plus beware the fact that all the stages are mechanically connected which may thermodynamically constraint the 1st stage (ending up with extra flow recycle).

 

[Justice100]

Thanks for your feedback, much appreciated. For this case the only process constraint I can see is the 3rd stage discharge pressure going into the manifold, this could of course vary somewhat. If I take the design flow it does give operational flexibility so that they can run in recycle at higher flows, not sure why anyone would want to as it is just a waste of energy (both compression and cooling), testing purposes perhaps? I must admit I am not too familiar with a cooler located on the recycle and surprisingly none of the design guides I have seem to discuss it.

 

[rotw]

Ok. Do you happen to have the specs / datasheet of the recycle/anti-surge valve? Maybe you could use the anti-surge valve specs as basis for sizing?

Is the subject valve specified for anti-surge protection only or is it foreseen for "dual" purpose, means anti-surge protection AND continuous recycle ?

I am raising this point to trigger some thinking / discussion... if the valve is for anti-surge protection only, I would suspect cooler must be sized correspondingly so that the max calculated Cv among all operating cases ((surge line intersecting MCS & mos speed curves) is covered - as minimum. But I can't state this for sure. Hope someone can help and provide further direction here.

Also when looking at the valve selected Cv, this one is normally selected between 1.8 to 2.2 times the maximum calculated Cv. The multiplier in question (1.8 to 2.2) does not mean that the flow through the valve is expected to run higher than the compressor surge flow. It is more of a measure to have the valve dynamic response cope with any surge event thereby ensuring machine is continuously protected.

I am tented to say use max calculated Cv case as sizing case for the recycle cooler when anti-surge protection applies only, but frankly it can be misleading. Like I said at this point, do not take this as granted but instead as a way to trigger discussion.

 

[georgeverghese]

The recycle cooler is most likely installed to enable a suction temperature that is close to ( or equal to ) the normal stage 1 feed gas temp, during recycle operation. The sizing basis for this cooler will depend on whether this recycle line is meant for antisurge purposes or for capacity recycle (or may both).
If it is meant for antisurge, then size for surge flow x 1.1 at max operating speed. If it is meant for capacity recycle, then size for the flow seen at min operating speed x 1.1.
Obviously you would take credit for the J-T effect across the recycle valve to get to the required stage 1S feed gas temp. So another set of parameters that would be applicable is to use the lowest gas mol wt and lowest stage 1D pressure. The gas with lowest mol wt may in some cases be that which is produced when the compressor is in full recycle with no fresh feed gas ( gas lean out effect when compressor is in full recycle).
Am assuming this stage 1R cooler gets its feed from upstream of the stage 1D cooler ( ie directly from compressor discharge). So another parameter to include would be some deterioration in polytropic eff in the stage 1 compressor that would appear after several years of operation or if the gas fouls the compressor impellers / diffusers.
It will be a lot less painful to get these cases studied if you have a process simulation program where you could load up the speed - polytropic head - flow curves.

 

[georgeverghese]

Just to clarify, the min operating speed referred to earlier for capacity recycle is not the min possible operating speed as limited by the machine driver. It is the min operating speed at which a flow of 1.1 x surge flow is realised at the min operating final pressure on stage 3D, which will be the same as the discharge PIC setpoint at this location which controls / resets machine speed.
Mass flow through the machine ( and hence the recycle gas flow ) at stage 1 will be highest at max stage 1S pressure - this is another parameter to take into consideration also.

 

[Justice100]

Thanks for your input.

No specs/DS for AS valve at the moment, that will probably come after I am off the job although could certainly be considered at the next phase. The recycle will be only used for AS in normal operation although it is not inconceivable they will operate at reduced flow rates far in the future where it is used as a recycle but currently that is not the design intent.

I have the curves in a simulator so I can simulate it in full recycle with the 3rd stage at discharge pressure (minimum and maximum), 10% above surge flow at each stage and the compressor at maximum speed, I think that is a reasonable design basis.

 

[georgeverghese]

Yes, it may take the capacity recycle stage 3D max set pressure controller be several seconds later (than the antisurge controller on stage 1 ) to kick in and reduce speed. But check that this surge flow x 1.1 at max speed does not exceed the PSHH setting on stage 1D, else the flow margin may need to be higher than 1.1 x surge flow.