3 replies to this topic

[Roark]

Dear all,

  

I am trying to fully understand an NGL recovery process based on Ortloff's GSP.  An inlet stream of dry, compressed feed gas is cooled down in several heat exchangers and a chiller (which runs on liquid propane refrigerant).  After this chilled, mixed liquid+vapor stream is separated on a low temperature separator, the outlet liquid feeds the demethaniser, and the vapour stream feeds a turboexpander.  As usual on a GSP, a slipstream of the vapour is condensed and fed to the demethaniser as reflux.  The cold overhead stream of the demethaniser cools down several process streams and is compressed in a residue gas compressor coupled with the turboexpander, and leaves the plant at a constant battery limit pressure.

   

Please find attached a sketch whil illustrates this process. 

  

I do understand most of the process control scheme shown in the sketch.  However, I cannot see how the pressure in the column is controlled.  Please note there is no pressure controller actuating on this process variable.  My intuition is that it is achieved indirectly by throttling the discharge of the residue gas compressor.  For example, if the compressor is unable to reach the B/L pressure, the throttle valve will close, the compressor runs up its curve increasing the discharge pressure (and decreasing the flowrate) and consequently increasing the column pressure.  Throttling PCV-2492 at the discharge of the compressor reduces the power requirements of the compressor.

  

Alas, as a result of this pressure increase in the column, the pressure drop in the expander is reduced, hence the power extracted by the EXP will also be lower, actually coupling both rotating machines (expander + compressor).

    

However, my main doubt is regarding the point highlighted in green.  Intuitively, if I throttle the discharge of the compressor its power requirements will go up, it's highly counter intuitive for me that closing a valve at the outlet of a compressor results in lower energy consumption.

   

I have been trying to learn about control of this process in many sources (GPSA, Ortloff's papers, Manning and Thompson's book...) but to no avail.  Any help or input from experienced engineers familiar with this sort of NGL recovery processes would be appreciated.

  

Regards

Attached Files

•  gsp.png   430.15KB   3 downloads

Edited by Roark, 28 January 2016 - 02:21 PM.

[Bobby Strain]

You should look for controls that regulate compressor inlet pressure. Apply same, but control the column pressure. Controls will vary depending on the compressor and driver. But application to column pressure control should be obvious to you. And, you should be able to find better examples than the one you have shown us.

 

Bobby

[Roark]

Hello Bobby,

  

Thanks for your reply.  Yes, controlling the pressure of the column via the inlet pressure of the compressor seems like a straightforward control method, probably the first thing that I could imagine.  However, the sketch shown is the actual process I am trying to fully figure out.  It's not that I aim to optimise or improve it: it's a proven NGL recovery process (firstly licenced by Ortloff but now it's open art) which I am trying to fully understand.

   

By the way, I have just realised that it may not be obvious from the sketch that the Turboexpander (TEX) and the residue gas compressor (RGC) are coupled, both impellers are mounted on the same shaft hence the work extracted by the expander (less some 1% friction losses) drives the residue gas compressor.

  

Regards

[Bobby Strain]

The overall plant controls can be quite complicated. There is typically a residue gas compressor to further boost the gas pressure. This is due to the limited upper pressure range that the NGL recovery column can operate. You must set the constraints to be met before you can implement controls for the plant. Including the column pressure control. In cases where the plant must process whatever inlet flow arrives, then the column pressure can be controlled by the expander capacity control, which is typically speed control. Residue gas flow/pressure control will then be manipulated to maintain the cold separator pressure. So, the answer to your query is "it depends....."

 

Bobby