1 reply to this topic

[SawsanAli311]

Dear All, 

 

I have the following design queries on reciprocating compressors:

1- Why numerous vendors tend to consider adiabatic efficiency over polytropic efficiency for reciprocating compressors, is it driven by the almost constant gas compressibility across the various stages in comparison to the multi-stage impeller centrifugal machines. 

2- In existing 4 stage compressor, I was surprised to see that vendor decided to set the discharge PSV at every stage's outlet to be less than the MAWP of the cylinders and the piping system. This reduction is not understood nor clarified by vendor. Additionally, increasing the capacity of the existing compressor is now deemed difficult since vendor stated that capacity increase can only be achieved by reducing the cylinder size (by adding liner) of the 1st and 2nd stage (due to motor load) while increasing cylinder size at the 3rd and 4th stage. 

3- Based on point 2, I am now questioning whether the design of reciprocating machine shall cater for future expansion increase in capacity and hence to adapt identical cylinder sizes across all stages although the normal design practice is to have the cylinder size of the 1st stages being progressively getting smaller for the subsequent stages as the pressure increases and gas volumetric rate decreases for the same throughput of gas. I understand this overdesign will come at the expense of more power, rod load and clearance pocket controls. 

 

Appreciate your advice based on your valuable experiences.

 

Regards,

 

 

[Bobby Strain]

In all my 50+ years as a practicing chemical engineer I never specified a reciprocating compressor for future debottlenecking. And this covers many different clients. It's usually difficult to control budgets without troublesome requirements. As to using adiabatic efficiency for reciprocating compressors, it closely describes the behavior of gases in the cylinder. And polytropic compression more closely describes the behavior in dynamic compressors. One can relate one to the other in a dynamic machine.

 

Bobby