Capacity control in reciprocating compressors 2

Depending on the system the best capacity control solution may be none of the above. My experience is limited to upstream oil & gas below 2000 hp, but for online capacity control I usually rely on varying engine speed and opening the bypass valve. On larger machines, I have used hydraulic VVCP's as a last resort (this wouldn't happen to be a Dresser HOS or VIP, is it?), but I prefer to use VVCP's as a last resort. They're a pretty slick system, but are really intended for getting around engine loading problems by reducing flowrate. Since they work by adding volumetric inefficiency to the head-end, as a result you can sometimes see increased discharge temperatures and reduced valve life. You also have to make sure you aren't going to cause rod loading (static & dynamic), reversal or interstage pressure problems.

I've never used finger-unloaders for activating valves for part of the stroke. My understanding is they're just a matter of activating or deactivating valves during operation for prolonged periods of time. Virtually none of the machines I see here in this particular field use them, if we need to go to single acting operating, we purge the machine and either pull the valve or install spacer rings under them for added clearance. I'm not sure I'd trust a pneumatic system to load and unload valves for part of their stroke - besides the fact I'd be surprised if you could get any kind of reliable timing, that has to be hard on the valve opening & closing time, I would think you'd see a lot of cracked valves. You also have to consider the same rod loading, reversal and interstage problems as with the VVCP.

You really have to look at what you are trying to accomplish. We have a bunch of 2 stage Ajax machines here in the southern Rockies with poppet VVCP to allow the compressors to react to changing conditons. They say the same machine is good from 200-600 Hp with these devices. I pay about $500k (US) for a 600 Hp recip and less than $175k for a 200 Hp recip -- why would I burden an operation that needs a $175k compressor with debt service, depreciation, and operating costs on a $500k machine? Doen't sound like good economics or good engineering to me.

My approach to fleet management has been to have 4 different size compressors (each one about half the Hp of the next larger machine), fabricate all the compressors in the fleet with the same size suction and discharge flanges located in the same place on the skid, and require all the skids have self-contained fuel gas. I allow zero flexibility on these points, and I can do a two-compressor swap and have both machines running in the new location on the same day. The fleet is all flooded-screw compressors under 600 Hp. With this design scheme I maintain over 70% Hp utilization across the fleet and spend less than $200k/year on compressor moves (68 compressors in the fleet rated at a total of 24,000 Hp). The fleet operations and capital costs are far less this way than to try to build a rubber compressor.

The VVCP idea seems sloppy to me.

David

Thanks a lot for your kind answers, my question has been done because we are thinking to install a better capacity control than the one we have(which only allows steps of 25%)in three reciprocating compressors (Peter Brotherhood) that compress Hydrogen. The discharge pressure is aproximately 70 Kg/cm2 in 2 stages (4 cylinders).
The system that unloads the valve for only a portion of its stroke seems to be the worse in order to guarantee valve life, but I think that it's the one which is going to be installed.I don't believe so much in actuate over the valves and that's why i posted this question.

P.D. Sorry about my English, I'm not fluent using it
Regards
Marc

Ah, never worked with Peter Brotherhood machines, just curious as I installed self-actuating hydraulic VVCP's on a Dresser Rand HOS frame a couple years ago.

I'm also curious though, if I understand right, your plan is to use finger unloaders to unload one or more suction valves of each cylinder for a fraction of each stroke, the idea being it will allow part of the hydrogen in the cylinder to be displaced back to suction for part of the stroke, and the rest to discharge, thus reducing overall capacity.

I must be missing something though, are these really low speed machines? I can't help but wonder how fast your machine runs and how you plan to synchronize the finger unloader with the compressor stroke so it only unloads the suction valve(s) for part of the stroke? All the recip machines I see have speeds in the order of 600 to 1800 RPM, with time between valve opening and closing in the order of milliseconds. That's a pretty small span of time to try and cut in half even further, especially with a system as spongy as one that's pneumatic driven.

You are completely right Scipio, because we work with compressors that run at 200/350 rpm and the system would be hydraulic not pneumatic.
Although, I don't trust so much in this system, and I hope not to have so many problems in the future.

Many thanks for your help!
Marc