Since, a control scheme that changes displacement is really just changing the clearance (since the cylinder, and piston are fairly rigid), you are decreasing the effeciency (i.e., increasing the energy cost per unit volume) when you increase clearance.1500 CFM (I think of it as 2.16 MMCF/day) is not a small energy load if you are running with significant compression ratios.  All other things being equal, I'd go with the variable speed.
My point is that it would probably be better to get the process to run Magnetic pump constantly at 1000 cfm demand and optimize your compressor for that volume rather than accomodating your production line at whatever they happen to want to run that day.... as far as designing the compressor goes.  Whether that is possible or not, is a question you should settle with production.  It is ALWAYS more efficient to optimize your setup for running at one constant speed, be that pumping, compressing or building widgets.  
You have to define what your "process" is.  The OP's process is casting, not compressing air.  Most anything he can do to optimize the casting process has a chance of positive economics.  Optimizing an air compressor for optimum efficiency probably has marginal to negative economics. The air compressor is a tool of production like a chain fall or a fork lift.  Tools of production need to be optimized to never be an impediment to production.  So you oversize them.  You make them tougher than they need to be.  You optimize them toward the high end of process needs.  The OP is putting in an air compressor that has significant over capacity in normal operations.  His air-demand is too high for storing a significant reservoir of compressed air, so his compressor needs to be flexible.  If he can provide the same process need for flexibility with either variable speed or variable capacity then he can save a few bucks by going with variable speed.Even when a compressor (or pump or widget) is the process (like in a natural gas compressor station for example), it will rarely have a constant throughput requirement.  Suction pressure will fluctuate.  Discharge pressure will fluctuate.  Required volume moved will fluctuate.  You can design for constant speed as these things change, but (for compressors, I don't know much about widgets), that comes at a cost of significant throttling of the suction or discharge flow to maintain a constant speed requirement--I contend that taking a 50 psid drop across a suction controller and a 200 psid drop across a discharge controller as I saw in a compressor station recently is not more effecient than letting the speed match currently prevailing needs.  I'm not buying your always statement.

2011-08-29

Back