Understanding pressure-related failures and how thoughtful design prevents dangerous surprises.

Compressors for refrigeration plants come in all shapes and sizes (and colors). My experience has mostly been with large reciprocating and screw compressors so I feel like I have a good idea of what’s going on in there. Scrolls are mysterious, but I can rationalize them as two-dimensional screw compressors. Centrifugal compressors are a wonder. I know they work, but I could never have figured them out for myself.

The rest of this column considers a few aspects of positive displacement compressors, meaning those that work by reducing the volume of a trapped quantity of gas. There are two key numbers that describe the performance of a positive displacement compressor: the isentropic efficiency and the volumetric efficiency.

When we talk about “the efficiency” of a compressor it is important to know which one is being discussed. They are as different as “miles” and “gallons”—both contribute to the miles per gallon figure for your car, but it doesn’t make sense to say how many gallons you drive to work each day or how many miles you bought at the gas station.

The isentropic efficiency is a measure of how well the compressor performs compared to an ideal machine and is defined as the work done in an ideal machine (when there is no increase in entropy) divided by the work done in the actual machine (where irreversible losses such as friction and turbulence cause the entropy to increase).

This definition of ideal divided by actual is not really an efficiency in the traditional sense of output divided by input, but it has been created to give the conventional understanding that higher numbers are good and lower numbers are bad. This masks the fact that lower efficiencies by this definition are not linear: a compressor with a 75% efficiency will consume two thirds as much energy as one with a 50% efficiency. It also means that, if an alternative type of compression is used, for example isothermal compression where heat can be transferred to and from the compression process, the work input can be less than would be required for the isentropic process and so the “efficiency” would appear to be more than 100%.

Volumetric efficiency is simply the ratio of the amount of gas delivered to the high pressure side of the system (expressed as volume at the suction condition of the compressor) compared to the theoretical swept volume (again at suction conditions). Strictly speaking, it is actually the mass of refrigerant delivered divided by the mass that would be contained in the ideal suction volume at suction conditions. Some clever compressor designs create a ramjet effect in the chamber inlet which increases the mass flow above the ideal maximum, so again the efficiency can appear to be more than 100%.

The volumetric efficiency of a reciprocating compressor is primarily influenced by the pressure ratio affecting the re-expansion of the clearance volume and so is a function of pressure ratio, whereas in a twin screw compressor it is primarily influenced by leakage across the tip seals, over the end of the rotors and through the “blowhole” (the gap between the rotors). These are all functions of the pressure difference. Conversely for a reciprocal, the frictional losses in bearings, valves and piston rings are a function of pressure difference whereas in a twin screw the isentropic efficiency is primarily affected by over or under-compression, so is a function of pressure ratio. Efficiency comparisons between different models of compressor should use pressure ratio as the basis for reciprocal volumetric and screw isentropic but should use pressure difference as the basis for recip isentropic and screw volumetric. Comparing screws and reciprocals across a range of operating conditions is notoriously difficult.