[ICanHas]

I have credible reference here that propane actually performs WORSE than R-22. You can not credit refrigerant as being inherently more efficient when other highly influential factors changed.

R-290 propane has different characteristics than R-22 and has a lower volumetric capacity. Common test method is based on dump gas and go method that free loads on efficiency gain attributable to reducing load on coils and compressor by reducing capacity.

In real life, you want the motor as close to fully loaded as possible for lowest production cost, but 75-80% load results in increased motor efficiency. So, using the same hp, higher displacement compressor effectively reduces the magnitude of impact.

To say R-290 is more efficient is like saying you get better mileage in 4th gear than 5th gear at the same RPM, because 4th gear in the gearbox is superior efficiency. This neglects the impact of the fact that road speed dropped here. The first paper makes a compensation comparable to changing the final drive so that 4th gear at same RPM provides the same road speed as 5th gear with original drive ratio.

According to Hwang, Gado and Radermacher as published in January 2003 ASHRAE journal, R-290 performs WORSE than R-290. A reduction of 3-6% in capacity, 5% reduction in steady state efficiency and a lower SEER value. They concluded that the poor performance applied to heating cycle as well until the outdoor temperature was below 17F. In this experiment, adjustments were made at the compressor and expansion device so that the system closely matches the original capacity of R-22 while using R-290.

Don't listen to turd polished BBQ gas sales people. Do your own research if you disagree with the above findings.

On two stage units, you'll find that EER on first stage at 67% the rated capacity is generally better than on 2nd stage at 100% capacity. Equipment manufacturers have raised the SEER on units by increasing the evaporator and capacitor size to BTU capacity ratio.

Any gains made due to changing coils to BTU size under same usage conditions can not be attributed as refrigerant being superior.

Also, any attribution of efficiency gain to using R290, its commercial equivalents such as Coolant Express 22a, Envirosafe ES22a or the like products that exploits the fact that existing system is oversized and giving the product the credit for gains made through increasing the run-time using a reduced system size is cheating.

In a paper by CHINNARAJ, GOVINDARAJAN and VIJAYAN published by THERMAL SCIENCE, Year 2011, Vol. 15, Suppl. 2, pp. S383-S390 suggests the above is true. They tweaked some hack chopped jury rigged window shaker and optimized charge levels for each of the three refrigerants they used, R22, R290 and R407C but used the same heat exchangers originally equipped with the unit.

I normalized their findings so that R-22 is referenced to 1.
The paper's data shows
R-290 shows a 11% reduction in capacity and a 4% rise in COP.

The reduction in capacity means the condenser and evaporator are both unloaded compared to R-22 which gave R-290 an unfair leverage. The test discussed in first paper took this away and found reduction in COP compared to R-22.

I've attached both papers for your reference.