Hello all,
I briefly started this discussion in another thread, but decided there is probably enough interest in aircon efficiency to justify a dedicated thread. My apologies to those who may already have read part of this in another thread.
I recently installed an inverter split system air conditioner, forcing me to study the efficiency of competing brands and models. What I found may be of interest. Aircon cooling efficiency is characterized by the EER (Energy Efficiency Ratio), which is the ratio of cooling power in kW, to the electrical power consumed in kW. For example, a typical value is 3.6, meaning that 3.6kW of cooling is produced for every kW of electrical input power. For reverse cycle heating, the equivalent ratio is known as the COP (Coefficient of Performance), which usually has a similar value. For simplicity I will refer only to the EER.
The most striking observation for me was that the smaller capacity units always have the highest EER ratings, which initially puzzled me, because in general, machines become more efficient as they are scaled up. As an extreme example, the most efficient inverter split-system aircon that I know of is the Mitsubishi Heavy Industries model SRK20ZJX, rated at 2kW cooling, and with an extraordinarily high EER of 5.7 If you lust for ultra-efficient cooling in a small bedroom or similar, then this unit (or the 2.5kW unit in the same series) would be worth a look, consuming just 350W at rated cooling.
Unfortunately, my need was for a largish (~7kW) unit and to my annoyance, I soon discovered that you cannot buy a larger capacity unit with a respectably high EER, so I reluctantly had to make do with an EER of 3.33, actually a Mitsibishi Electric MUZ-GE71. I even contemplated installing two smaller units for better efficiency, but rejected the idea because the installed cost would have been almost double. In general, the EER gets worse as the capacity increases, but why? Clearly it is technically possible to produce a large capacity unit with high EER, and I would have thought that a higher EER would be a strong selling point, so what is going on here? Being the curious engineering type, I had to investigate further. Below is a summary of the Mitsubishi Electric range of inverter split systems, which are typical of other manufacturers.
Capacity EER Dimensions (indoor unit)
2.5 kW 4.46 295x798x232
3.5 kW 3.47 295x798x232
4.2 kW 3.33 295x798x232
5.0 kW 3.05 295x798x232
Observe how the EER falls steadily as the capacity increases from 2.5kW to 5.0kW, and not by just a little bit either, but from 4.46 down to a dismal 3.05. Note also, that for reasons of manufacturing standardization and economy, the physical size of the indoor unit stays the same. The efficiency of an air conditioner depends largely on the size of the indoor heat exchanger relative to the cooling capacity, so it makes sense that the EER should fall as observed. The indoor heat exchanger is generously sized for the 2.5kW unit, but seriously undersized for the 5.0kW unit, so the efficiency suffers. As an engineering purist I find this a case of obscenely bad design, and if I had my way we would ban any aircon unit with an EER of less than 4.0. Consumers are partly to blame here as well, for preferring that everything be as small and compact (and therefore inefficient!!) as possible.
OK, if this theory is true, then the EER should rise again for the 6kW unit, which uses a physically larger indoor unit. Here are the next three capacities in the series:
6.0kW 3.40 325x1100x238
7.1kW 3.33 325x1100x238 (I own this one)
8.0 kW 3.12 325x1100x238
Yes indeed, the EER jumps up to 3.40, and then falls again as the capacity increases but the indoor unit stays the same. Note however, that the EER does not increase all that much, consistent with a modest increase in indoor unit length from ~800mm to 1100mm. If the length of the indoor unit scaled with cooling capacity, then the 8kW unit would need to be (8.0/2.5)x798 = 2550mm long to achieve an EER of 4.46, which would be difficult to mount, and flatly rejected by consumers as too big, ugly and expensive. The result is that we are forced to accept crap efficiency for the larger capacity units. Maybe we get what we deserve. Personally I would happily pay extra to be able to buy a 7kW unit with an EER of 4.0 or better, and I would happily put up with the larger size of indoor (and outdoor) unit that the higher efficiency requires. I would give my left, um, arm, to be able to buy a 7kW unit with an EER of 5.7, like the smallest of the Mitsubish Heavy Industry units. 
I have studied this trend for all major brands, and they are all more-or-less the same. The ‘star ratings’ reflect all that has been said, with the smaller capacities scoring the highest number of stars.
In practice, the low EER and COP of the higher capacity units may be partially offset by the fact that a larger unit may more frequently operate under partial load, and (for inverter units) the efficiency is generally higher under partial load. For example, a 7.4kW model manufactured by LG has a rated EER of 3.16 (yuk!), but under partial load, EER is given as 3.64. That makes me feel a bit better, for my 7.1kW unit rarely operates at full power, so the rated EER of 3.33 may be closer to 3.8 for most of the time.
