Clarifying boiling phenomena to improve evaporator performance and system reliability.

I was challenged recently to consider if I had to choose a single number to represent whether a refrigeration system was good, bad or indifferent, what would that single number be? At face value this is a daft question. Such systems are extremely complicated, and the circumstances surrounding their use vary so enormously that no single metric could cover all the variables that make one installation more or less appropriate than another.

However, this is a persistent question that keeps cropping up, so there is clearly an appetite for this single number approach. I was asked in the context of enabling the public to make more informed decisions about sustainable choices for refrigeration and heat pump systems and, in a sense, that’s fair. They shouldn’t be expected to understand all the technical nuances behind a decision, but obviously they want to “do the right thing.” So what options for appraisal are there?

Perhaps the earliest single number metric is coefficient of performance (COP, or sometimes in the U.S. energy efficiency ratio [EER]). I’m most familiar with the COP expressed as kW cooling (or heating) divided by kW electrical which gives a dimensionless ratio, usually in the range of one to 10 depending on the situation. It may, however, be expressed as ton/hp or more commonly (mostly in the U.S.) for chillers as hp/ton. I am usually quite good at mental arithmetic, so it’s surprising how difficult I find it to cope with a smaller number in hp/ton being better. I have the same trouble when converting from miles per gallon to liters per kilometer. COP has been very useful over the years but it has many shortcomings. It is an instantaneous value, so it bears no direct relation to the size of the energy bill. It is usually quoted for the design condition, which rarely happens.

The seasonal energy efficiency ratio (SEER) now commonly incorporated into efficiency legislation is more useful because it provides an indication of the likely size of the annual energy bill. Of course, it’s not a statement of what the bill will actually be because it’s based on typical weather history and a hypothetical use pattern but it ought to allow comparison between alternatives. A similar metric used for chillers is the integrated part load value (IPLV). IPLV applies a very crude load profile to chiller performance to give an indication of annual energy use, assuming a small percentage of the year at high load in maximum ambient, a significant proportion of the total at moderate load in moderate ambi-ents (at two load points) and a small percentage at low load in low ambient.

Both the SEER and IPLV metrics are valuable because they direct attention away from the design point to the condition at which most running hours occur but they still share some of the shortcomings of the COP metric. They are based on the “as-new” clean condition of the equipment and don’t reflect the effects of wear and tear, weathering or fouling. They also don’t account for the use pattern, so the SEER rating of a domestic air-conditioner doesn’t match the behavior of the occupants of the home and the IPLV of a chiller doesn’t work for a process load which is on 24/7 or on full load in low ambient.

Concerns about the ozone layer and global warming introduced a slew of additional single number metrics to the world. The ozone depletion potential gives a useful point of reference against the behavior of R-11 in the atmosphere, so it provides a “better or worse” overview but says nothing about what is acceptable or unacceptable. Global warming potential (GWP) does the same thing with carbon dioxide as the benchmark. The total equivalent warming impact made the valuable step of correlating GWP and COP to offer a single number metric which balanced the “direct” and “indirect” warming effects of a particular gas used in a particular system in a particular way. This required a heady mixture of scientific detail and guesswork; the former represented by the GWP and COP of the system and the latter by the number and severity of leaky joints and the embedded carbon factor used to calculate the effect of electricity use.

However, even the supposedly more scientific parameters were a bit woolly. For example GWP values are typically ±30% so while R-134a’s 100 year value might be quoted as 1,430 that means in reality somewhere between 1,000 and 1,860, not including the embedded carbon of the production process or the warming impact of the breakdown products in the atmosphere.

For several years now I have been studying and advocating the use of the specific energy consumption as a single number metric. I touched on this in the ASHRAE Journal in July and August 2019, but had already highlighted it in May 2012 when I was advocating “pies per kWh” as the preferred metric. I was actually suggesting that the ratio of “what you want to do,” for example, make pies, to “what you need to do,” for example, use electricity. This has several advantages over other metrics. It is flexible to suit the specific circumstances of the installation. It is adaptable to adjust performance to the condition of the plant as time goes by. It is readily understood and shines a light on changes in performance that might indicate that some maintenance is required to bring the system back to peak performance. Sadly, in the ten years since I first wrote about pies per kWh, I have not seen a single food manufacturing system that presents production throughput data to the refrigeration mechanics responsible for maintaining the system. It’s not difficult, but it’s also not happening.