Exploring where CO₂ refrigeration is unsuitable, borderline, or clearly the best choice.

<p>My last two columns (April and May) considered the delicate and subtle web of constraints that surround the design and use of refrigeration systems using highly flammable refrigerants. Now it’s time to go right to the other end of the scale and look at one of the very few refrigerants that is really nonflammable. In this case it is carbon dioxide, which is actually the product of combustion rather than a fuel and is widely used as a fire extinguisher.</p><p>There are 46 substances listed in ISO 817’s table C.1, and 25 of them are in flammability class 1, which is popularly referred to as “nonflammable.” However, the correct name for this class is “no flame propagation (when tested in accordance with ASTM E681).” The carbon-fluorine bond is exceptionally strong but will still break and allow the carbon to be oxidized if subjected to sufficiently high temperature. Therefore, the last resort for disposing of old refrigerants that can no longer be used due to legislative constraints is to incinerate them. The only truly nonflammable refrigerant apart from carbon dioxide in table C.1 is R-704, although ASHRAE Standard 34 used to list several more, including R-718 (water) that I would consider nonflammable.</p><p>Several years ago, I gave a presentation on carbon dioxide as a possible refrigerant and spoke about “the A B C.” “A” represents situations where R-744 is absolutely not the right thing. “B” covers cases where it is borderline; in other words using R-744 is perfectly feasible, but alternative solutions would also be viable. C is the rare set of circumstances where you would be crazy not to use CO₂ as the refrigerant.</p><p>There are many reasons why CO₂ may not be the right thing. In domestic refrigerators it was necessary to move away from ozone-depleting R-12 as the preferred refrigerant. Since the mid-1990s, the use of isobutane has built a phenomenal track record of safety in that market segment, so although there have been many research projects still looking for alternatives, there is no technical or financial imperative to do so.</p><p>In larger systems, for example integral display cases for supermarkets, CO₂ is not an easy solution for technical reasons. If an air-cooled solution is required, it will need to operate on the edge of the supercritical region. The air temperature in the store will probably be around 70°F (21°C) but the heat exchanger is likely to be small, tucked away in a corner and not kept clean, so operating a transcritical system efficiently is difficult. If a water-cooled unit is used, then the refrigerant charge can be low enough to use isobutane and it’s the same as the domestic situation; why would you?</p><p>Bottle coolers for retail are slightly different. The standard design needs to cope with a very wide range of operating conditions because the unit might be indoors in a controlled environment or it could be in the entrance hall or even out in the street and subject to both high and low ambients.</p><p>Split air-conditioners seem to me to be another case of “absolutely not.” In this instance the difficulty of providing both cooling and heating from a single unit makes CO₂ unattractive and while it might be technically possible to engineer a suitable unit, it is unlikely to be economically viable to do so.</p><p>I’m sure you can think of many more As. Next month we will explore the borderline Bs.</p>