Refrigerant Glide, and Why a Single Saturation Number Is Wrong

Superheat and subcooling are both differences between a measured line temperature and a saturation temperature looked up from pressure. That works cleanly for a single-component refrigerant, which boils at one temperature for a given pressure.

Zeotropic blends do not. They boil across a range, and the width of that range is called glide.

How wide

Wide enough to matter. For R-407C and R-448A the glide is around 10 °F; R-449A is close behind at about 9.8 °F. R-410A and R-32 have effectively none. R-454B, the A2L replacement now moving into comfort cooling, has about 2.5 °F.

Ten degrees is larger than most of the bands anyone judges a system against. Manufacturers specify subcooling in an 8-to-14-degree window. A 6-degree deviation from normal is a meaningful charge signal. Against numbers that size, a 10-degree lookup error is not noise — it is bigger than the thing you are trying to measure.

The two ends have names

The bubble point is where the first bubble of vapour appears — the liquid end. The dew point is where the last drop of liquid disappears — the vapour end. For a blend with glide these are ten degrees apart at the same pressure, and a generic pressure-temperature chart that gives you one number has quietly picked one for you.

The convention that makes the physics work is:

• Superheat from the dew point. Superheat asks how far past fully-vaporised the suction gas is, so the reference is where vaporisation completes.
• Subcooling from the bubble point. Subcooling asks how far below fully-condensed the liquid is, so the reference is where condensation completes.

Use one saturation figure for both and you put a systematic error into every reading, in opposite directions, on exactly the two numbers a charge diagnosis depends on.

Why it shows up in refrigeration first

Because that is where the high-glide blends live. R-448A, R-449A and R-407C are medium- and low-temperature refrigerants — coolers, freezers, display cases. Comfort cooling has largely been on R-410A, which has no glide, so the problem could be ignored. The A2L transition changes that: R-454B carries about 2.5 °F, small but no longer zero.

How we handle it

Every pressure-temperature table in our engine is generated with CoolProp, the open-source thermophysical property library, rather than transcribed from a chart — and superheat and subcooling are computed from the dew and bubble points respectively for all ten refrigerants we support.

There is an honest limitation to record alongside that. For R-454B, R-448A and R-449A our tables currently stop at 140 °F, which is below those refrigerants' own high-pressure cutouts. The pressure-based safety alerting still works above that point, so nothing dangerous is missed — but the derived diagnosis goes quiet, because condensing temperature and condenser split stop moving. We are fixing that by regenerating the tables, not by estimating past their edge.

This post describes how our own alert engine is built. Where it states a figure from published research the source is linked; where it describes our engineering decisions, that is what it is, and we have labelled it rather than dressing it up as a finding.