Answers, including the ones that are no

AHRI Standard 700 is the specification a fluorocarbon refrigerant has to meet to be sold as reclaimed. It is the document behind the phrase “meets AHRI 700”, and a verification analysis is judged against it.

It specifies 7 characteristics: air and other non-condensables, water, all other volatile impurities, high boiling residue, particulates and solids, acidity, chloride. 6 of the 7 are liquid-phase determinations.

Air and other non-condensables is the single vapour-phase measurement, and it is specified on a basis of % by volume @ 25.0 °C — which is why it needs its own vapour draw rather than being read off the liquid sample.

What the standard does not contain is a purity minimum. The characteristic is all other volatile impurities, maximum 0.5% by weight. “99.5% minimum” is the arithmetic complement of that figure, and it is only meaningful for a single-component refrigerant.

Limits are per refrigerant rather than one generic column. R-123 takes 20 ppm water and R-11 takes 20, where R-22 takes 10, and neither of the first two is subject to the non-condensables determination at all — the standard does not require it for fluids whose normal boiling point sits at or above room temperature. A single generic column applied to everything fails conforming material and passes non-conforming material, in both directions.

3 editions are in circulation — 40 CFR 82 subpart F, Appendix A; AHRI 700-2019; AHRI 700-2024 — and they do not agree on everything, so which edition a result was judged against is part of the result.

Because purity is not the test for a blend. Each named component carries its own two-sided composition window, and the verdict is whether every component sits inside its window — plus a cap of 0.5% by weight on anything that is not a named component.

Worked through: R-410A allows R-32 from 48.5 to 50.5% by weight. A cylinder assaying 47.9% R-32 is 99.5% pure and fails, because the composition is outside the window and purity was never the question.

The windows are not symmetric tolerances either. R-410A and R-507A are both nominally 50/50 and both are skewed, in opposite directions: R-410A allows R-32 from 48.5 to 50.5 while R-507A allows R-143a from 48.5 to 50.5. Anything implemented as a nominal value plus or minus a tolerance is wrong for both.

Physically, the usual causes are fractionation during recovery, topping off with something that is not the same blend, and a leak that removes one component preferentially. None of those move a purity figure much. All of them move the composition, which is what decides the result.

About 350 g of refrigerant, drawn from the Liquid valve into a sample cylinder rated to 400 psi, evacuated to 50 microns or better and filled to 75–80% of capacity.

The fill window is not arbitrary in either direction. Under-fill and there is not enough material left once the determinations that consume sample have run. Over-fill and you have a liquid-full cylinder with nowhere to expand, which is a safety problem before it is an analytical one.

For low-pressure refrigerant, the source tank may be pressurised through the Vapor valve with dry nitrogen.

Oil goes into a borosilicate graduated glass sample bottle, degassed before it is sealed. POE and Solest oils absorb moisture quickly, even in a sealed bottle. Send samples promptly.

Because a blend in a cylinder is two phases with two different compositions. The more volatile component is over-represented in the vapour, so a vapour draw reports a composition that is not the composition of the material in the tank — and for a blend, composition is the verdict.

That is why the procedure says to draw from the Liquid valve, and it is why 6 of the 7 characteristics in the standard are liquid-phase determinations. Moisture, high-boiling residue and acidity all sit in the liquid; a vapour draw leaves them behind in your tank and reports a cleaner sample than you have.

The non-condensables measurement is defined at 25.0 °C on a vapour-phase sample, so the source temperature at the time of sampling is part of the record.

A single-component refrigerant does not fractionate the same way, but the procedure does not change for it. The contaminant determinations are still liquid-phase, so the sample is still drawn liquid.

Yes. Say so when you register it.

It changes nothing about the limits — AHRI 700 has no burnout column, and the material is judged against the same figures as any other sample of that refrigerant. What it changes is what the result is for. On the refrigerant side, acidity and chloride are where a burnout shows itself. On the oil side, a burnout is the case condition analysis was built for, and the oil will usually tell you more than the refrigerant does.

One thing worth stating at registration: whether you are asking if the material meets reclaim specification, or asking whether the charge is fit to stay in the machine. Those are different questions. AHRI 700 answers the first one.

Yes, and on a system in trouble it is worth doing. They are two services, two samples, two kits and two reports — refrigerant into a pressure-rated cylinder drawn liquid, oil into a glass bottle — but they are two views of one event. Acid in the oil and acid in the refrigerant are the same chemistry seen twice, and moisture shows in both.

Register them against the same system and the same job so they reach the bench as a pair rather than as two unrelated samples.

One scope note, because it matters more than it sounds: only the refrigerant testing sits under the AHRI Certified® Refrigerant Testing Laboratory certification. The oil analysis does not, and we do not present it as though it did.

Either. Samples reach the laboratory by carrier or by hand, at 13564 E. Imperial Hwy, Unit D, Santa Fe Springs, CA 90670.

Hand delivery is what being in Santa Fe Springs is worth to a contractor working the basin: a cylinder drawn in the morning can be at the laboratory the same day, rather than in a carrier's network.

If you are driving one over, ring the laboratory first on (562) 404-9955 so somebody is expecting it. Hours are not published on this site.

One caution for anyone working from older paperwork: check the address printed on the form in your truck against the one above before you set off. Earlier Amalina documents give a different building.