SAMPLING & ANALYSISPurity and contamination testing to AHRI Standard 700. 7 characteristics, measured against the row of the table that belongs to the refrigerant you actually sent — not against one generic column applied to everything.
01What is measured
The basis matters as much as the number, so it is reported with it, and the determination named on each line is the one the standard defines for that characteristic. What each line tends to mean about the system it came out of is our reading, not the standard's.
01
vapor phaseBasis % by volume @ 25.0 °C
Determination Gas chromatography — separate vapor-phase sample
Air. It comes in through a low-side leak, a recovery machine left open between jobs, or a cylinder nobody evacuated. It is also the one characteristic that cannot be read off the liquid sample, which is why it needs its own draw.
02
liquid phaseBasis ppm by weight
Determination Karl Fischer titration
Moisture, and therefore acid before long — water hydrolyses with the refrigerant and with the oil. A water result over the limit usually predicts the acidity result on the line below it. A drier that has already changed color has told you half of this for free.
03
liquid phaseBasis % by weight
Determination Gas chromatography
Something present that is not the labeled refrigerant, most often another refrigerant. Recovery cylinders that were never dedicated to one fluid are the usual source, and the resulting mixture is worth far less than either fluid alone.
04
liquid phaseBasis % by volume or % by weight
Determination Gravimetric — the basis used must be recorded
Oil that traveled with the charge — typically off a recovery machine, or from pulling liquid out of a flooded system. It is measured gravimetrically, and the basis used is recorded because the standard permits two.
05
liquid phaseBasis pass/fail
Determination Visual
Metal fines, desiccant dust, rust off the inside of the cylinder. A visual determination against the standard's own wording, reported as that wording rather than as a number we invented.
06
liquid phaseBasis ppm by weight as HCl
Determination Titration — detection limit 0.1 ppm on a 50–60 g sample
Acid, reported as HCl. After a hermetic motor burnout this is the line that decides whether a recovered charge is worth reclaiming or is scrap, and it is the line a buyer will ask about first.
07
liquid phaseBasis pass/fail
Determination Turbidity — turbidity appears at roughly 3 ppm by weight
Chloride, as turbidity. It tends to appear alongside acidity when a chlorinated fluid has broken down, and it is a pass/fail determination — the standard does not define a reportable number for it.
2 of the 7 are pass/fail by definition and the rest are numbers. Note what is noton the list: there is no “purity, minimum” characteristic in AHRI 700. The figure everyone quotes is the arithmetic complement of the 0.5% by weight cap on other volatile impurities, and for a blend it is the wrong test entirely.
02Liquid draw, vapor draw
This is the part most sampling instructions skip, and it is the part that decides whether your non-condensables number means anything at all.
Draw 6 of 7 from the liquid valve
Water, All other volatile impurities, High boiling residue, Particulates and solids, Acidity, Chloride. All of them are liquid-phase determinations, which is why the procedure specifies the Liquid valve and a fill of 75–80%.
There is a second reason, and it matters more than the first: a zeotropic blend does not carry the same composition in the vapor as it does in the liquid. Draw vapor off a blend and you will measure a composition the cylinder does not contain, then fail it against a window it never left.
One characteristic needs its own sample
Basis: % by volume @ 25.0 °C. Determination: Gas chromatography — separate vapor-phase sample. Because the basis is defined on the vapor at a stated temperature, the source temperature at the moment of draw is part of the record — send it with the sample, and we will report it with the result. A non-condensables figure with no source temperature behind it is a number, not a measurement.
It is also the one line a laboratory can quietly not run. If a report shows a non-condensables result but the sampling instructions only ever asked for a liquid draw, the two statements do not fit together.
Our own procedure allows it: “For low-pressure refrigerant, the source tank may be pressurized through the Vapor valve with dry nitrogen.” Pushing a sample with nitrogen sounds like the last thing you would do before an air test — and it would be, except that the fluids needing that treatment are the same fluids AHRI 700 exempts from the non-condensables characteristic, because their normal boiling point is at or above room temperature: R-11, R-113, R-123, R-141b, R-245fa, R-1233zd(E). For those, non-condensables are reported as Not Applicable — never as a pass, never as a blank.
The corollary is the part to take away: never use nitrogen to move a high-pressure sample whose air result you intend to rely on. And check the edition — AHRI 700-2024 exempts 3 fluids the federally incorporated appendix does not (R-1336mzz(Z), R-1336mzz(E), R-514A), so the same sample can be Not Applicable under one edition and testable under another.
03Why the fluid changes the answer
Of the 22 single-component refrigerants in the published table, 7 carry at least one limit that differs from the set most reports print as universal.
| Refrigerant | Waterppm by weight | Acidityppm by weight as HCl | All other volatile impurities% by weight | Air and other non-condensables% by volume @ 25.0 °C |
|---|---|---|---|---|
| R-11 | 20 — differs from the common set | 1 | 0.5 | Not applicable |
| R-113 | 20 — differs from the common set | 1 | 0.5 | Not applicable |
| R-123 | 20 — differs from the common set | 1 | 0.5 | Not applicable |
| R-141b | 100 — differs from the common set | 1 | 0.9 — differs from the common set | Not applicable |
| R-142b | 15 — differs from the common set | 3 — differs from the common set | 0.5 | 2.0 |
| R-245fa | 20 — differs from the common set | 1 | 0.5 | Not applicable |
| R-1233zd(E) | 20 — differs from the common set | 1 | 0.5 | Not applicable |
Read the R-11 and R-123 rows in particular: a larger water allowance than the common set, and no non-condensables test at all. Hold a low-pressure charge to the common water limit and you fail material the standard passes; report a pass on non-condensables for those fluids and you have reported a test the standard never asked for. One generic limit column produces both mistakes.
Every blend shares the same contaminant limits. What decides the result is a two-sided window on each named component, plus the 0.5% by weight cap on anything that is not one of them. A blend can be 99.5% pure and fail — and a purity figure on its own cannot show you why.
The two below are both nominally 50/50 and both windows are skewed, in opposite directions. Anything implemented as nominal ± tolerance is wrong for both of them, which is why we store and publish explicit minima and maxima for all 25 blends.
| Component | Nominal | Allowable |
|---|---|---|
| R-32 | 50.0 | 48.5–50.5 |
| R-125 | 50.0 | 49.5–51.5 |
% by weight
| Component | Nominal | Allowable |
|---|---|---|
| R-125 | 50.0 | 49.5–51.5 |
| R-143a | 50.0 | 48.5–50.5 |
% by weight
04The certification
Amalina Technologies is an AHRI Certified® Refrigerant Testing Laboratory, and the only one in California. This service is what that certification covers.
AHRI publishes the participating laboratories in a directory that anyone can search. We do not print certificate numbers on this site; a number on a web page proves nothing about whether it is still current, and the register does.
Search the directory (opens in new tab)This service, and this service only. The program has no oil scope, so nothing on our oil analysis pages claims it. If you need a certified result, it is this service that carries one.
40 CFR 82 subpart F, Appendix A is the text incorporated federally; AHRI 700-2019 and AHRI 700-2024 are not identical to it or to each other. Say which one your obligation runs against in the notes when you register the sample, and the result will be read against that one.
Disclosure: we share common ownership with a refrigerant reclaimer in the adjacent unit of this building, and the purity we report on a batch is an input to what that company pays for it. That is exactly the kind of arrangement a customer should be told about rather than left to discover. How the laboratory is kept separate.
05Who sends us samples
Whether this batch can go out the door. The limits are the ones incorporated federally for Section 608 work — 40 CFR 82 subpart F, Appendix A — so the laboratory behind the number ends up in the batch record rather than in a file nobody reopens.
Lines that decide it
Every line, plus composition on every blend.
Whether a drum of recovered material has value or is a disposal cost. A test settles it before the drum moves; skipping it means finding out once the material has already been shipped, and moving the same drum twice.
Lines that decide it
All other volatile impurities first — mixed cylinders are the usual finding — then water and acidity.
Whether the charge sitting in recovery drums can go back into a machine, and whether the clean-up is finished. The refrigerant answers the first half; the oil sample answers the second, and no refrigerant result substitutes for it.
Lines that decide it
Acidity, chloride, water, high boiling residue.
06What comes back
Written to be read by the person who took the sample and by whoever asks them about it afterwards.
A result exactly at the limit passes. We publish that rule, and the rest of them, in advance — the time to agree how a close result will be read is before the sample is drawn, not after the number arrives.
Before you send anything
All 22 single-component fluids and 25 blends, with the decision rules that govern a close result. This is the bar your sample will be held to, published before you pay for anything.
Our report format, reproduced on an illustrative sample and annotated line by line, so nothing on the document that comes back needs explaining over the phone.
The two services compared side by side, keyed to what you are actually looking at on the machine.
07Your part
The sample is the half of this we cannot do for you, and a bad one produces a confident number about something that was never in the machine. The full procedure is published — this is the shape of it.
Log the fluid, the system it came from and where the report goes, before the cylinder travels. It arrives already identified, and you hold your own record of what you sent.
The laboratory-supplied cylinder is the standard route, and it comes back to us with the sample inside. A cylinder of your own is accepted only if it is rated to at least 400 psi, evacuated to 50 microns or better, and dedicated to sampling — a cylinder of unknown history carries whatever was in it last, and that is declared under Notes when you register.
Or better. Anything left in the cylinder is something we will find and report against your material — most obviously as air, which is the line evacuation exists to protect.
Liquid, not vapor. 6 of the 7 characteristics are liquid-phase determinations, and a blend sampled as vapor will report a composition the cylinder does not contain.
About 350 g. Enough material to run every characteristic, and enough headspace that the cylinder is not liquid-full when it warms up in a van.
At the moment of the draw, and send it with the sample. If the non-condensables result matters to you, that figure and a separate vapor draw are what make it a measurement.
Refrigerant analysis
Registered before it travels, judged against the published limits for the refrigerant you actually sent, and reported line by line with the basis of every number on it.