// Sampling / Refrigeration oil
Two decisions are made before the bottle is opened — which point, and in what running condition — and they matter as much as anything that happens afterwards. Then 7 steps, each with the reason attached.
POE and Solest oils absorb moisture quickly, even in a sealed bottle. Send samples promptly.
01Where the sample comes from
Oil analysis is a trend measurement. A number from a different point is not a worse number — it is a number about something else, and it will look like a change in the machine.
What a good point has
What a dead leg is, and why it lies
A dead leg is any volume the oil does not pass through in normal operation: a capped drain, a gauge line, an isolated port, the far side of a valve that is closed most of the year. The oil in it is a preserved specimen from whenever that leg was last flushed.
It has spent the interval in a small metal volume, in contact with whatever the fitting admits, so it tends to read older than the charge does — and the debris that settled out of it left long ago. Sampled without purging, it produces a result that is internally consistent, plausible, and about a fitting.
The fix is one actionRun oil to waste until the leg is cleared, then fill. That single step is what turns a reading of the port into a sample of the machine.
02Hot or cold
Not a worse one, necessarily — but a different one, and mixing the two conditions inside a single trend produces movement that has nothing to do with the machine. Pick a condition, record it, and keep to it.
| Property | At operating temperature | Cold, after shutdown |
|---|---|---|
| Wear debris | Suspended and circulating. What is in the bottle is roughly what is in the oil. | Settling under gravity from the moment the machine stops. Drawn off the top, the sample reads cleaner than the machine is; drawn off the bottom of a sump, it reads far worse. |
| Dissolved refrigerant | At the equilibrium the running condition sets. The sample gasses off briskly and then settles. | Higher. Refrigerant solubility in oil rises as temperature falls and pressures equalise, so oil that has sat overnight in a cold sump holds more of it. The bottle keeps releasing gas for longer, and viscosity read on it is the viscosity of oil plus refrigerant. |
| Water | Distributed through the charge at operating temperature, and that is the state the machine actually runs in. | Can separate as free water once the oil cools, which shows up as a layer rather than as a number and is easy to miss if the bottle is filled without looking at it. |
| What the result represents | The machine as it operates. | The machine plus however many hours of settling and migration have happened since it stopped. |
If you are investigating a failureA running sample is usually the more informative one, because the debris you want to see is still in suspension. Where the machine cannot be run, say so on the record — a settled sample interpreted as a running one under-reports the very thing you were looking for.
If you are building a trendConsistency beats the choice. Same point, same condition, same interval, recorded each time. A trend assembled from mixed conditions will show a step change at the point the practice changed and nowhere else.
03Procedure
Steps 4 and 5 are a pair that pull in opposite directions: cap the bottle too early and it pressurises in transit, leave it open too long and the moisture result becomes a measurement of the plant room.
Machine or compressor identity, sample point, date and time, oil family, and the running condition the sample was taken in. Write it while the bottle is clean and dry.
Why
A bottle with oil on the outside will not take a marker, and a bottle labelled afterwards is a bottle labelled from memory. An unidentified sample is a sample no result can honestly be attached to, and for a trend measurement the identity is not administrative detail — it is what makes this bottle comparable with the last one.
If this goes wrong
An unlabelled bottle arriving alongside three others from the same site is not a sample. It is four questions.
Open the sample point and let oil run to waste first — enough to clear the valve, the fitting and whatever length of line sits behind them. Then fill.
Why
The first oil out of any port is not circulating oil. It stopped moving the last time that point was used and has been sitting in a small metal volume ever since, with its own oxidation history, its own moisture and its own settled debris. None of that belongs to the machine you are asking about. Purging the point is the single act that turns a reading of one fitting into a sample of the charge.
If this goes wrong
A dead-leg sample is a perfectly accurate measurement of the wrong thing, and nothing in the numbers says so. It can send a technician looking for a fault that is not there — or, when the standing oil happens to look clean, hide one that is.
Fill to the graduation and leave the headspace above it empty. The bottle that comes in the kit is the vessel the procedure assumes.
Why
Borosilicate is specified because it is inert and it is transparent. Inert matters because the vessel must not contribute to a moisture or acidity result, and must not absorb what it is holding — plastic does both, and moisture is one of the things being measured. Transparent matters because colour, clarity, free water and visible debris are read before any instrument is involved. The graduation makes the fill volume repeatable so one sample is comparable with the next, and the headspace is where dissolved refrigerant goes when it leaves the oil.
If this goes wrong
A bottle filled to the neck has nowhere to put the gas. It either weeps past the cap or arrives with the sample pushed out of it.
Stand the open bottle somewhere clean and out of the wind, and wait for the gassing and foaming to subside before you reach for the cap.
Why
Oil from a working machine carries dissolved refrigerant, and at atmospheric pressure that refrigerant leaves. If it leaves inside a sealed bottle it pressurises it: the cap weeps, oil gets past it, the sample arrives short and the outside of the bottle is coated in what you were trying to send. Waiting also lets the foam collapse, which is the only way the fill level means anything.
If this goes wrong
A bottle capped while it is still working will not arrive as you sent it, and there is no way to tell afterwards how much left.
The moment the gassing stops, cap it and wipe the outside dry. Do not leave it standing open while you finish something else.
Why
This step and the one before it pull in opposite directions, and the technique is entirely in the gap between them. Sealed too early, the bottle pressurises in transit. Left open too long, the sample takes on the humidity of wherever it is standing, and the moisture result becomes a measurement of your plant room rather than of your machine.
If this goes wrong
Moisture picked up in the open bottle is indistinguishable, on the bench, from moisture that was in the compressor.
Oil family, refrigerant, hours on the oil and hours on the machine if you have them, running condition, sample point, and anything unusual you saw or smelled.
Why
An oil result is not measured against a published pass/fail table. It is read as condition — against the oil family, against the machine, and against the last sample from the same point. Take those away and the numbers are still true and no longer interpretable. The unusual smell you did not write down is often the most useful thing on the form.
Get the bottle moving the day it is drawn, and put the draw date and time on the record so the interval between the machine and the bench is known rather than assumed.
Why
A polyol ester will pull moisture out of whatever air it meets and keep it, and water in an ester does not sit inertly — it reacts, which moves acidity as well as moisture. Time in transit is therefore part of the measurement, not a delay before it. The date and time on the record are what let us tell drift apart from condition, which is the second reason step 1 is not administrative detail.
If this goes wrong
A bottle that waited on a shelf gives an honest reading of a bottle that waited on a shelf.
04Why promptness is not a formality
A sealed bottle is not a stopped clock. Which oil is in it decides how fast the clock runs, and it is the single reason the interval between the draw and the bench belongs on the record.
| Oil family | Affinity for water | What that means for your sample |
|---|---|---|
| Mineral | Low | Holds comparatively little water and gives it up slowly. The moisture figure is the one least disturbed by a delay — which is not permission to introduce one, because nothing else about the sample improves with age either. |
| Alkylbenzene | Low | Behaves much like mineral oil in the bottle. Treat the timing the same way. |
| POE / Solest | High | The reason the note above exists. A polyol ester takes up water readily and holds it, and water in an ester does not sit inertly — it reacts, so moisture and acidity both drift while the bottle waits. Time between the draw and the bench is part of what a POE moisture figure measures, which is why it belongs on the record. |
| PVE | High | Also takes up moisture readily. More stable than an ester once the water is in there, but the moisture number still moves, so the bottle still has to move. |
Service scope: Condition and wear analysis for chiller and refrigeration compressor oil — mineral, alkylbenzene, POE and PVE. If you do not know which oil is in the machine, say so rather than guessing — a wrong family on the form is worse than a blank one, because it silently changes what the numbers are compared against.
05Scope, stated plainly
This is the part a reader is entitled to hear from us rather than work out later.
The AHRI Refrigerant Testing Laboratory programme certifies refrigerant testing. It has no oil scope, and nothing on this page borrows it. Our refrigerant work is covered by that certification; our oil work is not, and we do not imply otherwise anywhere on this site.
AHRI Standard 700 sets limits for refrigerant. Nothing in it applies to a bottle of compressor oil, so there is no equivalent published table to hold your sample against. An oil report is read as condition — against the oil family, against the machine and against the previous sample from the same point. That is exactly why the record travelling with the bottle carries more weight here than it does on the refrigerant side.
The kit includes labelling. Confirm current packaging, marking and declaration requirements with your carrier before you ship — those requirements sit with the shipper and they change.
Amalina Technologies shares ownership with Chiller Services Refrigerants, a refrigerant reclaimer in the adjacent unit. It bears less on an oil result than on a purity result, and we state it here anyway. How the laboratory is kept separate from it.
Refrigerant purity to AHRI Standard 700, measured against limits that are published per refrigerant rather than as one generic column. The certification is listed by name in AHRI’s own directory, so it can be checked without asking us.
Check the directory ↗