Lubricants laboratory: tests, standards and instruments that approve every batch

Approving a batch means comparing a sample of the oil with the product's list of tests and ranges: if every result falls inside, the batch ships; if one falls outside, it stays. A batch is the oil blended at one time in one tank. At Kelpen Oil there are 25 types of test, and each batch goes through the ones its product requires. To see what to check when each delivery arrives, read how to buy oil direct from the factory.

What does a lubricant oil analysis laboratory do?

It does two jobs. At the plant, it checks that new oil is what the product promises. In maintenance, it examines oil in service, like a pool water test: it shows whether the water is still good and what has got into it.

The yardstick for a batch is the specification, the list of tests and ranges that product has to meet. At the Kelpen Oil laboratory, approval is part of the plant's management system:

• Traceability: every liter has a history, from raw material to destination.

• Retention sample: one bottle from each batch is kept, to settle questions later.

• Power of veto: no batch leaves without approval.

How does the laboratory approve or hold a batch?

First, what is in the tank: a lubricant, which is almost all base oil (see what base oils are), plus a small dose of additives, chemicals that give or strengthen a quality.

The best-known test is viscosity, how much the oil resists flowing (honey pours slowly; water pours fast). It is what gives the oil its number, like the 68 in "hydraulic 68".

That number is the ISO VG grade (from viscosity grade). It marks the center of a viscosity range at 40 °C, in a standard unit (cSt), which the ISO 3448 standard defines for each grade. Grade 68 oil has to fall inside the range for 68.

Each batch follows five steps:

1. Sample: taken from the batch, with its number.

2. Tests: the ones in the specification, with the result recorded in the system.

3. Comparison: each result against its range.

4. Decision: released, or held at the plant.

5. Record: the retention sample is kept and linked to the results, whether the batch is released or held.

For the buyer, the expected result is a new drum within the same quality range as the last one. If a doubt comes up, the batch number leads to the results and to the retention sample.

Figure 1. For ISO VG 68 (viscosity grade 68), a reading of 68 releases the batch; 58 holds it (example values).

For those who need the numbers: ISO 3448 allows 61.2 to 74.8 cSt (mm²/s) at 40 °C for ISO VG 68; the product's specification can be narrower, never wider. ISO 9001:2015, the quality management standard, requires in clause 8.6 that a product be released only after the planned checks, with a record.

What does physicochemical analysis of lubricant oil measure?

It measures how the oil behaves (how it flows in cold and heat, how it sheds water, how it protects metal) and what it is made of. Each test is called an assay, and the test standard is its written recipe (how much oil, what temperature, what instrument), so that two laboratories arrive at close numbers.

The standards come from ASTM (the American association), ABNT (which publishes the Brazilian NBR standards) and ISO (international); many NBR standards correspond to an ASTM one.

The 25 tests at the Kelpen Oil laboratory make up eight families, grouped into five blocks: cold and heat (6 tests); right product (4); recipe and base oil (4); clean, dry and unaged (4); water, foam, metal and quenching, the fast cooling that hardens steel (7).

Figure 2. Eight questions, five blocks, 25 tests.

How does the oil behave in cold and heat?

Every oil thins in heat and thickens in cold, like honey. That is why viscosity is measured at 40 °C and at 100 °C, and from the two the viscosity index (VI) is calculated: the higher it is, the less the oil changes between cold and hot.

The pour point is the lowest temperature at which the oil still flows. For engine oil, the cold cranking simulator (CCS) measures the viscosity of the cold oil under the hard shearing of an engine start, how much it drags on the starter motor.

The flash point is the lowest temperature at which the vapor of heated oil flashes when a flame passes over it. It is a safety figure for handling and does not say how far the oil can work.

In a new batch, off-range viscosity points to a wrong proportion of base oils or a leftover of another product in the piping; a low flash point points to contamination by a lighter product, such as a solvent.

How do you know the oil is the right product?

Four tests confirm the identity of the batch. Density is how much a given volume of oil weighs. Most lubricating oils weigh less than the same volume of water, so they float; a swapped or contaminated product changes that weight.

Color is read on a numbered scale (the ASTM one) or, for nearly transparent products, on the platinum-cobalt scale, made for clear liquids. Off-color points to a different raw material, contamination or degraded oil.

In infrared (FTIR), the oil receives a light the eye cannot see, and each type of chemical bond absorbs bands of its own. The resulting graph is the fingerprint of the formulation: if it does not match the product's reference, the batch may have a swap, a mix or contamination, and the laboratory investigates.

Figure 3. The batch spectrum over the reference: an extra peak can give away another product.

For those who need the numbers: ASTM E2412 was written to follow oil in service; for a new batch, the laboratory uses the same spectrum only to compare it with the product's reference.

How does the laboratory check the recipe and the base oil?

It measures the chemical elements that mark the additives. Many additives carry barium, boron, calcium, phosphorus or zinc, and their content shows whether the additive set (the package) went in at the right dose. Not every additive has them, and in plain base oil there is no package to check; there, what is checked is the sulfur of the base oil itself.

In ICP, or plasma spectrometry, the sample goes into a gas at thousands of degrees and each element gives off its own light, like the salt that turns a stove flame yellow. The intensity of that light gives the content, in ppm (parts per million).

In X-ray fluorescence (XRF), the sample is lit by that radiation and each element answers with X-rays of its own energy.

For the three elements both methods share (calcium, phosphorus and zinc), the two cross-check each other, as long as the oil has no barium or chlorine. Sulfur, only above about 0.05%, comes from XRF. In a new batch, an off-range element points to a package at the wrong dose or a swapped one.

Two tests in this block look at the base oil, not the additives. The aniline point is the lowest temperature at which equal volumes of oil and aniline (a reference solvent) mix completely. The lower it is, the more aromatic and the more solvent the oil is, and the more it tends to swell nitrile rubber seals.

Carbon distribution shows what kind of structure the base oil is made of: aromatic rings (CA), paraffinic chains (CP) and naphthenic rings (CN). Each abbreviation says how much of the carbon is in one of these forms (see paraffinic, naphthenic and aromatic base oil).

To learn more about how the refining level changes the base oil, Group 2 base oil has its own article.

For those who need the numbers: 1 ppm is 1 mg per kg of oil. The carbon distribution calculation (ASTM D2140 and D3238) combines viscosity, density and the refractive index, how much light bends in the oil.

At Kelpen Oil, ICP measures barium, boron, calcium, phosphorus and zinc; XRF measures calcium, phosphorus, sulfur and zinc, and the three in common serve as a cross-check at the contents both reach. ASTM D6481 was written for new oil and covers, by mass, phosphorus from 0.02 to 0.3%, calcium from 0.02 to 1.0%, zinc from 0.01 to 0.3% and sulfur from 0.05 to 1.0%. It does not apply to oil with barium or chlorine in the formula. The cross-check holds within those conditions, and sulfur comes only from XRF, within the scope of D6481.

Figure 4. At the Kelpen Oil laboratory, the two methods and the three elements in common, which serve as a cross-check.

Is the oil clean, dry and unaged?

Two tests measure water and dirt. Karl Fischer uses a reagent that combines with water in a fixed proportion, and the amount used gives the water content.

The particle counter passes the oil in front of a beam of light and counts the grains by size, like dust against the sun. The count becomes the three-number cleanliness code of ISO 4406.

In a new batch, high water points to moisture or condensation in the tank, piping or raw material; a high code points to dirt from the tank, piping, filter or packaging.

The other two measure acidity and alkaline reserve. The acid number (TAN) says how much acid the oil has; in oil in service, it rises when the oil oxidizes (ages by reacting with oxygen, more so in heat).

In new oil, part of the acidity comes from the additives themselves, and an off-range TAN points to an additive at the wrong dose or an oxidized raw material.

The base number (TBN) measures the alkaline reserve, the "antacid" of engine oil. Both come from titration: a reagent is dosed little by little until an electrode detects neutralization.

For those who need the numbers: the Kelpen Oil Karl Fischer is coulometric: the instrument electrically generates the iodine that reacts with the water, and the charge used gives the content. Mercaptans, ketones and certain additives interfere with Karl Fischer. In ISO 4406, each extra number is about twice as many particles.

How does the oil behave with water, foam, metal and in quenching?

The first test shakes oil with water, which should separate, like oil and vinegar in a dressing left standing. It gives two readings, which the list of 25 counts as separate tests. Demulsibility is the time to separate; the emulsion characteristics are the volumes of oil, water and emulsion (the mix that did not separate) at the end.

In detergent hydraulic oil (hydraulic oil with a detergent additive), the additive deliberately keeps the little water in fine droplets, so it does not collect in pools in the reservoir. That is why its specification does not ask for the fast separation of ordinary hydraulic oil.

In the foam test, air blown into the oil in three stages, two at 24 °C and one at 93.5 °C, must not leave too much foam or foam that is slow to disappear, like a shaken soda.

In a new batch, high foam points to antifoam at the wrong dose or a leftover of another product. Worse water separation points to a leftover of detergent oil in the piping or a demulsifier (the additive that helps water drop out) outside its dose.

Figure 5. The separation time and the volumes left at the end.

The copper strip stays in hot oil and is compared with a color scale, like a coin that tarnishes: it shows whether the oil attacks copper, bronze and brass.

The salt spray test compares the rust protection that a protective oil (the oil wiped on metal so it does not rust) gives in a salt chamber, like a car near the beach.

In quenching, the cooling curve shows how fast the oil, or the polymer dissolved in water that some processes use instead of oil, pulls heat from a standard metal probe dipped in place of the part.

For those who need the numbers: in ASTM D892, sequences I and III are at 24 °C and II at 93.5 °C. In ASTM D6200 and D6482, the probe is a nickel alloy, heated to about 850 °C.

Which tests decide the release of each type of oil?

The ones that usually decide are the tests that confirm product and dose, and the test tied to what the oil does in the machine. The most common are viscosity, density, color and infrared; depending on the product, flash point, acidity, water and the additive elements come in.

The table shows what weighs most for the most common types. The product standard (the list of requirements for the type of oil, such as DIN 51524 for hydraulic oil) also asks for performance tests; those are done once, in the qualification of the formula, the approval of the recipe.

What equipment is used in lubricant oil analysis?

Viscometer, refractometer, spectrometers, titrators, density meter, particle counter, colorimeter and automatic analyzers. At Kelpen Oil, all of them have traceable calibration, checked against reference standards. Among them:

Why is an automated test more reliable?

Because the instrument repeats the same protocol, varies less from one analyst to another and sends the data straight to the system. The small variation that belongs to each method remains. At Kelpen Oil, whatever the method allows is automated; copper strip, pour point, demulsibility, foam and salt spray need visual reading, by the criterion of the standard.

What does "approved batch" mean and how do you check it on the certificate?

It means that every test in that product's specification came out inside its range and was recorded with the batch number. Approved means the batch is the product the name says: an approved hydraulic 68 meets the specification of hydraulic 68, not that of a gear oil.

A batch certificate, also called a certificate of analysis, usually shows on each line: test, method (the standard), unit, result, specification (minimum, maximum or range) and status (conforming or not).

Figure 6. A certificate line in the ISO VG 68 example (example values).

1. Typical or specification? A product's data sheet gives typical values, what batches usually come out at; the certificate compares with the specification, the limits. A result far from typical but inside the range is still conforming.

2. Did your laboratory measure something different? Check that the method is the same: viscosity, for example, can come from another standard. Each method has a margin accepted between laboratories, the reproducibility, and a difference inside it does not prove the oil is another one. A larger difference can be settled with the batch's retention sample; talk to the Technical Team.

For those who need the numbers: ISO 4259-1 says how the reproducibility (R) of a method is obtained; ISO 4259-2 and ASTM D3244 say how to use it when supplier and buyer disagree. Viscosity measured by ASTM D445 (glass capillary tube) and by ASTM D7042 (Stabinger viscometer) is compared after the bias correction that D7042 itself provides.

What does analysis of oil in service reveal?

With almost the same tests, it shows whether the oil is still good and what has got into it: it is the pool test. The yardstick becomes the new oil, the earlier samples and the alert limits, values agreed for that machine at which you act.

• Viscosity that rises points to oxidation or thicker oil mixed in; one that falls, to fuel, thinner oil or a broken VI improver, an additive of long molecules that mechanical stress snaps.

• TAN rising points to advanced oxidation; TBN falling, to the alkaline reserve running out in engine oil.

• Water points to a heat exchanger leak (the equipment that cools the oil with water), condensation or entry through the seal.

• Cleanliness code above target points to insufficient filtration or dirt ingress.

• Additive elements different from the new oil suggest mixing with another oil.

One caution: spent additive usually stays in the oil, and ICP measures the element. Stable phosphorus and zinc do not prove the additive is still working; the consumption of some, such as the zinc antiwear additive that protects parts in rubbing contact, shows up in infrared.

One analysis is a snapshot. Samples from the same point, at regular intervals, form the trend, which, against the alert limits, decides between keeping, correcting or changing the oil. This is condition-based oil change: change when the analysis shows it is time, not by the calendar.

Condition-based change matters most where stopping the machine costs a lot, as in a tissue paper mill (see lubricants for the tissue industry).

The Kelpen Oil laboratory does this analysis as a paid service, one-off or under contract, on oil from any manufacturer, with a technical report and a recommendation. It also prepares the analyses that ANP (the Brazilian National Petroleum Agency) requires to register a lubricant. The laboratory also offers training in lubricant analysis.

For those who need the numbers: ASTM E2412 describes following oil in service by infrared, and ASTM D7720 describes how to set alert limits.

Frequently asked questions

How many tests does the Kelpen Oil laboratory run? 25 types; most of the tests follow ASTM, NBR and ISO standards. Each batch goes through the ones its product's specification asks for.

Why are there two methods to measure elements? Each reaches different elements: ICP (NBR 14786) measures barium, boron, calcium, phosphorus and zinc, and X-ray (ASTM D6481) measures calcium, phosphorus, sulfur and zinc. For the three in common, one checks the other.

How long has Kelpen Oil been ISO 9001 certified? The certification, to the quality management standard, has been continuous since 2003.

Glossary

• Batch: the oil blended at one time in a tank, with its own number.

• Specification: the list of tests and ranges a product has to meet.

• Retention sample: a bottle kept from each batch, to settle questions.

• Test standard: the written recipe of a laboratory test.

• Reproducibility: the difference accepted between two laboratories, same sample, same method.

Talk to the Technical Team

The Kelpen Oil Technical Team looks at the equipment, the lubricant, the hours since the last change, the temperature, the sampling point and the symptom, and builds the analysis plan with you. If the oil is already in service, the laboratory in Arujá (SP) runs the analysis of oil in service, as a paid service. Send this article to whoever looks after predictive maintenance (the kind that acts on the machine's condition, before it fails) or the receiving of lubricants.