Inside the oil · a field guide
What the
numbers reveal.
Understanding olive oil chemistry tests
Each test answers a different question. Read the result alongside its method, source and the other tests.
01 / Fruit & processing · titration
Free fatty acids
A clue to fruit condition and handling.
Free fatty acids have separated from the oil’s triglycerides. Higher free acidity can reflect hydrolytic damage from damaged fruit or poor handling before extraction.
A titration measures how much base neutralises the acids. For the same oil sample mass and base concentration, more base needed at the endpoint means more free fatty acids. The result is expressed as % oleic acid equivalent; it does not measure free oleic acid alone.
Doesn’t tell you: how sour the oil tastes, or whether it is fresh today.
IOC extra virgin reference: ≤ 0.80%. Standard & scope ↓ The endpoint can be detected with an indicator or instrumentally. The amount of base, its concentration and the sample mass determine the reported acidity. IOC free-acidity method.Method & source
10.00 g oil · 0.100 mol/L base
Previously neutralised solvent; base volume measured at the endpoint.
Endpoint reached at about 1.42 mL: this oil reports as 0.40% FFA.
FFA (%) = base (mL) × concentration (mol/L) × 28.2 ÷ oil (g)
1.42 × 0.100 × 28.2 ÷ 10.00 ≈ 0.40%
That is 0.40 g of oleic acid equivalent per 100 g of oil—below the 0.80% reference limit. With the same setup, a 0.80% oil would need about twice as much base: 2.84 mL.
The slider shows titration progress for this one oil. Stopping early or adding excess base does not change its true FFA; only the endpoint volume gives the result. Volumes are rounded.
02 / Oxidation · iodometric titration
Peroxide value
A snapshot of early oxidation products.
Lipid hydroperoxides form as oil oxidises. In the test, they liberate iodine from iodide; titration measures the iodine, giving peroxide value in meq O₂/kg.
Hydroperoxides also break down into secondary products. PV can rise, peak and later fall even while deterioration continues.
Doesn’t tell you: freshness on its own. A low PV can occur at different stages.
IOC extra virgin reference: ≤ 20 meq O₂/kg. Standard & scope ↓ PV measures substances that oxidise iodide under the specified test conditions. Read it with UV results and sensory evidence. IOC peroxide method.Method & source
— Hydroperoxides / PV┄ Secondary products
Hydroperoxides start to form. A single result cannot date an oil.
03 / Oxidation & processing · UV spectroscopy
UV K-values
Lower readings are generally preferable—within each test.
UV light passes through diluted oil in a quartz cell. Absorbance is normalised for concentration and path length to give specific extinction, K.
K232 responds to conjugated dienes, associated with oxidation. It complements PV.
K268 uses isooctane in the IOC method.
K270 uses cyclohexane. These longer-wavelength readings reflect conjugated structures associated with oxidation and/or refining.
Above the reference limit is an unfavourable result for that test. A lower result is more reassuring, but it does not prove freshness or grade by itself.
Compare like with like: K232 and K268/K270 have different limits. Keep K268 and K270 as reported; switching solvent is not a conversion.
IOC references: K232 ≤ 2.50*; K268/K270 ≤ 0.22. Standard & scope ↓ K = absorbance ÷ (concentration in g/100 mL × cell path in cm). These plots zoom into the selected wavelength and show normalised K, not raw instrument absorbance. Each comparison uses the same scale. IOC UV method.Method & source
K268: 0.12 is within the 0.22 reference; 0.30 exceeds it. Read the height at the marked wavelength, not the overall size of the curve.
A closer view of the selected wavelength. Both examples use the same scale. K268 and K270 use the same smooth illustrative shape; switching the solvent changes the wavelengths, not the example’s quality.
04 / Spectral shape · a calculation
Delta K, explained
The centre compared with its neighbours.
Three readings, not five: one at the centre and one 4 nm away on each side. “4” is the wavelength spacing, not four surrounding measurements. ΔK is how far the centre sits above their average.
ΔK = K(m) − [K(m−4) + K(m+4)] / 2
For K268, use 264, 268 and 272 nm. For K270, use 266, 270 and 274 nm. The centre must be no more than 0.010 above the neighbour average. A larger gap exceeds the IOC reference.
Doesn’t tell you: whether refining occurred by itself. The IOC criterion is an upper limit; negative ΔK is possible, not a separate failure or a “better oil” score.
IOC extra virgin reference: ≤ 0.01 specific extinction variation. Standard & scope ↓
┄ Neighbour average– – Average + 0.010
The bracket is ΔK. If the centre enters the shaded area, the gap exceeds 0.010. Lines join the three readings; they are not a measured spectrum.
- Average the neighbours: (0.090 + 0.110) / 2 = 0.100
- Subtract from the centre: 0.105 − 0.100 = 0.005
- Compare the gap with 0.010.
With these neighbours, the centre can be at most 0.110.
Try moving a neighbour: the average and the allowed centre height move too. ΔK is a relative gap, not simply a high K268 or K270 reading.
05 / Fruit handling · gas chromatography
Fatty acid ethyl esters
A trace of fermentation-related chemistry.
Fermentation can produce ethanol in olives. Ethanol can react with fatty acids to form ethyl esters, so elevated FAEE can indicate poor fruit handling before extraction.
Position tells you when a compound reaches the detector. Different compounds travel through the GC column at different speeds. Under a fixed method, retention time helps identify a peak; later does not mean worse.
Area tells you how much is present. The lab adds a known reference substance (the internal standard). Calibrated ester peak areas relative to that reference determine the FAEE total. The added reference is not part of that total.
Doesn’t tell you: the exact handling history. A result below the limit is not a complete quality verdict.
IOC extra virgin reference: ≤ 35 mg/kg. Standard & scope ↓ The IOC method uses capillary GC with flame ionisation detection. FAEE is distinct from fatty acid methyl esters (FAME) and combined alkyl ester totals. IOC FAEE method; FAEE formation and method validation.Method & source
Within reference
FAEE 15 mg/kg
6 mg/kg + 9 mg/kg = 15 mg/kg
Reference limit: ≤ 35 mg/kg
Above reference
FAEE 45 mg/kg
18 mg/kg + 27 mg/kg = 45 mg/kg
Reference limit: ≤ 35 mg/kg
Both samples have the same peak positions. The right sample has three times the ester area, with the same added reference, so its example total is three times higher.
The reference is a measuring aid added by the lab, not a “good oil” sample. Its area stays fixed and is excluded from the FAEE total. Only the ester areas change when you adjust the amount. For each example ester: (ester area ÷ reference area) × 25 mg/kg. This simplified illustration assumes equal response factors and uses two placeholder esters. Real analysis uses the specified peaks, calibration, sample mass and method conditions. The times here are invented; they do not indicate the oil’s age.How the example turns areas into amounts
06 / Composition · method matters
Phenolic content
A number needs a method.
Phenolic compounds contribute to bitterness, pungency and oxidative stability. “500 mg/kg” is incomplete without the test method and reporting basis.
HPLC separates compounds; Folin–Ciocalteu measures a colour response to reducing substances; NIR predicts a property from a spectrum and a calibration model.
Doesn’t tell you: an individual compound profile or a health benefit from a total alone. Results across methods are not interchangeable. HPLC totals depend on the extraction, compounds included and calibration basis. LC–MS can strengthen identification; qNMR can quantify defined compounds. NIR reliability depends on its calibration and sample coverage. IOC phenolic methods; studies of Folin calibration and NIR predictions against HPLC.Method & source
Separate & quantify
Calibrated peak areas quantify the compounds covered by the method. A total does not reveal every compound’s amount.
Measure a colour response
Reports reducing capacity against a reference, often gallic acid. Other reducing substances can contribute; this is not an HPLC total.
Predict from a fingerprint
The prediction depends on reference lab data and how well the sample matches the calibration set. It does not count individual phenols.
Read the tests together
Several clues. No single verdict.
One report describes one sample at one time. These results alone do not establish extra virgin grade: sensory assessment and the other quality and purity requirements also matter. Missing results mean unknown, not zero.
Reference limits & scope
Limits above refer to IOC COI/T.15/NC No. 3, Rev.22 (June 2026), extra virgin category, sections 4.2–4.9. Checked . They are educational references, not a statement of the law in every market.
* K232: the IOC table makes this determination optional for commercial partners; partners in the retail market may require compliance when the oil reaches the consumer.
For a particular oil, use its reported method and source. A reference threshold is not an “excellent / good / poor” scale. Explore reported results or return to the buyer’s guide.