A drum of refined hemp seed oil is only as good as the certificate of analysis travelling with it, and most buyers learn which lines on that document mattered only after a batch has turned in the warehouse. The specification sheet is where quality is defined; everything else is packaging.
Why "refined" is not a specification on its own
Refining is a sequence of unit operations, not a single state. Depending on the processor, an oil sold as refined may have been degummed, alkali neutralized, bleached, winterized and deodorized, or it may have passed through only two or three of those steps. Each stage removes something different: gums and phospholipids, free fatty acids and soaps, pigments and trace metals, waxes, and finally volatile odour and flavour compounds. Two oils carrying the same word on the label can behave very differently in a formulation and on a shelf.
So the negotiation belongs on the spec sheet. Ask for a batch-specific certificate tied to the lot number rather than a sheet of "typical values," and make sure each line states a limit you can reject against rather than an average. A number quoted without its test method is not a specification.
Free fatty acid and peroxide value: the two freshness numbers
Free fatty acid value measures how much of the triglyceride structure has hydrolyzed, releasing fatty acids into the oil, and is usually reported as a percentage expressed as oleic acid. A properly neutralized refined oil should sit in the low tenths of a percent or below. A figure creeping toward or past a full percent points to damp or late-pressed seed, oil that sat before refining, or a neutralization step cut short. High FFA lowers the smoke point, contributes sharp and soapy notes, and indicates conditions that will keep generating more of it.
Peroxide value measures primary oxidation, the hydroperoxides formed when oxygen attacks the double bonds, reported in milliequivalents of active oxygen per kilogram. Fresh, well-handled refined oil should be low single digits or under. The trap is that peroxide value is not monotonic: hydroperoxides eventually decompose into aldehydes and ketones, so a badly aged oil can present a flatteringly low figure. That is why the p-anisidine value, which captures those secondary breakdown products, belongs on the same sheet. A low peroxide value paired with a high anisidine value is the signature of an oil already past its oxidation peak. The combined Totox calculation compresses the two into a single figure.
What do colour and clarity actually tell a buyer?
Colour is normally reported as a Lovibond red and yellow reading or a Gardner value. Crude hemp seed oil is deep green from chlorophyll; refining and bleaching pull it toward pale gold with a green cast. The reading is a proxy for how thoroughly the bleaching step worked, and it flags a stability risk, because chlorophyll acts as a photosensitizer that accelerates oxidation whenever the oil sees light. A dark oil in clear packaging under retail lighting is a short-lived product. The opposite failure is equally real: aggressive bleaching strips tocopherols along with the pigment, and tocopherols are the oil's own antioxidant system. The palest oil is not automatically the best.
Clarity is a separate question. Winterization and polish filtration remove waxes and residual phospholipids that cloud the oil as it cools. The relevant lines are insoluble impurities, sediment, and a cold test holding the oil at refrigeration temperature for a stated number of hours without haze. Cloudiness is cosmetic in retail bottling and operational everywhere else: the same waxes blind filters downstream.
Moisture, volatiles and the trace metals nobody asks about
Moisture and volatile matter is measured by loss on drying or Karl Fischer titration and should be a small fraction of a percent. Water drives hydrolytic rancidity, so an elevated moisture figure at intake reliably predicts that free fatty acid value will climb in storage, and it causes foaming in heated processes.
Trace metals deserve more attention than they get. Iron and copper are catalytic pro-oxidants at parts-per-million levels, picked up from processing equipment, transfer lines or unlined storage. An oil with an excellent peroxide value today and elevated iron will not hold it. A supplier who quotes trace metals at all is a good sign; ask in the same conversation about vessel material, nitrogen blanketing and headspace control.
How stable is stable enough?
Hemp seed oil is dominated by polyunsaturated fatty acids, principally linoleic and alpha-linolenic acid with a smaller contribution from gamma-linolenic acid, and the omega-6 to omega-3 ratio is commonly cited near three to one. Those double bonds are simultaneously the nutritional selling point and the stability problem. Its iodine value sits in the range associated with drying oils, a compact way of saying this oil wants to react with oxygen.
Oxidative stability is measured by accelerated methods reporting an induction time in hours at an elevated temperature. Hemp's induction times are short compared with monounsaturated oils such as olive or high-oleic sunflower, which is expected rather than a defect. What matters is comparing like with like: an induction time quoted without its test temperature is meaningless, and figures run at different temperatures cannot be ranked against each other. A longer induction time reflects intact tocopherols, low starting peroxides, low catalytic metals and good inert-gas handling.
THC testing: the specification with legal consequences
Hemp seed does not itself produce cannabinoids. Trace THC in seed oil comes from external contamination, resin transferred onto the seed coat from bracts and leaf material during harvest and handling. Thorough seed cleaning and dehulling before pressing is what keeps the number low; no refining step substitutes for clean incoming seed.
Method matters enormously here. Gas chromatography decarboxylates THCA in the injector unless the sample is derivatized, so a GC result reports something closer to total THC, while liquid chromatography can separate the acid and neutral forms. Total THC and delta-9 THC are different numbers, and a contract that does not say which one it means is unenforceable in practice. Specify the analyte, the method, the limit in micrograms per gram and the accredited laboratory, and require testing on every lot rather than an annual certificate. Limits and permitted end uses differ by jurisdiction and product category, so write the specification against the destination market's rules, not the country of production.
Reading the sheet as a whole
Individual numbers matter less than the pattern they form. For related background on production and sourcing, see the rest of our article library.
| Specification | What it reports | What a poor figure implies |
|---|---|---|
| Free fatty acid | Hydrolysis of triglycerides | Damp or aged seed, incomplete neutralization |
| Peroxide value | Primary oxidation | Oxygen exposure, warm or lit storage |
| p-Anisidine value | Secondary oxidation | Oil past its peak, whatever the peroxide value |
| Colour | Pigment and chlorophyll load | Light-sensitive, or over-bleached and stripped |
| Moisture and volatiles | Residual water | Free fatty acid will rise in storage |
| Insolubles and cold test | Filtration and winterization | Haze in the bottle, blinded filters |
| Oxidative stability | Resistance to accelerated oxidation | Short shelf life from a clean start |
| Iron and copper | Catalytic pro-oxidant metals | Good numbers that will not hold |
| THC | Contamination from seed handling | Poor seed cleaning; market access risk |
Turn this into a working procedure rather than a preference. Write a purchase specification naming the analyte, method and limit for every line above, and require a lot-specific certificate matching the drum markings on arrival. Retain a sealed, headspace-minimized sample from each lot so a later dispute can be settled against physical evidence. For the first several shipments from any supplier, retest free fatty acid, peroxide value and moisture yourself at intake and record the gap between their certificate and your result. That gap, tracked over a handful of deliveries, tells you more about a supplier than any single specification sheet will.