Quality & Specification
Understanding Pulse Purity
What a purity percentage actually measures, why 98% and 99.5% describe different cleaning processes rather than different levels of care, and how purity relates to every other parameter in a contract.
· 7 min read
Purity is the headline figure in most pulse contracts and the most frequently misunderstood. It is not a measure of how clean a lot looks, nor a grade in its own right. It is an arithmetic statement: the percentage of a representative sample, by weight, that is sound, on-specification product of the contracted type. Everything else in the sample — foreign matter, other varieties, damaged and broken grain, immature seed — is what makes up the balance.
Purity is the inverse of the defect list
Because purity is defined as what remains after the defects, it is not independent of the other parameters in a specification. If a contract states 98% purity, 1% foreign matter, 2% damaged and 2% broken, those numbers describe overlapping territory and cannot all sit at their maximum simultaneously. Reading a specification therefore means reading it as a system rather than as a list of independent limits.
This is why experienced buyers tend to negotiate the individual defect tolerances rather than the headline purity figure. Two lots at the same purity can be commercially very different: one where the 2% shortfall is broken grain, and one where it is insect-damaged seed, will perform differently in a mill and will be valued differently.
What separates 98% from 99.5%
The gap between a typical machine-cleaned purity and a typical colour-sorted purity is not a gap in diligence. It is a gap in process capability. Mechanical cleaning separates on size, weight and density; it removes anything that differs from sound grain in those properties. What it cannot do is identify a seed that is the correct size, the correct weight and the correct density but the wrong colour.
Optical sorting adds exactly that capability. Cameras assess each grain against an accepted colour range and eject what falls outside it, which removes discoloured, stained, mould-marked and off-shade seed and lifts purity into the 99% and above range. The cost is yield: everything the sorter ejects leaves the contracted lot.
| Stage | Separates on | Typically removes |
|---|---|---|
| Pre-cleaning | Size | Straw, pod, stalk, mud, coarse trash |
| Aspiration | Weight | Dust, husk, chaff, shrivelled and hollow seed |
| Screening | Size | Oversize, undersize, broken pieces, small weed seed |
| Destoning | Density | Stone, soil pellets, dense foreign matter |
| Gravity separation | Density | Insect-hollowed and partly filled grain |
| Colour sorting | Colour | Discoloured, stained, off-shade and off-variety seed |
Purity and end use
The right purity for a contract is the one the end use requires, not the highest one available. A dal mill removes the husk and splits the cotyledon; discoloured seed coats are largely irrelevant to it, and paying a colour-sorting premium adds cost without adding value. A retail packer selling through a transparent pack is buying appearance, and a machine-cleaned lot will generate complaints regardless of its purity figure.
How purity is measured in practice
- A representative composite sample is drawn from across the lot, not from a few accessible bags.
- The composite is reduced to a working sample of a manageable size, using a divider or by quartering.
- The working sample is weighed.
- It is separated by hand into its component fractions: sound grain, foreign matter, other varieties, damaged, broken, insect damaged, immature.
- Each fraction is weighed and expressed as a percentage of the original working sample weight.
The arithmetic is simple. The judgement is not, and that is where disagreements arise. Whether a particular seed counts as "discoloured" or as "sound but off-shade" is an interpretation, and two competent assessors can differ. An approved sample sealed and held by both parties converts that interpretation into a physical reference, which is why it is worth the time it takes to agree.
What purity does not tell you
- Moisture — measured separately and the single most important arrival parameter.
- Size or count — a separate contracted parameter for mung beans, chickpeas and all beans.
- Crop year — purity says nothing about age, and age affects colour, cooking behaviour and insect load.
- Cooking and milling performance — assessed on a trial, not from a percentage.
- Germination, for sprouting buyers — tested on the actual crop.
See how the individual parameters appear in a working contract specification. Specifications hub