Coffee Bean Size

What determines bean size

1. 1Variety. Different Arabica selections have characteristically different seed geometry. This is genetic and consistent.
2. 2Growing conditions during fill. Water availability and nutrition while the cherry is developing set how much resource the seed gets.
3. 3Crop load. A heavily bearing tree distributes finite resources across more cherries, producing smaller seeds. Biennial bearing shows up in the size distribution.
4. 4Altitude. Slower maturation at altitude tends to produce denser, and often somewhat larger, seed — though this interacts with everything above.
5. 5Stress events. Drought or hail during development compresses the whole distribution downwards; the effect is visible across an entire season’s outturn.

Note what is not on that list: processing. Washing or drying a cherry does not change the width of the seed inside it. Size is set on the tree.

Distribution, not a single number

Every lot has a size distribution, not a size. A screen analysis reports what percentage of a sample is retained on each screen in the stack, and that curve is far more informative than the headline grade.

This is also the practical answer to "can you supply screen 18 in volume": it depends on how much of the crop is screen 18, which depends on the season. See screen 18 for the yield consequences.

Why uniformity matters in the roaster

Heat enters a bean through its surface and moves inward. Smaller beans have a higher surface-area-to-mass ratio, so at a given charge temperature and airflow they develop faster than larger beans in the same batch.

A wide size distribution therefore means that, at the moment the batch is dropped, the small beans are further along the roast than the large ones. Push for full development of the large beans and the small ones scorch; protect the small ones and the large ones stay underdeveloped. Neither outcome is good, and no amount of profile skill fully resolves it.

Size and quality: the honest position

Bean size and cup quality share some upstream causes — a well-nourished tree at suitable altitude in a good year tends to produce both larger seed and better coffee. That shared causation is why the association exists and why it survives in market pricing.

But shared causes are not a direct relationship. Fermentation, drying, sorting and storage all sit between the tree and the cup, and none of them touch bean width. A large bean processed badly is bad coffee. This is the same point made from the Kenya grades side, and it is worth making twice.