Rice Dehulling

Dehulling exists to remove a layer nobody can eat while leaving intact the kernel underneath. Those two requirements pull against each other, and the whole design of the step is the compromise between them: force sufficient to open every husk will also crack kernels, and force gentle enough to spare every kernel will leave husks on. Breakage that starts here is permanent — a kernel cracked at dehulling stays cracked through every step that follows, and eventually leaves the mill as brokens rather than whole grain.

The principle is differential shear. Two rubber rolls turn in opposite directions at different speeds, with a gap slightly narrower than a paddy grain. Each grain is caught, and because the two surfaces touching it are moving at different rates, it is not crushed but twisted — the husk splits along its natural seam and is stripped off. The kernel, which is not gripped in the same way, passes through. Rubber deforms around the grain rather than concentrating force on a point, which is why the roll is a gentler instrument than the abrasive designs it replaced.

Dehulling paddy yields brown rice and rice husk. Brown rice, however, is not modelled here as a product of the process, and the reason is worth stating plainly: brown rice is itself a traded commodity, bought and sold in its own right and stored, shipped, and specified as a commodity rather than as a mill output. This reference therefore models the paddy-to-brown-rice step as a relationship between two commodities rather than as a commodity-to-product edge. Dehulling’s declared product output is the husk; brown rice appears as the input commodity of the whitening step that follows, and the chain is walked through the commodities rather than around them.

That is a modelling decision about where commodity identity ends, not a claim that brown rice is unimportant — it is the more valuable of the two streams by a wide margin. It reflects the fact that the rice chain is a sequence of distinct traded commodities, each of which can leave the chain and be sold, in a way that flour and bran leaving a wheat mill cannot.

Brown rice

The dehulled kernel, still carrying bran and germ. A distinct traded commodity in its own right, and the input to whitening. Not modelled as a product of this method.

Rice husk

The removed outer covering. A by-product, not a waste: it is widely used as fuel — including to fire the mill itself — and as a source of silica and ash for building materials and other uses.

Immature and unhulled grain

Paddy that passes through unopened is separated and returned, and immature kernels are removed. These are handled within the mill rather than sold as streams.

What happens here governs what the mill can achieve later. Whole-kernel yield is the measure the whole rice trade turns on, and it is largely determined before whitening ever begins: kernels cracked at dehulling cannot be uncracked, and every one becomes a broken further down the line. Dehulling also cannot fix what the field and the dryer decided — a kernel already fissured by uneven drying will break here regardless of how well the rolls are set.

The characteristic loss points are therefore breakage and incomplete removal, and they trade against one another directly.

• Kernels cracked or broken at the rolls — permanent, and realised as brokens later
• Husks left unopened, requiring recirculation and a second exposure that risks further breakage
• Kernels already fissured by uneven drying, which break here whatever the mill does
• Whole kernels carried out with the husk at aspiration
• Roll wear widening the gap and degrading both removal and breakage together

Rice husk is the clearest case in the cereal chain of a residue that is not a waste. It is bulky and abundant, and it is widely burned as fuel — often to raise the steam and power for the mill producing it — while the silica-rich ash left behind has recognised uses in cement and other materials. Where husk is neither used nor properly managed it becomes a genuine disposal and air-quality problem, and both outcomes are common; which one occurs is a matter of local infrastructure rather than of the process.