Finishing
Draining is exponential in cycles; purging is exponential in volume. Three drains beat a continuous flush by half.
— nos
To bring carryover under the limit
The flow-through comparison assumes the trough is perfectly mixed, which is the pessimistic idealisation: a real trough has short-circuiting and dead corners, so some liquor leaves faster than the model says and some stubbornly does not, and the tail of a continuous purge is worse than exponential rather than better. That makes the case for draining stronger, not weaker. Drain efficiency is the input that decides everything and it is a machine property worth measuring once - fill with a traceable salt, drain, refill with clean water and titrate, and the heel fraction falls straight out. It is usually worse than the operator believes, because the heel hides in pipework, pumps and the doctor blade recess rather than in the visible trough. The concentration model tracks a soluble carried species and says nothing about a deposit: a pigment, a silicone or a crosslinked resin film on the trough wall does not dilute at all, and a changeover between chemistries that leave films needs mechanical cleaning that this arithmetic cannot substitute for. Where the next recipe is sensitive to a trace rather than a percentage - a white after a dark, or a hydrophilic finish after a fluorocarbon - set the carryover limit accordingly and expect the cycle count to rise steeply, since it grows with the logarithm of the ratio.
Finishing Trough Changeover: Purge Volume & Time — free, with the formula and a worked example, at Textile School.