Jar Test Dose Scale-Up

The jar test is the oldest instrument in water treatment and still the best: six beakers, six coagulant doses, a paddle stirrer, and an hour. The arithmetic that connects a pipette to a chemical feed pump is this one. Make up a stock at 10 g/L — 10 mg per mL — and 2 mL into a 1 litre jar delivers 20 mg of chemical per litre of sample, that is 20 mg/L, and the plant then feeds 20 mg/L on the raw water line. The convenient convention is a 1% stock (10 g/L) into 1 L jars, where each millilitre is exactly 10 mg/L.

Three habits separate a jar test that predicts the plant from one that misleads it. Use the raw water as it is at that moment, not a sample that has sat warming on a bench for four hours, because temperature changes both the coagulation chemistry and the settling velocity. Match the mixing to the plant: a flash mix of 100 rpm for a minute, then flocculation at 25–30 rpm for 20 minutes, then a settling period matched to the basin detention. And measure the right thing at the end — settled turbidity tells you about the sedimentation basin, but filtered turbidity and the residual coagulant tell you what the filters will see. If your bench optimum never works on the plant, the difference is nearly always mixing energy, not dose.