The velocity gradient selected for rapid-mixer design should be developed through the use of bench-scale or pilot plant testing. An alternative to laboratory testing is to utilize the experience of existing plants treating the same (or similar) water. In general, a short duration of high-intensity mixing often gives the best results.