(13c) Are You Feeling Lucky?: Stop Relying on Luck to Optimize Your Pilot Plant Processes

Luck plays a surprising part in process optimization. One-Factor-At-A-Time methods rely on luck, often producing neither the results you need nor a "map" of the process beyond the actual experiments performed. Design of Experiments(DOE) eliminates the element of luck and is proven to use the minimum resources to extract the most information from a pilot plant optimization study. Response Surface Methodology (RSM) enables you to understand the process at any level(s) of the factors studied.

Three purposes of pilot plant studies are to demonstrate the viability of new processes, generate design data, and produce market development samples. At the same time the pressure is on to get to market quickly with a robust manufacturing plan. Maximum efficiency of pilot plant efforts is critical. Engineers must understand the nature and scope of their factor interactions. In lieu of pushing the actual unit beyond its limits of performance, data can be extrapolated from a properly designed experimental analysis. Only RSM can easily point to the "Sweet Spot", that point where all response goals are simultaneously met.

Engineers must successfully communicate the results of their efforts. Conclusions demonstrated with contour plots visually compel decisionmakers to "see" the work performed, deductions and projections. Engineers who have struggled to convey the results of their optimization work will appreciate a better way to communicate it.

Examples from actual pilot plant optimization studies will be used to illustrate the necessity and effectiveness of DOE/RSM.