Pilot-Plant Scale-Up
Safety Evaluation
Scale-up incidents have a well-documented pattern: chemistry that runs safely at 100 mL in a stirred-tank lab reactor escapes thermal control at 100 L pilot scale because surface-area-to-volume ratio drops by an order of magnitude, mixing time extends, heat removal lags, and the previously-buffered exotherm escalates. T2 Laboratories 2007 (MCMT runaway, 4 fatalities), Concept Sciences 1999 (hydroxylamine), Bayer CropScience 2008, and a steady stream of pharma / specialty-chemical pilot-plant events all stem from inadequate scale-up safety evaluation. The discipline integrates calorimetric data (DSC, RC1, ARC, VSP2) translated through Stoessel criticality logic into the scale-target operating envelope, dimensionless heat-transfer analysis (Damköhler, Biot, Nusselt numbers), mixing-time and dead-zone characterisation (CFD or experimental tracer studies), continuous-vs-batch decision logic, engineering-controls inventory (cooling capacity, dosing rate, emergency relief, kill-system effectiveness), and HAZOP at each scale gate. Modern good-practice includes continuous-flow / microreactor evaluation at pilot stage as an inherently-safer alternative — particularly for energetic chemistry where the kg-scale residence time is the failure-mode driver.
