This paper discusses the core process models of a knowledge based supervisory control system designed to provide expert knowledge in the control and management of a sugar mill crystallization stage. Forecasting stock tank levels, within the sugar factory crystallization stage, is beneficial to ensure that there are sufficient quantities of materials in stock to allow completion of production schedules without disruption. Integration of projected vacuum pan feed rates with syrup rate production models and pan phase detection models allows forecasting of stock tank levels for liquor and molasses to forewarn of potential problems with the current operating strategies and advise corrective procedures. These models form an integral part of an overall pan stage knowledge based system to assist in providing a better decision-making strategy for crystallization stage operations.