Pulp Mill Set Point Control System

This paper describes the application of optimization techniques to an integrated Kraft paper mill to allocate material and energy flows throughout the mill in an optimal fashion. Each section of the mill (wood preparation, digestion, boilers, turbines, etc.) is modeled as a piecewise linear function and incorporated into a linear programming (LP) optimization code. The program determines the most economical plant unit process set points. Optimization includes such items as turbine extraction steam flows, wood blends, cooking time and chemicals, refining energy, and power allocation among the various boilers available. Model identification and parametric tracking are used to create models for certain sections of the mill. Local linearizations are then used to create piecewise linear models for optimization. Results of the LP model indicate significant savings are achievable through optimal control of process set points.