Optimizing Waste Fuel Boiler Control with Multivariable Predictive Control

Abstract

Multivariable predictive control, MPC, has been used in the continuous process industry for more than a decade. This strategy relies on a model created with test data from the process. The modeling produces a matrix of relationships between the "manipulated variables" and the "control variables" and "constraint variables". The MPC supervisory control software supplies the DCS, PLC or other regulatory controller with remote setpoints for the manipulated variables that will drive the control and constraint variables to their desired values. The control equations are solved simultaneously on a frequent interval to provide very tight control of the control variables and constraint limits. The technique can be applied to many pulp and paper applications including the waste fuel boilers that are an important part of the energy balance at today's mills. Significant improvements in efficiency have been achieved by optimizing the fuel and air flow to the boilers in order to minimize excess O2 and heat losses