This paper has the goal addressing the question that frequently rises when implementing a multivariable control strategy: What should be used? Structured PIDs OR Model Predictive Control (MPC)?
The paper gives an answer supported by the presentation of an application to an alkylation plant Deisobutanizer (DIB). It will describe how Material Balance Control was used to guarantee the stability, performance and robustness required to have the consistent steady state stability that is prerequisite to the implementation of any supervisory optimizing application such as Model Predictive Control (MPC). The paper will conclude by recommending that the answer to the question should be PIDs AND MPC and asserting that these are not competing but complementary control strategies.
