Linear Model Predictive Control of a Debutanizer column

Linear model predictive control studies have been an interesting field of research in the scope of chemical engineering. This research has been done in order to benefit the operations of PETRONAS Penapisan Terengganu Sdn. Bhd., PP(T)SB. The Debutanizer column at the plant is not achieving its desired output which is 30% propane and 70% butane. Therefore this study is conducted in order to obtain the optimum tuning parameters for Model Predictive Control, MPC controllers that can help achieve the desired output at the plant operation. The Debutanizer column has been simulated using HYSYS™ software according to real plant data. By conducting an open loop step test, relevant data were collected in order to proceed to MATLAB programming. By using the IDENT System Identification tool available in MATLAB, simple programming and coding were done to obtain necessary input information for the MPC controllers. ‘2 by 2 matrix’ was used. The findings from this research can be used to simulate MPC controllers in HYSYS™ software. The response and behavior of the simulated plant environment can be further studied and researched in order to achieve the desired output of the Debutanizer column.