(614c) Nonlinear Planning Model for FCC Operation: Simulation and Optimization

A nonlinear mathematical model for describing Fluid Catalytic Cracking - FCC unit operation is presented in this work. Operation of the regenerator was assumed to be constant in coke burning by specifying flue gas composition and Air consumption. Semi empirical correlations sets for the reactor were tuned to reproduce the yields reported in the backcasting information of four commercial FCC plants. Accuracy of the model in reproducing volume yields was verified and several simulations tested the procedure with new case studies. Tuning of the model guaranteed the correct representation of the FCC operation and prepared the model for its use in planning studies thereof. Mathematical structure of the model (i.e. nonlinear algebraic equations) allows both the properly representation of the cracking operation and the solution of the model.

One problem that many refineries face worldwide is the formulation of the appropriate feed composition when both different feed stocks and FCC units are available. Optimization of the routing for feeding four commercial FCC units was modelled by the use of the mathematical approach disclosed in this work. The problem of routing was expressed in terms of the MINLP topologies, and it was optimized by the use of the solver DICOPT enclosed in the General Algebraic Modelling System (GAMS). Gasoline maximization, LPG maximization, LCO maximization and profit maximization scenarios were analyzed for the real operation with the developed procedure. Results in all the cases coincided with the operational surfaces described by the backcasting files of each unit and with the expectations of the process experts.

The model allows treating the cracking operation nonlinearly while it can be easily adapted to the dynamic nature of the oil industry by the tuning of the parameters.