(363o) Nonlinear Model Predictive Control for the Dividing Wall Column

Dividing wall columns (DWCs^{[1, 2]}) are practical, effective, and promising among distillation process intensification technologies. Although the three-product Petlyuk DWC can reduce approximately 30% operation costs^[3], the fear of controllability problems impedes its large-scale industrialization. Therefore, a feasible Nonlinear Model Predictive Control (NMPC) scheme is studied in this paper to control the three-product Petlyuk DWC. NMPC has been applied in conventional distillation process control. As the Petlyuk DWC is strongly interactive and highly nonlinear, the NMPC may be more suitable than the traditional PI control for the DWC. The model is established based on Python and Pyomo platform.

Pyomo^{[4, 5]} has a flexible environment with inherent object-oriented aspects, along with advanced tools to automatically discretize general Differential-Algebraic Equation (DAE) optimizations. For a general Pyomo DAE model, Control and Adaptation with Predictive Sensitivity Enhancements (CAPRESE^{[6, 7]}) is a nonlinear optimization-based framework in Python for sensitivity-based NMPC and Moving Horizon Estimation (MHE) strategies. The NMPC scheme for the three-product Petlyuk DWC is illustrated in Figure 1. The controlled variables of the MPC are five input compositions, including impurity compositions of two outflows of the prefractionator (y_Dp,C, and x_Bp,A), and three product compositions of the corresponding product streams of the main section (x_D,A, x_S,B, and x_B,C). The manipulated variables of the MPC are five outputs, containing the reboiler duty to feed ratio (Q_R/F), reflux flow rate, reboiler duty, side product flow rate, and thermally coupled liquid and vapor flow rates to the prefractionator. The separation of a ternary equimolar saturated liquid mixture of ethanol (A), n-propanol (B), and n-butanol (C) is used as the studied case. Dynamic and economic performances show that the NMPC scheme are significantly improved. These demonstrates that the NMPC is a very feasible and effective scheme to control the three-product Petlyuk DWC.

References

[1] Dejanovic I., Matijasevic L. and Olujic Z.; Dividing wall column-A breakthrough towards sustainable distilling, Chemical Engineering and Processing, 2010, 49(6), 559-580.

[2] Kiss A. A., Landaeta S. J. F. and Ferreira C. A. I.; Towards energy efficient distillation technologies - Making the right choice, Energy, 2012, 47(1), 531-542.

[3] Triantafyllou C. and Smith R.; The design and optimisation of fully thermally coupled distillation columns: Process design, Chemical Engineering Research and Design, 1992, 70(A2), 118-132.

[4] Hart W. E., Laird C. D., Watson J.-P., Woodruff D. L., Hackebeil G. A., Nicholson B. L. and Siirola J. D.; Pyomo-optimization modeling in python, Vol. 67: Springer, 2017.

[5] Hart W. E., Watson J.-P. and Woodruff D. L.; Pyomo: modeling and solving mathematical programs in Python, Mathematical Programming Computation, 2011, 3(3), 219-260.

[6] Thierry D.; Nonlinear Optimization based frameworks for Model Predictive Control, State-Estimation, Sensitivity Analysis, and Ill-posed Problems: Carnegie Mellon University, 2019.

[7] Thierry D. and Biegler L. T.; Dynamic real-time optimization for a CO2 capture process, AIChE Journal, 2019, 65(7), e16511.