(595a) Design and Control of A High Purity N-Heptane Distillation System

Normal heptane (n-heptane) is a high volume chemical that is widely used as a solvent for the purification of numerous pharmaceutical products and as a solvent for electroplating. The industrial production of high purity n-Heptane is usually obtained from a multi-component feedstock using an energy intensive separation process such as distillation. In this case study, a particular existing multi-component hydrocarbon distillation system will be used as the demonstration platform to illustrate different control strategies and to show the coupling between design and control to meet the end objective of very high purity n-heptane.

The plant design and control tasks are usually carried out in series. The design is done to satisfy steady-state objectives of product quality and throughput while control structures maintain the process within dynamic constraints of hard actuator and sensor limits, safety, and environmental regulations. The separation of the two tasks may result in a process design that limits the control ora control strategy that never performs satisfactorily. Even if a control strategy can be generated, it has been shown that the control strategy is in general, sub-optimal when the tasks of design and control are carried out independently [1]. These factors motivate the study of the coupling between design and control of chemical plants.

There are some studies that address this issue using open-loop controllability measures [2] and others that focus on solving a design and control optimization problem simultaneously [3]. Recently, Shabde [4] demonstrated a bi-level optimization strategy of Ulsoy and coworkers [1,5,6] on the design and control of a highly nonlinear spray dryer that produces micro-particles. Shabde was able to demonstrate the coupling between the optimal design and the optimal controller and to compare the performance between the bi-level and serial solutions. This works builds on the results of Shabde [4] in that while the spray dryer is as single unit operation, the n-heptane distillation column of interest is in reality an integration of two columns with limiting instrumentation and constrained operations. This study will employ simulation and real operational data to compare different control solutions for different feeds and purity specifications and will attempt to quantify the coupling between design and control using the definition of optimal coupling as provided by [1]. Additionally, optimal column and control designs will be generated and their performance tested. Other aspects that will be investigated include the effect of tighter operational constraints and parameter uncertainty on the coupling and achievable control performance.

Literrature Cited

[1] H.K. Fathy, J.A. Reyer, P.Y. Papalambros, and A.G. Ulsoy. On the coupling between plant and controller optimization problems. In American Control Conference, 1864?1869, 2001.

[2] M.L. Luyben and C.A. Floudas. Analyzing the interaction of design and control- I & II. Computers &Chemical Engineering, 18:933?993, 1994.

[3] B.M. Russel, J.P. Henrikson, S.B. Jørgensen, and R.A. Gani. Integration of design and control through model analysis. Computers & Chemical Engineering, 26:213?225, 2002.

[4] V.S. Shabde. Optimal design and control of a spray drying process that produces hollow polymeric particles. Doctor of philosophy, Texas Tech University, Lubbock, TX, 2006.

[5] G.A. Brusher, P.T. Kabamba, and A.G. Ulsoy. Coupling between modeling and controller-design problems-part i: Analysis. J. of Dynamic Systems, Measurement and Control, 19(3):498?512, 1997.

[6] G.A. Brusher, P.T. Kabamba, and A.G. Ulsoy. Coupling between modeling and controller-design problems-part ii: Design. J. of Dynamic Systems, Measurement and Control, 19(2):278?283, 1997.