As a conventional unit operation based on vapor-liquid equilibrium with energy agent, distillation is often as a preferred method in the Petrochemical separation process, and synthesis of distillation columns sequences is also very important in process conceptual design. It has been proved that for a mixture separation the thermal couple configuration reduces the total energy consumption by 10 to 50% as compared to conventional systems. The synthesis of complex distillation column sequences has received particular attention because of the substantial energy savings that can be achieved through the use of complex columns instead of the more conventional simple column arrangements. Aggarwal and Floudas (1990) proposed a mixed integer nonlinear programming (MINLP) model by allowing the use of nonlinear cost functions for the synthesis of non-sharp splits, with the support of rigorous simulation, regression analysis was used to generate the parameters of nonlinear cost correlations. However, the superstructure of mixer-unit-splitter (M-U-S) is failed to attach any complex column substructure, only simple distillation column sequence is generated from this strategy. Yeomans and Grossmann (2000) proposed a generalized disjunctive programming (GDP) method for the rigorous design of complex column sequences, by means of the sequential column superstructure representation and overcoming numerical difficulties associated with common rigorous distillation models. The GDP techniques were also used in Caballero and Grossmann (2001), using a state task network (STN) formalism of the satellite column superstructure for the generation of simple and complex column sequences and using simple shortcut distillation models. In contrast to systems with conventional columns, the number of column sections is not fixed. Therefore, the trade off between the energy consumption and the extra number of column sections is happened. But it is difficult to capture all possible combination of different column sections under the framework of permanent and conditional tray. Through the Generalized Modular Framework (GMF), an alternative approach is proposed by Proios and Pistikopoulos (2005) for the systematic synthesis of complex column sequences, while avoiding potentially limiting simplifying assumptions. On principle of formal superstructure optimization techniques, although there are mix and/or split opportunities for a stream on upper and lower auxiliary block, the superstructure dos not supply any same opportunity on main module, so the model is incapable to evolve the connection structure between side draw and thermal couple. In addition, the more efficient configuration for adding or removing heat from an intermediate location of a thermally linked complex column is not taken into account in these previous works
In this work, the conception of Super Column (SC) is proposed based on state space (Bagajewicz and Manousiouthakis, 1992), a framework that takes multi-stream splitting and mixing possibilities into consideration, is presented for the design of complex distillation column sequences with both of external and internal connection. Super Column is a essential highly integrated architecture that has the ability to separate all desired products from the given feed by different segments of the body. More specifically, the overall complex distillation column sequence is viewed as a Super Column of two interconnected blocks (see Figure 1). One is referred to as the distribution network (DN), in which some mixers, splitters and the connections between them are embedded. The other is the described as process operator (PO), which can be further divided into a hierarchy of equilibrium stages with intermediate condenser and/or reboiler. There is obvious difference compared to the general structure of state space, the corresponding mixers and splitters are also arranged on the equilibrium stages. They either connect each other in PO for segment interior or connect to the junction of DN for segment exterior. All possible structure of complex thermal link and intermediate condenser and/or reboiler exchange match are merged into the Super Column. Depend on different design period, in which theoretical stage element (one phase equilibrium), segment unit (a series of connected independent stages) and super column system (a set of segments) are gradually produced and incorporated. Considering the Mass balance, Equilibrium relation, Summation check and Heat balances (MESH) for each of equilibrium stage and all mixers and splitters, the overall synthesis problem can be formulated as a mixed-integer nonlinear programming (MINLP), where the operating costs (including costs of cold and hot utilities) and equipment cost (including costs of distillation columns) are minimized simultaneously. Since (1) with the mechanism of arbitrary spit and/or mix in Super Column, the exchange opportunity of mass and heat between any stages is provided, and (2) the trade-offs between capital and operating costs can be properly carried out. So it is reasonable to expect that the TAC of the overall separation network can be reduced. In fact, from the comparison analysis of the solution data, not only the dramatic energy savings but also substantial capital reduced has been brought in our conceptual designs. A sample case study published in the literature was presented to illustrate the validity and advantages of the proposed approach.
Aggarwal, A.; Floudas, C. A. Synthesis of general distillation sequences - nonsharp separations. Comput. Chem. Eng. 1990, 14, 631.
Yeomans, H.; Grossmann, I. E. Optimal design of complex distillation columns using rigorous tray-by-tray disjunctive programming models. Ind. Eng. Chem. Res. 2000, 39, 4326.
Caballero, J. A.; Grossmann, I. E. Generalized Disjunctive Programming Model for the Optimal Synthesis of Thermally Linked Distillation Columns. Ind. Eng. Chem. Res. 2001, 40, 2260.
Proios, P.; Pistikopoulos, E. N. Generalized Modular Framework for the Representation and Synthesis of Complex Distillation Column Sequences. Ind. Eng. Chem. Res. 2005, 44, 4656.
Bagajewicz M, Manousiouthakis V. Mass/heat-exchange network representation of distillation networks. AIChE Journal. 1992; 38: 1769-1800.
Figure 1. The improved state-space superstructure