Process synthesis MINLP model

To illustrate some of the modeling guidelines, we consider in this section the MINLP model for the synthesis of heat-exchanger networks by Yee and Grossmann (1990), which uses as a basis the superstructure in Fig. 5. The notation that will be used is as follows. Process streams are divided into two sets, set... 

BaEban RC, Eha JA, Misener R, Floudas CA. Global optimization of a MINLP process synthesis model for thermochemical based conversion of hybrid coal, biomass, and natural gas to liquid fuels. Comput Chem Eng 2012 42 64-86. 

Mendez, C., Myers, J., Roberts, S., Logdson, J., Vaia, A., Grossmann, I., 2005. MINLP model for synthesis of paraxylene separation processes based on crystallization technology. In Puigjaner, L. (ed.), European Symposium on Gomputer Aided Process Engineering, Vol. 15. Elsevier, Amsterdam, the Netherlands. 

Szitkai, Z., Lelkes, Z., Rev, E., Fonyo, Z. (2002b). Handling of removable discontinuities in MINLP models for process synthesis problems, formulations of the Kremser equation. Computers Chem. Engng., in press. 

Gong, J., You, R, 2014. Optimal design and synthesis of algal biorefmery processes for biological carbon sequestration and utilization with zero direct greenhouse gas emissions MINLP model and global optimization algorithm. Ind. Eng. Chem. Res. 53, 1563-1579. 

Another approach that addresses the reduction of size of the MINLP is the state space approach by Bagajewicz and Manousiouthakis (1992). The basic idea of this strategy is to partition the synthesis problem into two major subsystems, the distribution network and the state space operator. The objective in the former is to make the decisions related to the distribution of flows in the superstructure, while the objective in the latter is to perform the optimization for the decisions selected in the distribution network. At the level of the state space operator one can consider the process either in its detailed level or simply as a pinch-based targeting model. While this strategy has the advantage of reducing the size of the MINLP, it is unclear how to develop automated procedures based on this approach. 

As was described in the review of previous work, over the last ten years MINLP optimization models have been reported for the synthesis of process flowsheets, heat-exchanger networks, separation sequences, reactor networks, utility plants, and design of batch processes. Rather than describing in detail each of these works, we will briefly highlight several examples from our research group at Carnegie Mellon to illustrate the capabilities and the current limitations of the MINLP approach. 

---