(235c) Novel Modular Design and Optimization Framework for Intensified Membrane Reactor Systems

The advancements in the design of intensified processes have brought about improvements in the efficiency and reduction in the size of many chemical processes. Although the primary objective of these design improvements has been to enhance the steady-state design performance, many control challenges arise when a process becomes intensified, notably a loss in the number of degrees of freedom when compared to a traditional process. This loss in degrees of freedom results in traditionally square control problems becoming underdefined, requiring set point control of some variables and set interval control of the remaining variables [1]. This suggests the traditional approaches to design are insufficient and that “process design, operation and control should be considered simultaneously, or in other term, they should be fully integrated” [2]. This observation was later confirmed in the literature by a rigorous justification [3]. The loss in degrees of freedom can be attributed to the interdependence between the multiple phenomena (i.e., reactions, separations, heat transfer, etc.) that occur simultaneously in intensified units. One approach to addressing this challenge can be to utilize a more complex control strategy, however the system is still subject to the described interdependence of the phenomena in the process. Therefore, the challenge of reduced degrees of freedom for intensified processes requires a novel approach to designing equipment that considers the control challenges during the design phase and utilizes a method that addresses the interdependence of phenomena in the process unit.

This work seeks to produce a novel design framework for intensified membrane reactor units to address the loss of degrees of freedom challenge in intensified process design and control. In this framework, the system model is constructed through the assembly of multiple modules (reactor, membrane separator, and membrane reactor) using the AVEVA SimCentral Simulation Platform [4] to allow for the decoupling of phenomena, thus facilitating the assessment of this problem. Concepts of the process operability approach [1, 5], such as the operability index, will be used as the assessment metric in conjunction with a mixed-integer nonlinear optimization (MINLP) formulation. The potential benefits of such a modular design approach in addressing operational challenges introduced through process intensification has been demonstrated in previous work [5]. To formulate the MINLP problem, each module is broken down into the base phenomena that occur in it (i.e., reactors have heat transfer and reactions, membrane reactors have heat transfer, reactions, and permeation, etc.) and assigned an array of integers to express the presence or absence of each phenomenon. Reformulating the problem in this way converts the MINLP to a constrained NLP problem as each phenomenon now can be fractionally expressed and the problem can be solved using gradient-based methods. The optimizer then adjusts these fractional values to maximize the operability index of the unit. However, the size of this proposed module assignment problem grows exponentially as the number of modules needed to optimize the operability index increases. To address this challenge, the optimizer performs an initialization step where optimal solutions to smaller versions of the module assignment problem are utilized as the initial guess in the optimization of the larger-sized problems. This step provides a warm start for the optimizer, thus decreasing the computational time required to search the design space for an optimum.

This approach to the design of membrane processes is part of a movement in the literature for new design paradigms specifically created for addressing the challenges through process intensification [3, 6–8]. Results of this approach show how the strategic coupling and decoupling of phenomena using the novel MINLP algorithm can maintain the efficiency desired from process intensification while simultaneously improving the process operability of the membrane reactor system.

References

[1] F. V. Lima, C. Georgakis, J. F. Smith, P. D. Schnelle, and D. R. Vinson, “Operability-based determination of feasible control constraints for several high-dimensional nonsquare industrial processes,” AIChE J., vol. 56, no. 5, pp. 1249–1261, 2010, doi: 10.1002/aic.11897.

[2] N. M. Nikačević, A. E. M. Huesman, P. M. J. Van den Hof, and A. I. Stankiewicz, “Opportunities and challenges for process control in process intensification,” Chem. Eng. Process. Process Intensif., vol. 52, pp. 1–15, Feb. 2012, doi: 10.1016/j.cep.2011.11.006.

[3] M. Baldea, “From process integration to process intensification,” Comput. Chem. Eng., no. 81, pp. 104–114, Mar. 2015, doi: 10.1016/j.compchemeng.2015.03.011.

[4] AVEVA, “SimCentral Simulation Platform: Process Simulation Reinvented.” AVEVA, 2019, Accessed: Aug. 08, 2019. [Online]. Available: https://sw.aveva.com/hubfs/pdf/datasheet/Datasheet_SE-LIO-SimSci_SimCent....

[5] B. A. Bishop and F. V. Lima, “Modeling, simulation, and operability analysis of a nonisothermal, countercurrent, polymer membrane reactor,” Processes, vol. 8, no. 1, p. 78, Jan. 2020, doi: 10.3390/pr8010078.

[6] K. P. Papalexandri and E. N. Pistikopoulos, “Generalized modular representation framework for process synthesis,” AIChE J., vol. 42, no. 4, pp. 1010–1032, 1996.

[7] J. A. Arizmendi-Sánchez and P. N. Sharratt, “Phenomena-based modularisation of chemical process models to approach intensive options,” Chem. Eng. J., vol. 135, no. 1, pp. 83–94, Jan. 2008, doi: 10.1016/j.cej.2007.02.017.

[8] P. Lutze, D. K. Babi, J. M. Woodley, and R. Gani, “Phenomena based methodology for process synthesis incorporating process intensification,” Ind. Eng. Chem. Res., vol. 52, no. 22, pp. 7127–7144, Jun. 2013, doi: 10.1021/ie302513y.