## Large-Scale Optimization of Complex Separator and Reactor Networks

- Author(s): Ghougassian, Paul Gougas
- Advisor(s): Manousiouthakis, Vasilios
- et al.

## Abstract

The generation of globally optimal designs which can minimize capital and/or operating cost expenditures is a highly sought after objective within the chemical industry. A methodology which can systematically generate such globally optimal solutions to objective functions commonly encountered in the chemical industry is the IDEAS framework. The IDEAS framework decomposes a process network into an operator, OP network, where the unit operations (reactors, distillation columns, heat exchangers, etc.) occur, and a distribution, DN network, where the flow operations (mixing, splitting, recycling, and bypass) occur. The optimal process network structure is identified through solution of an infinite linear program (ILP) that is formulated within the IDEAS framework. The ILP's solution is approximated by finite dimensional linear programs of ever increasing size. The global optimization of complex, multi-pressure distillation networks for the separation of azeotropic mixtures using the IDEAS framework, is presented in chapter 1. The objective function in this case aims at minimizing total network flow in an effort to directly (indirectly) reduce capital (operating) costs. The global optimization of chemical reactor networks is presented in chapter 2-4. There, interesting properties relating to energy consumption and entropy generation for isothermal/isobaric reactor networks are described in the context of the attainable region (AR). Given certain assumptions, namely that all reactors are either of the endothermic or exothermic kind, a proof is presented that energy consumption and entropy generation can be rigorously identified in the infinite space of chemical reactors, independently of the network's internals (chapter 2). For the case of isothermal/isobaric chemical reactor networks where both endothermic and exothermic reactors participate in delivering the desired outlet product composition, entropy generation minimization is synonymous with an objective function of minimum hot/cold utility cost, with the cost coefficient of hot (cold) streams being the inverse of the temperature of the cold (hot) reservoir to which it adds (removes) energy. For this scenario, the network's internal structure plays a key role in determining the optimal reactor network, which is determined using the IDEAS framework (chapter 3). A novel method to identify the sequence of isothermal mixed flow reactors (CSTR's) which globally minimizes a reactor residence time dependent objective function (able to represent such objectives as capital cost, volume, or total annualized cost), subject to a constraint dependent on the reactor sequence's exit concentrations, is presented in chapter 4. Finally, chapter 5 discusses a novel, heat-integrated, pressure-temperature-swing-adsorption (PTSA) process for the capture of CO2 from the flue gas of fossil-fueled power plants using MgO sorbents.