Multi-scale Optimization Frameworks for Integrated Process and Material Design and Intensification
Abstract
Meeting energy and chemical production demands while reducing costs and emissions is a grand challenge. Intensified processes which merge multiple tasks while maintaining performance can significantly reduce equipment footprint, energy input and costs. Effectively designing such processes requires balancing competing trade-offs on multiple levels.
A multi-scale framework is developed for simultaneous consideration of operational and material decisions by posing the intensified process design problem as an optimization formulation. Models and constraints related to process operations, process performance, product quality and material properties are incorporated into the framework.
The framework is applied to intensify the separation and storage of methane (CHv4) from feed-stocks by exploiting the preferential adsorption properties of zeolites. However, meeting constraints on CHv4 loss and purity while maximizing the storage capacity is a challenge requiring consideration of both process and material decisions. The complete dynamic process model and constraints along with adsorption isotherm models are posed as a nonlinear programming (NLP) problem. Adsorption isotherm data on 178 siliceous zeolite frameworks are obtained using Grand Canonical Monte Carlo (GCMC) simulations. An initialization strategy is developed to aid in optimizing the model using which the top candidate zeolites and their corresponding process conditions are determined for different feed compositions. The analysis is extended to obtain target material property maps by extensively sampling the material property space (Henry coefficient, deliverable capacity, isotherm parameters) using a Latin Hypercube based strategy. Data from publicly available zeolite databases are super-imposed onto these maps to identify the top zeolite structures for process performance and feasibility.
Another application studied is the design of a process to integrate COv2 capture and syngas pro-duction using methane feedstocks. The energy intensive periodic pressure changes employed for adsorbent-based COv2 capture are avoided by using a CHv4 rich purge feed to strip the adsorbed COv2 which then becomes feed for syngas production. A data-driven constrained optimization algorithm is applied to identify process conditions which satisfy process specifications and product quality requirements and to determine optimal process decisions for different objectives and feedstocks. The importance of the multi-scale optimization approach in designing novel intensified processes is demonstrated through these applications.
Subject
Multi-scaleOptimization
Integrated Process and Material Design
Inverse Design
Nanoporous Adsorbents
Separation
Citation
Iyer, Shachit Shankaran (2019). Multi-scale Optimization Frameworks for Integrated Process and Material Design and Intensification. Doctoral dissertation, Texas A & M University. Available electronically from https : / /hdl .handle .net /1969 .1 /184957.