(507a) Integrating CFD and Stochastic Optimization for High-Flux Membrane Design for Carbon Capture

Industrial CO₂ emissions are a primary contributor to global greenhouse gas emissions, underscoring the urgent need for innovative carbon capture solutions. Among the available technologies, membrane-based separation is emerging as a viable option due to its energy efficiency and scalability.

In this work, we introduce a superstructure optimization framework designed to streamline the development of multistage membrane processes for carbon capture. Our methodology employs reduced-order models, constructed from detailed membrane module simulations using Computational Fluid Dynamics, to capture essential operational parameters such as inlet flow rates, retentate and permeate pressures, and CO₂ concentration. These models rapidly predict key performance indicators like CO₂ recovery and permeate purity and are included in the superstructure mathematical program. The developed framework facilitates the efficient identification of both optimal process configurations and operating conditions of the multistage membrane process.

Recognizing the variability in feed compositions, our framework further integrates a two-stage stochastic programming approach with the superstructure optimization. This formulation minimizes the expected capture cost while accounting for the probabilistic distribution of key feed parameters, providing a more robust and realistic evaluation of process performance under uncertain operating conditions.

By combining superstructure optimization with stochastic analysis, this study offers a comprehensive tool for the systematic design of cost-effective and adaptable carbon capture systems. The proposed framework not only enhances process reliability across varying operational regimes but also supports the advancement of sustainable industrial practices, having the potential to be applied as a generalized framework to other similar technologies.