Research ActivitiesThe research program in the Systems Optimization and Multiscale Analysis group is directed towards making fundamental advancements in the area of Process Systems Engineering with applications to energy, the environment and sustainability. As such, our work lies at the interface of chemical engineering, applied mathematics/operations research, and computer science. The unified thrust of our research is to address fundamental problems and application areas through detailed mathematical modeling at the microscopic, mesoscopic and/or macroscopic level, rigorous optimization theory and algorithms, and large-scale computations on high performance clusters. We also develop scientific tools and software prototypes based on our research. Visit our Software and Webtools page for more information. Optimization Theory and AlgorithmsWe focus on developing theoretical, algorithmic and computational methods for both gradient/derivative-based and derivative-free optimization. In the realm of gradient-based optimization, we are primarily interested in global optimization methods for nonconvex problems. In the realm of derivative-free optimization, we are interested in constrained black-box, grey-box and/or simulation-based optimization using novel techniques such as univariate projection and lower-envelope optimization. Specially, black-box problems are challenging optimization problems due to their complexity, uncertainty, and our inability to generate accurate data using physical or computer experiments in reasonable time. Complex systems can be chemical, biochemical or biological. Examples vary from enzyme complexes to biological cells to process plants to supply chains. We are currently working to address many fundamental challenges which need to be resolved for accurate model prediction, optimization and control of multi-scale and complicated systems. We develop models which are "efficient enough" and at the same time "accurate enough". We also address the critical question on how to establish bounds solely based on input-output data toward optimizing complex systems to global optimality. Selected References
Systematic Process IntensificationProcess intensification refers to any dramatic improvement in process performance in terms of equipment sizes, waste generation, productivity or other factors. Despite the past research, there is a lack of systematic methods for the identification, creation and selection of intensification alternatives without exhaustive or sequential enumeration. It also remains a challenge to systematically identify the best combination of phenomena, functions, enabling materials and their arrangement for process intensification. In our group, we are developing an original method for systematic process design, synthesis and intensification, which does not depend on a priori postulation of process alternatives. We depart from the classical unit operation-based representation of process units, flowsheets and superstructures and propose a new representation using building blocks. The systematic arrangement of building blocks enables us to incorporate all plausible design and intensification pathways using a general block-based superstructure. It also enables a mathematical programming-based optimization method to simultaneously identify the best design, synthesis and intensification routes at the equipment and flowsheet levels. Funding
Selected References
Multiscale Energy Systems EngineeringOften the discovery and design of novel pathways for sustainable energy, environment, and health depend on addressing systems which are complex and vary across different time- and length-scales. An example of complex and multi-scale system is the CO2 capture, utilization and sequestration (CCUS) system where the overall cost of CO2 capture depends on multiple and often contradicting factors at materials-, process- and supply chain-levels. We are interested in developing and applying optimization techniques for the design and discovery of advanced materials and processes for clean energy, carbon capture and utilization, and sustainable fuels and chemicals from alternative feedstocks. All these address different facets of national security related to energy and climate change. The unified thrust of our research is to address fundamental problems through detailed mathematical modeling, rigorous optimization theory and algorithms, large-scale computations and experimental validation. We elucidate and combine materials-centric, process-centric and systems-centric understanding for the development of disruptive and transformative technologies for future energy. Funding
Selected References
Clean Energy: Carbon Capture, Utilization, and Storage (CCUS)As an application of multiscale systems engineering, we extensively work on CCUS systems towards discovering enabling technologies for producing clean energy from fossil fuels. While material selection is crucial for CO2 capture, the industrial scale deployment of CCUS requires material-, process- and supply chain network-level developments. Our work in CCUS extends across all three levels. At the materials level, we seek to identify the best materials (e.g., best amines for absorption, best polymeric and zeolitic materials for membrane, best zeolites and metal organic frameworks for adsorption) from the universe of materials for CO2 capture using multi-scale systems approach. At the process level, our research involves a detailed cost-based comparison of different technologies for post-combustion capture to elucidate the trade-offs between various costs related to capture. At the supply chain network level, the goal is to develop a nationwide CCUS structure that would use the most appropriate source plants, capture technologies and materials, transportation networks, and CO2 utilization and storage sites. A major question that we ask is whether (and how) it is possible to reduce 50% of the stationary CO2 emissions in the U.S. by CCUS in the most cost-effective manner. Funding
Selected References
Sustainability of Unconventional Energy, Fuels and Feedstocks |