(3es) Numerical Method Development for Process Optimization and Design

I am a 5^th-year Ph.D. candidate in the Chemical Engineering department at Massachusetts Institute of Technology. After defending in the next year, I am interested in pursuing a postdoctoral research position focused on the development of new numerical approaches to modeling and optimization problems in chemical processes. My Ph.D. work as a member of the Process Systems Engineering Laboratory has focused on applying recent advances in nonsmooth equation solving to develop new methods for modeling complex chemical systems that are too challenging to address with traditional approaches. In particular, I have been collaborating with the Jorf Lasfar phosphate-processing facility in Morocco to ensure that these tools can be effectively used to direct innovation in an industrial setting. I enjoy being able to use a rigorous foundation to develop novel numerical approaches that can solve concrete problems, and I hope to also have these qualities in my future work. For example, I am interested in the potential to develop tailored optimization algorithms, potentially using nonsmooth methods, to solve scheduling and supply chain problems.

Ph.D. Work:

The focus of my Ph.D. research has been developing novel methods for process integration, which are used to minimize the reuse of resources, such as heat, water, or other materials, by simulating a system with optimal resource reuse. This project was motivated by improving sustainability for the Jorf Lasfar phosphate-processing platform, a large industrial park that consists of a complex network of different units, including three major process types and onsite desalination and power production. However, because current integration approaches require mixed-integer programs that scale poorly with the number of resource streams, they are unable to address a platform of this size and complexity. To address this problem, I developed a new, computationally tractable approach that utilizes recent advances in solving nonsmooth equation systems [1]. Unlike current approaches, this method requires only equation-solving methods, remains a compact system of two equations regardless of the number of resource streams, and can solve for any unknown process variable. I am now using this approach to design new systems to improve resource use at Jorf Lasfar, including a new system for power cogeneration from waste heat, and I am developing nonsmooth optimization methods to be able to minimize the implementation cost of this system. Through this project, I have developed skills in designing and implementing process models, combining innovation and rigor to develop new modeling and optimization methods, and applying these models and methods to ensure they can have concrete benefits in an industrial setting. I hope to bring these skills to my future work as I continue to learn about numerical method development across different problems and industries.

[1] C. J. Nielsen and P. I. Barton. Ind. Eng. Chem. Res. 59(1): 253-264, 2020.