Rapid design and deployment of advanced process systems is necessary to address growing energy and water demands. Conventional methodologies for process design are focused on economies of scale, optimizing each installation uniquely. While effective for a custom design and installation, it can be inefficient when designing and deploying multiple instances across a variety of installation sites with different ambient conditions, fluctuating demands, etc. To leverage the advantages of reduced manufacturing costs and shared engineering effort through economies of numbers, we have developed a “Process Family Design” approach (Stinchfield et al., 2024). The problem formulation simultaneously designs a large “family” of process variants using a common platform of shared components. A major characteristic of the approach is the concurrent optimization of the shared component platform in conjunction with the process family. The formulation is a fundamentally complex, large-scale mixed integer nonlinear programming (MINLP) problem involving nonlinear expressions that describe the physics, performance, and cost of each process variant in the family. It includes binary variables that capture the distribution of platform components to the different process variants across the family. Previous work (Stinchfield et al., 2024), employed a discretization technique to transform the MINLP into a mixed-integer linear programming (MILP) problem; however, this approach required a significant number of simulations a priori to formulate the problem. In this work, we address the solution of the MINLP directly. Off-the-shelf global optimization solvers are not able to solve the problem at scale, and we present a decomposition approach that exploits the inherent block-angular structure of the problem and apply a reduced-space global optimization algorithm for nonlinear stochastic programs proposed by Cao and Zavala (Cao and Zavala, 2019). We decompose the larger process family into smaller family subproblems with coupling constraints that ensure that the common platform is the same across all the subproblems. Here, we present the process family design formulation that allows us to embed equation-oriented process models directly, and a computational performance comparison of the decomposition approach with an off-the-shelf global optimization solver.
Acknowledgement
This work was conducted as part of the U.S. Department of Energy’s Institute for the Design of Advanced Energy Systems (IDAES) supported by the Office of Fossil
Energy and Carbon Management’s Simulation-based Engineering Research Program.
Disclaimer
This project was funded by the Department of Energy, National Energy Technology Laboratory an agency of the United States Government, through a support contract. Neither the United States Government nor any agency thereof, nor any of its employees, nor the support contractor, nor any of their employees, makes any warranty, expressor implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof, or any of their contractors.
References
G. Stinchfield, J. C. Morgan, S. Naik, L. T. Biegler, J. C. Eslick, C. Jacobson, D. C. Miller, J. D. Siirola, M. Zamarripa, C. Zhang, and et al., “A mixed integer linear programming approach for the design of chemical process families,” Computers & Chemical Engineering, vol. 183, p. 108620, Apr 2024.
Y. Cao and V. M. Zavala, “A scalable global optimization algorithm for stochastic nonlinear programs,” Journal of Global Optimization, vol. 75, p. 393–416, Apr 2019.
