(460a) Multi-Objective Optimization for the Synthesis of Reliable and Inherently Safer Process Plants

The common approach to process safety relies on assessments made after the design phase is completed. Recently, the focus has changed to include these aspects in the early design stages. This new approach allows the generation of inherently safer designs as part of the process synthesis task [1]. Another consideration that has been explored recently is to include reliability as part of the process design. There are different approaches to consider for the improvement of the reliability of a chemical processing plant, one of them is the installation of an optimal number of backup units to prevent the shutting down of the plant due to the failure of critical units [2].

This work presents a multi-objective optimization model for the synthesis and design of chemical process systems accounting for safety, reliability, and economic aspects. The optimization model includes the selection of parallel standby units for critical process units to increase the system’s availability. The model also considers decisions regarding the structure of the process flowsheet and operating parameters. The trade-offs among the objectives arise as the selection of more standby units improves the availability of the system but at the expense of higher investment costs. Regarding safety, the presence of more units may increase the exposure to risk produced by the additional units; on the other hand, increasing availability reduces risk due to transition states (e.g. start-ups). Furthermore, the selection of the flowsheet structure and operating parameters generally gives rise to trade-offs between risk and economics; while extreme operating conditions may favor the efficiency of the process, operating under such conditions may increase the impact of hazardous events. Such outcomes are modeled through the application of quantitative risk assessment techniques [3]. The proposed multiobjective optimization model, which involves MINLP or Generalized Disjunctive Programming formulations, is applied to the design of air separation plants and to the synthesis of a flowsheet for methanol production. The results show that the proposed optimization model yields designs that exhibit optimal trade-offs between economics, reliability and safety.

References

[1] A.P. Ortiz-Espinoza, A. Jimenez-Gutierrez, M.M. El-Halwagi, 2017, Including inherent safety in the design of chemical processes, Ind. Eng. Chem. Res. 56, 14507–14517

[2] Y. Ye, I. Grossmann, J.M. Pinto, 2018, Mixed-integer nonlinear programming models for optimal design of reliable chemical plants, Comput. Chem. Eng. 116, 3-16

[3] AIChE, 2000, Guidelines for chemical process quantitative risk analysis