(258b) Process Conceptualization in a Carbon-Constrained World: Leveraging Optimization and Systems Engineering to Balance Multiple Objectives

Mathew, T. J., Purdue University
Jalan, A., ExxonMobil Research and Engineering
Goheen, C., ExxonMobil Research and Engineering
Kocis, G., ExxonMobil Research and Engineering
Narayanan, S., ExxonMobil Research and Engineering
Yang, L., ExxonMobil Research and Engineering
Tawarmalani, M., Purdue University
Agrawal, R., Purdue University
The search space for chemical process design can be remarkably complex and typically involves an interplay of both integer (e.g. what unit operations to consider?) and continuous decision variables (e.g. what operating conditions to run at?). Even for a seemingly simple unit operation like distillation, the number of potential configurations scales exponentially with the number of product streams (>80 million configurations for 6 products). As a result, it is no surprise that even seasoned practitioners rely heavily on heuristics and rules of thumb to “lock in” critical decision variables early on in the design process to keep the size of the search space tractable.

However, despite decades of experience, a heuristic-centric design philosophy can be overly restrictive (scanning a small part of the solution space) and lead to sub-optimal solutions. This is not only due to the highly non-linear nature of the search space but also because of the inherent complexity of handling multiple objective functions (emissions, economics, operational constraints etc.). Advances in computing and process systems engineering offer a route to systematically navigate the design space while accounting for critical design metrics and objectives.

This presentation will discuss the challenges in developing design tools for petrochemical separations focusing on distillation sequences. It builds on the pioneering work of Agrawal and co-workers (1-3) that relies on optimization models incorporating Underwood’s shortcut method to minimize the vapor duty of a configuration. Furthermore, we discuss approximations that enable comparison of stream temperatures using only information about the composition, component relative volatilities and pressure ratios of the streams, thus making it possible to identify energy efficient distillation configurations.

The approach outlined here has been successfully applied to separation schemes of interest to the petrochemical industry resulting in configurations that not only provide energy savings (reducing emissions and operating costs) but are also operationally feasible. Moreover the capability to provide multiple options gives process developers the ability to adapt quickly to evolving design objectives. These results provide the impetus for further efforts in developing tools that effectively leverage both heuristics and advanced optimization, and usher in a new paradigm in chemical process design.


  1. Agrawal, R., & Herron, D. M. (1997). Optimal thermodynamic feed conditions for distillation of ideal binary mixtures. AIChE journal, 43(11), 2984-2996.
  2. Nallasivam, U., Shah, V. H., Shenvi, A. A., Tawarmalani, M., & Agrawal, R. (2013). Global optimization of multicomponent distillation configurations: 1. Need for a reliable global optimization algorithm. AIChE Journal, 59(3), 971-981.
  3. Nallasivam, U., Shah, V. H., Shenvi, A. A., Huff, J., Tawarmalani, M., & Agrawal, R. (2016). Global optimization of multicomponent distillation configurations: 2. Enumeration based global minimization algorithm. AIChE Journal, 62(6), 2071-2086.