(4a) Identifying Energy-Efficient and Heat-Integrated Configurations for Multicomponent Distillation | AIChE

(4a) Identifying Energy-Efficient and Heat-Integrated Configurations for Multicomponent Distillation


Mathew, T. J., Purdue University
Tawarmalani, M., Purdue University
Agrawal, R., Purdue University
Multi-component distillation is one of the ubiquitous separation processes found in chemical and petroleum industries. Accounting for about 2.5% of the total energy use in the US [1], distillation is responsible for a significant part of the operating costs and greenhouse gas emissions in a chemical industry. Hence, it is an absolute necessity to explore possibilities by which the energy demand of the distillation configurations can be reduced.

One of the methods by which a reduction in energy consumption can be observed is through recycle of heat between condensers and reboilers. This mode of indirect transmission of heat (without mass transfer) is called as TCH (Thermal coupling via heat transfer) [2]. Since the feasibility and magnitude of the such heat integration links is strongly dependent on operating conditions, it is important that we consider them during the optimization of the configurations. The identification of such ‘beneficial’ and ‘realizable’ heat integration links requires a temperature predictor model. One major challenge is that detailed thermodynamic models for temperature prediction create convergence issues in the global optimization framework. In this research we use a simple to use equation, constructed on a theoretical basis, to predict temperatures of non-isothermal phase changing mixtures. We test our model on a few industrially relevant mixtures and validate the same.

Further, we incorporate our model into the global optimization framework to identify configurations with lowest vapor duty (a proxy for energy consumption of the configuration). To ensure finite heat transfer characteristics, we use the concept of MAT (Minimum Approach Temperature) in a heat exchanger. After adding the temperature predictor model and the MAT constraints, we demonstrate the power of our NLP (Non-Linear Programming Framework) by identifying energy-efficient and heat-integrated configurations for four and five component mixtures. These highly energy efficient configurations are simple to implement.


  1. S. Environmental Investigation Agency. Monthly Energy Review March 2021.
  2. Tony Joseph Mathew, Radhakrishna Tumbalam Gooty, Mohit Tawarmalani, and Rakesh Agrawal. A Simple Criterion for Feasibility of Heat Integration between Distillation Streams Based on Relative Volatilities. Industrial & Engineering Chemistry Research, 60(28): 10286–10302, 2021.