With rising carbon emissions and increased focus on efficiency, the synthesis and design of energy-efficient separation processes is a major challenge for a practicing engineer. All chemical industries require separation units to meet product quality, to recover harmful chemicals, etc. Quite often for a multicomponent separation, a sequence of separation devices based on distillation, membrane, adsorption, etc. are used. Some examples include crude petroleum distillation, ethylene recovery, and air separation. However, separation in general is an energy intensive process. It is estimated that separation processes account for 40 – 70% of chemical plant costs (Humphrey and Keller, 1997). In addition, for many separations like crude distillation, the energy input is sourced almost entirely from the combustion of fossil fuels; which results in significant contributions to carbon emissions. The energy consumption of such separation processes is heavily dependent on how the separation devices are sequenced or configured. Suboptimal sequences are known to result in energy penalty in excess of 50% (Shah and Agrawal, 2010). Therefore, it is essential to develop easy-to-use methods that will identify optimal separation sequences leading to large energy savings.
Our research is aimed at developing a systematic approach to generate energy-efficient separation schemes for several applications; with the current work using distillation and membranes as our technologies for separation. Rather than focusing on modifying design features such as sizing and number of trays, we work on the broader picture of skeletal structures. The first step is the synthesis of a search space of novel separation schemes (Shah and Agrawal, 2010; Shenvi et. al., 2013; Ramapriya et. al., 2018). After the search space is generated, we build and execute optimization models to identify which among the numerous skeletal structures is energy-efficient and attractive for a desired separation (Nallasivam et al, 2016; Gooty et al, 2019); where the objective can be minimizing energy consumption, exergy loss, or total cost. We work on incorporating various features in our model such as heat integration and impurities, and improving the optimization formulation to be able to converge to global optimality. Our goal is to provide easy-to-use tools for the synthesis of optimal separation process while considering all possible separation unit operations.