Synthesis of heat-integrated distillation sequences with mechanical vapor recompression by stochastic optimization
Introduction
Distillation has been the most widely used industrial separation technic with energy consumption up to 40–60% in the process industries (Humphrey and Siebert, 1992; Sholl and Lively, 2016), while its thermodynamic efficiency is usually less than 10% (Jana, 2010; Kiss et al., 2012). The synthesis of distillation sequences is an important and effective method to improve the energy and economic efficiency for multicomponent separation processes due to the significant economic differences between different separation sequences. And various models and methods for synthesis of distillation sequences constitute one of major achievements in chemical process systems engineering.
The studies of synthesis of distillation sequences started with simple models of sharp-split sequence problems (Qian and Lien, 1995; Thompson and King, 1972; Stephanopoulos and Westerberg, 1976). And then, increasingly more comprehensive modeling methods were proposed to realize the synthesis of distillation sequences with sharp or nonsharp-splits (Yeomans and Grossmann, 2000; Zhang and Linninger, 2006; Wang et al., 2016), and heat integration and/or thermal coupling were taken into account to obtain more significant reduction in energy consumption for multicomponent distillation separation (Yeomans and Grossmann, 1999b; Caballero and Grossmann, 2004, 2011, 2013; Zhang et al., 2018).
In addition to considering possible heat integration and thermal coupling in optimal synthesis of distillation sequence, mechanical vapor recompression (MVR) is also a practical energy saving technology, which can realize the energy reutilization by consuming a certain amount of compression work to augment the energy level of the system (Jana, 2014). MVR is often used in distillation processes, mostly appearing as heat pump distillation (Díez et al., 2009; Luo et al., 2015; You et al., 2016), but rarely considered simultaneously in the synthesis of distillation sequences with the other options like heat integration and thermal coupling. One of the reasons underlying the exclusion of the MVR from synthesis of distillation sequences is that its economic benefit may not be necessarily significant when the use of traditional thermal energy is relatively cheaper than the use of electricity, which is often accompanied by additional capital cost for compressors as well. However, in the context of global warming and under the pressure of carbon emission reduction, it becomes quite attractive to use green electricity to partially or even totally replace thermal energy that is produced by fossil fuel for distillation separation. Therefore, it becomes important to include MVR in the structure alternatives in the synthesis of distillation sequences to increase the use of electricity and limit the use of the traditional thermal energy. Another reason of not considering the MVR in distillation sequencing is that the size of the search space of the synthesis problem would increase dramatically if the MVR option is included and result in a large scale mixed-integer nonlinear programming (MINLP) problem that is difficult to model and solve effectively. These are why the application of MVR has been studied only for existing distillation configurations to improve energy efficiency and most for binary mixture separations, and, as a result, the problem of synthesizing MVR distillation sequences still remains open and challenging. For solving large scale complex distillation sequencing problems, several methods have been developed.
Mathematical programming is one of the most widely used methods, the synthesis problem is usually formulated as a mixed-integer linear programming (MILP) or MINLP problems based on superstructure, and deterministic algorithms were often used to solve the optimization problem (Biegler and Grossmann, 2004). Leeson et al. (2017) proposed a method for preliminary design of sharp heat-integrated distillation sequences based on MILP model. Yeomans and Grossmann (1999a) introduced a nonlinear generalized disjunctive programming method to overcome the difficulty in formulating the synthesis of sharp heat-integrated distillation sequences into an MILP problem. Caballero and Grossmann (2014) presented a MILP a two-level superstructure to solve a complex problem of synthesis thermally coupled distillation sequences. As demonstrated by many authors, the MINLP model for distillation sequencing is combinatorial in nature and sometimes non-convex. Gooty et al. (2019) solved an MINLP problem they proposed for thermally coupled distillation sequences using a branch-and-bound algorithm and decomposition strategy (Novak et al., 1996; Floudas and Gounaris, 2008), in fact, however, large-scale combinatorial problem is inevitable and a real challenge for all the deterministic mathematical programing methods to find a global optimum, even with the most advanced solvers available today.
The matrix method is another effective approach for the synthesis of distillation sequences. Shah and Agrawal (2010) proposed a matrix method for representing all basic configurations (distillation sequence with N–1 columns) and feasible thermally coupled distillation sequences. Nallasivam et al. (2016) and Jiang et al. (2019) applied this method to generate all distillation sequences for a given mixture, and conducted the optimization using enumeration based approach. Hou and Luo (2019) expanded the matrix method for distillation from configurations with N–1 to those with less than N–1 columns to include more advanced sequence configurations. The advantage of both the original matrix method and the improved matrix method is that it is easy to generate sequences, however, they are a kind of enumeration-based approach and heat integration was not considered in the synthesis of distillation sequences.
Compared with the above methods, stochastic algorithms can overcome the computational difficulties of deterministic optimization algorithms, as they can search the optimal solution automatically based on the stochastic optimization strategy in an evolutionary process, which allows stochastic optimization algorithms to efficiently solve the mathematical model with considering the non-linearity and non-convexity. The stochastic optimization algorithms can avoid numerous local optima and obtain the global optimum with high probability (Wang et al., 2008). Stochastic optimization methods have been shown powerful for solving large scale complex optimization problems for distillation sequence synthesis (Wang and Li, 2008; Jain et al., 2012). In a recent attempt, made by Zhang et al. (2021), a novel stochastic optimization method based on simulated annealing (SA) and particle-swarm optimization (PSO), namely SA-PSO method, was developed for automatic synthesis of large-scale nonsharp-split distillation sequence considering heat integration and thermal coupling. An efficient coding method was proposed in this method and enabled effective representations for all the alternative sequences possibly containing simple and complex distillation columns, heat integrations, and thermal couplings. The SA was used in an outer circle to explore sequence configurations by varying the codes and the PSO was used in an inner loop to optimize a sequence identified by the SA in each circle. The most promising feature of the SA-PSO is that, due to its coding technology and stochastic optimization strategy, it can easily incorporate energy-saving technologies such as heat integration and thermal coupling into the optimization framework, so that the continuous operating variables of columns and the discrete variables corresponding to distillation sequences and thermally coupled structures can be optimized simultaneously.
In the present work, a method of the synthesis of heat-integrated distillation sequences with mechanical vapor recompression (MVR-HIDSs) is proposed based on stochastic optimization method. In order to optimize MVR structures, a potential compressor ratio for each distillation column is introduced as a continuous optimization variable. The synthesis problem is formulated as an implicit MINLP problem, and solved by the SA-PSO algorithm to minimize the total annual cost (TAC) of the system. An example case of synthesizing five-component MVR-HIDSs is presented to validate the proposed method and demonstrate the necessity of involving MVR in synthesis of distillation sequences. The result proved that the distillation sequences with MVR have better economic efficiency compared with those sequences with only heat integration or thermally coupling considered, and our models can get the optimal sequence with high probabilities in a reasonable computing time.
Section snippets
Problem description
Given a zeotrope mixture of N components (N ≥ 3), the objective is to synthesize an optimal sharp or non-sharp MVR-HIDSs to obtain N pure components with the minimum TAC. To reasonably reduce the complexity of the problem, only basic configurations with N–1 columns are considered in the synthesis (Giridhar and Agrawal, 2010). Additional assumptions and specifications for the problem are as follows:
- (1)
Pressure drop and heat loss of heat exchangers are ignored.
- (2)
Throttling process is considered to be
Optimization framework
To optimize the MINLP problem and obtain the optimal MVR-HIDSs, we extend the SA-PSO optimization method proposed by Zhang et al. (2021) by adding optimization variable CRi for each column i as continuous variable. In the optimization method, the SA algorithm is used to optimize process structural parameters, namely the distillation sequence, while the PSO algorithm is used to optimize operating parameters to minimizes the TAC for a given sequence. The number of SA iterations is determined by
Illustration with five-component separation distillation process synthesis
In order to verify the advantages of the proposed MVR-HIDS synthesis, a five-component alcohol separation example which was addressed by Zhang et al. (2021) is studied for both the VR-HIDS and BF-HIDS cases. The five-component alcohol mixture contains (A) ethanol, (B) i-propanol, (C) n-propanol, (D) i-butanol, and (E) n-butanol. The feed mole fractions, relative volatilities of all components, and utility data are presented in Table 1. The feed is a saturated liquid with a flow rate of
Conclusion
In the present study, we proposed a systematic method to synthesize sharp or nonsharp-split MVR-HIDSs based on the SA-PSO stochastic algorithm. The purpose is to achieve the optimization for synthesis of distillation sequences not only considering heat integration, but also the MVR to increase economic efficiency of the processes where electricity can be used to replace partially or even totally the use of the traditional thermal energy. A set of compression ratios was introduced as new
CRediT authorship contribution statement
Haiou Yuan: Conceptualization, Methodology, Software, Validation, Formal analysis, Data curation, Writing – original draft, Visualization. Yiqing Luo: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Xigang Yuan: Supervision, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
This research was supported by the National Natural Science Foundation of China (No. 21978203).
References (38)
- et al.
Retrospective on optimization
Comput. Chem. Eng.
(2004) - et al.
Design of distillation sequences: from conventional to fully thermally coupled distillation systems
Comput. Chem. Eng.
(2004) - et al.
Optimal synthesis of thermally coupled distillation sequences using a novel MILP approach
Comput. Chem. Eng.
(2014) - et al.
Economic feasibility of heat pumps in distillation to reduce energy use
Appl. Therm. Eng.
(2009) - et al.
Synthesis of distillation configurations: I. Characteristics of a good search space
Comput. Chem. Eng.
(2010) - et al.
Synthesis of heat-integrated distillation sequence systems
J. Taiwan Inst. Chem. Eng.
(2012) Heat integrated distillation operation
Appl. Energy
(2010)Advances in heat pump assisted distillation column: a review
Energy Convers. Manag.
(2014)- et al.
Global optimization of multicomponent distillation configurations: global minimization of total cost for multicomponent mixture separations
Comput. Chem. Eng.
(2019) - et al.
Towards energy efficient distillation technologies – making the right choice
Energy
(2012)
Simultaneous design of separation sequences and whole process energy integration
Chem. Eng. Res. Des.
Simultaneous synthesis of distillation sequences in overall process schemes using an improved minlp approach
Comput. Chem. Eng.
Rule based synthesis of separation systems by predictive best first search with rules represented as trapezoidal numbers
Comput. Chem. Eng.
Studies in process synthesis—II: evolutionary synthesis of optimal process flowsheets
Chem. Eng. Sci.
Synthesis of heat-integrated complex distillation systems via genetic programming
Comput. Chem. Eng.
A formulation methodology for multicomponent distillation sequences based on stochastic optimization
Chin. J. Chem. Eng.
Nonlinear disjunctive programming models for the synthesis of heat integrated distillation sequences
Comput. Chem. Eng.
A systematic modeling framework of superstructure optimization in process synthesis
Comput. Chem. Eng.
Reducing process cost and CO2 emissions for extractive distillation by double-effect heat integration and mechanical heat pump
Appl. Energy
Cited by (7)
Optimal synthesis of heat-integrated distillation configurations using the two-column superstructure
2024, Chinese Journal of Chemical EngineeringEnhanced heat-integrated distillation processes: Simultaneous synthesis via structural modification utilizing the matrix method
2023, Journal of Cleaner ProductionEconomic and environmental assessment of heat pump integration in thermally coupled hybrid separation
2023, Chemical Engineering JournalMultiple utilities targeting in energy integration considering rigorous temperature-enthalpy relations
2023, Chemical Engineering ScienceSynthesis of Heat-Integrated Distillation Sequences with Nonsharp Splits Using a Sequential Metaheuristic Method
2024, Industrial and Engineering Chemistry Research