1. Introduction
Amid intense industrial growth, the inherent interconnectedness among water consumption, energy usage, and carbon emissions has emerged as an important role in discussions regarding water scarcity as well as climate changes, including global warming. As these issues become more urgent, there is an increasing need for innovative safety practices, advanced equipment, and the adoption of renewable raw materials and energy sources. These factors are essential to achieving sustainable development and mitigating the adverse impacts of industrial activities on both the environment and society [
1].
In this regard, environmental concerns have prompted the search for cleaner technologies that maintain productive competitiveness [
2]. Thus, process intensification (PI) has gained prominence, aiming to promote greater efficiency, enhanced safety operational conditions, and sustainability in various industrial segments [
3,
4]. A common definition for PI is the development of new techniques that, compared to the conventional arrangement, improve production performance [
4].
As a result, energy-intensive equipment with high capital investment and significant operating utility costs, such as distillation columns [
3,
4], has been the subject of numerous research efforts aimed at optimizing their performance [
2]. These studies focus on improving energy and water efficiency, also aiming to reduce CO
2 emissions and other environmental impacts [
5,
6]. The goal is to enhance process eco-efficiency, which is the efficiency related to raw material consumption, safety procedures, product quality, and environmental impacts, in addition to the traditional financial income [
3,
5,
6,
7,
8].
In this context, the dividing wall column (DWC) represents a significant advancement in multicomponent mixture separation as an important PI technology, with savings ranging from 30 to 40% in capital and operational expenditures, as well as a reduction of approximately 40% in the installation area as an extra benefit [
3]. Reducing or eliminating the remixing effect is the main factor contributing to increased energy efficiency [
4,
9]. The DWC is a modification of the Petlyuk layout in which the prefactionator and main distillation column are combined into a single tower with a vertical wall dividing it into two portions [
4]. This technology may also be used for extractive distillations, azeotropic separations, and reactive distillations, resulting in lower capital and operational costs [
3,
4]. A comprehensive review of DWC technology, including design operational parameters, advantages, and limitations, as well as industrial applications worldwide, was published by Segovia-Hernandez [
4].
Regarding industrial operations, one of the processes that can associate economic progress with sustainability is the separation of the benzene–toluene–xylene (BTX) system. This aromatic stream is often produced as a result of the catalytic reforming of naphtha, pyrolysis gasoline from steam cracking, and aromatization of liquefied petroleum gas (LPG) [
10]. These chemicals’ major applications include the synthesis of petrochemical products and other organic compounds such as ethylbenzene, benzoic acid, maleic anhydride, and monochlorobenzene, among others [
11]. As it is an important and large-scale chemical plant, its high-energy demand encourages the use of intensification technologies that have minor environmental consequences [
3,
8] and encourage the efficient use of water in its operations [
1,
5,
7].
Previous studies have suggested the DWC strategy to enhance the BTX separation system. Kim [
10] showed that such an application could save 35.8% and 32.2% of heating and cooling duties compared to the conventional counterpart. Such a result led to a reduction of 33.9% in utility costs, enhancing the investment’s payoff. In addition, Buitimea-Cerón et al. [
12] applied the minimum vapor, Triantafyllou and Smith, and Sotudeh and Shahraki methods, as well as proposed a modified version of the later one to carry out the optimum economic design of DWC based on total annualized cost (TAC) minimization.
Meanwhile, numerous earlier studies [
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25] have focused on developing control structures for this separation system. Tututi-Avila et al. [
25] also studied the Kaibel and Satellite configurations, since the authors considered n-heptane and tri-methyl benzene as light and heavy species, respectively, in addition to BTX ones. This same five-component mixture was also considered by Kiss et al. [
26]. The authors explored various separation configurations utilizing dividing-wall columns and also a Kaibel scheme. Their findings demonstrated that the two-DWC configuration achieved an approximate 14% reduction in both energy consumption and CO
2 emissions while maintaining high product purity, which represents a positive impact on the economy and environment. In turn, Si et al. [
27] offered another unique concept that examined the alignment of DWC and vapor recompression (VR) with the organic Rankine cycle (ORC). It was shown that the ORC applied to the VR-DWC could save up to 44.99% of energy demand and 33.57% in the TAC expenditure, which accounts for capital cost and operating cost, while reaching an exergy efficiency of 15%. Zhang et al. [
28], Zhao et al. [
29], and Ge et al. [
30] also developed studies of DWC assisted by VR strategies regarding the BTX system. The latter one also evaluated the synergy between the Kaibel configuration and the VR strategy through six different configurations, which resulted in complex separation schemes for four-component separation into a single column. All the proposed schemes were deemed economically infeasible because they increased the TAC criterion in comparison to the Kaibel scheme.
As recently reviewed by Segovia-Hernández et al. [
4], several studies have been conducted on the design, control, and optimization of DWC schemes in the last two decades, and the number of published papers per year increased from around 5 (year 2000) to 75 (mean value between 2018 and 2020). However, there are few studies devoted to the environmental analysis associated with DWC configurations, which motivates the present study.
Given the importance of thriving in sustainable development, a productive way to evaluate the benefits of process intensification strategies is the Eco-efficiency Comparative Index (ECI). This method integrates various eco-indicators, aligning several process aspects to its economic performance in order to provide a quantitative comprehensive metric that surpasses individual estimates in accuracy [
5,
7,
31]. The ECI has been successfully applied in research to improve key industrial chemical processes [
32,
33,
34]. Pereira et al. [
31] applied the ECI methodology to monitoring a real operating industrial petrochemical plant located in Rio de Janeiro, Brazil [
35]. The ECI metric was essential for quantifying the improvements of engineering actions in sustainability performance based on the joint evaluation of energy and water consumption savings, as well as the respective CO
2 emissions and wastewater/waste generation. Based on computer-aided process design and simulations tools, Caxiano et al. [
34] concluded that for the acetic acid purification process, VR and double-effect distillation (DED) strategies can reduce 81.35% and 45.6% of energy consumption, 73.91% and 58.5% of water consumption, as well as 85.77% and 43.2% of CO
2 emissions compared to the conventional counterpart, respectively, increasing its eco-efficiency by 67% and 45%, respectively. In turn, Figueiredo et al. [
32] and Figueiredo and Prata [
33] demonstrated an 84.57% and 96.21% eco-efficiency increase in the monochlorobenzene/benzene separation industrial process using a DED and VR configurations, respectively. A point in common among these works [
32,
33,
34] is the consideration of a utility plant for estimating water losses and enhancing the energy demand consumption, including the sensible heat parcel and the boiler combustion efficiency to generate steam. As a result, the CO
2 emissions, water consumption, and utility costs estimates are also considered more realistic as reported in the literature [
36].
Additionally, it is important to highlight that the safety metrics present in the social pillar of sustainability are perhaps the ones that are most overlooked during retrofit and revamping proposals through computer-aided process design and simulation tools [
37]. In this sense, including inherent safety principles during the early-stage design of chemical processes may be highly valuable in assessing the environmental and human health implications of discharges from industrial operations [
7,
34,
38]. Such an appraisal not only provides a measure of possible risks but can also, in the long term, reduce capital and operating costs [
39].
In light of this, the current work attempts to fill this important gap in the BTX separation system, specifically in the intensified design via DWC technology, and provide a comprehensive assessment of the sustainability and safety evaluation based on performance indicators (energy and water consumption, CO
2 emissions, and utility cost, among others), promoting the revamping of operating plants spread worldwide and reconciling the production sector with the United Nations’ Sustainable Development Goals established in the 2030 Agenda [
40]. Total energy consumption reduction, optimal economic design based on TAC criterion, control strategies to enhance product quality, energy-efficiency, and operational performance were the main research focuses previously reported in the literature regarding the BTX separation process [
10,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28,
29,
30], with only two works focusing on CO
2 emissions [
26,
27] as environmental burdens. Thus, there is a need to study environmental impacts and eco-efficiency performance in the BTX mixture separating system through the DWC strategy.
3. Methodology
The methodology used in this study was theoretical–computational performed in UniSim® Design Suite software version R490, in which six steps were preponderant:
- I.
Understanding and simulation of the conventional process (CP) and divided wall column configuration as proposed in Ling and Luyben [
13];
- II.
Simulation result validation (assuming tolerance boundaries);
- III.
Simulation of a utility plant for both separation schemes;
- IV.
Data survey and development of indicators (ECI and WAR methodologies) followed by a sensitivity analysis of key heuristic parameters from the utility plant;
- V.
Comparison and evaluation of both separation schemes;
- VI.
Conclusion.
The diagram shown in
Figure 1 depicts the strategy taken given the aforesaid steps, which will be supported and explored throughout this and the next sections.
The process proposed by Ling and Luyben [
13] was selected for this study for two primary reasons: it represents the largest industrial production scale (1 kmol/s = 336,749 kg/h) and is the most studied plant in terms of control strategy development for the BTX separation system among all other configurations [
10,
19,
22,
24].
Furthermore, the calculation and development of all performances index will be finally described.
3.1. Conventional Process Description
As proposed by Ling and Luyben [
13], the conventional separation process with two distillation columns in series has a feed of 1 kmol/s composed of 30 mol% benzene (X
B), 30 mol% toluene (X
T), and 40 mol% ortho-xylene (X
X) that enters the first column (with 30 stages; fed into the 14th stage; internal diameter of 6.19 m, spacing between plates of 0.6096 m). In the output streams, the distillate (D1) has a high purity of benzene (99 mol%), while the bottom product (B1) passes through a valve and is subsequently sent to a second column (with 28 stages; fed in the 14th stage; internal diameter of 8.22 m, spacing between plates of 0.6096 m). The separation resulting from the latter makes it possible to obtain a distillate (D2) rich in toluene (99 mol%) and a bottom stream composed mainly of ortho-xylene (99 mol%).
Figure 2 depicts such a process, as well as the simulation results.
3.2. DWC Separation Process Description
To improve the process’s energy efficiency, Ling and Luyben [
13] also proposed the BTX separation through the DWC intensification scheme, which is seen in further detail in
Figure 3 with the respective simulation results.
For this system, the input stream (Feed) is supplied to one side of the column with the same specifications as the conventional process, reaching the wall inside the tower (not necessarily in the center). As a result of mass and heat transfer principals, the lighter species (benzene) goes mostly to the top of the column, leaving in the distillate (D3), while the heavy key (o-xylene) is directed to the bottom of the column, being removed almost entirely as a bottom product (B3). Furthermore, the toluene initially flows both to the upper part of the tower and to the lower part; however, when it reaches the extremes of the column, it starts to move on the other side of the wall, being recovered through the side stream (SS).
In the proposed DWC scheme [
13] a total of 46 stages are required (including the condenser as first stage and the reboiler as the last one—Aspen Plus
® software notation), and the dividing wall is located between stages 10 and 33. The rectifying and stripping sections run from stage 2 to stage 9 and from stage 34 to stage 45, respectively. The feed enters onto stage 21 in the prefractionator side, and the intermediate product is withdrawn from stage 20 of the main distillation column (the other side of the wall). The diameter of the DWC is 7.37 m; the distance between the column shell and the wall on the prefractionator side is 5.63 m. A reflux-drum pressure of 0.37 atm is selected, resulting in a temperature of about 322 K so that cooling water can be used as utility in the overhead condenser. A 0.0068 atm pressure drop per stage is considered, resulting in a bottom pressure of 0.67 atm and a base temperature of around 404 K so that medium-pressure steam can be used as a heat source utility. These operational conditions represent the optimal economic design [
13]. The optimization conducted involved varying the number of trays in each of the four column sections, as well as adjusting the feed inlet location, sidestream withdrawal (SS), vapor split, and liquid split. The main goal was to identify the configuration that minimizes the total annualized cost while ensuring the desired product purities are achieved.
3.3. Utility Plant
In order to provide more accurate and consistent estimates of the demand for utilities in the aforementioned processes, a computational simulation of the utility plant was also conducted. Such a plant consists of two subsystems primarily driven by water: the cooling (CS) and the steam generation (SG) sections, without any heat integration or energy co-generation [
41,
42,
43], as shown in
Figure 4.
The cooling section of a utility plant plays a crucial role in heat removal, ensuring efficient operation of the entire facility. After absorbing heat from process streams in the main plant, the water undergoes chemical treatment and is then sent to the cooling tower. In this final mentioned piece of equipment, the warmed water is returned to the supply temperature by the actions of fans that induce a countercurrent air flow responsible for heat dissipation by means of evaporation and convection. The water, now cooled, is made available once again for the plant’s refrigeration activities. During such cycles, losses of water frequently occur due to possible leaks in the equipment present in the cooling operation (F
CL), evaporation (F
E), drift (F
D), and tower blowdown (F
CTB), which is required to avoid the accumulation of suspended and dissolved particulates. As a result, a makeup stream (F
MupCS) is required to replace the volume of water lost during the cooling operation [
41,
42].
As for the steam generation section, water undergoes pre-treatment in both cationic and anionic beds to deionize it. Subsequently, the deionized water is sent to the deaerator to remove gases, such as nitrogen, oxygen, and carbon dioxide. The liquid stream leaving the deaerator is the boiler feed water (BFW) (about 100 °C), which, after pressurization, is sent to the boiler, where the BFW acquires sensible and latent heat, shifting into the vapor state according to the desired pressure. In this way, the water, now heated, is made accessible to suit the demands of the industrial facility.
Furthermore, analogous to the cooling process, losses are also present due to possible steam leakage in transportation and the equipment present in the heating operation (F
HL), initial treatment (F
TL), and the boiler blowdown (F
BB), which is necessary to prevent the deposition of solids. Therefore, a second makeup stream (F
MupSG) is introduced to compensate for the amount of water lost. It is important to emphasize that in the SG section, different vapor classes can be generated [
41], such as low-pressure steam (lps—308 kPa and 407.7 K), medium-pressure steam (mps—1136 kPa and 458.7 K), and high-pressure steam (hps—4201 kPa and 527.1 K). This present study used lps and mps as a utility for the CP and DWC processes, respectively, since these classes fulfill the minimum approach of 10 °C with the process streams to be heated in the reboilers of the CP (378.2 K and 376.5 K) and DWC scheme (401.5 K), supporting the dynamic control of the main plant [
43]. Nevertheless, during the simulation of the utility plant, it is essential to consider the sensible heat source since it has the potential to result in an error of more than 18% in the overall combustion energy demand [
44], as well as the boiler combustion efficiency [
33,
41]. The operational conditions and heuristic values of water losses based on real operating utility plants adopted during its simulation are shown in
Table 1. Make-up process water was assumed to be supplied at 303 K and 101.3 kPa [
41].
Since the electrical power of the cooling tower’s fans (E
Fans) cannot be retrieved directly from the UniSim software environment, it was calculated using Equation (1) as suggested by Caxiano et al. [
34]. Meanwhile, the energy requested by the utility pumps was calculated by the software, and the efficiency was set at 75% [
42].
In Equation (1), E
Fans is computed using the specific area of the tower of 1.804 ft
2/(m
3/h) and the fan power per area, whose value is 8.05 × 10
−5 (GJ/h)/ft
2 [
34,
45]. This adjustment considers the air wet bulb temperature of 299.8 K (26.7 °C) (referent to Rio de Janeiro, Brazil), which allows the water inlet temperature to be 30 °C [
41], and accepts a cautious efficiency of 90% for the tower’s performance [
45], as well as 90% for the electricity driven fans (ƞ
Fans).
As described in the base process design books [
41,
42,
43], heat integration between the utility streams and those of the main process plant is not considered, nor is there cogeneration of electricity through turbines (only basic utility plant configuration is considered in this study). This last one is supported by the lack of need for high pressure steam in the analyzed processes.
3.4. Environmental Assessment and Eco-Efficiency Analysis
The purpose of the intensification techniques proposed for the BTX separation process is to boost its environmental and economic accomplishments, minimizing the emission of greenhouse gases and the generation of wastewater, as well as decreasing the expense of capital by reducing utility costs. In this context, the assessment of environmental, social, and operational impacts, especially in the initial stage of the project design, is essential to guarantee technical–economic feasibility and favor the examination of the process’s eco-efficiency [
5,
7].
Eco-indicators can assist in such evaluation, quantifying certain improvements [
5,
7]. They commonly relate a process aspect (for example, environmental, safety, social, and operational costs) to its economic performance (usually the production rate) [
31]. For instance, among the key indicators for analyzing the eco-efficiency of the BTX separation process, those that account for water consumption (WC), CO
2 emissions (CDE), and utility costs (SUC) are particularly relevant.
These aforementioned indicators are a common choice to evaluate eco-efficiency for three main reasons. First, they are easily calculated with real production data, mainly for process monitoring [
31], as well as with computer simulation data [
8,
32,
33,
34], without a substantial amount of quality data that might render it laborious, as reviewed [
7], principally during the early-stage design [
5]. Second, the utility cost, which represents a parcel on the TAC criterion, and CO
2 emissions are the most used metrics in the technical–economic evaluation of process retrofit/revamp projects [
9,
41] and environmental impacts [
7], respectively. Water and energy consumption, which include both combustion and electrical sources and have a direct correlation with CO
2 emissions, are the major drivers of utility costs. Finally, the reduction of water consumption, mitigation of greenhouse gas emissions, and process intensification are all major objectives contained in the #6 (clean water and sanitation), #12 (responsible consumption and production), #13 (climate action), and #9 (industry, innovation, and infrastructure) Sustainable Development Goals (SDGs) proposed by the United Nations during Agenda 2023, which influenced the performance indicators chosen for the eco-efficiency evaluation of the suggested configurations, as displayed in
Table 2. In addition, the CO
2 emissions and water are the most relevant metrics among the sustainability indicators [
46].
As presented in
Table 2, the chosen economic variable was the production rate, which was set as the sum of specific mass flow streams such as D1, D2, and B2 for the CP and D3, SS, and B3 for the DWC. F
Mup represents the total water requirement, obtained by adding F
MupSG and F
MupCS from the utility plant. Regarding energy consumption, there are two main constituents: combustion energy (E
Comb) and electricity (E
Elect). The former refers to the total boiler demand (including both sensible and latent heat), adjusted for its efficiency (η
Comb), while the latter accounts for the electric consumption related to the equipment’s, such as pumps and mechanical fans present in the utility plant, considering their respective efficiencies (η
Pump and η
Fans). Additionally, to further estimate the CO
2 emissions, conversion factors (ξ
Comb and ξ
Elect) were used to correlate the total energy per source spent with the mass of carbon dioxide produced. ξ
Comb was fixed at the international value of 0.0561 t
CO2/GJ [
47], and ξ
Elect was assumed to be 0.0234 t
CO2/GJ [
48], corresponding to Brazil’s mean value for the years of 2021–2022. Since this factor depends on the local energy matrix, its value may vary significantly from one country to another. Lastly, the total wastewater for treatment (F
ww) encompasses all remaining water lost within the utility system, excluding non-recoverable losses from drift and evaporation. A detailed breakdown of utility costs for each source is provided in
Table 3, supporting the SUC indicator calculation.
To enable the eco-efficiency index evaluation, the eco-indicators presented previously in
Table 2 are grouped and standardized on a common scale [
7,
31]. In this step, each eco-indicator is divided by its worst-case scenario (the highest value observed). This essentially flips the perspective, where a lower value now represents a better environmental performance. Such dimensionless analysis favors the plot of a radar chart, allowing a comparative analysis between different processes based on estimated areas calculated using the Law of Sines, as presented in Equations (2) and (3). Each alternative is represented by a polygon, where the distance from the center reflects its environmental impact. The smaller the area enclosed by a process, the better its eco-efficiency.
Within Equations (2) and (3), A refers to the area of the polygon with n indicators of each process, L is the normalized indicator, and A* is the area of the worst alternative polygon found. This last one essentially creates a reference point in the radar chart, allowing for comparison of the relative environmental performance of different processes.
This study employed a three-indicator evaluation, where the concept of cyclic permutation was not necessary due to the limited number of indicators, in which changing an indicator position (ordering) in the polygon does not affect the final result [
7,
34]. Additionally, all indicators were assigned equal importance (weight), reflecting their interconnected nature within the process and their role in representing the overall operational conditions [
31]. Modifications, including the importance (among indicators in the same category) and relevancy (among different categories, such as environmental x economic) could be further addressed. In this case, the radar chart shape is transforming into crown chart one [
7].
3.5. Environmental and Safety Assessment Using Waste Algorithm Reduction (WAR)
Recognizing the importance of balancing safety and environmental concerns, modern industry must prioritize sustainable practices alongside technical and financial considerations, principally in the early design stage [
3,
6]. To achieve this, the waste reduction algorithm (WAR) has been identified as a valuable tool for thoroughly assessing the potential impacts of chemical and petrochemical processing plants. This algorithm classifies processes according to eight impact indicators: global warming potential (GWP), acidification potential (AP), ozone depletion potential (ODP), photochemical oxidation potential (PCOP), aquatic toxicity potential (ATP), terrestrial toxicity potential (TTP), human toxicity potential by ingestion (HTPI), and human toxicity by exposure (HTPE).
The aforementioned metrics can be inferred by using an open-access graphical interface offered by the U.S. Environmental Protection Agency (EPA), which analyzes the designated control volume of the chemical plant in terms of the environmental impacts that cross its system boundaries [
50]. In this regard, simulation information concerning the composition and mass flow of inlet and outlet streams, as well as the total energy consumption, must be provided for the outcome evaluation [
6,
34]. All the potential environmental impacts (PEI) that are generated within or leaving the system are quantified by the mass of the product, expressed as PEI/t.
It is noteworthy that the GWP indicator was not included in the present study since it was replaced by the CO2 one. The CDE metric has a broader scope because it includes emissions from electrically driven equipment such as pumps and cooling tower fans. Nonetheless, both indicators take into account the boiler’s sensible heat source and combustion efficiency. This can result in comparable results due to the absence of significant electrically powered devices, such as compressors, and flare emissions in the BTX system.
The WAR algorithm indicators will complement the environmental-safety analysis of the BTX process, as well as CO2 emissions, water consumption, and wastewater indicators.
3.6. Computational Simulation
Process simulation has emerged as a powerful tool for evaluating process performance parameters such as exergetic, energetic, and environmental, as well as economic, safety, and control. Validating the previously reported results is critical for conducting process intensification studies and performance evaluations. Thus, to simulate the BTX processes and their respective utility plants, the UniSim
® Design Suite R490 software was used under steady-state conditions. The UniSim software is based on detailed, high-fidelity representative models such as mass and energy balance, thermodynamic equilibrium relationships, and constitutive equations (reaction and mass/heat transfer) and has been widely applied in the computational simulation of several chemical processes as recently reviewed [
32,
33].
In the reference work [
13], the thermodynamic package used was the Chao-Seader; however, aiming for greater compatibility between the data and smaller deviations, the Peng Robinson model was chosen, which is applicable for hydrocarbons at low pressures and was previously successfully applied to the BTX system [
28,
29]. For the utility plant, the UNIQUAC activity coefficient model was adopted, as well as the heuristics and operational conditions listed in
Table 1.
Furthermore, it is not possible to simulate the DWC setup using the Petlyuk column arrangement described by Luyben [
51]. Thus, DWC was developed using four interconnected columns [
13]: two absorbers to represent the prefractionation and main column (sections originated from the wall insertion; without any reboiler or condenser), a column with a top condenser to represent the rectifying zone (above the wall), and another with a reboiler to represent the stripping section (below the wall). UniSim
®’s Inside-Out solving numerical method (with standard tolerances for mass and energy balances) was applied to achieve convergence on each distillation column. In an inner loop, the algorithm uses equilibrium and enthalpy models to solve component and heat balances, whereas in an outer loop, rigorous calculations are used to update the thermodynamic models [
41]. Regarding the utility plant, the cooling tower simulation was performed by UniSim
®’s component splitter building block, which ignores the rigorous absorber separation model. Thus, the energy stream in this equipment needs to be replaced by the cooling fans power consumption described by Equation (1). In turn, the boiler simulation was performed in two steps. The former represents the sensible heat (setting the vapor fraction of the BFW stream to zero, after the pump pressurization), while the latter represents the latent heat (setting the BFW stream vapor fraction to one), transforming the BFW to saturated vapor.
Figures S1 and S2 in the
Supplementary Materials illustrate the conventional two-distillation-column process and the Divided Wall Column design, respectively, as developed in the UniSim Suite Design computational environment. Additionally,
Figure S3 shows the utility plant within the same computational framework.
Lastly, based on water and energy consumption data analysis, as well as the selected efficiencies, the related metrics (WC, CDE, and SUC) were determined through the Equations outlined in
Table 2 and the utility costs presented in
Table 3. These findings supported the evaluation of the ECI method, which was further supplemented by an overall environmental and safety assessment using the WAR algorithm. In addition, a sensitivity analysis was conducted with key heuristic values adopted for the utility plant simulation.
5. Conclusions
The present study was devoted to assessing the sustainability of the dividing-wall distillation column for separating a BTX mixture as a novelty. As revised, optimal economic design and control schemes were the main research focus previously reported by twenty other studies regarding DWC applied to this important petrochemical system. The conventional and DWC configurations were designed by computer simulation in the UniSim software. A utility plant with sections for cooling water and steam generation was also considered for more accurate results regarding water consumption, CO2 emissions, and utility costs since real heuristics for water losses, combustion efficiency, and sensible heat are included. The results showed that the DWC scheme is 37.52% more eco-efficient than the conventional counterpart, mainly due to an 15.59% and 30.31% savings on energy and water consumption, respectively, which provided a 15.48%, 30.05%, and 16.72% reduction on CO2 emissions, wastewater generation, and utility costs, respectively. Additionally, savings of about 16% were achieved regarding the other seven environmental indicators present in the well-known WAR algorithm, which includes two safety indicators, namely, human toxicity by exposure and human toxicity potential by ingestion. The influence of the sensible heat parcel was demonstrated since it accounts for 15.74%, 15.84%, 14.73%, and 18.94% of energy demand, pollutant emissions, operational costs, and SG make-up water, respectively. These results highlight the environmental advantages of the DWC configuration, aligned with the sustainable development goals of the 2030 Agenda of the United Nations, which represents a revamping option for existing BTX plants worldwide.
One limitation of this work worth noting is that only a preliminary safety analysis was performed. Since the DWC scheme operates at higher temperatures and requires a higher vapor class, when compared to the conventional counterpart represented by two connected columns, a higher pressurization of the boiled feed water stream before feeding the boiler at the utility plant is needed. Thus, it could be possible that some inherent safety indicators increase.
The quantification of economic and environmental benefits through the comparison of DWC with other intensification strategies, in addition to the conventional column sequence process, remains a challenge in the literature. Moreover, the development of more comprehensive and representative safety metrics is an important step in complementing the analysis presented here. Therefore, both aspects should be considered in future research.