Comparative study of conventional and process intensification by reactive distillation designs for glycerol carbonate production from glycerol and diethyl carbonate | Scientific Reports

Scientific Reports volume 15, Article number: 1753 (2025) Cite this article

1372 Accesses

2 Altmetric

Metrics details

Glycerol carbonate (GC) can be produced from glycerol (GL), a low-value byproduct in the biodiesel industry. In this work, continuous processes of GC production via transesterification from crude GL and diethyl carbonate (DEC) were developed using Aspen Plus. Two cases were considered, and their process performances were compared. In Case I, a conventional design consisted of a continuously stirred tank reactor for the reaction section and a distillation column for the purification section. In Case II, a process intensification design consisted of a reactive distillation column that could accommodate both reaction and purification within a single column. In both cases, the process optimizations were carried out by connecting the process models in Aspen Plus to MATLAB, using the Genetic Algorithm as the optimizer. The results showed that Case II was superior to Case I in terms of energy utilization, CO2 emissions, and economics with the specific energy consumption of 1.92 kWh/kg of diethyl carbonate, % internal rate of return of 274, payback period of 1.44 years, and CO2 emissions of 0.26 kg CO2/kg DEC. Lastly, the proposed process in Case II was compared with the GC production using dimethyl carbonate (DMC). It was found that using DEC was superior to DMC due to easier separation and glycidol avoidance.

Glycerol (GL) is the major byproduct of biodiesel production through the transesterification of vegetable oils or animal fats with methanol. In the last two decades, biodiesel industries have risen significantly due to the global demand for sustainable alternatives to petrochemicals and oil-derived fuels. A large and rapid growth of biodiesel at present causes a large amount of GL, which is inexpensive to sell. Hence, GL has been converted into high-value products, e.g., propanediol for automotive antifreeze additive, acetals as fuel additives in gasoline, and Epichlorohydrin for conversion to epoxy resins1,2. Glycerol carbonate (GC) is one of the high-value products that can be produced from GL by transesterification. GC can be used as a solvent and a surfactant in beauty and skincare products. Moreover, it may also be used as an electrolyte for lithium batteries and a reactant for synthesizing polymers such as polyesters, polyurethanes, and polycarbonates3.

Several methods can be used to produce GC from GL4. The first method involves reacting GL with phosgene (COCl2), but this process is hazardous because it involves a highly toxic gas. The second method requires urea as a feedstock and requires continuous ammonia removal to shift its chemical equilibrium using reactive distillation (RD). Reactive distillation (RD) is a unit operation, in which a reactor and a distillation column are combined in a single unit. RD can simultaneously facilitate chemical reactions and separation in a single column. RD has garnered significant interest, particularly for chemical equilibrium-limited reactions that typically require a large excess of one of the reactants. It can increase conversion by continuously removing the product from the reaction section, allowing the reaction to shift rightward and achieve higher conversion of reactants. Another advantage of RD is its ability to enhance selectivity, reduce energy consumption, lower capital expenses, and facilitate the separation of close-boiling components5. Lertlukkanasuk and coworkers6 used RD to remove ammonia continuously from the reaction zone. The results showed that the RD system gave better results than the conventional design system which used a continuous stirred tank reactor (CSTR) given the higher urea conversion from the rightward shift of equilibrium due to the ammonia removal.

The third method is the carboxylation of GL using CO/CO2. This method seems interesting because it utilizes two waste streams as reactants to create valuable products. However, the yield of this reaction is very low7. The last method is the transesterification reaction by reacting GL with dimethyl carbonate (DMC) or diethyl carbonate (DEC). This reaction also has a chemical equilibrium problem—it requires a technique that can shift the chemical equilibrium rightward to achieve a higher GL conversion.

Regarding the transesterification reaction of GL with DMC, Yu and coworkers8,9 studied this reaction using the RD system and compared it to the CSTR (conventional) system. Srivastava et al.10 generated their distillation model based on a residue curve map and compared the proposed model with reactive distillation extractive distillation (RDED) in Yu’s work9 and reactive distillation pervaporation (RDPV) in Sun’s work11. According to Srivastava’s results, their proposed model yielded better results than RDED and RDPV systems in all aspects of process performances.

Regarding the transesterification reaction of GL with DEC, Zhang et al.12 studied this reaction using Ce–NiO as a catalyst. Interesting findings based on their results revealed that an equilibrium conversion of 80% with 100% GC selectivity could be achieved by the proper selection of an operating temperature at 353 K at a reaction time of 4 h. In other words, the unwanted by-product, glycidol (GD) could be avoided leading to less separation equipment compared to the previous works9,10,11.

Yet, a simulation study of GC production using DEC and GL has never been studied. Therefore, the novelty of this work was to design the GC production process in which GL reacted with DEC. This work aimed to develop a continuous process for GC production from crude GL and DEC by designing two processes, including a conventional design (Case I) that employed a CSTR to accommodate a chemical reaction and a process intensification design (Case II) that employed RD. These two processes would be evaluated and compared in terms of the process performance indexes, e.g., GL utilization, energy utilization, economics, and CO2 emissions. In addition, the performance indexes from the superior case (e.g., Case I and Case II) would be compared with the previous works that used DMC as a raw material. The successful development of this process would reveal a new approach to GL conversion into a more valorized product such as GC.

To study feasibility and evaluate the performance of GC production between the two design cases, simulation models were developed in process simulation software, viz Aspen Plus V11. Details of the simulation are provided in this section.

Thailand had a biodiesel production rate of approximately 8.5 million liters per day, with 13 biodiesel plants registered with the Department of Energy Business in the year 202113. The average biodiesel production rate per plant was approximately 23,900 kg/h. For every 100 kg of biodiesel produced, 10 kg of crude GL is generated14, resulting in a supply of 2390 kg/h of crude GL. It was assumed in this work that 30% of crude GL was used to produce GC; thus, the feed rate of 716.4 kg/h of crude GL was used as a basis for simulation. Furthermore, DEC was provided in excess relative to GL with a ratio of 3:1 of DEC:GL in the feed stream. Ce–NiO was used as a catalyst in the amount of 5 wt% of GL, according to the study by Zhang et al.12.

According to Eric Carlson’s recommendation15, the UNIFAC model was used for the simulation of the crude GL pretreatment process due to the missing property of matter organic non-glycerol (MONG) and the existence of polar and non-polar compounds in the system. In this work, MONG was represented by tripalmitin, a triglyceride whose molecular structure is available in Aspen’s database. Since the molecular structure is available, other thermodynamic properties can be estimated by the UNIFAC model. For the transesterification of GL with DEC, the reaction was operated at low-pressure conditions, and there was a non-ideal mixture; hence, the activity coefficient models, e.g., NRTL, UNIQUAC, etc. should be selected. In this work, the Non-Random-Two-Liquid (NRTL) model was selected according to the study by Liu16, who reported the NRTL binary parameters for this system.

For the accuracy of the thermodynamics model, the NRTL binary parameters from Liu16 were entered into Aspen Plus. After that, two tertiary diagrams were simulated, which consisted of GL–DEC–GC and GL–DEC–EtOH systems for validation with the experimental data from Liu16. The validation is provided in "Thermodynamic model validation" section.

Since process simulation required input of several variables into the program, it was necessary to determine the optimal value for each variable so that the process could operate cost-effectively. In this work, the Genetic Algorithm (GA) was used to minimize the total annual cost (TAC), which was the objective function for the optimization17. The equation for the TAC is expressed in Eq. (1).

where the total capital cost was calculated based on the equipment and installation costs whereas the operating cost was calculated based on the sum of utility, labor, and maintenance costs. Please note that the GA was selected herein since the optimization problem was defined as a mixed-integer optimization problem. A list of parameters used in the optimization is provided in "Process economics" section for Case I and Case II.

Genetic Algorithm (GA) is a heuristic optimization method inspired by the process of natural selection and genetics. GA is used to find approximate solutions to complex problems by mimicking the mechanics of natural evolution, such as selection, reproduction, and mutation. The steps of GA for finding the optimal solution are summarized in Fig. 1a. GA optimization starts by creating the initial population and scoring which populations have a high probability of survival (the populations that produce a satisfactory solution). Then, the high-scoring populations are selected to generate the next generation according to the selection rule. The new generation can be created via two approaches. First, the crossover rule by combining the two parents to form a new child, and the mutation rule by slightly changing the parents to create a new child. The method to create a new generation is illustrated in Fig. 1b. Then, the new generation is scored and used to calculate the next generation in the optimization iteration18.

(a) Flow chart of GA18, (b) methods to create a new generation in GA18.

In this work, the tuning parameter in GA was the maximum number of generations (iterations). Herein, the maximum number of generations was set at 10. Please see "Optimization results of total annual cost (TAC)" section for further details.

Pandit19 described the method of communicating Aspen Plus with MATLAB using the ActiveX server as an intermediary for connection. MATLAB, which included the optimization algorithm, generated the input for the process and sent it to Aspen Plus through the ActiveX server. Then, Aspen Plus executed the simulation and sent the results back to MATLAB via the ActiveX server again. The data transfer between Aspen and MATLAB was repeated until the objective function was satisfied. Furthermore, on the MATLAB community website, Abril20 provided a code, connecting MATLAB with Aspen Plus via the ActiveX server to conduct a sensitivity analysis of a reactive distillation column. Herein, Abril’s code was utilized as the starting-point code that linked the two programs.

A crude GL pretreatment process was first designed, based on the works of Thanahiranya et al.21, Supramono and Ashshiddiq22, and Chang23. The components in crude GL are provided in Table 1.

The process began with feeding crude GL into a splitter to separate the solid phase from the crude GL. Then, the outlet stream from the splitter was pressurized to 0.1 bar and heated up to 95 °C in the flash tank to remove methanol and water. The remaining stream was then pressurized to 1.0 bar through a pump and sent to a decanter to separate MONG from the GL. The crude GL pretreatment process yielded a 97.4 wt% pure GL which was ready for esterification with DEC.

Zhang et al.12 studied the reaction of GC synthesis via the esterification of GL and DEC over the Ce–NiO catalyst. The main chemical reaction is expressed in Eq. (2).

The equilibrium constant (Kc) was calculated via experimentation. A reaction in a 50 mL isothermal batch reactor was carried out, in a temperature range of 338–358 K using a feed of GC and DEC in the ratio of 1:3 and catalyst in the amount of 5 wt% of GC12. The Kc was calculated based on Eq. (3), and the result was found by linear fitting of ln Kc versus 1/T shown in Eq. (4); where C is the molar concentration (mol/L), and T is the temperature (K).

Moreover, Zhang et al.12 also reported the effect of conversion and selectivity on the reaction time and temperature, as shown Fig. 2. Therefore, this work set the conditions for this reaction by controlling the operating conditions at a reaction time of 4 h and a temperature of 353 K, to avoid the formation of glycidol (GD) as an undesired product.

The effect of (a) reaction temperature, and (b) reaction time, on the conversion and selectivity at 353 K obtained from Zhang12.

The process flow diagram for Case I is provided in Fig. 3. In this case, the process began with a feed of 716.4 kg/h of crude GL into the GL pretreatment process to remove impurities from the crude stream. The purified GL was then mixed with the GL recycle stream and the fresh DEC stream with the DEC recycled stream. The mixed stream was subsequently fed into the CSTR where the transesterification occurred. To achieve a reaction close to its chemical-equilibrium limit without the occurrence of side reaction, e.g., the formation of GD, the amounts of GL and excess DEC were controlled at a molar ratio of 1:3 at 80 °C with the reactor’s residence time of 4 h, and atmospheric pressure. The Ce–NiO catalyst was loaded into the reactor with about 5% of the GL mass. According to the experiment by Zhang et al.12, the equilibrium constant of the reaction was used to calculate the reactor volume in this work.

Process flow diagram for the simulated process in Case I (CSTR).

The outlet stream from the reactor (R-201) was sent to the purification section, which consisted of three distillation columns (see Fig. 3). First, D-201 was used to separate DEC and EtOH from GL and GC. Second, the distillation column D-202 was employed to recycle any unreacted DEC to the reaction system and to purify EtOH to achieve the commercial sale purity of 95.0 wt%. Lastly, the D-203 distillation column was utilized to separate any unreacted GL and to purify GC to achieve a commercial sale purity of 97.0 wt%. Details of each unit operation along with the stream results of Case I are summarized in Supplementary Tables S2 and S4.

To determine the optimal conditions of the columns in Case I, Table 2 below provides a summary of all variables involved in distillation in Case I.

As seen in Table 2, there were three distillation columns. Each column had 5 input variables, e.g., Number of stages, feed stage, column pressure, distillate to feed (D/F) ratio, and reflux ratio. The total number of variables was 15. Since there were two targeted specifications for each column, the D/F and the reflux ratios were determined using DesignSpec features in Aspen. In summary, the number of variables was 15, and the number of specifications was 6. Thus, the degree of freedom was 9 in Case I as shown in Table 3.

Table 3 provides the summary of variables to be optimized along with the category (e.g., integer and float) of each variable. The optimization constraints were set as (1) × 1 − × 2 ≥ 1, (2) × 4 − × 5 ≥ 1, and × 7 − × 8 ≥ 1. The population size was set at 50 while the maximum number of generations was set at 10.

In Case II, the process began with a feed of 716.4 kg/h of crude GL to the GL pretreatment process and then to the reactive distillation column (RD-201). The process flow diagram of the RD process is shown in Fig. 4. The fresh DEC was mixed with unreacted DEC from the recycle stream and fed into RD-201 at the same ratio as in Case I. RD-201 was used instead of the reactor and the first distillation column in Case I (R-201 and D-201 in Fig. 3). The RD column was first set to separate EtOH to shift the reaction equilibrium forward. The operating pressure of RD-201 was set to operate at a vacuum condition to control the temperature in the reaction zone not to exceed 80 °C to avoid a side reaction (e.g., the GD formation)12.

The process flow diagram of the simulated process in Case II.

The GL conversion in this case appeared to reach nearly 100%, which showed that RD helped in the forward shifting of the chemical equilibrium. The bottom product from the reactive distillation column (RD-201) was sent to the separation system in D-201 for purification to achieve a purity of 97 wt% of salable GC (Fig. 4). The unreacted DEC was also recycled to the reaction system. Details of each unit operation along with the stream results of Case II are summarized in Supplementary Tables S3 and S5.

Similar to that in Case I, Table 4 summarizes all variables involved in distillation in Case II.

As seen from Table 4, there were two distillation columns. RD-201 had 6 variables whereas D-201 had 5 variables. The total number of variables was 11. Since there was one targeted specification in each column, the D/F ratio was determined using DesignSpec features in Aspen. In summary, the number of variables was 11, and the number of specifications was 2. Thus, the degree of freedom was 9 in Case II as shown in Table 5.

Table 5 provides the summary of variables to be optimized along with the category (integer and float) of each variable. The optimization constraints were set as (1) × 1 − × 3 ≥ 1, (2) × 3 − × 2 ≥ 2, × 6 − × 7 ≥ 1, The stage temperature (× 3 − 1) ≤ 80 °C, and DEC recovery at D-201 in the recycle stream ≥ 95%. The population size was fixed at 50 while the maximum number of generations was fixed at 10.

The designed processes (e.g., Cases I and II) were evaluated in terms of four performance indexes, e.g., GL utilization, energy utilization, economics, and CO2 emissions. The purities and prices of raw materials and products according to the commercial suppliers are provided in Table 6.

The utility costs were obtained from a process design textbook, namely “Analysis, Synthesis, and Design of Chemical Processes” by Turton et al.28. The costs of utility are summarized in Table 7.

Glycerol utilization in Eq. (5) was used to calculate how much GL was converted into GC. The higher the GL utilization, the more efficient the process is in terms of GL conversion.

The specific energy consumption (SEC) was used to calculate the energy required to produce 1 kg of GC. The SEC is calculated as per Eq. (6) 29.

In the economic evaluation, three economic indicators were determined: profitability index (PI), payback period, and internal rate of return (%IRR). The PI, calculated by dividing the net present value (NPV) by the initial investment, served as a measure of profitability, where a PI greater than one indicated a profitable process. The payback period represents the time it would take for the process to become profitable. Lastly, the %IRR is a useful indicator for comparing the two alternatives and determining which process is more economically appealing.

The environmental impact of the process was determined by calculating the amount of CO2 emissions using CO2 tracking available in Aspen Plus V11. For comparison between processes, the CO2 emissions per quantity of GC produced were calculated using Eq. (7).

This section is divided into four sections: model validation (Section "Model validation"), optimization results of total annual cost (Section "Optimization results of total annual cost (TAC)"), validity of the optimized operating conditions of distillation (Section "Validity of the optimized operating conditions of distillation: Case I and Case II"), and comparison of the performance indexes of the two cases (Section "Performance indexes comparison: Case I versus Case II").

The NRTL model and the NRTL binary interaction parameters for the system of GL, DEC, GC, and EtOH from Liu et al.16 were entered into Aspen Plus to describe the phase equilibrium behavior of these components. This section is dedicated to the validation of the NRTL model in accurately predicting the vapor–liquid and liquid–liquid equilibrium behaviors. The simulation results are shown in ternary diagrams of the two systems: GL-DEC-GC system, and GL-DEC-EtOH system, each at the temperatures of 343.2 K, 363.2 K, 383.2 K, and 403.2 K (Experiment data from16, Simulation results from Aspen Plus Figs. 5 and 6).

Ternary diagrams of GL-DEC-GC system at temperatures of (a) 343.2 K, (b) 363.2 K, (c) 383.2 K, and (d) 403.2 K.

Ternary diagrams of GL-DEC-EtOH system at temperatures of (a) 343.2 K, (b) 363.2 K, (c) 383.2 K, and (d) 403.2 K.

The simulation results are shown as the colored equilibrium tie-lines and the blue solubility curves in the ternary diagrams in Figs. 5 and 6. These results were compared with the black circle’s points, representing the liquid–liquid equilibrium experimental data obtained from Liu et al.’s experiments16. The calculated root-mean-square errors (RMSE) for GL-DEC-GC system was 0.016 and for GL-DEC-EtOH was 0.016. These RMSE values were consistent with the values reported by Liu et al.16. Hence, it was interpreted that the input binary interaction parameters were correctly applied in Aspen Plus which could be used for calculating and simulating the GC production processes.

The equilibrium constant shown in Eq. (3) was inserted into an RCSTR model in Aspen Plus to validate the reaction at various temperatures for comparison with the study by Zhang et al.12. The simulation results in Table 8 with the calculated root-mean-square error (RMSE) of 0.103 mol/L revealed that the reaction equilibrium constant could represent the experimental results and could be used to simulate the reaction in this work. Please note that the RMSE was calculated based on a square root of the average of squared errors. The squared errors were determined from the simulated result subtracted from its corresponding experimental result.

Herein, the maximum number of generations was varied to ensure that the optimized variables (in Tables 3 and 5) yielded the minimized total annual cost (TAC) expressed in Eq. 1. Figure 7a,b provide the evolution of solutions as generation proceeds. As seen in the figures, further evolution of the solution (e.g., 11, 12, and so on) did not significantly reduce the TAC value in both cases. Thus, the maximum number of generations was set at 10.

TAC minimization as the evolution of solution proceeded in (a) Case I and (b) Case II.

Since the design pressure of the column was one of the parameters in the optimization (see Tables 3 and 5), the pressure range was constrained within a range of 0.1–1.5 bar. For clarity, the operating temperatures of columns in Case I (CSTR) and Case II (RD) corresponding with the optimized column pressures were summarized in Tables 9 and 10.

As seen in the tables, utilities were available for all operating temperatures. This suggested that the optimized column pressures were appropriate since there exist working utilities that were subject to the minimized TAC value.

The processes in Case I and Case II were evaluated using three performance indexes, including glycerol utilization, energy utilization, economics, and CO2 emissions.

Table 11 below provides a comparison of GL utilization. As seen from the table, Case II outperformed Case I since the GL utilization approached 1.0. The result was consistent with the fact that using reactive distillation could shift the chemical equilibrium rightward resulting in the higher conversion of GL. Regarding the DEC feed amount, the feed in Case II was more than that of Case I. Since GL utilization in Case II was higher than in Case I, the required amount of DEC was higher, which was consistent with the higher conversion of GL observed in Case II. Additionally, Case II yielded a higher GC production rate compared to that of Case I. Although Case I incorporated a recycle stream to reuse unreacted GL, the unreacted GL was lost in D-203 (see Fig. 3), leading to a lower GC production rate.

Table 12 provides the energy utilization comparison between Case I and Case II. As seen from the table, Case II required lower energy consumption per product compared to Case I. This emphasized the advantage of process intensification—the incorporation of reaction and distillation in a single RD unit found in Case II.

In terms of process economics, the simulated processes in the two cases were evaluated using Aspen Economic Analyzer, as shown in Table 13.

According to Table 13, the total capital cost in Case II was lower than that of Case I because the distillation column to purify the GL recycle stream was not required in Case II. For the operating cost (a sum of the labor, maintenance, and utility costs), Case II was also lower compared to Case I due to the reduced number of unit operations and the lower energy consumption as seen in Table 12. Regarding the total cost of raw materials, Case II had a higher cost compared to that of Case I. This was consistent with the results in Table 11 since Case II required more DEC feed than Case I. Furthermore, since GC was obtained in Case II more than in Case I at the fixed value of GL feed, the total product sale in Case II was superior to that in Case I. In addition, the PI with values greater than 1 indicated that both cases were profitable. Additionally, when comparing the %IRR and payback period, Case II had a higher %IRR, a shorter payback period, indicating that Case II was more attractive for investment than Case I.

In terms of the carbon footprint, the total CO2 emissions from the utilization of utilities were used for comparison between the two cases. The results are presented in Table.

According to the CO2 emissions in Table 14, the results showed that Case II had lower CO2 emissions compared to that in Case I. This was because Case II consumed less energy than Case I, and the amount of energy used directly affected the amounts of utilities as well as the CO2 emissions.

Finally, the performance indexes comparison between using DMC10 and DEC (this work) as a choice of raw material was compared. Please note that the SEC obtained from Srivastava’s work was recalculated—both hot and cold utilities used in their process were combined so that the value could be compared with this work. As seen in Table 15, although the values of GLY/GC were comparable, the specific energy consumption obtained in this work was about 18% lower than that reported by Srivastava et al.10. Possible reasons that may explain the superiority in terms of energy consumption were that (1) the DEC/EtOH/GC system did not have an azeotrope (i.e., leading to an easier separation mixture than the DMC/methanol/GC system in which DMC and methanol form an azeotrope) as seen from the VLE predictions based on the validated BIPs values in Fig. 8a,b and (2) since GD was not produced in this work, the GD separating column was not required. Accordingly, the utilization of DEC in GC production appeared to be promising.

VLE prediction for (a) DEC/EtOH system at 1 bar, (b) DEC/GC system at 1 bar.

Cases I and II were modeled in Aspen Plus V11, evaluated in Aspen Economic Analyzer, and optimized in MATLAB using the Genetic Algorithm. The results showed that Case II was more efficient than Case I due to its ability to shift the reaction equilibrium, overcoming the chemical equilibrium problem. Additionally, Case II combined the reaction and separation steps into a single unit, resulting in significant cost savings in capital investments, operating expenses, and utility expenses. Moreover, economic evaluation results indicated that Case II had a shorter payback period, and a higher internal rate of return when compared to Case I, making it a more attractive option for investment.

Lastly, Case II was compared with a similar process in which DMC was used as a raw material. As seen from the comparison, the RD system that utilized DEC yielded better specific energy consumption since DEC/EtOH/GC did not form azeotrope – making the mixture easier to separate, compared to the DMC/methanol/GC system. In addition, GD was not formed if DEC was used as a raw material given that the reaction condition was properly set. Accordingly, DEC could be another potential feedstock for the production GC.

All data generated or analyzed during this study are included in this published article (and its Supplementary Information file).

Claudio J. A. Mota, B.P.P., Lúcia de Lima A. Glycerol A Versatile Renewable Feedstock for the Chemical Industry. (Springer, 2017).

Pagliaro, M., Properties, applications, history, and market 1–21 (Elsevier, 2017).

Das, A. et al. Microwave-assisted synthesis of glycerol carbonate by transesterification of glycerol using Mangifera indica peel calcined ash as catalyst. Fuel 330, 125511 (2022).

Article MATH Google Scholar

Sedghi, R. et al. Turning biodiesel glycerol into oxygenated fuel additives and their effects on the behavior of internal combustion engines: A comprehensive systematic review. Renew. Sustain. Energy Rev. 167, 112805 (2022).

Article MATH Google Scholar

Keller, T. Chapter 8 - Reactive Distillation. In Distillation (eds Górak, A. & Olujić, Ž) 261–294 (Academic Press, 2014).

Chapter MATH Google Scholar

Lertlukkanasuk, N. et al. Reactive distillation for synthesis of glycerol carbonate via glycerolysis of urea. Chem. Eng. Process.: Process Intensif. 70, 103–109 (2013).

Article Google Scholar

Kulal, N. et al. Zn-doped CeO2 nanorods for glycerol carbonylation with CO2. ACS Appl. Nano Mater. 4(5), 4388–4397 (2021).

Article MATH Google Scholar

Bor-Yih, Y. & Chen, W.-J. Rigorous Simulation and Optimization of a Plant-wide Glycerol Carbonate (GC) Production Process. In Computer Aided Process Engineering 283–288 (Elsevier, Netherlands, 2020).

MATH Google Scholar

Yu, B.-Y. Development of two plant-wide glycerol carbonate production processes: Design, optimization and environmental analysis. J. Taiwan Inst. Chem. Eng. 117, 19–25 (2020).

Article MATH Google Scholar

Srivastava, P., Gupta, A. & Kaistha, N. Glycerol valorisation to glycerol carbonate: Process synthesis, design and control. Comput. Aided Chem. Eng. 53, 1369–1374 (2024).

Article MATH Google Scholar

Sun, D. et al. Process design, economic optimization, and environmental assessment for the production of glycerol carbonate via reactive distillation combined with pervaporation. Ind. Eng. Chem. Res. 61(44), 16552–16564 (2022).

Article Google Scholar

Zhang, J. et al. Thermodynamic and kinetic studies for synthesis of glycerol carbonate from glycerol and diethyl carbonate over Ce–NiO catalyst. Chem. Pap. 72(11), 2909–2919 (2018).

Article ADS MATH Google Scholar

Tunpaiboon, N. Industry Outlook 2021–2023: Biodiesel. [cited 2023 26/4/2023]; Available from: https://www.krungsri.com/en/research/industry/industry-outlook/energy-utilities/biodiesel/io/io-biodiesel-21. (2021).

Yang, F., Hanna, M. A. & Sun, R. Value-added uses for crude glycerol–a byproduct of biodiesel production. Biotechnol. Biofuels 5(1), 13 (2012).

Article PubMed PubMed Central MATH Google Scholar

Carlson, E. C. Don’t gamble with physical properties for simulations. Chem. Eng. Progress 92(10), 35–46 (1996).

MATH Google Scholar

Liu, J. et al. Liquid-liquid equilibrium of immiscible systems involved in glycerol transesterification with diethyl carbonate. Fluid Phase Equilib. 558, 113455 (2022).

Article Google Scholar

Gao, X. et al. Novel heat pump reactive distillation and dividing-wall column reactive distillation processes for synthesizing isopropyl acetate to save TAC and reduce CO2 emissions. Chem. Eng. Process. Process Intensif. 171, 108746 (2022).

Article MATH Google Scholar

Mathwork. What Is the Genetic Algorithm? ; Available from: https://www.mathworks.com/help/gads/what-is-the-genetic-algorithm.html.

Pandit, S. R. & Jana, A. K. Transforming conventional distillation sequence to dividing wall column: Minimizing cost, energy usage and environmental impact through genetic algorithm. Sep. Purif. Technol. 297, 121437 (2022).

Article MATH Google Scholar

Abril, A.F. Aspen Plus - Matlab Link. Available from: https://www.mathworks.com/matlabcentral/fileexchange/69464-aspen-plus-matlab-link. (2018).

Thanahiranya, P. et al. Succinic Acid Production from Glycerol by Actinobacillus succinogenes: Techno-economic, environmental, and exergy analyses. J. Clean. Prod. 404, 136927 (2023).

Article Google Scholar

Supramono, D. and Ashshiddiq, J. F. A. Economic potential of the manufacture of acrolein and propylene glycol from glycerol. In AIP Conference Proceedings. (AIP Publishing LLC, 2021).

Chang, W.M., Glycerol (medical grade) preparation method using a by-product of a bio-diesel process. Google Patents (2011).

Hebei Disha Import and Export Trade Co., L. Crude Glycerine 80% 85% Industrial Grade. Available from: https://hbdisha.en.made-in-china.com/product/lQXYMihdCkWa/China-Crude-Glycerine-80-85-Industrial-Grade.html.

Anhui Eapearl Chemical Co., L. 99.99% Purity Diethyl Carbonate High Quality. [cited 2022 18/10/2022]; Available from: https://eapearlchem.en.made-in-china.com/product/RFutBspDhPVT/China-Best-Sale-99-99-Purity-Diethyl-Carbonate-High-Quality.html.

Guedes, P. H. P. S. et al. Process simulation for technical and economic evaluation of acrolein and glycerol carbonate production from glycerol. Biomass Bioenergy 168, 106659 (2023).

Article MATH Google Scholar

Xilong Scientific Co., L. Denatured Ethanol 95% Sda-40-B-190 Ethyl Alcohols for Disinfectant. Available from: https://xilongchemical.en.made-in-china.com/product/aXPJsKcxXrhZ/China-Denatured-Ethanol-95-Sda-40-B-190-Ethyl-Alcohols-for-Disinfectant.html. (2022)

Turton, R., et al., Analysis, synthesis and design of chemical processes. Fifth ed. 2021: Pearson Education.

Palamutcu, S. 2 - Energy footprints in the textile industry. In Handbook of Life Cycle Assessment (LCA) of Textiles and Clothing (ed. Muthu, S. S.) 31–61 (Woodhead Publishing, Cambridge, 2015).

Chapter Google Scholar

Download references

This research project is supported by the Second Century Fund (C2F), Chulalongkorn University. We also would like to acknowledge Thailand Science Research and Innovation Fund Chulalongkorn University (No. 6641/2566).

Bio-Circular-Green-Economy Technology and Engineering Center, BCGeTEC, Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, 10330, Thailand

Bushra Chalermthai, Chayanin Sriharuethai, Apinan Soottitantawat, Suttichai Assabumrungrat & Pongtorn Charoensuppanimit

Control and Systems Engineering Research Laboratory, Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, 10330, Thailand

Chayanin Sriharuethai

Department of Chemical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA, 02139, USA

Bradley D. Olsen

Division of Chemical Engineering, Faculty of Engineering, Rajamangala University of Technology Krungthep, Bangkok, 10120, Thailand

Kanokwan Ngaosuwan

Center of Excellence in Particle and Materials Processing Technology, Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, 10330, Thailand

Bushra Chalermthai, Apinan Soottitantawat & Pongtorn Charoensuppanimit

Center of Excellence in Catalysis and Catalytic Reaction Engineering, Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, 10330, Thailand

Suttichai Assabumrungrat

You can also search for this author in PubMed Google Scholar

B.C.: Writing—Original Draft. C.S.: Writing—Original Draft, Investigation. B.D.O.: Supervision. K.N.: Methodology. A.S.: Conceptualization, Methodology. S.A.: Supervision. P.C.: Supervision, Methodology, Writing-Review and Editing.

Correspondence to Pongtorn Charoensuppanimit.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Chalermthai, B., Sriharuethai, C., Olsen, B.D. et al. Comparative study of conventional and process intensification by reactive distillation designs for glycerol carbonate production from glycerol and diethyl carbonate. Sci Rep 15, 1753 (2025). https://doi.org/10.1038/s41598-025-85974-4

Download citation

Received: 04 July 2024

Accepted: 07 January 2025

Published: 12 January 2025

DOI: https://doi.org/10.1038/s41598-025-85974-4

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative