*3.2. Three-Step Process*

#### 3.2.1. Three-Step Process—Selecting Key Parameters

The three-step process was successfully implemented in Matlab. In Figure 8, CO2 single-pass conversion (*XCO*2,*SP*) (Figure 8a), the required feed excess (Figure 8b), the optimal temperature (Figure 8c), and the total recycle stream (Figure 8d) are described as a function of the number of reactor modules and the purge fraction. Since this process considers three reaction steps with intermediate cooling, the simulations were limited to multiples of three as the total number of reactor modules.

**Figure 8.** Three-step process—CO2 single-pass conversion (**a**), required feed excess (**b**), optimal temperature (**c**), and total recycle stream (**d**) as a function of the number of reactor modules and the purge fraction.

A significant improvement was seen in the three-step process in relation to the one-step approach. For similar conditions (i.e., the same total number of reactor modules and purge stream fraction), CO2 single-pass conversion had approximately doubled, the required feed excess decreased by 60–70%, and the total recycle stream decreased by 50–70%. The optimal values for the reactor cooling fluid remained close to the ones of the first approach (between 230 and 260 ◦C).

Similarly to the one-step process, a purge fraction equal to 2% was chosen here, having a good compromise between minimizing the feed requirements and minimizing the total recycle stream. With this fixed purge fraction, a number of reactor modules equal to three was selected, as further increasing this amount gave limited improvement in the required feed excess and the total recycle stream, while considerably increasing equipment and catalyst costs.

When analyzing different scenarios in Matlab, the same cooling fluid temperature (*Tw*) was considered for all reactors. A further optimization was possible by allowing this temperature to be independently operated in each reactor. This possibility was checked for the chosen condition (2% purge fraction, three reactor modules), but only a marginal improvement was obtained (see Table 3), probably not justifying the increase in plant complexity. Therefore, in the detailed analysis, the cooling fluid temperature of all the reactors was set to 258.5 ◦C.

**Table 3.** Three-step synthesis—Performance indicators for two process approaches: same cooling fluid temperature in all reactors, and independent optimization of the cooling fluid temperature in each reactor.


Similarly to the one-step process, the average heat transfer coefficients were obtained for each reactor and given to Aspen Plus: *<sup>U</sup>*<sup>1</sup> = 327 W·m−2·K−1, *<sup>U</sup>*<sup>2</sup> = 285 W·m−2·K−1, *<sup>U</sup>*<sup>3</sup> = 246 W·m−2·K−1. The decrease in the coefficient values is associated with a decrease in total flow, due to intermediate product removal. Still, the heat transfer coefficients were higher than in the one-step process (155 W·m−2·K−1), which had lower flows for each reactor module because of parallel operation.

#### 3.2.2. Three-Step Process—Detailed Plant Simulation and Process Analysis

A detailed flowsheet of the three-step process presented in Figure 4 was implemented in Aspen Plus, considering a 2% purge fraction, three reactor modules working in series with intermediate product condensation, and the previously optimized temperature of the reactor cooling fluid (*Tw* = 258.5 ◦C). A detailed plant description, stream properties, and a picture of the flowsheet in Aspen Plus are provided in the Supplementary Material (Section F).

In Figure 9, the concentration of the products along the length of the three reactors is shown, as well as the product removal through the intermediate condensation steps. In Reactor 1, CO entered at a low concentration (1.3% mol/mol), peaked at *z* = 2.5 m, and left the reactor with a higher concentration (2.7% mol/mol). This CO production through the rWGSR increased the water concentration (*yR*1,*out <sup>H</sup>*2*<sup>O</sup>* = 5.6% mol/mol) and slowed down methanol production (*yR*1,*out MeOH* = 4.7% mol/mol).

In Reactors 2 and 3, the CO inlet concentration was significantly higher (3.0% mol/mol for both cases), causing its concentration peak to come much sooner (at 1.8 m and 1.25 m, respectively). After that, the WGSR was faster than its reverse reaction and the CO concentration decreased, leaving both reactors with an overall positive CO consumption. Therefore, the water concentration in Reactors 2 and 3 was maintained at lower levels (*yR*2,*out <sup>H</sup>*2*<sup>O</sup>* = 4.7% mol/mol, *<sup>y</sup>R*3,*out <sup>H</sup>*2*<sup>O</sup>* = 4.5% mol/mol), enhancing the final methanol concentration (*yR*2,*out MeOH* = 5.6% mol/mol, *<sup>y</sup>R*3,*out MeOH* = 5.8% mol/mol).

**Figure 9.** Three-step process—methanol, water, and CO concentration along the length of each reactor, as well as in the intermediate condensation steps (C1 and C2).

Water is known to accelerate the deactivation of Cu-based catalysts [67]. Therefore, the lower water concentration of the three-step process in relation to the one-step process (*y*1*s*,*out <sup>H</sup>*2*<sup>O</sup>* = 7.2%) should not only benefit the reaction rates, but also the catalyst lifetime.

In Table 4, the operating conditions and split ratios of the intermediate condensation steps are provided, while the reactor information is summarized in Table 5. Methanol and water were almost fully removed from the gas phase, but at the cost of ca. 9–13% CO2 condensation. The split ratios of CO2 and methanol were strongly dependent on temperature, with the chosen values (*T*<sup>1</sup> = 45 ◦C, *T*<sup>2</sup> = 30 ◦C) being derived from a sensitivity analysis.

**Table 4.** Three-step process—operating conditions and split ratios of the intermediate condensation steps.


**Table 5.** Three-step process—heat transfer, inlet mole flow, mole fractions, and methanol production in the reactor modules.


The methanol production was similar in Reactors 1 and 2 (1616 and 1589 kmol·h−1, respectively), while it was 18% lower in Reactor 3 (1325 kmol·h<sup>−</sup>1). This shows the positive effect of a higher CO concentration in the reactor feed, despite the lower total feed flow and CO2 inlet concentration of Reactors 2 and 3 in relation to Reactor 1.

The CO2 single-pass conversion (*XCO*2,*SP*) was 53.9%, with a selectivity to methanol of 99.8% and a selectivity to CO of 0.2%. The feed excess was 2.35%, leading to an overall conversion of CO2 to methanol of 97.7%. These values are in agreement with the Matlab simulations (*XCO*2,*SP* = 54.1%, *Exc* = 2.42%, overall CO2 conversion to MeOH = 97.6%).

The three-step approach was significantly superior to the one-step process, even using only half the number of reactor modules (three vs. six). This superiority is clear when looking at the CO2 single-pass conversion (53.9% vs. 28.5%), leading to a considerably higher overall conversion to methanol (97.7% vs. 94.3%).

With the heat integration, the three-step process was also self-sufficient in heat, while electricity was produced through a water Rankine cycle, reducing the total power consumption from 42.7 to 21.8 MW. The chemical conversion efficiency was *η*3*<sup>s</sup> CCE* = 85.6%, which was higher than the value of the one-step process (*η*1*<sup>s</sup> CCE* = 82.3%) and, therefore, even closer to the maximum possible value (*ηCCE*,*max* = 87.6%).

In Figure 10, an exergy analysis of the process is presented. The exergy efficiency was *η*3*<sup>s</sup> Ex* = 78.8%, an improvement from the previous approach (*η*1*<sup>s</sup> Ex* = 76.4%), with the total exergy losses decreasing in 13% (245.3 vs. 281.9 MW). Although the total power consumption decreased (42.7 vs. 47.4 MW), the net power consumption increased slightly (21.8 vs. 17.6 MW). This occurred because power generation was significantly lower in the three-step approach (20.8 vs. 29.8 MW) due to the much lower heat duty of the fired heater, as less reactant was lost in the purge streams.

**Figure 10.** Three-step process—exergy analysis. (**a**) Global exergy balance (total exergy input = 1157.5 MW). (**b**) Distribution of exergy losses (total = 245.3 MW).

Chemical reactions with heat recovery at low temperatures was also the main cause of exergy losses in the three-step approach (reactor modules: 66.0%, fired heater: 6.0%). Both processes lost approximately the same exergy in the reactor modules and the distillation column. The main improvement in relation to the one-step process was a much lower exergy loss in the fired heater (14.7 vs. 41.4 MW), as the total purge stream flow decreased by 59% (455 against 1100 kmol·h<sup>−</sup>1). Despite the higher number of cooling and warming operations and the higher total heat transfer duty in the three-step process (357.1 vs. 310.2 MW), the exergy losses in the heat exchangers were slightly lower for the three-step process (29.4 vs. 31.4 MW). Finally, moderate improvements were also seen in the compressors and pump (8.2 vs. 9.1 MW) and in the valves and turbine (5.0 vs. 7.4 MW).

In Table 6, the data comparing both processes is summarized, once again emphasizing the superior performance of the three-step approach.

**Table 6.** Data comparison between the one-step and the three-step approach.


#### *3.3. Techno-Economic Analysis*

In Figure 11a, the distribution of the equipment costs (*EC*) is presented, with the reactor modules and the compressors representing the majority of the costs (>75%). The total EC was 85.5 and 66.1 M€ for the one-step and the three-step approach, respectively. This significant improvement of the three-step process was a consequence of the intermediate condensation steps, requiring a lower total reactor volume (due to an enhanced reaction velocity), lower compressor size (due to a lower recycle flow), and lower furnace, turbine, and generator size (due to a lower purge flow). The cost reduction in the aforementioned equipment was significantly higher than the additional costs of the heat exchangers and flash drums from the intermediate condensation units. The total fixed capital investment (*FCI*) was 415.9 and 321.4 M€ for the one-step and three-step approach, respectively. The detailed estimated capacity and price of each equipment is presented in the Supplementary Material (Section H).

**Figure 11.** Distribution of the costs. (**a**) Equipment Costs (EC). (**b**) Net Production Costs (NPC).

In Figure 11b, the distribution of the net production costs (*NPC*) is detailed. The main operating costs were the reactant expenses (78–80% of *NPC*), with *ACC* contributing with only 4–5%, while the catalysts and electricity consisted of less than 3% of the *NPC*. Due to the higher overall CO2 conversion to methanol, the *NPC* of the three-step process was 5.7% lower than the one-step approach. The detailed *OPEX* costs are presented in the Supplementary Material (Section H).

In Table 7, a summary of the overall costs is presented. The *NPC* was 920 and 868 €·ton−<sup>1</sup> for the one-step and the three-step process, respectively, corresponding to an improvement of 5.7% for the new process. Besides the hydrogen and carbon dioxide costs, the fixed capital investment (*FCI*) and the discount rate were the most sensitive parameters to the methanol selling price, as shown in the tornado analysis (see Figure 12).

**Table 7.** Summary of the costs of the one-step and the three-step process.


**Figure 12.** Sensitivity analysis of the main cost factors in relation to the net production costs (NPC). Variation of ± 20% in each factor. (**a**) One-step process. (**b**) Three-step process.

In Figure 13, the net production costs are plotted against the hydrogen price. Although the methanol market price in Europe was still significantly below the values (495 €·ton−<sup>1</sup> in February 2022), [68,69] the green methanol produced from the proposed process would become economically competitive if the green hydrogen price reached 1468 €·ton<sup>−</sup>1.

**Figure 13.** Net production costs of methanol as a function of green hydrogen price.

#### **4. Conclusions**

A detailed study of a methanol synthesis plant from H2 and CO2 with intermediate condensation units (the three-step process) is presented and compared with the conventional approach (the one-step process). The total production was fixed at 1.16 Mton MeOH·a−1. The processes were first implemented in Matlab in order to critically analyze the number of reactor modules, the purge fraction, and the reactor operating temperature. Using the most suitable process parameters, detailed plants of both approaches were implemented in Aspen Plus, including heat integration and a water Rankine cycle to make use of the reaction enthalpy. Finally, techno-economic analyses were applied. Both processes offered an excess of heat, which was used to generate electricity in our work, but could alternatively supply other plants (e.g., CCU, OME synthesis) in a larger process integration.

It was demonstrated that CO2 single-pass conversion almost doubled when including intermediate condensation steps (53.9 vs. 28.5%), resulting in a significantly higher overall conversion to methanol (97.7 vs. 94.3%) and in a higher exergy efficiency (78.8 vs. 76.4%). Because of the enhanced conversion, the new process required lower recycle and feed streams, decreasing net production costs by 61.2 M€·a−<sup>1</sup> (5.7%). Although additional equipment (i.e., heat exchangers and gas–liquid separators) is necessary, the improved

process was significantly more efficient than the conventional approach, requiring lower sizes of the main equipment (e.g., compressors, reactors, fired heater). Consequently, according to our analysis, the total investment costs were 94.5 M€ (22.7%) lower than for the conventional process.

Intermediate condensation steps are therefore highly recommended for methanol production from H2/CO2, reducing costs by improving CO2 equilibrium conversion to methanol while using commercially proven technology. Besides, since water accelerates the deactivation of Cu-based catalysts, product intermediate removal should increase catalyst lifetime, as the average water concentration in the reactor is significantly lower than in the conventional process.

With our proposed process, the methanol net production costs amounted to <sup>868</sup> €·ton<sup>−</sup>1, which are still significantly higher than the current market price (495 €·ton<sup>−</sup>1) but is believed to become economically viable with an effective reduction in the price of green hydrogen.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/pr10081535/s1, Figure S1: One-step process—Aspen Plus flowsheet; Figure S2: Three-step process—Aspen Plus flowsheet; Table S1: Parameters for the estimation of the specific heat capacity and specific enthalpy of selected components in the gas phase; Table S2: Liquid and gas fractions (% mol/mol) of the phase separation via flash drums and the separation via the distillation column in the one-step process. Values taken from Aspen Plus calculations and used for the Matlab simulations; Table S3: Liquid and gas fractions (% mol/mol) of the phase separation via flash drums and the separation via the distillation column in the three-step process. Values taken from Aspen Plus calculations and used for the Matlab simulations; Table S4: Aspen kinetic factor and Model-6p corresponding expressions; Table S5: Coefficients of the driving force constant and the corresponding expressions from Model-6p; Table S6: Concentration exponents (*υj*) of the driving force expression; Table S7: Adsorption constants and the corresponding expression of Model-6p; Table S8: Concentration exponents and the corresponding values of Model-6p; Table S9: Properties of the streams from the one-step process; Table S10: Molar composition (% mol/mol) of the streams from the one-step process; Table S11: Properties of the streams from the three-step process; Table S12: Molar composition (% mol/mol) of the streams from the three-step process; Table S13: Dimension of the flash drums of the one-step and the three-step processes; Table S14: Global heat transfer coefficients, heat transfer duty, and estimated surface area of the heat exchangers of the one-step and the three-step process; Table S15: Calculation of the Capital Expenses (CAPEX) depending on the total equipment costs (EC); Table S16: Estimation of indirect operating expenses (*OPEXind*); Table S17: Equipment characteristic dimensions and equipment costs (EC) of the one-step approach. All equipment was built with carbon steel. All equipment reference prices were taken from Peters et al., except for the power generator, whose ref. price was taken from Henning and Haase; Table S18: Equipment characteristic dimensions and equipment costs (EC) of the three-step approach. All equipment was built with carbon steel, and the costs included 10% delivery costs. All equipment reference prices were taken from Peters et al., except for the power generator, whose ref. price was taken from Henning and Haase; Table S19: Detailed operating expenditures (OPEX) of the one-step and the three-step approach.

**Author Contributions:** Conceptualization, B.L.d.O.C. and J.S.; methodology, B.L.d.O.C., K.J. and P.B.; software, B.L.d.O.C., K.J. and P.B.; formal analysis, B.L.d.O.C.; writing—original draft preparation, B.L.d.O.C.; writing—review and editing, K.J., P.B., K.H.D., S.P., N.D. and J.S.; visualization, B.L.d.O.C. and P.B.; supervision, K.H.D., S.P., N.D. and J.S.; funding acquisition, K.H.D., S.P., and J.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) (process number: 88881.174609/2018-01) and by the Helmholtz Research Program "Materials and Technologies for the Energy Transition (MTET), Topic 3: Chemical Energy Carriers". We also acknowledge the support from the KIT-Publication Fund of the Karlsruhe Institute of Technology.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
