A review of process intensified CO₂ capture in RPB for sustainability and contribution to industrial net zero

1 Introduction

The most challenging task currently is to control the increased levels of greenhouse gases (GHGs) in the atmosphere, which leads to global warming. Among the various GHGs, carbon dioxide (CO₂) significantly affects global warming. GHGs are being released into the atmosphere through multiple activities such as fuel combustion, manufacturing, construction, industrial process, transportation, and agriculture. CO₂ emissions from factories and power plants are the most significant sources of atmospheric pollution. Following the International Panel on Climate Change (IPCC) report, the CO₂ level will have increased to 570 ppm in the atmosphere by 2,100. This CO2 level will cause an increase in global temperature of approximately 1.9°C. (IPCC, 2022). There is a need to reduce these emissions of GHGs, and hence carbon capture, storage, and utilization (CCSU) is a solution. Though various technology options are available, such as pre-combustion, oxyfuel combustion, and post-combustion carbon capture (PCC), post-combustion is the widely preferred method as it can be used in the pre-existing plant compared to the other two methods (Marx-Schubach and Schmitz, 2019). The technologies available for the post-combustion process include membrane separation, solvent absorption, cryogenic fractionation, and adsorption using solid sorbent (Afkhamipour and Mofarahi, 2013; Dey et al., 2018). Chemical absorption is a widely observed post-combustion method used for capturing carbon compared to other techniques due to its high absorption rate (Joel et al., 2014).

The desirable properties of a suitable solvent are high thermodynamic capacity, high absorption rate, low energy for regeneration, less degradation, lower volatility, and lower corrosivity (Dey et al., 2020). Due to its desirable properties, the solvent used in industry is an amine-based solvent such as primary, secondary, tertiary, and sterically hindered amines (H. Li et al., 2013).

For years, researchers have studied post-combustion CO₂ capture through chemical absorption using different solvents in the conventional column. These include aqueous solutions of monoethanolamine (MEA), diethanolamine (DEA), methyl diethanolamine (MDEA), 2-amino-2-methyl-1-propanol (AMP), and blended amine AMP and piperazine. The most widely used alkanolamines in the literature include MEA, DEA, and MDEA (Dash et al., 2012). Currently, many experiments have been conducted using blended amine AMP and piperazine or its derivatives (Dash et al., 2022; M; Wang et al., 2011; Wu et al., 2020). Mathematical models and operational parameters of the conventional absorber have been studied, modeled, and simulated in various research papers (Pandya, 1983; Akanksha et al., 2007; Kvamsdal et al., 2009; Harun et al., 2012; Koronaki et al., 2015; Jin et al., 2019; Shahid et al., 2021). However, the large size of the conventional column has been cited as a hindrance to CO₂ capture. The problems are that it demands significant capital and operating costs, uses a lot of energy, and requires intercooling with heat integration (Chamchan et al., 2017). Thus, more efficient gas-liquid reactors employing new principles and technology for carbon dioxide capture are needed.

Process intensification technology is used to reduce capital costs and significantly improve process dynamics. The equipment can be made smaller using centrifugal forces, electrical fields, and microwaves or by downsizing to the meso- or microscale (Reay, 2008; Cortes Garcia et al., 2017). One of the most cutting-edge intensification technologies invented by Mallinson and Ramshaw (Mallinson and Ramshaw, 1981) is currently the rotating packed bed (RPB). Podbielniak, (1935) introduced the concept of reducing the height of the distillation column and designed a conceptual contactor for distillation. Mallinson and Ramshaw introduced the idea of rotating packed bed columns similar to that proposed by Pilo and Dahlbeck for distillation and absorption (Rao, 2022). The RPB is used in place of a packed bed absorber to capture CO₂ (M. Wang et al., 2015).

An RPB has been designed and installed at our laboratory to make a complete pilot plant for CO₂ capture. The process flow diagram of the RPB shown in Figure 1A, has the appearance of a doughnut, which spins to produce acceleration caused by centrifugal force. This force is far better than the acceleration caused by gravity applicable in conventional absorption columns for the promotion of micromixing and enhanced mass transfer. The spinning packed bed consists of a rotor with packing that has a considerable surface area and is shaped like a doughnut. A schematic diagram of the rotor is presented in Figure 1B. The solvent is first supplied to the rotor’s center. It is forced by the rotor force to flow in the circular direction via the space between the packing. Under the influence of the pressure gradient, the gas flow occurs concurrently in the direction opposite to the flow of the solvent. When the shaft is connected to a motor, the rotor may be rotated at high rates, creating significant gravitational acceleration. Using centrifugal force, a liquid distributor may transport the liquid to the packing’s inner edge (Ayash and Mahmood, 2022). This enhances the efficiency of mass transfer and is also an essential component of process intensification.

FIGURE 1. (A) Process Flow diagram RPB. (B) RPB Rotor with flow directions.

An RPB works on the principle of centrifugal acceleration ($\left.a_c\right)$, so an increase in ($a_c$) leads to a higher slip velocity, increasing flooding character and interfacial shear stress. This, in turn, enhances the mass transfer coefficient (Jassim et al., 2007). Due to the stronger centrifugal acceleration compared to gravity, packings having a greater contact area may be utilized, allowing the reactor to run at a higher flooding limit (Dhaneesh and Ranganathan, 2022). Using RPBs rather than packed bed absorbers shows a higher mass transfer rate. This is accomplished by expanding the interfacial area through a droplet, and film flow is attained via rotational motion, which produces a substantial decrease in the dimension and mass of the rigs (Guo et al., 2019). The reduction in the size of an RPB enables it to be built more inexpensively than traditional fixed-bed reactors. (Pan et al., 2017). A conventional column necessitates a substantial packed volume, resulting in higher capital costs. According to M. Wang et al. (2015), an RPB has excellent potential for minimizing operating and capital expenses for CO₂ capture compared to conventional columns. This article provides a detailed review of the current research on intensified post-combustion capture and discusses experimental rigs (intensified absorber, stripper, and heat exchanger), international laboratory research, solvent sampling, and simulated models. RPBs may be used as a highly efficient multiphase contactor due to the gravity created by the rotation of packing, which can approach 5,000 m⁻². Using an RPB absorber improves separation efficiency, as demonstrated by Chamchan et al. (2017), and causes a considerable drop in height transfer unit (HTU). It utilizes just 35% of the packed bed absorber volume.

However, RPB equipment is associated with more significant maintenance and repair expenses. Furthermore, the design of an RPB is more complex; design methodologies and relevant models are not as well defined as conventional packed columns. Despite this, an RPB is favored for CO₂ capture over a packed bed column as the HiGee technology has more advantages. Because of an RPB’s high-efficiency mass transfer performance and reduction in size in comparison to the conventional column, it has wide industrial applications. These applications are selective removal of H2S, desulphurization, dedusting, and deaeration of water, as well as an application in the pharmaceutical industry where higher purity of a product is required (Neumann et al., 2018).

Because of these benefits, it is good to have high-quality research on CO₂ capture with RPB absorbers through different experiments and modeling approaches to meet desirable environmental requirements. CO₂ capturing with RPB absorbers is comparatively unexplored and efficient. Thus, it has significant scope for research using different experiments and modeling approaches to meet desirable environmental requirements. Detailed and systematic modeling of RPBs is essential for large-scale CO₂ capture applications because experimental work is a long process and is limited. One can be robust in operation through modeling. Here, RPB models proposed by researchers during the last few years are reviewed and analyzed to predict CO₂ absorption efficiency using different characteristics and modeling approaches.

In this work, the first section reviews the different modeling approaches used in CO₂ capture utilizing RPB. The second contains sub-sections summarizing the study of hydrodynamic characteristics, the study of correlations of mass transfer characteristics of RPB, CO₂ capture modeling and process simulation using RPB, experimental review, and techno-economic analysis. The third section presents a conclusion and proposes fields for future work.

2 Modeling approaches

Various factors such as the rotational speed of RPB, CO₂ partial pressure, liquid-gas flow rate, the temperature of the liquid phase, absorbent concentration, and CO₂ loading solvent affect the working of an RPB. Numerous modeling techniques based on hydrodynamic analysis and the study of mass transfer characteristics using different operational parameters can be used to examine an RPB. A summary of these techniques is presented in Table 1. The analysis of fluid flow behavior in terms of the liquid holdup, flooding, pressure drop, and liquid flow pattern constitutes the hydrodynamic characteristics. The effect of mass transfer properties of the RPB can be investigated by both gas and liquid phase mass transfer coefficients and effective interfacial area. A numerical approach has been applied to study the process modeling of RPB. Different computational fluid dynamics (CFD) approaches, such as Eulerian-Eulerian or Eulerian-Largrangian, have been used to imitate the gas flow state, fluid drop path, and mass transfer coefficients in an RPB (Zhao et al., 2016). Furthermore, studies have analyzed turbulence models such as k-$\epsilon$, RNG k-$\epsilon$, and VOF (Xie et al., 2017; W; Yang et al., 2010; Y; Yang et al., 2015).

TABLE 1. Modeling approaches for RPB.

Additionally, researchers have also developed machine learning or artificial intelligence to simulate absorption in an RPB using an ANN (artificial neural network). The experimental and modeling approaches can be time-consuming at times. As a result, the ANN approach can be involved. An ANN can be used to predict the coefficient of mass transfer, hydrodynamic parameters, and thermodynamic properties of CO₂ in the absorption process (W. Li et al., 2016; Z. W; Liu et al., 2018; Zhao et al., 2014). Furthermore, expertise in techno-economic analysis is required to grasp the advantage of the proposed project, taking both benefits and costs into account. (T. L. Chen et al., 2020).

2.1 Hydrodynamics parameters

A thorough understanding of hydrodynamic parameters is vital for improving the performance of an RPB system. As a result of the high rotational conditions, the hydrodynamics of the RPBs are altered in ways that significantly affect the gas-liquid mass transfer coefficient. The effect occurs by changing the RPB’s hydrodynamics, flooding, liquid holdup, pressure drop, and fluid dispersion.

2.1.1 Liquid holdup (${\epsilon }_l$)

Liquid holdup describes the percentage of liquid volume in the bed after draining. Different correlations have been considered in the literature through modeling depending on the packing type used and flow theory (either film flow or droplet flow). The correlations were then validated with experimental data. These correlations deviate from the existing correlations for conventional columns because of the rotational flow in the RPB.

Basic and Dudukovic (Basic and Dudukovic, n.d.), working under the premise of film flow with complete wetting, presented a correlation for the liquid holdup. It was based on experimental data measured with the electrical resistance measurement method. The research discovered that this expression corresponds with the holdup data as a function of the operational parameters. Furthermore, (Burns et al., 2000) studied the ${\epsilon }_l$ correlation using the same electrical resistance measurement method with the packing of high porosity. They found that the local packing radius inversely affects the holdup. A practical approach for determining the liquid holdup in RPBs was developed by Y. H. Chen et al. (2004) and involved measuring the quantity of liquid that remained in the RPB after it had stopped operating. The correlation considered the gas flow, which previous researchers had not considered. However, its effect was not that strong and it was observed that ${\epsilon }_l$ was directly proportional to the liquid flow rate. Y. Yang et al. (2015) derived two correlations for ${\epsilon }_l$ using different packing. The inference made was that viscosity, liquid flow rate, and rotating speed affect ${\epsilon }_l$. Correlations of the liquid holdup are listed in Table 2 for the RPB.

TABLE 2. Liquid holdup correlations for RPB.

2.1.2 Pressure-drop ($\left.∆P\right)$

Pressure drop indicates the resistance a fluid encounters due to friction or other forces acting on it, such as gravity, and changes with the type of packing. Studies have been made on the empirical correlations of pressure drop with different radii, porosity, interfacial area, and type of packing. Correlations have been developed using these parameters. Keyvani (Keyvani and Gardner, n.d.) used a momentum integral balance on the rotor and the area between the rotor and container to simulate pressure drop. It was observed that $∆P$ rose with the rise to the square of rotating speed and when gas and liquid flow rates increased. Kelleher Kelleher and Fair, (1996) obtained a correlation between pressure drop and noted that $a_c$ and “Re” is the major contributor to the model. Sandilya et al. (2001) calculated the friction factor of an RPB by investigating the pressure drop. They studied it in two halves, $∆P_{0}and∆P_a$, using wire gauze. The impacts of gas and liquid flow rates and rotating speed on pressure drop were explored. Chandra et al. (2005) investigated the total pressure drop across the rotor as the sum of pressure drops due to the momentum ($\left.∆P_m\right)$, the friction provided by packing ($\left.∆P_f\right)$, and the centrifugal force ($\left.∆P_c\right)$. It was discovered that the total pressure drop was the sum of $∆P_f$; $∆P_c$, as $∆P_m$ was negligible. They also concluded that $∆P$ across the rotor was directly proportional to gas velocity and rotational speed. The correlations for a given RPB with distinct packing material are given in Table 3.

TABLE 3. Pressure drop correlations for RPB

Here, correlations are derived for a specific RPB with different structures and packing materials. The efficacy of the models or correlations to investigate pressure drops in various RPBs is still to be validated.

2.1.3 Flooding

Flooding occurs when the liquid stops flowing down the column due to the relative vapor flow rates being such that the drag force is higher than or equal to the gravitational force. The flooding phenomenon is different for an RPB compared to conventional columns due to differences in flow area and centrifugal acceleration (Rao et al., 2004). Due to the increased artificial gravity, RPBs may reach larger flooding capacities than conventional columns (Hendry et al., 2020). Two correlations have been used to study flooding: Sherwood’s correlation for the packed column and Lockett’s correlation for RPB.

Sherwood’s approach (Sherwood et al., 1938): $\frac{a_t{u}_G^2}{g{\epsilon }^3}{\left(\frac{{\rho }_g}{{\rho }_l}\right)}^{0.2}{\mu }_l^{0.2}$ vs. $\frac{L}{G}{\left(\frac{{\rho }_g}{{\rho }_l}\right)}^{0.5}$ and this gives the flooding limitation.

Wallis’s approach (Wallis, 1969) included some more parameters and proposed a relation:

Lockett’s approach (Lockett, 1995) for an RPB has the form: $C_G^{0.5}+m{C}_L^{0.5}=1.57{\beta }^{0.25}{a}_t^{-0.25}{\frac{{\mu }_L}{{\mu }_w}}^{-0.03}$

2.2 Mass transfer correlations

Mass transfer is defined as mass movement from one phase to another. The main reason behind using an RPB instead of the conventional column is its improved mass transfer efficiency. The mass transfer depends on the motion of fluid and hydrodynamic characteristics. Mass transfer of species changes with reaction, time, and space; hence, different mass transfer correlations are used to predict the mass flux of the process. Mathematical modeling of the RPB system requires studying the system of equations based on energy and mass balance, which uses a mass transfer coefficient in the gas-liquid phase and an effective interfacial area. Penetration theory, surface renewal theory, and two film theory are the three theories used to model mass transfer. Based on these theories, different empirical correlations are studied in the literature.

2.2.1 Gas phase mass transfer coefficient $\left(k_G\right)$

Y. S. Chen, (2011) developed a correlation utilizing the two-film theory. The bulk of the data could be accurately predicted using this correlation. The impacts of the flow rates of gas and liquid, the viscosity of the liquid, and centrifugal acceleration on $k_G$ were studied. Furthermore, a comparison of the $k_G$ of the conventional column and the rotating packed bed was made. The $k_G$ was found to be similar in both reactors because the behavior of the gas flow within the rotor was similar to the conventional column. Y. Li et al. (2017) utilized the volumetric gas-side mass-transfer coefficient ($k_{G}a$) to estimate the local gas phase mass transfer coefficient. This was calculated by air-stripping the ethyl alcohol from an alcohol-water mixture in RZB. They found that the values of $k_G$ was proportional to $R_s$, gas volumetric flow rate, and liquid volumetric flow rate, according to the experimental data. Research by Lin and Kuo, (2016) showed that mass transfer was achieved in the absorption of CO₂ using a solution of MEA with blade packings in the RPB. They observed that each RPB’s $k_G$ value was favorably impacted by the speed of rotation. Moreover, the $k_G$ value depends on the bed’s radius, and the flow rates of both the liquid and gas were accompanied by an increase in the $k_G$ values for each RPB. The correlations for $k_G$ found in the literature have been presented in Table 4.

TABLE 4. Gas phase mass transfer coefficient correlation in RPB.

2.2.2 Liquid phase mass transfer coefficient $\left(k_L\right)$

Absorption usually occurs in liquid, hence the $k_L$ model must be validated using experimental data to explain system behavior. Mass transfers are generally determined in RPBs using $k_{L}a$ instead of $k_L$ to overcome the difficulties in predicting the interfacial area, $a_e$. With the use of penetration theory, Tung and Mah (Tung and Mah, 1985) replaced gravitational acceleration with centrifugal acceleration to develop a correlation in an RPB for the ${\mathrm{k}}_{\mathrm{L}}$. This study did not use the effect of Coriolis acceleration and the geometry of the material used in packing. (Luo et al., 2012a) computed ${\mathrm{k}}_{\mathrm{L}}\mathrm{a}$ using surface renewal theory. They found a margin of error between the experimental results and the model predictions of nearly 15%. This was based on the premise that liquid droplets were present in blade packing. Y. S. Chen et al. (2005) examined the mass transfer effectiveness in an RPB with distinct packed bed radii and considered end effects for the calculation of ${\mathrm{k}}_{\mathrm{L}}\mathrm{a}$ in an RPB. This was applied to RPBs of various diameters and viscous Newtonian and non-Newtonian fluid systems. Similar correlations have been studied in the literature and are shown in Table 5.

TABLE 5. Liquid phase mass transfer coefficient correlation in RPB.

2.2.3 Effective interfacial surface area

Effective interfacial surface area $\left(a_e\right)$ is used to define the mass transfer properties of a reactor and is mostly affected by hydrodynamic properties. Correlations on the $a_e$ in conventional columns, studied by Onda et al. (1968); Billet and Schultes, (1999), have been used for RPBs by replacing “g" with “r ${\omega }^2$". Tsai and Chen, (2015) studied this correlation and noticed that the effective interfacial surface area was influenced by the speed of the rotor and $F_l$ . Chu et al. (2015) looked at the effective interfacial area using nickel foam packing and a combination of static rings and rotational packing. A rotor of two-stage counter-current RPB was used. The efficiency of CO₂ absorbed and $a_e$ were evaluated using cutting-edge “five-point” techniques. Experiments showed that the lower and upper packing zones were much better at absorbing CO₂ than the upper and lower cavity zones. This was true for distinct $R_s$ and liquid-gas flow rates. The correlations studied in the literature are tabulated here in Table 6.

TABLE 6. Effective interfacial surface area.

The correlation of ${\mathrm{k}}_G$, ${\mathrm{k}}_L$, and ${\mathrm{a}}_e$ are highly influenced by the experimental data and the fitting function used. As a result, generalizing the found correlations to other RPBs or operating situations is problematic.

2.3 Process modeling review

Process modeling is a technique used to predict the behavior of a system through computer techniques and graphical representation. It is further used to understand the influence of parameters on the system. Developing an RPB model involves chemical reaction equations, rate constants, and relations for calculating physical properties. These equations are included in the process modeling algorithm using tools for simulation and analysis and are shown here through a flowchart in Figure 2. For an RPB, a process model will be helpful for design, scaling up, debugging, process monitoring, mechanistic understanding, and optimal design and process evaluation. RPB modeling studied by various researchers is reviewed in this section.

FIGURE 2. Flowchart for the modeling procedure of RPB.

The assumptions made for the model and the calculation of the operational properties are different in different modeling studies. Since the shape of an RPB is circular, the common assumptions made by the researchers are that the flow of fluids is in counter-current and radial directions, shown in Figures 1A, B. To simulate the RPB, distinct theories have been proposed for mass transfer, such as penetration theory, film theory, and surface renewal theory. Sun et al. (2009) established a mathematical system for the absorption of CO₂ and NH₃ in water for an RPB. The mass equation of CO₂ in a droplet, as well as the film zone, was expressed in terms of a partial differential equation with Neumann boundary conditions. Through the model, an overall volumetric mass transfer coefficient $\left(K_{G}a\right)$ was predicted. Furthermore, it was observed that $K_{G}a$ was proportional to rotational speed, the volumetric flow rate of gas and liquid, and the NH₃/H₂O mole ratio. Kang et al. (2014) suggested an absorption model of CO₂, using MEA with the mass transfer, that uses two film theory. A system of differential equations based on material and energy balance equations has been developed for the gas and liquid phases. The boundary conditions were taken on the concentration of components and temperature of the gas and liquid phases at the inner-outer radii of the RPB. The proposed model took into account water evaporation and did not rely on the isothermal assumption. Joel et al. (2015) studied the rate-based model to analyze mass transfer using two sets of correlations in an RPB. The model has been implemented in FORTRAN and linked to ASPEN Plus. Through model validation, it was discovered that the second set of correlations was superior. Furthermore, it was determined that the amount of CO₂ captured and the flue gas flow rate rises with an increase in lean-MEA temperature. The model validation reports for the RPB, along with the modeling and simulation for CO₂ capture, are presented in Table 7.

TABLE 7. Modeling of RPB for CO₂ capture.

The reliability of these models is significantly dependent on the availability and effectiveness of existing correlations. As a result, developing trustworthy correlations is the foundation for adequately simulating the process using different modeling software. Various correlations with regard to mass transfer are tabulated in Tables 4, 5. The relation shows the internal connection of distinct operational parameters, physical properties of the process, and dimensionless numbers using the different packing materials. Other researchers studied the correlations of ${\mathrm{k}}_G$, ${\mathrm{k}}_L$, and ${\mathrm{a}}_e$ Taking different operational parameters, regressing constants’ values, and validating the relations with experimental data. These characteristics can also be interpreted and simulated using CFD as it can analyze the flow behavior in a better way. Hence, good predictions of these characteristics could be achieved.

2.4 Review of experimental data with RPB

In recent years, for the reduction of CO₂ emissions, a PCC process with the RPB technique has been investigated. Many researchers are working to determine CO₂ absorption efficiency in an RPB with different packing materials, dimensions of the equipment, and operational parameters. An experiment on the chemical absorption of CO₂ was conducted by Cheng and Tan, (2009) in an RPB utilizing a combination of amine solvents (MEA, AEEA, PZ, and AMP). The researchers found that higher temperatures increased absorption efficiency (303–333 K). It was also shown that a higher percentage of PZ in the mixture results in higher capture efficiency. For CO₂ absorption, Mohammadi Nouroddinvand and Heidari, (2021) performed experimental results employing an Arc-blade RPB in MEA (0.5–2.0 M) solvent. Wire mesh in the stainless-steel packaging was used. Two rectangular gas inlets, each 75 mm in length and 3 mm wide, were employed instead of circular designs. The obtained efficiency for capturing CO2 was 49%–78%, with $F_G=56\frac{l}{\mathit{min}},F_l=0.3-0.6l/\mathit{min},$; $R_S=300-1200rpm$. Utilizing the tri-solvent of N-methylpyrolidone (NMP), amorphous ammonium phosphate (AMP), and AEEA (2-(2-aminoethylamino) ethanol), Y. Wang et al. (2021) studied the absorption and desorption process for CO₂ capture in RPB. The measured CO₂ absorption efficiency was 89%. The findings reveal that this approach has promising industrial applicability, with a regeneration consumption of energy of 2.46 GJ/ton CO₂. Single and blended amines MEA, 2-methylamino-ethanol (MME), and PZ ranging from 30 to 100 wt% were tested for their ability to capture CO₂ in an RPB by Ma and Chen (Ma and Chen, 2016). The experimental findings revealed that the CO₂ removal efficiency was greatly enhanced by raising the alkanolamine concentration and the $R_S$ up to 1800 rpm. However, when the $F_g$ rose, the removal efficiency declined. Since MMEA interacts with CO₂ at a quicker rate than MEA, MMEA-based systems demonstrated greater efficiency in the removal process. Mixtures of alkanolamines with PZ were more efficient at sequestering CO₂ than either component used alone. Table 8 presents the summary of the experimental data from the literature, which gives a comparison of different performance studies using an RPB and would help in deciding on experimental conditions for future experiments.

TABLE 8. Experimental data review.

2.5 Techno-economic analysis

Implementing an RPB helps lower operating costs and environmental impact by reducing reactor size and capital expense. There are few studies in the literature that compare the environmental and economic effects of the RPB process to a conventional column process. There are benefits of using an RPB over the conventional column. Joel et al. (2014) compared conventional columns and RPBs and found that reduction in size had a crucial influence on absorption efficiency. Cheng et al. (2013) investigated the regenerator’s regeneration energy and found that an RPB with just a 10th of the volume of a conventional column could achieve the same regeneration efficiency. As a result of the enhanced heat transfer zone in the RPB, less vapor lean MEA needed to be transferred from the reboiler to the RPB. R. Zhang et al. (2017) used three mixed solvents MEA-MDEA-PZ, to study CO₂ absorption and desorption. An increase in the MDEA/PZ mole fraction among these blends resulted in a reduction in energy use and an increase in the rate of CO2 absorption. They also looked at the MEA-MDEA-PZ combination, which compared to 5M MEA, used 15.22%–49.92% less energy. M. Wang et al. (2015) noted that an RPB has good mass transport properties and is the most suited-technique for PCC. The performance of an RPB was methodically assessed from the engineering, environmental, economic, and energy perspectives by T. L. Chen et al. (2020). To propose a superior technical option compared to current CCSU facilities, a cost-benefit analysis was carried out utilizing running costs, carbon credit, and reductions in air quality. According to (Otitoju et al., 2023), packed bed absorbers’ and strippers’ size adds to the problems. If the absorbers and strippers can be replaced with RPB equivalents, CO₂ collection could be better and cheaper. A rate-based model in the steady state of the RPB absorber was proposed and verified for this study using Aspen Custom Modeller. It was used to perform technical and economic evaluations of operating big RPB absorbers using concentrated MEA (55–75 wt%). Compared to packed bed absorbers, the RPB absorber reduced the volume by a factor of 4–11 at a concentration of 55 wt% MEA, according to the technical evaluations, and had lower capital costs while having the largest volume reduction factors (3%–53% lower).

In conclusion, based on thorough and precise process models, it is necessary to estimate the entire costs (capital and operational expenses) utilized to capture CO₂ in intensified and traditional PCC processes. Different modeling approaches have been discussed in detail. The conclusion reached in this context is that the solution of the model for carbon absorption depends on the conservation of mass and energy balance equation with appropriate model assumptions and using appropriate relations for simplifying and solving the system of equations. The CO₂ mole fraction is predicted in outlet gas at numerous liquid flow rates, rotational speeds, and temperatures. The impact of different operating factors on the coefficient of mass transfer in RPB can be anticipated with the aid of a model. In addition to describing how mass transfer intensification occurs in an RPB, the gas-liquid mass transfer characteristic mathematical model also provides a theoretical foundation for the creation of RPBs and their actual application. Techno-economic analysis was discussed, which is necessary to optimize the entire expense utilized in CO₂ capture with an RPB having advantages over the conventional column. In chemical absorption, further research is needed to reduce solvent degradation, reducing the requirement for solvent make-up and, consequently, decreasing operational costs. New configurations and integration arrangements will contribute to cutting down energy consumption and, ultimately, minimizing expenses. Additives, for example, may be used to decrease surface tension and contact angles and thereby increase absorption rates utilizing CO₂ transfer and solvent wetting. The challenges that remain are actively being pursued across the policy, economic, environmental, and technology communities at all levels.

From the viewpoint of reducing emissions of CO₂ into the atmosphere, carbon capture utilization and sequestration (CCUS) technology needs to be retrofitted into power plants. Research has been done on the modeling and simulation of CO₂ capture for the conventional column, but an RPB is more efficient. More research is required in this domain to make the process of CO₂ capture faster with a lower cost. In the future, more modeling attempts are necessary for RPB absorbers in a steady and dynamic state with different solvents. Better operational parameters which influence the mass transfer of fluids are being investigated, which need to be studied to enhance the predictions of mass flux and CO₂ absorption efficiency. Furthermore, factors that influence the cost of the process need to be investigated for techno-economic analysis so that the process of CO₂ capture also becomes cost-effective. Utilizing CFD and an ANN are also promising approaches for studying hydrodynamic characteristics and mass transfer. The theoretical analysis and modeling of the RPB regenerator is the new scope of research work required. Understanding the complete absorption and desorption process through absorber and desorber together will scale up the CO₂ capture process.

3 Conclusion and future scope

RPB process intensification technology for the PCC process fits under sustainable development and industrial decarbonization. An RPB is a better option for CO₂ capture due to its advantages compared to the conventional column. However, it still has some disadvantages, such as large pressure drops and other mechanical wear and tear issues. The conventional column requires a considerable height, high capital cost, and more energy efficiency to regenerate the solvent. This equally applies to an RPB as it is a regenerative chemical absorption process. Both processes require energy-efficient solvents for cost-effective CO2 capture. Because the factors that affect the CO₂ absorption efficiency depend on hydrodynamic studies, mass transfer correlations, rotating speed, solvent characteristics, and fluid flow rates, this article reviews all these factors. The correlation of hydrodynamic parameters, mass transfer in the gas-liquid phase in RPBs, and process modeling of large-scale RPBs studied in the literature will help to determine the efficiency of CO₂ absorption in any amine-based solvent using RPB technology. A reasonable interpretation of the whole process can thus be made. The centrifugal surroundings have a significant influence on mass transport by modifying the hydrodynamics of an RPB. As a result, a thorough grasp of this is required to improve an RPB’s mass transfer performance.

Different hydrodynamic characteristics were discussed. The distinct correlations of liquid holdup and pressure drop in RPBs were summarised and showed that the parameters that affect pressure drop are rotational speed and centrifugal acceleration. Furthermore, studies were made to determine the influence of flow rates on pressure drop. The parameters that affect liquid holdup depend on velocity, rotor speed, and the packing structure. Hence, correlations using these parameters were discussed. The understanding of mass transfer is essential as it has a significant effect on the absorption of CO_2. Experimental studies with MEA and other amines/blended amines were explored. As pilot plant and scale-up studies are essential, this review provides a direction for a future CO₂ capture study in a pilot process with an energy-efficient solvent at the CO2 research center of our university.

Author contributions

CS: Analysis and interpretation; methodology; writing–original draft; review and editing. SD: Conceptualization; resources; project administration, supervision writing—review and editing. PM: Supervision; validation; writing review and editing.

Acknowledgments

The authors are thankful for financial support from the Department of Biotechnology (DBT), Ministry of Science and Technology, Govt. of India; through the project “Integrated Design and Demonstration of Intensified CO₂ Capture with cost-effective advanced Process. (INDIA-CO2),” No. T/PR31120/PBD/26/755/2019.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Nomenclature

${\mathit{A}}_{\mathit{c}}$ cross-sectional area (m²)

AF-1 air flow rate (kmol/s)

${\mathit{a}}_{\mathit{c}}$ centrifugal acceleration (m/s²⁾

${\mathit{a}}_{\mathit{e}}$ effective interfacial area (m⁻¹)

${{\mathit{a}}_{\mathit{p}}}^{\prime }$ surface area of the 2 mm diameter bead per unit volume of bead (m⁻¹)

${\mathit{a}}_{\mathit{t}}$ surface area of packing per unit volume of bead (m²/m³)

${\mathit{a}}_{\mathit{s}}$ specific surface area (m²/m³)

$\mathit{a}$ axial distance between discs of rotor (m)

$\mathit{c}$ width of square opening (mm)

${\mathit{c}}_{\mathit{g},\mathit{i}}$ molar concentration of component i in gas phase (kmol/m³)

${\mathit{c}}_{\mathit{l},\mathit{i}}$ molar concentration of component i in liquid phase (kmol/m³)

${\mathit{c}}_{\mathit{v},\mathit{g}}$ specific heat capacity in constant volume (kJ/kmolK)

${\mathit{c}}_{\mathit{p},\mathit{l}}$ specific heat capacity in constant pressure (kJ/kmolK)

${\mathit{c}}_{\mathit{p}\mathit{g}}$ specific heat capacity of gas phase (J/kmolK)

${\mathit{c}}_{\mathit{p}\mathit{l}}$ specific heat capacity of liquid phase (J/kmolK)

${\mathit{c}}_{\mathit{R}\mathbf{3}\mathit{N},\mathit{t}\mathit{o}\mathit{t}\mathit{a}\mathit{l}}$ total liquid bulk concentration of MDEA (kmol/m³)

${\mathit{D}}_{\mathit{G}}$ diffusion coefficient of gas phase (m²/s)

${\mathit{D}}_{\mathit{p}}$ nominal size of packing (m)

${\mathit{D}}_{\mathit{L}}$ diffusion coefficient (m²/s)

${\mathit{D}}_{\mathit{i},\mathit{L}}$ molecular diffusion of all species (m²/s)

$\mathit{d}$ diameter of stainless-steel fiber (mm)

${\mathit{d}}_{\mathit{p}}$ effective diameter of packing (m)

${\mathit{F}}_{\mathit{G}}$ molar flow rate of gas phase (kmol/s)

${\mathit{F}}_{\mathit{L}}$ molar flow rate of liquid phase (kmol/s)

${\mathit{F}}_{\mathit{V}}$ F-factor = $V_G.{\rho }_G^{0.5}$ $\left({kg}^{0.5}/m^{0.5}.s\right)$

${\mathit{f}}_{\mathit{w}}$ wetted resistance coefficient

F-1 liquid flow rate (kmol/s)

$\mathit{G}$ superficial mass velocity of gas (kg/m².hr)

${\mathit{G}}_{{\mathit{N}}_{\mathbf{2}}}$ flow rate of N₂ (m³/s)

$\mathit{g}$ gravitational acceleration (m/s²)

${\mathit{g}}_{\mathbf{0}}$ characteristic gravitational acceleration (m/s²)

$\mathit{H}$ Henry’s constant for absorption of CO₂ in MDEA (kPam³/kmol)

$∆{\mathit{H}}_{\mathit{r}\mathit{x}\mathit{n}}$ heat of reaction (kJ/kmol)

$∆{\mathit{H}}_{\mathit{v}\mathit{a}\mathit{p},\mathit{i}}$ heat of vaporization of i (kJ/kmol)

$\mathit{h}$ axial height of the packing (m)

${\mathit{h}}_{\mathit{g}\mathit{l}}$ heat transfer coefficient (kW/m³K)

${\mathit{K}}_{\mathbf{2}}$ equilibrium constant

${\mathit{K}}_{\mathbf{3}}$ equilibrium constant

${\mathbf{K}}_{\mathit{G}}$ overall volumetric mass transfer coefficient kmol/m³hr.kPa

K coefficient of contraction or expansion

${\mathbf{k}}_{\mathit{G}}$ gas phase mass transfer coefficient (m/s)

${\mathit{k}}_{\mathit{L}}$ liquid mass transfer coefficient (m/s)

${\mathit{k}}_{{\mathit{r}}_{\mathit{i}}}$ reaction rate constant of i reaction (kmol/m³/s)

${\mathit{N}}_{\mathit{i}}$ molar flux of component i (mol/m² s)

${\mathit{N}}_{{\mathit{C}\mathit{O}}_{\mathbf{2}}}$ absorption rate of ${CO}_2$ per unit volume (mol/m³/s)

$\mathit{P}$ total pressure of the system (kPa)

${∆\mathit{P}}_{\mathbf{0}}$ pressure drop outside the rotor (kPa)

P-1 pressure (kPa)

P-2 gas Pressure (kPa)

${\mathit{Q}}_{\mathit{g}}$ volumetric flow rate of gas (m³/s)

${\mathit{q}}_{\mathit{g}}$ heat transfer flux in gas phase (W/m²)

${\mathit{q}}_{\mathit{l}}$ heat transfer flux in liquid phase (W/m²)

$\mathit{R}$ universal gas constant (m³atm/kg.mol.K)

${\mathit{R}}_{\mathit{s}}$ rotational speed of RPB rpm

$\mathit{r}$ radial coordinate of the packed bed from the center (m)

${\mathit{r}}_{\mathit{i}}$ inner radius (m)

${\mathit{r}}_{\mathbf{0}}$ inner radius (m)

$\mathit{T}$ absolute temperature (K)

T-1 liquid inlet temperature (K)

T-2 gas inlet temperature (K)

T-3 gas cooler temperature (K)

T-4 liquid outlet temperature (K)

${\mathit{T}}_{\mathit{G}}$ gas phase temperature (K)

${\mathit{T}}_{\mathit{L}}$ liquid phase temperature (K)

${\mathit{U}}_{\mathbf{l}}$ liquid superficial velocity (m/s)

${\mathit{U}}_{\mathbf{0}}$ characteristic liquid superficial velocity (m/s)

${\mathit{U}}_{\mathbf{g}}$ gas superficial velocity (m/s)

${\mathit{V}}_{\mathbf{0}}$ volume between the outer radius of the bed and the stationary housing (m³)

${\mathit{V}}_{\mathit{G}}$ average gas superficial velocity (m/s)

${\mathit{V}}_{\mathit{t}}$ total volume of the RPB (m³)

${\mathit{V}}_{\mathit{i}}$ volume inside the inner radius of the bed (m³)

${\mathit{x}}_{\mathit{i}}$ mole fraction of component i in liquid phase (mol/mol)

${\mathit{y}}_{\mathit{i}}$ mole fraction of component i in gas phase (mol/mol)

Dimensionless symbols

${\mathit{F}\mathit{r}}_{\mathit{L}}$ liquid Froude number

$\mathit{G}\mathit{a}$ galileo number

${\mathit{G}\mathit{r}}_{\mathit{G}}$ gas Grashof number

${\mathit{G}\mathit{r}}_{\mathit{L}}$ liquid Grashof number

${\mathit{R}\mathit{e}}_{\mathit{G}}$ gas Reynolds number

${\mathit{R}\mathit{e}}_{\mathit{L}}$ liquid Reynolds number

${\mathit{S}\mathit{c}}_{\mathit{L}}$ liquid Schmidt number

${\mathit{W}\mathit{e}}_{\mathit{L}}$ liquid Weber number

Greek symbols

${\mathit{\mu }}_{\mathit{G}}$ viscosity of gas phase (kg/ms)

${\mathit{\rho }}_{\mathit{G}}$ density of gas phase (kg/m³)

${\mathit{\sigma }}_{\mathit{c}}$ critical surface tension (kg/s²)

${\mathit{\sigma }}_{\mathit{w}}$ surface tension of water (kg/s²)

$\mathit{\sigma }$ surface tension (kg/s²)

${\mathit{\epsilon }}_{\mathit{g}}$ hold up in gas phase (m³/m³)

${\mathit{\epsilon }}_{\mathit{l}}$ hold up in liquid phase (m³/m³)

$\mathit{\nu },{\mathit{\nu }}_{\mathit{g}}$ kinematic viscosity (mm²/s)

${\mathit{\nu }}_{\mathbf{0}}$ characteristic Kinematic viscosity (mm²/s)

$\mathit{\gamma }$ dynamic contact angle Degree

${\mathit{\gamma }}_{\mathbf{0}}$ characteristic Dynamic contact angle Degree

$\mathit{\omega }$ angular speed rad/s

$\mathit{\beta }$ constant