Post-combustion CO2 Capture Process Research Articles

On the development of steady-state and dynamic mass-constrained neural networks using noisy transient data

This paper presents the development of algorithms for mass-constrained neural network models that can exactly satisfy mass conservation laws of chemical process systems, even if the training data violates the same. As opposed to approximately satisfying mass balance constraints of a system by considering additional penalty terms in the objective function, algorithms have been developed to solve an equality-constrained optimization problem, thus ensuring the exact satisfaction of the overall mass conservation laws. For developing dynamic mass-constrained networks, hybrid series and parallel all-nonlinear static-dynamic neural network models are leveraged. The proposed algorithms for solving both the inverse and forward problems are tested by considering both steady-state and dynamic data in presence of varieties of noise characterizations. The proposed structures and algorithms are applied to the development of data-driven models of two nonlinear dynamic chemical processes, namely the Van de Vusse reactor system as well as a solvent-based post-combustion CO2 capture process.

Insights into Aerosol Emission Control in the Postcombustion CO2 Capture Process: From Cluster Formation to Aerosol Growth.

Aerosols produced in the amine carbon capture process can lead to secondary environmental pollution. This study employs molecular dynamics (MD) simulations to investigate cluster formation, amine behavior, and aerosol growth of amines, essential for reducing amine aerosol emissions. Results showed that the cluster evolution process can be divided into cluster formation and growth in terms of molecular content, and the nucleation rate for the present systems was estimated in the order of 1028 cm-3 s-1. CO2 absorption was observed alongside successful nucleation, with CO2 predominantly localizing in the cluster's outer layer postabsorption. Monoethanolamine (MEA) exhibited robust electrostatic interactions with other components via hydrogen bonding, leading to its migration toward regions where CO2 and H2O coexisted within the cluster. While MEA presence markedly spurred cluster formation, its concentration had a marginal effect on the final cluster size. Elevating water content can augment the aerosol growth rate. However, altering the gas saturation is possible only within narrow confines by introducing vapor. Contrarily, gas cooling introduced dual, opposing effects on aerosol growth. These findings, including diffusion coefficients and growth rates, enhance theoretical frameworks for predicting aerosol formation in absorbers, aiding in mitigating environmental impacts of amine-based carbon capture.

Resembling Graphene/Polymer Aerogel Morphology for Advancing the CO2/N2 Selectivity of the Postcombustion CO2 Capture Process.

The separation of CO2 from N2 remains a highly challenging task in postcombustion CO2 capture processes, primarily due to the relatively low CO2 content (3-15%) compared to that of N2 (70%). This challenge is particularly prominent for carbon-based adsorbents that exhibit relatively low selectivity. In this study, we present a successfully implemented strategy to enhance the selectivity of composite aerogels made of reduced graphene oxide (rGO) and functionalized polymer particles. Considering that the CO2/N2 selectivity of the aerogels is affected on the one hand by the surface chemistry (offering more sites for CO2 capture) and fine-tuned microporosity (offering molecular sieve effect), both of these parameters were affected in situ during the synthesis process. The resulting aerogels exhibit improved CO2 adsorption capacity and a significant reduction in N2 adsorption at a temperature of 25 °C and 1 atm, leading to a more than 10-fold increase in selectivity compared to the reference material. This achievement represents the highest selectivity reported thus far for carbon-based adsorbents. Detailed characterization of the aerogel surfaces has revealed an increase in the quantity of surface oxygen functional groups, as well as an augmentation in the fractions of micropores (<2 nm) and small mesopores (<5 nm) as a result of the modified synthesis methodology. Additionally, it was found that the surface morphology of the aerogels has undergone important changes. The reference materials feature a surface rich in curved wrinkles with an approximate diameter of 100 nm, resulting in a selectivity range of 50-100. In contrast, the novel aerogels exhibit a higher degree of oxidation, rendering them stiffer and less elastic, resembling crumpled paper morphology. This transformation, along with the improved functionalization and augmented microporosity in the altered aerogels, has rendered the aerogels almost completely N2-phobic, with selectivity values ranging from 470 to 621. This finding provides experimental evidence for the theoretically predicted relationship between the elasticity of graphene-based adsorbents and their CO2/N2 selectivity performance. It introduces a new perspective on the issue of N2-phobicity. The outstanding performance achieved, including a CO2 adsorption capacity of nearly 2 mmol/g and the highest selectivity of 620, positions these composites as highly promising materials in the field of carbon capture and sequestration (CCS) postcombustion technology.

Membrane CO2 Separation System Improvement for Coal-Fired Power Plant Integration

Even though there are numerous CO2 capture technologies (such as chemical and physical absorption), investigators are still trying to come up with novel methods that can minimize the energy requirements for their integration into thermal power plants, as well as the CAPEX and OPEX expenses. In this work, the technical and financial aspects of integrating two-stage polymeric membranes into a coal-fired power plant with a capacity of 330 MW were examined. The study researched the membrane post-combustion CO2 capture process utilizing CHEMCAD version 8.1 software with several parameters and an expander to decrease the total cost. The simulation showed promising results regarding reducing power consumption after using an expander for both a high capture rate (&gt;90%) and a CO2 concentration of more than 95%. Thus, the CO2 captured cost decreased from 58.4 EUR/t (no expander) to 48.7 EUR/t (with expander).

Effect of Fly Ash on the Mass Transfer Performance of CO2 Removal Using MEA and DEA Solutions in a Packed Tower

The accumulation of uncollected fly ash from flue gas in post-combustion CO2 capture processes is a significant concern in current coal-fired power plants due to its potential impact on the performance of CO2 absorbent and absorption towers. In order to determine the effect of fly ash on the mass transfer performance of CO2 absorption into monoethanolamine (MEA) and diethanolamine (DEA) aqueous solutions, experimental studies were carried out using a small-sized packed tower equipped with θ-ring random packing. These studies were conducted under various operating parameters, including solution temperature, liquid/gas ratio (L/G), packing height, and fly ash concentration. The results show that the effect of fly ash on the outlet CO2 concentration was primarily observed during the initial stages of the experimental process. Moreover, the presence of fly ash leads to a reduction in the volumetric overall mass transfer coefficient (KGav) when using MEA and DEA solution, and increasing the fly ash concentration further exacerbates this negative impact. However, the effect of fly ash on the reduction in KGav is not significantly related to its chemical composition but rather depends on the operational parameters. With increasing solution temperature, liquid/gas ratio (L/G), and packing height, the KGav values for different solutions exhibit an upward trend. The negative effect of fly ash on KGav remains relatively stable for MEA as solution temperature and packing height increase. Compared to MEA, fly ash has a greater negative effect on DEA solution under the same experimental conditions. The analysis reveals that the detrimental effect of fly ash on KGav primarily stems from its ability to alter the distribution state of the absorption liquid within the packed tower.

Modeling and assessing the integration of [formula omitted] capture in waste-to-energy plants delivering district heating

The waste-to-energy (WtE) sector is assuming an increasingly important role in net-zero pathways owing to its hard-to-abate emissions and its inherent potential for generating negative emissions. Thus, exploring effective strategies for CO2 capture and storage (CCS) integration with WtE plants that address the challenge posed by their coupling to district heating (DH) networks and the need to deliver both heat and electricity to customers is essential. This study presents a mass and energy balance model of a WtE plant with post-combustion CO2 capture process to assess the integrated system’s energetic feasibility. Results show that despite the extensive DH demand in heat-based WtE plants, enough excess heat is available to capture around 60% of the CO2 generated. As such, negative emissions can be achieved without external heat requirement at an efficiency reduction of 16%. By incorporating heat pumps and utilizing CCS waste heat for district heating, CO2 capture can be maximized at a total efficiency reduction as low as 36%. This is comparable to CCS in power-based WtE plants with no district heating. These findings highlight potential efficiency gains through the synergistic integration of CO2 capture and district heating, providing additional opportunities for WtE plants to contribute to the net-zero transition.

A comparative study of polyamine and piperazine as promoter for CO2 absorption performance in aqueous methyldiethanolamine blend system: 430 MW power plant data simulation and economic assessment

Fossil fuel power plants play a significant role in fulfilling global energy demands but are also responsible for high CO2 emissions. Amine-based post-combustion CO2 capture (PCC) is one of the technologies used in such power plants to reduce CO2 emissions. However, high energy consumption in amine-based PCC technology causes a reduction in the overall energy efficiency of power plants. The energy consumption significantly depends on the solvents used in the PCC process. Therefore, in this work, we performed CO2 absorption in the piperazine (PZ), triethylenetetramine (TETA), and N-methyldiethanolamine (MDEA) blend and tri-blend solvents. For the absorption performance and energy consumption calculation, a simulation is performed for the 430 MW power plant data in the DWSIM software. Initially, solvent screening was done, and four solvents are selected based on the optimum energy consumption of various solvents. Further, the solvents’ performance in terms of energy consumption is done based on the effect of different temperatures, CO2 composition in feed, and %CO2 recovery. Overall, the TETA+PZ MDEA tri-blend solvent showed optimum energy consumption for most of the conditions. Additionally, a techno-economy assessment was done considering annual capital cost and annual operating cost, and results revealed that the cost per ton of CO2 capture in 5 %TETA+ 5 %PZ+ 20 %MDEA solvent is 27.33 % lesser than industrial MEA solvent, and 1.56 % lower than 10 %PZ+ 20 %MDEA.

Cryogenic CO2 capture from oxy-combustion flue gas by a hybrid distillation + physical absorption process

Hybrid flowsheets are defined, in the context of process intensification, as alternatives suitable for replacing energy-intensive separation methods through the combination of more than one unit operation. Given the undoubted importance of carbon capture, utilization, and storage (CCUS) in the next decades in the path to net-zero emissions and even negative emissions scenarios, this work focuses on CO2 capture from an oxy-combustion flue gas by means of a distillation-based hybrid process that avoids CO2 freeze-out, which is capable of delivering a liquid CO2-rich stream that can in principle be pumped to transport pressures with lower energy consumption in comparison to conventional post-combustion CO2 capture processes based on chemical absorption. A distillation-based process and a hybrid distillation + physical absorption process have been studied based on simulations in Aspen Hysys® V11 using the Peng-Robinson Equation of State and the “Acid Gas – Physical Solvents” thermodynamic property package. Firstly, a CO2 capture process based on cryogenic distillation has been proposed and a sensitivity analysis on the operating pressure and specified CO2 recovery has been performed in order to determine the most convenient configuration, balancing capital and operating costs. Then, a hybrid capture process has been proposed and analyzed, where cryogenic distillation is the bulk removal step and physical absorption into a mixture of homologues of the dimethylether of polyethylene glycol (DEPG) is the finishing step. Energy and economic analyses have shown that the hybrid process, delivering a CO2 product respecting the stringent purity requirement on the oxygen content for application for Enhanced Oil Recovery (EOR) and in the food industry is less energy intensive than the standalone cryogenic distillation, though it is characterized by higher capital costs.

Flexible operation of Post-combustion CO2 capture process enabled by NARX-MPC using neural network

Post-combustion CO2 capture (PCC) process is expected to serve as the role of a bridge solution during the transition from fossil-based to renewable energy, but extensive energy consumption is a significant hurdle to its widespread commercial adoption. Operating a PCC process economically often requires large operating condition changes in response to power plant load changes and time-varying energy prices. The ‘flexible’ operation requires a controller that aids in such a dynamic operating environment of the PCC process. To balance model complexity with prediction accuracy, we adopt NARX-MPC which employs a neural network-based NARX model as a prediction model for MPC. Under various dynamic operation scenarios, the proposed NARX-MPC is evaluated and compared with a linear MPC (LMPC) controller. Our observations indeed reveal that the NARX-MPC delivers significantly superior closed-loop control performance over the LMPC, especially in scenarios involving large changes in the power plant load or inlet flue gas flow rate and simultaneous changes in the flue gas flow rate and the CO2 capture rate.

Hybrid Smart Strategies to Predict Amine Thermal Degradation in Industrial CO2 Capture Processes.

CO2 emission reduction is an essential step to achieve the climate change targets. Solvent-based post-combustion CO2 capture (PCC) processes are efficient to be retrofitted to the existing industrial operations/installations. Solvent degradation (and/or loss) is one of the main concerns in the PCC processes. In this study, the thermal degradation of monoethanolamine (MEA) is investigated through the utilization of hybrid connectionist strategies, including an artificial neural network-particle swarm optimization (ANN-PSO), a coupled simulated annealing-least squares support vector machine (CSA-LSSVM), and an adaptive neuro-fuzzy inference system (ANFIS). Moreover, gene expression programming (GEP) is employed to generate a correlation that relates the solvent concentration to the operating variables involved in the adverse phenomenon of solvent thermal degradation. The input variables are the MEA initial concentration, CO2 loading, temperature, and time, and the output variable is the remaining/final MEA concentration after the degradation phenomenon. According to the training and testing phases, the most accurate model is ANFIS, and the reliability/performance of its optimal network is assessed by the coefficient of determination (R2), mean squared error, and average absolute relative error percentage, which are 0.992, 0.066, and 2.745, respectively. This study reveals that the solvent initial concentration has the most significant impact, and temperature plays the second most influential effect on solvent degradation. The developed models can be used to predict the thermal degradation of any solvent in a solvent-based PCC process regardless of the complicated reactions involved in the degradation phenomenon. The models introduced in this study can be employed for the development of more accurate hybrid models to optimize the proposed systems in terms of cost, energy, and environmental prospects.

Kinetics of carbon dioxide absorption in aqueous blend of 1-(2-aminoethyl) piperazine using a stirred cell reactor.

Carbon capture utilization and storage (CCUS), the technology for decarbonizing carbon dioxide (CO2) from greenhouse gas emitters such as steel, cement, oil, gas, petrochemicals, chemicals, and fertilizers, has a critical role to play in the world to achieve industrial net zero targets by 2050. CO2 can be separated from industrial exhaust gases/flue gases using amine-based solvents utilizing the post-combustion CO2 Capture process. One most crucial solvent characterization for this application is the kinetics of CO2 absorption. This work identifies aqueous 1-(2-aminoethyl) piperazine (AEPZ) as a potential candidate for CO2 capture solvent. The kinetics of absorption of CO2 in aqueous AEPZ is studied using stirred cell reactor. The experiments are performed at temperatures ranging from 303 to 333K with weight fractions of AEPZ in an aqueous solution ranging from 0.1 to 0.4. One of the critical parameters of the kinetic study is Henry's constant which is determined experimentally using another stirred cell reactor at a similar temperature and pressure range. The experimental data shows that the overall rate constant is Kov = 2.52987 × 10-4mol/m2s-kPa for 0.1 wt fr. of AEPZ at 313K with an initial CO2 partial pressure of 10kPa. The temperature dependency relation of the second-order reaction rate constant, [Formula: see text] is found to be [Formula: see text] using the Arrhenius equation. The activation energy of 0.3 wt fr. AEPZ is found to be 42.42kJ/mol. In addition, the density and viscosity of the aqueous solvent are determined at a wide range of temperatures. The diffusivity of CO2 and physical solubility used in the model development has also been determined. The kinetic parameters obtained from this study are helpful in the process design of CO2 capture in a regenerative process with a blended solvent with AEPZ.

Development of novel waste tea-derived activated carbon promoted with SiO2 nanoparticles as highly robust and easily fluidizable sorbent for low-temperature CO2 capture

Waste-derived biomass-based sorbents being employed in many applications, e.g., post-combustion CO2 capture process, are in rising interest. This work investigates the CO2 capture technique by activated carbon (AC) derived from waste tea (WT) biomass. The synthesis protocol consists of carbonization followed by activation with KOH and NaOH, known as chemical activators, utilizing the wet impregnation method. With the aim of developing the AC sorbent fluidizable in fluidized-bed systems, an essential factor in the industrialization of the CO2 capture process, these sorbents are mixed with the easily fluidizable SiO2 nanoparticles (NPs). Prepared samples are characterized via SEM, EDS, BET, FTIR, and TG analysis. Based on the TG analysis assessed in three consecutive adsorption/desorption cycles, 457 %, 141 %, and 69 % higher average CO2 capture capacity is recorded for KOH-promoted AC (2.316 mmol/g) compared to raw waste tea (0.416 mmol/g), NaOH-promoted AC (0.961 mmol/g) and commercial AC (1.370 mmol/g), respectively. Fluidization experiments show that the WT-derived AC mixed with 2.5 wt% SiO2 NPs presents a 45 % higher average bed expansion ratio linked with a homogeneous and bubbleless fluidization regime than its silica-free counterpart, which presents the bubble- and channel-containing bed with a non-uniform regime. This modified sorbent shows a 9% more CO2 capture capacity (2.53 mmol/g) than the WT-derived AC and superior multicyclic stability during 25 multiple adsorption/desorption cycles.

Intensification of Sono-Assisted CO2 Stripping/Carbon-Rich Solvent Regeneration by Fe2O3 Hydrophobic Micronized Particles

In 2020, the amount of CO2 generated by coal combustion was estimated to be 15.32 Gt CO2. In a solvent-based postcombustion CO2 capture (PCCC) process, the solvent regeneration/CO2 stripping process uses thermal energy to regenerate the solvent. It has an adverse impact on the solvent’s thermal stability, besides causing a significant quantity of solvent loss and a high energy demand. In recent times, one method for PCCC cost savings has emerged: sono-assisted solvent regeneration/CO2 stripping. The present work focuses on enhancing the high-frequency ultrasound-assisted aqueous carbon-rich 30 wt % monoethanolamine (MEA) solvent regeneration/CO2 stripping process using Fe2O3 hydrophobic micronized particles (concentrations varied in the ranges of 0.005, 0.01, 0.05, 0.1, and 0.2 wt %) in a controlled-temperature environment at 12 °C using 360 kHz, 470 kHz, and 1 MHz (streaming-dominant frequencies). From the investigation, the lowest concentration (0.005 wt %) of micronized particles shows maximum enhancements of 12.29, 20.93, and 33.62%, thereby decreasing the solvent sensible energy requirements by 1.26, 2.4, and 2.9 times than without micronized particles for the tested frequencies. In addition, the observed CO2 stripping rate is much higher when compared to the case without micronized particles. The stripping efficiency is observed to be high in the initial stages of sonication (5 min).

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

Carbon dioxide (CO2), a significant greenhouse gas released from power plants and industries, substantially impacts climate change; minimizing it and achieving carbon net zero is essential globally. In the direction of reducing CO2 emissions into the atmosphere, post-combustion carbon capture from large point CO2 emitters by chemical absorption involving the absorption of this gas in a capturing fluid is a commonly used and efficacious mechanism. Researchers have worked on the process using conventional columns. However, process intensification technology is required because of the high capital cost, the absorption column height, and the traditional columns’ low energy efficiency. Rotating packed bed (RPB) process intensification equipment has been identified as a suitable technology for enhanced carbon capture using an absorbing fluid. This article reviews and discusses recent model developments in the post-combustion CO2 capture process intensification using rotating packed beds. In the literature, various researchers have developed steady-state mathematical models regarding mass balance and energy balance equations in gas and liquid phases using ordinary or partial differential equations. Due to the circular shape, the equations are considered in a radial direction and have been solved using a numerical approach and simulated using different software platforms, viz. MATLAB, FORTRAN, and gPROMS. A comparison of various correlations has been presented. The models predict the mole fraction of absorbed CO2 and correspond well with the experimental results. Along with these models, an experimental data review on rotating packed bed is also included in this work.

Process synthesis for amine-based CO2 capture from combined cycle gas turbine power plant

The study provides technoeconomic evaluations of advanced MEA-based Post-Combustion CO2 Capture (PCCC) process configurations applied to a 750 MW Combined Cycle Gas Turbine (CCGT) power plant. Rigorous rate-based model of the PCCC process developed in Aspen Plus was validated and then used to study synergistic effects of combining three process configurations in one flowsheet using energy and levelized cost of capture as key performance indicators (KPIs). The results of the energy and economic analysis elucidate that Absorber Inter Cooling (AIC) + Rich Solvent Split (RSS) + Lean Vapor Compression (LVC) is the optimal combination of MEA-based PCCC process as it provided minimum regeneration energy (2.80 GJ/tCO2) and levelized capture cost (72.7 $/tCO2) representing an overall energy and cost savings of 6.25% and 8.95%, respectively. The study demonstrates that the usage of single KPI such as energy savings can provide misleading results. For example, combination of AIC + RSS + Inter Heated Stripper (IHS) provided maximum equivalent energy savings (3.50%), however, it did not result in highest cost savings due to high capital and operating costs of additional heater and pump. Overall, the study underlines the potential of advanced mature technologies on energy and cost reductions.

Activated Carbon Produced from the Hydrothermal Treatment of Glucose with KOH Activation for Catalytic Absorption of CO2 in a BEA-AMP Bi-Solvent Blend.

The amine-based postcombustion CO2 capture (PCC) process involves absorption of CO2 into a solvent and then regenerating the solvent to produce CO2. In this study, the effect of an activated carbon (AC) catalyst, synthesized through hydrothermal treatment and furnace activation on CO2 absorption in a 4M BEA/AMP amine blend, was evaluated and compared with that of a KMgO/CNTs (1:4) catalyst. The physical and chemical properties of AC were investigated with a scanning electron microscope (SEM), CO2 temperature-programmed desorption (CO2-TPD), Brunauer-Emmett-Teller (BET), powder X-ray diffraction (XRD), and thermogravimetric analyzer (TGA) and compared with the KMgO/CNTs (1:4) catalyst. The results showed that when compared against noncatalytic CO2 absorption, AC enhanced the linear rate of CO2 absorption by 33.3%, while for KMgO/CNTs, it was reported as 25.5%. The relatively higher surface area, combined with the higher number and strength of basic sites of AC relative to the KMgO/CNTs (1:4) catalyst, provided effective basic reaction sites for CO2 absorption, thereby enhancing the rate of CO2 absorption into the amine. AC was also relatively easier to synthesize which would provide a good replacement for the CNT-based catalyst which has carcinogenic tendencies.

Attrition characteristics of limestone in gas-solid fluidized beds

Calcium looping is a potential energy-efficient post-combustion CO2 capture process that is competitive with existing commercially available technology like oxy-combustion. The attrition of CaO-based sorbents during alternating calcination/carbonation cycling in calcium looping (CaL) has the potential to escalate operation cost and contribute to environmental emisssions. Here, we measure jet mill attrition rates of four commercial limestones from Ghana and Canada and compare them with fluid catalytic cracking catalyst (FCC) and commercial vanadyl pyrophosphate (VPP) in a jet mill. We assessed the effect of orifice velocity, attrition time and calcination on particle attrition resistance. The attrition rates of the limestones increased exponentially from as low as 2.4 mg h−1, which is an acceptable rate for industrial applications, up to 38 mg h−1, which is prohibitively high as the orifice velocity increased from 180 m s−1 to 230 m s−1. The attrition rates of the FCC and VPP catalyst were somewhat higher at the low orifice velocity (5 5 mg h−1 and 8 mg h−1, respectively) but lower at the high orifice velocity (21 mg h−1−1 and 26 mg h−1, respectively). Within 3 h, Bui and Nauli limestones reached a steady attrition rate of 19 mg h−1 and 21 mg h−1, respectively whereas Oterpklu limestone reached a steady rate of 17 mg h−1 after 4 h. The calcined limestones attrition rate was twice as high as their carbonated counterpart. The limestones attrited predominantly throgh fragmentation.

Thin-film composite mixed-matrix membrane with irregular micron-sized UTSA-16 for outstanding gas separation performance

Low-cost, micron-sized particles still pose a barrier to their use in thin-film composite mixed-matrix membranes (TFC-MMMs) owing to their poor interfacial contact with the polymer matrix. Also, the particles are too large to be fabricated into the submicron-thick membranes. Herein, we report high-performing, TFC-MMMs based on a CO2-philic comb copolymer, poly (tetrahydrofurfuryl methacrylate)–co–poly (poly (oxyethylene methacrylate)) (PTO), and an irregular, micron-sized, CO2-selective metal-organic framework (MOF), UTSA-16. The PTO comb copolymer matrix exhibited excellent film-forming ability, adhesion properties and showed a good gas separating performance. The PTO comb copolymer also enhanced the dispersibility of UTSA-16 in an environment-friendly solvent mixture (i.e., ethanol/water), which did not adversely damage the underlying porous polymeric support. Despite the micron-scale particle size of UTSA-16, PTO copolymer completely covered the surface of UTSA-16 via strong interactions without any deep pore infiltration and exhibited excellent interfacial contact properties. Consequently, defect-free TFC-MMMs with a polymer thickness of 300 nm were successfully prepared on the porous support. The TFC-MMM with 10% filler loading exhibited excellent CO2 permeance and selectivity, i.e., CO2 permeance of 1070 GPU, CO2/N2 selectivity of 41.0, CO2/CH4 selectivity of 17.2, outperforming the TFC-MMMs prepared with commercially available Pebax. All PTO-based MMMs, with the exception of the low content of UTSA-16 (5%), exceeded the gas separation performance required for post-combustion CO2 capture process.

Agent-Based and Stochastic Optimization Incorporated with Machine Learning for Simulation of Postcombustion CO2 Capture Process

In this paper, a novel method is proposed for the incorporation of data-driven machine learning techniques into process optimization. Such integration improves the computational time required for calculations during optimization and benefits the online application of advanced control algorithms. The proposed method is illustrated via the chemical absorption-based postcombustion CO2 capture process, which plays an important role in the reduction of CO2 emissions to address climate challenges. These processes simulated in a software environment are typically based on first-principle models and calculate physical properties from basic physical quantities such as mass and temperature. Employing first-principle models usually requires a long computation time, making process optimization and control challenging. To overcome this challenge, in this study, machine learning algorithms are used to simulate the postcombustion CO2 capture process. The extreme gradient boosting (XGBoost) and support vector regression (SVR) algorithms are employed to build models for prediction of carbon capture rate (CR) and specific reboiler duty (SRD). The R2 (a statistical measure that represents the fitness) of these models is, on average, greater than 90% for all the cases. XGBoost and SVR take 0.022 and 0.317 s, respectively, to predict CR and SRD of 1318 cases, whereas the first-principal process simulation model needs 3.15 s to calculate one case. The models built by XGBoost are employed in the optimization methods, such as an agent-based approach represented by the particle swarm optimization and stochastic technique indicated by the simulated annealing, to find specific optimal operating conditions. The most economical case, in which the CR is 72.2% and SRD is 4.3 MJ/kg, is obtained during optimization. The results show that computations with the data-driven models incorporated in the optimization technique are faster than first-principle modeling approaches. Thus, the application of machine learning techniques in the optimization of carbon capture technologies is demonstrated successfully.

Techno-economic analysis and optimization of a CO2 absorption process with a solvent looping system at the absorber using an MDEA/PZ blended solvent for steam methane reforming

Advanced CO2 capture processes for steam methane reforming that supplies large quantities of H2 to industries are critical for cost-effective blue H2 and CO2 production. In this study, a pre-combustion CO2 absorption process with a novel solvent looping system was developed to capture CO2 from SMR gas at a flow rate of 6946.5 kmol/h and 21 bar. A techno-economic analysis of the advanced absorption process using MDEA/PZ blended solvent was performed with respect to five key operating parameters. For 90 % and 99 % CO2 capture rates, the minimal reboiler duties were 0.622 and 0.751 GJ/CO2 ton, which were approximately 70 % lower than those of conventional post-combustion CO2 capture processes. Despite a significant increase in the CO2 capture rate (90 % to 99 %), the increase in energy consumption and OPEX was only 4 % while using the novel blended-solvent looping system through a low-pressure flash drum, resulting in a lower CO2 capture cost per ton at 99 % CO2 capture rate. The economic analysis suggested that the minimal CO2 capture cost decreased from 44.89 to 42.56 USD/CO2 ton as the capture rate increased. Furthermore, DNN models could predict the CAPEX and OPEX under higher than 98 % accuracy. DNN-based optimization with a multi-variable mutation of features revealed the optimum operating conditions for a minimum CO2 capture cost (42.44 USD/CO2 ton) over a short computational time (12 s). Thus, this study provides insights on the blended solvent and guidelines for a high-pressure CO2 absorption process with a solvent looping system for cost-effective blue H2 production.