English

Unsteady-state modeling of a naphtha pyrolysis furnace considering coke formation and carbon emissions

To achieve carbon neutrality, it is crucial to reduce emissions from high-carbon-emitting processes. This study developed a fundamental model as the theoretical basis for an efficient carbon emissions reduction strategy, focusing on the naphtha pyrolysis process, which emits significant carbon dioxide owing to LNG combustion for heat supply. The model was constructed with a coupled firebox-tubular reactor structure, dividing the furnace into distinct phases (furnace wall, flue gas, tube wall, and process gas) to closely consider convective, conductive, and radiative heat transfer. In naphtha pyrolysis, key state variables, such as the olefin yield affected by coke layer formation in the tubular reactor and the phase-wise temperature distribution within the furnace, change dynamically over time. To analyze this behavior, unsteady-state governing equations incorporating the coke formation mechanism were rigorously formulated, and a precise numerical simulation procedure was proposed.

The developed model was calibrated through model fitting using actual operating data, and the results confirmed that the fitted model accurately predicts dynamic changes in key state variables such as ethylene and propylene yields and coke layer thickness within the reactor over time. Case studies quantitatively analyzed the effects of key operating conditions such as the combustion gas flow rate on coke formation, carbon emissions, and product yields. The developed model provides a core theoretical foundation for the integrated consideration of carbon emission reductions and process efficiency improvements. It is expected to serve as an important basis for sustainable process design and derivation of optimal operation strategies.

Modeling of a multistage fluidized bed reactor for hydrogen-based iron duction

With global warming and increasing CO2 emissions becoming critical issues, reducing carbon emissions from the steel industry has become an urgent priority. The steel industry is responsible for approximately 8% of global CO2 emissions, primarily generated from the blast furnace-basic oxygen furnace (BF-BOF) process. Hydrogen-based ironmaking processes have been proposed to address this challenge. In particular, the adoption of fluidized bed reactors (FBRs) offers high material and heat transfer efficiencies as well as the flexibility to utilize various iron ore raw materials. In this study, we developed a fundamental model of a multistage FBR for hydrogen-based iron ore reduction. Specifically, a two-phase model was employed in a single FBR to precisely describe the gas–solid interactions within the fluidized bed, and an unreacted shrinking core kinetic model was constructed to account for changes in the component (i.e., hematite, magnetite, wüstite, and iron) composition inside the iron ore particles. Subsequently, the single FBR model was extended to a multistage FBR to predict the iron ore reduction degree and the reaction gas composition in a counter-current flow system. Thereafter, the model parameters were optimized through model fitting using experimental data, with the predicted values confirmed to accurately reflect the actual process characteristics.

The developed model provides a theoretical framework for a detailed analysis of hydrogen-based iron ore reduction and mass transfer characteristics in a multistage FBR, thus providing a foundation for deriving optimal reactor designs and operating conditions for future applications. This study is expected to accelerate the commercialization of hydrogen-based ironmaking processes and contribute to efficient reducing carbon emissions in the steel industry.

Modeling of a microbial electrosynthesis system using porous electrodes for CO2-to-acetate conversion

Microbial electrosynthesis (MES) is an emerging carbon capture and utilization (CCU) technology that converts CO2 into value-added chemicals using microbial catalysts powered by electrical energy. This study presents a fundamental model for an MES system designed to produce acetate from CO2 incorporating real-world experimental conditions. Unlike existing models that focus on biofilm growth on nonporous metallic electrodes, the model emphasizes mass transfer, bioelectrochemical reactions, and biomass accumulation within porous graphite felt electrodes, which are widely used for their microorganism affinity and cost-effectiveness. Specifically, to account for biofilm growth within the porous electrode, the internal volume of the graphite felt electrode is divided into two distinct regions for which a mass balance equation is established: 1) a region devoid of microorganisms, and 2) a region where a biofilm is grown using carbon fibers as a support. A kinetic model based on the commonly used Nernst-Monod equation is introduced to consider the direct electron transfer between microorganisms and electrodes. In particular, the concept of accumulating the growth rate by position is introduced to account for volume changes in the continuously growing biofilm.

To ensure that the developed MES model accurately describes and predicts the characteristics of the actual system, model fitting was performed using experimental data. For MES systems, the model fitting process is essential for considering the complex interactions between the biological reactions of microorganisms and electrochemical reactions, while also reflecting the production process of the target product. Using parameters reasonably estimated through model fitting, an analysis of the simulation results confirmed that the fitted model accurately represents the dynamic behavior of the target MES system. The developed MES model can function as a virtual plant, enabling the analysis of optimal operating conditions to maximize acetate yield and improve energy efficiency. Furthermore, it is expected to significantly contribute to the commercialization of MES technology.

Simulation and optimization of a trickle bed reactor for hydrotreating of non-edible vegetable oil

Biofuels produced from biomass are gaining attention as a new energy source to replace fossil fuels. Particularly, non-edible crops like Jatropha are emerging as new sustainable feedstocks for biofuel production, distinct from edible vegetable oils, as they are independent of food production. However, biodiesel such as fatty acid methyl ester produced through the transesterification reaction of triglycerides has low oxidative stability and high CFPP(cold filter plugging point), limiting its blending ratio with petroleum-based diesel. On the other hand, HBD(hydrotreated biodiesel) produced through the hydrotreating of vegetable oils has high oxidative stability and a low CFPP, allowing for a higher blending ratio with petroleum-based diesel.
In this study, a computational model of the hydrotreating reactor was developed using Jatropha oil as a feedstock for HBD production. Kinetics validation was conducted using experimental data from a pilot reactor. This confirms that the kinetics model is applicable to various process conditions and reactor configurations using extrudate catalysts and accurately reflects the actual system characteristics. Subsequently, the model was expanded to a commercial-scale reactor to simulate a TBR(trickle bed reactor) with a quench zone added between the beds to prevent hot spot formation and maintain appropriate temperature conditions for HBD production.

Using the developed hydrotreating reactor model including the quench zone, optimization problems were designed and solved using variables including the inlet temperature, the position of the quench zone, and the degree of temperature drop in the quench zone to derive the optimal operating conditions in the HBD production process. In Case 1, operating condition was derived to maximize the yield of HBD. However, in this scenario, the amount of utility and hydrogen consumption in the process were not considered. Therefore, in Case 2, economic efficiency was optimized by considering the power, cooling water, and hydrogen consumption in the entire HBD production process including the hydrotreating reactor and other unit processes. The objective value of the economic optimization results was 14.6% higher compared to the base case and 4.22% higher compared to Case 1, indicating that more economically efficient operating conditions were achieved.

Multi-objective optimization of a CO2 hydrogenation process considering yield and energy efficiency

Due to the continuous increase in CO2 emissions and the resulting environmental problems, researches are conducted to create high-value substances using CO2 as a raw material. Among them, the light olefin production process via CO2 hydrogenation has been studied for its advantages of low environmental burdens and reduced dependence on petrochemicals. However, despite these efforts, the process still exhibits high energy requirements and low yields. To enhance process efficiency and achieve scale-up and commercialization, research on thermal design and optimization using modern thermodynamic optimization theory is needed. Therefore, in this study, we conducted a study on optimization methods that can simultaneously improve energy efficiency and yields.
The process was represented by a mathematical model of a packed-bed reactor where high-temperature Fischer-Tropsch reactions occur. Based on this model, we performed entropy generation rate minimization and product yield maximization, respectively, and confirmed that the two objective functions are in a trade-off relationship concerning the optimization variables (inlet temperature and pressure).
To achieve the simultaneous optimization of the two objective functions, we designed and solved the problem of maximizing the annual operating profit (AOP), which includes the economic loss due to the entropy generation rate and the economic gain due to the production of light olefin. However, the optimal operating conditions of the AOP changed according to the market prices of exergy and light olefin, which means that the fluctuation of market prices should be reflected in solving the optimization problem.

To reflect the characteristics of AOP, we used the weighted-sum method, a type of multi-objective optimization (MOO) methods, where weights are assigned to each objective function and linearly combined to find the optimal solution. The weighted-sum method progressively changes the assigned weights to perform optimization and find the Pareto front, the optimal set for the MOO problem. In this study, an objective function based on AOP was constructed by including the impact of the market prices of light olefin and exergy on the weights. In this case, each weight serves as an economic indicator according to market price changes. Based on this objective function, optimization was carried out by changing the weights to derive the optimal operating conditions at each point of the Pareto front. The Pareto front obtained by this method represents a set of optimal points that reflect changes in market prices.

Modeling and optimization of a naphtha pyrolysis process considering carbon emissions

In response to the recent climate crisis, research on greenhouse gas reduction is actively underway. One of the effective ways for carbon neutrality is to reduce carbon dioxide emissions from existing high-carbon-emitting processes. This study performs modeling study for naphtha pyrolysis processes at Naphtha Cracking Center that is a representative high-carbon-emitting process due to LNG combustion gas for heat supply. Specifically, since the interaction between a firebox and tubular reactors appears complex depending on the location inside the cracking furnace, this study proposes a coupled firebox and tubular reactor model to simultaneously predict product yield and carbon emissions according to operating conditions. In addition, model fitting is performed based on the operation data of an actual naphtha pyrolysis furnace to ensure that the developed model properly considers the characteristics of the actual system.

Utilizing the developed coupled firebox and tubular reactor model, optimization problems are designed and solved to compare the operating strategies under various scenarios. Existing optimization studies of the naphtha pyrolysis process have mainly focused on the maximization of profit from the products or minimization of operating cost without considering greenhouse gas emissions that has recently become a major environmental issue. Therefore, this study derives optimal operating conditions to maximize profit from ethylene and propylene while minimizing the carbon tax due to CO₂ emissions by considering the supplied naphtha and steam. Additionally, additional optimization is performed by applying various material prices to comprehensively investigate the impact of the carbon tax due to carbon emissions on the optimal operating conditions. The derived optimal operating conditions can increase the yield of the product and effectively reduce carbon emissions compared to the existing conditions. The need for the proposed economic optimization framework is expected to increase as the global carbon tax continues to rise.

Computational fluid dynamics based modeling and model predictive control of fluidized-bed dry reforming of methane process

In a fluidized bed DRM (dry reforming of methane) process, solid catalyst particles move at a low velocity in the bubbling fluidization zone at the bottom and descend with a high fluidization velocity at the top, with a non-uniform velocity profile depending on their location inside the reactor. Therefore, for precise fluidized bed DRM process control, a fundamental model that considers DRM reaction, mass and heat transfer as well as bubble and solid catalyst flow is required. STEP Lab is performing fundamental modeling of fluidized bed DRM process based on computational fluid dynamics (CFD)-based compartment modeling methodology. Specifically, the inside of the fluidized bed reactor is divided into hundreds of small zones that can be assumed as a single homogeneous reactor. Subsequently, mass and energy conservation equations for each zone and the entire reactor are formulated based on the DRM mechanism while CFD simulations are performed to calculate the mass flow between adjacent zones and account for the resulting heat and mass transfer.

Despite the obvious advantages from an environmental point of view, most of the existing studies for DRM processes are still on technologies such as catalyst and plasma reactor development, and there are few studies on process efficiency improvement through optimization and optimal control. STEP Lab is conducting research to improve process efficiency by establishing the model predictive control (MPC) system, which is one of the most advanced control methods applied to various chemical processes, in connection with fundamental modeling. Specifically, a MPC system is designed and verified by adopting the developed high-fidelity fundamental model for the fluidized-bed DRM reactor as a virtual plant. Then, the developed MPC system will be applied to an actual fluidized-bed DRM reactor through collaboration with companies or research institutes.

Bio-electrochemical system modeling and offset-free model predictive control

Recent research on bio-electrochemical system (BES) is in the stage of focusing on the possibility of the production of target molecules such as hydrogen, methane, and acetate from carbon dioxide at the cathode through the application of external electricity. In order to gradually move from the stage of development for basic technologies to the commercialization stage, research to improve the performance of the BES process by maximizing the carbon dioxide removal efficiency and productivity of the target molecules while reducing the overall energy consumption is essential. Therefore, STEP Lab is conducting research to optimize the operating conditions of BES and implement the optimal control system on the BES process. Specifically, we are trying to utilize the offset-free model predictive control technique that can effectively compensate for the model-plant mismatch in order to consider the error between the measured value from the practical process and the predicted value from the digital twin built via fundamental modeling.