Structure-oriented Lumping Research Articles

Advancing bioethanol production with a new kinetic model based on structure-oriented lumping for cellulose conversion

Bioethanol, as a significant renewable energy source, plays a crucial role in alleviating the energy crisis and reducing greenhouse gas emissions. This research constructed a kinetic model for the hydrolysis of cellulose into bioethanol using the Structure-Oriented Lumping (SOL) method. Three vector units were defined to represent the structure of cellulose and its products, and five reaction rules were used to generate an 18-step reaction network. Reaction rate constants for these reactions were determined for various catalysts using genetic algorithms, and catalyst performance was analyzed based on these rate constants. Results demonstrated a high degree of agreement between experimental data and model predictions over time for the yield of cellulose and its products across four model applications using 16 different catalysts, effectively predicting yield variations with the maximum mean absolute error (MAE) not exceeding 6.3 % and the minimum being as low as 0.8 %. The predictive accuracy of the model provides a robust basis for analyzing catalyst performance through four distinct model applications: 1. In the catalyst hydrolysis of cellulose to ethanol by 5Ru-25 W/HZSM-5, the model identified the hydrogenolysis of EG (k13) as the rate-determining step (RDS). 2. The model explored the impact of Cu-WOx/SiO2 catalysts loaded with various metals (X = None, Pd, Au, Ru), demonstrating how different metal loadings influence the RDSs for the production of EG and ethanol from cellulose. Notably, catalysts loaded with Ru significantly enhanced the RDS (km) for EG, thereby increasing its yield. 3. The model highlighted subtle influences on the RDS for cellulose hydrolysis to EG under catalysts prepared using different methods and carriers. However, fundamentally, these changes did not alter the rate-limiting step (k7), but merely affected the magnitude of the RDS rate. 4. For multiple diols, the model revealed that different catalysts have distinct rate-determining steps, yet the yields of these products consistently correlate with their respective RDS rates. In addition, distinct from traditional thermodynamic analyses that explain catalyst reactivity and efficiency, this study delved into the nuanced association between catalysts and key reaction steps from a kinetic perspective, offering a new dimension of understanding of the catalytic process.

Molecular-Level Catalytic Reforming Kinetic Model Based on Modified Structure-Oriented Lumping

Molecular-Level Catalytic Reforming Kinetic Model Based on Modified Structure-Oriented Lumping

Application of a property prediction model based on the structure oriented lumping method in the fluid catalytic cracking process

Based on the Structure Oriented Lumping (SOL) method and the Artificial Neural Network (ANN) algorithm, a SOL-ANN property prediction model was constructed to predict the properties of molecules and products in the fluid catalytic cracking (FCC) process. The properties of each structural vector in the molecular composition matrices of gasoline and diesel were calculated. The influences of reaction temperature on the properties of gasoline and diesel were investigated from the perspective of molecular composition. When the reaction temperature increased from 490 °C to 510 °C, the content of aromatics and olefins in gasoline and the content of aromatics in diesel increased, resulting in the research octane number (RON) of gasoline increasing by 2.96 units and the cetane number (CN) of diesel decreasing by 1.37 units. Using the molecular composition information of products to calculate the properties of molecules and products could guide the product quality evaluation and process optimization.

Molecular-level kinetic modelling for the pyrolysis of high-density polyethylene in a two-stage process

A two-stage process has been developed for the pyrolysis of high-density polyethylene to obtain light olefins, where probable mechanisms for each stage pyrolysis were proposed in our previous study. This work aims to establish molecular-level kinetic models for the first-stage pyrolysis at a lower temperature and for the second-stage pyrolysis at a higher temperature, respectively, to predict the overall reaction performances in accordance to the role of detailed mechanisms. A structure-oriented lumping (SOL) method based molecular-level kinetic model can effectively describe the first-stage pyrolysis, while a detailed molecular-level kinetic model is developed for the second-stage pyrolysis to track the evolutions of all alkanes, olefins, and diolefins, with special focus on the formation of light olefins. The established separate models can achieve the automatic generation of complex reaction networks and present good agreement with experimental data obtained from the first- and second-stage. In addition, the first-stage model indicates that the random scission plays an important role in the pyrolysis process at lower temperatures, while the chain-end scission mechanism dominates the second-stage reaction to produce light olefins at higher temperatures. As a whole, the molecular-level modelling scheme behaves excellently at a wide range of temperature, which provides guidance for upgrading the pyrolysis process.

Predicting the Physicochemical Properties of Molecules in Petroleum Based on Structural Increments

The structure-oriented lumping (SOL) method has been widely applied to molecular-level modeling and optimization in the refining industry. Calculating the physicochemical properties of products obtained by the SOL models can effectively guide the evaluation of product quality and optimization of the process. In this study, the SOL method was improved by extending the structural increments from 22 to 38 to describe the molecules in petroleum more accurately. According to the basic law that the properties and structures of molecules are inseparable, the structural increments with similar physical meanings were combined into a functional group to reduce the dimension to 11 characteristic parameters to describe the molecular structure, and a property prediction method that can be combined with the molecular-level refining models based on the optimized SOL method was constructed. In addition, the accuracy of the selection and definition of the 11 functional groups was verified by analyzing and discussing the effects of functional groups on the properties, which, in turn, ensured the accuracy of the property prediction models. Compared with other property prediction methods, this method can be combined with the SOL method to provide reliable predictions of the boiling point, critical temperature, critical volume, critical pressure, density, cetane number, motor octane number, and research octane number of molecules.

A coupling model of fluid catalytic cracking and diesel hydrotreating processes to study the effects of reaction temperature on the composition of diesel

The coupling optimization between different plants was the most difficult and key problem in refineries. To accurately describe the effective transfer of material flow information between Fluid Catalytic Cracking (FCC) and Diesel Hydrotreating (DH) processes and realize the coupling simulation for these processing units in refineries, a coupling model of FCC and DH processes using a fractionation tower simulation system was established to calculate the hydrocarbon compositions of diesel based on the Structure Oriented Lumping (SOL) method. The fractionation tower simulation system developed by MATLAB software was used to divide the molecular composition matrix of FCC product into the molecular composition matrices of dry gas, liquefied petroleum gas, gasoline, diesel, and oil slurry. The molecular composition matrix of diesel contained 73, 69, 183, 121 and 743 structural vectors representing paraffins, olefins, naphthenes, aromatics and non-hydrocarbon molecules, respectively. The fractionation ratios of the hydrocarbon group compositions with carbon numbers between 5 and 12 in the molecular composition matrix of gasoline and diesel were calculated. The effective transmission of material flow in the FCC and DH processes was accurately expressed by the transmission of the molecular composition matrix of diesel obtained by the fractionation tower simulation system. The effects of FCC and DH reaction conditions on the hydrocarbon compositions of diesel were calculated at the molecular level by the FCC-DH process coupling model. Through the FCC-DH process coupling model using a fractionation tower simulation system, the transformation laws of hydrocarbon and non-hydrocarbon molecules in different devices could be revealed to provide guidance for reducing the content of polycyclic aromatic hydrocarbons and sulfur in diesel, optimizing operating conditions and controlling product quality indicators.

Simulating polyethylene pyrolysis from a generalized molecular-level kinetic model

Waste plastics have emerged as a global environmental issue. Polyethylene, accounting for about 25% of global plastic production, has been acknowledged as the major cause of “white pollution”. In this paper, a generalized molecular-level kinetic model was proposed to understand the pyrolysis of polyethylene applying structure-oriented lumping, Poisson distribution, and genetic algorithm. The number of ordinary differential equations derived from the reaction kinetics were dramatically reduced by this simplified network with 14-rule design. The generalizability was confirmed by case studies without any change in the rule design. Under most conditions, the mean relative error between experimental and model data could be controlled within 5%, while abnormal deviations indicate the obstacle of heat transfer. The simulation curve could provide guidance for the experiments by locating the peak value of products. From 400–470 °C in Case 2, the increase in the reaction proportion of producing light oil explains the experimental results from the perspective of molecular-level reaction network. The comparison between PE and PS pyrolysis elaborates different pyrolysis mechanisms based on the different molecular structures.

Reaction laws of polycyclic aromatic hydrocarbons and heteroatomic compounds in hydrocracking process

Based on the Structure Oriented Lumping (SOL) method, a molecular-level reaction kinetic model for hydrocracking process was established to investigate the reaction laws of polycyclic aromatic hydrocarbons (PAHs) and heteroatomic compounds in hydrocracking process and calculate the molecular compositions of hydrocracking products. Molecules showed different reaction laws due to the differences of the initial contents, the structural increments and their positions in the reaction network. Since the hydrodenitrogenation reactions of indole and carbazole homologs were more difficult to carry out than the hydrodesulfurization reactions of benzothiophene and dibenzothiophene homologues, the denitrification rate was lower than the desulphurization rate at the outlet of the hydrotreating section. When the reaction temperature of the hydrotreating section was 360 °C, the conversion rates of sulfides and nitrides were 98.4 % and 91.3 %, respectively. The increase of reaction temperature intensified the hydrogenation saturation of aromatics, the ring opening of naphthenes and the cracking of paraffins. As the reaction temperature of the hydrocracking section rose from 370 °C to 410 °C, more isomerized carbon ions were generated due to the β-position cracking of paraffins. The mass ratio of methylbutane to n-pentane in light naphtha increased from 1.29 to 3.08. The increase of reaction temperature accelerated the side chain cracking reactions of naphthenes with 4 or more carbon atoms in the side chains, and generates more naphthenes with 0 to 3 carbon atoms in the side chains. Aromatics mainly underwent the ring-by-ring hydrogenation saturation reactions in the hydrotreating section, and bicyclic aromatic hydrocarbons was significantly easier to be hydrogenated than monocyclic aromatic hydrocarbons.

A molecular-level coupling model of fluid catalytic cracking and hydrotreating processes to improve gasoline quality

A molecular-level Fluid Catalytic Cracking (FCC)-Gasoline Hydrotreating (GH) process coupling model was established to guide the optimization of the hydrocarbon compositions in gasoline based on the Structure Oriented Lumping (SOL) method. 96 FCC reaction rules and 24 GH reaction rules were formulated, and a reaction network containing about 120,000 reactions was constructed. In order to establish the FCC-GH process coupling model, the effective transfer of composition information between the two processes was realized through the molecular composition matrix of gasoline. The molecular composition matrix of gasoline was obtained according to the classification rules of the molecular composition matrix of FCC products. The conversion laws of hydrocarbon molecules in gasoline were investigated by tracking their generation paths and reaction paths. The influences of reaction conditions on the distribution of hydrocarbons in the product gasoline could be calculated quantitatively by the FCC-GH process coupling model at molecular level.

Dry Glass Reference Perturbation Theory Predictions of the Temperature and Pressure Dependent Separations of Complex Liquid Mixtures Using SBAD-1 Glassy Polymer Membranes.

In this work we apply dry glass reference perturbation theory (DGRPT) within the context of fully mutualized diffusion theory to predict the temperature and pressure dependent separations of complex liquid mixtures using SBAD-1 glassy polymer membranes. We demonstrate that the approach allows for the prediction of the membrane-based separation of complex liquid mixtures over a wide range of temperature and pressure, using only single-component vapor sorption isotherms measured at 25 °C to parameterize the model. The model was then applied to predict the membrane separation of a light shale crude using a structure oriented lumping (SOL) based compositional model of petroleum. It was shown that when DGRPT is applied based on SOL compositions, the combined model allows for the accurate prediction of separation performance based on the trend of both molecular weight and molecular class.

Coupling simulation of delayed coking and hydrotreating process at molecular level

For the highly integrated system in the petrochemical industry, establishing a coupling simulation model for the units that were closely related at the molecular level was a big challenge due to the difficulty of massive and effective data interaction. Based on the structure-oriented lumping (SOL) method, the molecular composition matrix, which included material molecular information quantitatively, made it possible to build up the coupling simulation model. An interoperable molecular composition matrix including 1228 row vectors connected the delayed coking model and the hydrotreating model in series. The coupling simulation model of delayed coking and hydrotreating process was established at the molecular level. In addition to predicting the molecular composition of the products, the simulation model could track the reaction path and reflect the synergistic transformation effect of hydrocarbon and non-hydrocarbon molecules. And the simulation model could also quantitatively calculate the influence of changes in upstream reaction conditions on downstream products. The coupling simulation model provided useful explorations to achieve the petroleum molecular management throughout the whole petrochemical process.

Reaction network and molecular distribution of sulfides in gasoline and diesel of FCC process

In order to guide the accurate control of the sulfides in gasoline and diesel of Fluid Catalytic Cracking (FCC) process, a molecular-level model was established based on the Structure Oriented Lumping (SOL) method. According to the molecular composition characteristics of FCC feed oil, a molecular composition matrix containing 4,148 structural vectors was constructed with 24 structural increments. Using MATLAB software, the SOL reaction rules were compiled and a complex reaction network containing of about 110,000 reactions was established. According to the classification rules, the sulfides in gasoline and diesel represented by 32 and 136 structural vectors were divided into mercaptans, thioethers, monocyclic thiophenes, benzothiophenes and dibenzothiophenes, respectively. The conversion law of sulfides in gasoline and diesel was investigated by tracking their generation paths and reaction paths in the reaction network. The effects of the operation conditions on the sulfide content in gasoline and diesel were calculated quantitatively.

Calculation of reaction network and product properties of delayed coking process based on structural increments

The delayed coking reaction kinetic model based on the structure-oriented lumping method showed good accuracy in predicting the yield and composition of coking products at the molecular level. Based on the basic law that molecular structure determines molecular properties, the structural increment contribution method was constructed to expand the functions of established reaction kinetic model to predict the characteristic properties of coking gasoline and diesel products: distillation range, density, research octane number (RON) or cetane number (CN). The contributions of different structural increments of hydrocarbon molecules in gasoline to the RON and those in diesel to the CN could also be calculated by the model. The RONs of the hydrocarbon molecules decreased with the increase of R, and increased with the increase of br and me for the coking gasoline. The CNs of the paraffins molecules improved with the increase of R, but reduced with the increase of br and A6 for the coking diesel. The effects of all structural increments on the RON and CN could be quantitatively calculated through the established correlation formulas. The prediction of composition and properties of coking gasoline and diesel based on the model could guide the subsequent processing and process optimization of delayed coking plant.

Molecular-level reaction network in delayed coking process based on structure-oriented lumping

Based on the structure-oriented lumping method, a molecular-level reaction kinetic model of the delayed coking process, which adopted 24 structural increments to construct the feed molecular matrix containing 2944 molecules, was established with a reaction network containing 74,581 reactions using MATLAB. The reliability of the model was verified by experimental results. According to the discriminant rules of structural increments, 173 structural vectors in gasoline and 1132 structural vectors in diesel were classified into different group compositions, respectively. The model could track the reaction path of any specific molecule in the complex thermal cracking reaction network. The influences of operation conditions such as recycle ratio on the product distribution could be discovered through the calculation of the molecular-level model, which is helpful for the process optimization and precise regulation of product composition for the delayed coking plants.

Reaction network of sulfur compounds in delayed coking process

• A molecular-level model for delayed coking process was established. • The conversion behavior of sulfur in delayed coking process was investigated. • The sulfur compounds in gasoline were expressed as 57 structural vectors. • The sulfur compounds were classified according to the discriminant rules. In order to describe the reaction network and investigate the conversion behavior of sulfur compounds in delayed coking process, a molecular-level reaction kinetic model was established with a reaction network consisting of 81,081 reactions based on the structure-oriented lumping method. A feed molecular matrix containing 2,944 structural vectors was constructed with 24 structural increments based on the characteristics of the vacuum residue molecular composition. The reliability of the model was verified by industrial data. Through the calculation of the SOL model, the sulfur compounds in gasoline were expressed as 57 structural vectors, which were divided into thiophenes, mercaptans and thioethers according to the discriminant rules. The conversion law of sulfur compounds was investigated by tracking the reaction path of molecules in the reaction network. The SOL model could calculate the distribution of sulfur compounds in coking gasoline. The contents of thiophene molecules increased with the carbon numbers. Ethyl mercaptan and propyl mercaptan had higher content in mercaptans. The SOL model can reveal the influences of reaction conditions on sulfur compounds conversion in delayed coking process at the molecular level, which contributes to the accurate control of sulfur content in coking gasoline.

Online Optimization of Fluid Catalytic Cracking Process via a Hybrid Model Based on Simplified Structure-Oriented Lumping and Case-Based Reasoning

The mechanism models based on structure-oriented lumping (SOL) deliver a satisfactory prediction on the properties and yield distribution of the products from fluid catalytic cracking (FCC). Howeve...

Molecular level analysis on performance of diameter expanding reactor to improve gasoline quality in FCC process

A molecular level model for Fluid Catalytic Cracking (FCC) process was established based on the Structure-Oriented-Lumping (SOL) method to investigate the effects of the diameter expanding reactor on the product distribution. According to the molecular composition characteristics of feed oil, 24 structural increments were adopted to construct a molecular matrix containing 4539 molecules. Combined with the SOL reaction rules, a reaction network consisting of 118,272 reactions was established using MATLAB. The contents of 266 molecules in gasoline and 1189 molecules in diesel oil were calculated using the molecular level models of diameter expanding reactor and conventional reactor. The olefins content in gasoline, which is the summarization of 83 olefins molecules, increased as the reaction temperature went up. Compared with the conventional reactor, the ratio of isopentane to pentene in the diameter expanding reactor increased from 1.38 to 2.59 and the ratio of toluene to methylcyclohexane increased from 2.69 to 3.94. The model illustrated that the diameter expanding reactor could enhance the hydrogen transfer and isomerization reactions from the molecular level to reduce the olefins content and improve the iso-paraffins content in gasoline. The model could describe the molecular level product distribution from the reactor inlet to the outlet.

Hydrocracking: A Perspective towards Digitalization

In a world of fast technological advancements, it is increasingly important to see how hydrocracking applications can benefit from and adapt to digitalization. A review of hydrocracking processes from the perspective of modeling and characterization methods is presented next to an investigation on digitalization trends. Both physics-based and data-based models are discussed according to their scope of use, needs, and capabilities based on open literature. Discrete and continuous lumping, structure-oriented lumping, and single event micro-kinetic models are reported as well as artificial neural networks, convolutional neural networks, and surrogate models. Infrared, near-infrared, ultra-violet and Raman spectroscopic methods are given with their examples for the characterization of feed or product streams of hydrocracking processes regarding boiling point curve, API, SARA, sulfur, nitrogen and metal content. The critical points to consider while modeling the system and the soft sensor are reported as well as the problems to be addressed. Optimization, control, and diagnostics applications are presented together with suggested future directions of interdisciplinary studies. The links required between the models, soft sensors, optimization, control, and diagnostics are suggested to achieve the automation goals and, therefore, a sustainable operation.

A Fragment-Based Mechanistic Kinetic Modeling Framework for Complex Systems

There is great interest in developing chemical kinetic models with predictive power. A predictive kinetic model maintains high fidelity to the true chemistry usually via a full-detail molecule representation which distinguishes all the molecules and isomers. Unfortunately such full-detail representation quickly becomes intractable for large molecules because the number of isomers grows very rapidly with size. Several less detailed representations have been developed, including the popular Structure-Oriented Lumping (SOL) method, which allow one to model larger molecules. In this paper we first examine scalability for two existing representations (full-detail and SOL) and propose a new modeling methodology tailored for heavy systems: fragment-based modeling, which promises the best scalability. We show via a case study of heavy molecule pyrolysis that the new approach creates a much smaller reaction network but with similar prediction accuracy on feedstock conversion and products’ molecular weight distribu...

A kinetic model for in situ coking denitrification of heavy oil with high nitrogen content based on starch using a structure-oriented lumping method.

In this study, in situ coking denitrification technology was utilized to simplify the entire process by adding an appropriate quantity of denitrification agents to the delayed-coking tower without any further follow-up denitrification process. The effect of starch as the denitrification agent in the in situ coking denitrification process was studied. The distribution of oil products was characterized by GC of simulated distillation. The results indicated that the nitrogen compounds present in heavy oil transformed to coke via in situ coking conversion by reacting with starch. A molecular-level process model for the coking process was developed. Product distribution on the complex reaction network could be described accurately by the model.