A method for constructing a refinery cluster model

Abstract

The application discloses a kind of oil refining device group model construction method, it is related to petroleum refining technical field, method includes: the detailed characterization of crude oil obtains crude oil molecular composition information;According to the molecular structure characteristics of crude oil in each crude oil molecular composition information, using structure-oriented lumping method establishes crude oil molecular database;According to the reaction rule of the process flow to be executed, establish molecular level single device reaction process model;According to the process flow to be executed and the input and output of each molecular level single device reaction process model, determine material mixing rule;According to the process flow to be executed, material mixing rule and each molecular level single device reaction process model, through molecular level standard data chain makes each molecular level single device reaction process model coupling obtains molecular level oil refining device group model;According to molecular level oil refining device group model, predicts the product composition and property of each molecular level oil refining device group model in molecular level oil refining device group model.The application realizes the molecular management of whole cycle of oil refining process.

Description

Technical Field

[0001] This invention relates to the field of petroleum refining technology, and in particular to a method for constructing a model of an oil refining unit group. Background Technology

[0002] With social development, the upgrading of gasoline and diesel quality standards and the intensification of the trend of "reducing oil and increasing chemical and special products" are driving the refining and chemical industry to achieve molecular management of the entire life cycle of crude oil processing.

[0003] The petroleum refining process is an extremely complex and highly coupled system, involving tens of thousands of molecules, dozens of reactors, and hundreds of supporting devices. Changes in upstream process conditions inevitably lead to changes in the molecular composition of downstream materials and adjustments to downstream process conditions. This means that the petroleum refining process must be designed and innovated within an integrated system framework. The massive data interaction between highly interconnected operational units poses a significant challenge to establishing multi-process coupled models. Currently, mainstream petrochemical process simulation software, multi-lumped models, and data-driven models, due to the heterogeneity of their underlying architecture, struggle to achieve molecular-level massive data interaction between different process models, thus failing to effectively establish coupled models of multiple processes. Summary of the Invention

[0004] The purpose of this invention is to provide a method for constructing a model of an oil refining unit group, which enables molecular management of the entire oil refining process.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A method for constructing a model of an oil refinery unit group includes:

[0007] The crude oil was characterized in detail to obtain information on its molecular composition.

[0008] Based on the structural characteristics of crude oil molecules in the information on the composition of each crude oil molecule, a structure-oriented lumped method is used to establish a crude oil molecule database;

[0009] Based on the reaction rules of the process to be executed, establish a molecular-level single-device reaction process model;

[0010] Based on the process flow to be executed and the inputs and outputs of the reaction process models of each molecular-level single unit, determine the material mixing rules of the input materials of each molecular-level single unit reaction process model;

[0011] Based on the process flow to be executed, the material mixing rules, and the reaction process models of each molecular-level single unit, the reaction process models of each molecular-level single unit are coupled through a molecular-level standard data link to obtain a molecular-level refining unit group model.

[0012] Based on the molecular-level refining unit group model, predict the product composition and product properties of each molecular-level refining unit group model.

[0013] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0014] This invention couples the reaction process models of each molecular-level single unit through a molecular-level standard data link based on the process flow to be executed, thereby obtaining a molecular-level refining unit group model. Based on the molecular-level refining unit group model, the product composition and product properties of each molecular-level refining unit group model are predicted. This completes the underlying logic construction of the model at the molecular level and also realizes the effective interaction of material molecular composition information between different process models, thereby achieving molecular management of the entire refining process cycle. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of a method for constructing a model of an oil refining unit group provided in an embodiment of the present invention;

[0017] Figure 2 A diagram illustrating the refined characterization scheme for all fractions of crude oil provided in this embodiment of the invention;

[0018] Figure 3 A flowchart illustrating the construction of a molecular-level single-device reaction process model provided in this embodiment of the invention;

[0019] Figure 4 This is a process flow diagram of an oil refining unit group provided in an embodiment of the present invention;

[0020] Figure 5 This is a schematic diagram illustrating the transfer of molecular-level standard data links between the delayed coking model and the diesel hydrorefining model provided in this embodiment of the invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] The purpose of this invention is to provide a method for constructing a model of an oil refining unit group, which enables molecular management of the entire oil refining process.

[0023] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] like Figure 1 As shown in this embodiment, a method for constructing a refinery unit group model includes the following steps.

[0025] Step 101: Perform detailed characterization of crude oil to obtain information on its molecular composition.

[0026] Step 102: Based on the structural characteristics of crude oil molecules in the composition information of each crude oil molecule, a structure-oriented lumped method is used to establish a crude oil molecule database.

[0027] Step 103: Establish a molecular-level single-device reaction process model based on the reaction rules of the process to be executed.

[0028] Step 104: Based on the process flow to be executed and the inputs and outputs of the reaction process models of each molecular-level single unit, determine the material mixing rules of the input materials of each reaction process model of each molecular-level single unit.

[0029] Step 105: Based on the process flow to be executed, the material mixing rules, and the reaction process models of each molecular-level single unit, couple the reaction process models of each molecular-level single unit through the molecular-level standard data link to obtain the molecular-level refining unit group model; there is a molecular-level single unit reaction process model as a material information carrier between every two molecular-level single unit reaction process models with upstream and downstream reaction relationships.

[0030] In the molecular-level refining unit cluster model, the product molecular composition matrix calculated and output by the upstream molecular-level single-unit reaction process model is used as the raw material molecular composition matrix input by the downstream molecular-level single-unit reaction process model.

[0031] Step 106: Based on the molecular-level refining unit group model, predict the product composition and product properties of each molecular-level refining unit group model in the molecular-level refining unit group model.

[0032] Step 101 involves constructing a refined characterization scheme for the entire crude oil fraction to obtain information on the molecular composition of the crude oil, such as... Figure 2 As shown, step 101 specifically includes:

[0033] The crude oil is pre-separated using true boiling point distillation technology to obtain multiple fractions, including gasoline, diesel, atmospheric wax oil, vacuum wax oil, and vacuum residue.

[0034] Gas chromatography was used to obtain information on the gasoline molecular composition of the gasoline.

[0035] The molecular composition information of the diesel fuel was obtained using a two-dimensional gas chromatography-mass spectrometry system.

[0036] The carbon number and group composition information of atmospheric pressure wax oil and vacuum pressure wax oil were obtained by gas chromatography-mass spectrometry.

[0037] To address the complex composition and large molecular weight of vacuum residue, supercritical fluid extraction fractionation technology is employed to separate the vacuum residue into extractive oil and asphaltenes.

[0038] The carbon number-group composition information of the extracted distillate oil was obtained using Fourier transform ion cyclotron resonance mass spectrometry.

[0039] The average molecular structure information of the asphaltene was obtained by using nuclear magnetic resonance, mass spectrometry and elemental analysis.

[0040] Step 102, based on the structure-oriented lumped method, uses the Poisson distribution function and simulated annealing algorithm to establish a crude oil molecular database, specifically including:

[0041] Based on the structural characteristics of crude oil molecules in the information on the composition of each crude oil molecule, a structure-guided lumped method is used to generate 24 structural units that represent molecular functional groups or structural features.

[0042] A predetermined number of core molecules are selected, and methylene branches are added to these core molecules in an orderly manner, resulting in multiple raw material molecule matrices composed of multiple structural vectors. Each structural vector is represented by a combination of 24 structural units. More specifically, based on the structure-guided lumped method and the characteristics of crude oil molecules, 24 structural units representing molecular functional groups or structural features are designed. 150 types of core molecules, including alkanes, cycloalkanes, aromatics, sulfides, nitrides, oxides, and metalloporphyrins, are selected. Methylene branches are added to these core molecules in an orderly manner, ensuring that the number of homologues in each core molecule reaches a maximum of 85 carbon atoms. Then, some unreasonable molecules are removed, resulting in a 10000×24 raw material molecule matrix composed of 10000 structural vectors. A crude oil molecular structure database is constructed using a database language, meaning the crude oil molecular structure database includes 10000×24 raw material molecule matrices. The 24 structural units are A6, A5, A2, N6, N5, N4, N3, N2, N1, R, IH, br, me, AA, NS, NO, NN, RS, RO, RN, AN, KO, Ni and V. The structural formulas of the 24 structural units are shown in Table 1.

[0043] Table 1. Structural formulas of 24 structural units

[0044]

[0045] Wherein, A6 represents the benzene ring, A4 represents any aromatic ring other than the benzene ring in an aromatic hydrocarbon, A2 represents a polygonally fused aromatic ring in an aromatic hydrocarbon, N6 represents a six-membered ring in a cycloalkanes, N5 represents a five-membered ring in a cycloalkanes, N4 represents a cycloalkanes with 4 carbon atoms directly attached to an aromatic or cycloalkanes, N3 represents a cycloalkanes with 3 carbon atoms directly attached to an aromatic or cycloalkanes, N2 represents a cycloalkanes with 2 carbon atoms directly attached to an aromatic or cycloalkanes, N1 represents a cycloalkanes with 1 carbon atom directly attached to an aromatic or cycloalkanes, R represents the total number of carbon atoms in the alkyl group, and IH represents the saturation degree. An increase of 1 in IH represents the addition of two hydrogen atoms to the molecule, i.e., IH = 1 represents an alkane, IH = 0 represents a monoalkene, and IH = -1 represents a diene or alkyne. br represents the number of alkyl branches, me represents the number of methyl groups directly attached to the aromatic or cycloalkane ring, AA represents the number of bridging bonds between the six-membered rings in the benzene ring, the five-membered rings in the cycloalkane, NS represents the S atom in the cycloalkane or chain hydrocarbon that connects two carbon atoms, NO represents the O atom in the cycloalkane or chain hydrocarbon that connects two carbon atoms, NN represents the N atom in the cycloalkane or chain hydrocarbon that connects two carbon atoms, RS represents the number of mercapto groups in the molecule, RO represents the number of hydroxyl groups in the molecule, RN represents the number of amino groups in the molecule, AN represents a single atom on the aromatic ring, such as pyridine or quinoline, KO represents the number of methylene groups substituted by carbonyl groups, such as ketones, carboxylic acids, aldehydes, esters, etc., Ni represents the nickel atom in crude oil, and V represents the vanadium atom in crude oil.

[0046] In this invention, any oil molecule in the reactants-products of the petroleum processing reaction system can be represented by a structural vector composed of the above 24 structural units arranged in an orderly manner. The numerical value of the elements in the vector represents the number of the corresponding structural units in the molecule.

[0047] Considering the complexity of heavy oil molecular composition, this invention assumes that homologous molecules follow a Poisson distribution, and the Poisson equation for the feedstock distribution is as follows:

[0048]

[0049] Where n is the number of carbon atoms in the homologue molecule, C n X = k represents the relative mass fraction of molecules in a homologous series. n Let λ be the sequence number of a molecule in its homologue, and λ be the expected value of that homologue.

[0050] Based on the compositional information of each crude oil molecule, an objective function is constructed using the calculated and actual values ​​expressed in the crude oil molecular structure database. Using this objective function, a simulated annealing algorithm is employed to calculate the content of structure vectors in the raw material molecule matrix representing the compositional information of each crude oil molecule, thus establishing the crude oil molecular database. More specifically, based on the crude oil molecular compositional information, an objective function is constructed using the sum of squared residuals between the calculated and actual values ​​expressed in the crude oil molecular structure database. A simulated annealing algorithm is then used to calculate the content of each structure vector, thus establishing the crude oil molecular database. The crude oil molecular structure database can be understood as a raw material molecule matrix.

[0051] Information on the molecular composition of crude oil includes the elemental content, molecular weight, and molecular content of the molecules. The elemental content and molecular weight can be calculated based on 24 structural units; the molecular content is obtained by adding a column vector representing the molecular content after the 24 structural units.

[0052] The objective function is expressed as:

[0053] Where F1(w) is the objective function value, w is the parameter to be fitted (the column vector of the molecular content to be fitted in the crude oil molecular composition information), n is the carbon number of the homologue molecule, and P c P represents the calculated values ​​of various property indicators (homogeneous molecules with carbon number i in the crude oil molecular composition information). a These are the true values ​​of each property indicator.

[0054] In step 103, based on the process characteristics and reaction mechanism, reaction rules and reaction networks are designed to construct a molecular-level single-device reaction process model.

[0055] This invention targets processes such as atmospheric and vacuum distillation, catalytic cracking, delayed coking, hydrorefining, hydrocracking, and catalytic reforming. It combines reaction mechanisms such as carbocation reaction mechanism, free radical reaction mechanism, and hydrogenation catalysis, and designs reaction rules and constructs reaction networks based on the structure-guided lumped method.

[0056] Atmospheric and vacuum distillation is the first step in petroleum processing. It involves using atmospheric and vacuum methods to separate crude oil into distillate oils such as gasoline, diesel, atmospheric wax oil, vacuum wax oil, and vacuum residue oil according to their boiling range.

[0057] Catalytic cracking is a process in which heavy oil is used as feedstock, and in the presence of an acidic catalyst, at 500℃ and 0.1–0.3 MPa, a series of chemical reactions mainly following the carbocation reaction mechanism occur, producing gas, light oil, and coke.

[0058] Delayed coking is a processing method that uses vacuum residue as raw material and undergoes deep thermal cracking reaction at 490–505℃, mainly following the free radical reaction mechanism, to produce gas, gasoline, diesel, wax oil and coke.

[0059] Hydrorefining uses gasoline and diesel as feedstock, and processes the process on a metal-alumina catalyst at pressures of 3.0–8.0 MPa, temperatures of 250–400 °C, and space velocities of 1.0–5.0 h⁻¹. -1 Under reaction conditions of a hydrogen-to-oil ratio of 100–500, the process aims to remove sulfur, nitrogen, oxygen heteroatoms and metal impurities, while selectively hydrogenating and saturating olefins, dienes, aromatic hydrocarbons and fused-ring aromatic hydrocarbons.

[0060] Hydrocracking is a process that uses vacuum gas oil, coking gas oil, cracking cycle oil, and deasphalted oil as feedstocks. Under hydrogenation and high-pressure reaction conditions on a metal-zeolite molecular sieve catalyst, more than 10% of the molecules in the feedstock oil are reduced in size to produce high-quality light oil products such as gasoline, diesel, and aviation kerosene.

[0061] Catalytic reforming is a process that uses naphtha as a raw material to produce high-octane gasoline blending components and light aromatic hydrocarbons (benzene, toluene, xylene), while also producing hydrogen as a byproduct.

[0062] Based on the characteristics of the reaction process and the structure-oriented lumped method, reaction rules were designed and a reaction network was constructed. The main reaction rules designed are shown in Tables 2 to 6:

[0063] Table 2. Catalytic Cracking Reaction Rules

[0064]

[0065]

[0066] Table 3 Rules for Delayed Coking Reactions

[0067]

[0068]

[0069] Table 4 Hydrorefining Reaction Rules

[0070]

[0071] Table 5 Rules for Hydrocracking Reactions

[0072]

[0073]

[0074] Table 6 Rules for Catalytic Reforming Reactions

[0075]

[0076]

[0077] A Runge-Kutta-like method was used to solve the reaction network and establish a molecular-level single-device reaction process model. The Runge-Kutta-like method, borrowed from the Runge-Kutta method, divides the entire reaction time of a single process into several micro-reaction intervals. The initial reaction temperature and reactant molecular composition of each micro-reaction interval are the same as the termination reaction temperature and product molecular composition of the previous micro-reaction interval, and this process is repeated until the reactor outlet. More specifically, such as... Figure 3 As shown, the Runge-Kutta method includes: dividing the entire reaction time t of a single process into k reaction intervals, each interval being denoted as Δt. i (i = 1 to k). In the first reaction interval Δt1, the reaction temperature T1 and the raw material molecular composition matrix are input, and the corresponding reaction network is generated according to the reaction rules. The reaction rate constant is obtained by referring to the reaction rate constant correlation formula, and the product molecular composition matrix in the reaction interval Δt1 is calculated by combining the reaction time and reactant content. Then, the molecular structure vector set obtained by merging the unreacted reactant molecular structure vector in the reaction interval Δt1 with the product molecular structure vector at reaction time Δt1 is used as the reactant molecular composition matrix of the reaction interval Δt2. At the same time, the total heat of reaction in the reaction interval Δt1 is calculated and converted into a temperature difference ΔT1 to obtain the reaction temperature T2 of the next reaction interval, T2 = T1 + ΔT1. Then, the reaction process in the reaction interval Δt2 is calculated. The calculation is repeated until the reaction time reaches t, and the calculated single process product molecular composition matrix is ​​used as the output of the model. The above calculation process is similar to the Runge-Kutta method, so it is defined as a Runge-Kutta-like method.

[0078] Step 103 specifically includes: dividing the entire reaction time t of a single process in the process flow to be executed into k reaction intervals.

[0079] Within the i-th reaction interval, the product molecular structure vector of the i-th reaction interval is generated based on the reaction temperature, raw material molecular matrix, and reaction rules. The molecular structure vector set obtained by merging the unreacted reactant molecular structure vector and the product molecular structure vector in the i-th reaction interval is used as the raw material molecular matrix of the (i+1)-th reaction interval. The total heat of reaction in the i-th reaction interval is calculated. The reaction temperature of the (i+1)-th reaction interval is calculated based on the reaction temperature of the i-th reaction interval and the total heat of reaction. This calculation is repeated until the reaction time reaches the entire reaction time t. The molecular structure vector set obtained in the n-th reaction interval is used as the output of this single process, thus obtaining the molecular-level single-device reaction process model of this single process.

[0080] Step 104 involves designing mixing rules for material composition and properties based on the characteristics of the process flow. Specifically, this includes designing mixing rules for the molecular composition and properties of different materials based on the actual process flow of the refinery.

[0081] Generally, the molecular weight, density, and distillation range of petroleum products can be calculated based on the additive nature of mass or volume. There are two mixing rules: the first, for factors like molecular weight, density, and distillation range, can be directly calculated using mass addition; the second, for factors like the octane number of gasoline and the cetane number of diesel, requires calculation using the following formula.

[0082] For example, the density of No. 1 straight-run diesel oil is 0.83 km / m³. 3 The cetane number is 45, and the density of No. 2 coking diesel is 0.85 km / m³. 3 Both have a cetane number of 44. The typical process involves feeding both types of diesel fuel together into a diesel hydrorefining unit, which requires a mixing process.

[0083] If mixed in a 50%:50% ratio, the density would be 0.84 kg / m³. 3 However, the cetane number of diesel fuel is not necessarily 44.5; it is also affected by intermolecular forces and needs to be calculated specifically according to the following formula.

[0084] The mixing rules for properties such as the octane number of gasoline and the cetane number of diesel are as follows:

[0085]

[0086] Among them, ON mix The octane number or hexadecane number of the mixture, ON i It is the octane number or hexadecane number of the pure substance, C. i C is the concentration of pure substance i. j Let be the concentration of pure substance j, and B be the total deviation, as shown in the following formula:

[0087]

[0088] Among them, B i B j Let be the interaction parameters for molecules i and j, respectively, and α and n be the kinetic coefficients used to define the dipole moment. The strength of the resulting intermolecular interactions.

[0089] Step 105 involves constructing a molecular-level standard data chain that serves as the material information carrier for individual upstream and downstream reaction process models, thereby coupling reaction process models from different units and establishing a molecular-level refining unit cluster model. Specifically, this includes:

[0090] In the reaction process model established based on the structure-oriented lumped method, the molecular composition matrix is ​​the carrier of information on the molecular composition and properties of materials.

[0091] The molecular composition matrix includes both the raw material molecular composition matrix and the product molecular composition matrix. The product molecular composition matrix is ​​generally obtained by inputting the raw material molecular composition matrix into the reaction process model and performing calculations by the model.

[0092] Based on the actual upstream and downstream process flow of the refinery, the material flow between different units was sorted out. The product molecular composition matrix calculated and output by the reaction model of the upstream unit was used as the raw material molecular composition matrix input by the reaction model of the downstream unit, and named as the molecular-level standard data chain, as shown in Table 7.

[0093] Table 7 Molecular-level standard data link

[0094]

[0095] There is a molecular-level standard data link between each two process flows.

[0096] By establishing a molecular-level standard data chain between upstream and downstream units throughout the entire oil refining process and coupling reaction process models of different units, a molecular-level oil refining unit group model is thus established.

[0097] The refining processes involved in this invention include crude oil atmospheric and vacuum distillation, catalytic cracking, delayed coking, gasoline hydrorefining, diesel hydrorefining, wax oil hydrocracking, catalytic reforming, and other processes.

[0098] Figure 4 A process flow diagram of a refining unit group according to an embodiment of the present invention is disclosed, such as... Figure 4 As shown, the crude oil atmospheric and vacuum distillation unit is mainly divided into straight-run gasoline, straight-run diesel, wax oil, and residue oil in sequence. Straight-run gasoline is fed to the gasoline hydrorefining unit, straight-run diesel to the diesel hydrorefining unit, wax oil to the catalytic cracking unit, and residue oil to the delayed coking unit. Catalytic gasoline produced by the catalytic cracking unit is fed to the gasoline hydrorefining unit. Coking gasoline produced by the delayed coking unit is fed to the gasoline hydrorefining unit, coking diesel to the diesel hydrorefining unit, and coking wax oil to the wax oil hydrocracking unit. The refined gasoline produced by the gasoline hydrorefining, diesel hydrorefining, and wax oil hydrocracking units is further fed to the catalytic reforming unit.

[0099] For the entire oil refining process, which includes crude oil atmospheric and vacuum distillation, catalytic cracking, delayed coking, gasoline hydrorefining, diesel hydrorefining, wax oil hydrocracking, and catalytic reforming, a molecular-level standard data link is established between the upstream and downstream units. After coupling the reaction process models of different units, a molecular-level refining unit group model is thus established.

[0100] Figure 5 A schematic diagram of the molecular-level standard data link transfer between the delayed coking model and the diesel hydrorefining model is revealed. Figure 5 In the diagram, A is a coking furnace, B is a coke tower, C is a coking fractionation tower, D is a coking stripping tower, E is a hydrotreating furnace, F is a hydrotreating reactor, G is a separator, H is a hydrotreating fractionation tower, 1 is vacuum residue, 2 is coking gas, 3 is coking gasoline, 4 is coking diesel, 5 is coking wax oil, 6 is coke, 7 is hydrotreated liquefied petroleum gas, 8 is hydrotreated naphtha, and 9 is hydrotreated diesel.

[0101] Step 106 involves using a molecular-level refining unit cluster model to calculate the product composition and properties of each unit within the refining unit cluster. Specifically, this includes using the established molecular-level refining unit cluster model to calculate the product composition and properties of unit units such as catalytic cracking, delayed coking, hydrorefining, hydrocracking, and catalytic reforming within the refining unit cluster.

[0102] Product properties include density, distillation range, gasoline octane number, and diesel cetane number.

[0103] The calculation results of the molecular-level refining unit group model are shown in Tables 8 to 13:

[0104] Table 8 Catalytic Cracking

[0105]

[0106] Table 9 Delayed Coking

[0107]

[0108]

[0109] Table 10 Gasoline Hydrorefining

[0110]

[0111]

[0112] Table 11 Diesel Hydrorefining

[0113]

[0114]

[0115] Table 12 Hydrocracking of Wax Oils

[0116]

[0117]

[0118] Table 13 Catalytic Reforming

[0119]

[0120] In summary, the molecular-level refining unit cluster model established in this invention has good computational accuracy and can accurately predict information such as product yield, group composition and typical molecular composition of each process unit in the entire refining process. It also conforms to the reaction law of petroleum processing and can provide scientific data support and guidance for realizing molecular management of the entire life cycle of crude oil processing.

[0121] The present invention proposes a method for constructing a refining unit group model, specifically a method for constructing a refining unit group model based on molecular-level data link transmission. Based on the structure-oriented lumped method and molecular-level reaction mechanism, combined with the petroleum processing technology flow, a refining unit group model at the molecular scale is established, which has reliable calculation accuracy and conforms to the process characteristics and reaction laws of refining.

[0122] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0123] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for constructing a model of an oil refining unit group, characterized in that, include: The crude oil was characterized in detail to obtain information on its molecular composition. Based on the structural characteristics of crude oil molecules in the information on the composition of each crude oil molecule, a structure-oriented lumped method is used to establish a crude oil molecule database; Based on the reaction rules of the process flow to be executed, a molecular-level single-device reaction process model is established, specifically including: dividing the entire reaction time t of a single process in the process flow to be executed into k reaction intervals; Within the i-th reaction interval, the product molecular structure vector of the i-th reaction interval is generated based on the reaction temperature, raw material molecular matrix, and reaction rules. The molecular structure vector set obtained by merging the unreacted reactant molecular structure vector and the product molecular structure vector in the i-th reaction interval is used as the raw material molecular matrix of the (i+1)-th reaction interval. The total heat of reaction in the i-th reaction interval is calculated. The reaction temperature of the (i+1)-th reaction interval is calculated based on the reaction temperature of the i-th reaction interval and the total heat of reaction. This calculation is repeated until the reaction time reaches the entire reaction time t. The molecular structure vector set obtained in the n-th reaction interval is used as the output of this single process, thus obtaining the molecular-level single-device reaction process model of this single process. Based on the process flow to be executed and the inputs and outputs of the reaction process models of each molecular-level single unit, determine the material mixing rules of the input materials of each molecular-level single unit reaction process model; Based on the process flow to be executed, the material mixing rules, and the reaction process models of each molecular-level single unit, the reaction process models of each molecular-level single unit are coupled through a molecular-level standard data link to obtain a molecular-level refining unit group model. Based on the molecular-level refining unit group model, predict the product composition and product properties of each molecular-level refining unit group model.

2. The method for constructing a refinery unit group model according to claim 1, characterized in that, Detailed characterization of crude oil was performed to obtain information on its molecular composition, specifically including: The crude oil is pre-separated using true boiling point distillation technology to obtain multiple fractions, including gasoline, diesel, atmospheric wax oil, vacuum wax oil, and vacuum residue. Gas chromatography was used to obtain information on the gasoline molecular composition of the gasoline. The molecular composition information of the diesel fuel was obtained using a two-dimensional gas chromatography-mass spectrometry system. The carbon number-group composition information of atmospheric pressure wax oil and vacuum pressure wax oil was obtained by gas chromatography-mass spectrometry. The vacuum residue was separated into extractive oil and asphaltenes using supercritical fluid extraction fractionation technology. The carbon number-group composition information of the extracted distillate oil was obtained using Fourier transform ion cyclotron resonance mass spectrometry. The average molecular structure information of the asphaltene was obtained by using nuclear magnetic resonance, mass spectrometry and elemental analysis.

3. The method for constructing a refinery unit group model according to claim 1, characterized in that, Based on the structural characteristics of crude oil molecules in the composition information of each crude oil molecule, a structure-oriented lumped method is used to establish a crude oil molecule database, specifically including: Based on the structural characteristics of crude oil molecules in the information on the composition of each crude oil molecule, a structure-oriented lumped method is used to generate 24 structural units that represent molecular functional groups or structural features. A predetermined number of core molecules are selected, and methylene branches are added to the core molecules in an orderly manner to obtain multiple raw material molecule matrices composed of multiple structural vectors; the structural vectors are represented by combinations of 24 structural units; Based on the composition information of each crude oil molecule, an objective function is constructed by combining the calculated and actual values ​​expressed in the crude oil molecule structure database. According to the objective function, a simulated annealing algorithm is used to calculate the content of the structure vector in the raw material molecule matrix representing the composition information of each crude oil molecule, and a crude oil molecule database is established.

4. The method for constructing a refinery unit group model according to claim 3, characterized in that, The 24 structural units are A6, A5, A2, N6, N5, N4, N3, N2, N1, R, IH, br, me, AA, NS, NO, NN, RS, RO, RN, AN, KO, Ni, and V; Wherein, A6 represents the benzene ring, A5 represents any aromatic ring other than the benzene ring in an aromatic hydrocarbon, A2 represents a polygonally fused aromatic ring in an aromatic hydrocarbon, N6 represents a six-membered ring in a cycloalkanes, N5 represents a five-membered ring in a cycloalkanes, N4 represents a cycloalkane ring with 4 carbon atoms directly attached to an aromatic or cycloalkane ring, N3 represents a cycloalkane ring with 3 carbon atoms directly attached to an aromatic or cycloalkane ring, N2 represents a cycloalkane ring with 2 carbon atoms directly attached to an aromatic or cycloalkane ring, N1 represents a cycloalkane ring with 1 carbon atom directly attached to an aromatic or cycloalkane ring, R represents the total number of carbon atoms in the alkyl group, IH represents the degree of saturation, br represents the number of alkyl branches, and me represents the number of branches in the alkyl group. The number of methyl groups directly connected to the ring or cycloalkane ring; AA is the number of bridging bonds between six-membered rings in benzene rings, cycloalkanes, or five-membered rings in cycloalkanes; NS is the S atom connecting two carbon atoms in cycloalkanes or chain hydrocarbons; NO is the O atom connecting two carbon atoms in cycloalkanes or chain hydrocarbons; NN is the N atom connecting two carbon atoms in cycloalkanes or chain hydrocarbons; RS is the number of mercapto groups in the molecule; RO is the number of hydroxyl groups in the molecule; RN is the number of amino groups in the molecule; AN is a single atom on the aromatic ring; KO is the number of methylene groups substituted by carbonyl groups; Ni is the nickel atom in crude oil; and V is the vanadium atom in crude oil.

5. The method for constructing a refinery unit group model according to claim 3, characterized in that, The objective function is expressed as: ； in, Let be the objective function value, and n be the number of carbon atoms in the homologue molecule. This is the calculated value for the homologue molecule with carbon number i in the crude oil molecular composition information. This represents the true value of the homologue molecule with carbon number i in the crude oil molecular composition information.

6. The method for constructing a refinery unit group model according to claim 1, characterized in that, The product properties include density, distillation range, gasoline octane number, and diesel cetane number.

Images