A molecular-level atmospheric and vacuum corrosion simulation method and device

Abstract

The present invention discloses a molecular-level atmospheric and vacuum corrosion simulation method and apparatus. The method comprises obtaining first macroscopic property data of the current crude oil based on a crude oil physical property analysis model; obtaining first molecular concentrations of the corrosive molecules in the current crude oil based on a preset corrosive molecule library and the first macroscopic property data; obtaining virtual components of the corrosive molecules in the current crude oil based on a preset physical property database and the first molecular concentration; and performing iterative corrosion simulation on the atmospheric and vacuum device based on preset DCS data and the virtual components of the corrosive molecules, outputting simulation results of the distribution of the corrosive molecules in various parts of the atmospheric and vacuum device. Compared with the existing technology, the present invention can improve the accuracy and timeliness of the prediction of the distribution of corrosive molecules in various parts of the atmospheric and vacuum device.

Description

Technical Field

[0001] The present invention relates to the field of corrosion simulation of petroleum refining equipment, and in particular to a molecular-level atmospheric and vacuum corrosion simulation method and device. Background Art

[0002] In the petroleum refining process, distillation is the first step in crude oil processing. Atmospheric and vacuum distillation, which fractionates crude oil into components such as gasoline, kerosene, diesel, wax oil, and residual oil, is a key process in subsequent processing. However, with the increasing depth of crude oil processing, the proportion of heavy, high-sulfur, high-nitrogen, and high-salt crude oil has increased annually. This has led to increased corrosiveness in the acidic water of atmospheric and vacuum distillation units, causing severe salt buildup in equipment such as heat exchangers and air coolers, compromising the safe and stable operation of the units. To improve corporate profitability and reduce unplanned downtime, research on corrosion safety in atmospheric and vacuum distillation units has become an urgent issue. Corrosion in atmospheric and vacuum distillation units primarily stems from the thermal decomposition or hydrolysis of chlorides and sulfides in crude oil during the distillation process, generating corrosive media such as hydrogen chloride, hydrogen sulfide, and organic acids. These media react with the metal materials of equipment and pipelines, causing corrosion. Chlorine in crude oil exists in both inorganic and organic forms. While electrical desalters can remove 70%–100% of inorganic chlorides, small amounts of inorganic chlorides and some organic chlorides still enter downstream refining units. During processing, the hydrolysis of inorganic chlorides and the decomposition of organic chlorides generate hydrogen chloride gas, which can cause severe corrosion to production equipment and pipelines. Furthermore, nitrogen compounds in crude oil are easily oxidized to form colloids and sediments, which affect the oil's oxidative stability, poison catalysts, and generate ammonium salts, leading to under-deposit corrosion in equipment. Sulfides generate hydrogen sulfide during processing, triggering chemical corrosion and stress corrosion of equipment. Petroleum acids can also cause severe corrosive damage to equipment. Therefore, understanding the distribution of corrosive media such as chlorine, nitrogen, sulfur, and acids within the equipment is crucial.

[0003] Due to the complexity of crude oil composition, current atmospheric and vacuum distillation simulations typically use virtual components for characterization and calculation. These virtual components are primarily constructed based on boiling points and do not address the molecular level. To simultaneously address molecular-level information, molecular boiling point mapping is required. However, molecular boiling point mapping only considers the boiling point dimension. Furthermore, due to the wide carbon number distribution, wide boiling point range, and regular, concentrated, and trace content of corrosion molecules, using this approach can easily introduce subtle errors in property calculations. This makes it difficult to accurately characterize the impact of interactions between corrosion molecules and other molecules on concentration predictions, hindering the subsequent description of the thermal cracking of corrosion molecules during atmospheric and vacuum distillation. Specifically, traditional virtual component property calculations rely primarily on empirical regression formulas, typically with inputs consisting solely of the crude oil's macroscopic density and distillation temperature range. The output is the virtual component content and corresponding properties within a specific distillation temperature range. Due to the low concentration of corrosion molecules, empirical regression formulas rarely use their concentration distribution as input. This significantly impacts the calculation of virtual component properties in distillation ranges with high corrosion molecule concentrations. Furthermore, traditional virtual component partitioning primarily provides component contents and properties that vary with boiling point. However, due to the low concentration of corrosion molecules, the proportion of the content mapped to each virtual component is even smaller. If corrosion molecules are mapped to corresponding virtual components using traditional partitioning methods and then separated and calculated, and then the concentration distribution is calculated through inverse mapping, the distribution of corrosion molecules will be affected by the presence of the main hydrocarbon components due to the fixed nature of the mapping and the subjective nature of the property calculation.

[0004] In the actual industrial production process, there are certain difficulties in achieving molecular-level simulation. One of the reasons is that it is impossible to characterize crude oil in real time and in detail. Traditional atmospheric and vacuum simulation technology usually relies on low-frequency daily crude oil testing and lacks robustness to imported crude oil testing data. This leads to large deviations between process calculation results and actual results, making it difficult to cope with the dynamic changes in crude oil and testing methods of enterprises. Specifically, the frequency of additions and modifications to crude oil property items and the frequency of testing during the production process often cannot keep up with crude oil changes. As the processing capacity of atmospheric and vacuum equipment increases, the properties of crude oil processed in the same batch or even on the same day may vary. If you want to monitor changes in crude oil properties in real time and simulate their impact on production, it will increase testing costs and make it difficult to ensure timeliness. Summary of the Invention

[0005] The present invention provides a molecular-level atmospheric and vacuum corrosion simulation method and device, which improves the accuracy and timeliness of the prediction of corrosion molecule distribution in various parts of an atmospheric and vacuum device.

[0006] In a first aspect, an embodiment of the present invention provides a molecular-level atmospheric and vacuum corrosion simulation method, comprising:

[0007] Based on a first mid-infrared spectrum of the current crude oil, obtaining first macroscopic property data of the current crude oil through a crude oil physical property analysis model;

[0008] Searching for second macroscopic property data matching the first macroscopic property data in a corrosion molecule library, and obtaining a first molecular concentration of the corrosion molecule in the current crude oil based on a second molecular concentration of a historical process oil sample corresponding to the second macroscopic property data; wherein the corrosion molecule library is constructed based on historical process oil samples, and each historical process oil sample corresponds to one second macroscopic property data;

[0009] Classifying the corrosive molecules in the current crude oil according to different species, classifying the corrosive molecules of each species according to one or more physical property data, obtaining a lumped aggregate to which each corrosive molecule belongs, and obtaining a virtual component of the corrosive molecules in the current crude oil based on the lumped aggregate and a first molecule concentration; wherein the physical property data is obtained from a physical property database constructed based on historical processed oil samples;

[0010] According to the preset DCS data and the virtual components of the corrosion molecules, the corrosion iterative simulation of the atmospheric and vacuum device is performed, and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device are output; wherein, the preset DCS data is obtained in real time by the DCS system for the entire process of the atmospheric and vacuum process.

[0011] The embodiments of the present invention obtain the first macroscopic property data of the current crude oil, providing a basis for oil sample matching for the subsequent acquisition of the first molecular concentration of the corrosion molecules; by obtaining the first molecular concentration of the corrosion molecules in the current crude oil, a data foundation is provided for the subsequent acquisition of the virtual components of the corrosion molecules; by obtaining the virtual components of the corrosion molecules, the characteristics and distribution of the corrosion molecules can be more carefully reflected, thereby improving the accuracy of the simulation; by running the corrosion iterative simulation, the molecular concentration of the virtual components can be adjusted in real time, thereby accurately predicting the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device. Compared with the existing technology, the present application can improve the accuracy and timeliness of the prediction of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device.

[0012] Furthermore, before obtaining the first macroscopic property data of the current crude oil through the crude oil physical property analysis model, the method further includes:

[0013] A first mid-infrared spectrum of the current crude oil is obtained; wherein the first mid-infrared spectrum is obtained by measuring an attenuated total reflection accessory using a mid-infrared spectrometer, and the attenuated total reflection accessory is provided with the current crude oil.

[0014] The embodiment of the present invention can provide a basis for subsequently obtaining first macroscopic property data of the current crude oil by obtaining the first mid-infrared spectrum of the current crude oil.

[0015] Furthermore, the crude oil physical property analysis model is constructed based on historical processed oil samples, specifically:

[0016] Obtaining a second mid-infrared spectrum of the historical processing oil sample; wherein the second mid-infrared spectrum is obtained by measuring an attenuated total reflectance accessory using a mid-infrared spectrometer, and the attenuated total reflectance accessory is provided with the historical processing oil sample;

[0017] The macroscopic property data corresponding to each preset interval in the second mid-infrared spectrum is determined, and a crude oil physical property analysis model is constructed based on the correlation between each preset interval and the macroscopic property data.

[0018] The embodiment of the present invention uses a mid-infrared spectrometer to better capture complex molecules and chemical bond vibrations, and is more suitable for crude oil property analysis; the model can be simplified by selecting the interval of the mid-infrared spectrum.

[0019] Furthermore, the first molecular concentration of the corrosive molecules in the current crude oil is obtained based on the second molecular concentration of the historical processed oil sample corresponding to the second macroscopic property data, specifically:

[0020] The first macroscopic property data is compared with the second macroscopic property data, and the second molecular concentration of the corrosive molecules in the historical processing oil sample is adjusted according to the error, and the adjusted second molecular concentration is used as the first molecular concentration of the corrosive molecules in the current crude oil.

[0021] The embodiment of the present invention can quickly match the first molecular concentration of the current crude oil corrosion molecules through the preset corrosion molecule library and the first macroscopic property data of the current crude oil, reduce the molecular detection frequency of the crude oil, and meet actual production needs.

[0022] Furthermore, the corrosion molecule library is constructed based on historical processing oil samples, specifically:

[0023] Obtaining a second molecular concentration of all corrosion molecules in a historical processing oil sample; wherein the second molecular concentration is obtained by using a crude oil molecular detection technology;

[0024] A corrosion molecule library is constructed based on the second macroscopic property data and the second molecular concentration of the historical processing oil sample.

[0025] By constructing a corrosion molecule library, the embodiment of the present invention can quickly match the first macroscopic property data and the first molecular concentration of the current crude oil, providing a data basis for subsequent crude oil corrosion medium molecular characterization.

[0026] Furthermore, the virtual components of the corrosion molecules in the current crude oil are obtained based on the lumped and first molecular concentrations, specifically:

[0027] Based on the first molecular concentration and the physical property mixing calculation rule, obtaining each of the lumped first physical property data;

[0028] A virtual component of the corrosion molecules in the current crude oil is obtained based on the first molecular concentration and the first physical property data.

[0029] The embodiment of the present invention divides the corrosion molecules by species, thereby ensuring the independence of the distribution of each individual type of corrosion molecules; by dividing the corrosion molecules of each species by one or more physical property data, corrosion molecules with different characteristics of the same type can be obtained, and the chemical reaction process in the atmospheric and vacuum process can be described through the conversion between each other.

[0030] Furthermore, the physical property database is constructed based on historical processing oil samples, specifically:

[0031] Obtaining second physical property data of all corrosion molecules in the historical processing oil sample; wherein the second physical property data is obtained by using the group physical property calculation rule in the group contribution method;

[0032] A physical property database is constructed based on the second physical property data.

[0033] The embodiment of the present invention can quickly obtain the physical property data of different types of corrosion molecules by constructing a physical property database, providing a data basis for subsequent calculation of the physical properties of virtual components.

[0034] Furthermore, the corrosion iterative simulation of the atmospheric and vacuum device is performed based on the preset DCS data and the virtual components of the corrosion molecules, and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device are output, specifically:

[0035] According to a preset calculation sequence, the virtual components of the corrosion molecules of the current iteration and the DCS data of the current iteration, an iterative corrosion simulation is performed on the atmospheric and vacuum device until the DCS data error reaches a preset range, and the DCS data of the current iteration and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device of the current iteration are output; wherein, in each iteration, the virtual components of the corrosion molecules are updated according to the DCS data error of the previous iteration; the DCS data error is calculated based on the DCS data of the current iteration and the preset DCS data; and the calculation sequence is determined according to the equipment flow of the atmospheric and vacuum process.

[0036] The embodiment of the present invention ensures that the calculation of each device does not depend on the device for which the calculation has not yet been completed, by using a preset calculation sequence, thereby reducing calculation errors and improving calculation efficiency. By continuously adjusting the virtual component concentration of crude oil based on DCS data errors, the simulation results can be ensured to be closer to actual production conditions.

[0037] Furthermore, the virtual component of the corrosion molecule is updated according to the DCS data error of the last iteration, including taking the molecular concentration of the virtual component as an independent variable and the DCS data error as a dependent variable, and obtaining the optimal value of the independent variable through the SQP optimization algorithm, specifically:

[0038] ;

[0039] in, is the DCS data error; For the The error weight corresponding to the DCS data; For the The error corresponding to the DCS data is as follows:

[0040] ;

[0041] in, For the Number error mapping equation; For the The molecular concentration of the virtual component.

[0042] The embodiment of the present invention minimizes the error between the simulation results and the actual operation data through the SQP optimization algorithm, obtains the optimal virtual component molecular concentration, and thus obtains more accurate simulation results.

[0043] In a second aspect, an embodiment of the present invention provides a molecular-level atmospheric and vacuum corrosion simulation device, comprising:

[0044] A crude oil quick review module, configured to obtain first macroscopic property data of the current crude oil through a crude oil physical property analysis model based on a first mid-infrared spectrum of the current crude oil;

[0045] a molecular analysis module configured to search a corrosion molecule library for second macroscopic property data that matches the first macroscopic property data, and obtain a first molecular concentration of the corrosion molecule in the current crude oil based on a second molecular concentration of a historical process oil sample corresponding to the second macroscopic property data; wherein the corrosion molecule library is constructed based on historical process oil samples, each historical process oil sample corresponding to one second macroscopic property data;

[0046] a component mapping module for classifying the corrosive molecules in the current crude oil according to different species, classifying the corrosive molecules of each species according to one or more physical property data, obtaining a lumped aggregate to which each corrosive molecule belongs, and obtaining a virtual component of the corrosive molecules in the current crude oil based on the lumped aggregate and a first molecule concentration; wherein the physical property data is obtained from a physical property database constructed based on historical processed oil samples;

[0047] The simulation operation module is used to perform corrosion iterative simulation on the atmospheric and vacuum device according to preset DCS data and the virtual components of the corrosion molecules, and output simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device.

[0048] The embodiment of the present invention obtains the first macroscopic property data of the current crude oil through the crude oil quick evaluation module, providing a basis for oil sample matching for the subsequent acquisition of the first molecular concentration of the corrosion molecules; obtains the first molecular concentration of the corrosion molecules in the current crude oil through the molecular analysis module, providing a data basis for the subsequent acquisition of the virtual components of the corrosion molecules; obtains the virtual components of the corrosion molecules through the component mapping module, which can more carefully reflect the characteristics and distribution of the corrosion molecules, thereby improving the accuracy of the simulation; and runs the corrosion iterative simulation through the simulation operation module, which can adjust the molecular concentration of the virtual components in real time, thereby accurately predicting the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device. Compared with the existing technology, the present application can improve the accuracy and timeliness of the prediction of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic flow chart of a molecular-level atmospheric and vacuum corrosion simulation method provided by an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of the process of iterative corrosion simulation provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of a flow chart of a device calculation algorithm provided in an embodiment of the present invention;

[0052] Figure 4 A schematic diagram of a flow chart of a calculation error propagation path provided by an embodiment of the present invention;

[0053] Figure 5 A graph showing the results of high-resolution mass spectrometry-sulfur compound detection of a historical processing oil sample provided by an embodiment of the present invention;

[0054] Figure 6 The mid-infrared spectrum of crude oil provided by the embodiment of the present invention;

[0055] Figure 7 A unit flow chart of a refinery provided in an embodiment of the present invention;

[0056] Figure 8 A topological network diagram of a refinery unit process provided by an embodiment of the present invention;

[0057] Figure 9 A comparison chart before and after virtual component adjustment provided by an embodiment of the present invention;

[0058] Figure 10A schematic structural diagram of a molecular-level atmospheric and vacuum corrosion simulation device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0060] Please refer to Figure 1 , a molecular-level atmospheric and vacuum corrosion simulation method provided by an embodiment of the present invention includes steps S101 to S104, which are described in detail as follows:

[0061] Step S101: obtaining first macroscopic property data of the current crude oil according to a crude oil physical property analysis model.

[0062] In this embodiment, the first macroscopic property data of the current crude oil is obtained according to the crude oil physical property analysis model, specifically:

[0063] Obtaining a first mid-infrared spectrum of the current crude oil; wherein the first mid-infrared spectrum is obtained by measuring an attenuated total reflectance accessory using a mid-infrared spectrometer, and the attenuated total reflectance accessory is provided with the current crude oil;

[0064] Based on the first mid-infrared spectrum, first macroscopic property data of the current crude oil is obtained through a crude oil physical property analysis model; optionally, the first macroscopic property data includes specific gravity, water content, sulfur content, residual carbon, acid value, basic nitrogen and viscosity, etc., as shown in Table 1.

[0065] Table 1 - Part of the first macroscopic property data and reference standard table

[0066]

[0067] This embodiment uses the crude oil physical property analysis model to accurately calculate the first macroscopic property data of crude oil in a short period of time.

[0068] In this embodiment, the crude oil physical property analysis model is constructed based on historical processed oil samples, specifically:

[0069] Obtaining a second mid-infrared spectrum of the historical processing oil sample; wherein the second mid-infrared spectrum is obtained by measuring an attenuated total reflectance accessory using a mid-infrared spectrometer, and the attenuated total reflectance accessory is provided with the historical processing oil sample;

[0070] Determine the macroscopic property data corresponding to each preset interval in the second mid-infrared spectrum, and construct a crude oil physical property analysis model based on the correlation between each preset interval and the macroscopic property data; optionally, interval selection uses a correlation coefficient method; optionally, model selection includes partial least squares, neural network, similarity matching or kernel partial least squares method, etc., and the optimal crude oil physical property analysis model is automatically established and selected through the developed optimization algorithm.

[0071] This embodiment uses a mid-infrared spectrometer to better capture complex molecules and chemical bond vibrations, and is more suitable for crude oil physical property analysis; the model can be simplified by selecting the interval of the mid-infrared spectrum.

[0072] Step S102 : obtaining a first molecular concentration of the corrosion molecules in the current crude oil according to a preset corrosion molecule library and the first macroscopic property data.

[0073] In this embodiment, the first molecular concentration of the corrosion molecules in the current crude oil is obtained based on the preset corrosion molecule library and the first macroscopic property data, specifically:

[0074] According to the first macroscopic property data, searching for second macroscopic property data matching the first macroscopic property data in a corrosion molecule library; wherein each second macroscopic property data corresponds to a historical processing oil sample; optionally, the matching method uses clustering and similarity matching;

[0075] The first macroscopic property data is compared with the second macroscopic property data, and the second molecular concentration of the corrosion molecules in the historical processing oil sample is adjusted according to the error, and the adjusted second molecular concentration is used as the first molecular concentration of the corrosion molecules in the current crude oil; optionally, the first macroscopic property data of the single crude oil before mixing is used for matching in the corrosion molecule library to obtain the second molecular concentration of the corrosion molecules in the single crude oil, and then the crude oil is mixed according to the crude oil mixing ratio to obtain the second molecular concentration of the corrosion molecules in the mixed crude oil; optionally, the second macroscopic property data corresponds to the mixed crude oil, and is calculated based on the second molecular concentration of the corrosion molecules in the mixed crude oil and a physical property calculation algorithm.

[0076] This embodiment uses the preset corrosion molecule library and the first macroscopic property data of the current crude oil to quickly match the first molecular concentration of the corrosion molecules in the current crude oil, reduce the molecular detection frequency of the crude oil, and meet actual production needs.

[0077] In this embodiment, the corrosion molecule library is constructed based on historical processing oil samples, specifically:

[0078] Obtaining a second molecular concentration of all corrosive molecules in a historical process oil sample; wherein the second molecular concentration is obtained by a crude oil molecule detection technology; optionally, the crude oil molecule detection technology includes four-component analysis, elemental analysis, high-temperature simulated distillation, gas chromatography, or high-resolution mass spectrometry combined with various derivatization methods; optionally, the corrosive molecules include sulfur (molecules such as elemental sulfur, hydrogen sulfide, disulfide, and mercaptans in active sulfur, and molecules such as sulfides and thiophenes in inactive sulfur), nitrogen (molecules such as nitrogen, pyridine, pyrrole, and its derivatives), and acid (molecules such as alkanoic acid and cycloalkanoic acid);

[0079] Based on the second macroscopic property data and the second molecular concentration of the historical processing oil sample, a corrosion molecule library is constructed; optionally, the construction of the corrosion molecule library includes entering the SOL structure character strings of all corrosion molecules in the historical processing oil sample, and disassembling the SOL structure character strings based on a proprietary algorithm, wherein the proprietary algorithm rules include disassembling the molecules into group types and numbers according to group recognition priorities; optionally, the corrosion molecule library contains more than 20,000 corrosion molecules.

[0080] This embodiment constructs a corrosion molecule library to quickly match the first macroscopic property data and the first molecular concentration of the current crude oil, providing a data basis for subsequent molecular characterization of crude oil corrosion media.

[0081] Step S103 : obtaining virtual components of the corrosive molecules in the current crude oil according to a preset physical property database and the first molecule concentration.

[0082] In this embodiment, the virtual components of the corrosive molecules in the current crude oil are obtained according to the preset physical property database and the first molecule concentration, specifically:

[0083] The corrosive molecules in the current crude oil are divided into different species, and the corrosive molecules of each species are clustered according to one or more physical property data to obtain the cluster to which each corrosive molecule belongs; wherein the physical property data is obtained from a physical property database; optionally, the species include saturated hydrocarbons, aromatic hydrocarbons, sulfur-containing molecules, nitrogen-containing molecules, and acid-containing molecules; optionally, the physical property data include boiling point, critical properties, and density; optionally, the number of clusters for the clustering division can be given in advance;

[0084] Based on the first molecular concentration and the physical property mixing calculation rule, obtaining first physical property data of each of the lumped elements; optionally, the first physical property data includes an eccentricity factor (w), critical properties (Tc, Pc, Vc, and Zc), a saturated vapor pressure (Psat), a gas-liquid phase heat capacity, and other physical property data required for process simulation;

[0085] A virtual component of the corrosion molecules in the current crude oil is obtained based on the first molecular concentration and the first physical property data.

[0086] For example, when each type of corrosive molecules is divided according to the boiling point, and the boiling point division method is unified, each lumped first physical property data can be modified. Specifically, the actual boiling point curve (V-Tb) of crude oil or D86 experimental data, as well as the measured average density value of crude oil, are used to cut the current crude oil into temperature segments to obtain the volume, density and boiling point of the crude oil component (i.e., virtual component) in each temperature segment; based on the boiling point (Tb) and density of each crude oil component, an empirical formula for calculating crude oil physical properties, such as the TWU method or the LK method, is used to obtain the first physical property data of each crude oil component, wherein the first physical property data includes the eccentricity factor (w), critical properties (Tc, Pc, Vc and Zc ), saturated vapor pressure (Psat), gas-liquid phase heat capacity and other physical property data required for process simulation; compare the first physical property data of the virtual lumped at the same temperature section with the first physical property data obtained by regression of the empirical formula, and evaluate the specific temperature interval that obviously deviates from the empirical calculation law. If the non-hydrocarbon content of this interval is significantly higher or lower than the adjacent temperature interval, the result can be considered reasonable. If the non-hydrocarbon content of this interval is close to the adjacent temperature interval, the first physical property data of the virtual lumped need to be corrected to a certain extent; finally, the lumped flow direction corresponding to each molecule, the content of each lumped, the first physical property data of each lumped, and the first molecule concentration of each lumped can be obtained.

[0087] This embodiment divides the corrosion molecules by species, which can ensure the independence of the distribution of each individual type of corrosion molecules; by dividing the corrosion molecules of each species by one or more physical property data, corrosion molecules with different characteristics of the same type can be obtained, and the chemical reaction process in the atmospheric and vacuum process can be described through the conversion between them.

[0088] In this embodiment, the physical property database is constructed based on historical processing oil samples, specifically:

[0089] Obtaining second physical property data of all corrosion molecules in the historical processing oil sample; wherein the second physical property data is obtained by using the group physical property calculation rule in the group contribution method; optionally, the second physical property data includes boiling point (Tb), eccentricity factor (w), gas-liquid phase heat capacity, molecular formation enthalpy, and molecular Gibbs free energy;

[0090] A physical property database is constructed based on the second physical property data; optionally, the physical property database contains more than 100,000 pieces of second physical property data.

[0091] This embodiment can quickly obtain the physical property data of different types of corrosion molecules by constructing a physical property database, providing a data basis for subsequent calculation of the physical properties of virtual components.

[0092] Step S104 , performing iterative corrosion simulation on the atmospheric and vacuum distillation device according to preset DCS data and the virtual components of the corrosion molecules, and outputting simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device.

[0093] In this embodiment, the corrosion iterative simulation of the atmospheric and vacuum device is performed based on the preset DCS data and the virtual components of the corrosion molecules, and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device are output, specifically:

[0094] According to the preset calculation sequence, the virtual components of the corrosion molecules of the current iteration and the DCS data of the current iteration, the corrosion iterative simulation of the constant and vacuum device is performed until the DCS data error reaches the preset range, and the DCS data of the current iteration and the simulation results of the corrosion molecule distribution in each part of the constant and vacuum device of the current iteration are output; wherein, in each iteration, the virtual components of the corrosion molecules are updated according to the DCS data error of the previous iteration; the DCS data error is calculated based on the DCS data of the current iteration and the preset DCS data; the calculation sequence is determined according to the equipment flow of the constant and vacuum process; optionally, the DCS data includes the temperature, pressure, and steady-state data of the flow control instrument of the inlet and outlet pipelines of each equipment, as well as the reflux ratio, production volume and other parameters used for corrosion iterative simulation and other non-simulation parameters used for error comparison; optionally, before the corrosion iterative simulation is run, the equipment data of the constant and vacuum process needs to be obtained, and the equipment The data includes the design parameters of the tower equipment, the number and type of plates of the plate tower, the properties, specifications and distribution of the packing of the packed tower, the feed position, the middle reflux extraction tray number, the reflux tray number of the middle reflux, the side stripping extraction tray number, the side stripping reflux tray number and the side extraction tray number, etc.; optionally, the equipment data is obtained from the equipment design drawings; optionally, the DCS data is obtained in real time by the DCS system for the entire process of atmospheric and vacuum distillation; optionally, when obtaining the DCS data and equipment data, it is necessary to organize the variable set required for the stream calculation according to the equipment-stream relationship table, and link the variable set required by the unit with the algorithm variable name, and finally form a large json file, in which the variable name and DCS data corresponding to the DCS data used for error comparison need to be added; optionally, before running the corrosion iterative simulation, it is necessary to receive two files: the physical property data global variable json and the process input parameter json.

[0095] Exemplarily, the calculation sequence is determined according to the equipment flow of the atmospheric and vacuum process, specifically: according to the actual atmospheric and vacuum process flow, a unit flow chart is drawn; when drawing the unit flow chart, it is necessary to simplify the continuous heat exchange equipment and continuous pressure equipment, and the top condenser, tank, splitter and middle reflux, and the side stripping module need to be incorporated into the corresponding tower as one device, and the packed vacuum tower needs to be converted into a plate tower, and the equal plate height can be calculated according to the packing type to obtain the number of tower plates; according to the unit flow chart and the inseparable subsystem identification method, an algorithm connection topology is constructed, and an equipment-stream relationship table is obtained, and the topology will not have completely unrelated equipment and streams; the equipment-stream relationship table records the stream affiliation through 1 and 0, which describes the system The mutual relationship between each unit of the system, as well as the input and output relationship of logistics, its nodes represent equipment units, and directed lines represent logistics information flows; the equipment-stream relationship table is converted into an equipment adjacent matrix, and the equipment adjacent matrix consists of 0 and 1, and the grid corresponding to 1 represents the flow from the horizontal coordinate equipment to the vertical coordinate equipment; matrix multiplication operation is performed on the equipment adjacent matrix, and after obtaining the new matrix, the equipment corresponding to the rows of columns with all 0s are extracted as the priority calculation equipment, and the rows and columns are removed to form a new matrix, and the calculation is repeated until the last equipment; the corrosion simulation is iterated once to obtain the name of the current calculation equipment and the number of equipment that needs to be calculated. According to the number, it is determined whether it is the last equipment, and the corresponding parameters of the equipment are extracted and passed into the equipment calculation algorithm for calculation to obtain the result.

[0096] For example, the specific process of corrosion iteration simulation is as follows: Figure 2 As shown, the success of the calculation is determined according to the status information. If it is successful and the device is not the last one, the calculation continues to the next device; if it fails and the device is not the last one, the simulation is directly exited with an error; if it is successful and the device is the last one, the DCS data error is judged; if it fails and the device is the last one, the simulation is exited with an error; after the DCS data error judgment, if the error is less than the threshold, the calculation is terminated to obtain the simulation result, otherwise the crude oil adjustment algorithm is called to obtain the new crude oil composition and repeat the above steps for verification; according to the device type, the corresponding device calculation algorithm is used to obtain the output data of each device. The internal operation logic of the device calculation algorithm is as follows Figure 3 As shown; wherein, the mechanism model of the equipment calculation algorithm includes a distillation mechanism model, a pressure transformation mechanism model, a heat exchange mechanism model, and a simple processing mechanism model according to different equipment units. All models are based on the mass conservation equation, the energy conservation equation, the phase equilibrium equation, and the component equation to form their own numerical solution methods for the nonlinear equation group; the solution algorithms include the self-created basic flash evaporation iterative solution algorithm, the distillation iterative solution algorithm, the flash evaporation initial value given algorithm, and the distillation initial value given method.

[0097] This embodiment uses a preset calculation sequence to ensure that the calculation of each device will not depend on the device that has not yet completed the calculation, thereby reducing calculation errors and improving calculation efficiency; by continuously adjusting the virtual component concentration of crude oil through DCS data errors, it can ensure that the simulation results are closer to actual production conditions.

[0098] In this embodiment, the virtual component of the corrosion molecule is updated based on the DCS data error of the previous iteration, including taking the molecular concentration of the virtual component as the independent variable and the DCS data error as the dependent variable, and obtaining the optimal value of the independent variable based on the error minimization principle through mathematical iteration. This embodiment believes that the propagation path of the crude oil error affecting the final calculation error is mainly as follows: Figure 4 shown.

[0099] For example, assume that each computation layer propagates a nonlinear system of equations , whose variables are the results of the previous layer. In the subsequent calculation, the error mapping equation of the DCS data is nested according to the level to which the DCS data belongs, and finally a complex nonlinear equation group is formed. , the calculation formula is as follows:

[0100] ;

[0101] in, is the DCS data error; For the The error weight corresponding to the DCS data is specified by the user; For the The error corresponding to the DCS data is as follows:

[0102] .

[0103] In the corrosion iterative simulation process, the physical property calculation layer is the most basic layer. Since the strict calculation of physical properties uses highly nonlinear mathematical calculations such as square root and logarithm, the error equation should include square root and logarithm terms. Since the input range of the logarithm term is greater than 0 and does not include 0, it is directly replaced by the exponential term. In addition, the quadratic term, the linear term and the constant term are added. The error mapping equation of the physical property calculation layer is: The calculation formula is as follows:

[0104] ;

[0105] in, are the weights of each constituent equation; and is temperature and pressure; is the weight of temperature and pressure. When the input is crude oil temperature and pressure, the weight is 0; and is the enthalpy and equilibrium constant.

[0106] In the flash calculation layer and the unit calculation layer, since the core principle is the MESH equation (material balance, energy balance and normalization and energy balance equation), that is, input equals output, it can be considered as a linear equation group. Therefore, the error mapping equation is also simplified to a linear equation group in this embodiment. The error mapping equation of the flash calculation layer is The calculation formula is as follows:

[0107] ;

[0108] Unit calculation layer error mapping equation The calculation formula is as follows:

[0109] ;

[0110] in, is the molar flow rate.

[0111] Outermost error mapping equation The calculation formula is as follows:

[0112] .

[0113] for , the specific parameters of its nonlinear equations, such as is unknown, so the SQP optimization algorithm is used, that is, the error feedback method is constructed to calculate the error by continuously adjusting the value of the unknown variable. , by reducing and The difference between the two obtains the optimal nonlinear equation parameters; by setting the error weight, the calculation is obtained ; Obtained by SQP optimization algorithm Minimum concentration of virtual component molecules.

[0114] Optionally, in this embodiment, the strict mechanism model is combined with the basic algorithm to construct an algorithm for calculating corrosion parameters for streams based on the specific behavior and actual needs of corrosion, thereby improving the universality and accuracy of the corrosion model and providing corrosion results.

[0115] This embodiment minimizes the error between the simulation results and the actual operation data through the SQP optimization algorithm, obtains the optimal virtual component molecular concentration, and thus obtains more accurate simulation results.

[0116] In order to better demonstrate the beneficial effects of the present invention, this embodiment provides a specific example, taking a refinery in Shandong as the analysis object, to demonstrate the specific implementation process of the molecular-level atmospheric and vacuum corrosion simulation method.

[0117] In this specific embodiment, a corrosion molecule library and a physical property database are constructed based on historical processing oil samples, and a high-resolution mass spectrometry is performed on a single historical processing oil sample to obtain the second molecule concentration of each historical processing oil sample, and the simulated distillation results are recorded; the high-resolution mass spectrum is processed as follows Figure 5 As shown, the second macroscopic property data of the historical processing oil samples are shown in Table 2, and the results of the historical processing oil samples based on the group contribution method are shown in Table 3.

[0118] Table 2 - Second macroscopic properties data of some historical oil samples

[0119]

[0120] Table 3 - Calculation results of partial group contribution method

[0121]

[0122] In this specific embodiment, the current crude oil mixing ratio is crude oil a: crude oil b = 30:70, and the mid-infrared spectrum scanning result of the current crude oil after mixing is as follows: Figure 6 As shown, the first macroscopic property data in Table 4 are obtained by the crude oil physical property analysis model, and the crude oil physical property analysis model is constructed based on the historical processed oil sample.

[0123] Table 4 - Data on the primary macroscopic properties of some current crude oils

[0124]

[0125] In this specific embodiment, the first macroscopic property data of crude oil a and crude oil b have been measured when the crude oil is purchased into the factory, as shown in Table 5 respectively; second macroscopic property data matching the first macroscopic property data are searched from the corrosion molecule library; wherein each second macroscopic property data corresponds to a historical processing oil sample; the second macroscopic property data of the historical processing oil sample is shown in Table 6; the second molecular concentration of the historical processing oil sample is shown in Table 7, and the second molecular concentration of the historical processing oil sample is mixed according to the mixing processing ratio to obtain the initial mixing second molecular concentration.

[0126] Table 5 - Primary macroscopic properties of some crude oils a and b

[0127]

[0128] Table 6 - Second macroscopic properties of some reference historical processing oil samples

[0129]

[0130] Table 7 - Second molecule concentration of some reference historical processing oil samples

[0131]

[0132] In this specific embodiment, the first macroscopic property data of crude oil a and crude oil b are compared with the second macroscopic property data of the reference historical processed oil sample, and the second molecular concentration of the corrosive molecules in the reference historical processed oil sample is adjusted according to the error, and the adjusted second molecular concentration is used as the first molecular concentration of the corrosive molecules in the current crude oil. The first molecular concentrations before and after the adjustment are shown in Table 8; the comparison results of the second macroscopic property data of the reference historical processed oil sample and the first macroscopic property data of the current crude oil are shown in Table 9; virtual components are divided and mapped by combining boiling points of corrosive molecules of different species, and the first molecular concentration and first physical property data of each virtual component are calculated according to the molecular concentration and physical property mixing calculation rules, and finally the first molecular concentration and first physical property data of the virtual component are output as shown in Tables 10-11.

[0133] Table 8 - First molecule concentration table before and after some adjustments

[0134]

[0135] Table 9 - Comparison of the second macroscopic properties of some historical reference oil samples and the first macroscopic properties of current crude oil

[0136]

[0137] Table 10 - First molecular concentration and first physical property data of some virtual components

[0138]

[0139] Table 11 - First molecule concentration of some corrosion molecules

[0140]

[0141] In this specific embodiment, the equipment data of the refinery is obtained according to the equipment design drawings; the DCS data of the refinery is obtained according to the DCS system; and the actual process of the refinery is simplified to obtain the following Figure 7The unit flow chart is shown in FIG1 , wherein the packed tower VAC is mapped into 30 plates by equal plate height, and the number of CDU plates is rounded to 33 (including the top condenser) based on the full tower efficiency of 50%; the bit numbers of the DCS system are matched with the input parameters in the unit flow chart to form the results shown in Table 12; according to the requirements of the corrosion iteration simulation input variables, the algorithm variables of feed IN, STEAM, STEAM2, S11, S13, S15 and equipment B1, B2, B3, CDU, and VAC are sorted and integrated to form the stream setting table and device setting table shown in Tables 13-15; the preset DCS data for the crude oil adjustment algorithm are determined to form the results shown in Table 16.

[0142] Table 12 - DCS system bit number and process relationship and physical variable value table

[0143]

[0144] Table 13 - Stream setting variable value table

[0145]

[0146] Table 14 - Heat exchange and simple unit setting variable value table

[0147]

[0148] Table 15 - Tower unit setting variable value table

[0149]

[0150] Table 16 - Crude Oil Adjustment Preset DCS Data Table

[0151]

[0152] In this specific embodiment, the calculation algorithm of each device is configured according to the corresponding bit number, device data and virtual component, and the overall input stream of the process is configured according to the virtual component and steam data. The crude oil adjustment preset DCS data table is configured with the corresponding data, and finally two JSON format files are formed with the global physical property parameters and transmitted to the mechanism algorithm interface; the threshold of the crude oil adjustment algorithm is set to 5°C, that is, the absolute value of the absolute error between the simulated value and the actual value of the seven temperature variables is required to be within 5°C.

[0153] In this specific embodiment, construct Figure 8The algorithm scheduling topology network shown is converted into the device-stream relationship table shown in Table 17, and the device-stream relationship table is converted into a json file and passed into the mechanism algorithm interface; the device-stream relationship table is input into the outermost calculation sequence algorithm to generate the adjacent matrix shown in Table 18. After matrix operation, the calculation order of the corrosion iterative simulation is obtained as IN, STEAM, S11, S13, S15, STEAM2, B1, B2, B3, CDU, VAC.

[0154] Table 17 - Equipment-flow relationship table

[0155]

[0156] Table 18 - Equipment-Stream Adjacency Matrix

[0157]

[0158] In this specific embodiment, according to the calculation sequence, the flow stream algorithm is called to calculate the flow stream data of IN, STEAM, S11, S13, S15, and STEAM2; according to the calculation sequence, the equipment calculation algorithm is run on B1, B2, B3, CDU, and VAC in sequence, specifically:

[0159] The B1 equipment data is separated and combined with the virtual component and IN stream data to form the input parameters of B1. The scheduling algorithm calls the heat exchange equipment algorithm based on the equipment type to process the input parameters. The flash evaporation algorithm is called to obtain key parameters of the outlet stream IN2, such as temperature, pressure, flow rate, and molecular composition. The stream algorithm is called to calculate and obtain specific data of the outlet stream IN2, such as temperature, pressure, density, enthalpy, and volume.

[0160] The B2 device data is separated and combined with the virtual component and B1 outlet stream IN2 data to form the input parameters for B2. The scheduling algorithm calls the simple device algorithm based on the device type to process the input parameters. The flash evaporation calculation is used to obtain the temperature, pressure, and load conditions in the separation tank under the specified conditions. Based on the gas-liquid phase relationship, the fluid in the tank is phase-separated and the key parameters of each phase, such as temperature, pressure, flow rate, and molecular composition, are calculated. The stream calculation algorithm is called to obtain the specific data of the outlet streams IN5 and IN3.

[0161] The B3 equipment data is separated and combined with the virtual component and B2 outlet stream IN3 data to form the input parameters for B3. The scheduling algorithm calls the heat exchanger equipment algorithm based on the unit type to process the input parameters. The flash evaporation algorithm is called to obtain the key parameters of the outlet stream IN4, such as temperature, pressure, flow rate, and molecular composition. The stream calculation algorithm is called to obtain the specific data of the outlet stream IN4.

[0162] The CDU equipment data, outlet stream IN5 data, outlet stream IN4 data, virtual components, main tower steam stream STEAM data, and side tower steam stream S11 and other input data are used to form the input parameters of the CDU; the scheduling algorithm calls the distillation equipment algorithm according to the unit type to perform distribution calculation. The specific calculation process is as follows: calling the initial value calculation algorithm to obtain the CDU temperature, pressure, flow, molecular composition, and initial values ​​of the plate distribution; passing the initial plate distribution value and input parameters to the solution algorithm for iteration until the square error of the plate enthalpy calculated after two iterations is less than 0.0002; calling the result sorting and output algorithm to form the CDU output data, including equipment results such as the atmospheric tower main tower distribution, stripping tower distribution, mid-section reflux results, and condenser results, as well as product stream V, LIQUID, SP1, SP2, SP3, and AR data. The CDU corrosion molecule distribution is shown in the corresponding data before adjustment in Table 19.

[0163] The VAC equipment data, product stream AR data, product stream VAC data, virtual components, and steam stream STEAM2 data are used as VAC input parameters. The scheduling algorithm calls the distillation equipment algorithm based on the unit type to perform distribution calculations. The specific calculation process is the same as that of the CDU. Ultimately, the mid-section reflux results and tower distribution results, as well as product stream SIDE1, SIDE2, SIDE3, SIDE4, VR, and VP data are obtained. The VAC corrosion molecule distribution is shown in the corresponding data before adjustment in Table 20. Since the VAC is the last device and the calculation is successful, this corrosion iteration simulation ends and all results are summarized into a json file.

[0164] Calculate the relevant DCS data error. The error results are shown in Table 21. The average absolute error is greater than the threshold. The crude oil adjustment algorithm is called to obtain the following: Figure 9 Dummy components shown after adjustment;

[0165] The next corrosion iteration simulation is repeated to obtain the CDU and VAC corrosion molecule distribution data as shown in the adjusted corresponding data in Tables 19 and 20, and the error value as shown in the adjusted corresponding data in Table 21; if the error meets the requirements, the corrosion iteration simulation is terminated.

[0166] Table 19 - CDU Corrosion Molecular Distribution Data

[0167]

[0168] Table 20-VAC Corrosion Molecular Distribution Data

[0169]

[0170] Table 21 - DCS data error results for crude oil adjustment

[0171]

[0172] In this specific embodiment, based on the corrosion iterative simulation results, the corrosion dew point calculations were performed on the CDU and VAC tower tops, tower bottoms, and side products, and the results shown in Table 22 were obtained (reference pressure was 101.325 kPa).

[0173] Table 22 - CDU and VAC corrosion dew point calculation results

[0174]

[0175] This embodiment obtains the first macroscopic property data of the current crude oil, providing a basis for oil sample matching for the subsequent acquisition of the first molecular concentration of the corrosion molecules. By obtaining the first molecular concentration of the corrosion molecules in the current crude oil, a data foundation is provided for the subsequent acquisition of the virtual components of the corrosion molecules. By obtaining the virtual components of the corrosion molecules, the characteristics and distribution of the corrosion molecules can be more carefully reflected, thereby improving the accuracy of the simulation. By presetting DCS data and the virtual components of the corrosion molecules, the molecular concentration of the virtual components can be adjusted in real time, thereby accurately predicting the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation unit. Compared with the existing technology, this application can improve the accuracy and timeliness of the prediction of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation unit.

[0176] Please refer to Figure 10 , a molecular-level atmospheric and vacuum corrosion simulation device provided by an embodiment of the present invention, includes a crude oil quick evaluation module 1001, a molecular analysis module 1002, a component mapping module 1003, and a simulation operation module 1004, which are described in detail as follows:

[0177] The crude oil quick review module 1001 is used to obtain the first macroscopic property data of the current crude oil according to the crude oil physical property analysis model;

[0178] The molecular analysis module 1002 is configured to obtain a first molecular concentration of the corrosion molecules in the current crude oil based on a preset corrosion molecule library and the first macroscopic property data;

[0179] A component mapping module 1003 is configured to obtain virtual components of the corrosive molecules in the current crude oil based on a preset physical property database and the first molecule concentration;

[0180] The simulation operation module 1004 is used to perform corrosion iterative simulation on the atmospheric and vacuum distillation device according to preset DCS data and the virtual components of the corrosion molecules, and output simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device.

[0181] In this embodiment, the crude oil quick review module 1001 includes a spectrum scanning unit and a spectrum analysis unit, specifically:

[0182] a spectrum scanning unit, configured to obtain a first mid-infrared spectrum of the current crude oil; wherein the first mid-infrared spectrum is obtained by measuring an attenuated total reflectance accessory using a mid-infrared spectrometer, and the attenuated total reflectance accessory is provided with the current crude oil;

[0183] The spectrum analysis unit is used to obtain the first macroscopic property data of the current crude oil through the crude oil physical property analysis model based on the first mid-infrared spectrum.

[0184] This embodiment uses the crude oil physical property analysis model to accurately calculate the first macroscopic property data of crude oil in a short period of time.

[0185] In this embodiment, the molecular analysis module 1002 includes a crude oil matching unit and an error adjustment unit, specifically:

[0186] a crude oil matching unit, which searches, based on the first macroscopic property data, a corrosion molecule library for second macroscopic property data that matches the first macroscopic property data; wherein each second macroscopic property data corresponds to a historical processing oil sample;

[0187] An error adjustment unit is used to compare the first macroscopic property data with the second macroscopic property data, and adjust the second molecular concentration of the corrosive molecules in the historical processing oil sample according to the error, and use the adjusted second molecular concentration as the first molecular concentration of the corrosive molecules in the current crude oil.

[0188] This embodiment uses the preset corrosion molecule library and the first macroscopic property data of the current crude oil to quickly match the first molecular concentration of the corrosion molecules in the current crude oil, reduce the molecular detection frequency of the crude oil, and meet actual production needs.

[0189] In this embodiment, the component mapping module 1003 includes a physical property division unit and a physical property calculation unit, specifically:

[0190] A physical property classification unit is used to classify the corrosive molecules in the current crude oil into different species, and classify the corrosive molecules of each species according to one or more physical property data to obtain the lumps to which each corrosive molecule belongs; wherein the physical property data is obtained from the physical property database;

[0191] a physical property calculation unit, configured to obtain the first physical property data of each of the lumped elements based on the first molecular concentration and a physical property mixing calculation rule;

[0192] A component construction unit is used to obtain a virtual component of the corrosion molecules in the current crude oil based on the first molecular concentration and the first physical property data.

[0193] This embodiment divides the corrosion molecules by species, which can ensure the independence of the distribution of each individual type of corrosion molecules; by dividing the corrosion molecules of each species by one or more physical property data, corrosion molecules with different characteristics of the same type can be obtained, and the chemical reaction process in the atmospheric and vacuum process can be described through the conversion between them.

[0194] In this embodiment, the simulation operation module 1004 includes a corrosion iteration unit, specifically:

[0195] The corrosion iteration unit is used to perform corrosion iterative simulation on the atmospheric and vacuum distillation device according to a preset calculation sequence, the virtual components of the corrosion molecules of the current iteration and the DCS data of the current iteration until the DCS data error reaches a preset range, and output the DCS data of the current iteration and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device of the current iteration; wherein, in each iteration, the virtual components of the corrosion molecules are updated according to the DCS data error of the previous iteration; the DCS data error is calculated based on the DCS data of the current iteration and preset DCS data; and the calculation sequence is determined according to the equipment flow of the atmospheric and vacuum distillation process.

[0196] This embodiment uses a preset calculation sequence to ensure that the calculation of each device will not depend on the device that has not yet completed the calculation, thereby reducing calculation errors and improving calculation efficiency; by continuously adjusting the virtual component concentration of crude oil through DCS data errors, it can ensure that the simulation results are closer to actual production conditions.

[0197] The molecular-level atmospheric and vacuum corrosion simulation device described above can implement the molecular-level atmospheric and vacuum corrosion simulation method of the above-mentioned method embodiment. The optional options in the above-mentioned method embodiment also apply to this embodiment and will not be described in detail here. The remaining contents of the embodiments of this application can refer to the contents of the above-mentioned method embodiment and will not be repeated in this embodiment.

[0198] This embodiment obtains the first macroscopic property data of the current crude oil through the crude oil quick evaluation module, providing a basis for oil sample matching for the subsequent acquisition of the first molecular concentration of the corrosion molecules; obtains the first molecular concentration of the corrosion molecules in the current crude oil through the molecular analysis module, providing a data basis for the subsequent acquisition of the virtual components of the corrosion molecules; obtains the virtual components of the corrosion molecules through the component mapping module, which can more carefully reflect the characteristics and distribution of the corrosion molecules, thereby improving the accuracy of the simulation; runs the corrosion iterative simulation through the simulation operation module, which can adjust the molecular concentration of the virtual components in real time, thereby accurately predicting the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device. Compared with the existing technology, this application can improve the accuracy and timeliness of the prediction of the distribution of corrosion molecules in various parts of the atmospheric and vacuum distillation device.

[0199] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A molecular-level atmospheric and vacuum corrosion simulation method, characterized in that: Applicable to atmospheric and vacuum devices, including: Based on a first mid-infrared spectrum of the current crude oil, obtaining first macroscopic property data of the current crude oil through a crude oil physical property analysis model; Searching for second macroscopic property data matching the first macroscopic property data in a corrosion molecule library, and obtaining a first molecular concentration of the corrosion molecule in the current crude oil based on a second molecular concentration of a historical process oil sample corresponding to the second macroscopic property data; wherein the corrosion molecule library is constructed based on historical process oil samples, and each historical process oil sample corresponds to one second macroscopic property data; Classifying the corrosive molecules in the current crude oil according to different species, classifying the corrosive molecules of each species according to one or more physical property data, obtaining a lumped aggregate to which each corrosive molecule belongs, and obtaining a virtual component of the corrosive molecules in the current crude oil based on the lumped aggregate and a first molecule concentration; wherein the physical property data is obtained from a physical property database constructed based on historical processed oil samples; According to the preset DCS data and the virtual components of the corrosion molecules, the corrosion iterative simulation of the atmospheric and vacuum device is performed, and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device are output; wherein, the preset DCS data is obtained in real time by the DCS system for the entire process of the atmospheric and vacuum process.

2. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: Before obtaining the first macroscopic property data of the current crude oil through the crude oil physical property analysis model, the method further includes: A first mid-infrared spectrum of the current crude oil is obtained; wherein the first mid-infrared spectrum is obtained by measuring an attenuated total reflection accessory using a mid-infrared spectrometer, and the attenuated total reflection accessory is provided with the current crude oil.

3. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: The crude oil physical property analysis model is constructed based on historical processed oil samples, specifically: Obtaining a second mid-infrared spectrum of the historical processing oil sample; wherein the second mid-infrared spectrum is obtained by measuring an attenuated total reflectance accessory using a mid-infrared spectrometer, and the attenuated total reflectance accessory is provided with the historical processing oil sample; The macroscopic property data corresponding to each preset interval in the second mid-infrared spectrum is determined, and a crude oil physical property analysis model is constructed based on the correlation between each preset interval and the macroscopic property data.

4. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: The first molecular concentration of the corrosive molecules in the current crude oil is obtained based on the second molecular concentration of the historical processed oil sample corresponding to the second macroscopic property data, specifically: The first macroscopic property data is compared with the second macroscopic property data, and the second molecular concentration of the corrosive molecules in the historical processing oil sample is adjusted according to the error, and the adjusted second molecular concentration is used as the first molecular concentration of the corrosive molecules in the current crude oil.

5. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: The corrosion molecule library is constructed based on historical processing oil samples, specifically: Obtaining a second molecular concentration of all corrosion molecules in a historical processing oil sample; wherein the second molecular concentration is obtained by using a crude oil molecular detection technology; A corrosion molecule library is constructed based on the second macroscopic property data and the second molecular concentration of the historical processing oil sample.

6. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: The step of obtaining the virtual components of the corrosion molecules in the current crude oil based on the lumped and first molecular concentrations is specifically as follows: Based on the first molecular concentration and the physical property mixing calculation rule, obtaining each of the lumped first physical property data; A virtual component of the corrosion molecules in the current crude oil is obtained based on the first molecular concentration and the first physical property data.

7. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: The physical property database is constructed based on historical processing oil samples, specifically: Obtaining second physical property data of all corrosion molecules in the historical processing oil sample; wherein the second physical property data is obtained by using the group physical property calculation rule in the group contribution method; A physical property database is constructed based on the second physical property data.

8. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 1, wherein: The method of performing iterative corrosion simulation on the atmospheric and vacuum device according to the preset DCS data and the virtual components of the corrosion molecules and outputting simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device is as follows: According to a preset calculation sequence, the virtual components of the corrosion molecules of the current iteration and the DCS data of the current iteration, an iterative corrosion simulation is performed on the atmospheric and vacuum device until the DCS data error reaches a preset range, and the DCS data of the current iteration and the simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device of the current iteration are output; wherein, in each iteration, the virtual components of the corrosion molecules are updated according to the DCS data error of the previous iteration; the DCS data error is calculated based on the DCS data of the current iteration and the preset DCS data; and the calculation sequence is determined according to the equipment flow of the atmospheric and vacuum process.

9. The molecular-level atmospheric and vacuum corrosion simulation method according to claim 8, characterized in that: The updating of the virtual component of the corrosion molecule according to the DCS data error of the last iteration includes taking the molecular concentration of the virtual component as an independent variable and the DCS data error as a dependent variable, and obtaining the optimal value of the independent variable through the SQP optimization algorithm, specifically: ； in, is the DCS data error; For the The error weight corresponding to the DCS data; For the The error corresponding to the DCS data is as follows: ； in, For the Number error mapping equation; For the The molecular concentration of the virtual component.

10. A molecular-level atmospheric and vacuum corrosion simulation device, characterized in that: Applicable to atmospheric and vacuum devices, including: A crude oil quick review module, configured to obtain first macroscopic property data of the current crude oil through a crude oil physical property analysis model based on a first mid-infrared spectrum of the current crude oil; a molecular analysis module configured to search a corrosion molecule library for second macroscopic property data that matches the first macroscopic property data, and obtain a first molecular concentration of the corrosion molecule in the current crude oil based on a second molecular concentration of a historical process oil sample corresponding to the second macroscopic property data; wherein the corrosion molecule library is constructed based on historical process oil samples, each historical process oil sample corresponding to one second macroscopic property data; a component mapping module for classifying the corrosive molecules in the current crude oil according to different species, classifying the corrosive molecules of each species according to one or more physical property data, obtaining a lumped aggregate to which each corrosive molecule belongs, and obtaining a virtual component of the corrosive molecules in the current crude oil based on the lumped aggregate and a first molecule concentration; wherein the physical property data is obtained from a physical property database constructed based on historical processed oil samples; The simulation operation module is used to perform corrosion iterative simulation on the atmospheric and vacuum device according to preset DCS data and the virtual components of the corrosion molecules, and output simulation results of the distribution of corrosion molecules in various parts of the atmospheric and vacuum device.

Images