A method and model for rapid prediction of constant pressure side-line distillate yields and properties

Abstract

The application discloses a method for rapidly predicting the yield and properties of side cut oil of a constant pressure reduction device, characterized by estimating the initial value of a model separation factor according to the data of the constant pressure reduction device, giving an optimization solution range, calibrating the model separation factor by using actual production data, and predicting the yield of side cut oil of different crude oils after the constant pressure reduction processing based on the calibrated model. Compared with the high requirement of the traditional mechanism model calculation, the method establishes a simplified model and shortens the calculation time. By combining the virtual component division and the application of the distillation separation factor, the simplified calculation and rapid evaluation of the complex distillation process are realized. The method improves the accuracy and efficiency of the evaluation, and provides strong support for the production optimization and cost control of a refining enterprise.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of petroleum chemical industry, in particular to a method and model for quickly predicting the yield and properties of side-cut distillate oil of a crude oil distillation unit. TECHNICAL BACKGROUND

[0002] In the petroleum refining industry, the crude oil distillation unit is the leading unit for crude oil processing. The main role of this link is to effectively separate components with different boiling points in crude oil through the distillation process, and to obtain side-cut distillate oil for subsequent deep processing of petroleum, so the yield of side-cut distillate oil of the crude oil distillation unit affects the processing arrangement of the entire refinery. In the process of refinery planning and scheduling, people often want to quickly know the yield and properties of side-cut distillate oil of different crude oils after processing in the crude oil distillation unit of the plant, in order to guide the procurement of crude oil and arrange the processing plan in advance.

[0003] To obtain this information, a fractionation mechanism model is usually used for simulation and calculation, but the application process has the following problems. First, the calculation time of the mechanism model is generally long, which is difficult to meet the needs of quick decision-making. Second, the technology usually requires complex data input, and the initial value is required to be high. If the initial value is not set properly, the calculation result may be deviated or even failed.

[0004] In addition, there are some models using simple data regression, which can reduce the calculation complexity, but the regression result often lacks practical physical meaning and cannot correspond to the actual device, which makes it difficult to estimate the initial value and explain the result of these models. In the case of large noise in the calibration data, the resulting model is also difficult to generalize.

[0005] Based on the above situation, the present application proposes a simplified model of the crude oil distillation unit with practical physical meaning, which is used for quickly predicting the yield and properties of side-cut distillate oil, and can balance efficiency and accuracy, which has important value for refining production. SUMMARY

[0006] The present application discloses a method and model for quickly predicting the yield and properties of side-cut distillate oil of a crude oil distillation unit, which estimates the separation factor between each side line of the crude oil distillation unit, corrects the model coefficients by using the characteristic relationship between the raw material and the side-cut distillate oil, and obtains a side-cut yield and property prediction model suitable for a specific crude oil distillation unit.

[0007] The present application first discloses a method for quickly predicting the yield and properties of side-cut distillate oil of a crude oil distillation unit, which comprises the following steps:

[0008] 1. Based on the actual number of trays between each side line of the crude oil distillation unit, multiply the tray conversion factor to obtain the separation factor n i The initial value n i0The equivalent factors of the trays in different regions are shown in the following table:

[0009]

[0010]

[0011] The determination of the separation factor calibration solution range is ± 50% of the initial value.

[0012] 2. Collect the side cut oil yield and distillation range data of the atmospheric-vacuum distillation unit.

[0013] 3. Use spline interpolation to complete the data of the distillation curve of the side cut, and convert the distillation curve of each side cut oil into a TBP distillation curve;

[0014] 4. Based on the density and TBP distillation curve of the side cut oil, perform virtual component division; cut the full distillation range into j cut distillation ranges, and divide it into j virtual components; obtain the content of the side cut oil i virtual component j according to the temperature range of the divided virtual component; estimate the density of the virtual component using the 50% point temperature of the TBP distillation curve and the characteristic factor of the side cut oil i; and based on the TBP distillation curve, calculate the boiling point T j .

[0015] 5. Based on the yield of each side cut oil, accumulate the virtual component contents of all side cut oils to obtain the composition of the feedstock components, and further obtain the TBP distillation curve of the feedstock.

[0016] 6. Based on the yield of the side cut oil and the TBP distillation curve of the feedstock, calculate the cut temperature T c,i of the side cut oil i; and obtain the side cut oil yield prediction model using the following steps:

[0017] According to the overhead dry gas and i side cut oils in the atmospheric-vacuum distillation process, construct i sub-models, use one sub-model for the separation calculation of adjacent side lines, and calculate i sub-models from i to 1 step by step;

[0018] Based on the boiling point T j of the virtual component j and the cut temperature T c,i of the sub-model i, calculate the average relative volatility a i,j of the j virtual component in the sub-model:

[0019]

[0020] Based on the separation factor n i and a i,j , calculate the gas phase component x D,i,j and the liquid phase component x B,i,j of the component in the sub-model i:

[0021]

[0022] According to the material balance, the rising gas phase flow rate D of the sub-model i is calculated i and the produced liquid phase flow rate B i , the produced liquid phase flow rate B i is the flow rate of the side-cut distillate i:

[0023]

[0024] where F i is the feed of the sub-model i, F i = 0 for i = 1, and F i = D i-1 for i > 1, Z i,j is the content of component j in the feed of the sub-model i, and Z i,j = x D,i-1,j for i > 1.

[0025] The produced liquid phase flow rate Bi is the flow rate of the side-cut distillate i, and the flow rate of the side-cut distillate i divided by the total amount of the raw material is the yield of the side-cut.

[0026] 8. The TBP distillation curve of the side-cut distillate i is calculated based on the calculated properties of the oil product, where the 5% distillation temperature is denoted as T 5％,i , and the 95% distillation temperature is denoted as T 95％,i .

[0027] 9. The separation factor n i suitable for the device is obtained by optimization using an optimization algorithm to minimize the residual error err:

[0028]

[0029] where Y is the actual yield of the side-cut i, and T are the 5% and 95% distillation temperatures of the TBP distillation curve of the side-cut distillate i, respectively, w F and w T are the weight coefficients of the yield and temperature, respectively.

[0030] 10. For a new crude oil to be predicted, the above model is used to quickly predict the yield and properties of the side-cut distillate of the atmospheric and vacuum distillation unit based on the obtained separation factor n i and the cutting temperature T c,i .

[0031] The present invention also discloses a model for rapidly predicting the yield and properties of atmospheric and vacuum distillation side-stream oils. The model constructs i sub-models based on dry gas and i side-stream oils during atmospheric and vacuum distillation. The separation calculation of adjacent side-streams uses one sub-model, and the i sub-models are calculated step by step from i to 1.

[0032] Construct the following sub-model based on the boiling point T of component j. j and cutting temperature T c,i Calculate the virtual component j in this sub-model

[0033] The average relative volatility α in i,j :

[0034]

[0035] Construct the following sub-model based on the separation factor n i and α i,j Calculate the gas phase component x of sub-model i D,i,j and liquid phase group

[0036] x B,i,j :

[0037]

[0038] Using the following model, calculate the rising gas phase flow rate D of sub-model i. i and the produced liquid phase flow rate B i The output liquid flow rate B i The yield of side stream distillate i is the flow rate of side stream distillate i divided by the total feed volume.

[0039]

[0040] In the formula F i For the feed of sub-model i, for i=1, F i =0, for i>1, F i =D i-1 Z i,j Z represents the content of component j in the feed of sub-model i. For i > 1, Z i,j =x D,i-1,j ;

[0041] Based on liquid phase component x B,i,j The rules for calculating the mixing properties of oil products are used to calculate the properties of each side-stream distillate virtual component after mixing.

[0042] Beneficial effects:

[0043] The application discloses a method and a model for rapidly predicting properties of side-cut oil of a crude oil distillation unit, and compared with high requirements of calculation by using a traditional mechanism model, the method establishes a simplified model and shortens calculation time. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 Flow chart of the method and the model for rapidly predicting properties of side-cut oil of a crude oil distillation unit

[0045] Figure 2 Side-cut oil yield test results of the embodiment of the application

[0046] Figure 3 Naphtha side-cut oil test results of the embodiment of the application

[0047] Figure 4 Kerosene side-cut oil test results of the embodiment of the application

[0048] Figure 5 Heavy gas oil side-cut oil test results of the embodiment of the application DETAILED DESCRIPTION

[0049] The application will be further described below in combination with the drawings and specific embodiments, and the implementation effect of the method in predicting properties of side-cut oil of a crude oil distillation unit is illustrated by a specific operation process. The embodiment is implemented on the premise of the technical solution of the application, but the protection scope of the application is not limited to the following embodiment.

[0050] A domestic large-scale refining and chemical enterprise has a 10 million ton / year crude oil distillation unit, in order to rapidly predict the properties of side-cut oil obtained after processing a new crude oil by the unit, a prediction model suitable for predicting the properties of side-cut oil of the crude oil distillation unit is established by using the technical solution of the application, and a specific process is as shown in the following table. Figure 1

[0051] The unit can separate dry gas and seven side-cut oils, i.e. naphtha, kerosene, light gas oil, heavy gas oil, vacuum light oil, vacuum heavy oil and residual oil. The number of trays between the dry gas, the naphtha, the kerosene, the light gas oil and the heavy gas oil is 16, 12, 10 and 16 respectively, and the number of trays between the heavy gas oil, the vacuum light oil, the vacuum heavy oil and the residual oil is estimated to be 10, 8 and 8. Based on the tray conversion coefficient, the initial value n i of the separation factor n i0 is estimated as follows.

[0052] n i0 ​= [8, 7, 4, 6.4, 3, 2.4, 2.4]

[0053] Determination of separation factor n i Calibration solution range:

[0054] [4, 3.5, 2, 3.2, 1.5, 1.2, 1.2] ≤ n i0 ≤ [12, 10.5, 6, 9.6, 4.5, 3.6, 3.6]

[0055] The side cut distillate yield data of the crude oil processed by the atmospheric-vacuum distillation unit were collected for a period of time, as shown in Table 1.

[0056] Table 1 Side cut yield of the crude oil processed by the atmospheric-vacuum distillation unit

[0057] side-cut oil volume yield V% dry gas 3.57 naphtha 11.54 kerosene 6.33 light gas oil 9.75 heavy gas oil 19.70 vacuum light oil 18.63 vacuum heavy oil 4.93 residual oil 25.55

[0058] The side cut distillate yield data of the crude oil processed by the atmospheric-vacuum distillation unit were collected for a period of time, as shown in Table 1.

[0059] Table 2 TBP distillation curve of the side cut distillate of the crude oil processed by the atmospheric-vacuum distillation unit (°C)

[0060]

[0061] The virtual component division was performed on each side cut, and the virtual components of kerosene were shown in Table 3.

[0062] Table 3 Virtual components of kerosene side cut

[0063]

[0064]

[0065] The components of each side cut were accumulated to obtain the complete virtual component composition of the raw material, and the cutting temperature T between each side cut was calculated according to the yield. c,i :

[0066] T c,i = [45, 140, 180, 240, 360, 500, 540]

[0067] Using the data in Table 1 and Table 2, the separation factor n of each sub-model corresponding to each side cut was calibrated one by one, and the separation factor n was calculated. i :

[0068] n i = [8.1, 6.7, 5.1, 5.8, 5.1, 3.5, 3.1]

[0069] Using the crude assay data in Table 4, based on the calibrated separation factor n i and the cutting temperature T c,i , the yield of each side cut of the crude oil and the distillation range data were predicted.

[0070] Table 4 TBP distillation curve data of the crude oil to be processed

[0071] volume yield / % distillation temperature / °C 0 -6.55 5 58.73 10 98.39 30 218.09 50 349.17 70 495.29 90 671.65 95 766.09 100 870.87

[0072] To verify the model, the yield and property data of the side cut fractions of the crude oil processed by the crude oil unit were collected for a period of time after the new crude oil was processed, and were compared with the predicted data.

[0073] Figure 2 The predicted data of the yield model of each side cut was compared with the actual data of the refinery, and it can be seen from Figure 2 that the absolute deviation of the predicted yield of each side cut by the model is within 2%, and the yield of the side cut fraction of the crude oil unit can be well predicted.

[0074] Figures 3 to 5 The predicted data of the distillation range of naphtha, kerosene and heavy gas oil, which are the most concerned by the refinery, were compared with the assay data, and it can be seen that the average deviation of the predicted distillation range is 1.92℃, 0.75℃ and 3.37℃, respectively, which are all within the repeatability range of the assay, and meet the production requirements of the refinery and chemical industry.

[0075] In addition, the time required for model calibration and prediction using the method is within 30 seconds, while using the mechanism model under the same computer hardware conditions, the time required for model calibration and prediction is more than 10 minutes.

[0076] In summary, the method and model can quickly and accurately predict the yield and properties of the side cut fractions of the crude oil unit, meet the production accuracy of the enterprise, and improve the real-time performance.

Claims

1. A method for rapidly predicting the yield and properties of atmospheric and vacuum distillate side stream oils, characterized in that... The initial value of the model separation factor is estimated based on data from the atmospheric and vacuum distillation unit, and the optimization solution range is given. The model separation factor is calibrated using actual production data. Based on the calibrated model, the yield of side stream distillate oil after atmospheric and vacuum distillation of different crude oils is predicted. The process includes the following steps: 1) Calculate the separation factor n based on the number of trays between each side stream of the atmospheric and vacuum distillation unit. i initial value n i0 And determine that the separation factor calibration solution range is ±50% of the initial value; 2) Collect data on the yield and distillation range of the side stream distillate from the atmospheric and vacuum distillation unit; 3) Use spline interpolation to complete the data of the distillation curves of the side streams, and convert the distillation curves of each side stream fraction into TBP distillation curves. 4) Based on the density and TBP distillation curve of the side-stream oil, virtual component classification is performed to obtain the content of virtual component j and the boiling point T of virtual component j in side-stream oil i. j ; 5) Based on the yield of each side stream oil, the virtual component content of all side stream oils is summed to obtain the feed composition, and then the TBP distillation curve of the feed is obtained. 6) Based on the yield of the side-stream oil and the TBP distillation curve of the feedstock, the cut temperature T of the side-stream oil i was calculated. c,i ; 7) Construct a sidestream distillate yield prediction model using the following steps: Based on the dry gas and i side-stream distillate during atmospheric and vacuum distillation, i sub-models are constructed. The separation calculation of adjacent side-streams uses one sub-model. The i sub-models are calculated step by step from i to 1. Construct the following sub-model based on the boiling point T of component j. j and cutting temperature T c,i Calculate the average relative volatility α of virtual component j in this sub-model. i,j : Construct the following sub-model based on the separation factor n i and α i,j Calculate the gas phase component x of sub-model i D,i,j and liquid phase component x B,i,j : Using the following model, calculate the rising gas phase flow rate D of sub-model i. i and the produced liquid phase flow rate B i The output liquid flow rate B i The flow rate of side stream distillate i is given by dividing the flow rate of side stream distillate i by the total feed volume, which gives the yield of the side stream. In the formula F i For the feed of sub-model i, for i=1, F i =0, for i>1, F i =D i-1 Z i,j Z represents the content of component j in the feed of sub-model i. For i > 1, Z i,j =x D,i-1,j ; 8) Based on liquid phase component x B,i,j According to the rules for calculating the mixing properties of oil products, the TBP distillation range curve of the virtual components of each side-stream distillate is calculated, where the 5% distillation temperature is denoted as T. 5％,i The 95% distillation temperature is denoted as T. 95％,i ; 9) Use an optimization algorithm to minimize the residual error (err) to obtain the separation factor n suitable for the device. i : in The actual yield of lateral line i. and The 5% and 95% distillation temperatures of the TBP distillation profile obtained from the analysis of the side-stream oil i are respectively. F and w T These are the weighting coefficients for yield and temperature, respectively. 10) For new crude oil to be predicted, using the above model, based on the optimized separation factor n i and cutting temperature T c,i It can quickly predict the yield and properties of distillate oil from the side stream of atmospheric and vacuum distillation units.

2. The method for rapidly predicting the yield and properties of atmospheric and vacuum distillate sidestream oils according to claim 1, characterized in that... Estimate the separation factor n using the tray reduction factor. i The initial values ​​and the tray conversion factors for different regions are shown in the table below:

3. The method for rapidly predicting the yield and properties of atmospheric and vacuum distillate sidestream oils according to claim 1, characterized in that... The following method was used to perform virtual component partitioning on sidestream oil i: 1) Divide the entire distillation range into j cut distillation ranges, divide them into j virtual components, and obtain the content of virtual component j of side stream oil i according to the temperature range; 2) Estimate the density of the virtual component based on the 50% point temperature of the TBP distillation curve and the characteristic factor of the side-stream distillate i; 3) Calculate the boiling point T of virtual component j by integral calculation based on the TBP distillation curve. j .

4. A model for rapidly predicting the yield and properties of atmospheric and vacuum distillate sidestream oils, characterized in that... The model is constructed based on dry gas and i side-stream distillate oils during atmospheric and vacuum distillation. i sub-models are constructed, and the separation calculation of adjacent side-streams is performed using one sub-model. The i sub-models are calculated step by step from i to 1. Construct the following sub-model based on the boiling point T of component j. j and cutting temperature T c,i Calculate the average relative volatility α of virtual component j in this sub-model. i,j : Construct the following sub-model based on the separation factor n i and α i,j Calculate the gas phase component x of sub-model i D,i,j and liquid phase component x B,i,j : Using the following model, calculate the rising gas phase flow rate D of sub-model i. i and the produced liquid phase flow rate B i The output liquid flow rate B i The flow rate of side stream distillate i is given by dividing the flow rate of side stream distillate i by the total feed volume, which gives the yield of the side stream. In the formula F i For the feed of sub-model i, for i=1, F i =0, for i>1, F i =D i-1 Z i,j Z represents the content of component j in the feed of sub-model i. For i > 1, Z i,j =x D,i-1,j ; Based on liquid phase component x B,i,j The rules for calculating the mixing properties of oil products are used to calculate the properties of each side-stream distillate virtual component after mixing.

Images