Optimization Study on Enhancing Deep-Cut Effect of the Vacuum Distillation Unit (VDU)

: The vacuum distillation unit (VDU) is the key unit to produce vacuum gas oil and vacuum residue, which has a very important impact on the downstream secondary processing units. The optimization of deep-cut vacuum distillation seeks to improve the yield of heavy vacuum gas oil (HVGO) and its dry point temperature, which is related to the economic beneﬁts of the reﬁnery. In this study, we ﬁrst established a simple model of a VDU by using the Aspen HYSYS Process simulation software. Then, we built a rigorous model with fast convergence by using the initial values obtained by the simple model. The rigorous model can accurately reﬂect the reﬁnery’s operation and can make predictions. Then, based on the rigorous model, we increased the ﬂash section temperature (FST) to 420 ◦ C and the steam ﬂow rate (SFR) of the stripping to 26 t/h. We eventually increased the yield of HVGO by 6.3 percentage points to 43.4%, while increasing its D86 95%-point temperature by 31.9 ◦ C to 570.9 ◦ C. In this way, the reﬁnery can effectively optimize the deep-cut vacuum distillation and obtain greater economic beneﬁts.


Introduction
The vacuum distillation unit (VDU) is the key component of the crude oil distillation process and the leading device of the refinery [1,2].It can extract wax oil from atmospheric residue and provide essential raw materials for secondary processing units (such as the catalytic cracking unit and catalytic hydrogenation unit) [3,4].Since the beginning of the 21st century, the development of the oil refining industry has changed from high-speed development to high-quality development [5,6].Deep-cut vacuum distillation technology plays a key role in the VDU [7].Its purpose is to improve the recovery rate of vacuum gas oil (VGO) and the D86-95% point temperature of VGO.The extraction rate of heavy vacuum gas oil (HVGO) is the key factor in determining the quality of the VDU.Therefore, improving its extraction rate is conducive to the economic benefits of the refinery.Therefore, studying the VDU, various feeds and chemical reactions is necessary [8].It requires us to establish a rigorous mechanism model consistent with reality [9], and through this model, we can accurately determine how to optimize the deep-cut vacuum distillation.
Many scholars have studied the modeling and optimization of vacuum distillation before.The most typical is the research of the American KBC company and Dutch Shell company.KBC [10] simulated the VDU through the Petro-SIM simulation software to increase the cutting point temperature of vacuum distillation to a higher temperature [11].They strictly controlled the vacuum furnace tube through direct-contact heat transfer to operate for more than half a year at a higher furnace outlet temperature below 425 • C [12].Shell used deep-flash high vacuum unit technology to design empty columns, which reduced the vacuum in the column and allowed the real boiling point cutting temperature to reach the specified temperature [13].Shell's deep-cut vacuum distillation technology can reduce the packing and reduce the pressure drop of the whole column.However, it is only applicable to the design of new vacuum column units.Zhang Long [14] also used the new structured packing method of the vacuum column to improve the pull-out rate by 3%.Jiang Bin [15] used a quench oil cycling structure to reduce residue pyrolysis and polymerization.They used Fluent for the simulation and confirmed that using a quench oil distribution pipe could effectively ensure the uniform temperature distribution of residue.In addition, Liu [16] and Li [17] optimized the deep-cut vacuum distillation from the perspective of the type and amount of activator.Although the extraction rate improved to a certain extent, the range was only within 2~3%.Wei Zhong [18] simulated the atmospheric and vacuum distillation unit with Pro/II software in 2000, calculated the operation data of the atmospheric and vacuum distillation unit under specific conditions and compared them with the production practice.It is considered that it is feasible to simulate the atmospheric and vacuum distillation unit with Pro/II software.However, the products in the first line of decompression are quite different from the actual products.Cheng [19] and Hou [20] established the steady-state simulation process of an atmospheric and vacuum distillation unit on the Aspen Plus software platform, with the calibration data as the main input data and the product control index as the main constraint.Through the adjustment of operating parameters such as operating pressure and feed temperature, the crude oil extraction rate can be increased by 6.8 percentage points to maximize the economic benefits.
However, few scholars have used Aspen HYSYS to model the VDU, and even if they have, there are many deviations from the actual refinery operating environment.This has a certain adverse effect on deep-cut vacuum distillation operation optimization.At the same time, some researchers' optimization methods are singular, some are too complex, and some economic investments are too large, which greatly reduces the effect of actual optimization.The direct simulation of the vacuum column has the disadvantages of difficult convergence and great difference from the actual situation.This paper presents a method from simple simulation to rigorous simulation.Furthermore, the simple model proposed in this paper divides the vacuum column into four absorption columns for simulation one by one.The simple model can quickly and accurately identify the performance parameters of the VDU, especially the preliminary research of transformation.The model can converge according to the specified input parameters and provide a reliable initial value for establishing a rigorous model.The rigorous model established by this method is easy to converge and has high accuracy.We can further provide a model basis for deep-cut vacuum distillation optimization through this rigorous model.
Finally, we rely on the rigorous model of vacuum distillation established by Aspen HYSYS to find two efficient and convenient methods for optimizing deep-cut vacuum distillation, which are increasing the flash section temperature (FST) of the vacuum column and increasing the steam flow rate (SFR) of the stripping.The simulation software shows that the yield of HVGO and the temperature of its D86 95%-point can be improved greatly by using these two methods simultaneously.Furthermore, within a certain control range, this method can safely, reliably and economically improve the production efficiency of the refinery.

Case Study
This paper takes the VDU of a refinery in China as an example for simulation research.The refinery process flow diagram is shown in Figure 1.The atmospheric residue (AR) from the bottom of the atmospheric distillation column is preheated to 411 • C by the vacuum heating furnace and then enters the vacuum column.At the same time, low-pressure steam is introduced into the column bottom to strip AR.The vacuum column of the refinery is divided into four zones from top to bottom: vacuum distillate oil (VDO) zone, light vacuum gas oil (LVGO) zone, HVGO zone and stripping and flash zone [21].There is a mid-cycle in each of the first three zones.The pre-furnace circulation oil is pumped out in the lower part of the HVGO zone back to the heating furnace for reheating.The off-gas discharged from the top of the vacuum column then enters the collection unit, and the vacuum residue (VR) at the bottom of the column enters the downstream secondary processing unit.
is a mid-cycle in each of the first three zones.The pre-furnace circulation oil is pumped out in the lower part of the HVGO zone back to the heating furnace for reheating.The offgas discharged from the top of the vacuum column then enters the collection unit, and the vacuum residue (VR) at the bottom of the column enters the downstream secondary processing unit.

Simulation Methods
Before the simulation, we needed to adjust the device data.During this period, we collected the refinery data in the study case for up to one month.Due to a lack of data or instrument failure, we further averaged or extrapolated or interpolated the data to supplement and improve them and to ensure the integrity of the modeling data [22].At the same time, we also consulted with the field device engineer about the consistency of the data, and ensured that each complete data set did not contain abnormal operation or major operation adjustment.We usually need to adjust the trial operation data to perfectly match the material balance and energy balance [23].For this, we retested the original data.Table 1 shows the data requirements of the VDU model.

Simulation Methods
Before the simulation, we needed to adjust the device data.During this period, we collected the refinery data in the study case for up to one month.Due to a lack of data or instrument failure, we further averaged or extrapolated or interpolated the data to supplement and improve them and to ensure the integrity of the modeling data [22].At the same time, we also consulted with the field device engineer about the consistency of the data, and ensured that each complete data set did not contain abnormal operation or major operation adjustment.We usually need to adjust the trial operation data to perfectly match the material balance and energy balance [23].For this, we retested the original data.Table 1 shows the data requirements of the VDU model.We used the V11 version of Aspen HYSYS software to carry out the simulation [24,25].Next, we started by establishing a simple model and then built a rigorous model.Table 2 shows the setting data information of each column.The establishment of the simple model facilitates the convergence of the flowsheet.It allows reliable initial values to be obtained, thus laying the foundation for establishing the rigorous model, which is a closer simulation to the actual working conditions and has a key role in the subsequent study of the operational optimization of the working conditions.The following describes the establishment of the simple model and rigorous model.

Establishment of Simple Model
The simple model is established by dividing the vacuum column into four independent absorber modules, shown in Figure 2. The functions of these four absorption columns are: (a) stripping and flash, (b) washing and HVGO output, (c) LVGO output, (d) VDO output.In the simple simulation process, it is necessary to define the components, properties and oil product data according to the obtained data [26].The simulation data obtained by PR and GS physical property methods for the atmospheric tower are in good agreement with the actual production data, while BK10 is more suitable for the simulation calculation of the vacuum column [27,28].Therefore, we chose the BK10 method for simulation.Then, we defined the feed, gas stream and stripping steam, added four absorption towers in turn and set the relevant operation parameters.Because the previous data parameters are set correctly and the model is divided into four regions for sequential simulation, the model is easy to converge.A special point to note is the simulation of the pre-furnace circulating oil, which comes from the over-vaporization oil and the entrained oil in the stripping section, and its circulating amount needs to be controlled within 0.2%~5% of the feed amount of AR [29].The simple model is more likely to converge to the design specification than the rigorous model because the units are relatively independent [30].Figure 3 shows the simple model process flow in Aspen HYSYS.
Processes 2022, 10, x FOR PEER REVIEW 5 of 15 towers in turn and set the relevant operation parameters.Because the previous data pa-135 rameters are set correctly and the model is divided into four regions for sequential simu-136 lation, the model is easy to converge.A special point to note is the simulation of the pre-137 furnace circulating oil, which comes from the over-vaporization oil and the entrained oil 138 in the stripping section, and its circulating amount needs to be controlled within 0.2%~5% 139 of the feed amount of AR [29].The simple model is more likely to converge to the design 140 specification than the rigorous model because the units are relatively independent [30].141 Figure 3 shows the simple model process flow in Aspen HYSYS.towers in turn and set the relevant operation parameters.Because the previous data parameters are set correctly and the model is divided into four regions for sequential simulation, the model is easy to converge.A special point to note is the simulation of the prefurnace circulating oil, which comes from the over-vaporization oil and the entrained oil in the stripping section, and its circulating amount needs to be controlled within 0.2%~5% of the feed amount of AR [29].The simple model is more likely to converge to the design specification than the rigorous model because the units are relatively independent [30].Figure 3 shows the simple model process flow in Aspen HYSYS.

Establishment of Rigorous Model
The main data involved in simple model and rigorous model simulation are shown in Table 3.As the parameters of the rigorous model for the VDU are similar to those of the

Establishment of Rigorous Model
The main data involved in simple model and rigorous model simulation are shown in Table 3.As the parameters of the rigorous model for the VDU are similar to those of the simple model for the VDU, they include the flow rate and temperature difference for each mid-stage cycle, the column top temperature and the flash temperature.Therefore, according to the convergence results of the simple model, the rigorous model can converge quickly.Figure 4 shows the rigorous model of the VDU in Aspen HYSYS, and Figure 5 shows the column environment diagram of the vacuum column sub-model.cording to the convergence results of the simple model, the rigorous model can converge quickly.Figure 4 shows the rigorous model of the VDU in Aspen HYSYS, and Figure 5 shows the column environment diagram of the vacuum column sub-model.Processes 2022, 10, 359 7 of 14

Analysis of the Main Operating Parameters
In the previous work, we successfully established the rigorous model of the VDU.Next, we need further to determine its compliance with the actual working conditions.First, we check whether the simulated values of the main operating parameters are within the design control range, as shown in Table 4.The data information in Table 4 shows that the simulation data for this rigorous model are within the control range of the process design parameters.

Formatting of Mathematical Components
In order to obtain accurate products, it is very important to represent the feed of AR accurately [31].We focus on two requirements: (a) many virtual components represent AR, (b) high-quality AR analysis data.According to the collected limited feed distillation data, we use the statistical function in Aspen to extrapolate the distillation curve [32,33] and compare it with the simulated feed distillation information, as shown in Figure 6 below.It can be seen from Figure 6 that the distillation data of the simulated synthetic feed 179 and the actual feed are consistent.This also further ensures the accuracy of the simulation 180 and provides a prerequisite for obtaining authentic products in the follow-up.It can be seen from Figure 6 that the distillation data of the simulated synthetic feed and the actual feed are consistent.This also further ensures the accuracy of the simulation and provides a prerequisite for obtaining authentic products in the follow-up.

Analysis of Vacuum Column Tray Temperature and Pressure
For this simulation, it is necessary to analyze the contrast column environment further to determine whether the model is accurate.The two important analysis and measurement parameters for the vacuum column are the tray temperature and the tray pressure [34].The following analysis compares the temperature and pressure of the vacuum column tray of the rigorous model with the actual column tray temperature and pressure, as shown in Figure 7 below.It can be seen from Figure 6 that the distillation data of the simulated synthetic feed 179 and the actual feed are consistent.This also further ensures the accuracy of the simulation 180 and provides a prerequisite for obtaining authentic products in the follow-up.181

Analysis of Vacuum Column Tray Temperature and Pressure
For this simulation, it is necessary to analyze the contrast column environment fur-183 ther to determine whether the model is accurate.The two important analysis and meas-184 urement parameters for the vacuum column are the tray temperature and the tray pres-185 sure [34].The following analysis compares the temperature and pressure of the vacuum 186 column tray of the rigorous model with the actual column tray temperature and pressure, 187 as shown in Figure 7 below.Figure 7 shows that the temperature and pressure curve of the vacuum column tray in the model is consistent with the temperature and pressure curve fitted by the actual tray, which further proves that the simulated column environment is consistent with the actual situation.For the eleventh tray shown in Figure 7a, there is a certain degree of temperature difference because the tower bottom temperature needs to be adjusted to achieve model convergence in this simulation.The temperature change at this place has little impact on the actual products, especially on the HVGO studied.

Analysis of the Yield of the Main Products
The main products of the VDU include VDO, LVGO, HVGO and VR.We compared the calibration results of the simulated output and the actual working conditions to draw a histogram.It can be seen from Figure 8 that the deviation between the simulated and the actual results is very small.The rigorous model simulates the process consistently with the actual product output.
Processes 2022, 10, x FOR PEER REVIEW 9 of 15 Figure 7 shows that the temperature and pressure curve of the vacuum column tray 192 in the model is consistent with the temperature and pressure curve fitted by the actual 193 tray, which further proves that the simulated column environment is consistent with the 194 actual situation.For the eleventh tray shown in Figure 7a, there is a certain degree of tem-195 perature difference because the tower bottom temperature needs to be adjusted to achieve 196 model convergence in this simulation.The temperature change at this place has little im-197 pact on the actual products, especially on the HVGO studied.

Analysis of the Yield of the Main Products
The main products of the VDU include VDO, LVGO, HVGO and VR.We compared 200 the calibration results of the simulated output and the actual working conditions to draw 201 a histogram.It can be seen from Figure 8 that the deviation between the simulated and 202 the actual results is very small.The rigorous model simulates the process consistently with 203 the actual product output.In addition to the analysis of the yield of the main products of the VDU, whether one 209 can obtain qualified products is an important assessment index of the process flow, which 210 requires oil evaluation [35].Therefore, we compared and analyzed the distillation curves 211

Analysis of the Main Product Distillation Data
In addition to the analysis of the yield of the main products of the VDU, whether one can obtain qualified products is an important assessment index of the process flow, which requires oil evaluation [35].Therefore, we compared and analyzed the distillation curves of the main products, as shown in Figure 9 below.The distillation curves of the VDO, HVGO and VR are in good agreement with each other, which further indicates that the simulations obtained a qualified product that matches the properties of the actual product.There are some differences in the distillation curve in Figure 9b.It can be seen that the LVGO within 90% of the cutting point are relatively consistent, while the fitting of the parts beyond 90% of the cutting point needs to be significantly improved.Since the difference has little influence on the properties of LVGO and has little relationship with the research of HVGO, it can be ignored.The above data comparison and analysis prove that the simulation results of the rigorous model derived from the simple model can accurately reflect the actual operation of the refinery.Next, the strict model can be further used to study the optimization of deepcut vacuum distillation.

Optimization Analysis of Deep-Cut Vacuum Distillation
With the increasing demand for heavy oil processing in refineries, vacuum distilla- The above data comparison and analysis prove that the simulation results of the rigorous model derived from the simple model can accurately reflect the actual operation of the refinery.Next, the strict model can be further used to study the optimization of deepcut vacuum distillation.

Optimization Analysis of Deep-Cut Vacuum Distillation
With the increasing demand for heavy oil processing in refineries, vacuum distillation simulation has become an important industrial application to optimize the deep-cut processing of heavy crude oil [36].Production efficiency can be increased by increasing the D86 95%-point temperature of HVGO above 565 • C to produce more VGO for downstream units (such as the catalytic cracking unit).In contrast, for the rigorous model obtained above, the D86 95%-point temperature of HVGO is only 539 • C, and the yield is only 36.1%.There is still much room for optimizing and improving the refinery's production process.In the optimization process, it should be noted that the residual carbon content of HVGO and the light component content of VR lower than 538 • C should not exceed 5%.Next, we carry out the deep-cut vacuum distillation transformation of the VDU from the two aspects of increasing the feed vaporization rate and optimizing the stripping steam.Then, we improve the cutting point temperature and yield of HVGO.
In this optimization process, we use Aspen's case study tool [37,38].The tool can observe the response of dependent variables when the process-independent variables change in steady-state simulation.For each independent variable, the user specifies the up/down line and step size.Aspen HYSYS will change the independent variables and calculate the dependent variables according to the upper and lower limits and steps specified by the user.The details of the independent variables set in this study are shown in Table 5 below.The higher the vaporization degree of the feed, the better the distillation effect of the vacuum column, and the more HVGO will be separated [39].In vacuum distillation, the two key operating parameters affecting the feed vaporization rate are the flash section temperature and flash section pressure.In order to improve the gasification rate, it is necessary to increase the flash section temperature or reduce the flash section pressure.For these two operating parameters, the former is easier to operate and change, so we decided to use Aspen HYSYS's case study tool [37,38] to determine the effect of FST on the HVGO yield and D86 95%-point temperature.The FST setting was changed from 400 • C to 420 • C, and Figure 10 shows the experimental results.
Figure 10 shows that as the FST increases, the mass yield of HVGO can reach up to 39.6%, and the D86 95%-point temperature reaches up to 554.8 • C. Therefore, the higher the temperature in the temperature control range of the flash section, the more conducive it is to improving the HVGO yield and D86 95%-point temperature.
essary to increase the flash section temperature or reduce the flash section pressure.For 254 these two operating parameters, the former is easier to operate and change, so we decided 255 to use Aspen HYSYS's case study tool [37,38] to determine the effect of FST on the HVGO 256 yield and D86 95%-point temperature.The FST setting was changed from 400 °C to 420 257 °C, and Figure 10 shows the experimental results.

Optimizing Stripping Steam Flow Rate
The stripping section of the vacuum column plays a very important role [40].When the stripping steam enters the bottom of the column, the AR can be stripped and dispersed into fine droplets and then enters the upper end of the column for distillation.The more fully stripped, the better the fractionation effect [41].For the work of the stripping section, the stripping SFR is the key factor affecting the stripping process.Next, we use the case study tool [37,38] to conduct an experimental study on the effect of stripping SFR on the HVGO yield and D86 95%-point temperature.The SFR setting was changed within 10-26 t/h, and Figure 11 shows the experimental results.Figure 10 shows that as the FST increases, the mass yield of HVGO can reach up to 39.6%, and the D86 95%-point temperature reaches up to 554.8 °C.Therefore, the higher the temperature in the temperature control range of the flash section, the more conducive it is to improving the HVGO yield and D86 95%-point temperature.

Optimizing Stripping Steam Flow Rate
The stripping section of the vacuum column plays a very important role [40].When the stripping steam enters the bottom of the column, the AR can be stripped and dispersed into fine droplets and then enters the upper end of the column for distillation.The more fully stripped, the better the fractionation effect [41].For the work of the stripping section, the stripping SFR is the key factor affecting the stripping process.Next, we use the case study tool [37,38] to conduct an experimental study on the effect of stripping SFR on the HVGO yield and D86 95%-point temperature.The SFR setting was changed within 10-26 t/h, and Figure 11 shows the experimental results.It can be seen from Figure 11 that with the increase in SFR, the maximum yield of HVGO can reach 40.1%, and the maximum D86 95%-point temperature can reach 554.6 °C.

Analysis and Study with Comprehensive Consideration of FST and SFR
In order to maximize the extraction rate of HVGO, we use the case study tool [38,39] to optimize the two key operating parameters of FST and SFR.The rigorous model is used to simulate the process of the device under different FST and different SFR.Our final results are shown in Figure 12.It can be seen from Figure 11 that with the increase in SFR, the maximum yield of HVGO can reach 40.1%, and the maximum D86 95%-point temperature can reach 554.6 • C.

Analysis and Study with Comprehensive Consideration of FST and SFR
In order to maximize the extraction rate of HVGO, we use the case study tool [38,39] to optimize the two key operating parameters of FST and SFR.The rigorous model is used to simulate the process of the device under different FST and different SFR.Our final results are shown in Figure 12.At the beginning of the study, the HVGO yield at the original operating point (FST = 407 °C; SFR = 11 t/h) was 36.1%, and the D86 95%-point temperature was 539.0 °C.It can be seen from Figure 12 that the optimal HVGO yield and D86 95%-point temperature can be achieved under the operating condition of the FST of 420 °C and the SFR of 26 t/h.The optimum yield was 43.4%, and the D86 95%-point temperature was 570.9 °C.Therefore, we can conclude that increasing the FST and SFR within the control range can greatly improve the yield of HVGO and the D86 95%-point temperature to achieve a good decompression and deep-cut effect.

Conclusions
Because the direct establishment of a rigorous model of the VDU is difficult and cannot be consistent with reality, a new simulation method from a simple model to a rigorous model is proposed in this article.We first establish a simple model of the VDU connected by four absorption columns using Aspen HYSYS software.We obtain the accurate key initial values through the simple model, and then the rigorous model of the VDU is established quickly and accurately.After verification and comparison in many aspects, we prove that the rigorous model can accurately reflect the actual operation of the refinery and has a certain prediction ability.We conduct an optimization study of the deep-cut vacuum distillation under the original working conditions through the rigorous model.We directly use the case study tool to analyze the operating parameters.Finally, by increasing the FST to 420 °C and the SFR to 26 t/h, the HVGO yield is increased by 6.3 percentage points to 43.4%.At the same time, the D86 95%-point temperature of HVGO is also increased by 31.9 °C to 570.9 °C.Our optimization research on the deep-cut vacuum distillation of the refinery indicates that very good results can be achieved only through operation adjustment with strong operability, high safety and good effect.This study can provide a better simulation process for chemical industry practitioners, and this study has a strong guiding role for the actual production of the refinery and can directly promote the economic benefits of the refinery.
Author Contributions: Conceptualization, Jin, Q.; methodology, Li, Z.; software, Li, Z.; validation, Wang, B., Li, Z.; formal analysis, Jin, Q.; investigation, Li, Z.; resources, Jin, Q.; data curation, Li, Z.; writing-original draft preparation, Li, Z.; writing-review and editing, Li, Z.; visualization, Wang, At the beginning of the study, the HVGO yield at the original operating point (FST = 407 • C; SFR = 11 t/h) was 36.1%, and the D86 95%-point temperature was 539.0 • C. It can be seen from Figure 12 that the optimal HVGO yield and D86 95%-point temperature can be achieved under the operating condition of the FST of 420 • C and the SFR of 26 t/h.The optimum yield was 43.4%, and the D86 95%-point temperature was 570.9 • C. Therefore, we can conclude that increasing the FST and SFR within the control range can greatly improve the yield of HVGO and the D86 95%-point temperature to achieve a good decompression and deep-cut effect.

Conclusions
Because the direct establishment of a rigorous model of the VDU is difficult and cannot be consistent with reality, a new simulation method from a simple model to a rigorous model is proposed in this article.We first establish a simple model of the VDU connected by four absorption columns using Aspen HYSYS software.We obtain the accurate key initial values through the simple model, and then the rigorous model of the VDU is established quickly and accurately.After verification and comparison in many aspects, we prove that the rigorous model can accurately reflect the actual operation of the refinery and has a certain prediction ability.We conduct an optimization study of the deep-cut vacuum distillation under the original working conditions through the rigorous model.We directly use the case study tool to analyze the operating parameters.Finally, by increasing the FST to 420 • C and the SFR to 26 t/h, the HVGO yield is increased by 6.3 percentage points to 43.4%.At the same time, the D86 95%-point temperature of HVGO is also increased by 31.9 • C to 570.9 • C. Our optimization research on the deep-cut vacuum distillation of the refinery indicates that very good results can be achieved only through operation adjustment with strong operability, high safety and good effect.This study can provide a better simulation process for chemical industry practitioners, and this study has a strong guiding role for the actual production of the refinery and can directly promote the economic benefits of the refinery.

146 2 . 2 . 2 .
Establishment of Rigorous Model 147 The main data involved in simple model and rigorous model simulation are shown 148 in Table 3.As the parameters of the rigorous model for the VDU are similar to those of the 149

Figure 3 .
Figure 3. Process flow diagram for the simple model of VDU in Aspen HYSYS.

Figure 3 .
Figure 3. Process flow diagram for the simple model of VDU in Aspen HYSYS.

Figure 4 .
Figure 4. Process flow diagram for the rigorous model of VDU in Aspen HYSYS.

204 205█Figure 8 .
Figure 8.Comparison of main product yields of rigorous model.

Figure 8 .
Figure 8.Comparison of main product yields of rigorous model.

◼Figure 9 .
Figure 9.Comparison of D86 or D1160 distillation data of the main products of the rigorous model.(a) D86 distillation data of VDO; (b) D1160 distillation data of LVGO; (c) D1160 distillation data of HVGO; (d) D1160 distillation data of VR.

Figure 9 .
Figure 9.Comparison of D86 or D1160 distillation data of the main products of the rigorous model.(a) D86 distillation data of VDO; (b) D1160 distillation data of LVGO; (c) D1160 distillation data of HVGO; (d) D1160 distillation data of VR.

◆Figure 12 .
Figure 12.The effect of FST at different SFR on the effect of deep-cut vacuum distillation.(a) HVGO yield change; (b) HVGO D86 95%-point temperature change.

Figure 12 .
Figure 12.The effect of FST at different SFR on the effect of deep-cut vacuum distillation.(a) HVGO yield change; (b) HVGO D86 95%-point temperature change.

Table 1 .
Data requirements for VDU model.

Table 1 .
Data requirements for VDU model.

Table 2 .
Setting data information of each column.

Table 3 .
The main data involved in simple model and rigorous model simulation.

Table 3 .
The main data involved in simple model and rigorous model simulation.

Table 4 .
Comparison between design and simulation values of operating parameters.

Table 5 .
Parameter setting of independent variables in case study.