Sun Mengying; Fu Dianliang; Sun Lanyi

(1. State Key Laboratory of Heavy Oil Processing, College of Chemical Engineering,China University of Petroleum (East China), Qingdao 266580;2. School of Energy and Power Engineering, Shandong University, Jinan 250061)

Abstract: Crude distillation unit (CDU) is regarded as the main energy consumer in the entire refinery process. In this paper, the process simulation software and the energy management software are used to simulate the flowsheet and analyze the energy consumption, respectively. Stream data obtained from an existing CDU are applicable in the pinch analysis. To reduce the amount of cross-pinch heat transfer, three approaches of resequencing, repiping, and adding heat exchangers are adopted. Compared with the existing CDU, the results demonstrate that the inlet temperature of the furnace can be increased by 25.4 ℃, the amount of hot and cold utilities can be reduced by 15.1% and 19.6%, respectively. The economic evaluation indicates that the operating cost is saved by 8×106 $/a, and the payback period is about 9 months.

Key words: crude distillation unit; pinch analysis; heat exchanger network; retrofit; energy saving

1 Introduction

Crude distillation unit (CDU) is a typical crude processing unit and the first processing unit in all refineries. However, it is also the main energy consumer in a refinery, accounting for 20% of the total energy consumption of the refinery[1-2]. Therefore, energy saving is urgently needed to improve the efficiency of the CDU.As an important subsystem of energy recovery, heat exchanger network (HEN) optimization has been remarkably recognized for attaining maximum energy savings and economic interests[3-6]. Many researchers have studied the optimization of HEN of the CDU from different aspects. For example, Rossiter[7]used energy management software to carry out pinch analysis on the CDU, then eliminated heat exchangers operating with low efficiency, which are heat exchangers with large heat transfer duty across the pinch, and the retrofit achieved about 45% of the target savings for both energy and monetary value. Gadalla, et al.[8]presented a method based on rigorous simulation and optimization structure, which could optimize the operating conditions of distillation column (with the help of process simulation software)and HEN simultaneously, so as to make the most of the existing equipment save energy. Gu, et al.[9]proposed a HEN retrofit scheme with steam generation for the energy consumption problem of CDU, three heuristic rules were introduced to determine the heat flow suitable for steam production, it was showed that the retrofit scheme could save 9.65% and 18.22% of the existing cold and hot utilities, respectively.

The HEN optimization methods can be divided into two categories, namely the pinch analysis and the mathematical programming[10-14]. Pinch analysis is practical[15], which can determine the energy target of actual engineering problems and provide guidance for the design of HEN[16]. In addition, users can also make appropriate adjustments when matching streams according to the actual situation[17]. Chemical process simulation can be used for the development and research of new technological processes, new device design,device modification, and fault diagnosis[18]. Process simulation software can recognize the change of the heat capacity flow rate of the streams, and extract the streams data in segments[19]. Therefore, the process simulation and pinch analysis can be combined to design and optimize the HEN more accurately and conveniently[19-20].

In this paper, pinch analysis is used to determine energy targets, perform energy diagnosis, and explore energysaving opportunities of the present CDU capacity of 8.0 Mt/a. The simulation software Aspen HYSYS and the energy management software Aspen Energy Analyzer are combined to retrofit the HEN of CDU. The retrofitted HEN is compared with the existing HEN in terms of energy consumption and economics.

2 Process Description

The simplified process flow diagram for CDU is shown in Figure 1, which is composed of three distillation columns,two furnaces, and the main process streams related to the heat recovery network. The crude oil introduced from the storage tank is subject to three times of heat exchange,which are named the preheat train 1, then the preheat train 2, and the preheat train 3, respectively. The crude oil is first preheated by the preheat train 1 between the tank and the desalter. Afterwards, the desalted crude oil is further preheated by the preheat train 2 before entering the preflash column. The topped crude oil from the bottom of the pre-flash column flows through the preheat train 3 to the atmospheric column, and then heavy naphtha, kerosene,diesel, and gas oil are separated from the column. The product streams are transferred to the side strippers where the initial boiling point of the product is controlled by stripping steams. The atmospheric column is equipped with three pump-around circuits, viz.: the atmospheric top pump-around (ATPA), the atmospheric middle pumparound (AMPA), and the atmospheric bottom pumparound (ABPA), to recover heat for the preheat trains.The high-boiling bottom stream atmospheric residue (AR)from the atmospheric column enters the vacuum column to be further distilled. Then three kinds of products are extracted from the side lines, which are light vacuum gas oil (LVGO), medium vacuum gas oil (MVGO), and heavy vacuum gas oil (HVGO). The vacuum column also has three pump-around circuits, named the vacuum top pump-around (VTPA), the vacuum middle pump-around(VMPA), and the vacuum bottom pump-around (VBPA),respectively. The vacuum residue (VR) is obtained at the bottom of vacuum column.

Aspen HYSYS V9.0[21]is used to simulate the detailed process system. For refinery process simulation, the fluid package is specified to be PR (Peng-Robinson)[22]. The process simulations based on sub-flowsheet are shown in Figure 2, in which (a), (b), and (c) correspond to the three preheating trains in Figure 1, respectively.

Figure 1 Process flow diagram for CDU

Figure 2 Preheat processes of (a) preheat train 1, (b) preheat train 2, and (c) preheat train 3

3 Existing Heat Exchanger Network Analysis

3.1 Data extraction

Data extraction is the prerequisite for determining energy targets and retrofitting HEN[23]. The heat capacity flow rate of the process streams varies greatly in CDU, so it cannot be regarded as a constant[19,24]. Process simulation software can recognize the change of the heat capacity flow rate of the streams, and extract the streams data in segments. In this paper, comprehensive and accurate stream data can be extracted from Aspen HYSYS V9.0 to Aspen Energy Analyzer V9.0[21], which are listed in Table 1, including 14 hot streams and 4 cold streams.

3.2 Cost data

The purpose of HEN optimization is to reduce the heating and cooling costs by matching the hot and cold process streams, in a bid to minimize the capital cost and operating cost.

Table 1 The data extracted for the pinch analysis

3.2.1 Capital cost

The capital cost (CC, $) of heat exchanger based on the recommendation of Aspen Energy Analyzer is calculated by the following equation:

wherea(=30800 $) represents the installation cost of the heat exchanger;b(=750) andc(=0.81) are the duty/area-related cost set coefficients of the heat exchanger;A(m2) is the heat transfer area of the heat exchanger;Nshellrepresents the number of heat exchanger shells in the heat exchanger.

3.2.2 Operating cost

The operating cost (OC, $/a) represents the energy cost required for operating the equipment, and is shown as follows:

whereQi,HUandQj,CUare hot and cold utility consumptions, respectively (MW);Ci,HUandCj,CUare utility costs for hot and cold utility, respectively ($/MW a).The operating time is 8400 h annually, and the utility data of CDU are listed in Table 2.

Table 2 The utility data of CDU

3.3 Energy targets

The minimum energy targets are calculated based on the selected ΔTmin. With the decrease of ΔTmin, the capital cost of HEN (which increases as ΔTmingets smaller) and operating cost (which decreases as ΔTmingets smaller)also change. It is feasible to explore the trade-off between capital cost and operating cost, but in practice this is rarely done. Instead, the determination of ΔTminvalue is based on the application experience[7]. In this work,the ΔTminof CDU is set at 15 ℃. As shown in Figure 3, the temperature-enthalpy (T-H) diagram and grand composite curve (GCC) are plotted based on ΔTmin=15 ℃.TheT-Hdiagram and GCC show the pinch temperature is 295.7 ℃ (corresponding to a pinch at 303.2 ℃ on the hot streams and 288.2 ℃ on the cold streams), the minimum hot and cold utility duties are 86.22 MW and 62.54 MW, respectively.

Figure 3 (a) T-H diagram and (b) GCC of CDU

3.4 Energy diagnosis

The hot and cold utilities in existing HEN are 102.2 MW and 78.48 MW, respectively. TheT-Hand GCC diagrams of the existing HEN are shown in Figure 4. The dotted lines indicate theT-Hand GCC under the retrofit target, which is the information in Figure 3. Compared with the minimum consumption of hot and cold utilities determined, there is a potential energy saving of 15.94 MW, accounting for 15.6% and 20.3% of the existing HEN hot and cold utilities, respectively.

Heat transfer across the pinch leads to an increase in the usage of the hot and cold utilities. According to the three principles of HEN design (no cooler above the pinch,no heater below the pinch, heat must not be transferred across the pinch), the cross-pinch heat transfer can occur in three different ways[25]:

· Heat transfer between process streams across pinch,QPP

· Using hot utilities below pinch,QPH

Figure 4 (a) T-H diagram and (b) GCC of existing HEN

· Using cold utilities above pinch,QPC

The amount of heat transfer across the pinch (QP) is calculated by the following equation:

In order to save energy, the better way is to reduce or even eliminate the cross-pinch heat transfer. Judging from the grid diagram of the HEN of CDU shown in Figure 5, the heat exchangers used for heat transfer across the pinch can be clearly seen. There are 33 heat exchangers in the existing CDU, among which the heat exchangers E20, E22,E18, and F1 transfer heat across the pinch. The amount of cross-pinch heat transfer in each heat exchanger is listed in Table 3, denoting thatQPis 15.94 MW. Table 3 reveals that E20 and E22 have the largest heat transfer across the pinch,while F1 and E18 are significantly smaller, so the retrofit sequence is E20 and E22, and the heat transfer across the pinch of F1 and E18 is ignored in the retrofit process.

The existing HEN also has heat exchangers that seriously lose the driving force, which leads to the underutilization of heat. Hence, the high-temperature streams should be used to preheat the high-temperature crude as much as possible to reduce the consumption of utilities. Besides,since the existing CDU has complex configurations and interactions, it is necessary to identify the difficulty of retrofitting HEN, and maximize the use of existing equipment while considering the space limitation of the device.

Figure 5 Grid diagram of the current HEN of CDU with pinch line

Table 3 Cross-pinch heat transfer summary

4 Retrofit of Existing Heat Exchanger Network

4.1 Retrofit scheme

The approaches of resequencing, repiping and adding heat exchanger can be used to retrofit the HEN. Resequencing aims at adjusting one or both ends of the heat exchanger with another heat exchanger in the same hot and cold streams without changing the matched streams of the heat exchanger, which means it is necessary to reverse the order of the two heat exchangers. Repiping is similar to resequencing, but one or both of the matched streams can be different to those used in the current situation. Adding heat exchanger may change an existing matched up heat load. With the help of three methods, the HEN is retrofitted in order to solve the energy bottleneck proposed in the previous section. The detailed retrofit steps are as follows:1) The first step is to add a new exchanger N1 between HS-1 stream and CS-2 stream, in order to transfer the heat of VR in E20 above the pinch to crude oil at the bottom of the pre-flash column at F1 cold end. However,the heat load of E20 is reduced, resulting in a temperature reduction at the junction of crude oil at the bottom of preflash column branches. Resequencing exchangers E21 and N1, then crude oil at the bottom of pre-flash column is preheated below the pinch by using the VBPA. The crosspinch heat transfer of E20 is eliminated, and the furnace heat load is reduced from 72.27 MW to 63.94 MW by these modifications.

2) Using the heat load of the saved cold streams to cool the hot streams under the pinch, so as to reduce the consumption of cold utilities. A specific method is used to transfer the surplus cooling duties of crude oil at the bottom of pre-flash column through the path, then the heat exchanger N2 is added between C8 and C5 of the gas oil system to use the heat of the streams for balancing the heat load of desalting crude oil streams and reducing the cooling load of C5. However, these options make the temperature of the hot stream in E10 and the heat transfer driving force higher. The easiest way to modify is achieved through resequencing E10 and N2, and moving E10 to the cold end of N2. In addition, the heat load of C5 can be reduced by 8.33 MW after adopting these changes.3) Increasing the heat load of N1 and making full use of vacuum residuum to preheat crude oil at the bottom of pre-flash column above the pinch. Repiping the heat exchanger E22, so that its hot stream end is between E21 and E17 to reduce the temperature between the exchangers E21 and E17 on the VBPA, while reducing the heat load of E17. Repiping the exchanger E4 cold streams end to the N2 cold streams end is adopted to supply heat for the surplus cooling capacity. Then, the temperature level of the cold stream in the E3 is increased for transferring the remaining cooling capacity of the raw crude produced by repiping of E4. Because the temperature level of the ATPA,where E7 is located, is higher and the heat transfer driving force of the raw crude is higher, the E4 hot streams end is rearranged to the upstream of E7. Through the above improvements, the phenomenon of heat transfer across the pinch of E22 can be eliminated, and the temperature level of VTPA can also be increased.

4) Using the heat load of the saved cold streams to cool the hot streams below the pinch, so as to save the consumption of cold utilities. The heat exchanger N3 is added between the MVGO and raw crude to reduce the surplus cooling capacity of raw crude. Surprisingly, C7 can be removed. Add a heat exchanger N4 between diesel stream and raw crude stream to reduce the surplus cooling capacity of raw crude again, which can also reduce the heat load of C4. Finally, the surplus cooling capacity is eliminated and the heat load of C1 is reduced by increasing the heat load of E3.

After retrofitting, preheat networks of the preheat train 1, the preheat train 2, and the preheat train 3 are shown in Figure 6.

Figure 6 Preheat networks of (a) preheat train 1, (b)preheat train 2, and (c) preheat train 3

4.2 Retrofit results

Final grid diagram of HEN of CDU is shown in Figure 7.N1 and E18 still transfer heat across pinch, but the amount has been reduced from 15.94 MW to 0.57 MW, a reduction of 96.4%. The inlet temperature of furnace F1 is increased from 287.3 °C to 313.7 °C, and the reduction of heat load of F1 amounts to 15.37 MW. The amount of hot and cold utilities after retrofitting is shown in Table 4, which shows that the scheme has saved 15.41 MW of hot utility and 15.37 MW of cold utility, accounting for 15.1% and 19.6%of the existing hot and cold utilities, respectively.

4.3 Economic evaluation

Compared with 5.6×107$/a of operating cost of HEN beforeretrofitting, the operating cost after retrofitting is 4.8×107$/a, which can save 8×106$/a. The relevant parameters of optimal network design including area, duty, and capital cost are presented in Table 5. The total capital cost is 5.74×106$,and the payback period is calculated at about 9 months.

Table 4 Comparison of energy consumption before and after retrofitting

Figure 7 Final optimal modifications of grid diagram of the process

Table 5 Summary of heat exchangers to be adjusted during retrofitting

5 Conclusions

This paper analyzed the energy consumption of CDU, and pinch analysis was applied as a method for eliminating the heat transfer across the pinch in the HEN. Based on interactions between heat exchangers, three methods (viz.:resequencing, repiping, and adding new heat exchangers)were adopted to produce the retrofit scheme of CDU. The results reveal that the retrofit scheme can reduce 96.4%of the heat transfer across the pinch of the existing HEN,and the inlet temperature of the furnace can be increased by 26.4℃, while the consumption of hot and cold utilities can be reduced by 15.1% and 19.6%, respectively. It demonstrates that the payback period is about 9 months by economic analysis. Furthermore, the methods can give a reference for retrofitting HEN of other units.

Acknowledgments:This work was supported by the National Natural Science Foundation of China (Grant: 21878333).Furthermore, the authors are grateful to the editor and the anonymous reviewers for their helpful comments and constructive suggestions with regard to the revision of the paper.