Energy xxx (2014) 1e13

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Comparative analysis and evaluation of three crude oil vacuum

distillation processes for process selection
Wugen Gu a, Yuqing Huang a, Kan Wang a, Bingjian Zhang a, Qinglin Chen a, *,
Chi-Wai Hui b
a
School of Chemistry and Chemical Engineering/Key Lab of Low-carbon Chemistry & Energy Conservation of Guangdong Province, Sun Yat-Sen University,
Guangzhou 510275, China
b
Chemical and Biomolecular Engineering Department, Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong

a r t i c l e i n f o a b s t r a c t

Article history: There exists three practicable crude oil vacuum distillation processes and different processes have a
Received 18 January 2014 significant impact on material and energy performances, including product yield, economic potential,
Received in revised form heat recovery and the efficiencies of recoverable energy and recoverable exergy. Process selection is an
9 August 2014
important and difficult task for designers with various targets since the material and energy perfor-
Accepted 13 August 2014
Available online xxx
mances of a process do not coordinate with each other. In this work, an approach with simultaneous
considerations of material and energy performances is proposed to comparatively analyze and evaluate
the three processes, in order to provide insights for designers to screen a suitable process and vacuum
Keywords:
Crude oil distillation
furnace outlet temperature. The approach is conducted in three steps. In the first step, a simulation
Economic potential model is rebuilt to obtain basic material and energy data. In the second step, comparative analyses and
Exergy analysis evaluations are performed to measure the material and energy performances of three process options
Heat recovery under the same operating conditions. In the last step, the variations of the material and energy per-
formances are further investigated to determine the vacuum furnace outlet temperature. The results
indicate that the cycle process has the highest product yield and best economic potential but the lowest
efficiencies of recoverable energy and recoverable exergy, while the drawn process has the highest ef-
ficiencies of recoverable energy and recoverable exergy but the worst economic potential. The results
also demonstrate that the selection of a vacuum distillation process and the determination of the vacuum
furnace outlet temperature play a critical role in designing a crude oil vacuum distillation process.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction
The crude oil distillation unit (CDU) is a major process unit in a
refinery for separating crude oil into various product fractions. It
not only has the largest capacity, but also is one of the largest en-
ergy consumers in the refinery which consumes fuel at the equiv-
alent of 2% of the crude oil processed [1]. Consequently, it is a major
CO2 emitter in the refinery [2,3]. With high-energy costs and reg-
ulations on strict greenhouse gas emission, the CDU is required
urgently to improve the energy efficiency and economic profit.
Generally, a crude oil distillation process contains a preflash col-
umn, an atmospheric distillation column and a vacuum distillation
column. In a typical vacuum distillation process, an overflash oil is
* Corresponding author. Tel.: þ86 20 84113659; fax: þ86 20 84113731.
E-mail address: chqlin@mail.sysu.edu.cn (Q. Chen).
usually drawn out directly. With the processing requirement, two
modified vacuum processes are practically applied in which the
overflash oil is either sent into the stripping section of a vacuum
distillation column or cycled to a vacuum furnace. Different pro-
cesses have a significant impact on material and energy perfor-
mances, including product yield, economic potential, heat recovery
and the efficiencies of recoverable energy and recoverable exergy.
However, the material and energy performances of a process do not
coordinate with each other. Exploring the mutual relations can
provide insights for designers to determine a suitable vacuum
distillation process and vacuum furnace outlet temperature.
The energy analysis and optimization of crude oil distillation
processes have been a hot research topic and widely investigated.
The early crude oil distillation process only contained two-
distillation columns, an atmospheric distillation column and a
vacuum distillation column. Al-Muslim et al. [4e6] implemented
an energy and exergy analysis of this two-distillation column

http://dx.doi.org/10.1016/j.energy.2014.08.053
0360-5442/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
2 W. Gu et al. / Energy xxx (2014) 1e13
process and a single atmospheric distillation process, and pointed
out that the overall exergy efficiency of the former was 17.5% higher
than that of the latter. Bagajewicz and Ji addressed a heat demand-
supply diagram and used it as a tool to put forward a rigorous
procedure for the design of a vacuum fractionation unit [7]. The
column grand composite curve was an effective tool to quantita-
tively identify the targets for possible column modifications, and
widely used to retrofit crude oil distillation columns [8,9], a
methanol plant [10], a biodiesel production plant [11,12] and
multicomponent separation systems [13]. Subsequently, the pre-
distillation device was introduced into the process. Errico et al. [14]
and Benali et al. [15,16] employed exergy analysis to better un-
derstand the distribution of energy degradations in the distillation
column, and their studies showed that both energy consumption
and exergy loss were reduced when a flash drum was added. The
installation of a preflash vessel was also introduced to revamp a
crude distillation unit in Kamel's work [17,18]. Based on the former
work, a predistillation system for heavy crude oil was considered
and a two-preflash drum process was determined as an optimum
scheme by thermodynamic analysis [19]. Up to now, the conven-
tional three-column crude oil distillation process, including a pre-
flash/predistillation column, an atmospheric distillation column
and a vacuum distillation column, has been widely applied in re-
fineries. Ricardo et al. presented a method for the calculation of the
physical and chemical exergy of crude oil and its fractions [20], and
then carried out an exergo-economic analysis of all equipment
items in this conventional process to provide improvement po-
tential [21]. Mittal et al. [22] considered that crude oil blending had
a significant impact on energy, emission and economic profit (E3),
and developed an E3 methodology framework for better operating
the crude oil distillation system. Recently, the optimization work of
an atmospheric distillation column has been done by Arjmand et al.
[23]. The vapor feed was introduced into the upper stages of the
atmospheric distillation column, leading to a sizable energy saving.
Furthermore, several researchers paid attention to the multi-stage
crude oil distillation process in which one or more vacuum distil-
lations were added between a conventional atmospheric and a
vacuum distillation column. Gu et al. [24] proposed a general multi-
stage distillation energy consumption model to optimize the yield
distribution of distillation columns and the flowrates of side-
stream products on each distillation column were determined by
maximizing the thermal exergy recovery. Li et al. [25] investigated
the multi-stage effect on the crude oil distillation unit in thermo-
dynamics and recommended adding more vacuum distillation
columns for energy saving. According to the literature mentioned
above, it is inferred that the researchers paid more attention to
improving energy performance and providing solutions for the
predistillation column, the atmospheric distillation column and the
whole distillation system, but ignored the interactive relationship
between different vacuum distillation processes and material and
energy performances. It is an important and difficult task for de-
signers to select a suitable vacuum distillation process, and
comparative analyses and evaluations of three process options can
aid designers in process selection.
Process simulation, as an important aid tool in process design
and analysis, provides basic data for the energy analysis and opti-
mization of chemical processes. An equilibrium model or a rate
based model is generally used for the simulation of distillation
processes. The equilibrium model assumes that vaporeliquid
equilibrium is reached at each stage. The departure from the
equilibrium is accounted for by tray efficiency for tray columns or
the height equivalent of a theoretical plate for packed columns.
More et al. [26] used a distillation module based on the equilibrium
model in Aspen Plus to address the optimization of the crude oil
distillation system with a binary feed. Benali et al. [15,16] employed
Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
a distillation module based on the equilibrium model in PRO/II to
design an improved crude oil distillation process and determine
optimal parameters. The rate-based model [27e29] assumes that
the vaporeliquid equilibrium only occurs at the interface, and the
MaxwelleStefan equation is used to describe the mass transfer
between the vapor and liquid phases. Packed reactive distillation
columns [30], reactive dividing wall columns [31] and molecular
distillation [32] of green coffee oil were simulated with the rate
based models. In these studies, a single distillation module from the
simulation systems was employed to simulate the practical atmo-
spheric or vacuum distillation column. Actually, the vaporeliquid
separation of the feed only occurs in the feed section. Additionally,
the bottom of the rectifying section of a vacuum distillation column
is a blind tray, and the liquid from the bottom of the rectifying
section cannot directly flow into the stripping section. It is inac-
curate to represent a vacuum distillation column with a single
distillation module. A simulation model for the vacuum distillation
column must be properly rebuilt to provide reliable basic data.
This work is focused on providing insights for designers by
comparative analyses and evaluations of three crude oil distillation
process options. It is organized as follows. In Section 2, three
practicable vacuum distillation processes are depicted. An
approach for comparative analyses and evaluations of three process
options is presented in Section 3. Then, the results of the three
process options are discussed in detail by employing the proposed
approach in the following section. Finally, the conclusions are
given.
2. Process description
A CDU is generally the first process, and is also the largest ca-
pacity in a refinery complex. Different fractions are produced
depending on the difference in their boiling temperature, and are
further processed in the downstream plants for desired products.
An actual CDU in China is investigated in this paper, and its flow-
sheet is illustrated in Fig. 1. The crude oil, introduced from a storage
tank at atmospheric temperature, firstly passes through the pre-
heat train 1 which makes use of the heat of side-stream product
streams and pump-around circuits from the atmospheric and
vacuum distillation columns. The preheated crude oil then enters a
desalter to remove the trace salt and water. Afterwards, the
desalted crude oil enters the preheat train 2, and is preheated up to
about 200  C. Then it is pumped into a prefractionation distillation
column, and the light naphtha is separated. The topped crude oil
from the bottom of the prefractionation distillation column goes
through the preheat train 3, and then enters into the atmospheric
furnace and is heated up to about 365  C. Finally, the topped crude
oil is introduced to the atmospheric distillation column which
operates at atmospheric pressure. Superheated steam is introduced
from the bottom of the column, which is used to reduce the partial
pressure in the column and thus enhance vaporization and sepa-
ration of the crude oil. Heavy naphtha, kerosene, diesel, and gas oil
are drawn from the column. These side-stream products are
diverted into side strippers where the initial boiling point of the
product is controlled by stripping steams. This atmospheric column
is equipped with three pump-around circuits (APA1, APA2 and
APA3) to recover heat for the preheat trains. The atmospheric res-
idue (AR) is further heated by a vacuum furnace and then fed to a
vacuum distillation column. The AR is overflashed so that the total
vapor flowing up to the rectifying section is enough to produce
target products. The extra amount of vapor is known as an overflash
oil. It is worth noting that the bottom of the rectifying section is a
blind tray in the vacuum distillation column. The overflash oil from
the rectifying section of the distillation column cannot flow directly
into the stripping section inside the column, and it is drawn out at
 W. Gu et al. / Energy xxx (2014) 1e13 3

Fig. 1. A typical flowsheet of a crude oil distillation unit.

the bottom of the rectifying section. In a typical process, the over-
flash oil is drawn out directly and mixed with the vacuum residue,
which is called a drawn process (DP) here. With the consideration
of increasing product yields by extracting relatively light fractions
in the overflash oil, two modified vacuum processes have been
applied recently in practice. One process is to send the overflash oil
into the stripping section which is called the stripping process (SP),
as shown in Fig. 2(a). The other process is that the overflash oil is
cycled to pass through the vacuum furnace again which is called the
cycle process (CP), as shown in Fig. 2(b). After the vapor enters the
rectifying section, it is distilled as three side-stream products which
are light vacuum gas oil (LVGO), medium vacuum gas oil (MVGO)
and heavy vacuum gas oil (HVGO), respectively. The heat is
removed by these product streams and three vacuum pump-
around circuits (VPA1, VPA2 and VPA3). The steam is also injected
into the bottom to enhance the vaporization of light fractions.
Finally, the vacuum residue (VR) is produced in the bottom.
3. A comparative approach
An approach for comparative analyses and evaluations of three
crude oil distillation process options, a typical process (DP) and two
modified processes (SP, CP), is presented in this section. The
boundary of the investigated process here is the envelope with the
red dotted line in Fig. 1 (for interpretation of the references to
colour in this figure, the reader is referred to the web version of this
article), including a vacuum furnace, a vacuum distillation and the
heat recovery by the preheat train. The approach mainly contains
three steps: steady-state process simulation, exergy analysis and
economic potential evaluation, and the investigation of the varia-
tions of material and energy performances, which is shown in
Fig. 3.
In the first step, a simulation model is constructed based on the
PRO/II simulation package by considering three different function
sections of a vacuum distillation column, and the Braun K-10 (BK10)
model [15,26] is used to calculate the thermodynamic properties.
The real plant data are firstly adopted to examine the accuracy of
the model and the troubleshooting is performed when simulation
convergence fails, or simulation results do not match the plant data.
Then, these three processes are simulated by the proposed simu-
lation model. It should be noted that the operating conditions (for
example, distillation column pressure and product specifications)
are identical for these three processes which are from an actual
operating plant. In the second step, the material and energy data
required for exergy analysis and economic potential evaluation are
extracted from the simulation files. The data are then fitted into the
equations presented in Sections 3.2 and 3.3, and the material and
energy performances are measured. The main difference in these
three process options lies in the flow direction of the overflash oil,
and the flowrate of the overflash oil is associated with the vacuum
furnace outlet temperature. As a result, the vacuum furnace outlet
temperature is a critical parameter for these three processes, and

Fig. 2. Two modified crude oil vacuum distillation processes: (a) the stripping process (SP), (b) the cycle process (CP).

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
4 W. Gu et al. / Energy xxx (2014) 1e13
Fig. 3. The analysis and evaluation procedure.
the variations of the material and energy performances are further
investigated by altering the vacuum furnace outlet temperature in
the last step.
3.1. A simulation model
A vacuum distillation column is generally divided into three
parts: the rectifying section, feed section and stripping section, as
shown in Fig. 4(a). There exists a vaporeliquid equilibrium in the
rectifying section and different crude fractions are separated. The
Fig. 4. Simulation model for the crude oil vacuum distillation column: (a) the practical vacuum distillation column, (b) the simulation model.
Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
vapor and liquid phases of the feed are separated in the feed sec-
tion. The stripping section is designed to extract the relatively light
fraction from the residue through introducing high-temperature
steam at the bottom. It is not rational to represent a crude oil
vacuum distillation column with a single distillation module based
on either the equilibrium model or the rate based model. In this
work, we establish a virtually assembled process by combining a
flash module and two-distillation modules to represent the whole
vacuum distillation in PRO/II according to three different functional
sections, as shown in Fig. 4(b). The feed stream AR0 becomes a
vaporeliquid mixture after it is heated by a vacuum furnace. The
flashing vapor (Vf) and the flashing liquid (Lf) of the feed stream AR0
are separated in the feed section. The flashing vapor (Vf) is fed into
the rectifying section, and separated into various side-stream
products. The flashing liquid (Lf) contacts the vapor (Vs) from the
stripping section before it drops into the stripping section, but their
contact is not sufficient because of the limitation of time and space.
Thus, the feed section is regarded as an adiabatic flash. In the
stripping section, heat transfer exists between the dropping liquid
(Lf) from the feed section and the steam injected in the bottom. The
steam releases heat, and the liquid (Lf) with lower partial pressure
is heated to remove the light fractions. Finally, the remnant liquid is
VR. It is viewed as an equilibrium process approximately, and a
standard distillation column module with two theoretical trays is
chosen to represent the stripping section. In the rectifying section,
the rising vapor from the (N þ 1)th tray contacts fully the
descending liquid from the (N  1)th tray on the Nth tray, and it is a
typical vaporeliquid equilibrium process. The rectifying section is
modeled by using a standard distillation column module in the
PRO/II simulation package. The bottom product of the rectifying
section is the overflash oil. It is mixed with the VR in DP, sent into
the stripping section in SP, and cycled to the vacuum furnace in CP.
Implementing a steady-state simulation with recycle streams is
usually very difficult. The values of the recycle streams are usually
either zero or very large, which makes the simulation hard to
converge. The vapor stream at the top of the side column and the
liquid from the flash constitute a recycle loop (RL1) in the proposed
simulation model. In SP, there also exists a recycle loop (RL2) be-
tween the overflash oil introducing into the stripping section and
the feed of the rectifying section. The overflash oil entering to the
 W. Gu et al. / Energy xxx (2014) 1e13 5

furnace and the feed of the rectifying section form another recycle applied and the side-stream products are specified by the ASTM
loop (RL3) in CP. The tear stream approach [33] is applied to deal D1160 95% temperature. The simulated results from PRO/II are lis-
with the recycle stream, and a strategy involving four steps is ted in Table 1. It shows that the simulated values match the prac-
addressed in Fig. 5 to troubleshoot the simulation model. tical operation data quite well, and the proposed simulation model
can be employed to simulate the investigated processes.
Step 1: determine the calculation order of recycle loops. For the
processes with more than one recycle loop, the internal recycle
loop is firstly completed, and then extended outward. For 3.2. Exergy analysis
instance, in SP and CP process simulations, the calculation order
is RL1, RL2 and RL1, RL3, respectively. Exergy analysis is an effective thermodynamic method to eval-
Step 2: tear the recycle streams. The stream with fewer com- uate the performance of the processes based on the first and second
ponents is selected as the tear stream. In this work, the vapor laws of thermodynamics. There are three governing equations
stream (Vs) at the top of the side column and the overflash oil which are commonly used in exergy analysis of open systems,
are chosen as recycle streams (the red lines in Fig. 4; for inter- which are the conservation equations of mass, energy, and entropy
pretation of the references to colour in this figure, the reader is generation [6]. These equations are applied for the three process
referred to the web version of this article). options by neglecting changes in kinetic and potential energy, and
Step 3: obtain the initial value. The constraints are set in the written as:
simulation model according to the process specifications, and X X
the processes are simulated to get the initial value of the tear Min ¼ Mout (1)
stream.
X X X X
Step 4: connect the recycle streams. The tear stream of the first EM;in þ Qin ¼ EM;out þ Qout (2)
recycle loop is connected according to the calculation order
when it is convergent. On the contrary, the obtained value will X X X X
be a new initial value of the iteration. Afterwards, the remnant ExM;in þ ExQ ;in ¼ ExM;out þ ExQ ;out þ ExD (3)
recycle loops are connected in turn.
Exergy (Ex) is the maximum amount of work that might be
performed theoretically by bringing a resource into equilibrium
A practical CP is taken to examine the proposed simulation
with its surrounding through a reversible process [10,13,34], and is
model. The real plant data are collected and imported into the
defined as:
proposed simulation model. The thermodynamic method BK10 is
Ex ¼ DH  T0 DS (4)
The thermal exergy (ExQ) is considered for physical separation
processes, which is expressed as [35]:
Z  
T0
ExQ ¼ 1 dQ (5)
T
When the thermal exergy is calculated, the reference tempera-
ture (T0) and pressure (p0) states are 298.15 K and 1 atm, respec-
tively. For crude oil fraction streams represented by pseudo-
components, the exergy calculation can be found in Refs. [6,20].
The exergy destruction or internal exergy loss [21], caused by
the process irreversibility is deduced from the entropy generation
and the environment. According to the second law of thermody-
namics, the exergy destruction is always positive [36]. It is
expressed as the difference between the exergy input and the
exergy output.
X X
ExD ¼ Exin  Exout
X X  X X 
¼ ExM;in þ ExQ ;in  ExM;out þ ExQ ;out
(6)
Exergy efficiency is a well-known indicator to measure the
exergetic performance of a process. Different exergy efficiency
formulations were proposed for various processes in the literature
[37,38]. The total exergy efficiency proposed by Lior and Zhang
[37] is recommended for the process where a major part of the
exergy output is useful. In the crude oil vacuum distillation pro-
cess, the side-stream products and VR are further processed in
downstream units and converted into more commercially valuable
products. Besides, the heat is recovered by the preheat trains.
Thus, the total exergy efficiency is used to compare the degree of
thermodynamic perfection of the three processes in this paper,
which is expressed as the ratio of the total exergy output to the
Fig. 5. Strategy for the recycle loop. total exergy input.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
6 W. Gu et al. / Energy xxx (2014) 1e13

Table 1 [39]. A vertical line is drawn at the bottom of the cold composite
Plant data versus model data. curve to obtain an intersection. The heat of hot streams whose
Parameters Plant data Simulated value temperature is lower than the corresponding temperature of the
Feed, kg/h 331,800 331,800
intersection cannot exchange heat with cold streams, and is cooled
Furnace outlet temperature,  C 380 380 by cold utility. Therefore, the temperature of the intersection can be
Flash zone temperature,  C 362 358 defined as the recoverable temperature of hot streams, and the
Top tray pressure, mmHg 15 15 equations are expressed as:
Flash zone pressure, mmHg 30 30
Heavy diesel flowrate, kg/h 6000 6517 rec unrec
LVGO flowrate, kg/h 30,700 29,577
EM;out ¼ EM;out þ EM;out
X   X  
MVGO flowrate, kg/h 56,500 55,224
¼ Fi ci Ts;i  Trec þ Fi ci Trec  Tt;i (8)
HVGO flowrate, kg/h 65,170 65,779
Overflash oil flowrate, kg/h 34,183 33,000
LVGO drawn temperature,  C 164 162
X   X  
rec unrec
Qout ¼ Qout þ Qout ¼ Fj cj Ts;j  Trec þ Fj cj Trec  Tt;j
LVGO drawn temperature,  C 250 251
LVGO drawn temperature,  C 313 316 (9)
Total side-stream products yield, % 47.95 47.57

ExM;out ¼ Exrec unrec

M;out þ ExM;out
P P P Z   Z  
Exout ExM;out þ ExQ ;out T T
J¼ P ¼ P P (7) ¼ 1  0 dEM;out
rec
þ 1  0 dEM;out
unrec
(10)
Exin ExM;in þ ExQ ;in T T

From the viewpoint of heat recovery, the heat duty from the ExQ ;out ¼ Exrec unrec
Q ;out þ ExQ ;out
side-stream products and pump-around circuits can be classified Z   Z  
into two parts, as shown in Fig. 6(a). One part is recovered to pre- T T
¼ 1  0 dQM;out
rec
þ 1  0 dQM;out
unrec
(11)
heat the crude oil and is called the recoverable heat duty. The other T T
part is needed to be cooled with the cold utility and is called the In order to rationally evaluate heat recovery of the crude oil
unrecoverable heat duty. The energy quality of the recoverable and vacuum distillation process, the efficiency of the recoverable en-
unrecoverable heat duty is defined as the recoverable and unre- ergy (hrec) is defined as a ratio of the recoverable energy to the
coverable exergy, respectively. The recoverable heat duty and the energy input, and the efficiency of the recoverable exergy (Jrec) is
recoverable exergy are proposed to evaluate the heat recovery of a defined as a ratio of the recoverable exergy to the exergy input.
process. The recoverable heat duty is the heat recovery for the These two efficiencies are helpful in selecting a process with better
crude oil preheating trains from the start temperature (Ts) to the energy utilization performance.
recoverable temperature (Trec) while the unrecoverable heat duty is
P rec P rec P rec
the cooling heat with the cold utility from the recoverable tem- Eout EM;out þ EQ
perature (Trec) to the tank temperature (Tt). The start temperature is h rec
¼ P ¼ P P ;out (12)
Ein EM;in þ EQ ;in
available from the simulation files and the tank temperature is
specified by the corresponding product. The recoverable tempera- P rec P rec P
ture can be determined by the hot and cold composite curves based Ex ExM;out þ Exrec
Jrec ¼ P out ¼ P P Q ;out (13)
on pinch analysis of the whole unit, as shown in Fig. 6(b). The Exin ExM;in þ ExQ ;in
temperatureeenthalpy (TeH) diagram is frequently used to target
the hot and cold utilities. The cold and hot composite curves With regard to a vacuum furnace, by using Eq. (3), we obtain.
represent all the cold and hot streams in a process, respectively. The
ExAR þ Exfuel ¼ ExAR0 þ ExD;furnace (14)
overlap between the composite curves represents the maximum
amount of heat recovery possible within the CDU. The overshoot at
when it is used for CP, the overflash oil is added.
the bottom of the hot composite curve represents the minimum
amount of cold utility and the overshoot at the top of the cold ExAR þ Exoverflash oil þ Exfuel ¼ ExAR0 þ ExD;furnace (15)
composite curve represents the minimum amount of hot utility

Fig. 6. Heat recovery of CDU: (a) decomposition of the enthalpy of a stream according to heat recovery, (b) TeH diagram.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
 W. Gu et al. / Energy xxx (2014) 1e13 7

with regard to a vacuum distillation, we obtain
X X
ExAR0 þ Exsteam ¼ ExProduct i þ ExVPA j þ Exwater
i j
þ ExD;distillation
when it is used for CP, the overflash oil is also added.
X X
ExAR0 þ Exsteam ¼ ExProduct i þ ExVPA j þ Exoverflash oil
i j
þ Exwater þ ExD;distillation
For a vacuum distillation process, we obtain
X X
ExAR þExfuel þExsteam ¼ ExProduct i þ
i j
benefit is accounted for in terms of the reduced heat load of at-
mospheric furnace here. A weight factor a which is multiplied by
heat recovery is used here because heat recovery cannot be
completely transferred to the preheated crude oil due to heat loss.
(16) The economic potential expression (Eq. (26)) is specifically put
forward for a vacuum distillation process since heat recovery
cannot be used within the process. When the investigated
boundary is extended to the whole crude oil distillation process, Eq.
(25) is used. The operating conditions and product specifications
remain consistent in the three processes, but the vacuum furnace
outlet temperature and the stripping steam flowrate are manipu-
(17)
lated during the process simulation. Consequently, other operating
costs (management cost, labour cost, etc.) are assumed the same
and not involved in the gross economic potential in this work.
ExVPA j þExwater
Net revenue ¼ PB  FC  EC (25)

þExD;process (18)
Economic potential ¼ PB þ HR  FC  EC (26)
where where
X
ExProduct i ¼ Exheavy diesel þ ExLVGO þ ExMVGO þ ExHVGO (19) P
Product benefit PB ¼ pi  Fi
i
Heat recovery benefit HR ¼ a  QHR  pfuel
X Feed cost FC ¼ pAR  FAR
ExVPA i ¼ ExVPA 1 þ ExVPA 2 þ ExVPA 3 (20) Operation cost EC ¼ pfuel  Qfuel þ pcooling  Qcooling þ psteam  Fsteam
j

The exergy balance is found by comparing Eqs. (14), (16) 4. Results and discussion
and (18).
4.1. Basic data and process simulation
ExD;process ¼ ExD;furnace þ ExD;distillation (21)
The actual CDU in China (Fig. 1) is taken as a case study for the
As a result, the total exergy efficiency, the efficiencies of the comparative analysis and evaluation. A crude oil, Oman, is pro-
recoverable energy and recoverable exergy of a crude oil vacuum cessed in the unit and the capacity is 1000 t/h (175,331 bbl/d). The
distillation process can be calculated by: feed to the vacuum distillation column is an Oman atmospheric
P P residue. The TBP data of Oman are listed in Table A.1, and the basic
ExProduct i þ ExVPA j þ Exwater
i j data of the atmospheric distillation column for pinch analysis are
J¼ (22) given in Table A.3. After the light fractions are extracted in the at-
ExAR þ Exfuel þ Exsteam
mospheric distillation column, the flowrate of AR is about
P rec
P rec 549,651 kg/h, and the ASTM D1160 data are listed in Table 2. The
Eproduct i
þ EVPA j
i j
operating parameters and product specifications are listed in
rec
h ¼ (23) Table 3. The proposed model is used to simulate the three vacuum
EAR þ Efuel þ Esteam
distillation processes to achieve the stream parameters, and the
P P results are presented in Tables 4e6, respectively.
Exrec
product i þ Exrec
VPA j
i j
Jrec ¼ (24) 4.2. Quantitative comparisons under the same operating conditions
ExAR þ Exfuel þ Exsteam
4.2.1. Product yield
3.3. Economic potential The flowrates of all products are shown in Fig. 7. The total side-
stream product yields of the three processes are 57.69%, 59.06% and
The net revenue is generally based on product benefit, raw 60.16%, respectively. Fig. 8 indicates that all the side-stream prod-
material costs and energy costs which are the main variable oper- ucts comply well with the respective specifications. It is noted that
ating costs in a process, written as Eq. (25) [22,40,41]. In this DP is treated as a baseline for comparison. The product yields of SP
equation, the energy cost is dependent on heat recovery of the and CP are more than that of DP by 1.37% and 2.47%, respectively.
process studied. In the vacuum distillation process, the heat of side- The increased product flowrates reach 60,440 ton and 108,608 ton
stream products and pump-around circuits is recovered by the per year when the operating time is 8000 h/a. The ASTM D1160
crude oil preheat train. This reduces the heat load of the atmo-
spheric furnace and results in a decrease in fuel requirement. Table 2
However, it does not reduce the heat load of the vacuum furnace ASTM D1160 data of AR.
since heat recovery from the vacuum distillation column cannot be Percentage distilled, vol.% Temperature,  C
used to preheat AR. Therefore, heat recovery must be considered in
5 363
the economic evaluation for a crude oil vacuum distillation process. 10 407
A gross economic potential which considers simultaneously prod- 30 478
uct benefit, raw material, energy cost and heat recovery is intro- 50 524
duced in order to rationally evaluate the economic potential of a 70 580
90 807
vacuum distillation process, defined as Eq. (26). The heat recovery

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
8 W. Gu et al. / Energy xxx (2014) 1e13

Table 3 data of the VR in the three processes are listed in Table 7. It shows
Parameters and specifications of the vacuum distillation column. that the components of VR in SP and CP are both heavier compared
Parameters and specifications Value with DP, and more relatively light fractions are extracted in the
Total trays of the rectifying section 11
former two processes. These relatively light fractions enter MVGO
Feed tray 11 and HVGO whose yields in SP and CP are higher than that in DP, as
LVGO withdraw tray 2 shown in Fig. 7. The reason is that about 70 vol. % of the components
MVGO withdraw tray 6 of the overflash oil meets the product specifications of MVGO and
HVGO withdraw tray 9
HVGO (Table 7). Although both the product yields of SP and CP are
VPA1 withdraw, return tray 2, 1
VPA2 withdraw, return tray 6, 5 higher, their approaches are different. The relatively light fraction is
VPA3 withdraw, return tray 9, 8 obtained by vaporizing the overflash oil via the vacuum furnace in
Total trays of the stripping section 2 CP, while it is achieved with the stripping steam in SP. It is
Flash zone pressure, mmHg 30
concluded that the product yields of SP and CP are both higher than
Top pressure, mmHg 15
Overhead temperature,  C 80
that of DP under the same operating conditions, and the increment
Vacuum furnace outlet temperature,  C 400 is extracted from the overflash oil.
Steam flowrate/feed, wt% 2
LVGO D1160 95% temperature,  C 400 4.3. Energy performance
MVGO D1160 95% temperature,  C 500
HVGO D1160 95% temperature,  C 570
The stream data of these three processes are obtained from the
simulation files and listed in Tables 4e6, respectively. The energy
consumption of the vacuum furnace in DP and SP are both
Table 4
Stream parameters in DP. 21,209 kW while it is 23,085 MW in CP. There is 2056 kW higher in
CP because the overflash oil is cycled to the vacuum furnace to be
Stream Ts,  C P, mmHg F, kg/h E, kW Ex, kW
heated again. As the steam flowrate is specified to 2% of the feed,
AR 357 1177 549,651 133,382 70,273 the steam flowrate in CP is also a little more than that in DP and SP
Heavy diesel 80 15 2197 239 37 owing to the cycle of the overflash oil.
Water 80 15 10,990 8091 100
LVGO 144 17 30,264 2601 742
The energy and exergy performance results of the three pro-
MVGO 257 23 113,590 18,856 8252 cesses are presented in Table 8. It is found that the energy input in
HVGO 327 27 171,042 37,419 18,830 both DP and SP are the same in the three components, but CP has
VR 352 30 232,558 53,458 27,963 the larger values. This is due to more energy being consumed and
Steam 400 7356 10,990 9958 33,93
more steam injected in CP. From the perspective of exergy, although
VPA1 144e150 / / 10,882 41,44
VPA2 257e190 / / 13,153 8195 the exergy input in DP has the same value as that in SP, the exergy
VPA3 327e230 / / 19,724 13,383 output in DP is more than that in SP because the product yield in SP
Vacuum furnace 400 / / 21,209 21,209 is higher and it results in more exergy destruction in the distillation
process. For CP, both the exergy input and exergy output are a litter
higher due to more fuel consumption and steam injection. It causes
more flowrates via the vacuum furnace and more flowrates fed to
Table 5
Stream parameters in SP. the distillation column by cycling the overflash oil in CP, so exergy
destruction of these two components are both more than that in DP
Stream Ts,  C P, mmHg F, kg/h E, kW Ex, kW
and SP. Therefore, CP has the lowest exergy efficiency.
AR 357 1177 549,651 133,382 70,273 The heat duty of the side-stream products and pump-around
Heavy diesel 80 15 21,97 239 37
circuits is recovered by the crude oil preheat trains. The hot and
Water 80 15 10,990 8091 100
LVGO 144 17 30,394 2621 748 cold stream data of the three process options are listed in Table 9.
MVGO 257 23 114,233 19,011 8319 The composite curves are pictured in Fig. 9 based on pinch analysis
HVGO 327 27 177,826 38,989 19,620 by combining the data in Table 9 and Table A.3. When the minimum
VR 350 30 225,001 51,105 26,653 temperature approach (DTmin) is set to 20  C, the recoverable
Steam 400 7356 10,990 9958 3393
VPA1 144e50 / / 11,166 4247
temperatures of the three process options are 119.2  C, 119.4  C and
VPA2 257e190 / / 13,282 8254 120.4  C, respectively, and the recoverable energy and exergy per-
VPA3 327e230 / / 19,923 13,495 formance results are given in Table 10. The efficiencies of the
Vacuum furnace 400 / / 21,209 21,209 recoverable energy and exergy in SP are nearly the same as those in
DP. The decrease is mainly caused by more high-temperature vapor
Table 6
fractions being fed to the rectifying section and drawn out as side-
Stream parameters in CP. stream products with low temperature in SP. Compared with DP,
the recoverable energy and exergy of CP increase by 204 kW and
Stream Ts,  C P, mmHg F, kg/h E, kW Ex, kW
497 kW, respectively. However, the energy consumption including
AR 357 1177 549,651 133,382 70,273 fuel and steam in CP is 2792 kW more than that in DP, and the
Heavy diesel 80 15 2267 242 38
Water 80 15 12,000 8835 109
additional exergy input is 2188 kW. It means that the additional
LVGO 144 17 30,536 2629 750 heat duty is only recovered 7.31% from the streams while the
MVGO 257 23 114,747 19,070 8345 additional exergy recovery is only 22.71%, which are far lower than
HVGO 328 27 18,3121 40,240 20,282 the efficiencies of the recoverable energy and exergy in DP and SP. It
Overflash oil 359 30 50,448 12,317 6508
is proved in Fig. 9 that heat recovery is limited by a higher recov-
VR 351 7356 218,980 49,923 26,075
Steam 400 1177 12,000 10,874 3705 erable temperature in CP. Therefore, CP has the lowest efficiencies
VPA1 144e50 / / 11,684 4446 of the recoverable energy and exergy. We can also find that both the
VPA2 257e190 / / 13,840 8612 product yield and the exergy destruction increase in CP, but heat
VPA3 327e230 / / 20,748 14,077 recovery does not increase obviously, which should be considered
Vacuum furnace 400 / / 23,085 23,085
in the process design.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
 W. Gu et al. / Energy xxx (2014) 1e13 9

Fig. 7. Product flowrates of the three processes.

oil preheat train and the practical conditions of the case unit are
considered, the weight factor a is set to 0.7 here. The economic
potential in SP is 8144 CNY/h more than DP because the side-
stream product yield is higher while the energy consumption and
heat recovery are approximately identical. The product profit in CP
is more than that in DP, but the energy consumption also increases.
The total economic potential in CP is 14,042 CNY/h more than DP
due to the highest product yield. It is found that the profits of
products are dominant in the economic potential while the profits
of heat recovery only account for about 10% of the total economic
potential for all three processes. When the unit is expected to gain
more economic benefit, CP is suggested for its best economic
potential.

Fig. 8. ASTM D1160 distillation curves of the side-stream products in the three
4.5. Performance variations versus vacuum furnace outlet
processes.
temperature
4.4. Economic potential
In a vacuum distillation process, there are various operating
The prices of the raw material and products from the market are variables that affect the system, such as the vacuum furnace outlet
given in Table 11, and the economic potential is calculated ac- temperature, the distillation column pressure, top temperature of a
cording to Eq. (26), listed in Table 12. When heat loss in the crude distillation column and stripping steam flowrate, etc. In this work,
the vacuum furnace outlet temperature is chosen as a decision
variable for a critical reason that the main difference of the three
Table 7 process options lies in the flow direction of the overflash oil. The
ASTM D1160 data of VR and the overflash oil. flowrate of the overflash oil is associated with the vacuum furnace
ASTM D1160 vol.% VR,  C Overflash oil,  C outlet temperature, and it influences the material and energy
performances as discussed above. Thus, it is advisable to make a
DP SP CP
sensitivity analysis of the vacuum furnace outlet temperature and
5 523 539 544 483 further investigate the variations of material and energy perfor-
10 542 550 556 508
30 591 598 603 534
mance of the three process options. The temperature range is
50 656 663 669 550 approximately 380e420  C and a higher temperature is usually
70 766 773 779 572 forbidden owing to thermal cracking and coking of the crude oil.
90 932 935 937 605 Every process is simulated once with a step size of 10  C and the
results are summarized and analyzed.
Table 8 The variations of the product yields are pictured in Fig. 10. The
Energy and exergy performance results of the three processes. product yield grows as the temperature rises in the three pro-
Component Ein, kW Eout, kW Exin , kW Exout , kW ExD , kW J, % cesses. It is also found that the product yield gap between CP and
DP as well as between SP and DP is enlarging as the temperature
DP Vacuum 154,591 154,591 91,482 86,119 5363 /
furnace
Distillation 164,549 164,549 89,512 81,644 7868 / Table 9
column Stream data for the three processes.
Process 164,549 164,549 94,875 81,644 13,231 86.05%
SP Vacuum 154,591 154,591 91,482 86,119 5363 / Stream DP SP CP
furnace    
Ts, C T t, C Q, kW Ts, C Tt, C Q, kW Ts,  C T t,  C Q, kW
Distillation 164,549 164,549 89,512 81,474 8038 /
column LVGO 144 80 2601 144 80 2621 144 80 2629
Process 164,549 164,549 94,875 81,474 13,401 85.87% MVGO 257 90 18,856 257 90 19,011 257 90 19,070
CP Vacuum 168,784 168,784 99,848 94,027 5821 / HVGO 327 90 37,419 327 90 38,989 328 90 40,240
furnace VR 352 130 53,582 350 130 51,229 351 130 50,053
Distillation 179,658 179,658 97,732 89,243 8489 / VPA1 144 50 10,882 144 50 11,166 144 50 11,684
column VPA2 257 190 13,153 257 190 13,282 257 190 13,840
Process 167,341 167,341 97,045 82,735 14,310 85.25% VPA3 327 230 19,724 327 230 19,923 328 230 20,748

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
10 W. Gu et al. / Energy xxx (2014) 1e13

Table 11
Raw material costs and product prices.

Raw material Cost Product Price

AR, CNY/t 5400 Gas 2, CNY/t 4580

Steam, CNY/t 120 LVGO, CNY/t 7100
Fuel, CNY/kW 0.4 MVGO, CNY/t 6800
Cooling, CNY/kW 0.04 HVGO, CNY/t 6300
VR, CNY/t 5250

Table 12
Economic potential of the three processes.

Process PB, CNY/h FC, CNY/h EC, CNY/h HR, CNY/h Economic
potential, CNY/h

DP 3,295,841 2,968,115 10,680 35,663 352,708

SP 3,304,203 2,968,115 10,695 35,460 360,853
CP 3,310,775 2,968,115 11,630 35,720 366,750

outlet temperature, CP and SP are preferable to DP. Furthermore,

Fig. 9. Composite curves for the three processes: (a) DP, (b) SP, (c) CP.
from the viewpoint of the overall process in a refinery, SP and CP
are expected when the processing capacity of VR in the down-
stream units is lower. DP is recommended when the processing
capacities of side-stream products in the downstream units are
limited at a lower level.
Fig. 11 shows the variations of the total exergy efficiency. The
total exergy efficiency decreases when the temperature rises in the
three processes. At the same temperature, the total exergy effi-
ciency in CP is the lowest while the total exergy efficiency in DP and
SP are almost equivalent. The total exergy efficiency gap between
DP and CP is enlarging when the temperature is rising. It is proved
in Fig. 12 that the exergy destruction increases as the temperature
rises in the three processes, and it grows fastest in CP because the
increasing temperature makes increase in the vapor flowrate fed to
the vacuum distillation column. Over the temperature range, the
exergy destruction of DP and SP are all analogous while it is larger
in CP.
The recoverable energy and exergy do not increase obviously, as
shown in Fig. 13. The reason is that the recoverable heat duty is
determined by the crude oil preheat trains. When the temperature
rises, it also raises the recoverable temperature in the case of the
unchanged capacity of the whole crude oil distillation unit. In other
words, the recoverable energy and exergy almost remain un-
changed but both the energy input and the exergy input rise, so the
efficiencies of the recoverable energy and exergy decrease (Fig. 14).
In addition, the increasing vacuum furnace outlet temperature in-
creases the flowrate of the overflash oil, and it leads to a faster
increase of the energy input and the exergy input in CP than those

is rising. It is explained that the flowrate of the overflash oil in-

creases and more fractions are vaporized and fed into the recti-
fying section when the temperature rises. Over the temperature
range, the product yields in SP and CP are always higher than
those in DP, and CP always has the highest product yield.
Meanwhile, when the product yield is fixed (for example, 58%),
CP has the lowest vacuum furnace outlet temperature (TCP in
Fig. 10). It means that if a vacuum distillation process is designed
to get more distillate under a relatively lower vacuum furnace

Table 10
Recoverable energy and exergy performance results of the three processes.

Process Erec , kW Exrec , kW hrec , % Jrec , %

DP 127,367 73,767 77.40% 77.75%

SP 126,642 73,391 76.96% 77.35%
CP 127,571 74,264 76.23% 76.53%
Fig. 10. Variations of the product yield.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
 W. Gu et al. / Energy xxx (2014) 1e13 11

Fig. 11. Variations of the total exergy efficiency.

Fig. 14. Variations of the efficiencies of recoverable energy and exergy.

increase of the side-stream product flowrates associated with the

vacuum furnace outlet temperature. It is proved again that the
enhanced product profit is far higher than the additional energy
cost. It must be a considerable option in practice that greater eco-
nomic potential is returned at the expense of smaller energy cost in
the three processes. Compared to the energy performances in
Figs. 11 and 14, it is found that a process option of great economic
potential is not necessary for good energy performance. Hence, CP
under a higher temperature is suggested for designers when the
economic potential is targeted.

5. Conclusions

Three crude oil vacuum distillation processes are comparatively

analyzed and evaluated to aid designer for process selection. A
Fig. 12. Variations of the exergy destruction.
virtually assembled process with combination of different modules
in the PRO/II simulation package is established to represent the
whole vacuum distillation process, and this proposed simulation
model can accurately predict the parameters. By considering heat
of DP and SP. Consequently, the efficiency gap between DP and CP
recovery, the efficiencies of the recoverable energy and exergy as
widens when the temperature rises. As a result, from the energy
well as the economic potential are presented to rationally evaluate
performance, DP with a lower temperature is recommended when
the three processes. The results provide a guideline to screen a
the energy efficiency is targeted for this unit.
suitable process and vacuum furnace outlet temperature according
The economic potential is improving when the vacuum furnace
to the expected goals of the unit. The cycle process under a higher
outlet temperature increases, as shown in Fig. 15. The CP has the
vacuum furnace outlet temperature is suggested when the eco-
best economic potential over the temperature range. The economic
nomic potential is targeted, and the drawn process under a lower
potential gap between CP and DP is always enlarging with the

Fig. 13. Variations of the recoverable energy and exergy. Fig. 15. Variations of the economic potential.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
12 W. Gu et al. / Energy xxx (2014) 1e13

vacuum furnace outlet temperature is used when the exergy effi- Table A.2
ciency is targeted. From the viewpoint of the overall process in a Crude oil properties.

refinery, the stripping and cycle processes are expected when the Properties Value
processing capacity of the vacuum residue in the downstream units Density, kg/m3 861
is lower. The drawn process is preferred when the processing ca- API degrees standard 32.84
pacities of side-stream products in the downstream units are Sulfur content 1.39%
limited at a lower level. Furthermore, it is demonstrated that Viscosity (50  C), mm2/s 10.34
Watson characterization factor 12.2
greater economic potential is returned at the expense of smaller
energy cost in the three processes.

Table A.3
Acknowledgment The cold and hot stream data of the atmospheric distillation column.

The authors gratefully acknowledge the financial support from Stream Ts,  C Tt,  C Q, kW

the National Natural Science Foundation of China (No. 21376277), Crude oil 30 210 116,838
the project of the Science and Technology Star of Zhujiang of Topped oil 192 365 134,804
AR 357 400 21,208
Guangzhou City (No. 2013J2200006), and the Fundamental
Light naphtha 130 70 15,417
Research Funds for the Central Universities. Heavy naphtha 90 60 2869
Kerosene 149 70 6551
Nomenclature Diesel 225 70 9432
Gas oil 298 70 12,125
APA1 114 74 11,620
c heat capacity (kJ/t/ C) APA2 198 138 18,635
E energy (kW) APA3 271 191 25,507
Ex exergy (kW)
F flowrate (kg/h)
M mass flowrate (kg/h)
p price (CNY/t) References
P pressure (mmHg)
Q heat duty (kW) [1] Liu X, Chen D, Zhang W, Qin W, Zhou W, Qiu T, Zhu B. An assessment of the
T temperature ( C) energy-saving potential in China's petroleum refining industry from a tech-
nical perspective. Energy 2013;59:38e49.
[2] Li H, Mu H, Zhang M, Gui S. Analysis of regional difference on impact factors of
Greek symbol China's energy-related CO2 emissions. Energy 2012;39:319e26.
h energy efficiency (%) [3] Szklo A, Schaeffer R. Fuel specification, energy consumption and CO2 emission
in oil refineries. Energy 2007;32(7):1075e92.
J exergy efficiency (%) [4] Al-Muslim H, Dincer I, Zubair SM. Exergy analysis of single- and two-stage
crude oil distillation units. J Energy Res Technol 2003;125:199e207.
[5] Al-Muslim H, Dincer I, Zubair SM. Effect of reference state on exergy effi-
Subscripts and superscripts
ciencies of one- and two-stage crude oil distillation plants. Int Therm Sci
AR atmospheric residue 2005;44:65e73.
Cooling cooling utility [6] Al-Muslim H, Dincer I. Thermodynamic analysis of crude oil distillation sys-
D exergy destruction tems. Int J Energy Res 2005;29:637e55.
[7] Ji S, Bagajewicz M. Design of crude distillation plants with vacuum units:
fuel fuel consumed in the hot utility targeting. Ind Eng Chem Res 2002;41:6094e9.
HR heat recovery [8] Sharma R, Jindal A, Mandawala D, Jana SK. Design/retrofit targets of pump-
i i side-stream product around refluxes for better energy integration of a crude distillation column.
Ind Eng Chem Res 1999;38(6):2411e7.
j j pump-around circuit [9] Gadalla M, Jobson M, Smith R. Increase capacity and decrease energy for
rec recoverable existing refinery distillation columns. Chem Eng Prog 2003;99(4):44e50.
s start [10] Demirel Y. Retrofit of distillation columns using thermodynamic analysis. Sep
Sci Technol 2006;41(5):791e817.
steam steam consumed in the unit [11] Nguyen N, Demirel Y. Retrofit of distillation columns in biodiesel production
t target plants. Energy 2010;35(4):1625e32.
unrec unrecoverable [12] Nguyen N, Demirel Y. Using thermally coupled reactive distillation columns in
biodiesel production. Energy 2011;36(8):4838e47.
[13] Soares Pinto F, Zemp R, Jobson M, Smith R. Thermodynamic optimization of
Appendix. The basic data of the atmospheric distillation distillation columns. Chem Eng Sci 2011;66(13):2920e34.
column [14] Errico M, Tola G, Mascia M. Energy saving in a crude distillation unit by a
preflash implementation. Appl Therm Eng 2009;29:1642e7.
[15] Benali T, Tondeur D, Jaubert JN. An improved crude oil atmospheric distilla-
The TBP data and the properties of the crude oil Oman processed tion process for energy integration: part I: energy and exergy analyses of the
are listed in Tables A.1 and A.2. The cold and hot stream data of the process when a flash is installed in the preheating train. Appl Therm Eng
2012;32:125e31.
atmospheric distillation column are listed in Table A.3.
[16] Benali T, Tondeur D, Jaubert JN. An improved crude oil atmospheric distilla-
tion process for energy integration. Part II: new approach for energy saving by
use of residual heat. Appl Therm Eng 2012;40:132e44.
Table A.1 [17] Kamel D, Gadalla M, Ashour F. New retrofit approach for optimization and
TBP data of the crude oil Oman. modification for a crude oil distillation system. Chem Eng Trans 2013;35:
1363e8.
Percentage distilled, vol.% Temperature,  C [18] Gadalla M, Kamel D, Ashour F, Eldin HN. A new optimization based on retrofit
approach for revamping an Egyptian crude oil distillation unit. Energy Proc
5 44
2013;36:454e64.
10 93
[19] Wang Y, Hou Y, Gao H, Sun J, Xu S. Selecting the optimum predistillation
30 195
scheme for heavy crude oils. Ind Eng Chem Res 2011;50(18):10549e56.
50 355 [20] Rivero R, Rendon C, Monroy L. The exergy of crude oil mixtures and petroleum
70 501 fractions: calculation and application. Int J Appl Thermodyn 1999;2(3):
90 656 115e23.

Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
 W. Gu et al. / Energy xxx (2014) 1e13
[21] Rivero R, Rendon C, Gallegos S. Exergy and exergoeconomic analysis of a
crude oil combined distillation unit. Energy 2004;29:1909e27.
[22] Mittal V, Zhang J, Yang X, Xu Q. E3 analysis for crude and vacuum distillation
system. Chem Eng Technol 2011;34(11):1854e63.
[23] Arjmand M, Moreno M, Liu L. Energy saving in crude oil atmospheric distil-
lation columns by modifying the vapor feed inlet tray. Chem Eng Technol
2011;34(8):1359e67.
[24] Gu W, Zhang B, Wang K, Chen Q. Energy optimization for a multi-stage crude
oil distillation process. In: 26th International conference on efficiency, cost,
optimization, simulation and environmental impact of energy systems, ses-
sion 28; 2013. p. 1e13.
[25] Li X, Lin C, Wang L, Li H. Exergy analysis of multi-stage crude distillation units.
Front Chem Sci Eng 2013;7(4):437e46.
[26] More RK, Bulasara VK, Uppaluri R, Banjara VR. Optimization of crude distil-
lation system using aspen plus: effect of binary feed selection on grass-root
design. Chem Eng Res Des 2010;88:121e34.
[27] Taylor R, Krishna R. Multicomponent mass transfer. New York: Wiley; 1993.
[28] Koshy W, Amin SA, Shah JA. Rate based distillation. Int J Sci Eng Technol
2012;1(3):21e3.
[29] Demirel Y. In: Nonequilibrium thermodynamics: transport and rate processes
in physical, chemical and biological systems. 3rd ed. Amsterdam: Elsevier;
2014.
[30] Peng J, Lextrait S, Edgar TF, Eldridge RB. A comparison of steady-state equi-
librium and rate-based models for packed reactive distillation columns. Ind
Eng Chem Res 2002;41(11):2735e44.
[31] Mueller I, Kenig EY. Reactive distillation in a dividing wall column: rate-based
modeling and simulation. Ind Eng Chem Res 2007;46(11):3709e19.
Please cite this article in press as: Gu W, et al., Comparative analysis and evaluation of three crude oil vacuum distillation processes for process
selection, Energy (2014), http://dx.doi.org/10.1016/j.energy.2014.08.053
13
[32] Duran MA, Maciel Filho R, Wolf Maciel MR. Rate-based modeling approach
and simulation for molecular distillation of green coffee oil. Comput Aided
Chem Eng 2010;28:259e64.
[33] Luyben WL. Control of a multi-unit heterogeneous azeotropic distillation
process. AIChE J 2006;52:623e37.
[34] Demirel Y. Exergy analysis of distillation columns. Int J Exergy 2006;3:
345e61.
[35] Wang Y, Feng X, Cai Y, Zhu M, Chu KH. Improving a process's efficiency by
exploiting heat pockets in its heat exchange network. Energy 2009;34(11):
1925e32.
[36] Ghannadzadeh A, Thery-Hetreux R, Baudouin O, Baudet P, Floquet P, Joulia X.
General methodology for exergy balance in ProSimPlus process simulator.
Energy 2012;44:38e59.
[37] Lior N, Zhang N. Energy, exergy, and second law performance criteria. Energy
2007;32(4):281e96.
[38] Marmolejo-Correa D, Gundersen T. A comparison of exergy efficiency
definitions with focus on low temperature processes. Energy 2012;44:
477e89.
[39] Kemp IC. Pinch analysis and process integration: a user guide on process
integration for the efficient use of energy. Oxford: Elsevier; 2007.
[40] Inamdar SV, Gupta SK, Saraf DN. Multi-objective optimization of an industrial
crude distillation unit using the elitist non-dominated sorting genetic algo-
rithm. Chem Eng Res Des 2004;82(A5):611e23.
[41] Al-Mayyahi MA, Hoadley AFA, Smith NE, Rangaiah GP. Investigating the trade-
off between operating revenue and CO2 emissions from crude oil distillation
using a blend of two crudes. Fuel 2011;90:3577e85.