close

Вход

Забыли?

вход по аккаунту

?

Minimisation of fuel energy wastage by improved heat exchanger network designЧan industrial case study.

код для вставкиСкачать
ASIA-PACIFIC JOURNAL OF CHEMICAL ENGINEERING
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
Published online 9 October 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI:10.1002/apj.102
Research Article
Minimisation of fuel energy wastage by improved heat
exchanger network design – an industrial case study
Adel S. Ashaibani and Iqbal M. Mujtaba*
School of Engineering Design & Technology, University of Bradford, Bradford BD7 1DP, UK
Received 3 April 2007; Revised 4 August 2007; Accepted 13 August 2007
ABSTRACT: Petroleum refineries fulfil their energy (process heat) requirement by direct fuel firing. In this regard,
a heat exchanger network (HEN) is widely used to recover thermal energy that may be otherwise wasted. The HEN
can significantly reduce the overall energy consumption in industrial processes. In many industrial processes such as
oil refineries, crude oil needs to be heated to a required temperature. Here a study of a HEN problem of a crude oil
distillation unit of an African Oil Refining Company (AORC) is carried out. In this unit the crude is required to enter
the distillation column at 328 ◦ C, while the crude inlet temperature at the furnace is only 220 ◦ C. Therefore, a heater
must supply heat to raise the temperature by a third of the target temperature. In this work pinch technology is used
in the existing HEN in order to obtain a desirable energy saving and to identify the required additional modification.
With the addition of only one heat exchanger and re-construction of the HEN, significant utility savings are achieved.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
KEYWORDS: refinery; energy waste; heat exchanger network; design
INTRODUCTION
Petroleum refineries extract their energy requirement
in the form of direct fuel firing for process heat and
steam generation (for process use and electrical power
generation), as well as from electrical power. As shown
in Fig. 1, the distribution of the energy requirement for
a typical refinery was reported to be a follows: 65% of
the total energy used for process heat, 28% accounts
for steam generation and only 7% for electrical power
generation.[1] As a result of the increased energy cost
during the 1970s, much attention is being paid to energy
conservation.
Heat exchangers can be used to recover thermal
energy that may otherwise be wasted. It can reduce
the overall thermal energy consumption in industrial
processes. The design of a new chemical plant using
optimal heat recovery networks can often lead to a
reduction of both capital and operating costs. Most
industrial plants have hot utilities for heating cold
streams, and cold utilities for cooling hot streams. If
some of the hot and cold streams can be matched using
heat exchangers, then the necessity for hot and cold
utilities can be substantially reduced.
*Correspondence to: Iqbal M. Mujtaba, School of Engineering
Design & Technology, University of Bradford, Bradford BD7
1DP, UK.
E-mail: i.m.mujtaba@bradford.ac.uk
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
A certain number of cold streams to be heated
and a certain number of hot streams to be cooled
define the heat recovery network design problem. If
the required heating and cooling are achieved, then the
design is referred to as ‘minimum capital cost design’.
Alternatively, we can cool hot streams by using it to
heat cold streams, and if any stream does not reach its
target temperature we can use the hot or cold utility to
complete the job. This design is called the ‘maximum
heat recovery’ approach.
Heat exchanger networks (HENs) are widely used
to reduce the energy consumption in many process
industries such as oil refineries and petrochemical
and chemical industries. In these industries savings in
cost can be achieved if an optimal network design is
used. A number of methods based on simulation and
rigorous optimisation techniques have been reported in
the literature.
In this work, an HEN problem related to an oil
refining company in Libya is studied for the purpose
of developing an optimised HEN and energy saving.
Pinch technology (PT) is used in a systematic way to
achieve the objective. Note that it was not the intention
of the authors to use different HEN design techniques
available in the literature to see the adequacy of the
pinch method as compared to other methods; rather,
the intention was to bring industrial data from an
existing plant to the public domain and apply PT to
design/suggest an improved energy saving HEN.
576
A. S. ASHAIBANI AND I. M. MUJTABA
Asia-Pacific Journal of Chemical Engineering
where hi is the inside film coefficient and ho is the outer
film coefficient.
The log mean temperature difference is
Electricity
7%
Steam
Generation
28%
Tlm = (Th − Tc )/ ln(Th /Tc )
where Th is the temperature difference at the hot
terminal,
Process Heat
65%
Th = Ths − Tct , and
Tc is the temperature difference at the cold terminal,
Tc = Tht − Tcs .
Figure 1. Refinery energy distribution. This figure is
available in colour online at www.apjChemEng.com.
HEAT EXCHANGER MODEL
PINCH TECHNOLOGY
The design of a heat exchanger consists of two important parts: thermal design to ensure that the unit will
transfer the required amount of heat under the design
condition, and mechanical design to ensure that the unit
will withstand the pressure and loads. Figure 2 illustrates the temperature changes that occur in the two
streams along the length of heat exchangers of the
counter-flow configuration.[2]
In general, the heat lost by the hot stream = heat
gained by the cold stream
mh C ph (Ths − Tht ) = Q = mc C pc (Tct − Tcs )
(1)
where m is the mass flow rate; Cp is the specific heat,
which is supposed to be constant; Ths , Tcs are the supply
temperatures; and Tht , Tct are the target temperatures.
The total heat transfer Q can be related to an area
of heat transfer (A), overall heat transfer coefficient
(U ) and log mean temperature difference between the
streams (Tlm ) by
Q = U × A × Tlm
(2)
where U is overall heat transfer coefficient, which can
be estimated or calculated from the simple equation:
1/U = 1/hi + 1/ho
(3)
Temperature
Ths
Hot stream
Tht
Tcs
(4)
Tct
Cold stream
Length
Figure 2. Temperature distribution diagram.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Energy savings may come about in a number of
ways such as improvement of the temperature level
of the available heat source. It is possible to achieve
considerable savings in energy by conventional means
such as heat recovery from the feed and product streams
using heat exchangers.
In the literature, one of the most important considerations in building a modern crude unit is the design of
the crude pre-heat-exchanger train. Crude units are usually designed to process more than one type of crude.[3]
As a result, heat loads in the heat exchanger train can
vary over a wide range owing to change in product distribution and operating conditions. As heavier crude is
processed, the heat load shifts from the light product
exchangers to the heavy product exchangers. Also, the
heat transfer coefficients are generally lower because of
the higher viscosity of the heavier product.[4] Exchange
of thermal energy from a hotter fluid to a colder fluid
is one of the most prominent processes in the oil refining industry. Overall performance of the refinery can
be affected by the performance of heat exchange in
heat exchangers. Heat exchangers are exposed to various problems and phenomena, which can affect their
performance and efficiency. These are fouling, corrosion and plugging. In order to ensure optimum and
maximum heat transfer, with minimum down time, heat
exchangers are required to undergo inspection, cleaning,
both chemical and physical, and replacement if necessary. Fouling of heat exchangers may be defined as the
deposition of undesirable materials both on the tube or
shell side of the heat exchangers. This causes reduction
in the overall heat transfer coefficient and increase in
pressure drop, and thus poor performance. This can also
sometimes cause complete failure, which can stop the
flow and even cause rupture.
However, if the pre-heat-exchanger train fails to
perform at an optimum condition, the furnace (fired
heaters) would need higher duty, necessitating higher
consumption of fuel.
From the energy savings point view, the important
fields for energy use improvement are the HEN retrofit
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
AN IMPROVED HEAT EXCHANGER NETWORK DESIGN
projects which maximise the existing heat recovery.
Linnhoff and Flower[5] developed the tool known as PT
for designing the HEN in the late 1970s. The application
of PT on continuous processes is becoming more
attractive and provides a suitable tool for analysing
any process at different stages of the design. The PT
provides many helpful graphical representations that are
easy to use by the designer for analysis and for better
understanding of the problem. This technology has been
applied to many existing and new processes.
To illustrate the PT concepts, Fig. 3 shows a very
simple example of HEN (heat recovery) containing one
hot stream and two cold streams. The hot stream needs
to be cooled to its target temperature Tht , and both cold
streams need to be heated to their target temperatures
Tct . In Fig. 3 the hot stream exchanges heat with cold
streams in the heat exchangers, where the temperature
of hot stream is decreased to certain value, but to reach
its target temperature a cooler is used to reduce the
temperature of the hot stream to its target temperature.
After heat exchange with the hot stream, the first
cold stream requires extra heating to achieve its target
temperature and a heater is employed for this purpose.
Figure 4 shows the composite curves for the hot and
cold streams. The hot and cold streams are represented
as straight lines on the temperature–enthalpy (T –H )
diagram, with different heat capacity flow rates. Each
of them represents the cumulative heat sources and sink,
respectively, within the process region. Since the cold
stream is to be heated up and the hot stream is to be
cooled down to their target temperatures, the direction
of the cold stream line is from the low to the high temperature and that of the hot stream is in the opposite
direction.
For one hot stream and two cold streams, the T –H
diagram of Fig. 4 shows that the amount of heat that
can be recovered is represented by the overlap of the hot
stream with both cold streams. Additional heating and
cooling is required to achieve the target temperatures,
and these are denoted by Steam (St) and Cooling water
(Cw). The goal is to minimise the amount of St and
Cw. The pinch point is where the temperature difference
Heat source
T,
Temperature
St
Ths
Hot composite curve
Tct
∆Tmin at
the pinch
Heat recovery
Cooled composite curve
The pinch
Tht
Tcs
Cw
Heat sink
H, Enthalpy
Figure 4. Composite curve.
between those two curves is the minimum. This point
divides the system into two thermodynamically separate
subsystems; while only cold utility is required below
the pinch, only hot utility is required above the pinch.
In order to achieve maximum energy recovery (MER),
the following rules of Tjoe and Linnhoff and Linnhoff
et al .[6,7] should be followed:
• No cold utility to be used above the pinch point
• No hot utility to be used below the pinch point
• No process heat to be transferred across the pinch
For PT applications, first the data for cold and hot
streams are analysed, in which the supply temperature
(Ts ) and target temperature (Tt ) as well as the enthalpy
change (H ) and heat capacity flow rate (mC p) are
identified. It is important to have all the required data
for the existing energy consumption under the operating
conditions to do the pinch analysis. With the objective
to save energy, the PT will then be used to evaluate the
existing HEN of an industrial process.
PINCH DESIGN METHOD
Ths
Tcs1
Cold streams
Tcs2
E
E
Design task
Tht
C
Hot stream
Tct2
Cooler
H
Process exchangers
Heater
Tct1
Figure 3. Heat recovery network.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Consider a heat exchanger system, in which the process
streams need to be heated and cooled from their supply
(initial) temperature Ts to their target (final) temperature Tt . The streams that require cooling are defined as
‘hot’, and streams requiring heating are defined ‘cold’.
Hot streams are sources of heat, while the cold streams
are sinks.
Each stream is characterised by its
• supply Ts and target Tt temperatures (◦ C)
• mass flow rate m (kg/h)
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
577
578
A. S. ASHAIBANI AND I. M. MUJTABA
• heat content (enthalpy) Q (kJ/h)
• specific heat capacity C p (kJ/kg K)
Asia-Pacific Journal of Chemical Engineering
NOUT ≥ NIN?
NOUT ≥ NIN?
No
Yes
There are two methods for determining the utility
target: the composite curve method and the problem
table algorithm (PTA) (used in this case study). This
method[8] can be expressed in four steps:
Split streamOUT
1.
2.
3.
4.
Adjust for Tmin
Set up the temperature intervals
Calculate interval heat balance
Cascade for positive heat flows.
We restrict our considerations to the design of a
maximum heat recovery network.
The assumptions are as fiollows:
• The streams have constant heat capacity flow rates
• Tmin for heat transfer applies to all potential
matches
• There are no constraints on the streams being
matched
• Temperature intervals are defined by stream supply
and target temperatures, adjusted for Tmin
• Within any interval, hot and cold streams are at least
Tmin apart.
The interval boundary temperatures are set at 1/2
Tmin below the hot stream temperature and 1/2 Tmin
above the cold stream temperature.
Foran illustrative example on the PTA, see Douglas.[8]
Stream splitting
In more complex practical cases, such as the ‘multiple
utilities’, a more comprehensive set of rules and guidelines is required. These rules and guidelines constitute
a major part of the ‘Pinch Design Method’ of Linnhoff
and Hindmarsh.[9] For the MER design, the cooling utility must be used below the pinch. So all hot streams
must be cooled to their pinch temperatures by interchange with cold streams. In general, we first observe
whether the number of streams OUT is bigger or equal
to the number of streams IN (NOUT ≥ NIN where NOUT
is the number of streams run away from the pinch and
NIN is the number of streams run into the pinch). Then
observe if mC pOUT is bigger than or equal to mC pIN ,
for every match. Whenever the design runs into trouble
in matching hot and cold streams, it will be necessary to
follow the rules of pinch design. However, stream splitting does add complexity to the network as well as to its
flexibility. Stream splitting cannot reduce the number of
units below the target value. Step-by-step procedures for
finding stream splits are given in Fig. 5. This approach
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
mcpOUT ≥ mcpIN
for every match?
No
Split
StreamIN
Yes
Unsplit
solution
mcpOUT ≥ mcpIN
for every match?
No
Split
StreamIN
Yes
Place
matches
Figure 5. Design procedure using stream split.
has been used in the African Oil Refining Company
(AORC) refinery case study presented in this work.
INDUSTRIAL HEAT EXCHANGER NETWORK
OF AORC
Many industrial processes, such as in the AORC crude
oil refining, require heating to a required temperature. In
this process, the crude is required to enter the distillation
column at 328 ◦ C, while the heater inlet temperature is
only 220 ◦ C. To raise its temperature to the target of
328 ◦ C, fired heaters are needed with high consumption
of fuel. The objective of this case study is to evaluate
and optimise the pre-heat-train network to reduce the
consumption of fuel.
Problem representation
A variety of symbolic representations have been used
in the development of HEN. A simple representation is
that the hot streams are grouped together at the top and
run from left to right from their supply temperatures to
target temperatures.[10] Cold streams run counter to the
hot streams beneath. Process exchangers are represented
by vertical lines and circles on the streams matched.
Heaters are represented by circles on cold streams and
coolers are represented by circles on hot streams.
Problem statement
There is a set, Hs , of hot streams that are to be cooled
and a set, Cs , of cold streams that are to be heated. Each
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
AN IMPROVED HEAT EXCHANGER NETWORK DESIGN
Table 1. Process stream data.
Stream no.
and type
1.
2.
3.
4.
5.
6.
7.
8.
9.
COLD
COLD
COLD
HOT
HOT
HOT
HOT
HOT
HOT
Name of stream
m Flow rate
(kg/h)
Cp specific heat
(kJ/kg K)
Crude
Crude
water
FO
HGO
BPA
LGO
K
TPA
312 517
312 517
19 000
111 881
34 377
121 434
45 940
53 128
544 642
2.067
2.494
4.184
2.314
2.197
2.711
2.389
2.272
2.418
mcp heat capacity
flow rate (kJ/h)
T t target
temp ◦ C
Q (mC p. T )
heat load (kJ/h)
107
328
99
80
55
184
55
40
55
Net heating =
−56,199,390
−177 211 060
−1 510 500
62 910 270
16 766 550
26 665 200
17 560 000
15 450 880
81 650 280
−13 917 770
T s supply
temp ◦ C
645 970
801 860
79 500
258 890
75 525
329 200
109 750
120 710
1 316 940
20
107
80
323
277
265
215
168
117
Keys: FO, Fuel oil (bottom product); HGO, Heavy gas oil; BPA, Bottom pump around; LGO, Light gas oil; K, Kerosene; TPA, Top pump
around.
stream has a mass flow rate m, specific heat Cp , supply
temperature Ts and target temperature Tt .
The objective is to design a network of heat exchangers that maximises heat recovery and minimises the
consumption of fuel in the fired heaters.
In order to reduce the complexity of the problem, the
following assumptions are made:
• The minimum allowable temperature approach Tmin
for heat transfer is the same for all matches
• All streams are in single phase
• Stream properties are assumed to be constant with
respect to temperature
• There is only one hot utility and one cold utility
• There are no processing constraints, such as safety
and plant layout
9
117 °C
8
168 °C
7
215 °C
6
8
5
4
323 °C
C
40 °C
C
55 °C
184 °C
4
5
6
2
99 °C
328 °C
FH
C
3
265 °C
277 °C
55 °C
1
H
107 °C
55 °C
C
80 °C
3 80 °C
H
H
C
H
2 107 °C
H
H
H
1 20 °C
The existing heating = 89.24x106 kJ/hr and cooling = 93x106 kJ/hr
Figure 6. Existing HEN design and network pinch matches.
Problem formulation
In this refinery, the design capacity of crude throughput
is about 3.12 × 106 kg/h (60 000 bbl/day). The crude
is the major cold stream that is to be heated from an
ambient temperature to 328 ◦ C. In this process there
are three cold streams and six hot streams, many
heat exchangers (H) to allow process-to-process heat
transfer, one fired heater (FH) using fuel oil or fuel
gas as a hot utility and many coolers (C) using air or
water as cold utility. Figure 6 represents the existing
HEN of AORC.
The consumptions of existing hot and cold utilities
are 89.24 × 106 and 93 × 106 kJ/h, respectively, as
calculated from individual heater and coolers in the
existing HEN.
On the basis of the design data obtained from the
plant, all hot and cold streams have been identified.
A problem of a total of nine streams is formulated.
The process streams (mass flow rate, supply and target
temperatures and specific heats) all are listed in Table 1.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
On the basis of the first law of thermodynamics
analysis,[8] Table 1 shows that 13 917 770 kJ/h should
be supplied from utilities.
Problem analysis
A minimum utility algorithm as described before is
applied in order to locate the network pinch and to
predict the minimum amount of utilities required. Utility
targets are determined for different values of Tmin (at
5, 7, 10, 15 and 20 ◦ C) by using PTA. However details
are given only for a Tmin of 10 ◦ C as shown in Table 2.
Table 2 shows that each interval will have either a
net surplus (+Q) or net deficit (−Q) of heat. All the
surplus intervals reject heat to the cold utility and all
deficit intervals take heat from the hot utility. The net
difference between the heat available in the hot streams
and in the cold streams equal to −13 917 770 kJ/h,
which is identical to that obtained using the first-law
calculation (Table 1).
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
579
580
A. S. ASHAIBANI AND I. M. MUJTABA
Asia-Pacific Journal of Chemical Engineering
Satisfying the net heating and cooling requirements
in each interval involves simply supplying any heat
required from a hot utility and transferring any excess
heat to a cold utility (Fig. 7). This Fig. 7 shows that
we would need to supply 82 557 410 kJ/h from the hot
utility and would have to reject 68 639 640 kJ/h to the
cold utility. We also note that the difference is the same
as that obtained using the first law.
As can be seen from Fig. 7, the heat is being
transferred from the highest temperature intervals to a
cold utility directly, rather than supplying this available
heat to meet some of the energy requirements at lower
intervals. Thus it leads to high consumption of energy.
338 °C
Net = - 82557410
328 °C
-12027900
But if the arrangement shown in Fig. 7 can be improved,
energy consumption can be reduced. Figure 8 shows
how the heat available at the high temperature interval
level can be transferred to the next lower temperature
interval level.
From Fig. 8 it can be seen that instead of supplying 82 557 410 kJ/h we need to supply only 68 218
810 kJ/h and we reject only 54 301 040 kJ/h instead of
68 639 640 kJ/h. So there is a savings of 17.37% in hot
utility and 20.9% in cold utility.
From Fig. 8 we note that there is no heat transfer
between the seventh and eighth temperature intervals.
This point is known as pinch temperature (117 ◦ C for
338 °C
Net = - 68218810
-2496620
-2496620
-5609340
-5609340
-6912250
H
O
T
117 °C
U
T
I
L
I
T
Y
-6912250
C
O
L
D
-883345
-5723120
-12086235
U
T
I
L
I
T
Y
+9886760
+21970555
+12358450
+5861730
H
O
T
107 °C
117 °C
-5723120
-12086235
U
T
I
L
I
T
Y
0.0
+9886760
+21970555
+12358450
+5861730
U
T
I
L
I
T
Y
107 °C
-645715
-7878900
-7878900
30 °C
20 °C
-6459700
-6459700
Figure 7. Heat transfer to and from utilities.
St. No.
4
62910270
20 °C
Net =+ 54301040
Net =+ 68639640
HOT
UTILITY
68,218,810
C
O
L
D
-883345
+18562145
30 °C
328 °C
-12027900
Figure 8. Cascade diagram.
St. No.
5
16766550
St. No.
6
26665200
St. No.
7
17560000
12,019,420
St. No.
8
15450880
St. No.
9
81650280
1,510,500
25,838,74
St. No.
1
56199390
St. No.
2
177211060
St. No
3
1510500
COLD UTILITY
54301040
Figure 9. Minimum number of exchangers (based on first principles).
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
AN IMPROVED HEAT EXCHANGER NETWORK DESIGN
mcp (kJ/hr)
168°
117°
8
7
5
117°
215°
184°
323°
117°
245°
F
272°
117°
6
H
75,525
5
258,890
2
260°
328°
329,200
4
277°
4
109,750
3
265°
6
120,710
7
228°
205°
H
H
202°
125°
H
107°
2
801,860
H
260°
H
Figure 10. Above the pinch design. This figure is available in colour online at
www.apjChemEng.com.
the hot streams and 107 ◦ C for the cold streams, but
sometimes the average value of 112 ◦ C is used).
Network design
• Start with the biggest streams ‘IN ’–observe mCpOUT
≥ mCpIN
• Place all pinch matches first
• Maximise loads on all pinch matches
• Re-use as many existing matches as possible
After determining the minimum heating and cooling
requirements for the HEN and the pinch location,
the minimum number of heat exchangers required to
transfer the heat from the source to the sink (based
on first law) is calculated from the following relation,
and as shown in Fig. 9. The numbers in the rectangular
boxes are taken from Table 1.
Number of
Number of
Number of
=
+
Exchanger
Streams
Utilities
− 1 = 9 + 2 − 1 = 10
Figure 10 shows the complete design above the
pinch. The first point to realise is that, because there
are four hot streams and only one cold stream above
the pinch, the cold stream must be split into four ways
(parallel branches). Heat capacity flow rates of branches
are chosen to obey the mCp inequality as mentioned
before (Fig. 5) and to facilitate ticking off the streams.
Figure 11 shows the complete design below the pinch.
Figure 12 shows the complete design of the network,
which is obtained by coupling the design of hot end
Now commence the design using the ‘Pinch Design
Method’, keeping in mind that we want to maximise
the compatibility of the existing network. The hot end
design (above the pinch) consists of six streams: five
hot streams and one cold stream. The cold end design
(below the pinch) consists of seven streams: five hot
streams and two cold streams. Design is started at the
pinch and developed moving away from the pinch.
Applying the feasibility criteria according to Fig. 6, the
pinch topology is identified.
The following rules are used for the new design:
mcp (kJ/hr)
• Divide the problem at the pinch
• Start at the pinch and move away
• Check whether the number of steams OUT is bigger
than or equal to the number of streams IN
6,45,970
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
1,316,940
9
117°
55°
1
117°
1,20,710
8
1,09,750
7
75,525
5
25,890
4
8
105°
C
117°
79,500
C
117°
C
117°
C
99°
107°
H
H
80°
40°
55°
55°
80°
3
20°
1
Figure 11. Below the pinch design.
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
581
582
A. S. ASHAIBANI AND I. M. MUJTABA
Asia-Pacific Journal of Chemical Engineering
for calculating A is U , the overall heat transfer coefficient, which should be calculated or estimated.
The values of U for the process streams are estimated from different sources.[2,11 – 14] They provide the
typical overall heat transfer coefficient for various heat
exchanger types and services, and usually these are adequate for initial size and cost estimation.
In this work, both the values of U and the cost
estimation are estimated based on Hewitt et al .[13] Other
sources (as mentioned above) gave only the value of U .
Normally, the cost of a heat exchanger per unit area,
and hence the cost value, decreases with increasing
heat exchanger size. Co is given at specific values
of (Q/T ), and its value at any other intermediate
values of (Q/T ) may be estimated by logarithmic
interpolation.
Pinch
117°C
1
9
7
8
C
4
6
5
5
C
6
2
C
H
H
H
F
C
3
7
4
8
H
3
H
2
H
H
H
107°C
QHmin = 68.22 x106 kJ/hr
existing heat exchangers
1
QCmin = 54.30 x106 kJ/hr
new heat exchanger
RESULTS AND DISCUSSION
Figure 12. Final design for HEN. This figure is available in
colour online at www.apjChemEng.com.
Table 3 summarises the differences between the existing
design of the HEN with the proposed design. The data
shown in Table 3 is based on the plant design data (at
100% capacity). Table 3 gives the various parameters
of interest, such as the number of heat transfer units,
utility services required, total heat exchange area, heat
exchangers cost, etc. It can be seen that by improving
the amount of heat exchange (heater inlet temperature
was increased from 220 to 245 ◦ C) a reduction in
heating utility was achieved, and the cooling utility is
reduced accordingly. The energy saving is 21.03 × 106
kJ/h (23.6%). In addition, the cooling utility is also
reduced by 42%. This can be achieved by adding only
one heat exchanger, with a pay back time of only
2 months. Since the cooling duty was decreased, the
and cold end. As shown in Fig. 12, the modification
identifies the addition of a new heat exchanger match
between stream number 2 and stream number 8.
Area and cost estimation
Before the approximate heat exchanger cost can be
determined from the cost data, the surface area must be
calculated. Normally, the cost of a heat exchanger per
unit area decreases with increasing heat exchanger size.
The heat exchanger area, A, can be calculated from
Eqn (2). However, the remaining unknown parameter
Table 2. Temperature interval.
SUB
NETSTREAMS AND TEMPERATURE
WORK
T °C
1 2 3
SN1
SN2
SN3
SN4
SN5
SN6
SN7
SN8
SN9
SN10
SN11
SN12
SN13
SN14
SN = Stream No.
328
313
267
255
205
174
158
107
99
80
70
64
45
30
20
4
5
6
7
Q = [(mcp)h(mcp)c]. ∆T
(kj/hr)
∆T
8
9
323
277
265
215
184
168
117
109
90
80
74
55
40
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
15
46
12
50
31
16
51
8
19
10
6
19
15
10
Net Heating =
-12 027 900
-24 976 620
- 5 609 340
- 6 912 250
- 883 345
- 5 723 120
-12 086 235
+ 9 886 760
+21 970 555
+12 358 450
+ 5 861 730
+18562 145
- 7 878 900
- 6 459 700
-13 917 770
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
AN IMPROVED HEAT EXCHANGER NETWORK DESIGN
Table 3. Comparison between existing and new HEN (plant operating at 100% capacity).
No
1
2
3
4
5
6
7
8
Description
Total amount of heat exchange (kJ/h) ×106
Heating utility duty (kJ/h) ×106
Cooling utility duty (kJ/h) ×106
Number of heat exchange elements
Heat exchangers
Coolers
Heater
Heater inlet temp. (◦ C)
Total area of heat exchanger (m 2 )
Extra area cost ($)
Pay back time (month)
New design
Existing
design
168.37
68.21
54.30
153.8
89.24
93.0
8
5
1
245
3869
250 000
2
7
6
1
220
2946
–
–
Table 4. Comparison between existing and the new design (plant operating at 113% capacity).
No
1
2
3
4
5
6
7
8
a
Description
Total amount of heat exchange (kJ/h) ×106
Heating utility duty (kJ/h) ×106
Cooling utility duty (kJ/h) ×106
Number of heat exchange elements
Heat exchangers
Coolers
Heater
Heater inlet temp. (◦ C)
Total area of heat exchanger (m 2 )
Extra area cost ($)
Pay back time (month)
New design
191.78
83.32
63.6
8
5
1
245
3869
250 000
2
Existing design
Existing
designa
166.87
104.62
79.73
181.55
92.02
65.04
7
6
1
220
2946
–
–
7
6
1
235
2946
–
–
Existing design with improved energy and product recovery.
number of coolers is reduced: one cooler is eliminated
(Figs 6 and 12).
Ashaibani[3] has noted that from time to time the
refinery runs at 113% capacity. Thus, the evaluation of
the HENs with a higher capacity is considered. Table 4
shows a comparison between these two configurations.
The results obtained in this work (Table 4) clearly
show the flexibility of the existing HEN to operate at
different capacities with different flow rates. This will
allow the refinery process to handle different types and
amounts of crudes, which produce different distribution
of products. The new design shows a hot utility
savings of 20.36% and a cold utility savings of 20.23%
compared to those obtained by the existing HEN. The
use of existing HEN with optimised product recovery
and minimum utilities in the distillation column gives
a hot utility savings of 12.04% and a cold utility
savings of 18.42% compared to those obtained by
existing HEN.
Finally, the new HEN design eliminates one cooler
at the expense of the addition of one heat exchanger
and the cost of the cooler was not considered in the
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
evaluation of the economics. The cost of removing the
existing pipelines and installation of new pipelines in
the revamped design was not included. Also, the cost
of installation of the splitter and mixer in the new design
was not included. However, not including these will not
affect the savings in hot and cold utilities, which will
lead to energy savings, which was the main focus of
this paper.
CONCLUSIONS
In this work the existing pre-heat train of the crude
unit was analysed by applying pinch techniques. It was
found that about 23.6% savings in fuel energy cost
is achieved with the addition of only one new heat
exchanger. It is to be noted that any savings in hot utility
actually achieved will lead to an equivalent saving in
cold utility. A comparison between the modified and
existing structure of the plant HEN is shown in Fig. 13.
As a result of this, it is recommended that the refinery
modifies the existing HEN.
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
583
584
A. S. ASHAIBANI AND I. M. MUJTABA
Asia-Pacific Journal of Chemical Engineering
(a)
TPA2
LGO
FO
C1
CRUDE
E1
BPA2
C2
E2
HGO
E4
E3
FO2
LGO2
TPA1
C5
C4
C3
FEED
E5
BPA1
C6
E6
H1
FO1
HGO2
W.
K2
K..
K.
K
C4
E8
C5
W
Existing design of pre-heat train
(b)
K
FO
LGO
E8
HGO
BPA2
C1..
K.
E5
TPA2
C.1.
CRUDE
C1
C5
C4
C7
E4
E7
FEED
SP1
E1
C3
C1.
M1
C6
C
E6
H1
E3
TPA1
C.1
FO2
C2
FO1
E2
K2
LGO2
BPA1
HGO2
New design of pre-heat train
Figure 13. Comparison between proposed and existing design of HEN. This figure is available in colour online
at www.apjChemEng.com.
REFERENCES
[8]
[1] R.V. Elshout, Hydrocarbon Process., 1982; July: 109–116.
[2] D.Q. Kern, Process Heat Transfer, McGraw-Hill Inc., USA,
1950; p.871.
[3] A.S. Ashaibani, Modelling simulation and optimization of
refinery processes with energy conservation, PhD Thesis,
University of Bradford, UK, 2002; p.223.
[4] F. Huang, R. Elshout, Chem. Eng. Prog., 1976; 68–74.
[5] B. Linnhoff, J.R. Flower, AIChE J., 1978; 24, 633–642.
[6] T.N. Tjoe, B. Linnhoff, Chem. Eng., 1986; 47–60.
[7] B. Linnhoff, D.W. Townsend, D. Boland, G.F. Hewitt, B.E.A.
Thomass, A.R. Guy, R.H. Marsland, User Guide on Process
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
[9]
[10]
[11]
[12]
[13]
[14]
Integration for the Efficient Use of Energy, Institution of
Chemical Engineers, UK, 1982; p.41.
J.M. Douglas, Conceptual Design of Chemical Processes,
McGraw-Hill, Inc., USA, 1988; p.601.
B. Linnhoff, E. Hindmarsh, Chem. Eng. Sci., 1983; 38,
745–763.
B. Linnhoff, J.R. Flower, AIChE J., 1978; 24, 642–654.
A.S. Foust, L.A. Wenzel, C.W. Clump, L. Maus, L.B.
Andersen, Principles of Unit Operations, John Wiley and Sons,
Inc., USA, 1960; p.578.
O. Frank, Chem. Eng., 1974; May: 126–128.
G.F. Hewitt, G.L. Shires, T.R. Bott, Process Heat Transfer,
CRC Press, Inc., USA, 1994; p.1042.
D.R. Woods, Process Design and Engineering Practice, PTR
Prentice Hall, USA, 1995; p.543.
Asia-Pac. J. Chem. Eng. 2007; 2: 575–584
DOI: 10.1002/apj
Документ
Категория
Без категории
Просмотров
0
Размер файла
174 Кб
Теги
exchanger, fuel, heat, stud, network, industries, wastage, energy, case, minimisation, designчan, improve
1/--страниц
Пожаловаться на содержимое документа