Desalination: Muhammad Ahmad Jamil, Bilal Ahmed Qureshi, Syed M. Zubair

DES-13087; No of Pages 11
Desalination xxx (2016) xxxxxx
Contents lists available at ScienceDirect
Desalination
journal homepage: www.elsevier.com/locate/desal
Exergo-economic analysis of a seawater reverse osmosis desalination plant with

various retrot options
Muhammad Ahmad Jamil, Bilal Ahmed Qureshi, Syed M. Zubair
Mechanical Engineering Department, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia
H I G H L I G H T S
Seawater reverse osmosis desalination plant with four different retrot options is studied.
First- and second-law analyses are carried out to estimate energy requirements and second-law efciency.
The product cost is compared by performing exergo-economic analysis using reliable seawater properties.
Analysis revealed that with a pressure exchanger, energy consumption can be reduced by 24%.
It is also shown that post-treatment and distribution sections increase the product cost by about 20%.
a r t i c l e
i n f o
Article history:
Received 5 May 2016
Received in revised form 22 September 2016
Accepted 28 September 2016
Available online xxxx
Keywords:
Reverse osmosis
Seawater
Retrot
Pressure exchanger
Exergo-economic analysis
a b s t r a c t
The current study is focused on carrying out exergo-economic analysis of a seawater reverse osmosis (SWRO) desalination plant. The main objective is to compare the performance as well as the product cost of an existing
SWRO plant, including post-treatment and distribution sections, for four different retrot options made by coupling high-efciency pressure exchangers (PXs) in place of conventional energy recovery turbines. For this purpose, rst- and second-law analysis is carried out to estimate the energy requirements and second-law efciency
for each retrot option. Finally, the product cost is compared by performing an exergo-economic analysis using
appropriate seawater properties for the calculations. The analysis revealed that, by introducing a PX, the specic
energy consumption (SEC) can be reduced by about 24%; thus, increasing the second-law efciency. Besides this,
it is also demonstrated that the addition of post-treatment and distribution sections enhances the product cost by
almost 20%. Furthermore, the study suggested that using a booster pump with a PX (as used in retrot # 3) is best
suited for enhancing the plant capacity compared to retrot # 4 in which a PX is used in place of the pump. It has
the least product water cost among all the options discussed.
2016 Elsevier B.V. All rights reserved.
1. Introduction
Reverse osmosis (RO), a membrane-based desalination system is
one of the most frequently used techniques for treatment of seawater.
From 1970 till today, this technology has been widely used, studied
and improved over the time [1]. It has lower start-up time, decreased
environmental impacts (in terms of emissions) and easier operation
and maintenance. Energy analysis of RO systems operating under different capacities with and without energy recovery devices (ERDs) reveals
that their energy consumption can be greatly reduced by coupling ERDs
[26]. Coupling of Pelton turbines (as energy recovery turbines ERTs)
with RO systems is one of the oldest energy recovery methods [7,8]. Isobaric pressure exchangers (PXs) are relatively modern and better
Corresponding author.
E-mail address: smzubair@kfupm.edu.sa (S.M. Zubair).
devices in this regard [9]. A detailed discussion of the working and selection of ERDs is carried out by various investigators [1015].
Besides this, exergy analysis has been used as one of the most important tools by researchers [16,17] frequently to identify the components
with the greatest exergy destruction. Cerci [18] and Aljundi [19] analyzed two different RO plants using actual plant data and reported the
throttling valves and membrane modules to be the primary locations
for exergy destruction. Romero et al. [20] carried out a similar study
for a complete plant including pre-treatment, post-treatment and distribution sections. The above studies proposed that the second-law efciency of the plants can be improved by installing pump-motors
equipped with variable frequency drives and replacing throttle valves
on the brine stream with a PX.
Another useful way of analyzing the desalting systems is to combine
the exergy and cost analysis known as exergo-economic analysis.
Lozano and Valero [21] presented the theory of exergetic costs which
is considered to be one of the major approaches in this eld. Based on
http://dx.doi.org/10.1016/j.desal.2016.09.032
0011-9164/ 2016 Elsevier B.V. All rights reserved.
Please cite this article as: M.A. Jamil, et al., Exergo-economic analysis of a seawater reverse osmosis desalination plant with various retrot
options, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.09.032
M.A. Jamil et al. / Desalination xxx (2016) xxxxxx
this theory, Romero et al. [22] carried out an exergo-economic analysis

of an RO plant and reported the product cost to be 0.70 /m3. El-Emam
and Dincer [23] performed a similar analysis for different seawater salinities and estimated the product cost to be 2.45 $/m3 for a salinity of
35 g/kg. Spiegler and El-Sayed [24,25] contributed signicantly to the
eld of thermo-economics by developing the correlations for the rate
of xed cost of various components of desalination systems. They suggested that the main focus should be on the exergy destruction which
constitutes mainly the operating resources of any desalination system
rather than the making resources (xed cost). Some studies [2628]
were focused on analyzing solar-powered desalination systems. The
unit product cost for a small-scale solar-powered membrane distillation
unit was reported to be 15 $/m3 by [26] which is much higher than conventional systems. However, the unit product cost for a large scale PV/
RO system was estimated as 1.3 $/m3 by [27], which is slightly higher
than conventional systems (0.75 $/m3) due to higher electricity cost.
Penate and Rodriguez [29] proposed and analyzed four different retrot
options to provide upgradation opportunities for existing SWRO plants
working with conventional ERTs. The results of energy, exergy and thermo-economic analysis for all the retrot options were compared to
identify the best one with a minimum product cost.
1.1. On exergy calculation model
Fitzsimons et al. [30] examined and compared six different exergy
calculation models and showed that these models affect the nal results
signicantly. The study suggested that, among these, the electrolyte
model approach is not suitable because seawater is not an ideal mixture.
Similarly, the approaches used by Cerci [18] and Drioli et al. [3133] are
not suitable for desalination system analysis because of ideal mixture
assumptions and specic separation assumptions, respectively.
Sharqawy et al. [34] functions and Pitzer et al. [35,36] equations can
be used to calculate thermodynamic properties of seawater and other
electrolytes, respectively. A similar issue regarding the denition of second-law efciency is highlighted by various authors [17,3739] in their
studies. Some of them [14,40] dene it as the ratio of the total exergy
leaving to the total exergy entering the system, while others [16,17] as
the ratio of product to fuel exergies. Qureshi and Zubair [41] discussed
the applicability of these denitions and suggested that the second
one is more appropriate for desalination systems.
Based on the above discussion, the current study is focused on
reassessing and improving the work of Penate and Rodriguez [29] by
considering the following: (a) post-treatment and distribution sections
in the current analysis, (b) use of reliable and updated seawater properties recently compiled by Sharqawy et al. [34], (c) an appropriate denition of the second-law efciency suggested by Qureshi and Zubair [41],
and (d) plant performance as a function of important input parameters
such as unit electricity cost, feed salinity and high-pressure pump (HPP)
efciency.
i.e., P0 = 101.325 kPa, T0 = 20 C, S0 = 35 g/kg and operating temperature is considered constant throughout the system, (b) an overall pressure drop in RO modules, pipes and valves is considered to be 160 kPa,
(c) feed water pressure at HPP inlet is taken as 351.325 kPa and the recovery ratio is 45%, (d) effect of permeate back pressure, reverse salt diffusion, concentration polarization and system leakages are considered
negligible, (e) thermo-physical properties of seawater are based on
the correlations provided by Sharqawy et al. [34], and (f) efciencies
of the various components are, HPP = 78 % , BP = 77 % , FP = 78 % ,
DP =78% , Motor = 92 % , PX = 90% [29].
For numerical simulation, engineering equation solver (EES) software is used with updated seawater properties compiled by Sharqawy
et al. [34].
2.1. First-law analysis
To carry out the rst-law analysis, the mass balance (Eq. (1)) and the
solution balance (Eq. (2)) are applied. For a steady-state system, these
can be expressed as,
X
_
m
_
m
out
in
_
mS
_
mS
out
in
Pump and turbine work is calculated using Eq. (3) and Eq. (4), respectively as,
QP
WPump
Pump
WTB TB QP
The PX efciency is described as [41,42]:
PX
Q B;o P B;o Q F;o P F;o
Q B;i P B;i Q F;i P F;i
Specic energy consumption (SEC) is one of the important parameters for comparing plants working under different capacities because it
compares the energy requirement for a unit product. It can be expressed
as [41]:
SEC
Win
X
3600
Qp
out
2. System description and modeling

2.2. Second-law analysis
The system under consideration consists of a 10,000 m3/d seawater
reverse osmosis (SWRO) plant equipped with two membrane modules
of the same capacity coupled with two identical ERTs. The schematic for
this conguration is shown in Fig. 1, in which one HPP per train is used
to raise the feed pressure. The system is analyzed under four different
possible retrot options, calculations for each retrot option are performed and the results are compared with the standard conguration.
In the rst two retrot options, the plant capacity remains the same
while the focus is to minimize the energy consumption by replacing
ERTs with a PX. The last two options are proposed to increase the
plant capacity as well as minimize the energy consumption. The data
used for analysis of the plant is listed in Table 1. The analysis presented
in the paper is based on the following assumptions that are also considered by [29,41]: (a) the dead state is taken as the conditions of the feed,
This analysis measures the extent of irreversibility in terms of exergy

destruction, which is calculated by applying exergy-balance on each
component, separately:
X
_
X
fuel
X_ X_ D X_ L
products
The second law efciency is calculated, as described in [17,41]:
II
_ l; min
W
_ in
W
Fig. 1. Schematic of the standard SWRO plant.
3.1. Retrot options for constant capacity
Table 1
Operational data for standard conguration SWRO plant [29].
Parameter
Flow (m3/h)
Pressure (kPa)
Total feed
Feed per train
HPP inlet
Permeate stream per train
Total permeate
Brine stream/Turbine inlet
Brine discharge/Turbine Outlet
926
463
463
208
416
255
255
101.325
5961.325
351.325
101.325
101.325
5801.325
101.325
The concept of least work of separation and the minimum of least

work of separation is well explained by Mistry et al. [17]. The latter
can be written as:
_ l; min
W
X
_
X
X_
products
fuel
3. Retrot options
The retrot options proposed in this study are divided into two
categories:
These options are proposed with an aim to reduce energy consumption when the total plant capacity is to be maintained same as in the
standard conguration. In this category, the standard plant is retrotted
with high-efciency ERDs as explained below.
3.1.1. Retrot option # 1

It consists of one HPP instead of two (used in the standard conguration) and the two ERTs are replaced by a high-efciency isobaric energy recovery device. One HPP is sufcient for both the trains so this
conguration requires replacement of the HPP motor with a slightly
higher capacity and, thus, requires minor affordable changes in the electric wiring and other protection systems. The feed pump does not require any alteration because the same amount of feed water is to be
pumped at the same pressure. Neither the recovery ratio nor the ow
rates are modied. However, this conguration requires installation of
a booster pump (BP) for which new channeling of the feed water and
brine is required. Civil work is not required that much because the
ERD can be installed in the space left by ERTs. However, piping needs
to be modied because the ows are distributed over the plant in different ways as shown in Fig. 2. The feed-water is pumped to both the HPP
Fig. 2. Schematic of retrot option # 1.
Table 2
Operational data for SWRO plant retrot option # 1 [29].
Table 3
Parameter
Flow (m3/h)
Pressure (kPa)
Parameter
Flow (m3/h)
Pressure (kPa)
Total feed
Feed per train
HPP inlet
ERD feed
Total permeate
Brine stream/ERD brine inlet
Brine discharge/ERD brine outlet
926
463
416
510
208
416
510
510
101.325
5961.325
351.325
351.325
101.325
101.325
5761.325
221.325
Total feed
Feed per train
Low pressure BP inlet
HPP inlet
ERD feed inlet
Total permeate
926
463
416
463
510
208
416
510
510
101.325
5961.325
351.325
351.325
351.325
101.325
101.325
5801.325
101.325
and ERD inlet and its pressure is raised as it passes through these devices. Water from both circuits combine at state 8 and then distributed
evenly to each train. Permeate from both the trains is obtained as shown
in the gure. The high-pressure brine stream from each train is directed
to the ERD where it loses its pressure energy to raise the pressure of incoming feed water and then rejected back to the sea. The operational
data for this conguration is given in Table 2.
It consists of an isobaric pressure exchanger with a BP to raise the
pressure of the feed water and is proposed to avoid replacement of
the existing motors coupled to the HPP. One BP is also installed prior
to the HPP to maintain the required pressure, which avoids substantial
electrical modications by keeping the ow rates and the plant-capacity
constant. Flow arrangements remain same as discussed in the previous
option and schematic is shown in Fig. 3. Table 3 summarizes the operational data for this retrot option.
3.2. Congurations to increase the plant capacity
These options are proposed with an aim to increase the plant capacity by introducing high capacity RO trains with a higher number of pressure vessels. The plants are also retrotted with high-efciency ERDs to
minimize the energy consumption. In the current study, only one train
is shown for the sake of analysis because the two trains are identical.
These types of retrots are recommended for the cases where new investments can be made to upgrade the plants and sufcient space is
available to accommodate high capacity RO trains.
In this case, each HPP is retrotted to allow 35% reduction in the feed
ow rate because an isobaric ERD and a BP is used to manage the pressure of the remaining feed. The schematic for this option is shown in
Fig. 4. The new train consists of 23 pressure vessels and works with
the same recovery rate and the train capacity is increased from 5000
m3/day to 7200 m3/day. Pipe diameters are to be modied slightly
and more parallel pipes are to be installed. A new pipe is to be installed

for the low- and high-pressure feed ow to the ERD and BP. No doubt,
this conguration consumes more net energy but provides 40% more
permeate which reduces the SEC. The operational parameters used for
dening the new capacity are given in Table 4.
This retrot option is a modied form of the previous one (see Fig. 5)
in which an isobaric ERD is used as a second stage HPP. A portion of the
feedwater is pressurized by the HPP while the rest of it gets pressurized
by the ERD that uses the brine (from the rst stage) as the working uid.
This retrot produces about 50% more water than the standard conguration. It allows for the existing HPP and RO train to be used without any
new installation of energy consuming devices. The operational data for
this retrot is given in Table 5.
A sample calculation for the rst- and second-law analyses (of the
2nd retrot option) is given in Appendix A.
4. Exergo-economic analysis
For exergo-economic analysis, the plant is divided into three subsystems for each retrot option. Subsystem 1 consists of a feed pump and a
pre-treatment section and subsystem 2 consists of HPPs, BP (if needed)
and ERDs. Subsystem 3 includes the post-treatment and distribution facility. The schematic of subsystems for this analysis is shown in Fig. 6.
The rst step here is to calculate the exergy of each stream which is
the sum of physical and chemical exergies [34]. The other calculations
depend on the model used. For instance, the model used in the current
paper, which is based on the approach used by Romero et al. [22], requires the calculation of exergy destruction in each component for the
cost analysis. The only difference is that, in [22], the analysis for each
component is carried out separately but, in our case, the components
are combined in the form of subsystems. The calculations are given in
the following sections.
Table 4
Parameter
Flow
(m3/h)
Pressure
(kPa)
Total feed
Feed per train
HPP inlet
ERD feed inlet
Total permeate
1334
667
300
367
300
600
367
367
101.325
5961.325
351.325
351.325
101.325
101.325
5761.325
101.325
The stream leaving subsystem 1 will have certain pressure with reference to the dead state, so it will have certain exergy rate. Its unitary
exergy cost can be expressed as,
C x1
WSubsystem1
X 1
12
where WSubsystem1 represents the work supplied to the feed pump and
pre-treatment unit.
Based on the assumption stated above, the exergy cost of the blowdown is taken as 0 because it has no further utility.
C x2 0
13
4.1. Exergy costs of ow streams

The exergetic cost of any stream is based on the exergy required to
produce it. In the present study, the exergy cost calculations are based
on three assumptions [20]: (a) exergy cost of any input stream (from
the environment) is equal to its exergy value, (b) exergy cost of any useless ow (such as blowdown) is considered as zero, and (c) inlet and
outlet components of any fuel have the same unitary exergy cost.
The unitary exergy cost (a dimensionless parameter) of a stream
represents the ratio of its exergetic cost to power input. It tells us
about the exergy power required to produce that exergy stream. For different streams, the value of the exergy rate and unit exergetic cost are
different and can be calculated, as explained below.
The intake stream (0) is taken as the dead state so its specic exergy,
exergy rate and unitary exergy cost will be zero.
C x0 X0 0
11
The product stream has certain exergy rate with reference to the
dead state. Its unitary exergy cost can be calculated as
C x3
W Subsystem1 W Subsystem2 W Subsystem3
X3
W Subsystem;tot
X 3
14
where WSubsystem;tot represents the total work supplied to produce this

product stream. This includes work input to the feed pump, HPP and
BP (if required). The stream exergy (fuel, product, and losses) values
of the subsystems (see Fig. 6) are given in Table 7.
4.2. Exergo-economic costs of the ow streams
The next step in this analysis is to calculate the exergo-economic
costs of the ow streams. Tables 810 show the values used in the
exergo-economic analysis taken from the literature [29]. The unit
exergo-economic cost or the unit cost, in c$/MJ, of a stream can be
5. Results and discussion
Table 5
Operational data for SWRO plant retrot # 4 [29].
Parameter
Flow
(m3/h)
Pressure
(kPa)
Total feed
Feed per train 1
Feed per train 2
HPP inlet
ERD feed inlet
Permeate stream per train 1
Permeate stream per train 2
Total permeate
Brine train 1/ERD brine inlet
Brine train 1/ERD brine outlet
Brine train 2
1436
463
255
463
255
208
112
640
255
255
143
101.325
5961.325
5711
351.325
351.325
101.325
101.325
101.325
5761.325
101.325
101.325
calculated by applying the general formula which states that the unit
exergo-economic cost of any product stream is equal to sum of the
costs of fuel streams and xed cost of the components producing it.
This is expressed as [22]:
Cp C f
f
Xf
Xp
15
Xp
The rate of exergo-economic cost C p (in c$/s) is calculated as [22]:

C p C i C electricity X Z
The cost of fresh (desalted) water, in c$/m , for each retrot option is
calculated as:
Cp
Qp
This section compares the power requirements, specic energy consumptions and second law efciencies of the systems that are discussed
in the previous section.
5.1.1. Standard conguration plant

Referring to Table 6, it is obvious that, for the standard plant, the net
energy requirement reduces to 1573 kW from 2251 kW after installation of ERTs. SEC for the standard conguration is 3.78 kWh/m3. This
is expected to be reduced after installation of high-efciency ERDs. It
is important to emphasize that selection of any retrot option requires
the following parameters to be analyzed critically:
(a)
(b)
(c)
(d)
(e)
overall plant efciency,

new operational data and recovery rate,
civil works and hydraulic network,
capital investment associated with retrotting, and
management of the new system in terms of operation and maintenance, etc.
Keeping the above facts in mind, four different possible retrot options are analyzed for comparison purpose.
16
5.1. First and second law analysis
17
A sample calculation for the exergo-economic analysis is given in

Appendix B.

Table 6 illustrates that, for the same plant capacity, retrot option #
1 shows a considerable reduction in energy requirements compared to
the original plant. The table shows there is no change in the feed
pump work because this section is same for all the retrot options having same plant capacity. However, coupling of an isobaric PX reduces
overall energy consumption by 20.63% compared to the original plant.
Additionally, second-law efciency is also increased from 19.47% to
24.53% for this option. Therefore, one can say that this retrot is more
efcient from second-law viewpoint compared to the standard plant
because of better energy recovery.
Fig. 6. Schematic for exergo-economic analysis.
Table 6
First law analysis results.
Parameter
Standard plant
Retrot option #1
Retrot option #2
Retrot option #3
Retrot option #4
Feed pump work (kW)

HPP work (kW)
BP work (kW)
Product pump (kW)
Pelton turbine work (kW)
Total energy requirement (kW)
SEC (kWh/m3)
Energy saving (%)
II (%)
82.44
925
n/a
209.4
339.2
1573
3.78
n/a
19.51
82.44
816.1
57.03
209.4
n/a
1248
3.00
20.63
24.53
82.44
594.1
297.13
209.4
n/a
1268
3.05
19.31
24.15
118.8
1201
47.66
302.1
n/a
1788
2.98
21.15
24.70
127.8
1850
n/a
322.2
n/a
2472
3.86
2.18
19.06
n/a stands for not applicable; * represents energy produced by the system; and ** represents increase in SEC compared to the base conguration.

This retrot is obtained by slightly modifying the previous one with
an aim to avoid major replacements. The input power is distributed differently in this retrot because of an additional BP. From Table 6, it can
be seen that SEC for this option is reduced by 19.31% compared to the
standard plant. The second-law efciency, for this case, is 24.15%,
which is slightly lower than the previous retrot but higher than the
standard one. So, we may say that the energy consumption for both of
these retrot options lies in the same range because of the same plant
capacity. Hence, overall cost for modication is the only deciding factor
among these two.
The energy requirements for components like feed-pump, HPP and
distribution pump are higher for this retrot option because it has
higher plant capacity compared to the standard conguration. When
we compare the results (refer to Table 6), this retrot has the lowest
SEC value of 2.98 kWh/m3 and highest second-law efciency of 24.7%.
It gives energy saving of 21.15% compared to all other retrots. Therefore, this retrot is recommended where higher capacity RO trains can
be accommodated.
It is another retrot option that can also be used to enhance the plant
capacity. However, the analysis shows, (refer to Table 6) that this option
is not as efcient energetically as the previous ones. It can be seen from
this table that SEC for this retrot is 3.86 kWh/m3, which is the highest
among all, including the standard conguration. It shows an increase of
about 2% in the energy requirement compared to the base system. In addition, second-law efciency has the least value of about 19%. So, we can
say that this retrot is not suitable both from rst- and second-law
considerations.
5.2. Exergo-economic analysis
Table 11 summarizes both exergy values and unitary exergy costs of
ow streams. The exergy rates of streams leaving subsystem 1 and 2 are
same for the rst two retrot options because of their same capacity,
while these are different for the last two because of different capacities.
It is important to emphasize that unitary cost of stream 1 is same for all
the retrot options since subsystem 1 consists of the same equipment in
all cases. The difference, however, in work supplied due to different
mass ow rates is compensated by the product stream exergy and
Table 7
Stream exergies of the subsystems.
Stream
Subsystem 1
Subsystem 2
Fuel (f)
Product (p)
X_ 0 X_ 4
X_ 1
X_ 1 X_ 5
X_ 3
Losses (L)
X2
their ratio remains same. However, the unitary exergy cost of product
stream is different for all retrot options since different types of equipment are attached to each system with dissimilar capacities. Table 12
summarizes the nal product cost for all retrot options. It can be
seen that retrot option # 3 has the least product cost of 70.34 c$/m3
followed by option # 2 with a value of 72.24 c$/m3. Among the rst
two retrots, the second one has lower cost compared to the rst one.
This is because no change in HPP or any other component is required
and only a BP is introduced to meet the demand. We note that option
# 4 has the highest product cost of 83.07 c$/m3. This is primarily due
to the fact that it needs larger modications and has the highest energy
consumption and irreversible losses in the system components,
resulting in the lowest second-law efciency.
5.3. Comparison with literature
To assess the importance and effectiveness of the modications investigated in the current work, it is necessary to compare the present results with the one published in literature [29]. Figs. 7 to 9 compare the
SEC, second-law efciency and product cost for all the plant congurations with and without, post-treatment and distribution sections. The
Table 8
Data used in economic analysis [29].
Parameter
Value
Taxes
Amortization period
Lifetime
Loan Interest rate
Mean ination rate
Annual increasing of capital goods
above or below ination rate
Years in which the devices should be
replaced
Electricity cost
Annual increasing of O & M costs
Annual increasing of product cost
Annual availability
0.35
8 years
15 years
0.06
0.02
0.00
Intake and pumping once during lifetime,
while membrane every ve years
0.1344 ($/kWh)
0.00
0.00
0.95
Table 9
Input data costs for each retrot options analyzed.
Parameter
Retrot
option #1
Retrot
option #2
Retrot
option #3
Retrot
option #4
Subsystem 1 cost
Subsystem 2 cost
Specic O & M Cost
(insurance, labor,
overheads,
breakdowns, fuel
excluded)
0.2274 M$
0.2274 M$
0.4558 M$
0.4872 M$
1.1268 M$
1.167 M$
1.467 M$
1.596 M$
0.1512 $/m3 0.1512 $/m3 0.1456 $/m3 0.1568 $/m3
Table 10
Effective rate of xed costs for different subsystems [29].
Retrot option #1
Subsystem (in c$/s)
(1)
0.3523
Retrot option #2
Subsystem (in c$/s)
(2)
1.932
(3)
1.364
(1)
0.375
Retrot option #3
Subsystem (in c$/s)
(2)
1.818
(3)
1.364
(1)
0.772
Retrot option #4
Subsystem (in c$/s)
(2)
2.386
(3)
1.790
(1)
0.829
(2)
2.614
(3)
1.960
Table 11
Stream exergy values and unitary exergy costs of streams.
Parameter
X_ 0
(kW)
X_ 1
(kW)
X_ 2
(kW)
X_ 3
(kW)
Cx0
(--)
Cx1
(--)
Cx2
(--)
Cx3
(--)
Retrot option # 1
Retrot option # 2
Retrot option # 3
Retrot option # 4
0
0
0
0
64.31
64.31
92.4
99.72
0
0
0
0
306.2
306.2
441.6
473.3
0
0
0
0
1.282
1.282
1.282
1.282
0
0
0
0
3.369
3.433
3.342
4.54
Table 12
Product costs for the retrot options.
Parameter
Retrot option # 1
Retrot option # 2
Retrot option # 3
Retrot option # 4
Unit exergo-economic cost, C (c$/MJ)

Rate of exergo-economic cost, C_ p (c$/s)
35.85
8.37
35.90
8.34
37.31
11.73
42.07
14.78
Fresh water cost, p (c$/m3)
72.48
72.24
70.34
83.07
major outcomes of the comparisons can be summarized in the following

paragraphs.
Fig. 7 shows that SEC results for the current analysis (without additional sections) are in excellent agreement with the work of Penate and
Rodriguez [29] and conrms the model validity. It can be seen that after
introducing post-treatment and distribution sections, SEC values show
an increase of 15 to 20% due to additional energy consuming components. It is, however, important to note that second-law efciency
(refer to Fig. 8) values, calculated by using updated seawater properties
and an appropriate denition of the efciency, show a 50 to 60% decrease compared to the values reported in the literature for all the
cases invesitgated. In addition, second-law efciency values decrease
by 13 to 16% after incorporation of the post-treatment and distribution
sections.
Fig. 9 shows that the unit product cost (in c$/m3) calculated by the
current approach is higher than the one reported in [29]. It is, however,
close to the one reported in [22,43]. The possible reasons for this difference include the use of updated seawater properties and additional
post-treatment and distribution sections that are considered in the
present investigations. It should be noted from Fig. 9 that coupling of
the post-treatment and distribution sections increases the nal product
cost by 25 to 35% for all the cases.
5.4. Parametric investigation

In this section, the effect of input parameters like feed salinity,
pump efciency and input energy cost on the plant performance
are discussed. All the operating conditions that are used in this
study are mentioned in the corresponding gures.
Fig. 10 shows that, by introducing high-efciency HPPs, SEC of the
plant can be greatly reduced. The base system shows an abrupt reduction in SEC because two HPPs are involved and have the highest
energy consumption. While in the retrot options, due to the presence of PXs, the energy consumption of HPPs is already lower than
the base system. Thus, the increase in their efciency does not reduce the SEC as abruptly as in the base system. Similarly, Fig. 11
shows a signicant rise in second-law efciency for the base system
with an increase in HPP efciency compared to other retrot options
for the same reason. Though it is obvious that by introducing high-efciency components, lower SEC, and higher second-law efciency
can be obtained, the current analysis is carried out to give an idea
that how and how much the performance of a plant changes with
these parameters.
60
4.0
50
3.0
II (%)
SEC (kWh/m3)
3.5
2.5
2.0
1.5
40
30
20
1.0
10
0.5
0.0
0
Base System
Retrofit # 1
Retrofit # 2
Retrofit # 3
Retrofit # 4
Base System
Retrofit # 1
Plant Type
Literature
Current analysis
Current analysis with post-treatment & distribution
Fig. 7. Specic energy consumption for different plant congurations.
Retrofit # 2
Retrofit # 3
Retrofit # 4
Plant Type
Literature
Current analysis
Fig. 8. Comparison of second-law efciency for various plant congurations.
Specific product cost, p (c$/m3 )
90
80
70
60
50
40
30
20
10
0
Retrofit # 1
Literature
Retrofit # 2
Current analysis
Retrofit # 3
Retrofit # 4
Plant Type
Fig. 9. Comparison of specic product cost for various plant congurations.

Fig. 12. Second-law efciency vs feed salinity.
6. Concluding remarks
A seawater reverse osmosis plant discussed by Penate and
Rodriguez [29] is re-evaluated by using updated seawater properties
and an appropriate denition of the second law efciency for four
different retrot options along with the base system. Furthermore,
the study is also updated by adding the post-treatment and distribution sections that were not considered in the previous investigation.
The study provides reliable information about improving the
existing SWRO plant in terms of energy consumption by introducing
high-efciency PXs as well as upgrading the plant capacity. The
major ndings of the present study can be summarized as:
Fig. 10. Specic energy consumption vs high-pressure pump efciency.
Feed salinity is another parameter that affects the second-law efciency of the plants as shown in Fig. 12. When the feed salinity is increased, the recovery ratio decreases which is inversely proportion
to the input energy. The second-law efciency is obtained by dividing the minimum of least work of separation by input energy so it increases with the feed salinity.
It is important to understand the variation of the product cost
against the input energy cost because the price of electricity is different
in every locality. It can be seen from Fig. 13 that the nal product cost
increases linearly with the electricity cost. However, the rst and second
retrot options show an interesting shift at higher electricity cost. For
example, at an electricity cost 0.22 $/kWh, the second retrot option
has lower product cost while, for higher values of unit electricity cost,
the rst option gives better results from an economic standpoint.
motor = 92%, SF = 35g/kg, TF
B=
Compared to conventional energy recovery turbines, modern PXs

are more efcient and their installation resulted in 18 to 23%
reduction in SEC of the plant.
About 50 to 60% reduction in second-law efciency was observed
by using the updated and reliable seawater properties as well as
by using an appropriate relation for second-law efciency.
The installation of post-treatment and distribution sections increases the energy consumption as well as the product cost and decreases the second-law efciency of the plant. About 15 to 20%
increase in SEC, 25 to 30% increase in product cost and 13 to 16% reduction in second-law efciency is observed by introducing these
sections. This fact suggests that these should not be neglected for
a reliable design and analysis of desalination plants.
Among all the retrot options discussed in the paper, options 2 and
3 present better results compared to options 1 and 4 in terms of the
65g/kg
Fig. 11. Second-law efciency vs high-pressure pump efciency.
Fig. 13. Variation of the specic product cost vs electricity cost.
10
product cost.
For a constant plant capacity, option 2 is recommended for an electricity cost of 0.22 ($/kWh) since it has slightly lower product cost
compared to the rst option. While at higher electricity costs, the
rst option is more favorable.
Retrot option 3 is the best possible choice because of the lower
SEC and product cost. However, it requires higher capacity RO
trains which may not be affordable in every case.
Retrot option 4 shows the worst performance. It has the highest
product cost among all the retrot options because of higher SEC
and lower second-law efciency values.
The present study clearly shows that reliable seawater properties,
method of calculation and plant layout must carefully be selected
while analyzing any desalination system as they can affect the nal
results signicantly.
Nomenclature
C
unit exergo-economic cost (c$/MJ)
Cx
unitary exergy cost
C p
rate of exergo-economic cost (c$/s)
ESaving
energy saving (%)
_
m
mass ow rate (kg/s)
P
pressure (kPa)
Q
volume ow rate (m3/s)
S
salinity (g/kg)
T
temperature (C)
W
power requirement (kW)
X
exergy rate (kW)
Z
rate of xed costs (c$/s)
Acknowledgement
The authors acknowledge the support provided by King Fahd University of Petroleum & Minerals through the project IN151001.
Appendix A. First- and second-law analysis calculations
For sample calculations, retrot option # 2 is presented here because
it includes almost all the components that are discussed in this paper.
The pump work is given by,
QP
WPump
Pump
A 1
Now, we will calculate the individual pump works.

Feed pump work:
_ FP 0:2572 351:325101:325 82:44 kW
W
0:78
Low-pressure BP work:
_ LP;BP 0:1155 1951:325101:325 240 kW
W
0:77
BP work:
_ BP 0:14167 5961:3255651:325 57:03 kW
W
0:77
HPP work:
Greek letters
change in quantity
efciency
specic cost (c$/m3)
_ HPP 0:1156 5961:3251951:325 594:3 kW

W
0:78
Subscripts
0
dead state
B
brine
D
destroyed
F
feed
i
inlet
II
second law
in
input
l
least
L
loss
min
minimum
o
outlet
p
product
Rf#
retrot option
TB
turbine
Tot
total
_ DP 0:1155 1402101:325 192:599 kW

W
0:78
Abbreviations
BP
booster pump
DP
distribution pump
EES
engineering equation solver
ERT
energy recovery turbine
ERD
energy recovery device
HPP
high-pressure pump
LP
low-pressure pump
PVs
pressure vessels
PX
pressure exchanger
RO
reverse osmosis
SWRO
seawater reverse osmosis
SEC
specic energy consumption, (kWh/m3)
Distribution pump work:
With a motor efciency of 0.92, the total pump work comes out to
be,
_ tot 82:435 240 57:03 594:3 192:599 1267:8 kW
W
0:92
The SEC is given by,
SEC
W_ in
X _
Qp
3600
A 2
out
For retrot option # 2, it is found to be,

SEC
1267:8
3
3:05 kWh=m
416
The energy saving is then calculated as:

ESaving

3:783:05
100 19:31%
3:78
The second-law efciency is given as,
II
Wl; min
Win
A 3
For retrot option # 2, it is found to be,

II
306:2
100 24:15%
1267:4
Appendix B. Sample exergo-economics calculation

The sample calculations for exergo-economic analysis are given in
this appendix. In this regard, retrot option # 2 is presented here according to the procedure given in Section 4.
The unitary exergy cost of streams is expressed as,
C x#
WSubsystem#
B 1
X#
For stream 1,
C x1
89:603
1:393
64:31
For stream 3,
C x3
1267:8
4:13
306:8
Now, the unit exergo-economic cost, in c$/MJ, is described as
Cp C f
f
Cp
Xf
Xp
B 2
Xp
3:7931 105 100 89:603 0:3636 3:7931 105 100 968:51
64:31
64:31
306:2
1:818 3:7931 105 100 209:34 1:3636
35:92 c$=MJ
306:2
306:2
306:2
Now, the rate of product cost, in c$/s, is described as,

C p C i C electricity X Z
B 3
where C i represents the cost of inlet stream to that subsystem. This

gives,
C p 8:35 c$=s:
The fresh water cost, in c$/m3, is given by,
Rf #2;3
Cp
B 4
Qp
Therefore,
Rf #2;p Rf #2;3
8:35 3600
72:25 c$=m3
416
References
[1] R.W. Baker, Membrane Technology and Applications, 3rd ed. John Wiley & Sons, Ltd,
Chichester, UK, 2012.
[2] G.P. Narayan, R.K. McGovern, S.M. Zubair, J.H. Lienhard V, High-temperature-steamdriven, varied-pressure, humidication-dehumidication system coupled with reverse osmosis for energy-efcient seawater desalination, Energy 37 (2012) 482493.
[3] H. Cherif, J. Belhadj, Large-scale time evaluation for energy estimation of standalone hybrid photovoltaic-wind system feeding a reverse osmosis desalination
unit, Energy 36 (2011) 60586067.
[4] E.S. Hrayshat, Brackish water desalination by a stand alone reverse osmosis desalination unit powered by photovoltaic solar energy, Renew. Energy 33 (2008) 17841790.
11
[5] C. Fritzmann, J. Lowenberg, T. Wintgens, T. Melin, State-of-the-art of reverse osmosis

desalination, Desalination 216 (2007) 176.
[6] M. Li, Reducing specic energy consumption in reverse osmosis (RO) water desalination, Desalination 276 (2011) 128135.
[7] D.J. Woodcock, I.M. White, The application of Pelton type impulse turbines for energy recovery on seawater reverse osmosis systems, Desalination 39 (1981) 447458.
[8] S. Bross, W. Kochanowski, SWRO core hydraulic system: extension of the SalTec DT
to higher ows and lower energy consumption, Desalination 203 (2007) 160167.
[9] A.S. Dundorf, J. Macharg, B. Sessions, T.F. Seacord, Optimizing Lower Energy Seawater Desalination, The Affordable Desalination Collaboration, in: IDA World Congr.
2009, 127.
[10] J.P. Macharg, Retro-Fitting Existing SWRO Systems with a New Energy Recovery Device, San Leandro CA, USA, 2002.
[11] B. Penate, J.A. De La-Fuente, M. Barreto, Operation of the RO kinetic energy recovery
system: description and real experiences, Desalination 252 (2010) 179185.
[12] B.A. Qureshi, S.M. Zubair, Energy-exergy analysis of seawater reverse osmosis
plants, Desalination 385 (2016) 138147.
[13] B. Schneider, Selection, operation and control of a work exchanger energy recovery
system based on the Singapore project, Desalination 184 (2005) 197210.
[14] M.H. Sharqawy, S.M. Zubair, J.H. Lienhard V, Second law analysis of reverse osmosis
desalination plants: an alternative design using pressure retarded osmosis, Energy
36 (2011) 66176626.
[15] R.L. Stover, Seawater reverse osmosis with isobaric energy recovery devices, Desalination 203 (2007) 168175.
[16] Y. Demirel, Thermodynamic analysis of separation systems, Sep. Sci. Technol. 6395
(2010) 38973942.
[17] K.H. Mistry, R.K. McGovern, G.P. Thiel, E.K. Summers, S.M. Zubair, J.H. Lienhard V, Entropy generation analysis of desalination technologies, Entropy 13 (2011)
18291864.
[18] Y. Cerci, Exergy analysis of a reverse osmosis desalination plant in California, Desalination 142 (2002) 257266.
[19] I.H. Aljundi, Second-law analysis of a reverse osmosis plant in Jordan, Desalination
239 (2009) 207215.
[20] V. Romero-Ternero, L. Garcia-Rodriguez, C. Gmez-Camacho, Exergy analysis of a
seawater reverse osmosis plant, Desalination 175 (2005) 197207.
[21] M.A. Lozano, A. Valero, Theory of the exergetic cost, Energy 18 (1993) 939960.
[22] V. Romero-Ternero, L. Garcia-Rodriguez, C. Gmez-Camacho, Thermoeconomic analysis of a seawater reverse osmosis plant, Desalination 181 (2005) 4359.
[23] R.S. El-Emam, I. Dincer, Thermodynamic and thermoeconomic analyses of seawater
reverse osmosis desalination plant with energy recovery, Energy 64 (2014)
154163.
[24] K.S. Spiegler, Y.M. El-Sayed, The energetics of desalination processes, Desalination
134 (2001) 109128.
[25] Y.M. EI-Sayed, Thermoeconomics of some options of large mechanical vapor-compression units, Desalination 125 (1999) 251257.
[26] F. Banat, N. Jwaied, Economic evaluation of desalination by small-scale autonomous
solar-powered membrane distillation units, Desalination 220 (2008) 566573.
[27] G. Fiorenza, V.K. Sharma, G. Braccio, Techno-economic evaluation of a solar powered
water desalination plant, Energy Convers. Manag. 44 (2003) 22172240.
[28] A.S. Nafey, M.A. Sharaf, L. Garcia-Rodriguez, Thermo-economic analysis of a combined solar organic Rankine cycle-reverse osmosis desalination process with different energy recovery congurations, Desalination 261 (2010) 138147.
[29] B. Penate, L. Garcia-Rodriguez, Energy optimisation of existing SWRO (seawater reverse osmosis) plants with ERT (energy recovery turbines): technical and
thermoeconomic assessment, Energy 36 (2011) 613626.
[30] L. Fitzsimons, B. Corcoran, P. Young, G. Foley, Exergy analysis of water purication
and desalination: a study of exergy model approaches, Desalination 359 (2015)
212224.
[31] A. Criscuoli, E. Drioli, Energetic and exergetic analysis of an integrated membrane
desalination system, Desalination 124 (1999) 243249.
[32] E. Drioli, E. Curcio, G. Di Proo, F. Macedonio, A. Criscuoli, Integrating membrane
contactors technology and pressure-driven membrane operations for seawater desalination, Chem. Eng. Res. Des. 84 (2006) 209220.
[33] F. Macedonio, E. Drioli, An exergetic analysis of a membrane desalination system,
Desalination 261 (2010) 293299.
[34] M.H. Sharqawy, J.H. Lienhard V, S.M. Zubair, Thermophysical properties of seawater:
a review of existing correlations and data, Desalin. Water Treat. 16 (2016) 354380.
[35] K.S. Pitzer, Thermodynamics of electrolytes. I. Theoretical basis and general equations, J. Phys. Chem. 77 (2) (1973) 268277.
[36] K.S. Pitzer, J.J. Kim, Thermodynamics of electrolytes. IV. Activity and osmotic coefcients for mixed electrolytes, J. Am. Chem. Soc. 96 (1974) 57015707.
[37] A. Bejan, G. Tsatsaronis, M. Moran, Thermal Design and Optimization, JohnWiley &
Sons, Inc., NewYork, 1996.
[38] A. Bejan, Advanced Engineering Thermodynamics, 3rd ed John Wiley & Sons, Inc.,
New Jersey, 2006.
[39] K.H. Mistry, J.H. Lienhard V, Generalized least energy of separation for desalination
and other chemical separation processes, Entropy 15 (2013) 20462080.
[40] N. Kahraman, Y.A. Cengel, W. Byard, Y. Cerci, Exergy analysis of a combined RO, NF
and EDR desalination plant, Desalination 171 (2004) 217232.
[41] B.A. Qureshi, S.M. Zubair, Exergetic analysis of a brackish water reverse osmosis desalination unit with various energy recovery systems, Energy 93 (2015) 256265.
[42] Energy Recovery Inc. ERI Power Model. http://www.energyrecovery.com/resource/
power-model/ [accessed 18.09.16], (2016).
[43] V.G. Gude, Energy consumption and recovery in reverse osmosis, Desalin. Water
Treat. 36 (2011) 239260.

Desalination: Muhammad Ahmad Jamil, Bilal Ahmed Qureshi, Syed M. Zubair

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Desalination: Muhammad Ahmad Jamil, Bilal Ahmed Qureshi, Syed M. Zubair

Diunggah oleh

Hak Cipta:

Format Tersedia

DES-13087; No of Pages 11

Desalination xxx (2016) xxxxxx

Contents lists available at ScienceDirect

Exergo-economic analysis of a seawater reverse osmosis desalination plant with

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

this theory, Romero et al. [22] carried out an exergo-economic analysis

The PX efciency is described as [41,42]:

Q B;o P B;o Q F;o P F;o

Q B;i P B;i Q F;i P F;i

2. System description and modeling

This analysis measures the extent of irreversibility in terms of exergy

The second law efciency is calculated, as described in [17,41]:

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

Fig. 1. Schematic of the standard SWRO plant.

3.1. Retrot options for constant capacity

The concept of least work of separation and the minimum of least

3.1.1. Retrot option # 1

Fig. 2. Schematic of retrot option # 1.

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

and more parallel pipes are to be installed. A new pipe is to be installed

Fig. 3. Schematic of retrot option # 2.

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

Fig. 4. Schematic of retrot option # 3.

4.1. Exergy costs of ow streams

W Subsystem1 W Subsystem2 W Subsystem3

where WSubsystem;tot represents the total work supplied to produce this

Fig. 5. Schematic of retrot option # 4.

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

5. Results and discussion

The rate of exergo-economic cost C p (in c$/s) is calculated as [22]:

5.1.1. Standard conguration plant

overall plant efciency,

5.1. First and second law analysis

A sample calculation for the exergo-economic analysis is given in

5.1.2. Retrot option # 1

Fig. 6. Schematic for exergo-economic analysis.

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

Feed pump work (kW)

5.1.3. Retrot option # 2

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

Unit exergo-economic cost, C (c$/MJ)

Fresh water cost, p (c$/m3)

major outcomes of the comparisons can be summarized in the following

5.4. Parametric investigation

Current analysis with post-treatment & distribution

Fig. 7. Specic energy consumption for different plant congurations.

Current analysis with post-treatment & distribution

Fig. 8. Comparison of second-law efciency for various plant congurations.

Specific product cost, p (c$/m3 )

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

Fig. 9. Comparison of specic product cost for various plant congurations.

motor = 92%, SF = 35g/kg, TF

Compared to conventional energy recovery turbines, modern PXs

Fig. 11. Second-law efciency vs high-pressure pump efciency.

Fig. 13. Variation of the specic product cost vs electricity cost.

M.A. Jamil et al. / Desalination xxx (2016) xxxxxx

Now, we will calculate the individual pump works.

specic cost (c$/m3)

_ HPP 0:1156 5961:3251951:325 594:3 kW

_ DP 0:1155 1402101:325 192:599 kW

Distribution pump work:

For retrot option # 2, it is found to be,

The energy saving is then calculated as: