sustainability

Article
Techno-Economic Optimization of a Solar–Wind Hybrid System
to Power a Large-Scale Reverse Osmosis Desalination Plant
A. M. Soliman 1,2, *, Abdullah G. Alharbi 3 and Mohamed A. Sharaf Eldean 2

1 Department of Mechanical Engineering, College of Engineering, Jouf University, Sakaka 72388, Saudi Arabia
2 Department of Mechanical Engineering, Faculty of Engineering, Suez University, Suez 41522, Egypt;
swazy20@[Link]
3 Department of Electrical Engineering, College of Engineering, Jouf University, Sakaka 72388, Saudi Arabia;
agalharbi@[Link]
* Correspondence: amsoliman@[Link]

Abstract: Solar-wind hybrid systems have grown to become a pivotal option for powering membrane
desalination processes, especially because they have zero harmful emissions. In this work, solar
photovoltaic (PV) and horizontal wind turbine (HWT) systems were used to drive a reverse osmosis
(RO) desalination process to produce large-capacity fresh water. Moreover, an investigation into a
hybrid PV–HWT system combined with RO was also conducted. The proposed systems are compared
technically and economically with the solar organic Rankine cycle (SORC) for RO. Technical and
analytical optimization methods were performed to minimize the unit product cost (USD/m3 ).
The results revealed that photovoltaic-powered RO is recommended over wind energy operations.





 However, for large capacities, both thermal and wind farms dominate.
Citation: Soliman, A.M.; Alharbi,
Keywords: photovoltaic system; horizontal wind turbines; solar organic Rankine cycle; reverse osmosis
A.G.; Sharaf Eldean, M.A.
Techno-Economic Optimization of a
Solar–Wind Hybrid System to Power
a Large-Scale Reverse Osmosis
Desalination Plant. Sustainability 2021, 1. Introduction
13, 11508. [Link] The Middle East and North Africa (MENA) have the world’s lowest per capita avail-
su132011508 ability of water resources [1]. Fortunately, most MENA countries have a renewable energy
potential that encourages the application of solar and wind to drive desalination units.
Academic Editor: Renewable energy to drive the desalination units is a vital solution to water scarcity in
Pablo García Triviño
remote areas that lack conventional energy sources like heat and electricity grid [2]. Reverse
osmosis (RO) is a modern process to desalinate water for a wide range of applications [3].
Received: 14 September 2021
For this technique, mechanical energy in the form of a high-pressure pump, not
Accepted: 13 October 2021
thermal energy, is required. In recent decades RO has gone through a remarkable develop-
Published: 18 October 2021
ment. The number of RO plants has increased and the capacity of the units has reached
3,000,000 m3 /day [4]. The reliability and the availability of combining renewable energy
Publisher’s Note: MDPI stays neutral
resources such as solar and wind power are reasons for using RO instead of thermal distil-
with regard to jurisdictional claims in
lation processes. Photovoltaic (PV) panels are the common form of solar energy generation
published maps and institutional affil-
iations.
in RO units. Essam Mohamed et al. [5] investigated technically and economically a photo-
voltaic system to drive RO desalination. The productivity of that system was 0.35 m3 /day
with a specific power consumption of about (4.6 kWh/m3 ). The cost of fresh water was
about (EUR 15–20/m3 ), which is high because of the need for a large number of batteries to
stabilize the pressure and flow rate for the membranes [5]. An economic feasibility study
Copyright: © 2021 by the authors.
of PV power on RO within low specific power consumption was established by Helal [6]. It
Licensee MDPI, Basel, Switzerland.
was conducted using three alternative configurations of an autonomous PV–RO unit in the
This article is an open access article
United Arab Emirates (UAE). In that study, a diesel generator was used during periods of
distributed under the terms and
conditions of the Creative Commons
low sunlight. The productivity of PV-RO was not more than 20 m3 /day (10 h). In Helal’s
Attribution (CC BY) license (https://
study, the environmental impacts of diesel generator emissions were ignored. The technical
[Link]/licenses/by/ characteristics of PV-RO desalination systems and an economic comparison were presented
4.0/). by Manolakos [7]. The total peak power of the PV system was 846 W and consisted of

Sustainability 2021, 13, 11508. [Link] [Link]

Sustainability 2021, 13, 11508 2 of 20

18 Arco-Solar mono-crystalline PV panels. The productivity of the system was 0.1 m3 /h

and the specific energy consumption has been found to be in the range of 3.8–6 kWh/m3 . It
is found that the cost of 1 m3 of fresh water was about EUR 7.77. Ahmed [8] found that the
cost of the productivity for a PV-powered small-scale RO water desalination system is USD
3.73/m3 . In Morocco, Tzen [9] studied an autonomous PV–RO system for the potable and
other water needs of a rural community and found that the specific power consumption
(SPC) was 15 kWh/m3 ).
Wind energy is also used in this kind of operation, and some of the literature is
provided in this section. Liu et al. [10] studied a RO system driven by wind energy for
wastewater treatment without an economic study. An experimental RO plant connected
directly to a wind system without energy storage was presented by Pestana [11]. The system
was designed based on 21 kW of power, to produce 3.6 m3 /h. Dehmas et al. [12] found
that wind energy can successfully power a desalination plant. Their studies included the
economic analysis of 5 Bonus 2 MW wind turbines. Garcia-Rodriguez et al. [13] presented
the influence of the climate, nominal power of the wind turbine, salt concentration, design
arrangement, operating conditions, and plant capacity on the cost of fresh water. Romero-
Ternero et al. [14] presented exergo-economic analysis for wind-powered seawater RO
system the unit cost of freshwater was EUR 0.76/m3 . Wind RO plants have a wide range
capacity of 12–2500 m3 /day with nominal power of 4–200 kW [15]. Dimitriou et al. [16], and
Ruiz-García and Nuezb [17], investigated the performance of RO under variable operating
conditions, and another study by Ruiz-García and Nuezb [18] investigated the long-term
intermittent performance of a brackish water RO desalination plant.
From the literature review, the hybrid wind–PV system for driving RO has shown ad-
vantages: a wide range of capacities, minimal energy loss, easy maintenance and reliability.
However, such systems were not investigated from a problem optimization or techno-
economic perspective. Optimization criteria are very important because they reduce the
total power consumption on the RO high-pressure pump (HPP), thereby, reducing the area
of the solar field and the total cost. It is obvious from the literature that the possibility of
producing a large water capacity (>3000 m3 /day) is far away because of the cost limitations
of PV and wind technologies. Furthermore, electric storage for PV and wind power is still
unaffordable. In this study, under the steady regime, the design and optimization of the
various capacities of a PV-horizontal wind turbine (HWT)–RO system are investigated to
come to a clear decision about the feasibility of using PV and HWT regardless of location or
weather condition. Moreover, the comparison between the proposed systems and the solar
thermal power cycle of a RO operation is performed. The optimization of techno-economic
issues is the focus of this work to decrease power consumption and system costs. The work
proceeds as follows:
• Process configurations are presented, design limits are investigated, and the mathe-
matical model of the proposed systems is presented.
• The optimization method is performed to study the effect of cost minimization.
• A comparison of four cases (solar direct–three configurations vs. solar indirect–one
configuration) is performed.

2. The Process Configurations

The combination of solar power with RO was achieved by one of two methods. The
first is electrical (PV and HWT), and the second is via the thermal solar Rankine cycle. The
proposed configuration related to the electrical method was modeled using SDS software
and shown in Figure 1 [19]. SDS is a developed software library as a part of the REDS
program library developed by Sharaf et al. [16]. The model configuration contained a PV,
HWT, inverter unit, battery bank (storage unit) and a control room for power switching
between PV and HWT to operate the RO system.
 and shown in Figure 1 [19]. SDS is a developed software library as a part of the REDS
program library
and shown developed
in Figure bySDS
1 [19]. Sharaf
is aetdeveloped
al. [16]. Thesoftware
model configuration contained
library as a part of the aREDS
PV,
HWT, inverter
program unit,
library battery bank
developed (storage
by Sharaf et [Link])
[16].and
The amodel
control room for power
configuration switching
contained a PV,
Sustainability 2021, 13, 11508 between PV andunit,
HWTbattery
to operate 3 of 20
HWT, inverter bankthe RO system.
(storage unit) and a control room for power switching
between PV and HWT to operate the RO system.

Figure 1. The proposed system model browser of the PV–HWT–RO using the SDS program. The system contains: 1, HWT;
PV system;
Figure
Figure 3, control
1. The room
proposed unit;model
system 4, RO desalination plant
browser of the block.
PV–HWT–RO
PV–HWT–RO using the
using the SDS program.
program. The system contains: 1, HWT;
PV system;
PV system; 3,
3, control
control room
room unit;
unit; 4,
4, RO
RO desalination
desalination plant
plant block.
block.
The RO unit was operated by HWT, PV or both (PV–HWT). Concentrated solar
powerThe (CSP) ROisunit
RO shown
unitwas in
was Figure 2.
operated
operated byThis
by
HWT, configuration
HWT,PV orPVboth is presented
or both (PV–HWT).
(PV–HWT). basedConcentrated
on thermal
Concentrated solarpower
solar
power
to(CSP)
operate
power the isorganic
is(CSP)
shown shown Rankine
in Figure
in Figure cycle (ORC)
2. configuration
2. This This to generate
configuration powerbased
is presented
is presented to operate
based onthe ROpower
thermal
on thermal system.
powerto
This configuration
operate
to operate thethe
organic contained
Rankine
organic a cycle
cycle
Rankine parabolic
(ORC) to trough
(ORC) generate collector
power
to generate to(PTC)
power operatefield
thefor
to operate thethe
RO ROthermal
system. This
system.
power, a boilercontained
configuration
This configuration heat-exchanger
contained aunit
a parabolic (BHX),
troughpumps,
parabolic collector
trough ORC
(PTC)turbine,
collector field
(PTC) recuperator
for the thermal
field for theforpower,
regen-a
thermal
eration
boiler and energy
power,heat-exchanger recovery, a condenser
unit (BHX),
a boiler heat-exchanger pumps,
unit unit
(BHX), ORCfor heat
turbine,
pumps, rejection
ORC and
recuperator the pre-heating for of
for regeneration
turbine, recuperator salt
and
regen-
water
energystream
eration andathe
recovery,
and energy RO system.
condenser
recovery, unit Therminol-VP1
unit for[20]
for heat rejection
a condenser andwas
heat theused as heat
pre-heating
rejection and transfer
theof oil (HTO)
salt water
pre-heating stream
of salt
inand
thethe
water fourth
stream configuration.
RO system. Toluene
and theTherminol-VP1
RO system. was
[20]used as the
was used
Therminol-VP1 asworking
[20] heat substance
wastransfer
used as oil
heat through
(HTO) the(HTO)
in the
transfer oil ORC
fourth
[2,20,21].
configuration. Toluene was Toluene
in the fourth configuration. used as the
wasworking
used as substance
the working through the ORC
substance through[2,20,21].
the ORC
[2,20,21].

Figure
[Link]
Solar ORC underSimuLink
ORC model under SimuLinksoftware
softwarebrowser
browser
byby
thethe
useuse of SDS
of SDS program:
program: PTC,PTC,
field,
field,
BHX BHX
unit, unit,
pumps,pumps, turbine
turbine unit, unit,
heat heat exchanger
exchanger recuperator
recuperator and and
RO RO
plantplant
[Link].
Figure 2. Solar ORC model under SimuLink software browser by the use of SDS program: PTC,
field, BHX unit, pumps, turbine unit, heat exchanger recuperator and RO plant block.
3. Simulation Methodology
The calculated parameters were area, dimension, mass flow rate, and temperature.
For the desalination, the fresh water capacity was specified, after which the electrical load
was calculated for either PV or the HWT design specifications. The calculated parameters
Sustainability 2021, 13, 11508 4 of 20

and design limits of the proposed units are illustrated in Table 1. The mathematical model
representing the process configurations and validity is listed in the Appendix A.

Table 1. The specified parameters based on the design technique of modeling concept.

RO Model
Specified Calculated
Fresh water productivity, m3 /day Feed and brine mass flow rates, kg/s
Seawater temperature, ◦ C Pressure on the HPP, bar
Seawater salinity, ppm Average pressure, bar
HPP efficiency, % Product and brine salinities, ppm
Booster pump efficiency, % Salt rejection percentage, %
Membrane fouling factor, % HPP power, kW
Number of elements/number of
SPC, kWh/m3
pressure vessels
Element area, m2 Membrane area, m2
Recovery ratio, %
Pressure exchanger (PEX) efficiency, %
HWT Model
Specified Calculated
Starting and average wind speeds, m/s, and
Power/set, kW
air mass flow, kg/s
System total power, kW Rotor diameter, m
Ambient temperature, ◦C Hub height, m
Ambient pressure, bar Rotor speed, rpm
Swept area, m2
Axial force, kN and Torque, kN
Power coefficient, %
Number of wind turbines
Spacing between turbines in winddirection, m
Spacing cross turbines in wind direction, m
Total farm area, km2
PV Module
Specified Calculated
The open circuit voltage, V and the short
Solar radiation, W/m2
circuit current, A
Power/panel, W The maximum voltage and current
Total system power, kW The cell and module efficiencies, %
The number of cells and modules of the system
The module and system weights, kg, and
areas, m2
The battery bank capacity, A
Sustainability 2021, 13, 11508 5 of 20

Table 1. Cont.

S-ORC Model
Specified Calculated
Solar radiation, W/m2 Solar Fields, condenser, recuperator areas, m2
Parabolic trough collector (PTC) high
Solar field dimensions and design
temperature, ◦ C
Boiler heat exchanger (BHX) effectiveness, % Mass flow rates, kg/s
ORC turbine efficiency, % ORC mass flow rates, kg/s
Solar field mass flow rate, loop design, No.
Recuperator effectiveness, %
of collectors
Condensation temperature, ◦ C Heat rejected power, kW
Condenser effectiveness, % PTC pump power, kW
All thermo physical properties of all streams

4. Results and Comments

4.1. Non-Optimization Comparison Results
There are many ways to judge process performance results such as by field area, fresh
water production rate, salinity range or specific costs. The main performance indicator
studied in this work was the unit product cost (UPC) in USD/m3 . This indicator is very
important, giving an early evaluation before the optimization procedures. In this section,
the process results are obtained at an RO production rate of 3500 m3 /day as a pre-stage
before optimization. The costs of the system units are illustrated in the Appendix A. The
user has to specify total productivity and then all the required parameters will be calculated
relatively and iteratively.
Table 2 shows the data results for the HWT-RO system for 3500 m3 /day. The results
revealed that to generate 1131 kWe, two 600-kWe HWTs were used. HWT-600 kW was used
because the expected wind speed in the Suez Gulf region is suitable. The average wind
speed is 11–12 m/s. Technical analyses showed that the hub height was about 41 m with a
43 m rotor diameter. The wind farm occupies about 0.063 km2 of land; the total annual cost
is about 6.31 × 105 USD/year; and the unit product cost (UPC) is about 0.5541 USD/m3 .
Table 3 presents the results of the PV–RO system that produced the same amount
of fresh water (3500 m3 /day). The expected power load was 1131 kWe for the RO plant,
and a 220 W module is chosen for the PV site. In addition, 350 W/m2 of solar radiation
was chosen as a known input parameter to calculate the PV site area based on the worst
winter conditions [22]. Excess power in summer could be stored in other facilities such
as the control room cabinet or lighting issues in the plant. Compared to the HWT–RO,
the PV–RO had a lower UPC (USD 0.53 vs. 0.55/m3 ) and a smaller solar field site (about
0.02 vs. 0.06 km2 ). Based on current results, the PV–RO system is considered to have a
lower UPC than the HWT–RO system.
Table 4 shows the results of the hybrid system in which the power demanded by the
RO plant is equally divided between the HWT farm and the PV site. The results revealed
that the hybrid took up a larger area (HWT farm and PV site area) and the UPC increased to
USD 0.5744/m3 due to the HWT number increase. It is obvious from the related tables that
the UPC could be optimized to a minimum value with respect to the site area by selecting
the right PV module power or the number of wind turbines. Moreover, the PV–RO system
was considered attractive against the remaining systems according to the minimum area
required and UPC in USD/m3 . However, the 280 W module may consume a larger area
with a higher cost per site. Therefore, optimization should be implemented to identify the
greatest effect of each system on the UPC in USD/m3 .
Sustainability 2021, 13, 11508 6 of 20

Table 2. Preliminary data results for HWT–RO system for 3500 m3 /day.

Environmental Conditions
Ambient temperature, ◦ C 15 (winter)
Solar radiation, W/m2 350–400 (winter) [22]
Air pressure, bar 1.01
Seawater temperature, ◦ C 20
RO Plant Results
Specific power consumption, kWh/m3 7.68
Power, kW 1131
Feed mass flow rate, m3 /h 485.9
Production flow rate, m3 /h 145.83
Brine flow rate, m3 /h 340.1
Brine salinity, ppm 64180
Fresh water salinity, ppm 250
Salt rejection value 0.9944
RO high pressure, kPa 6850
HWT Farm Results
Total power for RO plant, kW 1131
Turbine power/wind power, kW 600/4762
Starting wind speed, m/s 4–11.95
Rated wind speed, m/s 17.62
Hub height, m 41.6
Rotor diameter, m 43.32
Swept area, m2 1474.08
Air mass flow rate, kg/s 3.173 × 106
Torque (Nm)/rpm 1.542 × 105 /37.15
No. of wind turbines 2
Farm area, km2 0.06368
Cost Results
Plant life time/interest rate 25/5%
DCC of wind turbines, USD 1.174 × 106
DCC of RO plant, USD 3.5 × 106
Annual total costs, USD/year 6.31 × 105
Unit product costs, USD/m3 0.5541

Table 3. Preliminary data results for PV–RO system for 3500 m3 /day.

Environmental Conditions
The environmental conditions Presented in Table 2
RO Plant Results
The RO results Presented in Table 2
PV Site Results
Open circuit voltage/short circuit current, V/A 58.6/8.49
Maximum voltage/maximum current, V/A 47.4/4.641
Module efficiency/cell efficiency, %/% 15.45/17.5
No. of cells per module/No. of total modules 96/5135
Module dimensions/width, m3 /mm 0.095/45
Net weight, kg 21.5
Cell area/module area, cm2 /m2 423.8/4.06
Total system area, km2 0.02089
Battery storage, Wh 3.766 × 106
Battery capacities, Ah 7.945 × 104
No. of batteries—12 volt system 4
Sustainability 2021, 13, 11508 7 of 20

Table 3. Cont.

Cost Results
Plant life time/interest rate 25/5%
DCC of PV, USD 7.912 × 105
DCC of RO plant, USD 3.5 × 106
Annual total costs, USD/year 6.099 × 105
Unit product costs, USD/m3 0.5305

Table 4. Preliminary data results for HWT–PV-RO system for 3500 m3 /day.

Environmental Conditions
The environmental conditions Presented in Table 2
RO Plant Results
The RO results Presented in Table 2
HWT Farm Results
Total power for RO plant, kW 1131
Turbine power/wind power, kW 100/596.2
Starting wind speed, m/s 5.6
Rated wind speed, m/s 14.77
Hub height, m 20.03
Rotor diameter, m 19.64
Swept area, m2 303.1
Air mass flow rate, kg/s 5467
Torque (Nm)/rpm 1.288 × 104 /74.14
No. of wind turbines 6
Farm area, km2 0.03924
PV Site Results
Open circuit voltage/short circuit current, V/A 58.6/8.49
Maximum voltage/maximum current, V/A 47.4/4.641
Module efficiency/cell efficiency, %/% 15.45/17.5
No. of cells per module/No. of total modules 96/5135
Module dimensions/width, m3 /mm 0.095/45
Net weight, kg 21.5
Cell area/module area, cm2 /m2 423.8/4.06
Total system area, km2 0.01045
Battery storage, Wh 1.88 × 106
Battery capacities, Ah 3.97 × 104
No. of batteries (12 volt system) 4
Cost Results
Plant life time/interest rate 25/5%
DCC of HWT, USD 1.1 × 106
DCC of PV, USD 3.958 × 105
DCC of RO plant, USD 3.5 × 106
Annual total costs, USD/year 6.604 × 105
Unit product costs, USD/m3 0.5744

4.2. Comparison Results Based on Optimization

It was becoming very hard to recognize the main cause for the UPC. Previous results
indicated that the PV–RO system had the lowest UPC; however, such a result needs more
investigation based on many factors such as the number of wind turbines, modules of watts
of PV, and RO operating conditions. For that purpose, technical and analytical optimization
techniques were implemented. The technical method was implemented for the RO part,
but the analytical method was implemented for the RO, HWT, and PV systems.
Sustainability 2021, 13, 11508 8 of 20

4.2.1. Technical Optimization Results

In this section, the solution to minimize the objective function (UPC, USD/m3 ) was
implemented for the RO part. Optimizing it influenced the HWT and PV. To minimize the
UPC, it was necessary to reduce the HPP power load and maintain the same fresh water
production rate. Increasing the number of stages decreased the HPP power load. For the
same case study (3500 m3 /day), the number of stages was increased to nine. Table 5 shows
that the increased number of stages decreased the power from 1131 to 917.47 kW (18–20%).
Then the calculated SPC became 6.3 vs. 7.7 kWh/m3 in the basic case. Based on a previous
study [3,21], increasing the number of stages decreased the HPP power load and the PEX
device dominated operations in RO plants, reducing power consumption by 60–65% [22].
Table 5 shows the data results of the RO plant for 3500 m3 /day based on the different
technical devices. Therefore, the PEX device is recommended in this study.

Table 5. Results comparison for different energy recovery devices.

Parameter: Power, kW SPC, kWh/m3 RO ∆P, bar Power Reduction, %

Basic 1131 7.7 68.66 –
Stages = 9 917.47 6.3 35.8 18–20%
PEX 380–394 2.704 68.74 60–65%

4.2.2. The Analytical Optimization Results

Based on the technical results in the previous section, the RO-PEX technique is rec-
ommended for the analytical method. Figure 3 shows the effect of the RO-PEX inlet
feed/splitter ratio and the total system productivity on the UPC. The figure addressed
the case of HWT-RO-PEX without the operation of the PV solar field. Decreasing the inlet
feed splitter ratio to 10% increased the demanded power by the HPP by more than three
times (390 to 1532 kW). The UPC was directly proportional to this effect, increasing to
0.683 USD/m3 at 3500 m3 /day. Furthermore, the effect on total productivity was also
notable with respect to the feed/splitter ratio. The number of HWTs in the wind farm was
a very important parameter that affected the UPC (Figure 3b). Increasing the number of
Sustainability 2021, 13, x FOR PEER REVIEW 9 ofare
wind turbines (low power per unit) increased the UPC. Therefore, fewer than two units 21

recommended for this operation.

Figure
Figure 3.3.(a)
(a)Effect
Effectof of feed
feed splitter
splitter ratioratio and RO-PEX
and RO-PEX productivity
productivity on the
on the UPC, UPC,3 ; USD/m
USD/m (b) effect;of
3

(b) effect of feed/splitter

feed/splitter ratio and no. ofratio
HWTs and
onno.
theof HWTs
UPC, on the
USD/m 3 . UPC, USD/m . 3

Figure 4b shows the effect of various system production rates from31000 to 3500 m 3/d
Decreasing the quantity of total fresh water from 3500 to 1000 m /day would also
on the UPC. The optimized point is obvious at 0.31 against 0.5573USD/m3. Decreasing total
increase the UPC for the same operating conditions. At 3500 m /day and a feed/splitter
productivity while increasing the inlet feed/splitter ratio led to a decrease in the UPC. For
ratio of 70%, the UPC is about 0.548 USD/m3 . Figure 4a,b represents the effect of the
the operation of the PV field, the 35 W module is effective because of the individual unit
price. Moreover, increasing rather than decreasing the inlet feed/splitter ratio is recom-
mended for the following reasons:
• It decreases the electrical load on the HPP unit;
Sustainability 2021, 13, 11508 9 of 20

inlet feed splitter, PV module power, and total productivity on the UPC. Figure 4a,b
demonstrates the effect on the UPC without the operation of the HWT farm. In Figure 4a,
results are obtained at total productivity of 3500 m3 /day. The figure shows that it was
important to increase the feed/splitter ratio related to the RO-PEX part because it decreased
the HPP power demand and hence lowered the UPC. At the same time, selecting a module
with a higher rate of power increased the UPC. The results revealed that the PV 35 W
module was the best choice with a feed/splitting ratio of more than 70%. However, the
PV 280 W module increased the UPC. For a ratio of 70% and PV 35 W, it was about
0.517 USD/m3 .
Figure 4b shows the effect of various system production rates from 1000 to 3500 m3 /day
on the UPC. The optimized point is obvious at 0.31 against 0.557 USD/m3 . Decreasing
total productivity while increasing the inlet feed/splitter ratio led to a decrease in the UPC.
For the operation of the PV field, the 35 W module is effective because of the individual
unit price. Moreover, increasing rather than decreasing the inlet feed/splitter ratio is
recommended for the following reasons:
• It decreases the electrical load on the HPP unit;
• The operation proceeds without any need for more pressure vessels or membranes;
• The gain power is fully loaded on the PEX unit without any excess power from the
brine loss.
Suppose that the investor or the designer had to choose between the PV solar field or
the HWT farm or both with respect to the control room load (CRL) distribution. From the
results, the operation of the PV is recommended over the HWT operation based on the UPC.
For both operations, increasing the load percentage meant increasing the dependent load
on the HWT farm. Figure 5 shows the variations in the electric load distribution between
the PV and the HWT. It is evident that the UPC was minimized by a load percentage of
less than 5%; i.e., fully operated by the PV solar field. Decreasing the load on the HWT
farm lowered the UPC. The operation with less than a 5% load on the HWT farm was
remarkable and cost about USD 0.518/m3 . Criteria to consider when selecting the type of
operation are
• The PV module power and number of HWTs,
• The total productivity with respect to the inlet feed/splitter ratio, and
Sustainability 2021, 13, x FOR PEER REVIEW 10 of 21
• The site operating conditions, which decide the number of HWTs in the farm based
on the power of each turbine.

Figure
Figure 4. (a) UPC,
4. (a) UPC,USD/m
USD/m [Link]
3 3vs. PVpower,
power,WW
andand RO–PEX
RO–PEX feed feed splitter
splitter ratio,
ratio, (b) UPC,(b)USD/m
UPC, 3 vs.
USD/m 3 vs. productivity,
3 m 3/d and RO–PEX feed splitter ratio (PV solar field without
productivity, m /day and RO–PEX feed splitter ratio (PV solar field without HWT farm).
HWT farm).
 Figure 4. (a) UPC, USD/m3 vs. PV power, W and RO–PEX feed splitter ratio, (b) UPC,
Sustainability 2021, 13, 11508 USD/m3 vs. productivity, m3/d and RO–PEX feed splitter ratio (PV solar field without
10 of 20
HWT farm).

Theeffect
Figure5.5. The effect of
of control
control room
room load distribution
distribution and the UPC, USD/m 3 (PV = 35 We module
Figure USD/m3 (PV = 35 We module
andHWT
and HWTNo. No.==11turbine).
turbine).

Accordingto
According toEquations
Equations(A1)
(A1)and
and(A14)
(A14)ininthetheAppendix
AppendixA, A,increasing
increasingthethepower
powerload
load
depends on system productivity, and the feed mass flow rate and the
depends on system productivity, and the feed mass flow rate and the HPP pressure de- HPP pressure depend
on the
pend ondesign parameters
the design of theofRO
parameters themodules.
RO [Link], Therefore,the power is a multidimensional
the power is a multidimen-
functional
sional of parameters:
functional of parameters:

 ∆P ∆𝑃==f 𝑓(𝐴
( Ae ,𝑒n, 𝑛e ,𝑒N, 𝑁V𝑉,,k𝑘w𝑤,,FF,
𝐹𝐹,TCF
𝑇𝐶𝐹))


𝑃𝑜𝑤𝑒𝑟 = 𝑓 { 𝑀
M 𝑓=𝑓{ 𝑑 𝑀 (1)
Power = f M 𝑅𝑅d
(1)
 f = f {
RR


where Ae , ne , NV are the design parameters for the RO, and their effect is notable on the
RO pressure.
In this optimization, the constraints of the algorithm are FF = 0.85; TCF = f(Tf );
Ae = 35.5 m2 ; ne = 8 elements, and the limits of the system are changed around the Md , NV ,
and the ∆P. The results revealed that more productivity meant more power at the same
recovery ratio (RR) of 0.3. It was obvious that by increasing system productivity up to
200,000 m3 /day the NV would directly affect the UPC at the same pressure category (e.g.,
40 bar). The UPC increased from 0.47 to 7.33 USD/m3 by the use of 20,000 NV . Increasing
the NV decreased the UPC by increasing the ∆P, thereby decreasing the power and the load
on the PV site. For moderate production (10,000 m3 /day), about 1450 pressure vessels at
40 bar were proposed; however, 750 pressure vessels to produce 100,000 m3 /day at 80 bar
were used. It is clear now that the choice should have been made according to the lower
cost for each pressure category. Therefore, 40 bar should be used, but the massive number
of pressure vessels is considered a hindrance to maintenance and system control. For less
control and fewer maintenance issues 80 bar is preferred; moreover, maximum production
could reach 200,000 m3 /day with 1500 pressure vessels compared with 20,000 pressure
vessels for 40 bar. According to the PV site, the operation of 200,000 m3 /day is considered
a massive challenge because of the relation of the site area to the PV site, which would have
to generate about 24 MWe. Therefore, because of cost, the 40-bar category was favored in
 is clear now that the choice should have been made according to the lower cost for each
pressure category. Therefore, 40 bar should be used, but the massive number of pressure
vessels is considered a hindrance to maintenance and system control. For less control and
fewer maintenance issues 80 bar is preferred; moreover, maximum production could
reach 200,000 m3/d with 1500 pressure vessels compared with 20,000 pressure vessels for
Sustainability 2021, 13, 11508 11 of 20
40 bar. According to the PV site, the operation of 200,000 m 3/d is considered a massive
challenge because of the relation of the site area to the PV site, which would have to gen-
erate about 24 MWe. Therefore, because of cost, the 40-bar category was favored in this
study. Figure
this study. 6 shows
Figure the the
6 shows data results
data of of
results thethe
effect ofofpressure
effect pressureand
andproductivity
productivityon
on the
the
system
system pressure
pressure vessel
vessel numbers
numbers and
and UPC.
UPC.

6. Effect of
Figure 6.
Figure of HPP
HPPpressure
pressureand
andsystem productivity
system on on
productivity (a) (a)
[Link].
of pressure vessels,
of pressure and (b)
vessels, andUPC
(b)
UPC 3 . 3.
in USD/m
in USD/m

4.3. Data
4.3. DataResults
ResultsCompared
Compared with
with Solar-ORC
Solar-ORC
The PV-RO
The PV-ROisiscompared
comparedwith withthethe S-ORC-RO
S-ORC-RO system
system underunder the same
the same operating
operating con-
conditions based on the 3500 m 3 /day capacity. For HPP = 40 bar category, the so-
ditions based on the 3500 m3/d capacity. For HPPRO = 40 bar ROcategory, the solar thermal
lar thermal cycle (PTC) consumed a larger area (4475 vs.a larger
3243 mUPC2 ) and a larger UPC
cycle (PTC) consumed a larger area (4475 vs. 3243 m 2) and (1.365 vs. 0.4705
3 ). Increasing the number of pressure vessels increased the in-
(1.365 vs. 0.4705 USD/m
USD/m ). Increasing the number of pressure vessels increased the indirect costs for both
3
direct costs
systems. for boththe
Moreover, systems.
power Moreover,
load on the the power units
thermal load on
wasthe thermal units
synchronized wasmore
with syn-
chronized with more thermal units, mass flow rates and pumps comparing with direct
thermal units, mass flow rates and pumps comparing with direct contact in the case of
contact in the case of PV-RO. Even though the HPPRO = 80 bar category harvested more
PV-RO. Even though the HPPRO = 80 bar category harvested more area, the UPC was lower
area, the UPC was lower for thermal system; however, it increased a little in the electric
for thermal system; however, it increased a little in the electric one. The HPP RO = 80 bar
one. The HPPRO = 80 bar category required more power, i.e., more area was needed.
category required more power, i.e., more area was needed. Figure 7 shows the data com-
Figure 7 shows the data comparison between both systems according to3 the area and UPC
parison between both systems according to the area and UPC at 3500 m /d. Increasing the
at 3500 m3 /day. Increasing the pressure on the RO reduced the number of pressure vessels
pressure on the RO reduced the number of pressure vessels (26 at 80 bar vs. 450 at 40 bar).
(26 at 80 bar vs. 450 at 40 bar). Therefore, the power increase to harvest more solar area
Therefore, the power increase to harvest more solar area caused an increase in the UPC.
caused an increase in the UPC. Figure 8 shows variations in the solar field area for both
Figure 8 shows variations in the solar field area for both sites (thermal and electrical) ac-
sites (thermal and electrical) according to the variations of HPP power. Increasing the
cording to the variations of HPP power. Increasing the power demand by the RO pump
power demand by the RO pump required more area to cover the load. The competition of
PV with solar thermal for power generation is still far away from implementation because
of limitations to the PV operation. Table 6 illustrates the comparison between the PV and
the CSP thermal power.
It is clear that the UPC of the PV–RO system was lower than that of the CSP–RO;
however, it may increase during operation for high rates of power (over than 10 MWe).
because of the high price of the PV panels. Reducing the capital costs of the PV panels may
reduce the UPC making it competitive with thermal systems. One possible way to reduce
the UPC is to increase the construction rate by building many PV–RO systems with higher
rates of productivity for larger remote areas. It may be noted that a PV-powered RO system
is suitable for remote regions. Generally, CSP is favorable for larger categories of power
generation up to 600 MW.
Figure 9 shows the UPC comparison between Toh et al. [23] and the present work
with different feed pressures. Increasing the pressure on the RO caused an increase in the
UPC, and the results showed compatibility and convergence.
 may reduce the UPC making it competitive with thermal systems. One possible way to
reduce the UPC is to increase the construction rate by building many PV–RO systems with
higher rates of productivity for larger remote areas. It may be noted that a PV-powered
RO system is suitable for remote regions. Generally, CSP is favorable for larger categories
Sustainability 2021, 13, 11508 of power generation up to 600 MW. 12 of 20

Table 6. CSP vs. PV solar power generation.

Table 6. CSPParameter:
vs. PV solar power generation.
CSP (Thermal) PV (Electric)
Parameter
Resource quality CSPkWh/m
2400 (Thermal)
2/yr 2445PV (Electric)
kWh/m 2/yr

Resource
Powerquality
type 2400
ThermalkWh/m2 /year
(indirect) 2445 kWh/m
Electrical
2
/year
(direct)
Power type
Desalination system to Thermal
All types (MSF,(indirect)
MED, MED– Electrical (direct)
Desalination system RO, MED-MVC
combine with to AllMED-MVC,
TVC, types (MSF, MED,
RO, ED RO, MED-MVC
combine with MED–TVC, MED-MVC, RO, ED
Levelised cost of energy
Levelised cost of energy 60–350 (USD 214 in 2030) 100–450 (USD 303 in 2030)
USD/MWh 60–350 (USD 214 in 2030) 100–450 (USD 303 in 2030)
USD/MWh
Construction
Construction period/life
period/life time
time 2/30
2/30years
years 1/30 years
1/30 years
Capacityfactor
Capacity factor 23–50%
23–50% 20%
20%
Powerproduction
Power production 600
600MW
MW 1010
MWMW
Stirling,
Stirling, Rankine,
Rankine, gas
gas turbines,
Heat
Heatengines
engines N/A
N/A
steam turbines
turbines, steam turbines

Sustainability 2021, 13, x FOR PEER REVIEW 13 of 21

Figure 7. Data comparison between PV-RO and S-ORC-RO: (a) area comparison, (b) UPC comparison.
Figure 7. Data comparison between PV-RO and S-ORC-RO: (a) area comparison, (b) UPC comparison.

Figure 8. The effect of HPP power, kWe on the solar filed area for both techniques thermal and electrical.
Figure 8. The effect of HPP power, kWe on the solar filed area for both techniques thermal and
electrical.

Figure 9 shows the UPC comparison between Toh et al. [23] and the present work
with different feed pressures. Increasing the pressure on the RO caused an increase in the
UPC, and the results showed compatibility and convergence.
Sustainability 2021, 13, 11508 13 of 20

Figure 9. Comparison of UPC between Toh et al. [23] and the present work with different feed pressures.

5. Conclusions
This work was concerned with powering a reverse osmosis (RO) desalination system
by using renewable energy sources such as solar photovoltaic and wind energy. Three
configurations for powering the RO were compared analytically by the use of SDS-REDS
software. The first was HWT-RO; the second was PV-RO, while the third was the hybrid
system HWT-PV-RO. To select the most cost-efficient one to be compared with the solar
thermal Rankine cycle (S-ORC-RO), a technical and analytical optimization had to be
implemented. The technical optimization recommended the use of the PEX technique,
which reduced power consumption by 65%.
For analytical optimization, the following items are recommended:
• Reducing the inlet feed/splitter ratio to 10% to more than triple the demanded power
by the HPP (390 to 1532 kW).
• Using fewer than two HWTs for this kind of an operation because increasing the
number of wind turbines (low power per unit) increases the UPC.
• Selecting a PV module that minimizes UPC. The results revealed that the PV 35 W
module was the best choice with a feed/splitter ratio of more than 70%, whereas the
PV 280 W module increased the UPC. The PV 35 W UPC was about 0.517 USD/m3 .
• Maintaining less than a 5% load on the HWT farm to minimize UPC. For both oper-
ations, increasing the load percentage meant increasing the dependent load on the
farm, and this work confirmed that was UPC is minimized when the load percentage
was less than 5%; i.e., fully operational by the PV solar field. This operation gave a
remarkable UPC of about 0.518 USD/m3 .
• Reducing the capital costs of PV panels to reduce the UPC so that it can compete with
thermal systems. In a comparison with S-ORC-RO, the UPC of the PV-RO system
was lower than that of the CSP-RO system; however, it may increase if the operation
generates power over 10 MWe.
Generally, the CSP was favorable for larger categories of power generation up to
600 MW. The PV-RO and HWT-RO were proven promising for producing sustainable fresh
water. PV-RO can compete against the HWT-RO or the S-ORC-RO for lower rates of fresh
water production. HWT-RO and S-ORC-RO consume larger areas; however, they could be
used for higher rates of power and fresh water production.

6. Recommendations
A wide range of varied operating conditions should be examined, and system perfor-
mance should be studied using more design parameter variation.

1
Sustainability 2021, 13, 11508 14 of 20

Author Contributions: Methodology: A.M.S. and M.A.S.E.; formal analysis: A.M.S. and M.A.S.E.;
software, M.A.S.E.; validation, A.M.S., A.G.A. and M.A.S.E.; formal analysis, A.M.S. and M.A.S.E.;
investigation, A.M.S., A.G.A. and M.A.S.E.; resources, A.G.A.; administration, A.M.S.; funding
acquisition, A.G.A.; writing—original draft preparation: A.M.S. and M.A.S.E.; review and editing:
A.M.S. and A.G.A.; supervision: A.M.S.; project administration: A.G.A. All authors have read and
agreed to the published version of the manuscript.
Funding: This work is supported by Ministry of Education in Saudi Arabia for funding this work
through the project number 375213500.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: The authors extend their appreciation to the Deputyship for Research & Inno-
vation, Ministry of Education in Saudi Arabia for funding this work through the project number
375213500.
Conflicts of Interest: Authors declare no conflict of interest.

Nomenclature

A Area: m2
Ac Cell area, m2
Ae Element area, m2
AH Battery capacity, Ah
Am Module area, m2
Ar Rotor swept area m2
At Total area, m2
BS Battery storage, Wh
Cb Battery cost, USD
Ct Total cost, USD
CP Turbine power coefficent
DCC Direct capital cost, USD
DOD Depth of discharge
Dr Rotor diameter, m
FF Fouling factor
FOBc Full over board cost, USD
Gb Solar flux, W/m2
HPP High pressure pump
Hh Hub height, m
HWT Horizontal wind turbine
k Permeability
M Mass flow rate, m3 /h, kg/s
n Number, #
ne Element number
NV Number of pressure vessels
NOB Number of batteries, #
NOC Number of cells, #
NOM Number of modules, #
NWT Number of wind turbines, #
OH Operating hours, h
P Power, Permeator, or Pressure, bar
Pm Module power, W
Pt Total power, W
Pw Wind power, kW
PV Photovoltaic
∆P Pressure, bar
RR Recovery ratio
Sustainability 2021, 13, 11508 15 of 20

RPMr Rotor speed, rpm

SPC Specific power consumption, kWh/m3
SR Salt rejection
T Temperature, ◦ C
Tor Torque, Nm
TCF Temperature correction factor
UPC Unit product cost, USD/m3
V Volt
Vws Start wind speed, m/s
Vwa Average wind speed, m/s
X Salinity, ppm
Subscripts
air Ambient
b Brine, battery
c Cell
d Distillate product
e Element
f Feed
m Module
ORC Organic rankine cycle
RO Reverse osmosis
t Turbine, total
v Vessel
w Water
Greek
η Efficiency, %
Π Osmotic pressure, kPa
ρ Density, kg/mr
ω Rad/s

Appendix A
A: The RO Model
The mathematical model for the proposed RO unit is written as follows [2,24]:
The feed flow rate Mf based on recovery ratio RR and distillate flow rate Md is

Md
Mf = (A1)
RR
The distillate product salt concentration Xd is

Xd = X f × (1 − SR) (A2)

where Xf is the feed flow rate salt concentration, and SR is the salt rejection percentage.
The rejected brine is found from
Mb = M f − Md (A3)

The rejected salt concentration kg/m3 is estimated by

M f × X f − Md × X d
Xb = (A4)
Mb

The average salt concentration kg/m3 is estimated as

M f × X f + Mb × X b
Xav = (A5)
M f + Mb
Sustainability 2021, 13, 11508 16 of 20

The temperature correction factor TCF is found by

  
1 1
TCF = exp 2700 × − (A6)
273 + t 298

Membrane water permeability kW is

k w = 6.84 × 10−8 × (18.6865 − (0.177 × Xb ))/(t + 273)) (A7)

The salt permeability ks is

  
k s = FF × TCF × 4.72 × 10−7 × 0.06201 − 5.31 × 10−5 × (t + 273) (A8)

where FF is the membrane fouling factor. The calculations of osmotic pressure for the feed,
brine, and distillate product are found as

Π f = 75.84 × X f (A9)
Πb = 75.84 × Xb (A10)
Πd = 75.84 × Xd (A11)
The average osmotic pressure on the feed side is
 
Π av = 0.5 × Π f + Πb (A12)

The net osmotic pressure across the membrane is

∆Π = Π av − Πd (A13)

The net pressure difference across the membrane is

 
Md
∆P = + ∆Π (A14)
3600 × TCF × FF × Ae × ne × Nv × k w

where Ae is the element area in m2 ; ne is number of membrane elements; and Nv is the

number of pressure vessels. The required power input in kW for the RO high pressure
pump (HPP) is estimated as

1000 × M f × ∆P
HPPpower = (A15)
3600 × ρ f × η p

where ρf is the feed flow rate density, and η p is the driving pump mechanical efficiency.
The specific power consumption in kWh/m3 is estimated as

HPPpower
SPC = (A16)
Md

B: The HWT Model

The Horizontal Wind Turbine (HWT) as a part of wind desalination library (REDS [25])
was modeled according to the specification data obtained from the manufacture manual
for many watt points, which vary from 0.5 to 8000 kW according to many companies. The
data were obtained from more than 50 companies involved in wind turbine manufacturing.
A developed model by Sharaf [26–29] was presented and correlated as a function of wind
turbine power (P) as follows:
The starting wind speed m/s as a function of turbine power (kW):
−5 × P )
Vws = 13.37 × e(1.698 − 10.72 × e(−0.003214× P) (A17)
Sustainability 2021, 13, 11508 17 of 20

The average wind speed m/s:

 
Vwa = 9.378 × P0.09862 (A18)

The rotor diameter m:  

Dr = 2.573 × P0.4414 (A19)

The tower (Hub) height m:

 
Hh = 1.437 × P0.5046 + 5.354 (A20)

Air density kg/m3 is calculated based on air and pressure temperature:

Pair × 100
ρ air = (A21)
0.287 × ( Tair + 273.15)

where Pair is in bar and Tair is in ◦ C.

The rotor swept area m2 is then calculated based on the rotor diameter Dr:

Ar = π × ( Dr/2)2 (A22)

The air mass flow rate kg/s is then calculated based on the density, rotor swept area
and average wind speed:
Mair = ρ air × Ar × Vwa (A23)
The required wind power kW:

( 1 × ρ air × Ar × Vwa 3


Pw = 2 (A24)
1000
The power coefficient is calculated from the assigned power P and the aerodynamic
power Pw:
P
CP = (A25)
Pw
The rotor speed in rpm:
 
rpmr = 347.6 × P−0.2909 − 16.91 (A26)

The rotor torque Tor in Nm based on the power of the turbine and the angular velocity (ω):

(2 × π × RPMr )
ω= (A27)
60
(1000 × P)
Tor = (A28)
ω
The number of wind turbines can be calculated related to the total demanded power
(TP kW) from the wind farm:
TP
NWT = (A29)
P
C: The PV Model
PV system was considered a very important power source in this work. It was
modeled according to the actual data presented through more than 150 data points from
the manufacturing manuals. The range of the operating modules type was from 5 to 280 W.
Each module watt type can calculate the module specification based on the data fed in the
table. Table A1 illustrates the inputs and outputs of the developed lookup table model
Sustainability 2021, 13, 11508 18 of 20

block. SDS program [16,26] library is used to model and visualize the PV system program.
The developed code is introduced to calculate the following:

Table A1. PV input and calculated parameters by the use of SDS program [25].

Inputs: Outputs:
1—Open circuit voltage (Voc ), Volt
2—Short circuit current (Isc ), A
3—Maximum voltage (Vm ), Volt
4—Maximum current (Im ), A
1—Operating hours (OH), h
5—Cell and Module efficiencies, %
2—Solar flux (Gb ), W/m2
6—Net weight, kg
3—Number of cloudy days factor
7—The dimensions, m2
4—System total power (Pt ), kW
8—Module price, USD/W
5—Module power (Pm ) (5–280 W)
9—Number of cells and modules (NOC)
6—Battery depth of discharge (DOD)
10—Cell area (Ac ), cm2
7—Battery voltage (Vb ), Volt
11—Module area (Am ), m2
8—Battery efficiency, %
12—Total system area (At ), m2
9—Battery unit price (Cb ), USD
13—Battery storage, Wh
14—Battery capacity, Ah
15—Number of batteries (NOB)
16—Full over board cost (FOBc ), USD.

When calculating the main specifications (Table A1, parameters from 1 to 9) based
on the module power, the following code was easily calculated. The number of modules
(NOM) based on total power and module power:

Pt
NOM = (A30)
Pm

The module area in m2 is then calculated based on module power Pm and efficiency ηm :

Pm
Am = 100 × (A31)
Gb × ηm

Then the total area in m2 was calculated:

At = Am × NOM (A32)

The cell area in cm2 based on the number of cells (NOC) that been calculated from the
lookup table.
Am × 103
Ac = (A33)
NOC
The battery storage in Wh based on the operating hours (OH), number of cells (NOC),
the total power (Pt ), battery efficiency and depth of discharge (DOD):

OH × NOC × Pt
BS = (A34)
DOD × ηb

If a 24 V system were chosen, the required (AH) of batteries would be 16,585/24,700.

BS
AH = (A35)
Vm

Number of batteries can be calculated as follows based on the maximum voltage and
the battery voltage:
Vm
NOB = (A36)
Vb
Sustainability 2021, 13, x FOR PEER REVIEW 20 of 21
Sustainability 2021, 13, 11508 19 of 20

Sustainability 2021, 13, x FOR PEER REVIEW 20 of 21

The system total costs in 𝐶(C

𝑡 t= (𝑃𝑡 ×are
, USD) 𝐹𝑂𝐵 𝑐) +
then (𝐶𝑏 × 𝑁𝑂𝐵)
calculated based on the full over board(A37)
costs of the modules (FOBc ) and the batteries costs (Cb ):
where the FOBc includes cables, connections, worker time, inverter unit, and maintenance
𝐶𝑡 = (𝑃𝑡 × 𝐹𝑂𝐵𝑐 ) + (𝐶𝑏 × 𝑁𝑂𝐵) (A37)
costs. Ct = ( Pt × FOBc ) + (Cb × NOB) (A37)
where the FOBc includes cables, connections, worker time, inverter unit, and maintenance
D: Monthly average irradiation and wind speed
costs.
where the FOBc includes cables, connections, worker time, inverter unit, and maintenance costs.
D: Monthly average
D: Monthly irradiation
average and wind
irradiation speed
and wind speed

Figure
Figure
Figure A1.A1.
A1. Monthlyaverage
Monthly average irradiation
irradiation for
forSuez
SuezGulf
Gulfregion.
region.

Figure A2. Monthly average wind velocity for Suez Gulf region.

References Figure
Figure A2. A2. Monthlyaverage
Monthly average wind
wind velocity
velocityforforSuez
SuezGulf region.
Gulf region.
1. Lamei, A.; van der Zaag, P.; von Münch, E. Basic cost equations to estimate unit production costs for RO desalination and long-
distance piping to supply water to tourism-dominated arid coastal regions of Egypt. Desalination 2008, 225, 1–12.
References
References
2.
1. Nafey,
Lamei, A.;
A.; Sharaf,
van derM. Combined
Zaag, P.; vonsolar organic
Münch, Rankine
E. Basic cycle with to
reverse osmosis desalinationcosts
process: Energy, exergy, and
1. Lamei, A.;evaluations.
cost van der Zaag, P.; Energy
Renew. von Münch,
2010,to costcost
E.2571–2580.
35, Basic equations
equations estimate
to estimate unit production
unit production costs for RO
for desalination
RO desalination and
and long-
long-distance piping to supply water tourism-dominated arid coastal regions of Egypt. Desalination 2008, 225, 1–12. [CrossRef]
distance
3. piping
Mark, W.; to supply
Craig, B. water to tourism-dominated
Optimization of seawater RO ariddesign.
systems coastalDesalination
regions of2005,
Egypt.
173,Desalination
1–12. 2008, 225, 1–12.
2. Nafey, A.; Sharaf, M. Combined solar organic Rankine cycle with reverse osmosis desalination process: Energy, exergy, and cost
2. Nafey,
4. A.; Sharaf,
Qasim, M. Combined
M.; Badrelzaman,
evaluations. Renew. EnergyM.; solar
Darwish,
2010, organic Rankine
N.N.; Darwish,
35, 2571–2580. cycleHilal,
N.A.;
[CrossRef] with [Link] osmosis
Reverse osmosisdesalination
desalination: process: Energy,review.
A state-of-the-art exergy, and
cost
3. Desalination
evaluations. 2019, 459,
Renew. 59–104.
Energy 2010, 35, 2571–2580.
Mark, W.; Craig, B. Optimization of seawater RO systems design. Desalination 2005, 173, 1–12.
3. Mark, W.; Craig, B. Optimization of seawater RO systems design. Desalination 2005, 173, 1–12.
4. Qasim, M.; Badrelzaman, M.; Darwish, N.N.; Darwish, N.A.; Hilal, N. Reverse osmosis desalination: A state-of-the-art review.
Desalination 2019, 459, 59–104.
Sustainability 2021, 13, 11508 20 of 20

4. Qasim, M.; Badrelzaman, M.; Darwish, N.N.; Darwish, N.A.; Hilal, N. Reverse osmosis desalination: A state-of-the-art review.
Desalination 2019, 459, 59–104. [CrossRef]
5. Mohamed, E.S.; Papadakis, G.; Mathioulakis, E.; Belessiotis, V. A direct coupled photovoltaic seawater reverse osmosis desalina-
tion system toward battery based systems—a technical and economical experimental comparative study. Desalination 2008, 221,
17–22. [CrossRef]
6. Helal, A.; Al-Malek, S.; Al-Katheeri, E. Economic feasibility of alternative designs of a PV-RO desalination unit for remote areas
in the United Arab Emirates. Desalination 2008, 221, 1–16. [CrossRef]
7. Manolakos, D.; Mohamed, E.; Karagiannis, I.; Papadakis, G. Technical and economic comparison between PV-RO system and
RO-Solar Rankine system. Case study: Thirasia island. Desalination 2008, 221, 37–46. [CrossRef]
8. Ahmad, G.; Schmid, J. Feasibility study of brackish water desalination in the Egyptian deserts and rural regions using PV systems.
Energy Convers. Manag. 2002, 43, 2641–2649. [CrossRef]
9. Tzen, E.; Perrakis, K.; Baltas, P. Design of a stand alone PV-desalination system for rural areas. Desalination 1998, 119, 327–333.
[CrossRef]
10. Liu, C.; Xia, W.; Park, J. A wind-driven reverse osmosis system for aquaculture wastewater reuse and nutrient recovery.
Desalination 2007, 202, 24–30. [CrossRef]
11. de la Nuez Pestana, I.; Latorre, F.J.G.; Espinoza, C.A.; Gotor, A.G. Optimization of RO desalination systems powered by renewable
energies. Part I Wind. Energy. Desalination 2004, 160, 293–299.
12. Dehmas, D.A.; Kherba, N.; Hacene, F.B.; Merzouk, N.K.; Merzouk, M.; Mahmoudi, H.; Goosen, M.F. On the use of wind energy
to power reverse osmosis desalination plant: A case study from Ténès (Algeria). Renew. Sustain. Energy Rev. 2011, 15, 956–963.
[CrossRef]
13. García-Rodríguez, L.; Romero-Ternero, V.; Camacho, C.G. Economic analysis of wind-powered desalination. Desalination 2001,
137, 259–265. [CrossRef]
14. Romero-Ternero, V.; García-Rodríguez, L.; Gómez-Camacho, C. Thermoeconomic analysis of wind powered seawater reverse
osmosis desalination in the Canary Islands. Desalination 2005, 186, 291–298. [CrossRef]
15. Ma, Q.; Lu, H. Wind energy technologies integrated with desalination systems: Review and state-of-the-art. Desalination 2011,
277, 274–280. [CrossRef]
16. Dimitriou, E.; Boutikos, P.; Mohamed, E.S.; Koziel, S.; Papadakis, G. Theoretical performance prediction of a reverse osmosis
desalination membrane element under variable operating conditions. Desalination 2017, 419, 70–78. [CrossRef]
17. Ruiz-García, A.; Nuez, I. Performance evaluation and boron rejection in a SWRO system under variable operating conditions.
Comput. Chem. Eng. 2021, 153, 107441. [CrossRef]
18. Ruiz-García, A.; Nuez, I. Long-term intermittent operation of a full-scale BWRO desalination plant. Desalination 2020, 489, 114526.
[CrossRef]
19. Nafey, A.S.; Sharaf, M.A.; García-Rodríguez, L. A new visual library for design and simulation of solar desalination systems
(SDS). Desalination 2010, 259, 197–207. [CrossRef]
20. Delgado-Torres, A.M.; García-Rodríguez, L. Preliminary assessment of solar organic Rankine cycles for driving a desalination
system. Desalination 2007, 216, 252–275. [CrossRef]
21. Nafey, A.; Sharaf, M.; Rodríguez, M.D.L.G. Thermo-economic analysis of a combined solar organic Rankine cycle-reverse osmosis
desalination process with different energy recovery configurations. Desalination 2010, 261, 138–147. [CrossRef]
22. Sharaf, M.A.; Nafey, A.S.; García-Rodríguez, L. Exergy and thermo-economic analyses of a combined solar organic cycle with
multi effect distillation (MED) desalination process. Desalination 2011, 272, 135–147. [CrossRef]
23. Toh, K.Y.; Lianga, Y.Y.; Lau, W.J.; Weihs, G.A.F. The techno-economic case for coupling advanced spacers to high-permeance RO
membranes for desalination. Desalination 2020, 491, 114534. [CrossRef]
24. El-Dessouky, H.T.; Ettouney, H.M. Fundamental of Salt Water Desalination; Elsevier Science: Amsterdam, The Netherlands, 2002.
25. Eldean, M.A.S. Design and Simulation of Solar Desalination Systems. Ph.D. Thesis, Suez Canal University, Suez, Egypt, 2011.
26. Soliman, A.M.; Eldean, M.A.S.; Miraouia, I. Experimental and Economical Analysis of an Autonomous Renewable Power Supply
System for Water Desalination and Electric Generation. Mod. Appl. Sci. 2019, 13, 43–64. [CrossRef]
27. Soliman, A.M.; Al-Falahi, A.; Mohamed AEldean, S.; Elmnifi, M. A new system design of using solar dish-hydro combined with
reverse osmosis for sewage water treatment: Case study Al-Marj, Libya. Desalination Water Treat. 2020, 193, 189–211. [CrossRef]
28. Eldean, M.A.S.; El Shahat, A.; Soliman, A.M. A new modeling technique based on performance data for photovoltaic modules
and horizontal axis wind turbines. Wind. Eng. 2017, 209–229.
29. Ansari, M.; Mudhar, A.; Obaidi, A.; Hadadian, Z.; Moradi, M.; Haghighi, A.; Mujtaba, I.M. Performance evaluation of a brackish
water reverse osmosis pilot-plant desalination process under different operating conditions: Experimental study. Clean. Eng.
Technol. 2021, 4, 100134. [CrossRef]