Dahioui and Loudiy 2013 WIND POWERED DESALINATION PAPER

Wind Powered Water Desalination
*
#
Youssef Dahioui , Khalid Loudiyi
"'
School o/Science and Engineering
Al Akhawayn University in Ifrane

/frane, Morocco
ly.dahioui@aui.ma
2k.loudiyi@aui.ma
Abstract-
Renewable energy,
and
more
specifically,
wind
energy powered desalination, has been going through an

upwards trend, especially during the last decade. Still, there is a
domain that has not been much researched 111; hence, this paper
tries to address wind powered independent desalination systems.
With its lower energy consumption and portability, Reverse
Osmosis (RO) method has been chosen as the desalination
technology that will be integrated with wind energy. In this work
a MA TLAB simulation is used to find out the effects of
fluctuating wind energy on a system that is designed to operate
under steady conditions. The results show that varying electrical
power leads to extreme fluctuation in feed water pressure,
beyond the operational range of the RO membranes. Still, this
does not draw a cross on the wind and desalination combo; in
fact, several methods exist to diminish the pressure variation,
including wind turbine de-rating, or use of pressure stabilizers.
Kwords-Desalination,
wind
energy,
reverse
osmosis,
MA TLAB, simulation
I. INTRODUCTION
Without water, nothing alive on this planet would have
existed. 75% of the surface of the earth is covered by water;
nevertheless, only 3% of that water is available as fresh water,
and only 13% of those 3% are directly available for drinking
and any other domestic, industrial or agricultural uses [2].
Just a few decades ago, fresh water was viewed as an eternal,
renewable and easily accessed resource; however, nowadays,
water shortage has become a serious issue that may be the
main cause of conflicts in the near future. One of the most
widely used indicators for defining water stress is the
Falkenmark indicator, illustrated in Table I and defined as the
entire annual water available for human use [3].
II. DESALINATION TECHNOLOGIES
A. Thermal Technologies
1) Multi-stage Flash Distillation: Multi-stage Flash
(MSF) is the most used thermal desalination technology,
worldwide; it represents about 50% of the installed capacity
[5]. Basically, this process is about evaporating feed water in
a group of chambers, each having a lower pressure than the
previous one. When getting into one stage, water flashes, or
evaporates instantaneously due to the low pressure implying a
lower evaporation point.
2) Multi-Effect
Distillation:
The
Multi-Effect
Distillation (ME) was the first process used for desalination of
seawater [6]. A quantity of water is heated up till becoming
vapor then goes through a heat exchanger. Feed water that is
going to be desalinated is sprayed in the heat exchanger
condensing vapor flowing through it. Latent heat released due
to condensation causes some of the feed water to evaporate
and flows to another heat exchanger for the process to take
place again until a significant quantity of condensate water
has been collected [6].
3) Vapor Compression: Vapor compression (VC)
process is a more recent method for water desalination; still, it
is based on a simple principle. As shown in Fig.l, it is usually
composed of three parts, a compressor, a heat exchanger and
an evaporator. Some of the vapor produced within the
evaporator is sent to the compressor, increasing its pressure,
and its temperature as well. The superheated vapor that leaves
the compressor, gets into the heat exchanger submerged
within the feed water, and causes its evaporation [7].
TABLE I
WATER SHORTAGE CATEGORIES
3
Index (m per capita)
>1700
1000 - 17000
500 - 1000
<
500
Category
No stress
Stress
Scarcity
Absolute Scarcity
To tackle this water shortage issue, water desalination has

represented for years, an effective, yet an energy consuming
method [4]. Thus, and taking into account the current energy
world market, with increasing fossil fuel prices, renewable
energy powered desalination is taking an interesting trend.
Compressor
--=====::==
LI
===:===i[:;;:;::::: -I.
-
{}
lW'
/=,-esh Water
Fig.
1 Vapor compression cycle
B. Membranes Technologies
978-1-4673-6374-7/13/$31.00 2013
IEEE
-= Feed Waler
1) Electrodialysis: Electrodialysis (ED) is the oldest

desalination membrane-based technology and has been used
all around the world for more than 40 years. More than 10
million m3 of water are produced on a daily basis using this
technology [8]. From Fig.2, it is a process in which ions are
attracted to their respective electrically charged electrode
through ion-selective semi-permeable membranes. Positively
and negatively charged dissolved salts in the aqueous solution
are attracted to the electrode with the opposite charge, and
membranes, that allow either cations or anions to pass
through, are installed in an alternative way, thus creating
concentrated and purified streams.
III. W[ND POWERED DESALlNAT[ON

One problem that has been facing water desalination is its
significant energy needs.
Consequently, to incorporate
renewable energy and more specifically wind energy that can
be considered nowadays as a mature technology, there are two
main aspects or factors that should be taken into
consideration:
Specific energy consumption (Table II) or ability to
produce as much water as possible from the available
energy during any period
Operability under variable conditions and this is what
actually keeps the thermal technologies away; they
usually require a long start-up time and significant
energy waste could result from frequent stops.
Concentrate
Oir uate
Concentrate
TABLE II
SPECIFIC ENERGY PER DESALINATION TECHNOLOGY
0
0
Feed Water
Fig. 2 Electro dialysis process
2) Reverse Osmosis: The movement of solvent

molecules from one space with low solute concentration to
another space through a semi-permeable membrane with a
higher solute concentration is what is called osmosis,
observed in 1748 [9]. What causes this movement is the
difference in chemical potential between the two solutions,
affected in its turn by three different factors: Salt
concentration, the higher the concentration the lower the
chemical potential; temperature, the higher the temperature,
the higher the chemical potential; pressure, the higher the
pressure the higher the chemical potential
This movement, or osmosis process, would carry on until
equilibrium in chemical potential between the two sides is
reached; this is the "osmosis equilibrium."
Therefore, in order to desalinate water, it is the reverse
process that has to be done, called "reverse osmosis." As
pointed up in Fig.3, some pressure has to be applied at the
beginning to initiate the flow from the saline side to the fresh
side; this is the "osmotic pressure."
T:U
- r
Osmotic PresS(Jre
Desalination Technology
Specific
Energy
3
(kWh/m )
Multi-stage Flash
Multiple Effect
6-9
Vapor Compression
Electrodialysis
0.7 - 2.2
Reverse Osmosis
3 - 13
10 - 14.5
15
Other factors could be taken into account as well such as

ease of maintenance and portability. At the end, RO seems to
be the technology with the highest potential.
In the literature review we have gone through [10], [11],
[12] very few implemented projects dealt with this
problematic.
Amongst these, a prototype system using
mechanical energy transmission between the wind turbine and
the water pump. This type of desalination using mechanical
energy instead of electrical faces several difficulties related to
more frequent failures due to the use of mechanical bearings,
and most importantly, the end result was not really satisfying,
since the water product quality was not high enough to be
drinkable even though it still could be used for irrigation.
Another prototype implemented in Coconut Island, Hawaii
uses electrical energy of the wind turbine, and a feedback
system that enables the control of water flow, thus stabilizes
its pressure. [t was able to achieve a cost of $5.4 per m3 [10].
A. Wind
Fresh Water
[15]
cathode
&
RO
Saline Water'
D'
Semi-permeable
Membrane
Fig. 3 Reverse osmosis process
Reverse Osmosis has been continuously improved over the

years making it one of the most energy efficient, and
representing more than half of the new desalination plants that
are being installed every year. This is mainly due to the
development of the main part of a RO system, its membranes.
Lots of efforts have been put into improving their
performance, their resistance to pressure fluctuations.

Eventually, to optimize their performance, they have to
undergo a specific flow rate and pressure, and this is the main
challenge with fluctuating wind energy [13], [14], [7].
under specific conditions; any extreme variation is expected to

cause mechanical fatigue and impact significantly the lifetime
of the membranes [11], [[6].
B. MATLAB Model
Thus, as demonstrated in Fig.4, that was made arbitrarily,
when designing the system, there are several conditions that
have to be respected to maximize the lifetime of the
membranes and avoid any significant deterioration:
Maximum feed Pressure (membrane resistance)
Maximum brine flow rate (membrane resistance)
Minimum brine flow rate (fouling problem)
Maximum product concentration (depends of osmotic
and applied pressure)
Since, there has not been much major research regarding

RO membranes operating under variable conditions, we
designed a MATLAB model simulation (Fig.S) in order to
investigate the effects of fluctuating wind energy on the
overall system.
1)
Wind Turbine
Pressure
F unct i on
Maxim um Pressure
Minimum
Brine Flow
Rate
RO membrane
O p erational
Maxi mum
Region
Brin e Flow
Consta nt
Rate
Fig.
Maxim u m
5 Wind turbine MATLAB model
Al lowe d
Co n c entration
Flow Rate
Fig. 4 RO membrane operating region under specific conditions
1) System with Backup: Basically, an additional energy

source (diesel generator, grid connection) will be used besides
wind energy. This will compensate for electricity coming
from the wind generator during low or no wind. This
represents an easy solution for the wind fluctuations; however,
this makes our system completely dependent on the additional
source. [n the case of energy shortage, or power cuts, the
system will simply stop operating.
The wind turbine model has been designed in a way to

gather wind speed data over a fixed period of time from a
spreadsheet file.
A saturation model was added, to limit the wind speed
input, assuming that a pitch mechanism would starts operating
when wind speed reaches 12 ms-1. In order to compute the
electrical power, we used the relation shown:
P ='h p A v3 Cp
(1)
Where the different parameters in the equation are:
Cp: Power coefficient (given a value of 0.4)

A: rotor area (computed for a turbine with 5.2m
diameter)
3
p: air density (l.225 kgm- )
v: wind speed (taken from an excel file)
2)
Whole System
2) Independent System (No Backup): To keep the

system operating at nearly steady conditions and without any
backup generators, there are three main ways:
Storage: either electrical storage with batteries or water
storage, by pumping water into a tanl<- when surplus of
electrical energy is produced
Switching ON/OFF modules: this requires the

availability of several independent reverse osmosis
systems all connected to the same wind energy
generator. Systems will be turned on or off depending
on the available energy [11], [\5].
Wind
Turbine Power limitation: this consists in
applying a pitch mechanism that would put a boundary
on the power produced from the wind turbine generator,
thus, limiting the power fluctuation [11], [15].
In the case of variable conditions, no major efforts are
taken into stabilizing the electrical power, flow rate or
pressure of the feed water going into the RO membranes.
Since, these filters have been designed to operate optimally
Fig. 6 Wind Powered Desalination System Model
This model (Fig.6) calculates, depending on the power

coming from the wind turbine generator, the feed water flow
rate and its pressure. The main purpose of this example is to
illustrate some of the effects of variable wind conditions on

our desalination system.
""
,
-Pr.uW'_
- .....nllSp.ad
'1ooI!'.,.
For the flow rate calculation we have used [17]:
With P representing the electrical power ( in kW), H the head

(given a value of 1m), g the gravitational acceleration (9.81
ms-\ and 11 the pump efficiency.
For the pressure calculation we used [18]:
QF being the feed flow rate calculated in the previous

equation, R the recovery rate (given a value of 0.5), p the
membrane permeability (given a value of 0.2464m3/m2h/bar),
Am the membrane area (given a value of 1m2), and n the
osmotic pressure
As it can be noticed in the graph resulting from the
simulation (Fig.7), wind speed variation leads to significant
pressure changes. Thus, from this simulation, a simple wind
powered desalination RO system may be regarded as not
feasible.
p,
-\\lndSpOIQd
rIawR:tc:
Fig. 8 De-rating impact on pressure variation
By reducing the maximum power produced by the wind

turbine, we simply diminish the pressure fluctuation;
therefore, we may reduce it down to an acceptable level that
would not result in mechanical damage.
Although this may be seen as an effective and easy
solution, it at the same time represents a waste of potential
electrical energy, due to pitching, or in other terms, waste of
investment.
C.
Parts a/the system
1) Variable Speed Drive: With an energy source as

intermittent as wind energy, use of variable speed pumps is a
necessity. This variable speed drive is controlled by the wind
turbine generator and includes: rectifier, DC link capacitor
and variable frequency inverter (Fig.9). This later component
will adjust the speed and torque of the pump depending on the
available power. This should allow the pump to continue
operating smoothly without sudden changes under variable
wind conditions.
Fig. 7 Wind speed impact on pressure and flow rate
The remedy for this situation could be de-rating the wind

turbine. Thus, by using the same model, but flattening the
electrical power coming from the wind turbine generator we
get the following results of de-rating shown in Fig.8.
Power Grid
Rectifier
DC Unk
In veTter f------j
Fig. 9 Variable Speed Drive
2) Pressure Stabilizer: Since, pressure seems to be the

most critical issue when it comes to wind powered
desalination with its irreversible effects on the membranes, it
is extremely vital for it to be dealt with. The simplest way that
comes to mind, in order to stabilize the pressure, would be
storage, but since, a minimum of pressure is required to be
able to produce fresh water through the reverse osmosis

system, a pressurized water tank would be needed [10].
From that point, a choice can be made on whether to have a
continuous or non-continuous operating system. In the non
continuous case feed water is stored and it is taped when
needed. On the other hand, the continuous system uses
solenoid valves (figure 10) controlled by pressure sensors.
The number of valves that open will depend on the pressure
within the tank [19]. The higher the pressure, the more valves
will open, to keep it within the acceptable operating range for
the membranes.
Feed Water
Pressure Sensors
Concerning, the operational costs, besides maintenance, an

annual change of RO membranes is assumed; a yearly amount
of $300 each was assumed. Thus, for a lifetime of about 20
years and an interest rate of 5%, the net present value of the
operational costs would be $4522.
Summing both the capital and operational costs would give
us, as a final total cost, a value of $18770.
With an average wind speed of 4m.s1 leading to a daily
average of around 3m3 of water produced, during the 20 years
lifetime of the project, the total water production would be
3
21 900 m .
Therefore, this wind powered desalination system will be
producing fresh water at a cost of 7.35 MAD/m3
IV. CONCLUSIONS
Valves
Fig.lO Pressure-Controlled Valves
3) Energy Recovery Device: One positive aspect of the

reverse osmosis process is that there is a very small drop
pressure across the RO vessels; in other words, the
concentrate stream keeps most of the initial pressure of the
feed water. Therefore, energy recovery devices have been
designed to make use of this significant amount of pressurized
brine that would, otherwise, be wasted through direct
discharge [20], [21]. That pressure will be transferred to some
of the feed water (Fig. I I), hence increasing the pressure
within the pressurized tanl<- faster than in typical conditions,
allowing for more water production.
High PresSlJre Feedwoter
Low Pressure Feedworer
----
By using a low wind speed, and a low water production, the

cost of water produced does not go beyond 7.35 MAD which
is a highly satisfying cost, especially when compared with
other countries, where the cost can go up to 30 MAD/m3, in
some of the Greek Islands, for example, where powered
desalinations systems with high wind integration have been
implemented [2].
Water Cost ($) per m3 Vs. Wind Speed (m.sl)
1
0,8
t;
0,6
0,4
0,2
o
4
High
'--_____L..----"
10
12
Pressure Brine
Wind Speed (m.sl)

Low Pressure Bnne
Fig. 12 Water production cost per wind speed

Fig. 11 Energy Recovery Device
Most ERDs incorporate positive displacement technology and

can be highly efficient going up to 96% of efficiency.
D. Cost Analysis
For the last part, the cost per unit of water produced will be
estimated.
Starting with the capital cost, the purchase of a 5.1 kW
wind turbine used for this analysis, a variable speed pump, a
small RO system, will cost around $13 700. According to
Energy Recovery Inc (ERI), ERDs cost around 4% of the
capital cost, which amounts to $548 [22].
Eventually, production cost could be much lower in windy

regions, and on higher implementation scale (figure 12).
Therefore, it is safe to say that independent wind powered
desalination systems will have a bright future in the coming
years.
For the Moroccan context, we believe that our
proposed system will be suitable for implementation in
regions where there's urgent need for water (either industrial
or sanitary); and thus all the coastal southern Saharan region.
In these regions all required conditions are present: important
wind resource (the Sahara trade winds), raw material (sea and
brackish water), in addition to a permanent need for fresh
water.
However, it is important to mention that this system stays

economically attractive only for low saline water desalination.
Thus, we propose for future work conducting additional
experimentation and research to effectively optimize the
system and make it economically attractive for all desalination
cases.
[19] C.K. Liu,
1. Park, M. Reef: Q. Gang. (2002). Experiments of a prototype
wind-driven reverse osmosis desalination system with feedback control.

[Online]. Available: https:llwiki.duke.edu.
[20]
1. Kaufler. (2006). Wind & Solar Powered Seawater Desalination.

[Online]. Available: http://events.exicon-intl.com.
[21] E.Kondili, K.Kaldellis. (nd). Wind Energy Based Desalination Process

and
Plants.
[Online].
Kondili.pdf.
REFERENCES
[22] Energy Recovery Inc. (2011). The Economics of Downtime. [Online].

Available: http://www.energyrecovery.com/whitepaper-.Jldfs
[1]
c.Generaal, "Wind Driven Reverse Osmosis Desalination for Small

Scale Stand-Alone Applications," M.S. thesis, Faculty of Aerospace
Engineering, Delft University of Technology, Delft, Netherlands, 2011.
[2]
E.Spang, "The potential for wind-powered desalination in water scarce

countries," M.S. thesis, Department of Law and Diplomacy, The
Fletcher School, Tufts University, Medford, Massachussets, 2006.
[3]
Matlock.
(2011).
Review
of
Water
Scarcity
Indices
and
Methodologies. The sustainability Consortium. [Online]. Available:

http://www.sustainabilityconsortium.org.
[4]
1. Chapman, D. Noll, A. Jackson, B. Ewing,
A.Swift, K.Rainwater,
L.Song, G.Ganesan, R. Marshall, V. Doon, P. Nash. (2009). Wind

Power and Water Desalination Technology Integration. Desalination
and Water Purification Research and Development Program Report No.
146.
[Online].
Available:
http://www.usbr.govlresearch/AWT/reportpdfs/reportI46.pdf.
[5]
E.Miller.
(2003).
Review
of
Water
Resources
and
Desalination
Technologies. [Online]. Available: http://prod.sandia.gov.

[6]
Tata. (2009). Dessalement de I'eau de mer : bilan des demieres avancees

technologiques ; bilan economique ; analyse critique en fonction des
Available
[Online].
contextes.
http://www.agroparistech.fr/IMG/pdfITATA-DUCRU_sr_final.pdf.
[7]
Department of Agriculture, Fisheries

Introduction
to
Desalination
Report.
& Forestry - Australia. (2002).
Technologies
in
Australia.
Summary
[Online].
Available:
http://www.environment.gov.au/water/publications/urban/pubs/desalinat
ion-summary.pdf.
[8]
H. Strathmann. (nd). Assessment of Electrodialysis Water Desalination

Process
Costs.
[Online].
Available:
http://gwri-
ic.technion.ac.iI/pdf/lDS/82.pdf.
[9]
Williams. (2003). A brief Review of Reverse Osmosis Membrane

Available:
[Online].
Technology.
http://www.eetcorp.com/heepm/RO_ReviewE.pdf.
[10] C.K. Liu. (2009). Wind-Powered Reverse Osmosis Water Desalination
for Pacific Islands and Remote Coastal Communities. Desalination and
Water Purification Research and Development Program Report No. 128.
Available:
[Online].
http://www.usbr.gov/researchlAWTIreportpdfs/report128.pdf.
[II] M.Marcos, "Small-Scale Wind-Powered Seawater Desalination Without

Batteries," PhD. dissertation, Univ. of Loughborough, Loughborough,
United Kingdom, 2003.
[12] E. Rabinovitch. (2008). Drinking with the wind.
[Online]. Available:
http://www.citg.tudelft.nl.
[13] O. Galal. (2011). Solar Desalination. [Online]. Available
http://dii
eumena.com.
[14]
1. Kaufler. (2012). Seawater Desalination (RO) as a Wind/Solar

Powered Industrial Process - Technical and Economic Specifics [Online]. Available: http://events.exicon-intl.com.
[15] M.Marcos, D.Infield. (2002). A wind-powered seawater reverse osmosis

system
without
batteries.
[Online].
Available:
http://www.spectrawatermakers.com/landbased/medialwindpowered_ro_nobatteries.pdf
[16] M. Goosen, H. Mahmoudi, N. Ghaffour, S. Sablani. (nd). Application of
Renewable
Energies
for
Water
Desalination.
[Online].
Available:
http://cdn.intechweb.orglpdfs/13755.pdf.
[17] GRUNDFOS. (nd). The centrifugal Pump. Grundfos Research and
Available:
[Online].
Technology.
http://www.grundfos.comlThe_Centrifugal]ump.pdf.
[18] Assimacopoulos. (nd). A tool for the design of desalination plants
powered
by
renewable
http://environ.chemeng.ntua.gr.
energies.
Available:
http://ikaros.teipir.gr/mechengiOPSIArchimedes/WRECX_wind_desal_
[Online].
Available:
International Journal of Green Energy, 4: 471481, 2007

Copyright Taylor & Francis Group, LLC
ISSN: 1543-5075 print / 1543-5083 online
DOI: 10.1080/15435070701583060
A WIND-POWERED SYSTEM FOR WATER

DESALINATION
Eyad S. Hrayshat
Tafila Technical University, Tafila, Jordan
A wind-powered reverse osmosis desalination system is proposed in order to assess the
potential of the development of water desalination in Jordan. A simulation model for the
prediction of the power delivered for a given value of wind speed is adopted. Based on the
average wind speed data and salinity of the feed water, the amount of water that can be produced at eight different sites is calculated. According to the annual amount of water produced, the selected sites can be divided into three different categories. The first one, which
includes Hofa and RasMuneef, is considered to be adequate for wind-powered reverse
osmosis desalination. Its annual amount of water output forms about 57% of all water produced at all the eight sites combined. The second category, which includes Safawy,
Twaneh, and Tafila, is considered to be promising. Its water output adds up to about 30%
of all water produced at all sites. The third category, which includes Jurf AlDaraweesh,
Aqaba, and Shoubak, is considered to be poor. Only about 13% of the water produced
from all sites combined can be obtained from these three sites.
Keywords: Reverse Osmosis; Wind; Desalination; Jordan
INTRODUCTION
The expanding population and the climatic and topographical conditions of Jordan
have exerted enormous pressure on the limited water resources and created a severe
water supplydemand imbalance where the renewable water resources are among the
lowest in the world, and are declining with time. Resources are already seriously limited
and are far below the water poverty line of 1000 m3/capita/year. Available water from
existing renewable sources is projected to fall from 160 m3/capita/year in 2002 to 90m3/
capita/year by the year 2025 (Malkawi, 2003). The supply-demand imbalance has influenced the quality of water resources where over-extraction from groundwater aquifers
exploited the aquifers at more than double their sustainable yield in the average.
Desalination of brackish or sea water now represents a consolidated system to
resolve the water emergency. The main drawback of this solution, however, remains the
high energy consumption. Considering their limited availability in Jordan, high cost and,
above all, the negative environmental impacts caused by their use, it is imperative to
search for new alternative sources to supplement or substitute for conventional fuels. In
view of the aforementioned problems, considering renewable energy resources such solar
Address correspondence to Eyad S. Hrayshat, Tafila Technical University, P.O. Box 66, Tafila 66110,
Jordan. E-mail: ehrayshat@yahoo.com
471
472
HRAYSHAT
and wind energies seems very attractive, especially for remote areas with no electricity
grid and which would be expensive to connect. Fortunately, Jordan is blessed with
abundant solar and wind energy sources (Hrayshat, 2002; Hrayshat and Al-Soud, 2004;
Hrayshat, 2005).
The use of solar energy for desalination plants in Jordan has been investigated by
many researchers (Al-Rawajfeh et al., 2003; Al-Rawajfeh et al., 2004; Mohsen and Jaber,
2001; Abdallaha et al., 2005). However, proper exploitation of wind energy as a source of
power for desalination plants in Jordan has not yet been investigated to the extent that
significant results and/or design methods could be obtained.
The objective of this paper is to assess the potential of wind-powered desalination as
a viable alternate water source for eight selected Jordanian sites. A wind-powered reverse
osmosis (RO) desalination system is proposed, and a simulation model is utilized for the
evaluation of the produced water amount based on the average wind speed data and salinity
of the feed water (TDS of 3000, 5000, 7000, and 10000 mg/L).
WIND DATA
The wind speed data used in this paper was measured and recorded at eight stations
distributed all over the country of the Jordanian Meteorological Department, at ten
meters above ground level, between 1990 and 2001. These stations are: Hofa, Ras
Muneef, Safawy, Twaneh, Tafila, Jurf AlDaraweesh, Aqaba, and Shoubak.
The data has been averaged over the twelve years. Each data is recorded every five
minutes and then averaged on an hourly basis and stored as hourly values. Monthly files
were obtained for each year, with the data recorded in four columns: month, day, hour,
and hourly mean wind speed. The hourly mean wind speed is the average of the twelve
pieces of data corresponding to the twelve periods of five minutes that make up each hour
of original data.
BRACKISH WATER IN JORDAN
In Jordan, two main sources are available to be desalted: the Aqaba Gulf and the
brackish deep groundwater in some basins. Preliminary studies showed that by the year
2010, more than 2 107 m3 of brackish water could be developed in central Jordan. This
figure may reach 7 107 m3 by the year 2040 (Jaber and Mohsen, 2001). According to the
water quality analysis conducted by the Japanese International Cooperation Agency
(JICA) on brackish water in Jordan, the total dissolved solids results (TDS) were in the
range of 5,00010,000 ppm (JICA, 1995). The salinity of water is around 500010,000 mg/L
as TDS with water temperatures of 3236C. NaCI is the main component of salt in water
(30006000 rag/L), besides the cations of Ca, Mg and the anions of SO4 and HCO3 that
are considered to be scaling substances existing in relatively high concentrations, and the
calculated total hardness is in the range of 15003000 mg/L as CaCO3. The Fe concentration is 515 mg/L. As for SiO2, which is a fouling substance for membranes, its
concentration is in the range of 1020 mg/L.
THE WIND-POWERED RO DESALINATION SYSTEM
The wind-powered RO desalination system consists of the membrane separation
section, which is fed via a high pressure reciprocating pump (pressurizing the feed
WIND-POWERED SYSTEM FOR WATER DESALINATION
473
stream up to the desirable pressure levels), and which is properly connected to a hydroturbine, for the recovery of energy by the brine stream leaving the process. The permeate stream leaving the membranes constitutes the lean product of the system. The high
pressure pump operates by means of a three-phase motor which is supplied by electrical
power. In this case, electrical power is available by the wind turbine. The technical data
of the utilized wind turbine is furnished in table 1. An advanced control system is
required for the regulation of the power source. The schematic diagram of the system is
shown in Figure 1.
The volume of a cylinder of air approaching a rotating wind turbine is given by the
following equation (Fanchi, 2004):
n = AL
(1)
The mass of the cylinder of air equals:

(2)
m = ra n = ra AL
Table 1 Technical data of the utilized wind turbine.

Cut in speed
(m/s)
2.5
Cut out speed

(m/s)
rated speed
(m/s)
Rated power
(kW)
Rotor diameter
(m)
CP
17
5.6
0.36
Product
D
Membrane
Wind Turbine
Turbine
Pump
M
Motor
Feed
Brine
Figure 1 Schematic diagram of the wind-powered reverse osmosis desalination system.
474
HRAYSHAT
Assuming that the cylinder of air is moving with speed v directly at the turbine, then
the kinetic energy of the moving air will be:
KE = 0.5mn 2 = 0.5ra ALn 2
(3)
The length of the cylinder of air that reaches the wind turbine in a time interval t is:
L = n t
(4)
Substituting eq. (4) in eq. (3) gives:
KE = 0.5ra An 3 t
(5)
The rate of air arrival is the wind power, which equals:
pw =
KE
= 0.5ra An 3
t
(6)
The area A is the surface area of the circle formed by the rotating tip of the rotor
blade, which equals:
A = pR 2
(7)
Using eq. (7), eq. (6) can be rewritten in the following form:
pw =
p
ra R 2n 3
2
(8)
Eq. (8) shows that wind power is proportional to the square of the radius of the fan
created by the rotating rotor blade.
The power in the wind is converted to mechanical power with an efficiency (coefficient of performance) Cp, which is transmitted to the generator through a mechanical
transmission with efficiency m and which is converted to electricity with an efficiency
g. The electrical power output is then:
pe = C p h m h g rw =
p
C p h m h g ra R 2n 3
2
(9)
Using the aforementioned model, the power delivered for a given value of wind
speed is predicted. The relationship between energy consumption and water salinity for
the RO system is shown in Figure 2. With brackish water of 2000 mg/L and 5000 mg/L
TDS, the amount of energy required is 1.1 kWh/m3 and 1.6 kWh/m3, respectively (JICA,
1995).
RESULTS AND DISCUSSION

Figure 3 exhibits the monthly average of the measured wind speeds between 1990
and 2001 for all the selected sites. Ras Muneef and Hofa have the most potential among
the selected sites in terms of their wind speed values.
SPECIFIC ENERGY CONSUMPTION [kWh/m3]
475
2.5
1.5
0.5
0
0
2000
4000
6000
8000
10000
TDS [mg/L]
Figure 2 Energy consumption of the reverse osmosis desalination system as a function of the total dissolved
solids content.
Tafila
Aqaba
Twaneh
RasMuneef
10
9
WIND SPEED [m/s]
Shoubak
Jurf Aldaraweesh
Safawy
Hofa
7
6
5
4
3
2
1
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
0
MONTH OF THE YEAR
Figure 3 Monthly average wind speeds for all the selected sites.
Using the proposed mathematical model, the power delivered for a given value of
wind speed is predicted. Then the relationship between energy consumption and water
salinity for the RO system (see figure 2) is used for the calculation of the amount of water,
produced by the system.
Figures 411 show the daily water production during a 1-year cycle at the selected
sites for different values of TDS (3000, 5000, 7000, and 10000 mg/L). As depicted in figures 4
and 5, Hofa is the most potential site for wind powered RO desalination, followed by Ras
476
HRAYSHAT
WATER PRODUCTION [m3]
12
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
10
8
6
4
2
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 4 The daily water production during a 1-year cycle at Hofa as a function of the total dissolved solids
content.
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
6
5
4
3
2
1
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 5 The daily water production during a 1-year cycle at Ras Muneef as a function of the total dissolved
solids content.
Muneef. With a TDS of 7000 mg/L (the actual brackish water salinity at these sites is in
the range of 5000 to 10000 mg/L), the amount of water that can be produced at these sites
during the month of July are 4.8 and 3.1 m3/ day respectively.
Figure 12 exhibits the annual amount of water, which can be produced at all the
eight selected sites for different values of TDS. Based on the obtained results, the selected
sites can be divided into three different categories: the first one, which includes Hofa and
3000 mg/L
5000 mg/L
7000 mg/L
477
10000 mg/L
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 6 The daily water production during a 1-year cycle at Safawy as a function of the total dissolved solids
content.
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 7 The daily water production during a 1-year cycle at Twaneh as a function of the total dissolved solids
content.
Ras Muneef, is considered to be adequate for wind-powered RO desalination. The

annual amount of water produced at these two sites forms about 57% of all water produced at all the eight sites combined. The second category is considered to be promising. It includes Safawy, Twaneh, and Tafila. Their water output adds up to about 30% of
all water, produced at all sites. The third category, which includes Jurf AlDaraweesh,
478
HRAYSHAT
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 8 The daily water production during a 1-year cycle at Tafila as a function of the total dissolved solids
content.
2.5
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
1.5
0.5
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 9 The daily water production during a 1-year cycle at Jurf AlDaraweesh as a function of the total
dissolved solids content.
Aqaba, and Shoubak is considered to be poor, because only about 13 % of the water
produced from all sites combined can be obtained from these three sites. It is obvious that
wind powered RO desalination at these sites is not a good option. Therefore, other alternatives for water desalination should be taken into account.
CONCLUSIONS
In terms of their potential for wind-powered RO desalination, some of the selected
sites were considered to be adequate. They include Hofa and Ras Muneef. Hofa is
1.8
3000 mg/L
5000 mg/L
7000 mg/L
479
10000 mg/L
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 10 The daily water production during a 1-year cycle at Aqaba as a function of the total dissolved solids
content.
0.3
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
0.25
0.2
0.15
0.1
0.05
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
MONTH OF THE YEAR

Figure 11 The daily water production during a 1-year cycle at Shoubak as a function of the total dissolved solids
content.
considered to be the best among all of the selected sites for wind-powered RO desalination. Other sites were considered to be promising. They include Safawy, Twaneh, and
Tafila. The rest of the sites studied, namely Jurf AlDaraweesh, Aqaba, and Shoubak are
considered to be poor for wind-powered desalination. Therefore, other alternatives for
water desalination should be taken into account at these sites.
480
HRAYSHAT
2000
3000 mg/L
5000 mg/L
7000 mg/L
10000 mg/L
ANNUAL WATER PRODUCTION [m3]
1750
1500
1250
1000
750
500
250
Shoubak
Aqaba
Jurf Aldarawwesh
Tafila
Twaneh
Safawy
Ras Muneef
Hofa
Figure 12 Annual water production for all the selected sites as a function of the total dissolved solids content.
NOMENCLATURE
A
CP
Pw
R
v
KE
L
m
t
V
The cross-sectional area

Power coefficient
Wind power
Radius of the rotor blade
Wind speed (m/s)
The kinetic energy of the moving air
The length of the cylinder of air
The mass of the cylinder of air
Time interval
Volume of a cylinder of air
Greek letters
a
Air density (kg/m3)
REFERENCES
Abdallaha, S., Abu-Hilal, M., Mohsen, M.S. (2005). Performance of a photovoltaic-powered reverse
osmosis system under local climatic conditions. Desalination 183: 95104.
Al-Rawajfeh, A., Glade, H., Ulrich, J. (2003). CO2 release in multiple effect distillers controled by
mass transfer with chemical reaction. Desalination 156: 109123.
481
Al-Rawajfeh, A., Glade, H., Qiblawey, H., Ulrich, J. (2004). Simulation of CO2 release in multipleeffect distillers. Desalination 166: 4152.
Fanchi, J. (2004). Energy: technology and directions for the future. London: Elsevier Academic
Press.
Hrayshat, E. (2002). Wind energy in Jordan: current status and future potential. Proc. World Renewable Energy Congress-VII. Germany.
Hrayshat, E. (2005). Wind availability and its potential for electricity generation in Tafila/Jordan.
Renewable and Sustainable Energy Reviews 9: 111117.
Hrayshat, E., Al-Soud, M. (2004). Solar energy in Jordan: current state and prospects. Renewable
and Sustainable Energy Reviews 8: 193200.
Jaber, J.O., Mohsen, M.S. (2001). Evaluation of non-conventional water resources supply in Jordan.
Desalination 136: 8392.
JICA. (1995). Final report on brackish ground water desalination in Jordan. Amman, Jordan.
Malkawi, S.H. (2003). Wastewater Management and Reuse in Jordan. Proc. First Regional Water
Reuse Conference. Jordan.
Mohsen, M.S., Jaber, J.O. (2001). A photovoltaic-powered system for water desalination. Desalination 138: 129136.
Desalination 277 (2011) 274280
Contents lists available at ScienceDirect
Desalination
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / d e s a l
Wind energy technologies integrated with desalination systems:

Review and state-of-the-art
Qingfen Ma a,, Hui Lu b
a
b
Department of Mechanical Engineering, Hainan University, Danzhou, 571737, China

Institute of Environment and Plant Protection, Chinese Academy of Tropical Agriculture Sciences, Danzhou, 571737, China
a r t i c l e
i n f o
Article history:
Received 13 January 2011
Received in revised form 11 April 2011
Accepted 14 April 2011
Available online 6 May 2011
Keywords:
Renewable energy
Wind energy
Seawater and brackish water
Desalination
a b s t r a c t
Throughout the world, desalination is intensively used as a means to reduce current or future water scarcity,
especially for the coastal areas. However, the dramatic increase in desalinated water supply will create a series
of problems, the most signicant of which are those related to energy consumption and environment impacts.
Renewable energy provides an energy security and environmental friendly option simultaneously when
decreasing global reserves of fossil fuels threatens the long-term sustainability of global economy. Thus, the
integration of renewable resources in desalination and water purication is becoming increasingly attractive.
In this paper an attempt has been made to present a review, in brief, work of the highlights that have been
achieved during the recent years worldwide and the state-of-the-art for most important efforts in the eld of
desalination by wind energy, which is one of the most common form of renewable energies. The wind energy
transform patterns, modeling and experimental studies of various wind energy powered desalination plant,
and the prototypes established worldwide are majorly discussed. Moreover, two important technological
problems in wind utilization are discussed, and the present or potential countermeasures for the intermittent
characteristic and direct utilization of wind energy are presented.
Crown Copyright 2011 Published by Elsevier B.V. All rights reserved.
1. Introduction
Water and energy are two inseparable items that govern our lives
and promote civilization. The social and economic health of the
modern world depends on sustainable supply of both energy and
water. As of today, about three billion people have no access to a
secure source of fresh water and about 1.76 billion people live in areas
already facing a high degree of lacking water [1]. Meanwhile, with the
increase of population, industrial and agricultural activities, available
water resources has been excessively exploited and severely polluted.
The need for fresh water is at the top of the international agenda of
critical problems, at least as rmly as climate change.
Because of the growing scarcity of freshwater, a trend to intensied
use of desalination as a means to reduce current or future water
scarcity can be observed. Seawater or brackish water can be desalted
and supplied in large quantity, but this will create a severer series of
problems, the most signicant of which are those related to energy
consumption and environment impacts. If desalination is accomplished by conventional technology, it will require the burning of
substantial quantities of fossil fuels, which will aggravate the energy
crisis worldwide and environment pollution. The particular environmental impact in desalination system is not well known yet, many
Corresponding author. Tel.: + 86 898 31132006.

E-mail address: mqf0920@gmail.com (Q. Ma).
environmental studies related with desalination technologies are

being conducted [2].
Renewable energy provides a variable and environmental friendly
option and national energy security at a time when decreasing global
reserves of fossil fuels threatens the long-term sustainability of global
economy. The integration of renewable resources in desalination and
water purication is becoming increasingly attractive. However, at
present, total worldwide renewable desalination installations amount
to capacities is less than 1% of that of conventional fossil fuel
desalination plants [3]. This is due mainly to the high capital and
maintenance costs required by renewable energy, making these
desalination plants noncompetitive with conventional fuel desalination plants. However, the cost of renewable energy systems has been
signicantly reduced during the last decades. Therefore, future
reductions as well as the rise of fossil fuel prices could make possible
the competitiveness of seawater or brackish water desalination driven
by renewable energies.
Solar thermal and photovoltaic (PV) systems, wind power,
biomass, oceanic, geothermal and nuclear energy etc. are the basic
kinds of renewable energy used nowadays. Among above renewable
energies, wind energy has been maturely used for power production
and wind turbines are commercially available on a wide range of
nominal power. The electrical or mechanical power generated by a
wind turbine can be used to drive desalination plants. The wind
powered desalination systems are one of the most frequent
renewable desalination plants, especially for coastal areas presenting
0011-9164/$ see front matter. Crown Copyright 2011 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.desal.2011.04.041
Q. Ma, H. Lu / Desalination 277 (2011) 274280
a high availability of wind energy resources [4]. In addition, according

to some authors, among the various renewable energy resources, the
integration of desalination with wind energy had the least impact on
the environment, with an important environmental impact reduction
of 75% [5]. Thus, in high wind-potential areas where desalination is
also required, wind energy is the preferred energy source option.
In this paper, the status and development of coupling wind energy
systems with desalination units are reviewed. The following discussion concentrates on the main wind-driven desalination models and
experiments as well as prototypes and implementations, aiming to
trace the development process and the problems arising, and light up
their perspective characteristics and trends.
2. Matching wind energies with desalination units
2.1. Desalination technologies
Currently available desalination technologies can be mainly categorized into two groups:
(1) Thermal desalination (phase change process) that involves
heating the feed (seawater, brackish water or other impaired
water) to boiling point at the operating pressure to produce
steam, and condensing the steam in a condenser unit to
produce freshwater. Thermal desalination process includes
multi-stage ash (MSF), multi-effect distillation (MED), mechanical/thermal vapor compression (MVC/TVC), membrane
distillation (MD) and solar distillation (SD).
(2) Membrane desalination (non-phase change process) that
involves separation of dissolved salts from the feed waters by
mechanical or chemical/electrical means using a membrane
barrier between the feed (seawater or brackish water) and
product (potable water). In the membrane desalination, the
reverse osmosis (RO) and electro-dialysis (ED) are technologies used frequently.
The dominant desalination processes are MSF and RO44% and
42% of world wide capacity, respectively. The MSF represents more
than 93% of the thermal process production, while RO process more
than 88% of membrane process production [4].
2.2. The coupling interface between wind energy and desalination unit
Wind energy and desalination plants are two different technologies, which can be coupled in various ways. The interface between the
wind energy system and the desalination system is met at the place/
subsystem where the energy generated by wind energy is promoted to
275
the desalination plant. Considering that the energy requirements for

desalination continues to be a highly inuential factor in system costs,
the integration of renewable energy systems with desalination seems
to be a natural and strategic coupling of technologies. Currently, wind
energy can power desalination plants directly or indirectly through
four types of energy media: electricity, thermal energy, gravitational
potential energy and kinematical power (shaft power). Fig. 1 shows
the existing interfaces between wind energy and desalination unit
which can be found in the following discussions.
Electricity is the most commonly used energy form as the interface
between wind energy and desalination process. After having changed
into electricity, the energy from wind plant can be employed to drive
desalination processes such as green house, RO, ED and MVC [69]. The
wind plant can be on or off the grid. For the intermittent characteristic
of wind power, usually backup facilities like battery, water tank,
ywheel system might be integrated into the system to store or release
energy when the wind speed exceeds or cannot achieve the required
value.
The technique of direct conversion from wind energy to thermal
energy has been studied for room heating and hot-water supply
systems because the efciency of direct wind-thermal conversion is
higher than that of wind-electricity conversion and their structures
are simpler. Nakatake and Tanaka proposed a newly designed,
maritime lifesaving small distiller. The wind energy was directly
converted to frictional thermal energy to heat the distiller. The
proposed distiller could be driven by wind only and was predicted to
produce 1.5 kg/d or more when a 6 m/s wind blew steadily all day on a
sunny or cloudy day [10].
To reduce the energy loss caused by the wind-electricity
conversion, gravitational energy has also been used as the interface
between wind energy and desalination process. Fadigas and Dias
designed an alternative conguration to conventional RO desalination
systems by incorporating the use of gravitational potential energy,
without using either electricity or fossil fuels. The gravitational
potential energy, presented by water stored in a reservoir above a
certain height, was converted by wind energy from windmills (or
wind turbines) [11].
Besides, interesting experimental research about directly coupling
the kinematical power from wind turbines and a desalination unit has
been carried out. Projects AERODESA I and AERODESA II of Canary
Islands Technological Institute included the direct coupling of wind
energy and RO unit by means of shaft power [4]. In Coconut Island off
the northern coast of Oahu, Hawaii, a brackish water desalination
wind-powered RO plant was established. The system was driven
directly by the shaft power of a windmill using a high pressure pump.
The water production rate can be maintained at 13 l/min for wind
Fig. 1. Existing interfaces between wind energy and desalination unit.
276
Fig. 2. Breakdown of renewable energy powered desalination system technologies

implemented worldwide [19].
speed of 5 m/s [12]. Witte et al. proposed WindDeSalter Technology

and analyzed its feasibility by calculation. The core of this technology
was using the substantial part of the available kinematical energy
from a WEC (wind energy converter) directly to drive the compressor
of an MVC plant or the high-pressure pump of a RO plant. The WEC
was also integrated with all necessary functional elements such as
seawater reservoir, ltering installation, pump units, desalination
units, compressor, heat exchanger and drinking water storage tank
[13].
3. Wind + desalination units
3.1. Wind + RO
Reverse osmosis (RO) is a pressure-driven process that separates
two solutions with different concentrations across a semi-permeable
membrane [14]. RO system major components include membrane
modules, high-pressure pumps, power plant, and energy recovery
devices as needed. RO is one of the most efcient desalination
technologies, requiring about 310 kWh of electric energy per m3 of
freshwater produced from seawater [15]. Since RO is the desalination
process with the lowest energy requirements and coastal areas present
a high availability of wind power resources, according to some authors,
wind powered RO plants appear to be one of the most promising
alternatives of renewable energy desalination [1618]. Fig. 2 [19]
shows a breakdown of renewable energy powered desalination
system technologies implemented worldwide. The most common
wind powered water treatment systems in the past were RO and windpowered RO systems made up approximately 19% of total RES
desalination facilities, second only to photovoltaic-powered RO units
(32%).
3.1.1. Feasible and economic evaluation
Several simulation studies have been done to discuss the feasibility
of wind powered RO technologies, based on various models with
different emphases. Feron was among the rst to evaluate the economic
feasibility of a wind-powered RO plant by mathematical modeling
analysis under some assumptions. The author concluded that the
economic use of a wind-powered RO plant might be restricted to areas
with high wind speeds and fuel prices. However, it could become more
economic because of current developments such as decreasing RO plant
costs and wind turbine cost, and steady or increasing fuel costs [20].
Later, Habali and Saleh conducted a cost analysis of a wind-assisted
RO system for desalinating brackish groundwater in Jordan. The
authors stated that it would cost less to desalinate brackish water with
a wind-assisted RO system than with a conventional diesel-powered

system [21]. Kiranoudis et al. performed a detailed analysis of a windpowered RO plant. Not only different wind turbines and membranes
were analyzed, but also seawater and brackish water feed were
considered. Moreover, generalized design curves for process structural
and operational parameters were derived [22].
Voivontas et al. developed a method to evaluate the potential
market for RES (renewable energy resources) powered desalination
systems. The results showed that Aegean Islands were the most arid
areas in Greece, with abundant wind energy. Thus most windpowered RO desalination plants could operate economically in most
of the arid Aegean Islands [23]. They also explored a computer-aided
design tool as a means to compare the alternative options on the basis
of economic indicators, combining technologies that guarantee the
desalination energy needs. Using the model, the effects of critical
design parameters on the water selling price were analyzed for a
wind-powered RO plant [24].
Garca-Rodriguez et al. analyzed the inuence of the main
parameters on the cost of fresh water: climatic conditions, nominal
power of the wind turbine, salt concentration of seawater or brackish
water, design arrangement, operating conditions, plant capacity, cost of
RO modules and cost of wind turbines [16]. Romero-Ternero et al. [25]
quantied the unit cost of fresh water generated from representative
wind-powered seawater RO system as well as the exergy efciency of
the process by means of thermoeconomy. The unit cost of freshwater
was determined exclusively by considering a wind-powered desalination system. The exergoeconomic analysis showed that the wind-driven
seawater RO desalination was cost-effective for the representative wind
power site with medium plant capacity and the unit cost of freshwater
was 76 c/m3 [25].
In the past ve years, with the rapid development of both wind
energy and RO technologies and the increasing aggravation of the
conventional energy crisis, feasibility and economic analyses of wind
powered RO plant appear to be more important to assist the design,
site selection, water production, and cost/price estimation etc. Koklas
and Papathanassiou [26] proposed a logistic model to provide insight
in the component selection criteria of an autonomous wind-driven RO
plant. The simulation of the system operation was performed
employing a variety of different congurations with respect to the
size of its main components (wind turbine, RO plant and batteries).
For each case, the annual water production was calculated, an
economic assessment performed and the water production cost
estimated [26].
Forstmeier et al. [18] developed physics-based system models to
conrm the technical feasibility of using wind as the power source for
desalination, including both RO and MVC units. The results showed
that the costs were in line with what was expected for a conventional
desalination system, proving to be particularly cost-competitive in
areas with good wind resources that had high costs of energy. Thus
wind-powered desalination could be competitive with other desalination systems, providing safe and clean drinking water efciently in
an environmentally responsible manner [18]. An integrated model
incorporated in the REDDES software for the use of renewable
energies (wind, solar) in the desalination of seawater was developed
by Koroneos et al. [27]. The desalination technologies (mainly RO and
MVC) were coupled with RES power systems to produce potable
water at the lower possible cost. The results indicated that water
production costs of an RES-desalination conguration depended
heavily on the available RES potential. The greater the RES potential
the smaller the energy production cost from the RES unit and thus
smaller water production cost from the desalination unit.
Spyrou and Anagnostopoulos [28] proposed a RO desalination unit
powered by wind and solar electricity production systems and by a
pumped storage unit. A specic computer algorithm was developed to
simulate the entire plant operation and perform economic evaluation
of the investment. Design optimization studies of the plant for various
objectives were conducted, like the minimization of fresh water

production cost or the maximization of water need satisfaction [28].
Koutroulis and Kolokotsa [29] presented a methodology for the
optimal sizing of PV modules and wind-generator powered RO
systems. Among a list of commercially available system devices, the
optimal number and type of units were determined so that the
20-year round total system cost was minimized, while simultaneously
the consumer's water demand was completely covered. The corresponding optimal sizing results indicated that the total cost of the RO
system was highly affected by the operational characteristics of the
devices comprising the system [29].
Bourouni et al. [30] proposed a new model based on the Genetic
Algorithms allowing the generation of several individuals (possible
solutions) for coupling small RO unit to RES to minimize the total
water cost. A particular interest was focused on the hybrid systems
(PV/WIND/Batteries/RO), and a case of PV/RO unit, installed since
2007 in Ksar Ghilne village of southern Tunisia was studied [30].
3.1.2. Prototypes and installations
The prototypes of wind-powered RO desalination system have
been reported in many regions of the world and a range of
experiments has been conducted with various concentrations. So
far, most of these installations, either connected to a utility network or
operating in a stand-alone mode, have been installed in Europe.
In France, as early as 1982, a small system was set at Ile du Planier.
It was a 4 kW turbine coupled RO desalination unit with production
rate of 0.5 m3/h. The system also was designed to operate via batteries
[31]. Another case where wind energy and RO was combined is that of
the Island of Drence in 1990. The wind turbine, rated at 10 kW, was
used to drive a seawater RO unit [32]. A very interesting experience
was gained at a test facility in Lastours, where a 5 kW wind turbine
provided energy to a number of batteries (1500 Ah, 24 V) and via an
inverter to an RO unit with a nominal power of 1.8 kW.
In Spain, the desalination leader of Europe [33], a pilot windpowered RO plant was installed at Canary Island, in 1984. It was
connected to the grid as auxiliary energy when the wind power was
not enough for plant operation, with production rate of 200 m3/d and
energy consumption of 5 kWh/m3 [34]. In 1993, a 56 m3/d hybrid
diesel-wind-RO plant providing fresh water and electricity for local
people and a battery-less wind-RO plant started operating at Pjara,
Fuerteventura Island. The system consists of two diesel engines and a
wind turbine of 225 kW. The Canary Islands Technological Institute
developed the concept AEROGEDESA based on the long-term
experience accumulated at Canary Islands, referring to a compact,
stand-alone wind-RO system with capacities between 5 and 50 m3/d
[35]. A wind/RO system without energy storage was developed and
tested within the JOULE Program in 2001 by the University of Las
Palmas. The RO unit had a capacity of 43113 m3/h, and the W/G had a
nominal power of 30 kW [17]. Recently, experience of a seawater RO
plant with capacities of 5000 m3/d in Gran Canaria, Canary Islands has
been reported. The power produced by the wind generators was
variable throughout the year; at times excess power was sold to the
conventional power network in place, and sometimes the RO plant
consumed supplementary power from the network grid [36].
Additionally, a 500 l/h seawater RO unit driven by a 2.5 kW wind
generator without batteries was developed and tested by the Centre
for Renewable Energy Systems Technology (CREST) UK. The system
operated at variable ow, enabling it to make efcient use of the
naturally varying wind resource, without need of batteries [7].
Excellent work on wind/RO systems has been done by ITC within
several projects such as AERODESA, SDAWES and AEROGEDESA [37].
A great job on the combination of wind/RO has also been done by
ENERCON, the German wind turbine manufacturer. ENERCON provides modular and energy-efcient RO desalination systems driven by
wind turbines (grid-connected or standalone systems) for brackish
and seawater desalination. Market-available desalination units from
277
ENERCON range from 175 to 1400 m3/d for seawater desalination and
350 to 2500 m3/d for brackish water desalination. These units
combine with other system components, such as synchronous
machines, ywheels, batteries and diesel generator, supply and
store energy and water precisely according to demand [38].
Other wind-driven RO systems in Europe are as follows:
A RO system driven by a wind power plant, in Island of the County
Split and Dalmatia [39];
Island of Suderoog (North Sea), with 69 m3/d [40];
Island of Helgoland, Germany, with 2 480 m3/d [40];
Island of St. Nicolas, West France, hybrid wind-diesel [40];
Island of Drenec, France, with wind energy converter of 10 kW [40];
Ile du Planier, France Pacic Islands, with production rate of 0.5 m3/h
[40].
Except Europe, engineers of other regions of the world have also
made efforts to install and test the wind-RO desalination units. In
1986, the installation of a RO plant in the Middle East began. It was a
25 m3/d plant connected to a hybrid wind-diesel system [41]. When a
second RO plant of 168 m3/d was commissioned by the Water
Authority of Western Australia, at Denham in Shark Bay, in 1991, the
power requirements exceeded the diesel grid's capacity and a
supplementary 30 kW West wind turbine was installed to power
the plants. Being grid connected, the plant imported power when
required and exported power back to the grid when excess power was
generated [42]. In Coconut Island off the northern coast of Oahu,
Hawaii, a brackish water desalination wind-powered RO plant was
installed. The system was using directly the shaft power production of
a windmill with the high pressure pump and RO. In particular a
constant fresh water production of 13 l/min can be maintained for
wind speed of 5 m/s [12].
3.2. Wind + MVC

Although mechanic vapor compression (MVC) consumes more
energy than RO, it presents fewer problems due to the uctuations of
the energy resource than RO. MVC systems are more suitable for
remote areas since they are more robust, and they need fewer skilled
workers and fewer chemicals than RO systems. In addition, they need
no membrane replacement and offer a better quality product than RO.
In case of contaminated waters, the distillation ensures the absence of
microorganisms in the product.
Few applications have been implemented using wind energy to
drive a mechanical vapor compression unit. A pilot plant was installed
in 1991 at Borkum Island (Germany), where a wind turbine with a
nominal power of 45 kW was coupled to a 48 m3/d MVC evaporator,
with a 36 kW compressor [43]. The experience was followed in 1995
by another larger plant at the Rgen Island, in Baltic Sea, with a
capacity of 360 m3/d and wind energy production capacity of 300 kW
[44].
Additionally, a 50 m3/d wind MVC plant was installed by Instituto
Tecnologico de Canarias (ITC) in Gran Canaria, Spain, within the Sea
Desalination Autonomous Wind Energy System (SDAWES) project
[45]. The wind farm was composed of two 230 kW wind turbines, a
1500 rpm ywheel coupled to a 100 kVA synchronous machine, an
isolation transformer located in a specic building, and a 7.5 kW
uninterruptible power supply located in the control dome. A detailed
analysis of the inuence of the main parameters of wind powered
MVC systems was performed by Karameldin et al [46]. The study
indicated that the operating evaporator temperature and temperature
difference recommended were 50 C and 3 C respectively. Under the
average prevailing wind speed in these areas, the system productivities were 203, 398 and 938 m3/d when the wind turbine
diameters were 20, 28 and 43 m respectively [46].
278
3.3. Wind + ED
Finally, ED process is interesting for brackish water desalination
since it is able to adapt to changes of available wind power and it is
more suitable for remote areas than RO. Modeling and experimental
results of on-grid tests of installed such system at the ITC, Gran
Canaria, Spain was presented by Veza et al. The main goal of this
project was to test and identify the most suitable desalination systems
for connection to the medium off-grid wind farm. The capacity range
of this plant was 19272 m3/d [47]. Later, they developed an
operational envelope for the electrodialysis reversal unit, off-grid,
i.e., only coupled to the wind farm. The desalination unit showed good
exibility, adapting smoothly to variations in wind power, even when
sudden drops or rises occurred [48].
4. Challenges and emerging/potential countermeasures to the
utilization of wind energy in desalination process
4.1. Intermittent characteristic and emerging countermeasures
Since the intermittent characteristic of wind energy, the desalination system driven by completely wind energy is affected by power
variations and interruptions. The power variations, however, have an
adverse effect on the performance and component life of certain
desalination equipment. Hence, back-up systems might be integrated
into the system to reduce the effect. Meanwhile, there is also another
solution, to integrate wind energy with other energy source, either
conventional or renewable energy, such as solar PV or thermal, diesel
[49], etc.
4.1.1. Integration wind energy with other energy source
The complementary features of wind and solar resources make the
use of hybrid wind-solar systems to drive a desalination unit a
possible alternative. Solar energy desalination is generally the
collecting of solar thermal energy that is used for desalination directly
in solar stills, or that is converted to electricity by photovoltaic (PV)
process rst and then used in either thermal of membrane processes
for desalinations [50,51].
4.1.1.1. Wind/PV hybrid system. As early as 1979, Petersen et al.
reported two RO-desalination plants with the GKSS-Research Centre
(Germany) plate module system supplied by a 6 kW wind energy
converter and a 2.5 kW solar generator for remote areas [52]. Later,
they reported another two such prototypes which were installed in
the Northern part of Mexico (Concepcin del Oro) and in a small
island at the German coast of the North Sea (Soderoog) [53]. The
Cadarache Centre (France) designed another unit that was installed in
1980 at Borj Cedra (Tunisia). The system consisted of a 0.1 m3/d
compact solar distiller, a 0.25 m3/h RO plant and an ED plant for 4 g/l
brackish water. The energetic system consisted of a photovoltaic eld
of 4 kW peak and two wind turbines [31].
Test results of a PV/wind powered brackish water RO plant
installed in Israel were reported by Weiner et al [54]. Its production
was 3 m3/d and expected life-span was 15 years. Two-day battery
storage and a diesel generator were built to serve as the back-up of the
system. The test results showed that the optimum ratio between the
power of a desalination unit and PV/wind peak power was of the
order of 3050% [54].
Kershman et al. presented a hybrid wind/PV powered RO
desalination plant implemented on Libya's coast of the Mediterranean
Sea. The nominal production of the plant was intended to be 300 m3/d
to supply a village with potable water. While the expected nominal
power load for the operation of the RO desalination system was
70 kW (net power after recovery), the solar PV system was designed
for 50 kW, and the WEC for 200 kW nominal outputs. The facility
design was exible for the integration of a diesel generator and

electrochemical storage [55,56].
Except for the existing desalination plants, some novel ideas or
analyses for wind/PV hybrid systems have also been made. Mohamed
et al. developed a simplied method for sizing and simulating a hybrid
wind-PV powered RO desalination unit based on a techno-economic
analysis. The water production cost calculated (5.21 /m3) was very
promising compared to the water transportation cost by tankers in
some Greek islands that can reach 612 /m3, far below transportation cost of 20 /m3 in Algeria [57,58]. Gilau and Small analyzed the
cost-effectiveness of a stand alone small-scale renewable energypowered seawater reverse osmosis (SWRO) system for developing
countries by a new methodology and an energy optimization model.
Applying the model, using the wind and solar radiation conditions for
Eritrea, East Africa, they computed that for a two-stage SWRO system
with a capacity of 35 m3/day, the specic energy consumption was
about 2.33 kW h/m3, which was a lower value than that achieved in
most of the previous designs [59].
4.1.1.2. Wind/solar thermal hybrid system. Recently, wind energy has
been attached with the solar still to power the desalination process
together with solar energy. Mohamed and Zhao [60] designed,
fabricated and evaluated a new hybrid desalination system that
constituted of wind turbine and inclined solar water distillation
integrated with main solar still. A small wind turbine was used to
operate a rotating shaft tted in the main solar still to break boundary
layer of the basin water surface. The system can produce distilled and
hot water. It was estimated that the electricity annual savings was
192.22 RMB/kWh/m2, and the quality of distilled water as well as hot
remaining water was good enough for domestic usage [60].
Moreover, wind and solar thermal can also be combined together
to drive the thermal desalination process. Fernndez-Lpez et al. [61]
analyzed an integrated desalination scheme consisting of two
sequential systems: a MED plant and a MVC system based on
evaporator equipment. The MED stage was driven by thermal solar
collector, whereas the energy consumption of MVC was fuelled by
wind-powered turbines. The nal products were dry salt and fresh
water, with the desalted water production of 100 m3/h and the price
of 0.59 /m3 [61].
4.1.1.3. Wind/multi-renewable energy hybrid system. Regarding to
other renewable sources combined with wind energy, an interesting
designa oating island was proposed by Stuyfzand and Kappelhof
[62], although no such plant was implemented [63]. The plant was an
articial, oating island 10100 km from the shore, 0.060.65 km2 in
size with hexagonal shape, 0.11 km in diameter and 20 m deep. RO
unit was driven by a combination of renewable energy sources
including wind, solar, tidal, wave and hydrothermal gradient. A
storage reservoir aboard was used for stabilization and coping with
uctuations in energy supply and water demand. The plant was
estimated to produce high-quality fresh water of 5500 Mm3/year at a
cost of 0.881.32 /m3.
4.1.2. Exploitation of wind-suited desalination unit
The disconnect relationship between the variable power production
of wind and the need for consistent energy input for most desalination
systems is important to improve the efciency of wind-powered
desalination units. If desalination units could be designed to respond
effectively to variable energy input, i.e., the desalination units would
operate at variable capacities based on the available wind, there is no
need of backup energy storage or integration with other energy sources.
A company based in Germany, ENERCON GmbH, has addressed this
specic limitation of RO technology. As a company focused primarily on
wind energy products, ENERCON's operations have expanded to include
desalination technologies. They have developed an RO technology that
involves a piston system used for energy recovery that also enables
279
Table 1
Basic information about some wind-driven desalination applications.
Plant location
Water type
Desalination unit, capacity
W/T Nominal power
Commissioning year
Ile du Planier, France

Canary Island, Spain
Island of Drence, France
Pjara, Fuerteventura Island,
Spain
Denham, Shark Bay, Australia
Borkum Island, Germany
Rgen Island, Germany
Gran Canaria, Spain
SW/BW
SW
SW
SW
RO, 0.5 m3/h

RO, 200 m3/d
RO,
RO, 56 m3/d
4.0 kW
42 kW
10 kW
225 kW
1982
1984
1990
1991
BW
SW
SW
SW
30 kW
45 kW
360 kW
460 kW
1991
1991
1995
1999
Coconut Island, Hawaii

Canary Island, Spain
CREST, UK
ENERCON, Germany
BW
SW
SW
SW/BW
RO, 130 m3/d

MVC, 48 m3/d
MVC, 360 m3/d
RO, 200 m3/d
MVC, 50 m3/d
ED, 72192 m3/d
RO, 2.7 l/min
RO, 5000 m3/d
RO, 500 l/h
RO, 1751400 m3/d (SW);
3502500 m3/d (BW)
1.2 kW
2.64 MW
2.5 kW
200 kW
1999
2002
2004
2006
Unit water cost;

energy consumption
5.0 kWh/m3
12.5 cent $/kWh

2.1 kWh/m3
RO, 7.5 kWh/m3
MVC, 14.4 kWh/m3
ED, 2.4 kWh/m3
19.865 cent /m3; 2.9 kWh/m3
2.6 $/m3
22.25 kWh/m3
Note: SW: Seawater; BW: Brackish water.
variable levels of energy input. ENERCON plants have no xed operating

point and the water production can range from max. 12.5% to 100% of
the nominal capacity by adjusting the piston speed according to
demand. This has two main advantages: rstly, operation is possible
with a uctuating energy supply, and secondly, output can be adjusted
exibly to water demand without shutting down the plant [38].
4.2. The potential of direct utilization of wind energy
The technologies have been developed to some extent to integrate
wind energy directly or indirectly to the desalination process such as
RO, MVC, ED and solar still. Although RO is the major desalination
process connected to wind power, it is not always the appropriate
one. For the remote areas which are short of fresh water but abundant
of wind energy and seawater, more robust, easily operated wind
powered desalination system should be explored. Considering
environment protection, the system needs to discharge fewer
chemicals and offer high quality fresh water but not the proper
water from RO unit. Thus, in the long term, thermal desalination units
directly powered by wind power are more attractive due to their little
impact on environment, high quality of treated water and energy
saving feature.
So far, MVC is the major thermal desalination process integrated
with wind. However, due to the higher operation temperature
(compared to the seawater temperature), additional heat source is
always necessary. In the single wind powered desalination systems,
the transformation of wind energy to heat must reduce the energy
utilizing efciency. Therefore, if the wind powered desalination unit
can operate at lower temperature, eliminating the transformation
from wind energy to heat, the desalination cost and energy utilizing
efciency might be improved. From this point of view, two potential
developments of wind-desalination system are suggested by authors
as follows:
The humidicationdehumidication process presents several
attractive features which are proper for the wind power, including
modest level of technology employed, simplicity of design,
relatively high efciency compared to other thermal processes,
and, most of all, the ability to combine with low temperature
renewable energy source. Thus, it is of great importance to work
further on problems related to the coupling interface of wind energy
and desalination unit, design and optimization of operational and
structural parameters, and cost evaluation systems as well.
Vacuum distillation might be another alternative as the desalination
unit powered by wind energy. In the process of vacuum distillation,
the phase changes at low temperature, which makes the extra
thermal source unnecessary. In addition, the wind power can drive

the equipments such as vacuum pumps and compressor directly,
avoiding the energy loss caused by energy conversion process. Thus,
this kind of system might increase the usage efciency of wind
energy, and further research should focus on system design,
parameter optimization and feasibility analysis, etc.
5. Conclusions
The use of wind energy for desalination appears nowadays as a
reasonable and technically mature option towards the emerging and
stressing energy and water problems. In spite of intensive research
worldwide, the actual penetration of wind-powered desalination
installations is still low. During the recent past, there has been a rather
intense attempt to develop effective small or medium scale
desalination plants, mainly powered by wind energy. The technologies have been developed to some extent to integrate wind energy
directly or indirectly to the desalination process such as RO, MVC, ED
and solar still. So far, RO is the major desalination process connected
to wind power and MVC is the major thermal desalination process
powered by wind. The basic information of major desalination
prototypes or plants mentioned in this paper is concluded in
Table 1. Through these activities, considerable experience has been
gained.
Similar to other renewable energies, overcoming the intermittent
characteristic and improving the energy utilizing efciency of wind
energy are two important technological problems in the present or
future research. At present, the solutions for the intermittent
characteristic of wind energy are mainly integrating the wind energy
with other kind of energy and designing exible desalination unit to
t the variation of wind. As to the improvement of the energy utilizing
efciency of wind energy, the authors has pointed out two potential
kinds of wind-powered desalination units, possibly utilizing the wind
energy directly with less energy loss.
Acknowledgement
The nancial supports by the National Natural Science Foundation
of China (51009044), Scientic Research Fund of Hainan Provincial
Education Department (Hjkj2011-05) and Start-up Fund Project of
Hainan University (kyqd1106) are gratefully acknowledged.
References
[1] G.M. Asmerom, J.S. Mitchell, Designing cost-effective seawater reverse osmosis
system under optimal energy options, Renewable Energy 33 (2008) 617630.
280
[2] P. Sandeep, M.M. Farid, J.R. Selman, Said AI-hallaj, Solar desalination with a
humidication -dehumidication techniquea comprehensive technical review,
Desalination 160 (2004) 167186.
[3] E. Delyannis, Historic background of desalination and renewable energies, Solar
Energy 75 (2003) 357366.
[4] L. Garca-Rodrguez, Renewable energy applications in desalination: state of the
art, Solar Energy 75 (2003) 381393.
[5] R.G. Raluy, L. Serra, J. Uche, Life cycle assessment of desalination technologies
integrated with renewable energies, Desalination 183 (2005) 8193.
[6] H. Mahmoudi, N. Spahis, M.F. Goosen, S. Sablani, S.A. Abdul-wahab, N. Ghaffour, N.
Drouiche, Assessment of wind energy to power solar brackish water greenhouse
desalination units: a case study from Algeria, Renewable and Sustainable Energy
Reviews 13 (2009) 21492155.
[7] M.S. Miranda, D. Ineld, A wind-powered seawater reverse-osmosis system
without batteries, Desalination 153 (2002) 916.
[8] V.G. Gude, N. Nirmalakhandan, S. Deng, Renewable and sustainable approaches
for desalination, Renewable and Sustainable Energy Reviews 14 (2010)
26412654.
[9] S.A. Kalogirou, Seawater desalination using renewable energy sources, Progress in
Energy and Combustion Science 31 (2005) 242281.
[10] Y. Nakatake, H. Tanaka, A new maritime life saving distiller driven by wind,
Desalination 177 (2005) 3142.
[11] E.A.F.A. Fadigas, J.R. Dias, Desalination of water by reverse osmosis using gravitational
potential energy and wind energy, Desalination 237 (2009) 140146.
[12] C.C.K. Lui, J.W. Park, R. Migita, G. Qin, Experiments of a prototype wind-driven
reverse osmosis desalination system with feedback control, Desalination 150
(2002) 277287.
[13] T. Witte, S. Siegfriedsen, M. El-Allawy, WindDeSalter Technology Direct use of
wind energy for seawater desalination by vapour compression or reverse osmosis,
Desalination 156 (2003) 275279.
[14] C. Fritzmann, J. Lwenberg, T. Wintgens, T. Melin, State-of-the-art-of reverse
osmosis desalination, Desalination 216 (2007) 176.
[15] D. Colombo, M. de Gerloni, M. Reali, An energy-efcient submarine desalination
plant, Desalination 122 (1999) 171176.
[16] L. Garca-Rodrguez, V. Romero-Ternero, C. Gmez-Camacho, Economic analysis
of wind- powered desalination, Desalination 137 (2001) 259265.
[17] I. de la Nuez Pestana, F.J.G. Latorre, C.A. Espinoza, A.G. Gotor, Optimization of RO
desalination systems powered by renewable energies, Part I: Wind energy,
Desalination 160 (2004) 293299.
[18] M. Forstmeier, F. Mannerheim, F. D'Amato, M. Shah, Y. Liu, M. Baldea, A. Stella,
Feasibility study on wind-powered desalination, Desalination 203 (2007)
463470.
[19] E. Tzen, Overview of the desalination technologies powered by renewable
energies (RE), Proceedings of International Seminar Desalination Systems
Powered by Renewable Energy, Amman, Jordan, 2006, Information on:, http://
www.adu-res.org.
[20] P. Feron, The use of wind power in autonomous reverse osmosis seawater
desalination, Wind Engineering 9 (1985) 180199.
[21] S.M. Habali, I.A. Saleh, Design of stand-alone brackish water desalination wind
energy system for Jordan, Solar Energy 52 (1994) 525532.
[22] C.T. Kiranoudis, N.G. Voros, Z.B. Maroulis, Wind energy exploitation for reverse
osmosis desalination plants, Desalination 109 (1997) 195209.
[23] D. Voivontas, K. Yannopoulos, K. Rados, A. Zervos, D. Assimacopoulos, Market
potential of renewable energy powered desalination systems in Greece,
Desalination 121 (1999) 159172.
[24] D. Voivontas, K. Misirlis, E. Manoli, G. Arampatzis, D. Assimacopoulos, A. Zervos, A
tool for the design of desalination plants powered by renewable energies,
Desalination 133 (2001) 175198.
[25] V. Romero-Ternero, L. Garca-Rodrguez, C. Gmez-Camacho, Thermoeconomic
analysis of wind powered seawater reverse osmosis desalination in the Canary
Islands, Desalination 186 (2005) 291298.
[26] P.A. Koklas, S.A. Papathanassiou, Component sizing for an autonomous winddriven desalination plant, Renewable Energy 31 (2006) 21222139.
[27] C. Koroneos, A. Dompros, G. Roumbas, Renewable energy driven desalination
systems modeling, Journal of Cleaner Production 15 (2007) 449464.
[28] I.D. Spyrou, J.S. Anagnostopoulos, Design study of a stand-alone desalination system
powered by renewable energy sources and a pumped storage unit, Desalination 257
(2010) 137149.
[29] E. Koutroulis, D. Kolokotsa, Design optimization of desalination systems powersupplied by PV and W/G energy sources, Desalination 258 (2010) 171181.
[30] K. Bourouni, T. Ben M'Barek, A. Al Taee, Design and optimization of desalination
reverse osmosis plants driven by renewable energies using genetic algorithm,
Renewable Energy 36 (2011) 936950.
[31] A. Maurel, Desalination by RO using RE (solar & wind): Cadarache Center
Experience, Proceedings of the New Technologies for the Use of RE Sources in
Water Desalination, Greece, 1991, pp. 1726.
[32] A. Peral, A.G. Contreras, T. Navarro, IDM-Project: Results of one year's operation,
Proceedings of the New Technologies for the Use of RE Sources in Water Desalination,
Greece, 1991, pp. 5680.
[33] C. Graber, Desalination in Spain, Technology ReviewInformation on, http://

www.technologyreview.com.
[34] J.M. Veza, A. Gmez-Gotor, Status of desalination in Canary Islands. Energy related
aspects, Seminar on New Technologies for the Use of Renewable Energies in
Water Desalination, Athens, Commission of the European communities, DG
XVIIfor Energy, CRES (Centre for Renewable Energy Sources), 1991.
[35] L. Garca-Rodrguez, Seawater desalination driven by renewable energies: a
review, Desalination 143 (2002) 103113.
[36] S. Rybar, M. Vodnar, F.L. Vartolomei, R.L. Mndez, J.B.L. Ruano, Experience with
renewable energy source and SWRO desalination in Gran Canaria. SP05-100,
International Desalination Association World Congress, 2005, Information on,
http://www.membranes.com.
[37] CORDIS Database, 2005.
[38] K. Paulsen, F. Hensel, Design of an autarkic water and energy supply driven by
renewable energy using commercially available components, Desalination 203
(2007) 455462.
[39] R. Vujcic, M. Krneta, Wind-driven seawater desalination plant for agricultural
development on the islands of the County of Split and Dalmatia, Renewable
Energy 19 (2000) 173183.
[40] European Commission, Desalination guide Using Renewable Energies, 1998.
[41] M. Stah, Small wind powered RO seawater desalination plant design, erection and
operation experience, Seminar on New Technologies for the Use of Renewable
Energies in Water Desalination, Commission of the European Communities, DG
XVIIfor Energy, Centre for Renewable Energy Sources, Athens, 1991.
[42] S. Brown, Deham wind-turbine-Project completion report, Water Authority of
Western Australia, Internal Report, Perth, 1991.
[43] C.R. Henderson, J.G. McGowan, J.F. Manwell, A feasibility study for a wind/diesel
hybrid system with desalination for star island, NH, Global WINDPOWER
Conference and Exhibition, 2004.
[44] U. Plantikow, Wind-powered MVC seawater desalination-operational results,
Desalination 122 (1999) 291299.
[45] A.J. Carta, J. Gonzalez, V. Subiela, The SDAWES project: an ambitious R&D
prototype for wind-powered desalination, Desalination 161 (2004) 3348.
[46] A. Karameldin, A. Lotfy, S. Mekhemar, The Red Sea area wind-driven mechanical
vapour compression desalination systems, Desalination 153 (2003) 4753.
[47] J.M. Veza, B. Penate, F. Castellano, Electrodialysis desalination designed for wind
energy (on-grid test), Desalination 141 (2001) 5361.
[48] J.M. Veza, B. Penate, F. Castellano, Electrodialysis desalination designed for offgrid wind energy, Desalination 160 (2004) 211221.
[49] C.R. Henderson, J.F. Manwell, J.G. McGowan, A wind/diesel hybrid system with
desalination for Star Island, NH: feasibility study results, Desalination 237 (2009)
318329.
[50] S. Kumar, G.N. Tiwari, Performance evaluation of an active solar distillation
system, Energy 21 (1996) 805808.
[51] M. Thomson, D. Ineld, Laboratory demonstration of a photovoltaic-powered seawater
reverse osmosis system without batteries, Desalination 183 (2005) 105111.
[52] G. Petersen, S. Fries, J. Mohn, A. Mller, Wind and solar-powered reverse osmosis
desalination unitsdescription of two demonstration projects, Desalination 31
(1979) 501509.
[53] G. Petersen, S. Fries, J. Mohn, A. Mller, Wind and solar-powered reverse osmosis
desalination unitsdesign, start up, operating experience, Desalination 39 (1979)
125135.
[54] D. Weiner, D. Fisher, E.J. Moses, B. Katz, G. Meron, Operation experience of a solarand wind powered desalination demonstration plant, Desalination 137 (2001)
713.
[55] S.A. Kershman, J. Rheinlnder, H. Gabler, Seawater reverse osmosis powered from
renewable energy sourceshybrid wind/photovoltaic/grid power supply for
small scale desalination in Libya, Desalination 153 (2002) 1723.
[56] S.A. Kershman, J. Rheinlnder, T. Neumann, O. Goebel, Hybrid wind/PV and
conventional power for desalination in Libya-GECOL's facility for medium and
small scale research at Ras Ejder, Desalination 183 (2005) 112.
[57] E.S. Mohamed, G. Papadakis, Design, simulation and economic analysis of a standalone reverse osmosis desalination unit powered by wind turbines and photovoltaics,
Desalination 164 (2004) 8797.
[58] E.S. Mohamed, G. Papadakis, E. Mathioulakis, V. Belessiotis, An experimental
comparative study of the technical and economic performance of a small reverse
osmosis desalination system equipped with a hydraulic energy system,
Desalination 194 (2006) 239250.
[59] A.M. Gilau, M.J. Small, Designing cost-effective seawater reverse osmosis system
under optimal energy options, Renewable Energy 33 (2008) 617630.
[60] E.A. Mohamed, Z. Zhao, Wind turbine-inclined still collector integration with solar
still for brackish water desalination, Desalination 249 (2009) 490497.
[61] A. Viedma Fernndez-Lpez, R. Herrero, A.S. Kaiser, Seawater integrated
desalination plant without brine discharge and powered by renewable energy
systems, Desalination 235 (2009) 179198.
[62] P.J. Stuyfzand, J.W.N.M. Kappelhof, Floating, high-capacity desalting islands on
renewable multi-energy, Desalination 177 (2005) 259266.
[63] C. Charcosset, A review of membrane processes and renewable energies for
desalination, Desalination 245 (2009) 214231.
DESALINATION
ELSEVIER
Desalination 153 (2002) !%I6
A wind-powered seawater reverse-osmosis system without

batteries
Marcos S. Miranda*, David Infield
Centerfor Renewable Energy Systems Technology CREST, Loughborough University LEll3TW, UK
Tel. +44 (I J09) 22il44; Fax +44 (ISO@ 610031; email: MS.&iranda@lboro.ac.uk
Received 20 April 2002; accepted 30 April 2002
Abstract
The development of small-scale stand-alone desalination systems is important to communities on islands and in
isolated inland areas, In such places, electricity supplies are often expensive and unreliable, while the wind resource
is abundant. The system presented here comprises a 2.2 kW wind turbine generator powering a variable-flow Reverse
osmosis (RO) desalination unit. It is highly efficient, rugged, built with off-the-shelf components and suitable for
use in remote areas. Operation at variable-flow allows the uncertainty and variability of the wind to be accommodated
without need of energy storage. Batteries, which are common in stand-alone systems, are avoided and water production
is dependent on the instantaneous wind speed. A model-based control strategy is used to independently maximize
both the energy extracted from the wind and the water output of the RO unit. A computer model of the system has
been developed based on component models, identified through laboratory testing. Performance predictions are
presented and discussed.
Keywork: Wind power; Reverse osmosis; Energy recovery; Seawater desalination; Renewable energy
1. Introduction
There is no need to dwell on the importance
of energy in the daily life of modern society and
that its availability relies mostly on the existence
of fossil fuels. Increasing awareness ofthe depletion

of current sources has led to a global effort in the
research and development of renewable energy
technologies, such as wind, solar, tidal and geothermal energy.
*Corresponding author.
Presented at the EuroMed 2002 conference on Desalination Strategies in South Mediterranean Countries:
Cooperation between Mediterranean Countries of Europe and the Southern Rim of the Mediterranean.
Sponsored by the European Desalination Society and Alexandria University Desalination Studies and Technology
Center, Sharm El Sheikh, Egypt, May 4-6, 2002.
00 l l-9164/02/$- See front matter CQ2002 Elsevier Science B.V. All rights reserved
PII: SOOIl-9164(02)01088-3
10
M.S. Miranda, D. Infield/Desalination
This motivation for using renewable energy

is even greater if stand-alone desalination applications are considered. This is because the energy
required for the process is particularly expensive
in the remote areas where fresh water is required.
Renewable energy sources can provide a reliable
energy supply alternative for water desalination.
Initial cost and resource availability are the most
significant limitations.
In the context of the utilization of the more
established renewable energy sources: the sun
(thermal and PV), and the wind, stand-alone
desalination systems have been widely discussed
[l-5]. Since desalination is an especially promising application that involves broad fields of
study, many different solutions have been proposed.
Even if one focuses on one particular renewable
source and a specific desalination method, there
may still be many options available in terms of
the final system configuration.
One of the critical limiting factors to the wider
implementation of renewable energy driven desalination systems is the intermittence and unpredictability of the renewable source. Two distinct
problems have been identified in an earlier study
[5]. The first is that most desalination technologies
are not suitable for operation at variable power,
and the second is that lack of continuity of energy
supply, or even limited power availability over
variable periods of time may cause the demand
not to be met. A common solution to these problems
is the use of energy storage, which can accumulate
energy surpluses (long-term storage) and/or smooth
out shorter-term variations in the supply (shortterm storage). On the negative side, the use of such
devices results in increases to both capital and
operating costs.
A summary of the technically possible combinations of renewable energy sources and desalination techniques is given in Table 1. Wind energy
is a most attractive source in the short-term since
the technology is well developed and relatively
cheap. It will be the subject of the work presented
here.
I53 (2002) 9-16
Table 1
Applicability
of renewable
desalination techniques
energy sources to water
Wind
Solar
Tidal
Geothermal
RO
vc
FS
ED
ME
MSF
ED
RO
SD
RO
VC
FS
ED
MSF
ME
vc
RO
ED
Source Hanafi [ 11.
2. Wind power systems

The use of wind energy for electricity generation
can be divided into two main application areas.
The first and foremost of these is the commercial
generation
of bulk electricity
through gridconnected systems.
The second category is the electricity generation
within stand-alone systems. In contrast to grid
connected systems, these are built to be used in
sites where maintenance may be sporadic and
technical assistance unavailable, thus greater
robustness is required in their design, incurring
higher capital costs than otherwise. Grid connected
turbines are used as an additional supply, complementing the conventional base load power
system. In stand-alone systems, the wind can often
be the sole source of energy and this should be
fully taken into consideration during the design
stage.
Due to the inherently random characteristic of
the wind, certain key aspects must be attended to
in the design process of stand-alone systems.
Besides the wind resource potential, the nature
of the electricity load needs to be given careful
consideration, in particular whether disconnection
is acceptable, and if so, for how long.
This consideration relates to the presence and
sizing of any storage system that may be included
in the system. Energy storage plays an essential
role in determining system performance, as well
M.S. Miranda, D. Infield / Desalination 153 (2002) 9-16
11
as the aforementioned influence on capital and

maintenance costs.
For the design of a small stand-alone system,
a key challenge is to find a good compromise
between reliability and system complexity that
meets the economical constraints. This is not a
simple matter, and it will mostly depend on the
type of load and the local resources.
I
(2 (limin)
3. Reverse osmosis systems

Reverse osmosis (RO) is now a well-established
technology for the desalination of water and in
particular seawater. Nevertheless, the use of RO
in small stand-alone systems (in the range of a few
m3/d) is still an area of developing technology.
Islands and isolated inland areas, where the
electricity supply may be a problem, form an ideal
application for such stand-alone systems. In these
remote places, electricity is often supplied by a
weak grid or even generated locally, by means of
diesel generators.
Despite the many advances in RO membrane
technology over recent years, such as higher
rejection rates and improved tolerance to pressure
fluctuations, considerable attention still needs to
be given to operational conditions. It is believed that
higher efficiencies can be obtained when operating
at certain flow and pressure conditions.
An important contribution dealing with variable
flow and pressure operation of RO membranes,
which looks at the feasibility of wind-powered
systems, is that of Feron [2].
In this work, an operational window is established (Fig. 1)determiningthe operational parameter
variation to which a membrane can be safely submitted. The four limits that define this window are:
Maximum feed pressure - determined by the
membrane mechanical resistance;
Maximum brine flow rate - should not be
exceeded to avoid membrane deterioration;
. Minimum brine flow rate - it should be observed to avoid precipitation and consequent
membrane fouling;
l
Fig. 1. RO membrane operational window.
Maximum product concentration - if the

applied pressure is less than a determined value,
permeate concentration will be too high.
4. Wind-R0 configuration possibilities

A classification of the different wind powered
reverse osmosis systems found in the literature has
been made.
This was based on some of the points previously
discussed: the existence of an alternative electrical
supply (weak grid or diesel generator); the matching
of the available wind energy to the load; and the
operational characteristic of RO membranes.
4.1. Systems with back up (diesel/grid)
In these systems, an additional energy source
is provided (a diesel-powered generator or even
the local grid) so that the power supplied to the
RO is constant. The back-up generation complements the power generated from the wind
turbine to match the RO unit power consumption.
The main benefit of these systems, as in any
hybrid wind-diesel configuration is the achievement of fuel savings, which may increase the
generator availability and reduce overall energy
costs.
On the other hand, problems such as fuel
shortages, diesel generator maintenance, interruptions or power cuts in the supply, may lead to
MS. Miranda, D. Infield/Desalination 153 (2002) 9-16
12
of the RO system since it cannot

be powered by the wind turbine alone.
unavailability
4.2. Systems without back up

Systems without an external energy source can
be divided into two categories, with emphasis on
the RO unit operation: systems which run under
approximately constant operating conditions; and
those that experience variable operational conditions.
4.2.1. Near constant operating conditions
This first type of operation can be implemented
by three different means: on/off switching of the
RO units; usage of storage devices; and de-rating
the wind turbine. In all three cases, an attempt is
made to supply the individual RO modules with
approximately constant power.
4.2.1. I. Storage devices
In this strategy, storage devices are employed
to accumulate energy surplus during periods when
the power generated by the wind turbine is greater
than the load demand from the desalination unit.
This surplus would then be used later when the
generated power is insufficient to tneet the load
demand.
One common way of storing the surplus energy
is by using batteries [3]. In this case, the relation
between operational pressure, storage sizing and
average wind speed should be considered in the
design stage. In addition, capital and maintenance
costs should be carefully assessed. A disadvantage
of this approach to the system design is the rating
of the energy storage system, since this can make
it economically unattractive at higher power levels
due to the sizing of the battery bank.
demand to the total power generated instantaneously by the turbine. There is no limitation
concerning the system Rower rating, and this approach is feasible up to power levels of hundreds
of kilowatts.
Although, frequent cycling of RO units is not
usually recommended, this problem can be overcome by implementing different types of configuration. Rahal [4] uses a higher power wind
turbine operating at near constant speed Cgenemtion
management) connected to many equally smaller
RO units switching on/off (load management). To
smooth out the fluctuations, short-term energy
storage (a flywheel in this instant) is used. Varying
the pitch angle of the wind turbine blades controls
the generated power.
Another possibiiity [5], suitable for smal1er
systems (with medium/low power turbines of less
than 50 kW rated), is the switching of few (two/
three) desalination units with distinct power levels.
Additionally, some auxiliary loads (such as pumping/heating and dump loads) can be implemented
to absorb any power surplus, keeping the system
voltage and frequency constant.
4.2.1.3. Wind turbine de-rating
This approach consists of making use of the
flat end of a pitch controlled wind turbine power
curve to operate the RO unit at approximately
constant power [6}. An implication of this configuration is that, since the turbine rated power is
only achieved at high wind speeds, it would have
to be de-rated by changing the settings of the pitching
mechanism. This will cause the generated power
to be flattened at lower wind speeds and consequently to have lower values. Therefore, the original
rating of the turbine rotor should be considerably
higher than the RO unit rated power making the
system more expensive.
4.2.1.2. RO unit switching

This strategy is based on the use of a higher
power wind turbine connected multiple smaller
RO units. The power control is achieved by
switching the units on and off so as to match the
4.2.2. Variable operating conditions

In contrast to systems that operate under constant
conditions, another operational strategy is based
on the establishment and imposition of certain
MS. Miranda, D. Infield/Desalination 153 (2002) 9-16
13
operational limits [63, as previously illustrated in

Fig. 1. This means that, based on the input power
to the RO unit (flow times pressure), a control
strategy is determined which .imposes a fixed
operating point on the system that lies within the
allowed region (i.e. the operational window of
the RO unit shown in Fig. 1).
By doing this, an attempt is made to operate
the system autonomously over a wider power
range, without the need to use a back up unit or
storage devices. The overall effect is to reduce
capital and operating costs. One aspect that should
be emphasized is that very little is known about
the consequences of variable operation of RO
membranes. It is recognized that mechanical fatigue
can occur and that lifetime of the RO elements
may be shortened and performance impaired.
less system operating from a variable energy source

such as the wind.
Extensive laboratory testing was carried out
so as to develop suitable mathematical models for
individual components over a wide operating range.
Based on these individual models, a complete
system model was built to understand and assess
system behavior as well as predict its performance.
Based on these analyses, optimal performance
trajectories were determined. These are the operating conditions that maximize, at the same time,
wind power conversion for a given wind speed
and desalinated water production with the available power. The control strategy makes the system
follow these trajectories by controlling the speed
of the positive displacement pumps.
5. Proposed wind-R0
6. Performance predictions
system
The proposed system comprises a small (2.2 kW)

wind turbine directly supplying a reverse osmosis
desalination unit. The pumping system is composed
of two variable speed drives (inverter/induction
motor) driving a medium and a high-pressure pump,
each. The speed of each pump is individually
controlled to maximize both wind power capture
and product water flow. A Clark pump is used to
recover the energy from the brine-stream, making
the system highly efficient over a wide operating
range.
A detailed description of the system and its
components is presented by CREST, Loughborough University [6].
5. I. Control strategy
The use of two positive displacement pumps
indirectly enables the control of feed pressure and
flow independently. This characteristic makes it
possible to operate the system at any point within
the operational window shown in Fig. 1, provided
a suitable control strategy is employed. This is
critical in maximizing the efftciency of a battery-
The following performance predictions are

taken from a detailed Matlab-Simulink model of
the entire system. This model is based on extensive
in-house component characterization, supplemented
by manufactures data. The salinity of the seawater
feed is taken to reflect that of the Red Sea at
40,000 ppm, which is isosmotic with 34,300 ppm
of pure NaCl. Additionally, the feed temperature
is constant and equal to 25C.
6.1. Long-term (steady-state) performance

The first analysis carried out aimed at establishing the input-output characteristic ofthe system,
i.e. the relationship between wind-speed and fresh
water production, as depicted in Fig. 2.
The importance of this characteristic relies on
the fact that it can be used in a system siting study.
Similarly to a wind turbine power curve, it would
give the expected output from the system, given
the resource in a certain location. A statistical
analysis can be conducted by applying the wind
speed probability (Weibull distribution) for this
location against the characteristic of Fig. 2. The
14
6.2. Short-term (dynamic) performance
43.
2.
1
4
10
12
14
16
14
16
0-W
Fig. 2. Product water flow vs. wind speed.
n3
r.
3
10
12
PW
Fig. 3. System specific energy consumption.
result is the water production probability expected

for this location.
Fig. 3 shows the system specific energy as a
function of the wind speed. It presents an overall
value below 3.5 kWh/m3, being slightly higher at
lower wind speeds. Nevertheless,
it is almost
constant throughout the operating range. This
translates in the fact that water production water
is directly related to the available power.
A dynamic model of the system was developed

to verify its performance during normal transient
operation, under turbulent wind. This is particularly important in the determination ofthe control
strategy and fine-tuning of the controller parameters.
The following figures show the behavior of
the system for a particularly high wind speed data
series. The time-series used is 5 min long, with
an average wind speed of 8.3 m/s.
Fig. 4a shows the wind speed time series used
as an input to the system. The desalinated product
water flow is plotted in Fig. 4b. Its average value
is 8.5 m3/d. The dependence between both curves
is obvious. Contrary to the product water flow,
which is proportional to the net driving pressure,
the salt passage through the membranes is mostly
a function of the difference in concentration
between the feed and the product streams (approximately constant). Therefore, the product water
concentration (Fig. 4c) will be inversely proportional to the flow. The concentration shown is
resultant of the mixing of the different product
concentrations of the four modules used with an
average value of almost 300 ppm.
Finally, Fig. 4d presents the system instantaneous specific energy. This figure is obtained by
dividing the instantaneous wind turbine output
power (in kW) by the product water flow (m3/h).
It should be pointed out that desalinated water
has a higher energetic content than seawater and
therefore represents an energy storage medium
itself.
Such low figures for the specific energy emphasize the importance of assessing individual
component efficiency, in an attempt to minimize
system overall losses.
The average value of about 3.4 kWh/m3 for
the system specific energy seems remarkably low
when compared with the literature standard of 1O15 kWh/m3 achieved in standard (no energy recovery) systems. Additionally, it should be mentioned
15
A high-efficiency configuration was proposed,

having a Clark pump as the concentrate stream
hydraulic energy recovery device. One of the main
objectives of this analysis was to verify the integration of the several components -which were
individually tested and modeled - and their performance over a wide operational range. More
important still, was to demonstrate the proposed
configuration as a promising alternative for the
desalination of seawater.
Finally, a more representative validation of the
system will be only possible after prototype testing,
which will follow in the next phase of the project.
Prototype testing will provide a deeper understanding of the system performance, particularly
considering the transients present in a system supplied by an extremely variable and unpredictable
source, such as the wind.
Acknowledgement
This work was carried out in conjunction with

Dulas Ltd, Machynlleth, UK and was supported
by ETSU, DTI, UK and CNPq, Department of
Science and Technology, Brazil.
0
50
100
160
200
250
300
(9
Fig. 4. Wind speed, product flow, product concentration

and specific energy.
References
111 A. Hanafi, Desalination
that energy recovery - using turbines, reverse

running pumps or other devices - is a relatively
common practice in bigger systems, with rated
outputs of at least 100 times higher. This is also
due to the improved efficiencies associated to the
increased sizing of components. Nevertheless,
such efficiencies would not be easily achieved in
smaller components such as the ones needed to
implement a system of the same size of the one
proposed in this work.
[21
7. Conclusions
151
This work presented the analysis of a small

wind-powered reverse osmosis desalination system.
161
I.31
[41
using renewable energy

sources, Desalination, 97 (1994) 339-352.
S. Alawaji, M.S. Smiai, S. Rafique and B. Stafford,
PV-powered water pumping and desalination plant
for remote areas in Saudi Arabia, Applied Energy,
52(2-3) (1995) 283-289.
H. Ehmann, A. Wobben and M. Cendagorta, PRODESAL - development and pilot operation of the
first wind powered reverse osmosis seawater desalination plant, Mediterranean Conference on RES for
Water Production, Santori, Greece, 1O-l 2 June 1996,
pp. 84-87.
C.T. Kiranoudis, N.G Voros and Z.B. Maroulis, Wind
energy exploitation for reverse osmosis desalination
plants, Desalination, 109 (1997) 195-209.
CRES, Greece, Desalination Guide Using Renewable
Energies, THERMIE - DG XVII, European Commission Report, 1998.
P. Feron, Use of windpower in autonomous reverse
16
[7]
[8]
[9]
MS. Miranda, D. Infield / Desalination 153 (2002) 9- I6

osmosis seawater desalination, Wind Engineering,
9(3) (1985) 180-199.
D.G Infield, Performance analysis of a small wind
powered reverse osmosis plant, Solar Energy, 61(6)
(1997)415-421.
Z. Rahal and D.G Infield, Wind powered stand alone
desalination, EWEC97, Oct. 1997, pp. 802-806.
AS. Neris, GB. Giannakopoulos and N.A. Vovos,
Autonomous wind turbine supplying a reverse

osmosis desalination unit, Wind Engineering, 19(6)
( 1995) 325-346.
[lo] M. Thomson, M.S. Miranda and D.G Meld, A smallscale seawater reverse-osmosis system with excellent
energy efficiency over a wide operating range,
Euromed, 2002, Desalination Strategies in South
Mediterranean Countries.
DESALINATION
ELSEVIER
Desalination 171 (2004) 257-265

www.elsevier.com/locate/desal
Preliminary experimental study of a small reverse osmosis

wind-powered desalination plant
Fabrizio Moreno, Alvaro Pinilla*
Department of Mechanical Engineering, Universidad de los Andes, Cra 1 4, No. 184, 10 Bogota, Colombia
Tel. +57 (1) 332-4322; Fax: +57 (1) 332-4323; emaih apinilla@uniandes.edu.co
Received 15 April 2004; accepted 4 June 2004
Abstract
The paper describes the work carried out in the development of a small wind-powered desalination plant. An
alternative control system was studied to serve as a direct interphase between a reverse osmosis desalination plant and
a small wind energy conversion system. The main purpose was to reduce or eliminate the need for an energy storage
system (usually, a battery bank). In order to achieve this objective, an experimental prototype of a desalination plant
and a wind generator simulator were developed. The systems were evaluated under laboratory-controlled conditions
and subjected to field trials. The experimental plant desalinates highly saline seawater (35,000 mg/L) at a rate of
approximately 0.4 m3/d. This amount of potable water is sufficient to supply the basic water demands in a small
community in an isolated location. The paper also describes the identification of technical problems associated with
operating a desalination plant with an intermittent source of energy (wind).
Keywords: Wind energy; Simulation; Reverse osmosis; Seawater desalination
1. Introduction
Reverse osmosis (RO) is a process used to
desalinate salt water. The process has the advantage that it requires low energy consumption
compared to other desalination processes. Feron
and Smulders o f Eindhoven Technical University
(Netherlands) found in the 1980s that RO has the
lowest energy consumption amongst most
methods o f desalination [ 1].
*Corresponding author.
In general, desalination processes are required

in coastal or isolated locations where there are no
water sources other than water from the sea, or
the water source are in deep wells that produce
salty water. A convenient alternative is the supply
o f potable water through independent and selfsufficient solutions such as that offered by thJis
combination o f technologies (a RO plant and the
wind energy conversion system).
Depending upon the concentration o f salts
dissolved in water, the pressure required for the
0011-9164/04/$- See front matter 2004 Elsevier B.V. All rights reserved
doi: 10.1016/j.desal.2004.06.191
258
F. Moreno, A. Pinilla /Desalination 171 (2004) 257-265
RO process varies from 1.4 to 8.3 MPa (2001,200 psi). There is a practical relationship that
allows the calculation of the minimum required
osmotic pressure: at least 7 kPa (1 psi) is required
for every 100 mg/L of salts dissolved in the
water; therefore, if the seawater has 35,000 mg/L
of salts, at least 2.45 MPa (350 psi) are needed
for the desalination process to take place [2].
Desalination plants usually run on fossil fuel
(diesel or motor gasoline) generators or they are
connected to the local electricity network. Sometimes energy is generated by renewable energies
(solar or wind) which use energy storage systems
(battery banks). The battery bank is expensive in
RO plants for this type of application. Consequently, if a control system can be devised to
allow the direct operation of the plant, initial
costs will fall substantially, and the system will
be simpler and more feasible to use.
The RO desalination plant is composed of five
systems (Fig. 1): (1) a pre-treatment system,
whose function is to reduce substances harmful to
the RO membrane; (2) the desalination system,
composed of a high-pressure pump and the RO
membrane; (3) the power supply system, whose
function is to generate the power required, (4) a
control system, i.e., the interphase between the
power supply and the desalination system; and
(5) the post-treatment system to make the water
of optimum quality for human consumption.
2. Pretreatment systems
The function of a pre-treatment system is to
eliminate agents that will block the membrane.
Most of these pretreatment systems are complex,
however, and what is needed is a simple and
effective form ofpre-treatment which requires no
electricity for operation (a passive system).
A granular medium slow pre-filter was chosen
[3] due to its simplicity in construction, operation
and low associated costs. The first two stages of
the pre-filter are anthracite (amorphous fragile
coal) beds and river sand. They allow suspended

colloid particles to filter out by deposition on the
grains of the filter medium. The final stage of the
pre-filter uses granulated activated coal, which is
used to free the filtered water of oxidizer agents
and bacteria that might form a biological block.
The results of tests on this type of pre-filter
show that it is effective for removing agents
harmful to the RO membrane. The design parameters of the pre-filter are determined by the type
of medium, size and depth of the filtration beds,
the surface area of the filters, the static pressure
head available to act as the driving force and the
method of operation of the pre-filter including
cleaning.
The performance of this type of pre-filter is
determined by the head loss through the filter and
the resulting quality of the water. The pre-filter
should be changed each time the maximum
known head loss is reached or each time the
quality of the pre-filtered water indicates that this
should be done. The change ofpre-filter is easy,
and it demands no more than removing the
blocked-up filtration media and adding new filtration beds [4].
3. Pump-RO membrane system

This system is composed of a commercial RO
membrane and a high-pressure pump. The membrane was chosen, beating in mind, two main
constraints: it must be designed to desalinate seawater and it must have the lowest possible feed
flow rate. The membrane selected was the SW302521 (Dow Chemicals FILMTEC TM Division),
with a maximum flow rate of 0.37 L/s and a maximum operating pressure of 6.9 MPa (1,000 psi).
The pump used was a Hydra-Cell M-03-E
(Warner Engineering) positive displacement piston pump; its main features are a maximum flow
of 0.14 L/s and 8.2 MPa maximum operating
pressure. The pump is made of stainless steel
(essential for operation in salt water).
F. Moreno, A. Pinilla / Desalination 171 (2004) 257-265
Saltwater
Purewater
.................................................. 1. Pre-Treatment
System
259
.........
Electrical Energy.~.~.-..--m,m
nk
Prefilter
=n=l~
Desalination
Tank
!.................
2. Desafination System
Pu'rn-.......................
.......................................
~.,i,.~'+~._,',_17~
e : ~ l : ~
,,
3. Energy
supply system
r a
"--
-..~i ~J_. ~
= " pe~;::e
I-
I. . . . . . . . . . . . . . . . .
Concentrate
-t
-~"
Post-treatmen
I
Pe~mn~ate i
L .................
I
4. Control System
Windgenerator
Fig. 1. Simplified scheme of a RO desalination plant and a wind energy supply system.
The variables that affect the performance of

the membrane are:
feeder concentration (Ca) - - This is assumed
to be constant since the plant is designed for
desalination of seawater with concentration of
35,000 mg/L;
operating pressure - - this affects the quality
of the water varying the permeate concentration (Cp) and permeate flow rate (Qp)];
feeder flow rate, Qa (affects Qp and Cp).
The reverse osmosis system application (ROSA)
simulation program, supplied by the manufacturer, was used to determine the magnitude of
the effect of these operating factors on the
variables of membrane output and predicted
performance.
The results of the ROSA simulation determined the operating limits of the membrane
(Fig. 2). The left limit for the minimum concen-
trate flow rate was suggested by the manufacturer

(0.066 L/s). The right limit for the maximum
feeder flow rate was 0.37 L/s. These two limits
are caused by ionization and polarization of the
RO membrane. The upper limit depends on the
maximum concentration of salt permitted in water
for human consumption (1,000 my/L), as established by the World Health Organization. Based
on these results, two operating parameters for the
desalination plant were established: an operating
pressure above 3.4 MPa in order to obtain a water
quality with less than 1,000 mg/L and a minimum
feeder flowrate of 0.066 L/s for the membrane to
perform correctly.
4. Strategies for electricity supply

Two alternative sources of supply were
evaluated for the RO membrane-pump combina.-
260
Left L i m i t
Rght limit
Optim u m O p e r a t i n g Z o n e
.
14oo
t200
. Upper limit
__
= 1000 . . . .X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
0
800
8e,
8
3,4 MPa (500psi)
60O
,= ~ ,
_.........
4,1 IvPa ~600psi/
400
200
0
0.00
5,5 t,Pa (soopsi)
1.
T
0.05 ~
'
0.10
0.15
..........
0.20
0.25
0.30
0.35
~0,40
0.45
0,50
Feed Flowrate [ L / s ]
Fig. 2. Determination of desalination limits obtained by reverse osmosis system application (ROSA) simulation.
tion. One is by connecting the system to an

external AC supply allowing variations in voltage
and frequency for the excitation of a conventional
electric motors. The other was to operate the
system from a DC source, using an AC-DC
rectifier.
4.1. Alternating current electricity supply

This alternative was evaluated using a theoretical model, which was subsequently supported
by laboratory tests. Some results are provided to
show that it is a viable alternative.
A study was made of the performance of a
permanent magnet synchronous generator by
theoretical analysis [5] and experimentation. The
analysis defined the behaviour of the electrical
characteristics of the generator such as voltage,
frequency and power in relation to changes in the
rotational speed of the electricity generator.
Tests were made on three-phase motors under
variable electrical conditions. Three 3-phase

220 V AC/60Hz motors of different power
(0.67 kW, 0.9 kW and 1.1 kW) were tested. They
were connected directly to the output of a small
commercial three-phase 900 W wind generator in
the wind simulator bench. The tests for the
motors connected directly to the electrical generator suggested that the motor could work under
variable conditions but would not supply the
torque required by the load due to the demand for
voltage. Two alternatives were proposed to solve
this problem: a three-phase transformer could be
used at the output of the wind generator or the
wind generator should generate at a high voltage.
It is worth noting that the performance of the
generator depends on the type of electrical load to
which it is connected. Since the motors used were
squirrel's cage induction motors, the load was
inductive, and this produces a lag of voltage in
relation to current. This means a fall in the
voltage supplied by the generator [6].
261I
The results obtained by using transformers

allowed the conclusion that transformers can be
used to increase generator output voltage, and
thus to increase the torque capacity that the motor
can deliver. The characteristics of the wind generator used, however (a wind generator used to
charge batteries, low voltage, high current), do
not provide an acceptable solution. Also, since
the load is inductive, the global power factor of
the plant is further reduced.
Nonetheless, the use o f high-voltage wind
generators shows that they can operate a desalination plant simply and efficiently since they
eliminate the need for additional elements that
consume power such as rectifiers, transformers
and battery banks.
4.3. Modelling and theoretical simulation o f the

AC alternative
,4.2. Direct current electricity supply
Time, s
The evaluation of this scenario was made with

a theoretical model of a DC motor. A permanent
magnet DC motor was selected and it copes with
the power requirements o f the system that met the
electrical restrictions. A forced commutation
rectifier was used, given the advantages that this
type of device has with small generators [5].
The theoretical analysis of the motor was
supported by the operating data supplied by the
manufacturer. The model shows a reduction in
speed with an increase in motor torque, but this is
:not significant at high speeds. The performance
o f speed as the DC current is increased over the
armature, for different voltages, shows a minor
dependence on changes in current, but a high
dependence on variations of voltage.
In conclusion, the advantages of this scenario
are the easy speed control and the torque-speed
characteristics, flexibility in operating conditions
with changing voltage and current and the
response to changes in power, without the need or
an electronic control system. However, size and
cost are greater than that o f an equivalent AC
motor, and a rectifier is required.
Wind speed,
m/s
Theoretical potable
water flowrate,
dm3/s
30
60
90
120
150
180
210
240
270
300
330
360
390
420
450
480
510
540
570
600
630
8.83
7.55
8.11
7.48
8.16
8.04
7.66
8.56
8.51
7.04
8.79
8.71
7.86
7.05
7.67
7.97
7.33
8.98
8.88
7.10
8.12
9.49
9.03
9.26
8.99
9.28
9.24
9.08
9.42
9.40
8.75
9.48
9.46
9.17
8.76
9.08
9.21
8.92
9.53
9.51
8.79
9.27
With the power characteristics of a commercial 1.5 kW wind generator with a rotor diameter
of 3 m, a theoretical curve relating wind rotor
speed to wind speed was developed [7,8].
In accordance with the theory of three-phase
motors and tests run on them, a motor speed as a
function o f wind generator output was calculated.
Since the pump is a positive displacement device,
the speed determines the feeder flowrate regardless of working pressure.
Table 1
Theoretical data of plant performance at a given wincl
speed profile
262
F. Moreno, A. Pinilla / Desalination t 71 (2004) 25 7-265
12.9
~Wind
Speed (m/s) +
9,60
Theoretical PotableWater Flow rate
9.40
11.9
E
9,20 ~'~
10.9
9.00 E
o
~.
9.9
0
8.80 "~
8.9
8.60 ~=
7.9
8.40
6.9
8.20
30
90
150
210
270
330
390
450
510
570
630
Time [s]
Fig. 3. Theoretical water flowrate produced by the desalination unit operating under varying wind speed profiles.
The theoretical model can determine the

amount of potable water produced over time,
considering fluctuating wind speeds (Fig. 3). This
theoretical simulation assumes a salt water
concentration of 35,000 mg/L and a resulting
permeate with less than 1,000 mg/L, at a constant
system feed pressure of 3.4 MPa. The figure
shows that for a wind speed variation from 7 m/s
to 9 rn/s (typical wind speed variation at the site
of the future location of the plant), the theoretical
potable water flowrate varies from 8.8 dm3/s to
9.5 dm3/s (it is almost constant for variations in
wind speed). Table 1 summarizes the theoretical
data.
5. Laboratory testing of the prototype desalination plant

An experimental prototype was constructed
(Fig. 4) and the plant was then subjected to a
number of laboratory tests. The pro-treatment

system is comprised of an intake tank (1)
connected to two pro-filters (2) arranged in parallel with a flute (3). The pro-filtered water (profiltrate) is carried down the pro-filter discharge
flute (4) to a pro-filtrate storage tank (5). Finally,
the pro-filtrate is feed into the desalination system
(6).
The desalination system is bolted to a steel
structure. A three-phase 0.9 kW motor (7), the
positive displacement pump (8), the RO membrane (9) and the control and measurement panel
(10) are fitted to the structure. The control panel
is composed of two pressure gauges, one on the
suction side of the pump and the other on the
output side, with two flow-meters, one measuring
the flow rate of permeated water and the other
measures the flow rate of rejected water.
The photograph also illustrates the wind generator simulator assembly (11) composed of the
263
Desalination
inlet Tank (6)
'Inlet
)
Fee
Motol
Pump
Fig. 4. Photograph of the

experimental prototype of the
Instrl
contn
RO wind-powered desalination plant.
1200
~1000
.E
SirnLdation
Expenmenlal
~=8o0
6oo
8i
400
~oo
3,00
3.50
4,00
4.50
5+00
Feed Pressure [MPa]
5,50
6.00
Fig. 5. Validation of operation of the prototype and

ROSA simulation (feed flow rate with 35,000 mg/L of
salts).
generator supplying AC current to the desalination plant. The prime mover of the simulator is
connected to a frequency controller that allows
variations in speed in accordance with the windspeed variations simulated by a computer through
a PLC.
The desalination unit was subjected to tests in
the laboratory at a constant electricity supply
using a solution of salt water (close to

35,000 mg/L) and tea in a proportion of 1:3. In
order to validate the plant's performance, the
ROSA simulated results were compared with the
experimental laboratory tests.
In Fig. 5 the experimental laboratory results
show an adequate desalination plant output with
regard to feed pressure variation, securing a permeate concentration of less 1,000 mg/L of salts
dissolved. Table 2 shows a comparison of ROSA
simulation and the experimental laboratory results
on the desalination plant with respect to recovery
and salt rejection rates. It is worth noting the
close agreement in the results, both simulated and
experimental, as well as, a good salt rejection and
recovery rates not only when compared by
numerical values but in the general tendency.
This experimental prototype will permit more
conclusive information on the overall performance of the desalination plant when tested in
field trials matched to a wind generator, as the
next step of the project.
264
F. Moreno, A. Pinilla / Desalination 171 (2004) 25 7-265
Table 2
Experimental laboratory data of the desalination unit compared with simulation results
Feed pressure,
MPa
ROSA simulation (feed
concentration: 35,000 mg/L):
3.4
4.1
4.8
5.5
Laboratory experimentation
(feed concentration: 35,725 mg/L):
3.6
4.0
4.5
5.0
5.5
Feed
Permeate
Recovery,
flowrate, L/s flowrate,L/s %
Permeate
Salt
concentration, mg/L rejection, %
0.19
0.19
0.19
0.19
0.006
0.010
0.014
0.018
3.0
5.3
7.7
9.7
771
462
339
271
97.8
98.7
99.0
99.2
0.133
0.133
0.133
0.133
0.133
0.003
0.005
0.008
0.013
0.017
2.5
3.8
6.3
10.0
12.5
935
707
409
308
211
97.3
98.0
98.8
99.1
99.4
6. Conclusions and recommendations

This phase o f the project explored the possibility of operating a small desalination plant
without using any form o f electronic controls.
The most appropriate alternative for operating a
desalination plant is the high-voltage wind generator, connected directly to the motor. It allows
variations of voltage and frequency in the
generation of electricity to excite conventional
electrical motors. This is a simple and efficient
way of operating a plant since it eliminates additional power-consuming devices such as rectifiers, transformers and battery banks.
Although the DC electricity supply allows
control of speed and torque-speed characteristics,
and it offers flexibility in operating characteristics
with variations in voltage and current, the system
is larger and more expensive than the AC equivalent. The DC supply also needs a rectifier.
Therefore, this option provides a solution but
does not meet the requirement o f simplicity that
the project intends to provide.
The immediate phase o f the project will allow
more representative validation o f the system after
testing the prototype under variable electrical

conditions. The testing o f the prototype will give
a deeper understanding o f the performance of the
system, particularly considering wind speed
variations.
Further work will be carried out to perform a
broader set of experiments on the pre-treatment
system, an in-depth study of the effect of temperature on the performance of the RO membrane
and the analysis of the most appropriate posttreatment system.
Acknowledgements
The authors wish to express their thanks to
Instituto Colombiano para el Desarrollo de la
Ciencia y la Tecnologia ( B I D - COLCIENCIAS)
and the Colombian commercial company
Acquaire Ltda., for their cooperation and funding
of this research project. The authors would also
like to thank the company Severn Trent Universal Aqua for its kind donation of the RO
membrane.
References
[ 1] P. Smulders and P. Feron, Seawater desalination and
wind energy, Commission of the European Communities, Hamburg, 1984.
[2] A. Pinilla and F. Moreno, Desarrollo de un sistema
de control para los procesos de desalinizaci6n de
agua y refrigeraci6n, con base en plantas de osmosis
inversa y refrigeradores, operados con sistemas de
energia no conveneional, Informe T6cnico No 1,
COLCIENCIAS, Bogota, 2003 (in Spanish).
[3] J.C. Botero, Estudio de un sistema de desalinizaci6n
de agua marina por osmosis inversa, Proyecto de
Grado, Departamento de Ingenierla Mechnica.
Universidad de Los Andes, Bogota, Colombia, 2002
(in Spanish).
[4] G. Tchobanoglous and E. Schroeder, Water Quality,
Addison-Wesley, New York, 1985.
265
[5] A. Grauers, Design of direct-driven permanentmagnet generators for wind turbines, Chalmers
University of Technology, G6teborg, Sweden, 1996.
[6] B. Guru and H. Hiziroglu, Electric Machinery and
Transformers, 2nd ed., Saunders College, 1995.
[7] S. Espinosa, Disefio y construcci6n de un control de
velocidad del eje de un motor, para simular las
condiciones de incidencia de viento en el rotor de un
molino, Proyecto de Grado, Departamento de Ingenieria Mec~inica. Universidad de Los Andes, Bogota,
Colombia, 1998 (in Spanish).
[8] A. Zapata, Caracterizaci6n de un generador de
imanes permanentes, Proyecto de Grado, Departamento de Ingenieria Mec6xtica, Universidad de Los
Andes, Bogota, Colombia, 2001 (in Spanish).
Energy 36 (2011) 4372e4384
Contents lists available at ScienceDirect
Energy
journal homepage: www.elsevier.com/locate/energy
Assessment of a stand-alone gradual capacity reverse osmosis desalination plant

to adapt to wind power availability: A case study
Baltasar Peate a, *, Fernando Castellano a, Alejandro Bello a, Lourdes Garca-Rodrguez b,1
a
b
Water and Renewable Energies Departments, Canary Islands Institute of Technology (ITC), Playa de Pozo Izquierdo s/n. 35119 Santa Luca e Las Palmas (Spain)
Departamento de Ingeniera Energtica, Universidad de Sevilla, ETSI, Camino de Los Descubrimientos, s/n. 41092-Sevilla (Spain)
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 September 2010
Received in revised form
25 March 2011
Accepted 3 April 2011
Available online 30 April 2011
Desalination driven by renewable energies is an interesting technology in isolated coastal areas. Its
feasibility and reliability are guaranteed by innumerable designs implemented and experiences carried
out, mainly focused on small capacity systems. However, only mature and efcient technologies are
suitable for medium or large scale desalination. In the case of seawater desalination, wind-powered
reverse osmosis is the most efcient, mature and cost-effective technology. This paper assesses the
most suitable design for seawater reverse osmosis desalination driven by off-grid wind energy systems.
A high innovative design based on gradual capacity with nominal production of 1000 m3/d is compared
to a conventional xed capacity desalination plant. Due to the intermittent wind resource, the gradual
capacity desalination plant is able to t the available energy and maximize the annual water production.
2011 Elsevier Ltd. All rights reserved.
Keywords:
Wind energy
Reverse osmosis desalination
Medium scale
Gradual capacity
1. Introduction
The Instituto Tecnolgico de Canarias (ITC) e a public research
company of the Canary Islands (Spain) e coordinated a pioneering
project called SDAWES (Seawater Desalination with an Autonomous Wind System) [1]. It was the rst initiative intended to test
a SeaWater Reverse Osmosis (SWRO) system (8 25 m3/d) fed
directly with wind power, alongside other technologies (electrodialysis and mechanical vapour compression). The eight Reverse
Osmosis (RO) modules were connected or disconnected depending
on the available wind power resources, thus resulting in a gradual
production capacity system. Subiela et al. [1] analysed the lessons
learnt of this experience and concluded that SWRO technology was
the best of all the desalination technologies tested using wind
energy. Nevertheless, an improved design of a gradual production
capacity RO system should be conceptually and experimentally
analysed.
This paper deals with a comparative analysis of two different
wind-powered reverse osmosis plants for seawater desalination,
based on the experience gained by the ITC within SDAWES project.
* Corresponding author. Tel.: 34 928727511; fax: 34 928727590.

E-mail addresses: baltasarp@itccanarias.org (B. Peate), lourdesg@esi.us.es
(L. Garca-Rodrguez).
1
Tel.: 34 954487231; fax: 34 954487233.
0360-5442/$ e see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2011.04.005
The two plants analysed in this paper are driven by an off-grid wind
energy system:
The rst system is an energy efcient design of a SWRO plant
with 1000 m3/d of nominal production capacity. The plant
could operate within a restrictive range of power consumption,
only when wind resource is enough to supply all the energy
required by the RO plant.
The second system is an energy efcient design of a 1000 m3/d
gradual production capacity SWRO plant. This system consists of
three RO racks, two of them of 400 m3/d capacity and the other
one of 200 m3/d. The three racks are able to operate independently, being connected or disconnected depending on the
amount of energy available.
The selected capacity for the analysis can be considered as
medium capacity within the framework of renewable energypowered desalination. However, analyses of SWRO powered by
wind energy have been normally focused only on small scale
desalination [2,3]. Garca-Rodrguez [4] reviews desalination
technologies driven by wind power. Romero-Ternero et al. [5]
analyzes different parameters for a medium capacity SWRO
driven by wind energy in the Canary Islands. The product cost
ranges obtained prove that this technology is the most costeffective compared to other systems based on renewable energydriven desalination [6].
B. Peate et al. / Energy 36 (2011) 4372e4384
4373
Fig. 1. Seawater reverse osmosis desalination plant driven by wind energy design process.
The connection of a wind system to a SWRO plant may be

possible either, through electrical or mechanical coupling. Previous
experiences in mechanical coupling [7e10] have not obtained
satisfactory results for medium capacity systems. In this case, the
high-pressure pump of the RO systems is connected directly to
a Wind Turbine (WT) through a shaft linkage or by a hydraulic
system.
Therefore, wind energy conversion into electricity is the only
feasible option to operate a medium or large scale desalination
plant. Product cost and the inuence of design and operational
parameters are analysed. In addition, there are some pilot schemes
on a small scale that demonstrate the technical feasibility of such
systems for medium and large capacity.
Several experiences of RO desalination powered by WTs have
been developed. Within JOULE Programme, the OPRODES project
combined a variable 43e113 m3/h RO unit with a 30 kW nominal
power wind generator in 2001. This experience was coordinated by
the University of Las Palmas de Gran Canaria [10]. Moreover, Carta

et al. [11] describes an initiative made within VALOREN Programme
of the European Commission in the isolated village of Puerto de la
Cruz e at the southern of the island of Fuerteventura. A wind/diesel
RO system fully covers the energy and potable water requirements
in stand-alone conditions. The power system consisted of a 225 kW
WT e two 160 kVA diesel engines with ywheel and synchronous
generator of 75 kVA each, to produce electricity for village lighting,
sh preserving in refrigerated chambers, sanitation and a 56 m3/d
SWRO plant. Each one of the two diesel sets is mechanically connected through an electromagnetic clutch to the ywheel. Each
ywheel can deliver half of the envisaged peak power demand e
100 kW e during 30 s approximately, without frequency decreases
below the minimum level allowed (48 Hz).
In general, all the experiences require a studied design and
power control system to ensure optimized operation and stability
in the system functioning. In general, the use of wind energy to
Fig. 2. Basic scheme of the wind energy system connected to the seawater reverse osmosis desalination plant.
4374
Fig. 3. Diagram of a gradual capacity seawater reverse osmosis plant.
supply a stand-alone SWRO system, adapted to medium or large

water demands, entails taking into account a number of technical
considerations:
Use of synchronous wind turbines with power control: this
generator is able to adapt to load variations quickly and
generate more energy according to needs. The latest WT
developments for medium and high power propose an energy
topology production that combines a synchronous generator of
permanent magnets with a bi-directional power converter (full
converter). This topology allows controlling the WT power ow
with the electrical grid connected.
Short-term (ywheel) and mid-term (batteries) appropriate
dimensioned of storage system. First one is useful for primary
regulation, providing active and reactive power control and
Table 1
Main design parameters of 1000 m3/d xed capacity seawater reverse osmosis
plant.a
Main streams
Volumetric
ow (m3/d)
Pressure
(MPa)
Pipe dimension
(mm)
Feed water intake (A)

Product water (I)
Brine water (H)
High-pressure pump outlet (C)
K-200 inlet (G)
Booster pump inlet (D)
Booster pump outlet (E)
RO rack inlet (F)
2500
998
1502
1000
1496
1494
1494
2494
0.25
0.07
0.10
5.77
5.57
5.51
5.77
5.74
160
100
160
101
127
101
101
127
(4-inch)
(5-inch)
(4-inch)
(4-inch)
(5-inch)
a
Recovery rate, 40%; fouling factor, 0.85; energy recovery device: differential
pressure high-pressure side 60 kPa, differential pressure low-pressure
side 30 kPa; leakages, 0.4%.
frequency stability to the system. Second one warranties time

of continuous operation.
RO plants connected to an off-grid wind farm have got to
consider a charge control system. This system must allow
absorb or disconnect the power consumed in operation.
Table 2
1000 m3/d xed capacity seawater reverse osmosis rack nominal characteristics.
RO rack characteristicsa,b,c
No. pressure vessels
Pressure vessel conguration
Total No. elements
Total active membrane area, Am (m2)
Average ux, qvp/Am (l/(m2$h))
No. ff ERD K-200 required
Energy recovery efciency, hERD
K-200 energy savings, kW
High-pressure pump electrical
power required, kW
Booster pump electrical power
required, kW
Intake pump electrical power
required, kW
RO process specic energy
consumption, kWh/m3
Total specic energy consumption,
kWh/m3
a
11
2 SW30HRLE-400i
5 SW30ULE-400i
77
2861.32
14.54
2
0.98
104.0
70.2
6.4
35.4
1.93
2.78
Recovery rate, 40%; fouling factor, 0.85.

Hydraulic performance of high-pressure pump, 0.92; BOoster Pump (BOP)
performance, 0.77; intake seawater pump performance, 0.74; variable frequency
drives performance, 0.98 and motor efciencies, 0.95.
c
Isobaric energy recovery device (ERD) was designed according to the following
criteria: 30 kPa ERD low-pressure site differential; 60 kPa ERD high-pressure side
differential; 0.4% lubrication ow e leakages.
b

Table 3
Volumetric ows of 1000 m3/d gradual capacity seawater reverse osmosis plant.a
Main streams
Feed water intake (A)

Product water (I)
Brine water (H)
High-pressure pump
outlet (C)
K-200 inlet (G)
Booster pump inlet (D)
Booster pump outlet (E)
RO rack inlet (F)
Volumetric ow (m3/d)
Pressure
(MPa)
Train
No. 1
Trains
No. 1 2
Trains
No. 1 2 3
571
199
372
200
1714
599
1113
600
2857
997
1854
1000
0.25
0.07
0.10
5.66
370
370
370
570
1111
1110
1110
1710
1852
1850
1850
2850
5.38
5.32
5.66
5.63
a
Recovery rate, 35%; fouling factor, 0.85; energy recovery device: differential
pressure high-pressure side 60 kPa; differential pressure low-pressure
side 30 kPa; leakages, 0.4%.
Water storage, preferably in height, to meet the water demand

in long temporary spaces of lack of wind resource or serious
breakdowns of the system.
According to the expertise of the ITC, for designing an optimized
stand e alone RO-WIND system the following points should be
considered:
Variable frequency drives for pumps to achieve slight ow
variations must be included in the process. It reaches slight
changes in power consumption. In this case, it should be
assessed the effects of harmonic generation over the grid [1]
and analysed the installation of devices to minimize this
effect over the stand-alone grid.
Design a control system and gradual operation mode. This
connection requires a variable capacity to modify quickly the
water production in the case of signicant variations of the
wind resource. This gradual production regulation mode
allows adapting the energy consumption progressively. Such
systems require and adequate balance between the generated
Table 4
1000 m3/d gradual capacity seawater reverse osmosis rack nominal characteristics.
Train
No. 1
No. pressure vessels
Pressure vessel conguration
No. elements
Total active membrane area, Am (m2)
No. K-200 needed
power required, kW
required, kW
required, kW
consumption, kWh/m3
Total specic energy consumption
a
Trains
No. 1 2
Trains
No. 1 2 3
2
6
10
2 SW30HRLE-400i 5 SW30ULE-400i
14
42
70
520.84
1561.32
2601.8
15.94
15.98
15.96
1
1
2
0.98
0.97
0.97
41.3
123.9
206.5
14.5
43.5
72.5
2.1
6.3
10.5
8.1
24.3
40.5
2.96
2,97
Product volumetric ow, m3/d

power required, kW
required, kW
required, kW
consumption, kWh/m3
Train
No.1
Trains
No. 1 2
Trains
No. 1 2 3
139
11.11
0.98
26.1
9.2
429
11.45
0.97
81.1
28.5
719
11.51
0.96
136.1
47.8
1.0
3.2
5.4
5.6
17.4
29.1
1.76
1.77
2.73
2.74
2.75

b
c
power, the energy consumed in the RO plant and the available

energy stored.
Variable power system operation will generate frequent and
signicant uctuations in the main operation parameters e
ow and pressure basically. It is recommended to select high
quality materials, especially for high-pressure water pipelines
and to assess carefully the choice of the RO membranes.
2. Work hypothesis and design procedure

The systems proposed are stand-alone desalination plants. In
both of them, a SWRO plant driven by a wind farm, able to adapt its
production to the available power. Loads in this case are referred to
SWRO system. There are two major subsystems:
1. No. 1: The Wind Energy Conversion Subsystem (WECS) is
composed by the following items:
WTs that generate electricity,
Energy Storage System (ESS), which could be either,
mechanical (ywheel) or electrical (batteries) or both,
depending on the necessity of short-terms or medium/longterms storage and control requirements,
Electrical connection between WT and the RO subsystem,
Master power electronic equipment, which controls the
main parameters of the grid (voltage, frequency and reactive
Table 6
Summary of energy requirement in a 1000 m3/d xed capacity seawater reverse
osmosis design.
2,97

c
b
Table 5
Minimum operation points of 1000 m3/d gradual capacity seawater reverse osmosis
rack characteristics.
Energy requirements
1.99
4375
Product water volumetric ow, m3/d

Recovery rate, R
Electrical power required, kW
consumption, kWh/m3
1000 m3/d xed capacity

Nominal operation
point
Minimal
operation point
998
0.40
112
1.93
748
0.34
82.8
1.73
2.78
2.73
4376
Table 7
Summary of energy requirement in a 1000 m3/d gradual capacity seawater reverse osmosis design.
Energy requirements
1000 m3/d gradual capacity

Trains No. 1 2
Train No.1
Product water volumetric ow, m3/d

Recovery rate, R
Electrical power required, kW
RO process specic energy consumption, kWh/m3
Trains No. 1 2 3
Nominal
operation point
Minimal operation
point
Nominal
operation point
Minimal operation
point
Nominal
operation point
Minimal operation
point
199
0.35
24.7
1.99
139
0.35
15.8
1.76
599
0.35
74.1
1.99
429
0.35
49.1
1.77
997
0.35
123.5
1.99
719
0.35
82.3
1.77
power) and manages active power ow between WT,

storage system and RO subsystem,
Control system: it regulates the RO modules start-up and
stop, and the charge/discharge regulation of the ESS through
the master power electronic equipment.
2. No. 2: The SWRO plant uses the electricity generated by the
WECS subsystem.
The calculation procedure is summarized in Fig. 1. The design
begins with the dimensioning of the SWRO plant and the calculation of its energy consumption. These data determine the size of the
energy system. The power of the WT depends on the wind resource
available at the chosen location. The ESS capacity depends on the
WT power and the energy consumption of the selected RO plant.
Energy production, equivalent hours, capacity factor and annual
water production are calculated for the WT model selected. Three
different models of WTs were preselected in order to identify which
is the model with best adaptability to the system requirements. As
mentioned before, it is advisable to use an ESS that smoothes
uctuations in wind resource. The choice of the type of ESS and its
maximum capacity is a key factor in the system design but it is not
considered in this work, due to the specic solution, whether
batteries, ywheel or both, will depend on the characteristics of the
WT and the RO power response.
Fig. 2 shows the system as a whole, showing the WECS and
SWRO plant subsystems main ows and their equipments.
3. Seawater reverse osmosis desalination plant
Technical and design characteristics of the desalination unit
should allow to achieve the highest energy efciency and high
quality of the product water (permeate). Two SWRO plant design
possibilities are considered and compared below.
Fixed Capacity (FC) SWRO plant: one train of 1000 m3/d. The FC
system is able to operate only when the wind resource is
enough to supply all the energy required by the RO plant in its
low-energy operation point.
Gradual Capacity (GC): Three RO racks with production
capacities of 200 m3/day (Rack No. 1) and 2 400 m3/day
(Racks No. 2 and 3), respectively. The GC system is able to
operate with production gradually adapted to the available
wind power. According to that it is obtained the previous rack
capacities. It most adequate to prevent a low capacity rack
instead of three identical ones with the aim to produce water
even with low wind power periods.
According to this design the following treatment line raises:
Seawater intake through beach-wells: in the case of GC case
will be necessary to install two intake pumps for supplying the
feed water demanded depending on the capacity required.
RO pre-treatment: Antiscalant dosing and ltration with sand

and cartridge lters.
RO process: Each RO train (one in FC case and three in the GC
design) consists of a High-Pressure Pump (HPP), Energy
Recovery Device (ERD), BOoster Pump (BOP) and RO rack. The
latter is integrated by several pressure vessels with seven
membrane elements each. The design of each pressure vessel
consists in seven-element hybrid membrane interstage design
(2 SW30HRLE-400i 5 SW30ULE-400i membrane elements).
Each of the pumps is equipped with a Variable Frequency
Driver (VFD) which allows slightly adapting the ows to the
wind power available.
Brine discharge: consisting of a brine pipe drain by direct
drainage to sea.
Clean-in-place system: Chemical cleaning pump and tank.
The general description of each RO train is similar to the desalination unit shown in Fig. 2. Fig. 3 shows the proposed design for
the GC system, with three RO trains and different pumps and ERDs.
The isobaric ERDs transfer the energy from the membrane reject
stream directly to the membrane feed stream [12]. In the case of GC
an ERD which permits recovery and ux variations is required. The
degree to which centrifugal ERD performance varies as a function of
recovery and membrane ux changes depending upon the characteristics of a particular device and must be considered in the
SWRO design process. In this GC design the selection of a exible
ERD is important. Isobaric ERDs must allow varying membrane ux
and recovery independently. If the ow rate of the booster pump is
set with a variable frequency drive to be equal to the ow rate of the
HPP, the system will operate at 50% recovery rate. If the ow rate of
the booster pump is increased to double the ow rate of the HPP,
the system will operate at 33% recovery rate. These operation
conditions do not change signicantly the HPP ow rate or the
permeate ow rate [13]. In this sense, the RO Kinetic ERD is
proposed.
Table 8
Monthly wind speed average and typical deviation at Pozo Izquierdo (Gran Canaria
island) at 10 m height.
Month
Speed average, m/s
Typical deviation
Jan
Feb
March
Apr
May
Jun
July
Aug
Sept
Oct
Nov
Dec
Annual
5.86
7.09
5.08
5.6
8.44
8.82
12.56
10.22
9.2
5.06
5.86
4.47
7.36 m/s
3.34
3.71
2.93
3.67
4.03
4.06
3.31
3.13
4.39
3.14
2.88
3.11
4377
Monthly wind speed average (10 m)

(m/s)
14
Wind speed
12
Annual wind speed av.

7,36 m/s
10
8
6
4
2
0
Jan
Feb
March
Apr
May
Jun
Jul
Aug
Sept
Oct
Nov
Dic
Fig. 4. Monthly wind speed average at Pozo Izquierdo e Gran Canaria island (3500 equivalent hours/years) at 10 m height.
Therefore, the GC design allows operating with, one, two or

three racks depending on the available wind power. In turn,
installed ERD should be exible and able to be quickly adapted to
the ow and recovery rate desired. Each RO train should work at
variable recovery rate, ow and pressure, within ranges dened by
the membranes and installed equipments. Besides, the FC design
can only work until limited minimum recovery rate that will be
dened by the lower limit of feed ow to the RO train.
For designing the RO trains, the leading membrane manufacturer Filmtec was selected. All simulations were gauged using
Filmtec RO design software Rosa v.7.01 [14]. The SWRO trains
design parameters are the following:
Seawater characteristic: 38,170 mg/l TDS (east Atlantic seawater
beach-well), temperature: tsw 20 C, pH 7.2, density:
rsw 1.03 kg/l, kinematic viscosity: nsw 1.25$106 m2/s
Brine density: rb 1.05 kg/l, brine kinematic viscosity: nb
1.80$106 m2/s.
Fouling factor: 0.85.
Seawater intake pumps: submerged centrifugal pump;
hydraulic performance: hIP 0.74.
Seawater pressure required (from intake to high-pressure
pump inlet): pIN 850 kPa.
Recovery rate: variable R 30e45%.
HPP: axial piston pump; hydraulic performance, hHPP 0.92.
BOP: centrifugal pump; hydraulic performance, hBOP 0.77.
Variable frequency drives performance: hVFD 0.98.
Motor performance: hVFD 0.92.

ERD: K-200 model [15]; brine volumetric ow rate, 15e30 m3/
hr; low-pressure site differential, pLP 30 kPa; high-pressure
side differential, pHP 60 kPa, lubrication ow e leakage
rate, 0.4%.
Based on aforementioned data, Tables 1 and 2 give the nominal
characteristics of the FC RO plant. Table 1 shows the nominal
operation point data of the plant. The nomenclature of the mass
and energy streams is related to Fig. 3. A 40% recovery rate is obtained and the high pressure at nominal operation is 5.74 MPa. The
RO train conguration consists of 11 pressure vessels (203 mm (8inch) diameter and 6.9 MPa (1000 psi) nominal working pressure).
The pressure vessels will house 77 SWRO membranes e see Table 2.
Very low specic energy consumption is achieved for the desalination process. Simulations get the minimum operation point for
FC design, in which is able to work respecting the product water
quality parameters, minimum working pressure and the design
ow of hydraulic pipes installed. In this case, the minimum point
corresponds to a production of 748 m3/d and R 34%. The reduction of the energy consumption is up to 26% of the required energy
in comparison with the nominal operation point.
The following Tables 3e5, show the design parameters of the
SWRO plant with a GC design. Tables 3 and 4 show the RO characteristics, the number of pressure vessels in operation is variable.
It increases from 2 to 10 depending on the wind power availability,
corresponding to a number of membranes from 14 to 70. Each of
Table 9
Monthly and annual power generation, capacity factor and equivalent hours of three wind turbines. Pozo Izquierdo e Gran Canaria Island location.a
Wind generator
FUHRLANDER FL100 (100 kW)
Month
Jan
Feb
March
Apr
May
Jun
July
Aug
Sept
Oct
Nov
Dec
Annual
MWh
23.80
31.02
23.74
22.49
58.05
46.54
71.27
58.77
48.00
25.00
22.65
14.76
446.10
Equiv. hours
238.00
310.20
237.41
224.90
580.53
465.40
712.70
587.70
480.00
250.01
226.50
147.60
4,460.96
VESTAS V27 (225 kW)

Cap. Factor
0.32
0.46
0.32
0.31
0.78
0.65
0.96
0.79
0.67
0.34
0.31
0.20
0.51
MWh
45.59
58.90
45.79
42.97
109.31
87.48
133.84
110.12
89.86
47.98
43.59
28.38
843.80
Equiv. hours
202.62
261.78
203.51
190.98
485.81
388.80
594.84
489.42
399.38
213.23
193.73
126.13
3,750.24
ENERCON E32 (300 kW)

Cap. Factor
0.27
0.39
0.27
0.27
0.65
0.54
0.80
0.66
0.55
0.29
0.27
0.17
0.43
MWh
68.56
88.30
70.25
64.55
159.92
127.61
185.88
160.18
128.05
73.48
66.83
43.22
1,236.83
3500 equivalent hours/years; 7.36 m/s annual wind speed average; Rough length: z0 0.0024; wind turbine reliability 0.9.
Equiv. hours
228.54
294.33
234.16
215.16
533.08
425.37
619.59
533.94
426.84
244.92
222.78
144.06
4,122.77
Cap. Factor
0.31
0.44
0.31
0.30
0.72
0.59
0.83
0.72
0.59
0.33
0.31
0.19
0.47
4378
Table 10
Annual simulation results for three wind turbines coupled to a 1000 m3/d xed
capacity seawater reverse osmosis design.
Annual simulation
Table 11
Annual simulation results for three wind turbines coupled to a 1000 m3/d gradual
capacity seawater reverse osmosis design.
Annual simulation
Results
FL100
V27
E32
Results
FL100
V27
E32
Total water produced, m3

Annual average product
volumetric ow, m3/d
Total water not-produced, m3
Total plant operation hours, h
Annual operation rate, %
Total plant stop hours, h
Energy produced by wind
turbine (kWh)
Energy consumed by the
desalination plant (kWh)
Excess energy (kWh)
Productivity ratio, qvp/PWT
172,896
473.69
230,318
631.01
250,865
687.30
159,166
436.07
223,090
611.21
245,540
672.71
191,457
4487
51%
4273
468,674
134,035
5688
65%
3072
886,511
113,488
6135
70%
2625
1,299,430
204,822
6464
74%
2296
468,674
140,899
6844
78%
1916
886,511
118,448
7062
81%
1698
1,299,430
467,161
620,901
676,008
468,599
660,850
728,179
1512
0.37
265,610
0.26
623,421
0.19
Total water produced, m3

Annual average product
volumetric ow, m3/d
Total water not-produced, m3
Total plant operation hours, h
Annual operation rate, %
Total plant stop hours, h
Energy produced by wind
turbine (kWh)
Energy consumed by the
desalination plant (kWh)
Excess energy (kWh)
Productivity ratio, qvp/PWT
74
0.34
225,661
0.25
571,250
0.19
the 3 possible design points could reduce the energy required by

varying the pressure and feed ow. In addition, the minimum
operation points of each rack are given in Table 4.
4. Wind energy conversion subsystem
From above section the loads of the wind energy system are
known. A summary of the minimum and nominal energy required
by the RO plants is shown in Tables 6 and 7 for both, FC and GC
designs, respectively.
Once the energy consumption is dened, the next step is the
design of the WECS. The WT power will be dened depending on
wind resource and consumption range of the RO plant. Three WTs
with a nominal power around the calculated power range of the
desalination plant are selected for the comparative analysis. In this
case three wind generators of 100 kW, 225 kW and 300 kW of
nominal power are considered. All simulations were developed
using the ITC software CE2000. This software has been designed
within a Regional Program for the wind resources management of
the Canary Islands. The theoretical fundaments used for this
calculation proceeding is shown in the Appendix A.
To design the WECS the following parameters are used:

Wind speed, v (m/s): The location at Pozo Izquierdo (Gran
Canaria, Canary Islands) for the whole system was chosen. The
wind speed for a model year is represented in Table 8 and Fig. 4.
Rough length: z0 0.0024 m.
Air density: rai 1.204 kg/m3 (at 20 C).
Air dynamic viscosity: maid 1.80$105 (kg/m$s)
Air cinematic viscosity: maic 1.50$105 (m2/s)
Relative air energy: 244.2 W/m2 (at rai 1.225 kg/m3 and 15 C).
WT reliability 0.9
WT 1: FUHRLANDER FL100 (100 kW) [16]
WT 2: VESTAS V27 (250 kW) [16]
WT 3: ENERCON E32 (300 kW) [16]
ESS: previous consideration have to be considered and revised
with further calculation:
BMinimum total ESS considered: 3 h (estimation for the
normal operation of the RO plant and maintain tasks).
BTotal useful ESS capacity: 1200 Ah
BTotal ESS performance: 0.7
Fig. 5. Annual wind energy produced and water obtained with a Vestas 27 WT coupled to a 1000 m3/d xed capacity seawater reverse osmosis design.
4379
Fig. 6. Annual wind energy produced and water obtained with a Vestas 27 wind turbine coupled to a 1000 m3/d gradual capacity seawater reverse osmosis design.
The values of energy produced, equivalent-hours and capacity

factor for each WT over one year are represented in Table 9 (equations in the Appendix A). The capacity factor is the relationship
between the real energy produced by a wind generator and the
energy theoretically produced during a dened period. Equivalenthours are the number of hours that a WT would have been operating
to the nominal power to produce the same amount of energy over
the same period of time, i.e. the capacity factor expressed in hours.
Values given in Table 9 determine how the WT ts to the
proposed site. A proper capacity factor value must exceed 0.4,
including aspects of the WT availability. The FL100 wind generator
proposed exhibits the best capacity factor, however annual energy

production is somewhat less than the energy provided by the other
two proposed WTs.
The energetic and hourly water production calculations are
performed after obtaining the specic WT data for the selected site.
From those, the monthly and annual calculations are carried out.
The objective is to identify which combinations offer the best
solution to obtain the highest water production and operation hour
rates. For this purpose the operation range of both desalination
plants is divided in different power required steps. Two steps for
the FC plant and ve steps in the case of GC plant. The energy
W in d en erg y p ro d u ced vs. W ater o b tain ed (F eb ru ary)

250
100
Wind machine
225
90
200
80
175
70
150
60
125
50
100
40
75
30
50
20
25
10
P roduct volum etric flow (m 3/h)
Average hourly energy (kW )
Water pro duced (m3/h)
0
1
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Da ys
Fig. 7. Wind energy produced and water obtained with a Vestas 27 wind turbine coupled to a 1000 m3/d xed capacity seawater reverse osmosis design (February simulation).
4380
W in d en erg y p ro d u ced vs. W ater ob tained (Feb ru ary)

250
100
Wind m achine
225
90
200
80
175
70
150
60
125
50
100
40
75
30
50
20
25
10
P rod uct vo lum etric flo w (m 3/h)
Average hou rly energy (kW )
Water pro duced (m 3/h)
0
1
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Da ys
Fig. 8. Wind energy produced and water obtained with a Vestas 27 wind turbine coupled to a 1000 m3/d gradual capacity seawater reverse osmosis design (February simulation).
balance of the whole system is realised every hour following

a sequence calculation:
taking into account the investment that involves the desalination

plant, the option of FL100 with 51% of annual operation rate could be
Step 1

Step 2

Power Storagen2
Power Windn1
a
Power demanded SWROn kW Step . kW/m3 =h a Power Storagen1 KW a
Power excessn1 kW
Power available SWROn1 kW
Step .
Step 5
Tables 10 and 11 summarize the resulted of this sequence

simulation for the two RO plant designs. Results obtained for the FC
system (Table 10) shows that the three wind generators cover more
than 50% of the maximum output capacity of the plant. However,
dismissed. Regarding the E32/300 kW WT, the values of production

and coverage are higher than those presented by the V27/225 kW
WT. However, the cost of the increase in water production and
higher excess of energy are considered insufcient to justify the
W in d en erg y p ro d u ced vs. W ater o b tain ed (Ju ly)

Water pro duced (m3/h)
Wind machine
100
225
90
200
80
175
70
150
60
125
50
100
40
75
30
50
20
25
10
P ro duct vo lum etric flow

(m 3/h)
Averag e hourly energy (kW )
250
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Da ys
Fig. 9. Wind energy produced and water obtained with a Vestas 27 wind turbine coupled to a 1000 m3/d gradual capacity seawater reverse osmosis design (July simulation).
4381
E n erg y b alan ce (F eb ru ary)

250
2.000
Wind machine
225
1.875
SWRO plant
B attery cap. (A h)
175
Averag e h ou rly en erg y (kW )
1.750
B alance (kW)
1.625
150
1.500
125
1.375
100
1.250
75
1.125
50
1.000
25
875
750
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
-25
625
-50
500
-75
375
-100
250
-125
125
-150
Battery cap acity (Ah )
200
Da ys
Fig. 10. Energy balance of a Vestas 27 wind turbine coupled to a 1000 m3/d xed capacity seawater reverse osmosis design (February simulation).
increase in the investment cost. For that reason, the V27/225 kW WT

could be considered like the optimum WT to install.
Besides that, Table 11 shows results of GC system analysis. FC
plant allows the production of a greater amount of water per year in
comparison with GC design, but the desalination unit operates less

number of hours along the year. Consequently, the GC design
produces between 2 and 8% water production fewer than FC design.
However, the annual operation rates are higher with the GC design
E n erg y b alan ce (F eb ru ary)

250
2.000
Wind machine
225
1.875
SWRO plant
B attery cap. (A h)
175
Average hourly energy (kW)
1.750
B alance (kW)
1.625
150
1.500
125
1.375
100
1.250
75
1.125
50
1.000
25
875
750
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
-25
625
-50
500
-75
375
-100
250
-125
125
-150
Da ys
Fig. 11. Energy balance of a Vestas 27 wind turbine coupled to a 1000 m3/d gradual capacity seawater reverse osmosis design (February simulation).
Battery capacity (Ah)
200
4382
2.000
225
1.875
200
1.750
175
1.625
150
1.500
125
1.375
100
1.250
75
1.125
50
1.000
25
875
0
-25
750
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
-50
625
Averag e h ourly en erg y (kW )
E n erg y b alan ce (Ju ly)

250
500
Wind machine
-75
375
SWRO plant
-100
B alance (kW)
250
-125
B attery cap. (A h)
125
-150
Da ys
Fig. 12. Energy balance of a Vestas 27 wind turbine coupled to a 1000 m3/d xed capacity seawater reverse osmosis design (July simulation).
and the excess energy is lower. It could be considered as an

advantage in order to reduce investment costs due to the highest
production-excess energy relation obtained.
The following graphs (Figs. 5 and 6) represent the annual
evolution of the energy produced by the WT V27/225 kW and
water production. It can be observed for both RO plant designs that
in months with many calm days the operation of the plant is more
intermittent. On the other hand, in days with excellent wind
resource the plant is operated for entire days at nominal operation
point. These periods show temporary spaces with surplus energy.
Non-operation period are not related to the large power of the
WT installed, but hours of calm of the site. For this reason, these
periods do not reach overall percentages of more than 80% of
operating hours of the plants and these are determined to the wind
frequency distribution of the site selection. Besides that, the GC
design presents a number of operating hours higher than FC design,
but lower at the nominal operation RO.
Figs. 7e10 are related to the same systems, they present the
behaviour of the system as a whole. The two SWRO plant designs
are represented in specic months, operated by using the V27/
225 kW WT. Firstly, Figs. 7 and 8 depict the operation of the system
in a low wind resource month (February) for FC and GC plants,
respectively. Secondly, the corresponding operation within
a month of high wind resource (July) is given in Figs. 9 and 10.
Considering Figs. 7 and 8, the GC design is able to operate more
than 80% of the month, although it produces a 3.3% less water than
250
2.000
225
1.875
200
1.750
175
1.625
150
1.500
125
1.375
100
1.250
75
1.125
50
1.000
25
875
0
-25 1
750
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
625
-50
Wind machine
500
-75
SWRO plant
375
-100
B alance (kW)
250
-125
B attery cap. (A h)
125
-150
Da ys
Fig. 13. Energy balance of a Vestas 27 wind turbine coupled to a 1000 m3/d gradual capacity seawater reverse osmosis design (July simulation).
Average hourly energy (kW )
E n erg y b alan ce (Ju ly)
the FC design. The FC design produces 18488.3 m3 in 456 h of

operation. It represents an average production of 660.3 m3/d,
opposite to 638.6 m3/d of GC design.
Besides that, regarding results obtained for July (Fig. 9), the
differences between both designs are minor. The FC design is able
to work a total of 715 h while the GC design achieves 97.7% of the
month. The average daily productions are similar, 957.4 m3/d for
the FC design, and 954 m3/d to the GC design, i.e. both designs are
working at the nominal point without any interruption over 96% of
the month. It should be taken into account that months with high
and constant wind resources generate a surplus energy that denes
the choice of the appropriate WT.
Finally, Figs. 10e13 depict the energy balances from the representative months previously considered, February and July.
These graphs show how the variations in the energy produced by
V27 WT cause batteries loading and unloading in periods between 1
and 5 h. This transfer of energy allows that the system could manage
the response times of the RO plant operation in processes of energy
decit. In the case of GC design the system is kept running even
during hours of low or null wind resource. These graphs show how
the design of the SWRO plant with the ability to operate at variable
ow and pressure is able to adapt quickly the energy required to the
energy available for any raised designs. Besides that, in periods of
wind resource shortage (Figs. 12 and 13) the ESS plays his role as
regulator of the system. Even, the ESS allows for adapting the RO
operation point e from nominal to minimum operation point e
without needing to stop working. In this sense, the GC design would
increase the useful life of the ESS by permitting the consumption of
less battery life cycles. As said above, during the high wind resource
months the two layouts behave in the same way.
5. Summary of results
This section focuses on the energy efcient design and comparison of a high innovation 1000 m3/d gradual and xed capacity
SWRO plants powered by an off-grid wind energy system. In the rst
case the desalination plant is more versatile because it is able to
operate racks independently or jointly, depending on the amount of
energy available. In the second case, the plant could only operate
within a more restrictive range of power consumption when wind
resource is enough to supply all the energy required by the RO plant.
Three commercial WTs close to the calculated RO plant power
and energy ranges are selected for the comparative analysis (100,
225 and 300 kW). The simulation is performed with wind data of
Pozo Izquierdo location (Gran Canaria Island).
Fixed capacity plant allows the production of a greater amount of
water per year in comparison with gradual capacity design, although
the desalination unit does not operate more hours during the year.
The gradual capacity design produce between 2 and 8% water less
than xed capacity design. However, the annual operation rates are
higher with the gradual capacity plant and the excess energy is lower.
Both designs reach rates above 95% production hours in months of
high wind resource. Besides that, in months with low wind resources,
the xed capacity design produces more water compared with the
gradual capacity design. In addition, gradual capacity systems should
be designed by using synchronous WTs with power control, capable
to quickly adapt to uctuations in demand by adjustment in power
production. Besides, a good dimensioning of the ESS (ywheel,
batteries) that allows storing or transferring the energy during the
transitional periods, as well as storing the minimum amount of
energy consumed for the boot process.
In the case of an off-grid system the power control component is
more important than the priority of producing the maximum
volume of water. This is because the power control technology of the
energy generation systems still requires the contribution of demand
4383
control systems for good tuning between variable renewable energy

power generation and the power demand. Given this approach,
gradual capacity designs are a major contribution to the design and
optimization of desalination powered by renewable energy sources.
Currently, the development of wind generation technology allows
having generation systems with primary power control and
supplementary services that make possible coupling to desalination
technology without power control, i.e. xed capacity RO plant.
Besides, this contribution allows to achieve reductions in the
ESS capacity (batteries, where appropriate). With an adequate
estimation of the wind resource (using prediction systems) and the
control of the desalination plant power demand, the requirements
of ESS can be minimised. Only the need of providing power in
unintended electrical cuts and maintain electrical parameters on
the margins of quality assurance of supply to the load.
The power available from the renewable energy resource dene
the performance of the system, due to the improvements in the RO
system are normally associated with reductions in the wind
generation system performance. This means that it should be
studied the xed capacity desalination plant production (and
performance) enhancements regarding the variable capacity. In
parallel, the losses in the performance of the energy generation and
the need to increase the ESS should be thoroughly studied.
For a system similar to that designed in this paper, but connected to a weak electrical grid, whose target is to supply water to
a site with minimal impact on the grid, the priority criteria should
be water production and the whole system performance. In this
case, it would be necessary to simulate different variable capacity
operation RO plant points and check that the variable performance
of the system from the optimal one means a capacity reduction.
However, the development of an adequate synchronous WT
power control means that the power loss to the maximum that can
be generated, despite the improvements that it entails for the
control of the combination SWROeWECS. This loss must be
determined in each case due to the dependence of the performance
of the WT and the performance of the electronic devices installed.
6. Conclusions
Desalination driven by renewable energies is an attractive chance
for fresh water production in arid regions. Only mature and efcient
technologies are suitable for medium to large scale desalination.
The design of two possible combinations useful for renewable
energy-powered SeaWater Reverse Osmosis desalination plant
(SWRO) are addressed in this paper. An energy efcient design and
comparison of a high innovative 1000 m3/d Gradual Capacity (GC)
and Fixed Capacity (FC) SWRO plants powered by an off-grid wind
energy system is performed.
In the rst case, the plant is more versatile because it is able to
operate trains independently or jointly, depending on the amount of
energy available. In the second case, the plant can operate within
a more restrictive range of power consumption when wind resource
is enough to supply all the energy required by the RO plant.
FC plant allows the production of a greater amount of water per
year in comparison with gradual capacity design, but the desalination unit does not work more hours during the year. The GC
design produces between 2 and 8% water production fewer than FC
design. However, the annual operation rates are higher with the GC
plant and the excess energy is lower. Both designs reach rates above
95% production hours in months of high wind resource. In months
with low wind resource, the FC design produces more water
compared with the GC design.
In the case of an off-grid system, the power control component
becomes a priority on producing the maximum volume of water.
The power control technology of the energy renewable generation
4384
systems still requires the contribution of demand control systems

for good tuning between variable renewable energy power generation and the power demand. So that, the GC is a major recommendation for the design and optimization of the SWRO
desalination powered by wind energy.
In general, the designs proposed require of an itemised study of
the annual local water needs and the wind availability with the aim to
ensure the optimized operation and stability in the functioning of the
system. The use of wind energy to supply a stand-alone SWRO system,
adapted to medium or large water demands, obliges to take into
account the water storage as the key energy storage system. This
solution warranties to meet the water demand in periods of lack of
wind resource or serious breakdowns of the system.
Table A1 (continued).
Energy generated by
a wind turbine (W)
Energy generated: Mean power

multiply by the period.
Acknowledgements
The authors wish to thank the Spanish PROFIT Programme for
their nancial assistance with DEREDES project e Desarrollo de la
desalacin con energas renovables (FIT-310200-2006/2007-175)
and the European Commission for their nancial assistance with
the POWERSOL project e Mechanical power generation based on
solar heat engines e (FP6-INCO2004-MPC3-032344) within the
International Cooperation Activities Programme.
Energy generated
by a wind farm (W)
Table A1
Summary of equations.
Vertical wind variation

(Mikhail & Justus)
Ej Energy produced by a wind

turbine j and period t
Ej P j $t
3f wind turbine reliability
4estji Correction factor of the wind
turbines j in each area i.
References
K1 h v i
K
K
v
$e C
f v; K; C
C
C
K shape parameter
C scale parameter, m/s
n wind speed, m/s
a
Hh
vh va
Ha
a vertical exponent.

1
0:0881
va
ln
:a

Ha
6
Hg
1 0:0881ln 10
ln
H0
Hg geometric mean elevation between

two vertical elevations (a, h);
Hg(Ha$Hh)1/2, m
H0 Rough length, m
va mean wind speed (m/s) at elevation Ha
vh mean wind speed (m/s) at elevation Hh.
Wind energy (W)
Total energy produced by several wind

generators connected in an specic area:

Pi NAdif PJ NGdif
ET
NGj Ej 4estji 3f
i1
j1
NAdif No. of different areas (i) envolved
in the wind farm
NGdif No. of wind turbines (model j)
Appendix A. Theoretical fundaments used for dimensioning

the wind energy systems.
Probability density
function of the general
Weibull distribution
Mean power generated:

RN
P 0 Pv$f vdv
P(V) wind generator characteristic curve
(two specic parts of the curve):
P
Pi
v vi Pi
Pv i1
vi1 vi
f(v) wind speed distribution in the site
Weibull distribution law:

PN1 R vi1 K v
K1 Cv K
P
$e
i1 v
C C
#

Pi1 Pi
v vi Pi dv
vi1 vi
Pi
1 2
1
rV AVi rAVi3
2 i
2
Vi wind speed at i time (m/s)

A area, m2
r air density (kg/m3)
Wind power density (W/m2):
Pi
1
rVi3 W=m2
2
A
Wind power at different intervals:
R t2 1
Et1t2 t1 rVi3 dt
2
V3 following Weibull law

R
3
N
< V 3 > 0 v3 f vdv c3 G 1
K
[1] Subiela VJ, Carta JA, Gonzlez J. The SDAWES project: lessons learnt from an
innovative project. Desalination 2004;168:39e47.
[2] Liu C. Experiments of a prototype wind-driven RO desalination system with
feedback control. Desalination 2002;150:277e87.
[3] Miranda M, Ineld D. A wind powered seawater RO system without batteries.
Desalination 2002;153:9e16.
[4] Garca-Rodrguez L. Desalination by wind power. Wind Engineering 2004;28:
453e66.
[5] Romero-Ternero V, Garca-Rodrguez L, Gmez-Camacho C. Thermoeconomic
analysis of wind powered seawater reverse osmosis desalination in the
Canary Islands. Desalination 2005;186:291e8.
[6] Lior N. Sustainable energy development: the present (2009) situation and
possible paths to the future. Energy 2010;35:3976e94.
[7] Clark C, Liu K, Jae-Woo P, Reef M, Gang Q. Experiments of a prototype winddriven reverse osmosis desalination system with feedback control. Desalination 2002;150:277e87.
[8] Witte T, Siegfriedsen S, El-Allawy M. WindDeSalter technology direct use of
wind energy for seawater desalination by vapour compression or reverse
osmosis. Desalination 2003;156:275e9.
[9] Subiela VJ, de la Fuente JA, Piernavieja G, Peate B. Canary Islands Institute of
Technology (ITC) experiences in desalination with renewable energies
(1996e2008). Desalination and Water Treatment 2009;7:220e35.
[10] de la Nuez I, Garcia FJ, Argudo C, Gmez A. Optimization of RO desalination
systems powered by renewable energies. part I: wind energy. Desalination
2004;160:293e9.
[11] Carta JA, Gonzlez J, Gmez C. Operating results of a wind-diesel system
which supplies the full energy needs of an isolated village community in the
Canary Islands. Solar Energy 2003;74:53e63.
[12] Peate B, Garca-Rodrguez L. Energy optimisation of existing SWRO
(seawater reverse osmosis) plants with ERT (energy recovery turbines):
technical and thermoeconomic assessment. Energy 2011;36:613e26.
[13] Peate B, de la Fuente JA, Barreto M. Operation of the RO kinetic energy recovery
system: description and real experiences. Desalination 2010;252:179e85.
[14] Filmtec. RO design software Rosa v.7.01 (dow water solutions company); July e
August 2009.
[15] Barreto M. Personal communication; September 2009. Web references
[16] Windpower,
wind
turbines
and
windfarms
databasehttp://www.
thewindpower.net/21-windpower-data.php; January 2009.

Dahioui and Loudiy 2013 WIND POWERED DESALINATION PAPER

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Dahioui and Loudiy 2013 WIND POWERED DESALINATION PAPER

Caricato da

Copyright:

Formati disponibili

Wind Powered Water Desalination

Al Akhawayn University in Ifrane

energy powered desalination, has been going through an

II. DESALINATION TECHNOLOGIES

To tackle this water shortage issue, water desalination has

1 Vapor compression cycle

1) Electrodialysis: Electrodialysis (ED) is the oldest

III. W[ND POWERED DESALlNAT[ON

Fig. 2 Electro dialysis process

2) Reverse Osmosis: The movement of solvent

Other factors could be taken into account as well such as

Fig. 3 Reverse osmosis process

Reverse Osmosis has been continuously improved over the

performance, their resistance to pressure fluctuations.

under specific conditions; any extreme variation is expected to

Since, there has not been much major research regarding

5 Wind turbine MATLAB model

Fig. 4 RO membrane operating region under specific conditions

1) System with Backup: Basically, an additional energy

The wind turbine model has been designed in a way to

Cp: Power coefficient (given a value of 0.4)

2) Independent System (No Backup): To keep the

Switching ON/OFF modules: this requires the

Fig. 6 Wind Powered Desalination System Model

This model (Fig.6) calculates, depending on the power

illustrate some of the effects of variable wind conditions on

For the flow rate calculation we have used [17]:

With P representing the electrical power ( in kW), H the head

QF being the feed flow rate calculated in the previous

Fig. 8 De-rating impact on pressure variation

By reducing the maximum power produced by the wind

Parts a/the system

1) Variable Speed Drive: With an energy source as

Fig. 7 Wind speed impact on pressure and flow rate

The remedy for this situation could be de-rating the wind

Fig. 9 Variable Speed Drive

2) Pressure Stabilizer: Since, pressure seems to be the

able to produce fresh water through the reverse osmosis

Concerning, the operational costs, besides maintenance, an

Fig.lO Pressure-Controlled Valves

3) Energy Recovery Device: One positive aspect of the

Low Pressure Feedworer

By using a low wind speed, and a low water production, the

Wind Speed (m.sl)

Fig. 12 Water production cost per wind speed

Most ERDs incorporate positive displacement technology and

Eventually, production cost could be much lower in windy

However, it is important to mention that this system stays

[19] C.K. Liu,

1. Park, M. Reef: Q. Gang. (2002). Experiments of a prototype

wind-driven reverse osmosis desalination system with feedback control.

1. Kaufler. (2006). Wind & Solar Powered Seawater Desalination.

[21] E.Kondili, K.Kaldellis. (nd). Wind Energy Based Desalination Process

[22] Energy Recovery Inc. (2011). The Economics of Downtime. [Online].

c.Generaal, "Wind Driven Reverse Osmosis Desalination for Small

E.Spang, "The potential for wind-powered desalination in water scarce

Methodologies. The sustainability Consortium. [Online]. Available:

1. Chapman, D. Noll, A. Jackson, B. Ewing,

L.Song, G.Ganesan, R. Marshall, V. Doon, P. Nash. (2009). Wind

Technologies. [Online]. Available: http://prod.sandia.gov.

Tata. (2009). Dessalement de I'eau de mer : bilan des demieres avancees

Department of Agriculture, Fisheries

& Forestry - Australia. (2002).

H. Strathmann. (nd). Assessment of Electrodialysis Water Desalination