Energy Conversion and Management 51 (2010) 2245–2251

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier .com/ locate /enconman

Sustainable desalination using solar energy

Veera Gnaneswar Gude *, Nagamany NirmalakhandanNew Mexico State University, Las Cruces, NM 88003, USA

a r t i c l e i n f o

Article history:Received 4 August 2009Accepted 21 March 2010Available online 14 April 2010

Keywords:DesalinationSolar energySustainable desalinationProcess modelPrototype systemPhotovoltaics

0196-8904/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.enconman.2010.03.019

* Corresponding author. Tel.: +1 530 751 6061.E-mail address: john_us@nmsu.edu (V.G. Gude).

a b s t r a c t

Global potable water demand is expected to grow, particularly in areas where freshwater supplies arelimited. Production and supply of potable water requires significant amounts of energy, which is cur-rently being derived from nonrenewable fossil fuels. Since energy production from fossil fuels alsorequires water, current practice of potable water supply powered by fossil fuel derived energy is not asustainable approach. In this paper, a sustainable phase-change desalination process is presented thatis driven solely by solar energy without any reliance on grid power. This process exploits natural gravityand barometric pressure head to maintain near vacuum conditions in an evaporation chamber. Becauseof the vacuum conditions, evaporation occurs at near ambient temperature, with minimal thermal energyinput for phase change. This configuration enables the process to be driven by low-grade heat sourcessuch as solar energy or waste heat streams. Results of theoretical analysis and prototype scale experi-mental studies conducted to evaluate and demonstrate the feasibility of operating the process using solarenergy are presented. Predictions made by the theoretical model correlated well with measured perfor-mance data with r2 > 0.94. Test results showed that, using direct solar energy alone, the system could pro-duce up to 7.5 L/day of freshwater per m2 of evaporator area. With the addition of a photovoltaic panelarea of 6 m2, the system could produce up to 12 L/day of freshwater per m2 of evaporator area, at efficien-cies ranging from 65% to 90%. Average specific energy need of this process is 2930 kJ/kg of freshwater, allof which can be derived from solar energy, making it a sustainable and clean process.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Increasing demand for potable water due to population growthand rapid development is a major concern nationally and globally.According to the Energy Information Administration (EIA) projec-tions, the US population is expected to grow by about 70 millionby 2030 [1]. The direct domestic water demand and the indirectindustrial, agricultural, and environmental water needs to sustainthis growth is expected to place serious strains on the currentlyavailable water resources. At the same time, this growth in popula-tion is expected to increase the electricity demand by approximately50% [1], which will place additional demands on available water. Forexample, in 2000, thermoelectric power plants accounted for 48% ofthe total water withdrawal in the US; consumptive use of water forelectricity production could more than double from 3.3 billion gal-lons/day in 1995 to 7.3 billion gallons/day in 2030 [2]. Although thisconsumptive use is not high compared to the total US consumptionof 100 billion gallons/day, large volumes of water are to be dedi-cated to thermoelectric power plant operation.

Future demand for potable water will be much higher in theglobal context. According to the World Health Organization, nearly

ll rights reserved.

2.8 billion people (�40% of the world population) currently haveno access to safe drinking water and, water-borne diseases accountfor 90% of all infectious diseases in the developing world. TheWorld Resources Institute predicts that by 2025, at least 3.5 billionpeople will experience water shortages. Global agencies (includingWHO, UNDP, UNICEF, etc.) expect that 24 of the least developedcountries, many of them along coastal areas without access tofreshwater and electricity, need to more than double their effortsto reach the Millennium Development Goals (MDGs) for basichealth, sanitation, and welfare.

Provision of clean water inevitably requires energy, which iscurrently being provided essentially by nonrenewable fossil fuels.Total energy demand for providing the US water needs is reportedas 123 � 106 MW h/year [3]. It has been estimated that productionof 1 m3 of potable water requires the equivalent of about 0.03 tonsof oil [4]. Extraction and refining of fossil fuels and production ofenergy not only places additional demands on water, but also re-sult in pollution of water sources and air. Thus, the projected globaldemand for clean water supply for the future will significantlyaccelerate not only depletion of fossil fuel reserves but also theconcomitant environmental damage and emission of greenhousegases.

One of the solutions to this dilemma is to develop sustainableapproaches that can utilize renewable water and energy sources

Nomenclature

A surface area, m2

cp specific heat, kJ/kg KC concentration of solute, kg/kghL(T) latent heat at temperature T, kJ/kgI(t) solar insolation as a function of time, kJ/h m2

m mass flow rate, kg/hM total daily mass of distillate, kgQ heat flow rate, kJ/ht time, hT temperature, �CV volume, m3

Greek symbolsa absorptivity (–)j experimental constant (10�7 � 10�6 kg/m2 Pa s K0.5)

q mass density, kg/m3

s transmissivity (–)g efficiency (–)

SubscriptsBB battery bankEC evaporation chamberg glassl lossesPV photovoltaic panel

2246 V.G. Gude, N. Nirmalakhandan / Energy Conversion and Management 51 (2010) 2245–2251

without consuming any nonrenewable resources (fossil fuels andwater) and without causing any environmental harm. Even thoughwater is one of the most abundant resources covering three-fourths of the planet’s surface, about 97% of this volume is saline,and only 3% is fresh water suitable for humans, plants, and ani-mals. The amount of water in the oceans, however, can serve asan inexhaustible and equitable source for the planet’s freshwaterneeds, if sustainable and cleaner technologies can be developedfor desalination.

While a range of mature technologies are available for desalina-tion, most of them are cost-prohibitive, energy-intensive, and fossilfuel-dependent. Table 1 summarizes the energy requirements andgreenhouse gas emissions associated with currently available tech-nologies. With increasing costs and uncertainties of fossil fuel sup-plies and the environmental impacts associated with energyproduction, currently available desalination technologies are notreliable, affordable, and sustainable solutions for meeting futurewater needs, particularly for low-income rural and remotecommunities.

Use of renewable energy sources (such as wind, solar, geother-mal) to drive desalination processes can be a sustainable andaffordable approach to reclaim potable water from seawater andbrackish waters. Solar energy, in particular, has been identified asa convenient renewable energy source for this application, becauseit is more widely available and can be stored in batteries via pho-tovoltaic (PV) arrays, and converted to heat or mechanical energywith reasonable efficiency. Although solar energy is ‘‘free” thehardware necessary for capturing it, converting it to useful forms,and storing it can add significantly to the cost. Additional costs willincur depending on the type of desalination technology that isused. Economic factors are the main barriers to the use of solar en-ergy for desalination. However, for rural and remote applications,where grid power or fossil fuels to generate energy may not beavailable at affordable costs, solar energy-driven desalinationmay be economically attractive. Thermal desalination technologies

Table 1Comparison of proposed process with traditional desalination processes.

MSF MED

Specific energy (kJ/kg) Thermal 294 123Mechanical 44 26Total 338 149

CO2 emissions (kg CO2/kg H2O) 0.09 0.04

MSF – multi-stage flash distillation; MED – multi-effect distillation; MVC – mechanical vphotovoltaic, this study [4].

require large quantities of energy. Traditionally, fossil fuels havebeen used to provide the energy requirements for desalination ofseawater or brackish waters. In an effort to conserve fossil fuel re-sources, desalination industry has been adopting several energy-saving measures in recent years. Examples include recovery andrecycling of energy as in the case of staging, low temperature desa-lination, and utilization of waste heat or renewable energy.

In this study, a new low temperature desalination process hasbeen developed which can utilize low-grade heat sources such aswaste heat releases or solar energy. Since the process operates atlower temperatures than traditional thermal desalination pro-cesses, energy losses and hence the net energy requirements forthis process are lower and its thermodynamic efficiency is higher[5]. As this process utilizes waste heat releases and renewable en-ergy, it does not contribute directly to any greenhouse gas emis-sions, and can be considered a sustainable process.

2. Proposed system

The premise of the proposed process can be illustrated by con-sidering two barometric columns at ambient temperature, onewith freshwater and one with saline water. The headspace of thesetwo columns would be filled by the vapors of the respective fluidsat their respective saturated vapor pressures. Suppose these head-spaces are connected to one another. Since the vapor pressure offreshwater is slightly higher than that of saline water at ambienttemperature, water vapor will distill from the freshwater columninto the saline water column.

However, if the temperature of the saline water column ismaintained slightly higher than that of the fresh water column toraise the vapor pressure of the saline water side above that ofthe fresh water side, water vapor from the saline water column willdistill into the fresh water column. A temperature differential ofabout 10–15 �C is adequate to overcome the vapor pressure differ-ential to drive this distillation process. Such low temperature dif-

MVC RO ED SS PV

0 0 0 0 0192 120 144 3.6 0192 120 144 3.6 00.051 0.032 0.38 �0 0

apor compression; RO – reverse osmosis; ED – electrodialysis; SS – solar still; PV –

V.G. Gude, N. Nirmalakhandan / Energy Conversion and Management 51 (2010) 2245–2251 2247

ferentials can be achieved using low-grade heat sources such as so-lar energy or process waste heat.

2.1. Process configuration

A schematic of the process configuration built on the aboveprinciples is shown in Fig. 1. The major components of the desali-nation unit are an evaporation chamber (EC), a condenser (CO), aheat exchanger HE, and three 10-m tall columns. These three col-umns serve as the saline water column; the brine withdrawal col-umn; and the desalinated water column, each with its own holdingtank, SWT, BT, and DWT, respectively. The EC is installed atop thethree columns at a height of about 10 m above ground level to cre-ate vacuum naturally in the headspaces of the feed, withdrawal,and desalinated water columns. When heat is supplied to the EC,water evaporates and condenses into the desalinated water col-umn. At the same time, a set amount of brine is drained from theEC to maintain the salt level. As the water level in the EC drops,atmospheric pressure drives the saline water into the EC to main-tain the set level. In this manner, the desalination process contin-ues without any mechanical pumping [6,7]. A heat exchanger (1–2) is used to recover some of the heat carried away by the brineto preheat the saline water feed.

With this configuration, saline water can be desalinated atabout 40–50 �C, which is lower than the 60–100 �C range in tradi-tional solar stills and other distillation processes. In a previousstudy, we have illustrated the use of heat rejected by an absorptionrefrigeration system (ARS) in driving this process [8]. In that studyit was shown that the heat rejected by an ARS of cooling capacity of3.25 kW (0.975 tons of refrigeration) along with an additional en-ergy input of 208 kJ/kg of desalinated water was adequate to pro-duce desalinated water at an average rate of 4.5 kg/h. In this paper,we present two configurations for utilizing solar energy: (A) a lowcost system using direct solar energy in a solar evaporation cham-ber (SEC) and (B) a high efficiency system using solar photovoltaicpanels. In configuration A, the EC doubles as a solar still to harvestsolar energy during sunlight hours to provide the heat for phasechange in the EC. In the configuration B, additional PV panels har-vest solar energy during sunlight hours to charge a battery bank viaa charge controller, which in turn, powers a DC heater to heat theEC to maintain it at the set temperature throughout the day.

Theoretical modeling and prototype scale demonstration of thetwo configurations are presented in the following sections. De-

Reflector PV panel

CO

3 4

2 5 Heater

Saline Auxiliarywater 6 powerinlet 1

Saline Desalinatedwater tank water tank

EC

tankBrine

HE~ 10 m

Charge controller/

Battery bank

Fig. 1. Schematic of the proposed process configuration.

tailed theoretical analyses of different heat addition configurationsand model simulations have been presented elsewhere [9]. Here,we present limited data on configurations A and B to validate theprocess model under dynamic conditions.

2.2. Theoretical modeling

Mass and heat balances around the evaporation chamber (EC)yield the following coupled differential equations, where the sub-scripts refer to the state points shown in Fig. 1. The variables aredefined in the Nomenclature.

Mass balance on volume of water in EC:

ddtðqVÞEC ¼ m2 �m5 �m3 ð1Þ

Mass balance on solute in EC:

ddtðqVCÞEC ¼ m2C2 �m5C5 ð2Þ

Heat balance for volume of water in EC:

ddtðqVcpTÞEC ¼ Q in þ ðmcpTÞ2 � ðmcpTÞ5 �m3hLðTÞ � Q l ð3Þ

where Qin is the rate of heat input to the EC and Ql is the rate of heatloss from the EC. In configuration A, the heat input is via direct solarinsolation on the EC (=I(t)ASE awsg). In configuration B, additionalheat, QBB, is input by the battery bank, only when the evaporationtemperature T3 falls below the set temperature, Tset. The heat inputrate by the battery bank, QBB is assumed to be a constant (whenT3 < Tset); while the charging rate of the batteries is equal to I(-t)APV gPV (when T3 > Tset). The excess energy is stored in the batterybank. Hence,

Qin ¼ IðtÞASECawsg þ Q BB ð4Þ

Desalination efficiency is defined as

g ¼ MhLðTÞ

RðQ inDtÞ ð5Þ

where

hLðTÞ ¼ 3146� 2:36ðT þ 273Þ ð6Þ

Evaporation rate as a function of pressure and temperature[10]:

m3 ¼ AECj fðCECÞpðTECÞ

ðTEC þ 273Þ0:5�

pðT3Þ

ðT3 þ 273Þ0:5

" #ð7Þ

where

pðTÞ ¼ ½expð63:02� 7139:6=ðT þ 273Þ� 6:2558 lnðT þ 273Þ� � 102 Pa ð8Þ

The above equations are solved using Extend (ImagineThat Inc.)simulation software.

3. Results

3.1. Configuration A: using direct solar energy

Initially, configuration A was simulated with the followingparameters: solar energy incident on evaporation chamber (SEC)area of 1 m2; water depth in the EC of 0.05 m; and the referencetemperature of 25 �C. Based on model simulations, configurationA could produce up to 5.25 L/day of freshwater, which is more thantwice the productivity of a flat basin solar still of comparable areaunder comparable solar insolation. This advantage over the solarstill is due to the lower evaporation temperature whereby signifi-

2248 V.G. Gude, N. Nirmalakhandan / Energy Conversion and Management 51 (2010) 2245–2251

cant energy need for the sensible heat has been averted. Desalina-tion efficiencies of 60–70% could be achieved by this configuration.

The experimental prototype system had an evaporation cham-ber area of 0.2 m2. Since the evaporation rate has a factor of areaas expressed in Eq. (7) at given temperatures and pressures, theexperimental results from this system are extrapolated to evapora-tion area of 1 m2 to enable comparisons between the experimentalresults and theoretical simulations. Experimental data from a typ-ical run starting from a ‘‘cold start” are shown here to demonstratethe adequacy of the model presented earlier. Fig. 2 compares thetemperature of the EC predicted by the model against the mea-sured temperature and the ambient temperature in configurationA. During this test, the solar insolation reached a peak of 1150 kJ/h m2 over the 8-h photoperiod. The maximum ambient tempera-ture recorded was 36 �C and the maximum temperature of theEC was 52.75 �C. The predicted maximum temperature was 52 �C.As shown in Fig. 2, EC temperature declined after the sunlight per-iod, and approached ambient temperature after sunset. The corre-lation between the predicted and measured EC temperature wassatisfactory with r2 = 0.943, F = 2358.2, p < 0.001.

The predicted distillate volume during the above test is com-pared against the measured distillate volume in Fig. 3. Cumulativevolume predicted by the model for a 24-h period was 5.25 L/day m2

while the measured value was 4.95 L/day m2. The difference (of5.5%) in the cumulative distillate volume is mainly due to theassumption that the entire volume of the distilled vapor condensedon the freshwater side whereas, during the test it was observed thatsome of the vapor condensed on the roof of the evaporator andtrickled back to the evaporation chamber. Correlation between

0

10

20

30

40

50

60

0 200 400 600 800 1000 1200 1400Time [min]

Temperature profiles [oC]:1- Evaporation chamber, measured2- Evaporation chamber, predicted3- Ambient

1

2

3

Fig. 2. Typical temperature profiles in configuration A over 1-day period.

0

1

2

3

4

5

6

0 200 400 600 800 1000 1200 1400

Time [min]

Daily distllate production [L]:1- Measured2- Predicted

2

1

Fig. 3. Daily distillate production in configuration A: measured vs. predicted.

the predicted and measured distillate volume as a function of timewas strong with r2 = 0.988, F = 11,839.4, p < 0.001. The process effi-ciency as a function of time predicted by the model is compared inFig. 4 against the efficiency calculated using the measured distillatevolume from Eq. (5). The predicted efficiency averaged 64% whilethe observed efficiency averaged 61% over this test period. Correla-tion between the predicted and measured efficiency was strongwith r2 = 0.985, F = 538.7, p < 0.001.

The above results demonstrate the feasibility of the proposedconcept in maintaining the near vacuum pressure in the EC andmaintaining continuous flow of the fluids without any mechanicalenergy input. However, the yield and the efficiency of this systemdeclined when there was no incident solar energy. The perfor-mance of configuration A was improved slightly when a reflectorwas installed to increase the incident energy. Fig. 5 compares thedistillate production rate as a function of time with and withouta reflector. During the tests with the reflector, average cumulativeproduction of 7.5 L/day m2 was obtained over 24 h. This output isequivalent to three times the distillate produced by normal solarstill. The increased production is due to higher energy input as wellas due to the slightly longer period of production because of thesensible heat stored in the EC.

3.2. Configuration B: using PV panel and direct solar energy

The limitation of configuration A was overcome by using a PVpanel/battery bank to heat the EC during non-sunlight hours. Inour experiments, a standard PV panel area of 6 m2 rated at

0

20

40

60

80

100

0 100 200 300 400 500 600 700

Time [min]

Distillation efficiency [%]:1- Measured2- Predicted

12

Fig. 4. Distillation efficiency in configuration A: measured vs. predicted.

0

20

40

60

80

100

0

200

400

600

800

1000

1200

1400

1600

0 200 400 600 800

Time [min]

Distillate production [mL]-Every 15 minutes:

1- without reflector2- with reflector

Cumulative:3- without reflector4- with reflector

1

2

3

4

Fig. 5. Distillate production with and without a reflector in configuration A.

0

10

20

30

40

50

60

0 200 400 600 800 1000 1200 1400Time [min]

Temperature profiles [oC]:1- Evaporation chamber, measured2- Ambient

1

2

Fig. 7. Temperature profiles in configuration B with PV/battery.

V.G. Gude, N. Nirmalakhandan / Energy Conversion and Management 51 (2010) 2245–2251 2249

185 W (Sharp NT-S5E1U) was used to charge a 12-V battery bankwhich provided power to a thermostatically controlled 12-V DCheating coil installed in the EC. The efficiency of the PV modulesis 14%. Even though this configuration could be driven round theclock by a thermal energy storage system backed by solar collec-tors, the approach described above was used in this study for easeof control and measurements.

The energy flows during a typical test under this configurationare shown in Fig. 6 the incident solar insolation in Fig. 6a; the en-ergy produced by the PV panel in Fig. 6b; the energy flow to/fromthe batteries in Fig. 6c; and the energy provided to the EC in Fig. 6d.The temperature profiles during a typical test under this configura-tion are shown in Fig. 7. Photovoltaic energy generated during theday was sufficient to produce freshwater of 4–5 L/day m2 duringnon-sunlight hours. Specific energy required for this process toproduce 1 kg of freshwater was 2926 kJ. Freshwater productionrates up to 10 L/day m2 have been obtained from this configurationover 24 h, by maintaining the evaporation temperature nearly con-stant at the set value throughout the 24-h period as shown inFig. 7. Comparing the temperature profile for configuration A in

200

400

600

800

1.000

1.200(a) Incident soalr energy [W/m2]

-200

-100

0

100

200

300

400(c) Energy to/from battery bank [W]

-200

-100

0

100

200

300

400

12:00 AM 06:00 AM 12:00

(d) Energy to evaporation chamber [

-200

-100

0

100

200

300

400(b) Energy from PV panel [W]

To battery

From battery

Fig. 6. Energy flows in configuration

Fig. 2 with that for configuration B in Fig. 7, the benefit of addingthe PV/battery system is obvious. However, it has to be noted thatthe performance of configuration B is limited in this case by the

PM 06:00 PM 12:00 AM

W]

B over a typical 1-day period.

17,137

11,300

0

5,000

10,000

15,000

20,000

12:00 AM 06:00 AM 12:00 PM 06:00 PM 12:00 AM

Cumulative power [kJ/day]: 1- produced by PV panel 2- consumed by desalination process

Excess energy

1

2

Fig. 8. Energy produced, energy consumed, and excess energy.

2250 V.G. Gude, N. Nirmalakhandan / Energy Conversion and Management 51 (2010) 2245–2251

evaporation area rather than the PV panel area. In fact, the PV pa-nel used in this study was oversized, and was able to provide moreenergy than what is required for evaporation as shown in Fig. 8.

Fig. 9 shows the freshwater production as a function of time inconfiguration B. Comparing the production in configuration A(Fig. 5) with that in configuration B (Fig. 9), the benefit of addingthe PV/battery system in extending the desalination period isobvious.

Grid energy requirements for the proposed process are com-pared with the following commonly used desalination processesin Table 1 [11]: multi-stage flash distillation (MSF); multi-effectdistillation (MED); mechanical vapor compression (MVC); reverseosmosis (RO); and electrodialysis (ED). The process developed inthis study eliminates green house gas emissions by using solar en-ergy for heating as well as fluid transfer while the other processesconsume nonrenewable energy sources for providing the thermaland mechanical energy requirements with green house gas emis-sions contributing to global warming. In this comparison, 30% pro-duction efficiency for the production of electricity from fossil fuelsis considered. The carbon dioxide emissions in Table 1 are esti-mated based on the assumption that 1 kW h electricity productionresults in 0.96 kg of CO2 emissions [12].

4. Conclusions

Round the clock operation of a low temperature desalinationsystem powered by both solar and photovoltaic modules has beendemonstrated. The experimental results prove that the proposedsystem has the ability to produce two times of the distillate that

0

20

40

60

80

100

0

1000

2000

3000

4000

5000

0 200 400 600 800 1000 1200 1400

Time [min]

Distillate production [mL]-1- Every 15 minutes2- Cumulative

1

2

Fig. 9. Distillate production in configuration B with PV/battery.

can be derived from a simple solar still configuration and threetimes the distillate of a solar still when fitted with a reflector. Elec-trical energy generated by the photovoltaic modules was utilizedsparingly to drive the desalination process during non-sunlighthours which favors higher freshwater production rates due to low-er ambient temperatures. The system combined with photovoltaicmodules produced 12 L/day of freshwater for an evaporator area of1 m2 and 6 m2 photovoltaic area. The energy and cost require-ments for a single stage thermal desalination process can be pro-hibitively high especially when the renewable energy sources areconsidered. However, the process economics can be recovered toa greater extent if a multi-effect configuration is considered. Envi-ronmental benefits offered by renewable energy sources as well asreliable freshwater source are the two factors that outweigh thecost factor of a renewable energy powered desalination system.These preliminary studies indicate the opportunities to considerdifferent economically feasible combinations such as thermal en-ergy storage tanks backed by solar collectors and other continuouslow-grade heat sources which can source the energy for desalina-tion process. It can be concluded from the above analysis that thissystem could therefore be most suitable for remote areas withoutelectrical grid, in other words, where the scarcity for freshwaterand electricity is highly pronounced. However, the benefit of utiliz-ing natural vacuum principle to save mechanical energy needs hasto be evaluated at a large scale to validate the process feasibility.While the proposed process has lower specific energy require-ments compared to other single stage evaporation units, its perfor-mance can be further improved by adding multi-effectconfiguration. Ability of this process to utilize renewable energysources can minimize greenhouse emissions that contribute to glo-bal warming, making this a sustainable process. Recovering wasteheat from other processes such as air-conditioning systems andpower plants to drive this process can significantly improve theoverall economies of the combined processes.

Acknowledgement

The research reported here was supported by Grants from theNew Mexico Water Resources Research Institute (WRRI).

References

[1] EIA, Energy Information Administration. Annual energy outlook 2006: withprojections to 2030. Washington, DC; 2006.

[2] Hoffmann J, Forbes S, Feeley T. Estimating freshwater needs to meet 2025electricity generating capacity forecasts. National Energy TechnologyLaboratory; June 2004.

V.G. Gude, N. Nirmalakhandan / Energy Conversion and Management 51 (2010) 2245–2251 2251

[3] EPRI, Electric Power Research Institute. Water and sustainability: US waterconsumption for power production—the next half century, No. 1006786.California, Palo Alto; 2002.

[4] Kalogirou SA. Economic analysis of a solar assisted desalination system. RenewEnergy 1997;124:351–67.

[5] Gude VG, Nirmalakhandan N. Desalination at low temperatures and lowpressures. Desalination 2009;244:239–47.

[6] Al-Kharabsheh S, Goswami DY. Analysis of an innovative water desalinationsystem using low-grade solar heat. Desalination 2003;156:323–32.

[7] Al-Kharabsheh S, Goswami DY. Experimental study of an innovative solarwater desalination system utilizing a passive vacuum technique. Sol Energy2003;75:395–401.

[8] Gude VG, Nirmalakhandan N. Desalination using low-grade heat sources. ASCEJ Energy Eng 2008;134:95–101.

[9] Gude VG. PhD Dissertation. New Mexico State University, Las Cruces, NM, USA;2007.

[10] Bemporad GA. Basic hydrodynamic aspects of a solar energy baseddesalination process. Sol Energy 1995;54:125–34.

[11] Kalogirou SA. Seawater desalination using renewable energy sources. ProgEnergy Combust Sci 2005;31:242–81.

[12] CO2 emissions per kilowatt hour of electricity production; 2009. <http://www.eia.doe.gov/cneaf/electricity/page/co2_report/co2report.html#electric2000>