Desalination 351 (2014) 236–242

Contents lists available at ScienceDirect

Desalination

j ourna l homepage: www.e lsev ie r .com/ locate /desa l

A study of the bubble column evaporator method forthermal desalination

Muhammad Shahid ⁎, Richard Mark PashleySchool of Physical, Environmental and Mathematical Sciences, University of New South Wales, Canberra, Australia

H I G H L I G H T S

• High temperatures do not affect inhibition of bubble coalescence in salt solutions.• Enhanced bubble column evaporation was observed using high inlet gas temperatures.• Surfactant monolayer coatings were found to improve the evaporation rate.• Latent heats of vaporization can be obtained even at high inlet gas temperatures.

⁎ Corresponding author. Tel.: +61 2 6268 8290.E-mail address: m.shahid@student.adfa.edu.au (M. Sh

http://dx.doi.org/10.1016/j.desal.2014.07.0140011-9164/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 13 February 2014Received in revised form 11 July 2014Accepted 12 July 2014Available online xxxx

Keywords:Bubble column evaporatorNon-ionic surfactantMonolayerSupersaturation

A simple bubble column evaporator can be used to evaporate water from concentrated salt solutions withoutboiling. The process is made more effective by the inhibition of bubble coalescence caused by the presence ofsome concentrated salts, such as NaCl. This work examines the effects of high bubble temperatures on thiscoalescence inhibition and its effects on the efficiency of water vapor collection. A continuous flow of hot dryair, at 275 °C, produced about 10% higher rate of water vaporization than that expected from the equilibriumvapor pressures. Also, the use of a non-ionic surfactant monolayer bubble coating further improved the evapora-tion efficiency, by up to 18%, apparently due to supersaturation. In addition, the steady state temperature of thebubble column evaporator can be used to estimate the latent heat of vaporization even for inlet air temperaturesof up to 275 °C.

ahid).

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Producing drinkingwater from seawater has a long history. Aristotle(384–322 BCE) commented that pure water can bemade by the evapo-ration of seawater. In ancient times, many civilizations used distillationto produce drinking water on their ships. Aristotle also carried outsome experiments on removing salt from seawater by filtration andion exchange, by flowing it through a high surface area porous materialsuch as sand and clay. Simply by digging a hole near the seashore andallowing the seawater to percolate through the sand reduce the saltlevel [1]. The availability of water has been a major influence on culturein the Middle East, for thousands of years [2]. They support habitationand agriculture, in some parts of the Middle East, until very recently,when pumped bores began to significantly deplete the water table [3,4]. The Arab states of the Persian Gulf were among the first to adoptindustrial-scale desalination, and have the largest proportion of theworld's installed desalination capacity. This is partly because of theavailability of abundant fossil fuels, proximity to the sea, and the limited

natural fresh water [5,6]. Other countries, such as Singapore [7],Australia [8] and Spain [9], also have increasing levels of investment inseawater desalination. Demand for water in Asia and the Middle Eastis expected to increase sharply in the long term, due to rapid populationgrowth and economic growth. Much of this demand is expected to bemet by desalination [10]. Predicted climate change may also have asignificant impact on the availability of conventional water supplies.Water stress is expected to increase in the future, due to populationgrowth alone [11].

The most cost and energy effective production of drinking watercomes from the collection, storage and treatment of natural rainfall[12]. However, this source of water is not always readily available nextto regions of high population and so for coastal regions two mainprocesses have been developed for seawater desalination, which arebased on the ancient methods of boiling and filtration. These methodsare called thermal desalination and reverse osmosis (or membranefiltration) desalination [12,13]. Thermal desalination methods areseverely limited by their high thermal energy demand and somulti-stage flash (MSF) and multi-effect distillation (MED) havebeen developed [12] to re-use the psychrometric (or vapor potential)energywhichmust be collected from the water vapor on condensation.

237M. Shahid, R.M. Pashley / Desalination 351 (2014) 236–242

More effective methods of obtaining this heat through the vapor com-pression process [14] have recently been developed.

In large scale, commercial thermal desalination processes, boiling isusually carried out under a reduced pressure to depress the boilingpoint. This process is very effective at producing cleanwater but it is cost-ly in terms of energy requirements and is usually only cost effectivewhencombined with an available source of waste industrial heat, for examplefrom a power station. The second process of increasing importance iscalled reverse osmosis (RO) filtration [14]. In this process an asymmetricfiltration membrane is used which contains a thin surface layer of poresso fine that only water molecules can pass through. Unfortunately, highpressures, in the range of 50 to 80 bar have to be used to force the seawa-ter through the pores at a reasonable rate and the pores easily becomeclogged and so this process is also fairly costly as care has to be taken topre-filter and clean the seawater prior to RO filtration [14]. Ion exchangeis normally used for the desalination of brackish water [15].

Thermal evaporative methods have some advantages in terms of areduced need for high quality feed water (compared with RO, wherethe membranes are readily fouled) and also the rejected salt can havea much higher concentration. In addition, evaporative methods can beused to treat heavily contaminated industrial waste water and hypersa-line feed water. The cost of desalinatedwater is considerable, comparedwith that of treated reservoir water.MSF andMEDdesalination requiresabout 150–300 MJ m−3 for stand-alone units where waste heat is notused [13] and falls to about 17–43MJ m−3 when using waste industrialheat [16]. Thesefigures should be comparedwith the enthalpy of vapor-ization ofwater, of about 2.26GJm−3 at 100 °C and 2.45GJm−3 at roomtemperature [17], which clearly demonstrates the importance ofthermal energy recovery in these evaporative methods. This enthalpyof vaporization values is not altered much by the addition of NaCl.Hence, the thermal recovery efficiency has to be about 90% or better.Even with this energy demand, the cost of water production, per cubicmeter, is modest, at typically between US$0.5 and $2 [18]. These costsare largely determined by the efficiency of the plant, and are not affectedsubstantially by feed water quality.

It is possible to obtain high quality drinking water from seawaterwithout the need for boiling using the fact that the water–air interfaceis a natural semi-permeable membrane, since it allows water vapor toescape but not dissolved ions. It turns out that this simple process canbe further enhanced using a remarkable but still unexplained propertyof salt in water which was first discovered by Russian mineral flotationengineers in the 1930s. They found that adding salt to a flotation cham-ber significantly reduced bubble size and hence improved its efficiency[19]. The formation of a bubble in water requires somework because ofthe surface tension of the water. We can see this for ourselves whenweuse our lungs to blow bubbles in water through a straw. When two airbubbles are forced together in water, they tend to coalesce to formone bigger bubble. This is what we would expect because a single bub-ble has a smaller total surface area than two small bubbles of the samevolume. However, within a bubble column containing large numbersof continuously colliding bubbles it becomes clear that salt has the abil-ity to inhibit bubble coalescence, hence allowing the formation of a highvolume fraction (N50%) column of bubbles [20,21].

A suitable high density bubble column can be produced by pumpingair continuously through a 40–100 μm pore size glass sinter (sinter size2) to produce a continuous stream of bubbles within a column filledwith water or salt solution. When using an aqueous NaCl solution ofabout 0.15 M, or more, finer bubbles are produced (of about 1–3 mmdiameter) giving an opaque column, because of the salt inhibition effect[20,21]. These bubbles rise at a limited rate of between about 15 and35 cm/s in quiescent water because they undergo oscillations in shapeand rise trajectory, which dampen their rise rate [22,23]. These oscilla-tions also increase the rate of transfer of water vapor into the bubblesand enhance the rate of water vapor collection. Equilibrium vaporpressure within the bubbles is therefore attained quite quickly, withina few tenths of a second [22] and the bubbles will therefore reach

saturated vapor pressure within a travel distance of about 5–10 cm.Larger bubbles are limited at a similar rise rate and so the use of largerbubbles has no advantage and has the disadvantage that they will takelonger to reach water vapor equilibrium, simply due to their largersize [22–24]. Smaller sized bubbles, below about 1mm, have increasing-ly slower rise rates, which will significantly reduce the efficiency of thevapor transfer process [23,24]. The relatively short travel distance foroptimum sized (1–3 mm) bubbles to rapidly reach water vapor satura-tion will have a significant influence on the design of any larger scalevapor separation process based on a bubble column evaporator.

This bubble column evaporator process can be described by Eq. (1),which is based on the energy balance within the column at steady stateequilibrium [25].

½ΔT � Cp Teð Þ� þ ΔP ¼ ρv Teð Þ � ΔHv Teð Þ in units of J=m3� �

ð1Þ

Cp(Te) is the specific heat of the gas flowing into the bubble columnat constant pressure; Te is the steady state equilibrium temperature ofthe column; ρv is the vapor density at Te; ΔT is the temperature differ-ence between the gas entering and leaving the column; and ΔP, the ad-ditional correction term, is equal to the hydrostatic differential pressurebetween the gas inlet into the sinter and atmospheric pressure at thetop of the column, which represents the work done by the gas flowinginto the base of the column until it is released from the solution.Eq. (1) describes the process by which heat is supplied fromwarm bub-bles for vaporizing water in solutions in units of Joule per unit volume.

These combined factors mean that a simple bubble column can beused to efficiently collect water vapor over a modest distance andtime period. Unfortunately, however, only very limited data is currentlyavailable on the rise rate at different bubble sizes and for higher columntemperatures. Although there is a lack of detailed information on thefundamental processes involved in the bubble column evaporator, thetechnique has recently been used in the development of several newapplications. These include sub-boiling desalination [26], measurementof the latent heat of vaporization of concentrated salt solutions [25],evaporative cooling [25] and, most recently, low temperature steriliza-tion [27]. The bubble column evaporator (BCE) method also has greatpotential for applications in the treatment of heavily contaminated in-dustrial water and high salinity water. This is because the evaporationsurfaces are continually being produced and destroyed, that is, thefresh bubbles. The buildup of scale in conventional thermal desalinationprocesses presents a major problem [28] which is almost completelyremoved by the use of the BCE process.

When no other materials are present, air bubbles in water readilycoalesce with each other and break at the surface of the water becauseof the release of their surface energy/tension. However, bubbles canbe stabilized by adding soap to the water. Soap molecules adsorb atthe bubble surface and form a monolayer film, which is often chargedand lowers the surface tension. Colliding bubbles are prevented fromcoalescing by the adsorbed soap molecules and instead form a foam.This is one of the most important and well known characteristics ofsoap solutions and is the basis of the ‘bubble persistence test’. Whenwater is vigorously shaken, the bubbles break in less than a second,unless there is soap contamination. Even a very low level of residualsoap will extend bubble lifetimes to several seconds. This is the basisof the simple and practical ‘bubble persistence test’.

The ability of monomolecular soap layers to reduce evaporationrates is also well known and used to reduce water loss in arid areas[29]. However, we also know that lipid bilayers in biological cells readily(and importantly) allow the transport of water across membranes inboth directions [29], hencewater can be transmitted along hydrocarbonchains within bilayers. In the experiments reported here we havestudied the effects of high temperature gases on water vapor collectionin a bubble column evaporator combined with a study of the additionaleffect of surfactant monolayer coatings at the bubble surface.

Fig. 1. Schematic diagram of the bubble column evaporator (BCE) desalination unit.

238 M. Shahid, R.M. Pashley / Desalination 351 (2014) 236–242

2. Materials and methods

A high surface area air/water interface was continuously producedby pumping air through a 40–100 μm pore size glass sinter into a120mmdiameter open top glass column filledwith solution to a heightof about 50mm to allow complete and efficient removal of water vapor.The apparatus used to study desalination using the bubble columnevaporator (BCE) process with a high temperature gas (air) flow isshown in Fig. 2. This system enables the use of inlet dried air tempera-tures at more than 150 °C. The inlet air temperature was varied usinga Tempco air heater with a thermocouple temperature monitor and anACVariac electrical supply. The actual temperature of the dry airflowinginto the solution was measured by a Tenmars thermometer (±1.5 °C),before solution present, just at the center of the sinter. The air wasproduced using a HIBLOW air pump through a silica gel dessicator anda BOC gas flow meter. The temperature of the column solution wasalso continuously monitored using a thermocouple positioned at the

Fig. 2. Photograph of a bubble column containing distilled water with inlet gas flow (dryair) at 275 °C.

center of the column solution. The high temperature air flow, of up to300–600 °C, needed to produce air temperatures in the range 150–275 °C just above the glass sinter, necessitated the use of steel andbrass connectors for the downstream output from the heater and theuse of FM Insulation Rock Wool as an insulating material. (See Fig. 1.)

In the first series of experiments, the hot gaswas passed through theempty, dry column and then a known weight of the solution (i.e. 200 gof 0.5 m NaCl) was quickly added. The temperature of the solution wasthenmeasured everyminute throughout the 30min bubbling runs and,in each case, the column solution quickly approached the predictedsteady state equilibrium operating temperature given by Eq. (1).After 30 min the column and remaining solution were detachedand weighed. Since the dry weight of the column was known, thetotal amount of water vapor removed in each experiment was easilymeasured.

In a second series of experiments the same procedure was repeatedbut with the addition of 0.002 g of a non-ionic surfactant (octaethylene

Fig. 3. Photograph of a bubble column containing 0.5 m NaCl solution with inlet gas flow(dry air) at 275 °C.

Table 1Desalination efficiency with inlet gas temperature.

Inlet gastemp (°C)

Weightof waterloss (g)

Calculated weightusing measuredcolumn temp (g)

Expectedtheoreticalweight (g)

% measured vaporizationcompared to the expectedequilibrium vapor pressure

150 41.7 40.8 40.4 2.2175 48.5 47.1 46.8 3.0200 56.2 54.0 53.5 4.1225 63.7 60.2 59.8 5.9250 73 68.1 67.3 7.2275 80.4 72.9 72.1 10.3

ecte

d

239M. Shahid, R.M. Pashley / Desalination 351 (2014) 236–242

glycol monododecyl ether) in known weight of solution (i.e. 200 g of0.5 m NaCl). The temperature of the solution was then again mea-sured every minute throughout the 30 minute bubbling runs. Bothseries of experiments were carried out using different inlet gas tem-peratures ranging from 150 °C to 275 °C, in an attempt to optimizethe bubble column desalination rate. The gas flow rate (at room tem-perature) used in these experiments was fixed at about 23 L/min.

3. Results and discussion

The effects of high inlet (dry) air temperatures on salt coalescenceinhibition were studied using bubble columns filled with distilledwater and with 0.5 m NaCl solution. Fig. 2 illustrates the degree ofbubble coalescence typically observed with distilled water and a highinlet gas temperature of 275 °C. In this case the bubbles were in thecm size range. However, using the same high inlet gas temperaturebut with added salt (i.e. 0.5 m NaCl) the bubbles produced were signif-icantly smaller, in the mm size range (see Fig. 3). Comparison of theseobservations clearly demonstrates that even using a high temperatureinlet gas, at up to 275 °C, has little or no effect on the inhibition of bub-ble coalescence caused by the addition of salt. The solution temperatureof the column quickly approaches the steady state value of about 56 °C.Earlier studies have shown that bubble columns using cool nitrogen gasbut with the column solution heated to about 60 °C actually showed aslight enhancement of the inhibition effect [21].

As two air bubbles collide coalescencemust be influenced by the rateof thinning of the intervening water film. This rate depends on the vis-cosity of the intervening solution as well as on the rate of approachand the size of the bubbles, since these factors will also determine thedegree of deformation of the bubbles [30]. Within a chaotic bubblecolumn collisions occur continuously and within fractions of a second.If the film can drain during this collision time the bubbles will coalesce.Increasing the temperature of the column solution will significantlyreduce its viscosity [31] and this will act to increase the drainage rateof the film between approaching bubbles [32]. This will also allow thefilm to drain before the bubbles are bounced apart within the turbu-lence of the column. Thus the combined effect of an increased columnsolution temperature and the very hot gas bubbles which must, atleast close to the sinter in the bubble column evaporator (BCE), producesignificant local heating in the solution surrounding the bubbles,should significantly reduce film viscosity and hence enhance bubblecoalescence. However, this effect was not observed in the experi-ments reported here. The effect of salt clearly dominates the degreeof coalescence and temperature effects can only be either absent orof negligible importance.

In these studies the heated air flowwas continued for about 30 minand the total weight loss of the solution measured. The actual gas flowrate within the column can be estimated from the room temperatureflow rate using the ideal gas equation, due to the change of air

0

20

40

60

80

100

120

140

0 10 20 30 40 50 60

Vap

Den

sity

(gm

/m3 )

Column Temp (°C)

Fig. 4.Water vapor density as a function of temperature for aqueous 0.5 m NaCl solution.

temperature and pressure. For example, the flow rate in the columnat around 40–60 °C is given by the simple relation: flow rate throughcolumn = inlet (room temp) flow rate × (Tcolumn / Troom) with T inKelvin units. In addition, the flow rate also increases within the columndue to the pressure drop across the sinter. In these experiments, theair passed through the sinterwith a 4.2% higher pressure than the atmo-sphere, which was measured using a digital manometer. The combinedcorrection factor of temperature and pressure was used to calculate the‘expected’ total vapor transfer, obtained by summing the equilibratedwater vapor densities during each minute of the 30 minute runs. Thevapor densities for 0.5mNaCl solutions (given in Fig. 4)were calculatedfrom the ideal gas equation, using the vapor pressure data of Clarkeet al. [33]. At the measured column temperatures, the vapor densityfor each minute was multiplied by the air flow rate in the columnover each of the 30 min of the run. In these experiments, desalinationefficiency was monitored over a wide range of different inlet gastemperatures.

The results given in Table 1 were obtained using inlet air tempera-tures ranging from 150 to 275 °C and the column was filled with 200 gof 0.5 m NaCl solution. The results obtained indicate that the amountof water carried over increased with inlet air temperature, relative tothe expected vapor carry over, obtained from the measured columntemperature and the corresponding vapor pressure. At the highestinlet air temperature, the observed total water transfer at 275 °C wasfound to be about 80.4 g and the calculated weight (calculatedfrom the observed column temperatures) was 72.9 g. By comparison,a column operating at the expected thermal equilibrium tempera-ture would have produced a total weight of about 72.1 g. These re-sults show that at this temperature there was a 10.3% highermeasured water vapor carry over compared to that expected if thevapor pressure in the bubbles was the equilibrium value correspond-ing to that column temperature. The results obtained over a range ofinlet air temperatures is summarized in Fig. 5.

0

2

4

6

8

10

12

150 175 200 225 250 275

% V

apor

izat

ion

Com

pare

d w

ith E

xpE

quili

briu

m V

apor

Pre

ssur

e

Inlet Gas Temperature (°C)

Fig. 5. Improved % vaporization of 0.5 m NaCl solution with increase in inlet gas (dry air)temperature, compared with that expected from the column's operating temperature.

Table 2Desalination efficiency with inlet gas temperature and added surfactant.

Inlet gastemp (°C)

Weightof waterloss (g)

Calculated weightusing measuredcolumn temp (g)

Expectedtheoreticalweight (g)

% measured vaporizationusing surfactant comparedto the expected equilibriumvapor pressure

150 41.6 40.3 40.2 3.2175 49.4 47.1 46.8 4.9200 57.8 53.6 53.5 7.8225 66.2 60.0 59.8 10.3250 75.5 65.6 65.0 15.1275 83.9 71.1 70.5 18.0

20

25

30

35

40

45

50

55

60

65

0 5 10 15 20 25 30

Tem

p (°

C)

Time (min)

Experimental trend

Expected theoretical trend

Fig. 7. Measured solution temperatures within a bubble column containing 0.5 m NaClsolution with 0.002 g of added surfactant (octaethylene glycol monododecyl ether) over30 min, for 275 °C inlet gas (dry air). The measured weight of water evaporated wasfound to be 83.9 g and the expectedweight from the column temperatures was calculatedto be 72.8 g. By comparison, the theoretically expected weight (that is, with the columnoperating at thermal equilibrium) was 72.2 g.

240 M. Shahid, R.M. Pashley / Desalination 351 (2014) 236–242

In all of the experiments carried out in this study, over this range ofinlet temperatures, themeasured total water vapor transferredwas sig-nificantly greater than that expected for the corresponding equilibriumvapor pressures. The results obtained also clearly demonstrate that thiseffect increases with inlet air temperature. Given the relatively shortresidence time of each bubble within the column, of a few tenths of asecond, and the mm size of the bubbles, it appears likely that a slightdegree of supersaturation within the bubbles may have been createdat high bubble temperatures. This supersaturation was not observedin earlier studies carried out at lower air temperatures, i.e. closer toroom temperature. The effect seems to disappear when the inlet airtemperature falls below the boiling point of water. These results alsoshow that themeasured columnweight losswas not due to any artifact,such as aerosol loss into the air above the column, but was entirely dueto water vapor transfer into the bubbles.

In the second series of experiments the effects of the presence of asurfactant monolayer at the bubble surface on the water vapor carryover rate were studied over a range of different inlet gas temperatures(150–275 °C).

Similar column experiments were carried out using the sameamount of saline solution butwith 0.002 g of addednon-ionic surfactant(octaethylene glycol monododecyl ether) over 30 min. This surfactantconcentration was chosen, in part, because it did not produce toomuch foaming at the top of column. The results of these studies aregiven in Table 2, where the total measured weight is compared withthe weight expected from the measured column temperatures and thecorresponding vapor pressures, as well as the weight expected if the

Fig. 6. Photograph of a bubble column containing 0.5 m NaCl solution with 0.002 g ofadded surfactant (octaethylene glycol monododecyl ether) with a gas (dry air) inlettemperature of 275 °C.

columns were operated under conditions of thermal equilibrium.Fig. 7 gives an example of the analysis of a typical column experiment.In this case, at the highest inlet air temperature of 275 °C, themeasuredweight of water evaporated was found to be 83.9 g and the expectedweight from the column temperatures was calculated to be 71.1 g. Bycomparison, the theoretically expectedweight (that is, with the columnoperating at thermal equilibrium) was 70.5 g. The observed weighttransfer was therefore found to be about 18% higher than that expectedfrom the equilibrium vapor pressures. At all air inlet temperatures theeffect of added surfactant was to further enhance the water vapor col-lection rate.

The amount of this surfactant added to the column should be morethan sufficient to fully coat all of the bubbles formedwithin the column,noting that as bubbles collapse others are formed at the sinter and asteady state situation arises, where the bubbles occupy roughly 50% ofthe solution volume at any given time. Given that the column contained200 g of solution before bubbling, the actual bubble/solution volumewas about 400 ml, with 50% occupied by 1 mm radius bubbles. The

0

2

4

6

8

10

12

14

16

18

20

150 175 200 225 250 275% V

apor

izat

ion

Com

pare

d w

ith E

xpec

ted

Equ

ilibr

ium

Vap

or P

ress

ure

Inlet Gas Temperature (°C)

with surfactant

without surfactant

Fig. 8.Enhanced % vaporization as a function of gas (dry air) inlet temperature for columnscontaining 0.5 m NaCl solution, with and without added C12EO8 surfactant. The % wascalculated by a comparison with the vaporization expected from the equilibrium vaporpressures at the operating temperatures of the bubble column measured during the30 min runs.

9

11

13

15

17

19

150 175 200 225 250 275

Eva

pora

tion

laye

r th

ickn

ess a

t bub

ble

surf

ace

(nm

)

Inlet gas temperature (°C)

Fig. 9. The relationship between the temperature of the inlet gas and the calculated thick-ness of the water layer around the surface of the bubbles, of diameter 1 mm, required toevaporate to produce the saturated vapor pressure within the initially dry bubbles.

241M. Shahid, R.M. Pashley / Desalination 351 (2014) 236–242

steady state total bubble surface area would then be about 0.6 m2. Inthe current studies, using 0.002 g of surfactant of molecular weight540 g/mol and assuming each molecule of surfactant occupies roughly1 nm× 1 nm area at the bubble surface, then there is clearly a sufficientadded surfactant (i.e. of total area 2.2m2) to coat the 0.6m2 bubble areawithin the column. At higher levels of surfactant foaming becomes aproblem but at the concentration chosen the foaming level was man-ageable (see Fig. 6). Since the surfactant remains in the column, aseach bubble reaches the surface and collapses, there was no need toadd additional surfactant during the runs.

1mm air

bubble

la

Surfactant C12EO8

Tail Head

Fig. 10. Schematic diagram of adsorbed C12EO8 molecules and the corresponding

The air bubbles are apparently able to capture more water vaporthan under normal conditions and this produces a uniform, efficient ex-change of water vapor into the bubbles, which could then be condensedand collected as purewater. Fig. 8 gives a comparison between the exper-imentswith andwithout added surfactant. The effects of added surfactantare small at modest gas inlet temperatures, around 150 °C but increasedsubstantially at higher temperatures, reaching an 18% improvementover the expected vapor transfer at 275 °C. This represents a substantialincrease in water vapor carry over at inlet air temperatures above theboiling point of water Thus, in effect, the surfactant molecules appear tobe acting somewhat like a “surface molecular diode” either by facilitatingthe transfer of water vapor into the air bubbles or by inhibiting thewater vapor from returning to the solution. In this way the bubbles canbecome supersaturated with pure water vapor. The combined effect of ahigh gas inlet temperature and surfactant monolayer coating appears toproduce a substantial increase in vapor collection, which increases withtemperature. Reducing the amount of added surfactant reduces thisenhanced vaporization.

It is possible to calculate the thickness of water around the surface ofa 1 mm diameter air bubble, which must vaporize to produce the satu-rated vaporwithin the initially dry bubble, at the operating temperatureof the bubble column. Fig. 9 shows the almost linear relationship be-tween the inlet gas temperature and this evaporation layer thicknessfor a 1mmdiameter bubble. Thiswater layermust be evaporated aroundthe bubble surface and this should be considered relative to the thicknessof the adsorbed surfactant monolayer, as illustrated in Fig. 10. Thus, thepresence of a monolayer of non-ionic surfactant might be expected tohave some influence on the vapor collection rate and also the rate ofwater vapor return to the solution. An imbalance of these rates mightlead to a transient water vapor supersaturation within the bubble. Thismechanism is consistentwith the results obtained in this study. It shouldalso be noted that this evaporation process will prevent the surfactant

18.5nm

evaporation

yer thickness

0.5m NaCl

solution

Increasing

thickness

with

temperature

thickness of the evaporation layer required for dry inlet bubbles at 275 °C.

242 M. Shahid, R.M. Pashley / Desalination 351 (2014) 236–242

coating from reaching high temperatures, much above the boiling pointof water, due to the high enthalpy of vaporization. This is supported bythe observation that there was no visible change in evaporation rate orfoaming during the column experiments.

In each experiment the steady state operating temperature of thecolumn can be used to calculate the effective latent heat of vaporizationof the solution using Eq. (1). As an example, even under the mostextreme conditions, with an inlet temperature of 275 °C, the calculatedlatent heat of vaporization (at a column operating temperature of55.5 °C) for the salt solution was found to be 42.7 kJ/mol, close to theliterature value of 42.9 kJ/mol. In general, the observed steady stateoperating temperatures of the columns were found to be close to thevalues expected for thermal equilibrium. However, with added surfac-tant and at the highest inlet air temperature the significant increase inwater vapor carry over, of about 18% (see Fig. 8), means that theseconditions offer the greatest potential for a lower energy thermaldesalination process. Assuming that 85% of the vapor enthalpy can becollected on condensation, a simple calculation (based on the thermalenergy required to heat the air and the pressure work done) suggestsan energy cost, under these conditions, of roughly 290 MJ/m3, whichis comparablewithMSF processes [13].Waste industrial heat and directwind power could also be combined for use with this process to furtherreduce energy costs.

4. Conclusions

The bubble column evaporator process for sub-boiling, thermaldesalination presented here is based on some unusual properties ofwater and salt solutions, which enable the use of high bubble densitiesin concentrated NaCl solutions. The initial results reported here indicatethat a slight but significant degree of supersaturationwithin the bubblesmay have been created when using high bubble temperatures. In addi-tion, the use of adsorbed surfactant monolayers further increases thiseffective supersaturation and creates a substantial increase in watervapor carry over rate. Commercially, waste heat in the form of hotwaste vent gases is widely available at low cost and these could be usedto desalinate large volumes of water using this non-boiling process.

References

[1] B.A. Adams, E.L. Holmes, Adsorptive properties of synthetic resins, Int. J. Soc. Chem.Ind. 54 (1935) 1–6.

[2] G. Meerganz von Medeazza, Water desalination as a long-term sustainable solutionto alleviate global freshwater scarcity? A north–south approach, Desalination 169(2004) 287–301.

[3] P. Beaumont, Qanats on the Varamin Plain, Iran, Trans. Inst. Br. Geogr. 45 (1968)169–179.

[4] D. Mustafa, M.U. Qazi, Transition from karez to tubewell irrigation: development,modernization, and social capital in Balochistan, Pakistan, World Dev. 35 (10)(2007) 1796–1813.

[5] L.F. Greenlee, B.F. Lawler, B.D. Freeman, B. Marrot, P. Moulin, Reverse osmosisdesalination: water sources, technology and today's challenges, Water Res. 43(2009) 2317–2348.

[6] A.D. Khawaji, I.K. Kutubkhanah, J.M. Wie, Advances in seawater desalination tech-nologies, Desalination 221 (2008) 47–69.

[7] C. Tortajada, Water management in Singapore, Water Resour. Dev. 22 (2) (2006)227–240.

[8] I. El Saliby, Y. Okour, H.K. Shon, J. Kandasamy, In.S. Kim, Desalination plants inAustralia, review and facts, Desalination 247 (2009) 1–14.

[9] A. Hurlimann, S. Dolnicar, When public opposition defeats alternative water pro-jects: the case of Toowoomba Australia, Water Res. 44 (2010) 287–297.

[10] C. Fritzmann, J. Lowenberg, T. Wintgens, T. Melin, State-of-the-art reverse osmosisdesalination, Desalination 216 (2007) 1–76.

[11] F. Fung, A. Lopez, M. New,Water availability in+2 °C and+4 °Cworlds, Phil. Trans.R. Soc. A 369 (1934) (2011) 99–116.

[12] J. Zhou, V.W.-C. Chang, A.G. Fane, Life cycle assessment for desalination: a review onmethodology feasibility and reliability, Water Res. 61 (2014) 210–223.

[13] T. Mezher, H. Fath, Z. Abbas, A. Khaled, Techno-economic assessment and environ-mental impacts of desalination technologies, Desalination 266 (2011) 263–273.

[14] M. Marcovecchio, P. Aguirre, N. Scenna, S. Mussati, Global optimal design of me-chanical vapour compression (MVC) desalination process, in: S. Pierucci, G. BuzziFerraris (Eds.), 20th European Symposium on Computer Aided Process Engineering—ESCAPE20, Elsevier, 2010, pp. 1261–1266.

[15] E.R. Reahl, Half a Century of Desalination with Electrodialysis, GE Water & ProcessTechnologies, USA, http://www.gewater.com/pdf/Technical%20Papers_Cust/Americas/English/TP1038EN.pdf (retrieved January 2010).

[16] R. Semiat, Energy issues in desalination processes, Environ. Sci. Technol. 42 (22)(2008) 8193–8201.

[17] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 88th ed., CRC Press, FloridaUSA, 2008.

[18] I.C. Karagiannis, P.G. Soldatos, Water desalination cost literature: review and assess-ment, Desalination 223 (2008) 448–456.

[19] V.I. Klassen, V.A. Mokrousov, An Introduction to the Theory of Flotation,Butterworths, 1963.

[20] V.S.J. Craig, B.W. Ninham, R.M. Pashley, Effect of electrolytes on bubble coalescence,Nature 364 (1993) 317–319.

[21] V.S.J. Craig, B.W. Ninham, R.M. Pashley, The effect of electrolytes on bubble coales-cence in water, J. Phys. Chem. 97 (1993) 10192–10197.

[22] I. Leifer, R.K. Patro, P. Bowyer, A study on the temperature variation of rise velocityfor large clean bubbles, J. Atmos. Ocean. Technol. 17 (2000) 1392–1402.

[23] R. Clift, J. Grace, M. Weber, Bubbles, Drops and Particles, Academic Press, New York,1978.

[24] E. Klaseboer, R. Manica, D.Y.C. Chan, B.C. Khoo, BEM simulations of potential flowwith viscous effects as applied to a rising bubble, Eng. Anal. Bound. Elem. 35(2011) 489–494.

[25] M. Francis, R.M. Pashley, Application of a bubble column for evaporative cooling anda simple procedure for determining the latent heat of vaporization of aqueous saltsolutions, J. Phys. Chem. B 113 (2009) 9311–9315.

[26] M. Francis, R.M. Pashley, Thermal desalination using a non-boiling bubble column,Desalin. Water Treat. 12 (2009) 155–161.

[27] M. Shahid, R.M. Pashley, R.A. Mohklesur, Use of a high density, low temperature,bubble column for thermally efficient water sterilization, Desalin. Water Treat.(2013) 1–9.

[28] P. Budhiraja, A.A. Fares, Studies of scale formation and optimization of antiscalantdosing inmulti-effect thermal desalination units, Desalination 220 (2008) 313–325.

[29] R.M. Pashley, On the Surface of Things, Connor Court Publishing, Ballan, Vic, 2012.[30] R.G. Horn, L.A. Del Castillo, S. Ohnishi, Coalescence map for bubbles in surfactant-

free aqueous electrolyte solutions, Adv. Colloid Interf. Sci. 168 (2011) 85–92.[31] D.R. Lide, CRC Handbook of Chemistry and Physics 2012–2013: A Ready Reference

Book of Chemical and Physical Data, 2012.[32] S.P. Frankel, K.J. Mysels, On the “dimpling” during the approach of two interfaces,

The Journal of Physical Chemistry 66 (1962) 190–191.[33] E.C.W. Clarke, D.N. Glew, Evaluation of the thermodynamic functions for aqueous

sodium chloride from equilibrium and calorimetric measurements below 154 °C, J.Phys. Chem. Ref. Data 14 (1985).