chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

journa l h om epage: www.elsev ier .com/ locate /cherd

Heat extraction and brine management fromsalinity gradient solar pond and membranedistillation

Kamran Manzoor, Sher Jamal Khan ∗, Yousuf Jamal,Muhammad Aamir ShahzadInstitute of Environmental Sciences and Engineering, School of Civil and Environmental Engineering, NationalUniversity of Sciences and Technology (NUST), Islamabad, Pakistan

a r t i c l e i n f o

Article history:

Received 24 July 2016

Received in revised form 27

November 2016

Accepted 21 December 2016

Available online 29 December 2016

Keywords:

Heat extraction

Brine management

Desalination

Permeate flux

Brine pre-heating

a b s t r a c t

The problem associated with the reverse osmosis (RO) system is its brine disposal. Exper-

imental salinity gradient solar pond (SGSP) with surface area 4.65 m2 was established to

extract and apply heat on laboratory scale direct contact membrane distillation (DCMD).

In this study, the performance of the SGSP and DCMD system under Pakistan’s climatic

condition of Islamabad was evaluated. The heat extraction was carried out using internal

heat exchanger by passing fresh water through it at different flow rates in summer and

winter. Maximum temperature of 37 ◦C in summer and 28.5 ◦C in winter was extracted. The

extracted heat from SGSP can be used to pre-heat the brine for temperature driven desalina-

tion processes. Least drop in lower convective zone (LCZ) temperature of SGSP was observed

at flow rate of 7.5 L/min. In DCMD, two temperatures obtained from SGSP (28.5 and 37 ◦C)

at feed side were maintained to investigate the permeate flux, percentage salt rejection

and total dissolve solids (TDS). Two further temperatures 50 and 60 ◦C were maintained to

investigate the DCMD performance. Flux increased as temperature difference between feed

and permeate increased. 28.5 ◦C in terms of SGSP temperature was also feasible for DCMD

process as flux was maintained over time.

© 2016 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction

The demand of fresh water is increasing worldwide with the increase

in population to maintain the desired living standards (Matti et al.,

2010). Pakistan is also facing challenges of water shortfalls as well as

ground water contamination in terms of total dissolved solids (TDS)

ranging from 1500 to 4000 mg/L in many areas including Punjab, Sindh

and Balochistan provinces (Ahmad and Chand, 2015; Kahlown et al.,

2004). Reverse osmosis (RO) is the most commercially available mem-

brane desalination technology and common method to increase fresh

water production around the world because of relatively cheaper mem-

brane and specific water production costs (Hassan and Fath 2013; Zhu

et al., 2010). Brine disposal is the main drawback of RO process. Surface

water discharge, deep well injection, land application and evaporation

∗ Corresponding author.E-mail addresses: s.jamal@iese.nust.edu.pk, sherjamal77@gmail.co

http://dx.doi.org/10.1016/j.cherd.2016.12.0170263-8762/© 2016 Institution of Chemical Engineers. Published by Elsev

ponds are commonly used disposal options for RO concentrate that are

expensive and have negative impact on the water environment due

to high salinity (Afrasiabi and Shahbazali, 2011; Pérez-González et al.,

2012; Rosado and Bernaola, 2014; Shannon et al., 2008; Subramani and

Jacangelo, 2014). For better brine management and more fresh water

production from RO concentrate, advance methods of temperature

driven membranes are needed.

Membrane distillation (MD) is an emerging technology among

desalination processes and brine management techniques (Cath et al.,

2004; Cipollina et al., 2012; Pantoja et al., 2015). In MD, hydrophobic

membrane transports vapor molecules by temperature driven sepa-

ration process. The advantage of MD is to combine both membrane

technology and evaporation process in one unit (El-Bourawi et al.,

2006; Hwang et al., 2011). MD is classified in four different types,

direct contact membrane distillation (DCMD), airgap membrane dis-

tillation (AGMD), sweeping gas membrane distillation (SGMD) and

m (S.J. Khan).

vacuum membrane distillation (VMD). DCMD is a suitable technology to

concentrate the aqueous solutions and for desalination processes

ier B.V. All rights reserved.

chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237 227

(

b

a

l

p

w

i

p

r

l

g

D

t

g

v

(

a

e

m

c

t

o

f

e

c

e

(

t

l

f

t

p

i

a

r

a

2

c

d

n

c

f

c

t

t

z

w

b

c

h

d

g

l

o

n

i

T

o

t

T

c

L

i

d

s

f

2

f

a

Camacho et al., 2013). Though the MD process is old, recently it has

een widely studied for saline water and wastewater treatment such

s thermal distillation and production of more drinking water due to

ow cost and simple design (He et al., 2011). DCMD can be used for com-

lete rejection of all non-volatile components in feed solution due to

ater transported through the membrane only in vapor phase exhibit-

ng nearly 100% salt ions rejection (Nghiem et al., 2011). In separation

rocesses, DCMD membrane requires low operating pressures than

everse osmosis membrane (Cath et al., 2004). Because DCMD requires

ow operating pressure than conventional distillation, it can utilize low

rade heat as hot feed (Bui et al., 2010; Lawson and Lloyd, 1997). In

CMD process, hot feed solution and cold distillate are in direct con-

act with membrane on both sides with driving force of vapor pressure

radient. The advantage of DCMD is to obtain high fluxes and disad-

antage is heat losses due to conduction through membrane surface

Camacho et al., 2013; Qtaishat and Banat, 2013). The main problem

ssociated with DCMD is the energy requirement, particularly heat

nergy requirement i.e. the energy needed to heat the feed brine on

embrane side (Bouguecha et al., 2015; Boukhriss et al., 2015). SGSP

an be considered as one of the promising renewable energy sources

hat can provide low grade heat to DCMD. It allows DCMD for remote

peration with very little electrical input (least equipment) by reducing

ossil fuel consumption and increasing fresh water production (Suarez

t al., 2015). SGSP can store heat for long term which can be used during

loudy days and night time. SGSP has low cost per unit area of collector,

ssential storage capacity and can easily be constructed over large area

Lu et al., 2001; Nakoa et al., 2015). For useful heat, the annual collec-

ion efficiency is around 10–15%. Smaller ponds are less efficient than

arger ponds due to losses at the pond edge. Solar ponds are suitable

or desalination plants as waste brine from desalination can be used as

he salt source for establishing density gradient. Salinity gradient solar

ond (SGSP) can be attractive renewable source to heat the brine used

n DCMD process (Qtaishat and Banat, 2013). Arid and semi-arid zones

re well matched to this method due to maximum availability of solar

adiations (Seckler et al., 1999). Many parts of Pakistan get solar radi-

tions ranging from 4.7 to 6.2 KWh/m2/day annually (Gadiwala et al.,

013).

A large water body containing different concentration of salts (also

alled salinity gradients) between 2 to 5 m deep is called salinity gra-

ient solar pond (SGSP). SGSP contains lower convective zone (LCZ),

on-convective zone (NCZ) and upper convective zone (UCZ). The upper

onvective zone (UCZ) has low saline water and its temperature is uni-

orm, close to ambient. To compensate for evaporation, UCZ needs

ontinuous flushing with fresh water which allows the top surface

o flush away the rising salts by the natural process of salt diffusion

hrough non-convective zone (NCZ). Below the UCZ, the non-convective

one (NCZ) comprises a variation of salt concentrations such that the

ater near to the upper layer of NCZ is always less dense than the water

elow in NCZ. The heat in NCZ cannot be transmit by convection pro-

ess. The main purpose of NCZ is to act as an insulator which prevent

eat from escaping to the UCZ and maintain higher temperatures at

eeper layer. In this layer, the temperature is not uniform and increases

radually with depth to form temperature gradient. The last bottom

ayer is the lower convective zone (LCZ) also called storage zone (SZ)

r energy storage zone (ESZ). In this layer, the density is uniform and

ear to saturation. This layer has the function of absorbing and stor-

ng the heat (Gietas et al., 2009; Leblanc et al., 2011; Rizvi et al., 2015;

undee et al., 2010). LCZ stores solar radiation for long term in the form

f thermal energy which can be utilized for heating applications like

hermal desalination by direct contact membrane distillation (DCMD).

o compensate the losses of salt from NCZ and LCZ, solid salt or high

oncentration brine should be regularly added to LCZ. Solar energy in

CZ can only escape by conduction. The thermal conductivity of water

s moderately low and also heat escapes upward from the LCZ slowly

ue to NCZ. This makes the solar pond perform as a long term storage

ystem (Nie et al., 2011). SGSP is a least expensive and suitable option

or storage of heat to make it attractive for many applications (Lu et al.,

002, 2004). SGSP has been established and studied around the world

or about half century in different countries such as Israel (Beit HaAr-

va, West Bank), US (El-Paso, Texas), India (Bhuj, Gujarat) and Australia

(Pyramid Hill, Victoria) (Akbarzadeh et al., 2008; Kumar and Kishore,

1999; Newell et al., 2000; Tabor and Doron, 1990). The heat from LCZ

can be extracted through heat exchangers (Leblanc et al., 2011; Suarez

et al., 2010, 2015).

DCMD is simple in terms of construction and operation as it requires

only a membrane module, low grade heat source and two low pressure

pumps to couple with SGSP for solar powered thermal desalination

(Suarez et al., 2015). Heat has been provided for several kinds of desali-

nation processes by using solar ponds in different studies (John and

Walton, 2001; Qtaishat and Banat, 2013; Suarez et al., 2010). In some

studies, the performance of DCMD/SGSP coupled system was inves-

tigated to achieve solar power thermal desalination (Lu et al., 2004;

Suarez et al., 2010, 2015). Brine management and production of fresh

water from different techniques of MD has been investigated (Adham

et al., 2013; Afrasiabi and Shahbazali, 2011; Duong et al., 2015; Pérez-

González et al., 2012; Rosado and Bernaola, 2014; Subramani and

Jacangelo, 2014). Guillén-burrieza et al. (2011) studied the AGMD pro-

cess for brine management. They fed the system with sodium chloride

solutions between 1 and 35 g/L at temperatures up to 85 ◦C in the

feed and up to 75 ◦C in the refrigeration (cold stream). For heat pur-

pose, a static solar collector field was provided. Martinetti et al. (2009)

studied vacuum enhanced direct contact membrane distillation (VED-

CMD) with two RO brines used as feed with TDS concentrations of

7500 and 17,500 mg/L. They showed that recovery factor up to 81%

was achieved but this factor was limited by precipitation of inorganic

salts on the membrane surface. They also showed the cleaning tech-

niques to remove the scaling layer from membrane surface to restore

the water permeate flux almost to the initial level. Ji et al. (2010) inves-

tigated the bench scale membrane distillation crystallization (MDC)

for sodium chloride crystallization and water recovery. They showed

that MDC has ability to concentrate RO brines. They also investigated

that industrial scale of MDC process for managing large volume of

brines with no technical difficulty. Tun and Groth (2011) investigated

MD along with a crystallizer to concentrate RO brine having conductiv-

ity of 15 mS/cm from industrial wastewater. They obtained recovery of

95% from overall feed and flux of 3–5 L/m2 h. One of the method that

can be adopted for brine management is DCMD and heat required for

DCMD can be obtained from SGSP. This method has not been inves-

tigated to the extent earlier. Therefore, further understanding of this

concept is necessary.

In this applied research, effort has been made to evaluate the

performance of semi-pilot-scale SGSP in terms of heat extraction

and lab-scale DCMD in terms of brine management separately under

Islamabad’s climatic condition. The objectives of this work were to

investigate heat extraction from SGSP in summer and winter and to

investigate the fluxes, percentage salt rejection and total dissolve solids

(TDS) in feed with respect to time in DCMD. The significant factors per-

tinent to this semi-pilot-scale SGSP and lab-scale DCMD will permit

the successful preparation of heat extraction and brine management

under pilot scale research and to couple the system of DCMD/SGSP.

2. Material and methods

2.1. Salinity gradient solar pond (SGSP) design andconstruction

Salinity gradient solar pond (SGSP) with area of 4.65 m2 havingdepth 1.9 m and capacity of 8500 L was designed and con-structed as shown in Fig. 1. A small distinct section was alsoconstructed which was separated by mirror to observe thepond and temperatures of gradient layers. Thickness of thewalls were kept 0.25 m. The walls of the pond were constructedwith hollow blocks due to the presence of air pockets which actas a thermal insulator between the pond and the surroundingsoil. A layer of 25.4 mm plaster followed by 25.4 mm layer ofchips were used in order to make the pond water resistant.

The glass was made from three tampered glass (eachsize of 0.012 m) sandwiched together and sealed from the

228 chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237

Fig. 1 – Schematic diagram of salinity gradient solar pond (SGSP).

Fig. 2 – (a) Thermocouples in glass and sampling points at various heights (b) solar pond filled with brine for operation.

sides. In order to ensure prevention of any leakages, a spe-cial sealant (Sika HP 2 bond) was lined on both sides ofthe glass. There were holes punched into the glass in orderto insert thermocouples at various heights. The samplingpoints were installed at different heights on the brick wallas shown in Fig. 2(a). The internal walls of the SGSP werelined with black, smooth and stiff surface high densitypolyethylene (HDPE) liner with a thickness of 1 mm whichallows maximum absorption of the solar radiation in thepond.

Stainless steel pipe grade 304, Schedule-10S and nominalouter diameter 21.1 mm of 11.6 m length was bent to makean internal heat exchanger. Solar pond filled with brine foroperation is shown in Fig. 2(b).

For measuring the temperature values, two digital meters(T4WM, Autonics Series, Korea) (5 thermocouples K-typein 1) were used. Moreover, three extra digital temperatureindicators (TPM-900, SANHNG, China) having accuracy ofup to ±0.1 ◦C were used to measure the ambient temper-ature and temperatures of water entering and exiting theheat exchanger. Specific gravity was measured by using twohydrometers having ranges (1.000–1.100) and (1.100–1.200).The accuracy of hydrometers reached up to 0.001. Densitywas calculated by using Eq. (1) in g/cm3, taking liquid sam-ple using 25 mL density flask weighing on analytical balance

(AAA 160LE, ADAM, USA) and then value in g/cm3 convertedinto kg/m3.

Density = Mass of liquidvolume of liquid

(1)

1.5 HP centrifugal pump (PW-600 M, Wilo, Korea) was usedhaving flow range up to 70 L/min. Variable frequency drive(VFD) (RM5G, RHYMEBUS, Japan) was connected to pumpalong with the flowmeter (PF2A721-03-27M, SMC, USA) rangefrom 3.5 to 30 L/min. Changing the frequency of VFD resultedin changing the frequency of the pump rotation and flowrates as per requirement and the adjusted flow was measuredthrough the flow meter placed after the pump.

2.1.1. Diffuser designDiffuser design included two semi-circular disc separated bya constant slit as shown in Fig. 3.

The discs were made of metal plates separated by nutsof width 2 mm. Flowrate required for the density gradient tobe established from fixed injection technique also known asre-distribution method was calculated through a Froude equa-tion by putting Froude number constant at 18 (Zangrando,1980). If Froude number is maintained smaller than 18, theinjected water rises and mixes above the diffuser level. If

Froude number is greater than 18, the injected water mixesbelow the diffuser level (Leblanc et al., 2011; Zangrando, 1980).

chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237 229

Fig. 3 – Diffuser used for gradient establishment.

Bw0t

F

w

g

2(

2AMepts2

2FTr4a2Dut(ss

tion was increased in winter for further experimentations.The entire process of gradient establishment consumed 8 h in

ased on the flowrate available of up to 30 L/min, the slit widthas kept at 0.0025 m and the diameter was maintained at

.08 m. By using Froude equation putting Froude number 18,he injection velocity was found by solving Eq. (2).

roude number =V√

gd × √( ��−�i�i

) (2)

hereV = velocity of fluid (m/s)g = gravitational acceleration (m/s2)d = diffuser slit width (m)�� = density of fluid environment (kg/m3)�i = density of injected fluid (kg/m3)Commercial salt sodium chloride (NaCl) was used for the

radient establishment as it was cheaper and locally available.

.2. Lab-scale direct contact membrane distillationDCMD) setup

.2.1. Membrane flat sheet microporous hydrophobic membrane (Porousembrane Technology, Ningbo, China) was used in this

xperiment. The major membrane characteristics wereolypropylene (PP) supporting layer laminated with 12 �mhick polytetrafluoroethylene (PTFE) active layer having poreize of 0.2 �m and porosity of 70% respectively (Duong et al.,015).

.2.2. Synthetic feed characteristicseed side synthetic brine solution of concentration 7560 mg-DS/L was prepared based upon 40% recovery from pilot-scaleeverse osmosis (RO) plant having feed concentration of500 mg/L operated at the National University of Sciencesnd Technology (NUST), Islamabad, Pakistan (Khanzada et al.,017). Distilled water was used on the permeate side ofCMD system. Analytical grade salts and distilled water weresed in synthetic brine solution preparation with concen-rations of sodium chloride (1922.4 mg/L), calcium chloride2041.5 mg/L), magnesium chloride hexahydrate (2127.0 mg/L),odium nitrate (99.0 mg/L), sodium sulfate (1333.7 mg/L), andodium bicarbonate (40.6 mg/L).

2.2.3. Experimental set up and procedureA closed loop laboratory-scale direct contact membrane dis-tillation (DCMD) setup used in this study is shown in Fig. 4.

A flat sheet membrane module was used made of acrylic.Dimensions of the module included 14 cm length and 6 cmwidth. The active module dimensions in which membranewas operated included 9.5 cm length, 4 cm width and 0.1 cmdepth as shown in Fig. 5.

The membrane module comprised of two sections, the feedside and the permeate side. The module was fixed horizon-tally so that feed brine solution (hot) flowed through bottomside of cell and permeate (cold) flowed through upper side ofcell. The feed and permeate were isolated through flat sheethydrophobic microporous membrane with membrane effec-tive area of 380 cm2 (0.00038 m2). The flow rate at both sides(feed and permeate) was kept constant at 150 mL/min. Themembrane was placed in the module having the active layerof membrane encountering the hot stream. The feed and per-meate were contained in 10 L plastic reservoir and desiredflow through membrane module was maintained by a peri-staltic pump (Model 7524-45, Master flex, USA). A chiller wasused to maintain permeate temperature at 20 ◦C. Two TDSmeters (KOMATSU, Japan) ranging from 0 to 5000 mg/L withtemperature sensor were used for permeate in and out. Twotemperature indicators (TPM-900, SANHNG, China) were usedto monitor the feed inlet and outlet temperatures. Stainlesssteel heating rod, magnetic contactor (GCM-22, LS, Korea) andtemperature controller (XMTG-131 China) was used for heat-ing and temperature control in feed reservoir. Feed TDS wasmonitored by TDS meter (Sension5, HACH, USA). Permeatevapors were collected through membranes and returned toreservoir where it was measured by digital balance (UX 6200H,SHIMADZU, Japan). Digital balance was set for every 5 minreading connected with desktop computer.

3. Results and discussion

Punjab Saaf Pani (Clean Drinking Water) Company Pakistanis installing several RO plants in different districts of Pun-jab province, so the effective treatment and disposal of ROconcentrate is necessary requirement for these plants. Alsothis region has solar potential in terms of solar radiationsand convenient availability of ordinary salt for installing SGSPto provide heat for DCMD system. Results were divided intotwo separate phases for better understanding of the processof brine management in sub-continental areas especially inPakistan’s environment. The results of these two separatephases will permit the successful preparation of heat extrac-tion and brine management under pilot scale research and tocouple the DCMD and SGSP systems.

3.1. Phase 1: salinity gradient solar pond (SGSP)

3.1.1. Gradient establishment and density measurementGradient layer was established separately in salinity gradi-ent solar pond (SGSP) for both summer and winter seasons.In both seasons, densities for gradients establishment weredifferent. Water density for summer was kept at 1750 kg/m3.Gradient establishment for winter was established again dueto monsoon rainfall which spoiled the previously establishedgradient and it was maintained at 1190 kg/m3. Concentra-

230 chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237

Fig. 4 – Schematic diagram of lab scale direct contact membrane distillation (DCMD) system.

odul

Fig. 5 – Flat sheet acrylic m

summer and 10 h in winter. In summer, height of LCZ, NCZ andUCZ was maintained at 0.508, 0.762, and 0.56 m, respectively.While for winter, height of LCZ, NCZ and UCZ was maintainedat 0.635, 0.863, and 0.330 m, respectively. In winter, UCZ waskept less than that of summer as UCZ height should be inbetween 15–25% of total height of the solar pond (John andWalton, 2001).

For summer, SGSP was filled up to a height of 0.89 m whilefor winter, SGSP was filled up to a height 1.06 m. The diffuserwas placed at the interface of LCZ and NCZ for establishmentof salinity gradient (Leblanc et al., 2011). Diffuser was posi-tioned along the glass side so that its height could be seenfrom the scale calibrated on the glass. Only horizontal mixingwas observed during the whole process due to Froude num-ber of 18. Other studies found that no vertical mixing occurreddue to Froude number maintained at 18 (Leblanc et al., 2011;Zangrando, 1980). The diffuser level was raised to the topof NCZ and its level became equal to the height of water atthat instant. Comparing the experimental density profile withanticipated density profile, it was investigated that same slopewas obtained, however a little deviation from the experimen-tal profile was observed due to a small difference in densitiesof NCZ and LCZ as shown in Fig. 6. Density profile slope aftergradient establishment in this study complemented otherstudies using gradient establishment techniques, either on alarge scale or small scale setup having different salt concen-

trations (Busquets et al., 2012; John and Walton, 2001; Leblancet al., 2011).

e (a) top view (b) side view.

Density measurements were undertaken on weekly basisfor both summer and winter as shown in Fig. 7. It showsthat there were no such variations in LCZ due to high saltconcentration but some variations in UCZ were due to someevaporation loss but these variations did not affect the gradi-ent layer. Long term operation of solar pond should be takenunder careful consideration as changes may occur both in LCZand UCZ with time (Jaefarzadeh and Akbarzadeh, 2002).

3.1.2. Daily average temperatures in LCZ and UCZ duringsummer and winterDaily average temperature was the average temperatureobtained after every two hours starting from early morningin day time. It was calculated by the average formula by sum-ming all of the numbers in a list divided by the total number

in that list given by A = 1n

n∑

i=1

xi. The temperature variations

and heat extraction from lower convective zone (LCZ) in thesolar pond during summer and winter were investigated. Withincrease in time, the temperature of LCZ and non-convectivezone (NCZ) increased whereas a little variation is shown inupper convective zone (UCZ), depending upon the weather. Amaximum temperature of 44 ◦C was observed at steady statecondition in LCZ with a �T (LCZ–UCZ) of 14 ◦C for summeras shown in Fig. 8 (a). A maximum temperature of 35 ◦C wasobserved at steady state condition in LCZ with a �T (LCZ–UCZ)

of 10 ◦C for winter as shown in Fig. 8 (b). Based on Fig. 8, 35 ◦Cand 27 ◦C are the initial temperatures that the pond started

chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237 231

Fig. 6 – (a) Density profile for summer (b) density profile for winter.

Fig. 7 – (a) Weekly specific gravity investigation for summer (b) weekly specific gravity investigation for winter.

Fig. 8 – (a) Daily temperatures for summer (b) daily temperatures for winter.

232 chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237

Fig. 9 – (a) Heat extraction at 3.5 L/min (b) heat extraction at 5.5 L/min (c) heat extraction at 7.5 L/min (d) comparison of heat

extraction in summer.

to gain heat from solar radiation. 37 ◦C in summer at outlet ofheat exchanger was achieved after 30 days when the temper-ature of the pond was about 44 ◦C. 28.5 ◦C in winter at outlettemperature was achieved after 30 days when temperature ofpond was about 35 ◦C. Other studies investigated the effect ofambient temperature in LCZ with time found that the temper-ature gain was between 38–50 ◦C (Dah et al., 2010; Jaefarzadehand Akbarzadeh, 2002; Nakoa et al., 2015; Nie et al., 2011) thatwell matched the temperature gained in LCZ in this study.Moreover, Leblanc et al. (2011) investigated the temperaturevariation in LCZ and UCZ and reported that daily averagetemperature increase in LCZ was 0.7 ◦C. In this study, it wasobserved that the average temperature of 0.4–0.8 ◦C increasedin LCZ in summer and 0.3–0.5 ◦C increased in winter on dailybasis.

3.1.3. Heat extraction from LCZ in summer

Heat extraction was carried out for three different flow ratesof 3.5, 5.5 and 7.5 L/min. Fig. 9(a) shows that for 3.5 L/min, drop

of 6 ◦C in LCZ was reported and maximum heat extraction of7451 kJ was achieved. This was due to the increased contacttime that allowed greater heat transfer. Fig. 9(b) shows thatfor 5.5 L/min, drop of 4 ◦C in LCZ was obtained and maximumheat extraction of 6593 kJ was achieved. Fig. 9(c) shows thatfor 7.5 L/min, drop of 2 ◦C in LCZ was observed and maximumheat extraction of 5651 kJ was achieved. Also at this temper-ature, least drop in LCZ was observed. The reason is that asthe velocity of water increases, contact time of water throughheat exchanger pipe decreased. Fig. 9(d) shows the compari-son of all three flow rates in summer. Outlet temperatures ofthe heat extraction fluid for summer were 37 ◦C for 3.3 L/min,35 ◦C for 5.5 L/min and 33 ◦C for 7.5 L/min.

3.1.4. Heat extraction in winterHeat extraction was carried out for three different flow ratesof 7.5, 9.5 and 11.5 L/min. These flow rates were different

than those previously investigated for summer in order tooptimize the flow rate of heat extraction from LCZ. Since

chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237 233

Fig. 10 – (a) Heat extraction at 7.5 L/min (b) heat extraction at 9.5 L/min (c) heat extraction at 11.5 L/min (d) comparison ofheat extraction in winter.

Fig. 11 – (a) Flux (L/m2 h) at (28.5, 37 50 and 60) ◦C (b) flux (L/m2 h) at temperature difference (8.5, 17 30 and 40) ◦C.

234 chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237

Fig. 12 – Flux (L/m2 h) at temperatures 28.5, 37, 50 and 60 ◦Cfor 24 h duration.

major drop in LCZ temperatures were found for 3.5 and5.5 L/min flow rates in summer, these flow rates were notconsidered for the following winter. Fig. 10(a) shows that for7.5 L/min, drop of 2.5 ◦C in LCZ was reported and maximumheat extraction of 5180 kJ was achieved. Slightly more dropin LCZ is due to the winter weather as surrounding tempera-ture decreased. Fig. 10(b) shows that for 9.5 L/min, drop of 2 ◦Cin LCZ was reported and maximum heat extraction of 3770 kJwas achieved. Fig. 10(c) shows that for 11.5 L/min, drop of 1.5 ◦Cin LCZ was obtained and maximum heat extraction of 2930 kJwas achieved. Fig. 10(d) shows the comparison of all three flowrates in winter. In order to retain the temperature of LCZ, thetime of the contact hour was fixed to 3 h and the consequentchanges were recorded. In this study, the optimum flowratewas found to be 7.5 L/min based upon heat extractions of 5651and 5180 kJ for summer and winter seasons with least drop inLCZ. Outlet temperatures of the heat extraction fluid for sum-mer were 28.5 ◦C for 7.5 L/min, 26 ◦C for 9.5 L/min and 23 ◦Cfor 11.5 L/min. In other pertinent studies, flow rates for heatextraction between 3 and 6 L/min were reported (Leblanc et al.,2011; Lu et al., 2004; Saleh et al., 2011; Suarez et al., 2015).Tundee et al. (2010) theoretically investigated the heat extrac-tion using heat exchanger in solar pond for 3 h. They foundthat only 1 ◦C loss in LCZ temperature was observed. Anotherstudy experimentally conducted heat extraction using inter-nal heat exchanger with 4 m2 pond area and depth of 1.1 mand found that using pump discharge of 1 L/min, temperaturedrop in LCZ for first four days declined from 59 to 40 ◦C. Afterincrease in LCZ temperature, temperature drop in LCZ for next2 days declined from 47 to 38 ◦C and then become stable forthree days (Jaefarzadeh, 2006).

3.2. Phase 2: direct contact membrane distillation(DCMD) process

DCMD experiment was carried out to investigate permeatefluxes at four different feed water temperatures. In this phase,four different temperatures (28.5, 37, 50, 60 ◦C) at feed sidewere maintained to investigate permeate flux, percentage saltrejection and feed TDS concentration. Constant temperatureof 20 ◦C was maintained at permeate side and 24-h operationtime was maintained under each feed side temperature condi-tion. Fig. 11(a) shows the behavior between fluxes versus timeat different temperatures. It shows that as temperature of feedincreased, flux also increased linearly. At temperature of 60 ◦C,the temperature influence was more significant on perme-ation flux as compared to 28.5 ◦C depicting that vapor pressureand temperature has an exponential relation. Same trendwas shown in Fig. 11(b) between flux vs �T (8.5, 17, 30, and40 ◦C). It shows the effect of change in temperature betweenfeed and permeate water on the performance of DCMD. Withincrease in temperature and vapor pressure difference acrossthe membrane, increased the conductive heat flux and driv-ing force across the membrane surface. Nakoa et al. (2015)investigated the DCMD performance at different temperaturescoupled with solar pond. They maintained the permeate tem-perature at 20 ◦C and feed temperatures varied between 29 and45 ◦C. They reported that at higher feed temperature, perme-ate flux was higher. Suarez et al. (2015) investigated the DCMD

coupled with solar pond and observed steady state conditionin first 2.5 h at feed temperature of about 36 ◦C.

3.2.1. Effect of temperature on permeate flux propensityPermeate flux, feed TDS and percentage of salt rejectionbehaviors with respect to time by using PTFE membrane arediscussed in this section. Fig. 12 shows that flux increased astemperature difference between feed and permeate increased.28.5 ◦C is also feasible for DCMD process as flux was stable withpassage of time.

Fig. 13(a) shows that feed TDS increasing with respect totime as temperature of feed increased. Fig. 13(b) showed thatafter 24 h, the percentage of salt rejection is acceptable for alltemperatures even at 28.5 ◦C which is about 98.7%. Percentageof salt rejection was calculated by using Eq. (3).

Percentage of salt rejection= (Cf − Cp) /Cf × 100 (3)

where Cf is the concentration in feed and Cp is the concentra-tion in permeate.

Cath et al. (2004) studied the performance of DCMD forsodium chloride and synthetic feed salt solution at 40 ◦C fordifferent type of membranes with different thickness. Theyfound that salt rejection was 99.9% and the salt concentra-tion in the feed had only minor effect on the water flux. Fardand Manawi (2014) studied the sea water desalination usinghydrophobic PTFE membrane in DCMD process for 17 h. Theyfound in their experiments that salt rejection for all salts were99.9%. By using PTFE membrane in this experiment, it wasobserved that low fluxes were recorded at all temperatures.Other commercial polyethylene (PE) membrane can be usedto enhance flux. PE membranes have intrinsic hydrophobic-ity with low surface energy and low thermal conductivity. PEmembranes having sponge like morphology, porosity, appro-priate pore size, thickness and reasonable wetting resistancemake these membranes good candidates for MD process. Athigher feed flow rates, these membranes lead to a higherpermeate flux and reduced temperature polarization effect(Zuo et al., 2016). In this study, after 24 h of continuous oper-ation, the permeate TDS concentration was well below thepermissible limit of 1000 mg/L as per WHO Guidelines forDrinking Water Quality (WHO, 2011WHO Guidelines, 2011).These results showed that even low SGSP extracted temper-ature of 28.5 ◦C was also feasible for MD process as flux wasstable with time and TDS concentration increased with reduc-

tion of feed side volume.

chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237 235

) pe

atmsfpscer

4

IutIooipcsrdftmTfwccsnbamtt

Fig. 13 – (a) Feed TDS vs time (b

Sustainable heat requirement to MD process can bechieved based upon the design of SGSP adjusted in the wayhat energy recovered through the heat exchanger in a batch

ode can be stored in a feed tank and then fed to the MDystem operated continuously. Scale up factor can be used inuture study for sustainable heat provided for continuous MDrocess based upon the amount of hot water generated fromolar pond and amount of water utilized by MD process. Toorrelate both SGSP and DCMD in our case, solar pond cov-red at least 20 times requirement of DCMD at minimum flowate of heat extracted via the internal heat exchanger.

. Conclusions

n SGSP, flow rate of 7.5 L/min was found to be optimum basedpon minimal drop of 2.0–2.5 ◦C in LCZ. Due to this reason,he LCZ heat can be used to pre-heat the concentrated brine.n summer, temperature of 37 ◦C was observed at the outletf heat exchanger while in winter, temperature of 28.5 ◦C wasbserved after passing through the heat exchanger. It can be

nferred that the heat extracted from SGSP can be used tore-heat the brine for temperature driven desalination pro-ess for longer time period at low temperatures as flux wastable with time. In DCMD, low temperatures of 28.5 and 37 ◦Ceduced the flux and thermal efficiency but flux was stableuring the operational duration for given feed volume. Theeed TDS concentration increased with DCMD operation dueo temperature induced flux over 24 h operation and maxi-

um TDS concentration of 8400 mg/L at 60 ◦C was observed.he permeate flux increased with increasing temperature dif-

erence through the membrane side while the permeate TDSas below the permissible limit of 1000 mg/L. As Pakistan is

ontinuously facing shortage of drinking water, DCMD processan provide highly concentrated brine as well as productionafe drinking water. SGSP can be considered as a suitable tech-ology for heat recovery as well as for brine managementecause of abundant solar radiation available in arid and semi-rid regions. Also availably of salt lakes can make solar pondore efficient as existing brine can be replenished by concen-

rated brine in LCZ. This study of two separate phases reflectshat the combined SGSP and DCMD process can be a viable

rcentage salt rejection vs time.

option for heat recovery as well as brine management underpilot scale research in further study.

Acknowledgments

Authors would like to express their gratitude to WaterAidPakistan (WAP) for their financial support throughout theproject (Partnership Agreement # 6IN03). Authors would liketo acknowledge Professor Long Nghiem from University ofWollongong, Australia for providing hydrophobic flat sheetmembrane distillation (MD) membrane for research work.

References

Adham, S., Minier-matar, J.E., Janson, A., Adham, S., Hussain, A.,Matar, J.M., Dores, R., Janson, A., 2013. Application ofmembrane distillation for desalting brines from thermaldesalination plants desalination plants. Desalination 314,101–108, http://dx.doi.org/10.1016/j.desal.2013.01.003.

Afrasiabi, N., Shahbazali, E., 2011. RO brine treatment anddisposal methods 3994. Desalin. Water Treat. 35, 39–53,http://dx.doi.org/10.5004/dwt.2011.3128.

Ahmad, M., Chand, S., 2015. Spatial distribution of TDS indrinking water of Tehsil Jampur using ordinary and Bayesiankriging. Pak. J. Stat. Oper. Res. 3, 377–386.

Akbarzadeh, A., Andrews, J., Golding, P., 2008. ‘Solar ponds’, insolar energy conversion and photo energy systems. In:Encyclopedia of life support systems (EOLSS), DevelopedUnder the Auspices of the UNESCO. EOLSS Publishers, Oxford,UK http://www.eolss.net.

Bouguecha, S.T., Aly, S.E., Al-Beirutty, M.H., Hamdi, M.M.,Boubakri, A., 2015. Solar driven DCMD: performanceevaluation and thermal energy efficiency. Chem. Eng. Res. Des.100, 331–340, http://dx.doi.org/10.1016/j.cherd.2015.05.044.

Boukhriss, M., Bacha, H.B., Zarzoum, K., Zhani, K., 2015. Study ofmodeling and simulation of direct contact membranedistillation. Int. J. Sci. Eng. Res. 6, 1317–1325.

Bui, V.A., Vu, L.T.T., Nguyen, M.H., 2010. Simulation andoptimisation of direct contact membrane distillation forenergy efficiency. Desalination 259, 29–31,http://dx.doi.org/10.1016/j.desal.2010.04.041.

Busquets, E., Kumar, V., Motta, J., Chacon, R., Lu, H., 2012. Thermalanalysis and measurement of a solar pond prototype to study

the non-convective zone salt gradient stability. Sol. Energy 86,1366–1377, http://dx.doi.org/10.1016/j.solener.2012.01.029.

236 chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237

Camacho, L.M., Dumée, L., Zhang, J., Li, J.-de, Duke, M., Gomez, J.,Gray, S., 2013. Advances in membrane distillation for waterdesalination and purification applications. Water 5, 94–196,http://dx.doi.org/10.3390/w5010094.

Cath, T.Y., Adams, V.D., Childress, A.E., 2004. Experimental studyof desalination using direct contact membrane distillation: anew approach to flux enhancement. J. Membr. Sci. 228, 5–16,http://dx.doi.org/10.1016/j.memsci.2003.09.006.

Cipollina, A., Sparti, M.G.Di., Tamburini, A., Micale, G., 2012.Chemical engineering research and design development of amembrane distillation module for solar energy seawaterdesalination. Chem. Eng. Res. Des. 90, 2101–2121,http://dx.doi.org/10.1016/j.cherd.2012.05.021.

Dah, M.M.O., Ouni, M., Guizani, A., Belghith, A., 2010. Theinfluence of the heat extraction mode on the performanceand stability of a mini solar pond. Appl. Energy 87, 3005–3010,http://dx.doi.org/10.1016/j.apenergy.2010.04.004.

Duong, H.C., Gray, S., Duke, M., Cath, T.Y., Nghiem, L.D., 2015.Scaling control during membrane distillation of coal seam gasreverse osmosis brine. J. Membr. Sci. 493, 673–682,http://dx.doi.org/10.1016/j.memsci.2015.07.038.

El-Bourawi, M.S., Ding, Z., Ma, R., Khayet, M., 2006. A frameworkfor better understanding membrane distillation separationprocess. J. Membr. Sci. 285, 4–29,http://dx.doi.org/10.1016/j.memsci.2006.08.002.

Fard, A.K., Manawi, Y., 2014. Seawater desalination for productionof highly pure water using a hydrophobic PTFE membraneand direct contact membrane distillation (DCMD). Int. J.Environ. Chem. Ecol. Geol. Geophys. Eng. 8, 398–406,http://scholar.waset.org/1999.6/9998440.

Gadiwala, M.S., Usman, A., Akhtar, M., Jamil, K., 2013. Empiricalmodels for the estimation of global solar radiation withsunshine hours on horizontal surface in various cities ofPakistan. Pak. J. Meteorol. 9, 43–49.

Gietas, M.C., Pina, H.L., Milhazes, J.P., Tavares, C., 2009. Solar pondmodeling with density and viscosity dependent ontemperature and salinity. Int. J. Heat Mass Transf. 52,2849–2857,http://dx.doi.org/10.1016/j.ijheatmasstransfer.2009.01.003.

Guillén-burrieza, E., Blanco, J., Zaragoza, G., Alarcón, D.,Palenzuela, P., Ibarra, M., Gernjak, W., 2011. Experimentalanalysis of an air gap membrane distillation solardesalination pilot system. J. Membr. Sci. 379, 386–396,http://dx.doi.org/10.1016/j.memsci.2011.06.009.

Hassan, A.S., Fath, H.E.S., 2013. Review and assessment of thenewly developed MD for desalination processes. Desalin.Water Treat. 51, 574–585,http://dx.doi.org/10.1080/19443994.2012.697273.

He, K., Hwang, H.J., Woo, M.W., Moon, I.S., 2011. Production ofdrinking water from saline water by direct contact membranedistillation (DCMD). J. Ind. Eng. Chem. 17, 41–48,http://dx.doi.org/10.1016/j.jiec.2010.10.007.

Hwang, H.J., He, K., Gray, S., Zhang, J., Moon, I.S., 2011. Directcontact membrane distillation (DCMD): experimental studyon the commercial PTFE membrane and modeling. J. Membr.Sci. 371, 90–98, http://dx.doi.org/10.1016/j.memsci.2011.01.020.

Jaefarzadeh, M.R., Akbarzadeh, A., 2002. Towards the design oflow maintenance salinity gradient solar ponds. Sol. Energy 73,375–384, http://dx.doi.org/10.1016/S0038-092X(02)00114-7.

Jaefarzadeh, M.R., 2006. Heat extraction from a salinity-gradientsolar pond using in pond heat exchanger. App. Therm. Eng.26, 1858–1865,http://dx.doi.org/10.1016/j.applthermaleng.2006.01.022.

Ji, X., Curcio, E., Al Obaidani, S., Di Profio, G., Fontananova, E.,Drioli, E., 2010. Membrane distillation-crystallization ofseawater reverse osmosis brines. Sep. Purif. Technol. 71,76–82, http://dx.doi.org/10.1016/j.seppur.2009.11.004.

John, H.L., Walton, C., 2001. Desalination coupled to salinitygradient solar ponds. Desalination 136, 13–23.

Kahlown, M.A., Tahira, M.A., Sheikh, A.A., 2004. Water QualityStatus in Pakistan, Second Report 2001-2003. Pakistan Council

of Research in Water Resources (PCRWR), Ministry of Scienceand Technology, Islamabad, Pakistan.

Khanzada, N.K., Khan, S.J., Davies, P.A., 2017. Performanceevaluation of reverse osmosis (RO) pre-treatment technologiesfor in-land brackish water treatment. Desalination 406, 44–50,http://dx.doi.org/10.1016/j.desal.2016.06.030.

Kumar, A., Kishore, V.V.N., 1999. Construction and operationalexperience of a 6000 m2 solar pond at Kutch, India. Sol. Energy65, 237–249.

Lawson, K.W., Lloyd, D.R., 1997. Membrane distillation. J. Membr.Sci. 124, 1–25 (Review).

Leblanc, J., Akbarzadeh, A., Andrews, J., Lu, H., Golding, P., 2011.Heat extraction methods from salinity-gradient solar pondsand introduction of a novel system of heat extraction forimproved efficiency. Sol. Energy 85, 3103–3142,http://dx.doi.org/10.1016/j.solener.2010.06.005.

Lu, H., Walton, J.C., Swift, A.H.P., 2001. Desalination coupled withsalinity-gradient solar ponds. Desalination 136, 13–23,http://dx.doi.org/10.1016/S0011-9164(01)00160-6.

Lu, H., Walton, J.C., Hein, H., 2002. Thermal Desalination UsingMEMS and Salinity-Gradient Solar Pond Technology.University of Texas at El Paso, El Paso Texas.

Lu, H., Swift, A.H.P., Hein, H.D., Walton, J.C., 2004. Advancementsin salinity gradient solar pond technology based on sixteenyears of operational experience. J. Sol. Energy Eng. 126, 759,http://dx.doi.org/10.1115/1.1667977.

Martinetti, C.R., Childress, A.E., Cath, T.Y., 2009. High recovery ofconcentrated RO brines using forward osmosis andmembrane distillation. J. Membr. Sci. 331, 31–39,http://dx.doi.org/10.1016/j.memsci.2009.01.003.

Matti, K., Philip, J.W., Hans de, M., Olli, V., 2010. Is physical waterscarcity a new phenomenon? Global assessment of watershortage over the last two millennia. Environ. Res. Lett. 5,34006, http://dx.doi.org/10.1088/1748-9326/5/3/034006.

Nakoa, K., Rahaoui, K., Date, A., Akbarzadeh, A., 2015.Sciencedirect an experimental review on coupling of solarpond with membrane distillation. Sol. Energy 119, 319–331,http://dx.doi.org/10.1016/j.solener.2015.06.010.

Newell, T.A., Cowie, R.G., Upper, J.M., Smith, M.K., Cler, G.L., 2000.Construction and operation activities at the university ofillinois salt gradient solar pond. Sol. Energy 45, 231–239.

Nghiem, L.D., Hildinger, F., Hai, F.I., Cath, T., 2011. Treatment ofsaline aqueous solutions using direct contact membranedistillation. Desalin. Water Treat. 32, 234–241,http://dx.doi.org/10.5004/dwt.2011.2705.

Nie, Z., Bu, L., Zheng, M., Huang, W., 2011. Experimental study ofnatural brine solar ponds in Tibet. Sol. Energy 85, 1537–1542,http://dx.doi.org/10.1016/j.solener.2011.04.011.

Pantoja, C., Pantoja, C.E., Nariyoshi, Y.N., Seckler, M.M., 2015.Membrane distillation crystallization applied to brinedesalination: a hierarchical design procedure. J. Ind. Eng.Chem. 54, 2776–2793, http://dx.doi.org/10.1021/ie504695p.

Pérez-González, A., Urtiaga, A.M., Ibánez, R., Ortiz, I., 2012. Stateof the art and review on the treatment technologies of waterreverse osmosis concentrates. Water Res. 46, 267–283,http://dx.doi.org/10.1016/j.watres.2011.10.046.

Qtaishat, M.R., Banat, F., 2013. Desalination by solar poweredmembrane distillation systems. Desalination 308, 186–197,http://dx.doi.org/10.1016/j.desal.2012.01.021.

Rizvi, R., Jamal, Y., Ghauri, M.B., Salman, R., Khan, I., 2015. Solarpond technology for brine management and heat extraction: acritical review. J. Fac. Eng. Technol. 22, 69–79.

Rosado, D., Bernaola, F.J., 2014. Comparative study of brinemanagement technologies for desalination plants.Desalination 336, 32–49,http://dx.doi.org/10.1016/j.desal.2013.12.038.

Saleh, A., Qudeiri, J.A., Al-Nimr, M.A., 2011. Performanceinvestigation of a salt gradient solar pond coupled withdesalination facility near the Dead Sea. Energy 36, 922–931,http://dx.doi.org/10.1016/j.energy.2010.12.018.

Seckler, D., Barker, R., Amarasinghe, U., 1999. Water scarcity inthe twenty-first century. Int. J. Water Resour. Dev. 15, 29–42,http://dx.doi.org/10.1080/07900629948916.

Shannon, M.A., Bohn, P.W., Elimelech, M., Georgiadis, J.G.,Marinas, B.J., Mayes, A.M., 2008. Science and technology for

chemical engineering research and design 1 1 8 ( 2 0 1 7 ) 226–237 237

S

S

S

T

T

membrane distillation. J. Membr. Sci. 497, 239–247,http://dx.doi.org/10.1016/j.memsci.2015.09.038.

water purification in the coming decades. Nature 452,301–310, http://dx.doi.org/10.1038/nature06599.

uarez, F., Ruskowitz, J.A., Tyler, S.W., Childress, A.E., 2015.Renewable water: direct contact membrane distillationcoupled with solar ponds. Appl. Energy 158, 532–539,http://dx.doi.org/10.1016/j.apenergy.2015.08.110.

uarez, F., Tyler, S.W., Childress, A.E., 2010. A theoretical study ofa direct contact membrane distillation system coupled to asalt-gradient solar pond for terminal lakes reclamation. WaterRes. 44, 4601–4615,http://dx.doi.org/10.1016/j.watres.2010.05.050.

ubramani, A., Jacangelo, J.G., 2014. Treatment technologies forreverse osmosis concentrate volume minimization: a review.Sep. Purif. Technol. 122, 472–489,http://dx.doi.org/10.1016/j.seppur.2013.12.004.

abor, H.Z., Doron, B., 1990. The Beith Ha’Arava 5 MW solar pondpower plant. Progress report. Sol. Energy 45, 247–253.

un, C.M., Groth, A.M., 2011. Sustainable integrated membranecontactor process for water reclamation, sodium sulfate salt

and energy recovery from industrial effluent. Desalination283, 187–192, http://dx.doi.org/10.1016/j.desal.2011.03.054.

Tundee, S., Terdtoon, P., Sakulchangsatjatai, P., Singh, R., 2010.Heat extraction from salinity-gradient solar ponds using heatpipe heat exchangers. Sol. Energy 84, 1706–1716,http://dx.doi.org/10.1016/j.solener.2010.04.010.

WHO, 2011. Guidelines for Drinking-Water Quality, 4th ed. WorldHealth Organization (WHO).

Zangrando, F., 1980. A simple method to establish salt gradientsolar ponds. Sol. Energy 25, 467–470.

Zhu, A.H., Rahardianto, A., Christofides, P.D., Cohen, Y., 2010.Reverse osmosis desalination with high permeabilitymembranes—cost optimization and research needs. Desalin.Water Treat. 15, 256–266,http://dx.doi.org/10.5004/dwt.2010.1763.

Zuo, J., Bonyadi, S., Chung, T.S., 2016. Exploring the potential ofcommercial polyethylene membranes for desalination by