lable at ScienceDirect

Journal of Cleaner Production 92 (2015) 343e353

Contents lists avai

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

Performance enhancement of a Flat Plate Solar collector usingTitanium dioxide nanofluid and Polyethylene Glycol dispersant

Z. Said a, e, M.A. Sabiha a, R. Saidur a, b, *, A. Hepbasli c, N.A. Rahim b, S. Mekhilef d,T.A. Ward a

a Department of Mechanical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysiab UM Power Energy Dedicated Advanced Centre (UMPEDAC), Level 4, Wisma R & D, University of Malaya, 50603 Kuala Lumpur, Malaysiac Department of Energy Systems Engineering, Faculty of Engineering, Yasar University, 35100 Izmir, Turkeyd Department of Electrical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysiae Department of Engineering Systems and Management, Masdar Institute of Science and Technology, PO Box 54224, Abu Dhabi, United Arab Emirates

a r t i c l e i n f o

Article history:Received 26 November 2014Accepted 3 January 2015Available online 17 January 2015

Keywords:NanofluidPressure dropEnergyExergyThermal conductivityTiO2

* Corresponding author. Department of Mechanicagineering, University of Malaya, 50603 Kuala Lumpu7611; fax: þ60 3 7967 5317.

E-mail addresses: saidur@um.edu.my, saidur912@y

http://dx.doi.org/10.1016/j.jclepro.2015.01.0070959-6526/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

The use of TiO2ewater nanofluid as a working fluid for enhancing the performance of a flat plate solarcollector has been studied. The volume fraction of the nanoparticles was 0.1% and 0.3% respectively,while the mass flow rates of the nanofluid varied from 0.5 to 1.5 kg/min, respectively. Thermo-physicalproperties and reduced sedimentation for TiO2-nanofluid was obtained using PEG 400 dispersant. Theresults reveal the impact and importance of each of these parameters. Energy efficiency increased by76.6% for 0.1% volume fraction and 0.5 kg/min flow rate, whereas the highest exergy efficiency achievedwas 16.9% for 0.1% volume fraction and 0.5 kg/min flow rate. Results showed that the pressure drop andpumping power of TiO2 nanofluid was very close to the base fluid for the studied volume fractions.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

The fast development in population and economy accomplish-ments in tropical countries has resulted in an increase in energyintake, which accelerates the reduction of available energy re-sources. The reduction of conventional energy sources andincreasing concerns on global warming are supporting the use ofrenewable and clean energy sources (Davis andMartín, 2014). Solarthermal energy is considered as one of the cleanest energy sources(Koroneos and Nanaki, 2012) and the production cost of solar en-ergy is reducing with advances in technology (Fu et al., 2015). Solarcollectors are the main components of solar energy system whichtransform solar radiation into heat which is transferred to a me-dium. Various types of solar collectors are available, but the mostproductive and commonly used is flat plate solar collector (FPSC).FPSC is cheaper and based on simpler technology, but it hascomparatively low efficiency and outlet temperatures. There areseveral conventional approaches to improve the collector's

l Engineering, Faculty of En-r, Malaysia. Tel.: þ60 3 7967

ahoo.com (R. Saidur).

efficiency. For example, the size can be optimized or the glazingmaterials altered, but these modifications are often inconvenient.One of the most convenient and effective methods to increase theefficiency is to substitute pure water (acting as the functioningfluid) with a higher thermal conductivity fluid. Solid nanoparticlescan be suspended in a base fluid in order to obtain a high thermalconductivity. These fluids, known as ‘nanofluids’, can greatlyenhance the heat transfer performance of conventional fluids(Yousefi et al., 2012a).

The highest possible useful work during a process, in which thesystemwith a heat reservoir is brought into equilibrium, is definedas exergy of a system (Nguyen et al., 2014). Irreversibility in aprocess can reduce the exergy of a system. To optimize complexthermodynamic systems, an exergy analysis is used. The termexergy was proposed in 1956 by Rant, but the idea was not estab-lished until 1873 by Gibbs (1873). Exergy analysis has become apowerful method to evaluate the efficiency of thermodynamicalsystems by reducing energy related system losses, while maxi-mizing energy savings and significantly help society to move to-wards sustainable development (Xydis, 2013).

Solar radiation which converted to heat is projected to delivernearly 20% of overall heat demand by 2030 and 50% by 2050 in theEuropean Union (EU). According to ‘Vision 2030’ by ESTTP

Abbreviations and nomenclature

Ac collector area, m2

Cp specific heat, J/kg Kd diameter of pipe, m_Exin exergy rate at inlet, W_Exout exergy rate at outlet, W_Exdest rate of irreversibility, W_Exheat exergy rate received from solar radiation, W_Exwork exergy output rate from the system, W_Exmass;in Exergy rate associated with mass at inlet, W_Exmass;out exergy rate associated with mass at outlet, W_Sgen entropy generation rate, W/K_Qsun;in energy gain rate, Wt shear stress;kp thermal conductivity of nanoparticle, W/m KK loss coefficient (dimensionless)_m mass flow rate, kg/s_W work rate or power, Wq convective heat transfer rate, Wk thermal conductivity, W/m K

_Qo heat loss rate to the ambient, W_Qs energy rate engrossed, WTa ambient temperature, KR ideal gas constant, JK�1 mol�1

hin specific enthalpy at inlet, J/kghout specific enthalpy at outlet, J/kgm coefficient of viscosityTs sun temperature, KTsur surrounding/ambient temperature, Km viscosity, N s/m2

t transmittancea absorptance4 nanoparticles volume fraction, %sa entropy generation to surrounding, J/kg Ksin entropy generation at inlet, J/kg Ksout entropy generation at outlet, J/kg Kr density, kg/m3

s overall entropy production, J/kg Kf friction factorh specific enthalpy, J/kg_g shear strain rate.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353344

(European Solar Thermal technology Platform), they will designand establish “Active Solar Building” as a standard of new buildingswhere solar energy will meet 100% of their heating and coolingdemand and “Active Solar Renovation” as a standard for therefurbishment of existing buildings where solar thermal energywill meet at least 50% of heating and cooling demand of therenovated buildings (ESTIF, 2011). Solar radiation can also be con-verted to electricity by photovoltaic technology which aims toreduce 30% overall emission in Netherland by 2020 (Vasseur et al.,2013). Energy performance can be enhanced by using a hybridphotovoltaic and thermal (PVT) collector tool, which uses water as acoolant. The electrical and thermal efficiencies of both can beincreased by improving the thermal contact between the thermalabsorber and the PV module (He et al., 2006). Solar collectors andthermal energy storage systems are two important applications ofthis technology (Tian and Zhao, 2013).

Several articles examine the prospects of improving collectorefficiency by using nanofluids on FPSC. According toDuangthongsuk and Wongwises (2009), most of the solids havehigher thermal conductivity than that of conventional coolingfluids (e.g. water, oil and ethylene). Otanicar and Golden (2009)compared the environmental and economic influences of conven-tional and nanofluid solar collectors. Otanicar et al. (2010) con-ducted experiment on direct absorption solar collector usingdifferent nanofluids. They used nanofluids as absorption mediumand found that the efficiency improved by 5% in a solar thermalcollector. The experimental result was also compared with a nu-merical model. A novel approach for utilizing nanofluids to improvethe energy density and to enhance heat transfer from solar collectorto storage tanks was proposed by Das et al. (2008). Natarjan andSathish (2009) also used nanofluids in place of conventional heattransfer fluid to enhance the efficiency of solar water heater. Tiwariet al. (2013) used a Al2O3 nanofluid to improve the thermal effi-ciency of a solar collector by 31.64%, when using an optimum Al2O3particle volume fraction of 1.5%. A similar experiment was done byYousefi et al. (2012b) to investigate the effect of Al2O3eH2O nano-fluid on the efficiency of FPSC. They improved the efficiency of asolar collector by 28.3%, while using a volume fraction of 0.2% Al2O3

in their nanofluid. Yousefi et al. (2012b) increased the efficiency oftheir solar collector by 28.3% with a Al2O3 nanofluid; and 35% usinga nanofluid based on multi wall carbon nanotubes based.

Gangadevi et al. (2013) conducted experiments that showedthat the thermal efficiency of a FPSC could be improved byincreasing the nanoparticle volume fraction and therefore thenanofluid's thermal conductivity. Efficiency increased to about 30%by using Al2O3 water nanofluid as aworkingmedium. Masuda et al.(1993) studied the thermo physical properties of (Al2O3 and TiO2)immersed in water. They reported that the thermal conductivity ofAl2O3 water nanofluids and TiO2 water nanofluids at a volumefraction of 4.3% are about 32% and 11% higher than pure water,respectively.

Numerous additional authors (Reddy et al., 2012) (Turgut et al.,2009; Wamkam et al., 2011; Xie et al., 2010) have studied the effectof increasing the thermal conductivity by increasing the concen-tration of nanoparticles in fluids. Wang et al. (2007) showed thatthe thermal conductivity of nanofluids can be improved byreducing the nanoparticle's size. Viscosity is directly related topumping power and pressure drop for a solar collector. Demir et al.(2011) obtained numerical results showing that a high concentra-tion of nanoparticles (Al2O3 or TiO2) in water yielded higher heattransfer rates and pressure drops. Fotukian and Nasr Esfahany(2010) experimentally examined the influence of dilute Al2O3/wa-ter nanofluid on the heat transfer and the pressure drop inside acircular tube under turbulent flow conditions. His work showedthat increasing the volume of nanoparticles to a base fluid,improved the heat transfer coefficient. However, addition ofnanoparticles also increased the viscosity of fluid, requiring greaterenergy for pumping the fluid in the circuits. This consequentlydecreased the improvement in heat transfer efficiency.

Research has been conducted using Al2O3 based nanofluids, butlimited studies have been carried out on TiO2 nanofluids. Consid-ering the increasing global demand for solar energy, it is veryimportant to find new, effective, and convenient approaches toimprove the efficiency rate of a FPSC. Thus, the aim of this experi-ment is to study the effect of TiO2 nanofluid as a working fluid andits effect on performance.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353 345

2. Hypothetical background

All the analysis has been carried under steady-state and steady-flow conditions. This investigation begins by analyzing solar waterheating (SWH) using the first and second laws of thermodynamicsapplied to varying absorbing surfaces; in other words, an assess-ment from the energetic and exergetic points of view.

2.1. First law of thermodynamics (energy) analysis

A theoretical model can be derived using a thermal energybalance equation (shown in Eq. (1)).

MpCp�dTp;ave

�dt�þ _mCpðTout � TinÞ ¼ h0IAc � Uc

�Tp;ave � Te

�Ac

(1)

The thermal efficiency of the FPSC (h) is therefore:

h ¼ _mCpðTout � TinÞ=IAc (2)

2.2. Second law of thermodynamics (exergy) analysis

The exergy efficiency (hex) can be derived from the first andsecond laws of thermodynamics. This is a steady flow system thatoperates under steady state conditions. Both the potential and ki-netic energy are negligible. The thermo physical properties ofnanofluids flowing in and out of the collector are constant. Thework transfer from the system and the heat transfer to the systemare positive. The loss coefficient only accounts for the entry effect.

Therefore the general exergy balance for a steady state andsteady flow process is:

_Exheat � _Exwork � _Exmass;in � _Exmass;out ¼ _Exdest (3)

Substituting terms into this equation yields:

X�1� Ta

Tsur

�_Qs� _Wþ

X_minJin�

X_moutJout ¼ _Exdest (4)

X�1� Ta

Tsur

�_Qs � _m½ðhout � hinÞ � Taðsout � sinÞ� ¼ _Exdest (5)

where _Qs is the total rate of the exergy received from the solarradiation by the collector absorber area (Esen, 2008);

_Qs ¼ IT ðtaÞAc ¼ S$Ac (6)

The change in enthalpy and entropy of the nanofluid in thecollector is:

Dh ¼ hout � hin ¼ Cp;nf�Tf ;out � Tf ;in

�(7)

Ds ¼ sout � sin ¼ Cp;nf lnTf ;outTf ;in

� R lnPoutPin

(8)

Replacing Eqs. (6)e(8) in Eq. (5) yields:

�1� Ta

Tsur

�IT ðtaÞAc � _mCp;nf

�Tf ;out � Tf ;in

�þ _mCp;nf Ta ln

Tf ;outTf ;in

� _mRTa lnPoutPin

¼ _Exdest

(9)where _Exdest is the exergy loss (or irreversibility) rate defined as:

_Exdest ¼ Ta _Sgen (10)

Therefore, the exergy efficiency is given by the followingequation:

hex ¼ 1� Ta _Sgen½1� ðTa=TsÞ� _Qs

(11)

2.3. Pumping power and pressure drop

A pump is needed to circulate nanofluids all theway through thesystem. The pumping power and pressure drop were calculatedusing Eq. (12).

Dp ¼ frV2

2Dldþ K

rV2

2(12)

The loss coefficient K is calculated using formulas that consist ofthe density and kinematic viscosity of the heat transfer fluid.Pumping power is calculated by Eq. (13) (Said et al., 2013b):

Pumping power ¼

_mrnf

!� Dp (13)

3. Methodology

Commercial spherical shape TiO2 (99.5% trace metal, SigmaAldrich, Malaysia) having an average diameter of ~21 nm wereused. Polyethylene glycol 400 (PEG) (Sigma Aldrich) was used as asurfactant. Distilled water was used as a base fluid.

3.1. Stability characterization and data collection

Nanoparticles were suspended in base fluid to reduce the ag-gregation of TiO2. Two methods were executed in this study. Thefirst method used polyethylene glycol 400 (PEG, Sigma Aldrich) as adispersant. The second method used a highly pressurized homog-enizer (capacity of up to 1700 bar) to optimize the dispersion ofnanoparticles (0.1 vol% and 0.3 vol%) into the distilled water basefluid. The homogenizer dispersed the well-isolated primary parti-cles (Bobbo et al., 2012). The TiO2 nanoparticles with 0.1 vol% and0.3 vol% were added to the base fluid to obtain a homogenouslydispersed solution, after passing the solution through 30 cycles in ahigh pressure homogenizer. For thewatereTiO2 nanofluid, the ratiobetween the nanoparticles and the dispersant mass was 1:2.

The morphological characterization of the nanoparticles wasaccomplished by a field emission scanning electron microscopy.The SEM pictures of TiO2 are shown in Fig. 4. The actual dimensionsof the TiO2 nanoparticles are between 20 and 40 nm. A Zeta-seizerNano ZS was employed to examine the average size of the nano-particles in the base medium, along with the zeta potential value.The stability time of TiO2eH2O is further supported by visual im-ages presented in this paper. The density of the nanofluid wasdetermined using the DA-130N Density Meter (Kyoto Electronics).The thermal conductivity of the nanofluid was measured using aKD2 Pro thermal property analyzer (Decagon Devices).

3.2. Location of the study

The solar collector was investigated at the University Malaya,Malaysia for the experiment. The tilt angle has major impact on theamount of energy that the system can capture, for non-trackingsolar collectors. The tilt angle used for this experiment is 22�,which is the optimal tilt angle to obtain the highest average dailyradiation.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353346

3.3. Experimental approach

The schematics of the solar collector and the experiment arepresented in Figs. 1 and 2, respectively. The dimensions of the solarcollectors are listed in Table 1. Table 1 also includes the physicalproperties of the TiO2 nanoparticles and water used in the calcu-lations. Table 2 presents the specifications of the FPSC that arestudied in this experiment. For the force convection system, anelectric pump is used in the solar water heating system. Fig. 2shows that the tank with a capacity of 50 L absorbs the heat loadform the collector cycle. The heat load of the solar cycle is trans-ferred to the water by a heat exchanger, which is used outside thetank. Fig. 2 illustrates that a flow meter is linked to the water pipebefore the electric pump.

A simple valve is connected after the electric pump, whichregulates the mass flow rate of the working fluid. Five K thermo-couples are used to assess the fluid temperatures at: the inlet andoutlet of the solar collector, the left and right panels of the solarcollector, and the outside environment. A 10 channel data logger isconnected to the sensors. The Li-COR Pyranometer (TES 1333R SolarPower Meter) was used to record the total solar radiation whereasPROVA (AVM-07) Anemometer was used to record the speed of thewind. A pressure sensor was used to measure the pressure differ-ence between the inlet and outlet of the solar collector. All of thedata were later transferred into the computer via USB interfaces.Measurements were taken multiple times and averaged to reduceexperimental error.

3.4. Testing process

The ASHARE Standard 93-2003 (Standard, 1977) was used toevaluate the thermal performance of FPSC. The incident radiation,ambient temperature, outlet fluid temperature and inlet fluidtemperature are used to obtain the thermal performance of thesolar collector.

3.5. Inaccuracy analysis

Two potential sources of error were considered. The first errorarises from the direct measurement parameters such as:DGc, DT ,DP, and m. The second error arises from indirect measurementssuch as: energy and exergy efficiencies. The subsequent relations

Fig. 1. Experimental setup for the study conducted

are derived from the Luminosu et al. (Luminosu and Fara, 2005)method:

Dhex ¼DI:

E:xheat

þ I:E:xheat

E:x2heat

(14)

and

Dhen ¼ Dqa:

Gcþ qa

:DGc

G2c

(15)

where each error component can be evaluated through thefollowing relations:

DExheat ¼ DTTs

þ TaDTT2s

!AcðtaÞGc þ

�1� Ta

Ts

�AcðtaÞDGc (16)

DI:¼ TaDS

:

gen þ S:

genDT (17)

DS:

gen ¼�R ln

PoutPin

þCp lnTinTout

þCpTout þTin

Ta

�Dm

:

þGcAcðtaÞDTT2a

þm:Cp

1

Toutþ 1Tin

þ 2Ta

þðTout þTinÞT2a

!DT

þm:R�

1Pout

þ 1Pin

�DPþAcðtaÞ

�1Ts

þ 1Ta

�DGc

(18)

Dq:

a ¼ Cp

Dm

: ðTout þ TinÞ þ 2m:DT

Ac

(19)

4. Results and discussion

The errors are ±2.25 W/m2, ±1 �C, ±0.00144 kg/min and ±1.1%,in measuring the solar radiation, temperature, pressure and massflow rate, respectively. Hence, the maximum errors (uncertainties)associated with energy and exergy efficiencies are estimated to be±0.04 and ±0.14 using Eqs. (14) and (15).

: (a) Front part & (b) Back part; photographs.

Fig. 2. The presentation of the experimental setup in schematic diagram.

Table 2Specification and environmental conditions for the FPSC.

Parameters of collector Value

Frame Aluminum alloyGlazing 4 mm tempered texture glass

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353 347

4.1. Nanofluids stability characterization

Fig. 3 displays the particle size distribution according to theintensity obtained from the Zeta-seizer for 30 days. Fig. 4 presentsSEM and TEM images for characterization of the nanoparticles. Thebest appropriate preparation technique and surfactant type andconcentration have been used in order to optimise the stablenanofluids by examining the average size distribution of the par-ticles with respect to time. In fact, after dispersing in the base fluids,these nanoparticles can aggregate and settle down, resulting inlower stability of the nanofluid and therefore affecting theirapplication. On the other hand, the surrounding face of the TiO2nanoparticles is adsorbed by the PEG molecules, creating a com-pressed layer around the particles. Steric effects stabilize the fluidand result in the formation of more dense aggregates (Alphonseet al., 2009).

Fig. 5 shows the visual appearances of the nanofluids with nosign of aggregation for a period of a month.

Since these fluids will likely be used in a FPSC, where forcedcirculation takes place resulting in a continuous mixing condition,the settling effects of the nanoparticles are negligible. Measure-ments revealed that the nanoparticles were larger than the sizelisted by the supplier. The measured average diameter is 127 nmand 136 nm for of 0.1vol% and 0.3vol%, respectively. This shows atendency of Titania particles in liquid media to aggregate, howeverthey are still distinctly nanometric.

Table 1Physical characteristic of TiO2 and base fluid (Bayat and Nikseresht, 2012; Kamyaret al., 2012; Said et al., 2013a; Sridhara and Satapathy, 2011).

Particle &base fluid

Average particlesize (nm)

Actual density(kg/m3)

Cp(J/kg K) K (W/mK) Viscosity(mPa s)

TiO2 21 4230 692 8.4Water 997.1 4179 0.605 0.89

Nanofluids with an average particle size of 225.9 nm were ob-tained for the samples after 30 days, with high zeta potential valueof 41.8 mV (presented in Table 3). Zeta potential higher than 30 mVwas obtained for all the samples prepared.

4.2. Density of TiO2 nanofluid

The density of the nanofluids is proportional to the volume ratioof nanoparticles (solid) and base fluid (liquid) in a system. The basefluid also plays a significant role in the density of the nanofluids,whereas the other parameters, such as nanoparticles shape, size,zeta potential and additives, do not affect the density of thenanofluids. Pak and Cho (1998) conducted an experiment (at asingle temperature of 25 �C) for Al2O3 and TiO2 nanofluids up to 4vol% to prove the mixing theory. Not enough density measurementdatawas available for various nanofluids at varying temperatures inthe literature. Therefore, comprehensive measurements were car-ried out to obtain this data, as well as verify the applicability of Eq.

Working fluids in flow ducts Water and TiO2 based nanofluidAbsorption area, Ap 1.84 m2

Wind speed 5 m/sCollector tilt, bo 22�

Absorption rate 0.94Emittance 0.12Heat transfer coefficient 4.398Header material Copper TP2Header tube size F22 mm � t0.6 mm (2 pcs)Riser tube material Copper TP2Riser tube size F10 mm � t0.45 mm (8 pcs)

Fig. 3. Size of watereTiO2 0.1% nanofluids with Polyethylene glycol 400, for a period of30 days.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353348

(20). Density with respect to concentration and temperature arepresented in Fig. 6 and Fig. 7.

rnf ¼�mV

�nf

¼ mf þmp

Vf þ Vp¼ rf Vf þ rpVp

Vf þ Vp¼ �1� fp

�rbf þ fprp

(20)

where, 4p ¼ Vp=Vf þ Vp is the volume fraction of the nanoparticles.The density of the nanofluids of TiO2 using distilled water as the

base fluid was measured at different temperature and differentconcentrations (vol%). It was found that the density decreases aseither temperature or volume concentration increases. The densityvalues at 25 �C of TiO2 nanofluid showed an outstanding agreementwith Pak and Cho equation (1998). The highest deviance betweenthe experimental values and the model is 0.22% for TiO2, which isvery small and below the minimum acceptable limit (1%).

4.3. Thermal conductivity

Previous research shows that the stability of nanoparticles re-duces when the concentration of nanoparticles increases, becausethe nanoparticles tend to agglomerate (Said et al., 2013b) (Saidet al., 2013a) (Vatanpour et al., 2011) (Said et al., 2013c). In orderto get stability for a longer period of time, nanoparticles with alower volume fraction can be immersed into a base fluid, which is

Fig. 4. (a) SEM image of TiO2 nanoparticles. (b) TE

comparatively more stable than the nanofluids with higher volumefraction.

Fig. 8 shows the improvement of thermal conductivity at avolume fraction of 0.1%e0.3%, respectively using TiO2eH2O nano-fluids. The figure shows that thermal conductivity rises withincreased volume concentration and also temperature. These re-sults closely matched with the results obtained by Fedele et al.(2012). The experimental outcomes demonstrated a peak in theimprovement factor in this range of volume fractions in the tem-perature range calculated, which suggests that an optimum sizeexists for dissimilar nanoparticle and base fluid mixtures. Thisoccurrence can be neither projected nor clarified using the hypo-thetical models presently existing in the literature.

The thermal conductivity of 0.3 vol % of TiO2 enhances up to 6%and from the measured data, the thermal conductivity is found tobe directly proportional to the volume fraction. The thermal con-ductivity surges to 6% from2.4%. Therefore, the correlation betweenthe volume fraction and the thermal conductivity is directly pro-portional. The values used in the experimental results presentedare confirmed with other researchers in Fig. 8 (Chon and Kihm,2005; Das et al., 2003; Fedele et al., 2012; Mintsa et al., 2009) (Liand Peterson, 2006). To get more heat rate from the solar collec-tor, it is necessary to have higher thermal conductivity, which ispossible when the concentration of nano particles is higher. It wasalso observed that the difference between the inlet and the outlet ishigher comparing to water which is the ultimate result of theincreasing volume fraction. The heat loss is less from the collectorusing nanofluid than water and thus the nanofluid can be used toobtain as much higher heat transfer rate. From Fig. 8, it is obviousthat the increased volume fraction and the rising temperaturesincrease the thermal conductivity as well. In conventional collec-tors, the increasing fluid temperature may cause an increase in theheat loss.

4.4. Viscosity of TiO2 nanofluid

In this content, an important matter is to achieve nanofluidphysical information, and one of the possible techniques is througha comprehensive rheological investigation (Chen and Ding, 2009).In this experiment, three kinds of studies, namely viscosity as afunction of shear rate, temperature as well as at different massconcentrations, have been conducted.

Uncertainty in experimental data is illustrated in Fig. 9. Theexperimental values of the viscosity of the nanofluid and the

M image of TiO2/water using control pH ¼ 9.

Fig. 5. Prepared TiO2 nanofluid solutions (a) Samples on the first day of preparations (b) Samples after 30 days of preparations.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353 349

ASHARE data match fairly with a highest deviation of ±2.0% withtemperature ranging from 0 �C to 80 �C. The machine error resultsin the deviation at low shear rates.

The governing equation for the Newtonian behavior of a fluid isrepresented by

t ¼ m _g (21)

The relation between the shear stress and the shear rate is linearin a Newtonian fluid, passing through the origin; the constant ofproportionality is called the coefficient of viscosity. In a non-Newtonian fluid, the line never passes through the origin. From55 �C, the TiO2 nanofluid with 0.1 vol% was Newtonian, whereasbelow this temperature, TiO2 nanofluid with 0.1 vol% was non-Newtonian. For 0.3 vol% of TiO2, a Newtonian behavior wasobserved in the entire temperature range in Fig. 10. It is noticedfrom the results that the viscosity for the fresh samples and that of

Table 3Zeta potential, particle size and pH values of TiO2/water 21 nm particles suspendedin water.

Nominalparticle size

Zetapotential (mV)

Particle size (nm)from DLS usinghigh pressurehomogenizer

No. of days

21 nm 48.6 126.9 1st21 nm 46.2 160.7 7th21 nm 45.8 202.0 21st21 nm 41.8 225.9 30th

Fig. 6. Density vs. concentration graph of TiO2ewater nanofluid at differentconcentrations.

the samples after running in FPSC are not the same anymore. Thedeviation among the values for the similar volume fractions isobtained. This finding is not explained anywhere else in the liter-ature, therefore opening scope for new research field of nanofluidsand their effects on applications in the long run.

4.5. Time range for experiment

The hourly time range of the experiment was 9:00 am to5:00 pm (MST) in order to utilize the availability of the sun inMalaysia. To minimize the variance of solar radiation, the experi-ment was done in consecutive days. Solar radiation data (on both aclear day and a cloudy day) are included in Fig. 11.

Numerous test runs were performed by distributing all thepresented data. Quasi steady conditions were used in each test run(divided into several test periods) (Ucar and Inallı, 2006).

4.6. Mass flow rate

The mass flow rate of a nanofluid was varied using a flowmeter.Nanofluids with 0.1 vol% and 0.3 vol% nanoparticle were used asworking fluids. Experiments were performed at mass flow rates of0.5, 1.0 and 1.5 kg/min to investigate the efficiency of the collector.

Fig. 7. Density vs. temperature graph of TiO2ewater nanofluid at differentconcentrations.

Fig. 8. Thermal conductivity of TiO2eH2O nanofluid with different volume fractionand changing temperature. Fig. 10. Viscosity of TiO2ewater nanofluid with different volume concentration and

different temperature.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353350

The results, presented in Figs. 12e16, show that the mass flow ratehas an impact on the efficiency of the collector. The relationshipbetween solar energy and the mass flow rate is shown in Eq. (22).

_Qu ¼ _mCpðTout � TinÞ (22)

Eq. (22) shows that the solar energy is directly proportional tothe mass flow rate.

Fig. 11. Solar radiation readings for a clear day and on a cloudy day.

4.7. Energy and exergy efficiencies

Fig. 12 shows the energy and exergy efficiency of the FPSC. Thiswas determined by varying the mass flow rate of the nanofluid (forvarious volume fractions) by 0.5, 1.0 and 1.5 kg/min. Each investi-gation was repeated for several days to determine measurementerror. This efficiency was evaluated using Eqs. (2) and (11),respectively and input from Tables 2 and 3.

From Fig. 12 it can be seen that the efficiency of the solar col-lector decreases as mass flow rate reduces. However, the efficiencyof the solar collector increases as the volume fraction of thenanoparticles increases. A similar result has been reported in a pastjournal article (Otanicar and Golden, 2009). This increase in effi-ciency is due to the increase in the nanofluid's thermal conduc-tivity, which elevates the convective heat transfer coefficient. Theenergy efficiency for the TiO2eH2O nanofluid increased by 76.6% for0.1 vol% and 0.5 kg/min, whereas energy efficiency increased by67.9% for 0.3 vol% and 0.5 kg/min. For water the highest energyefficiency that was obtained was 42.1% for a flow rate of 0.5 kg/min.As it seen in Fig. 12 reducing the flow rate to approximately 0.5 kg/min causes a considerable reduction in the absorber plate's tem-perature. The decrease in temperature gradient between the

Fig. 9. Uncertainty in viscosity of TiO2ewater nanofluid with 0.1% volume concen-tration and at 30 �C.

absorber plate and the environment reduces the overall heat losscoefficient, improving the thermal efficiency of the collector.

In a flat plate solar collector, entropy generation reduction ismore significant for higher temperature systems (Bejan, 1996)(Bejan et al., 1981). The reduction of the entropy generation rate isequivalent to an increase in power output. Exergy efficiency re-duces as the volume fraction reduces or mass flow rate increases.The TiO2eH2O nanofluid shows improved values of efficiencymatched to water as a base fluid. By using the TiO2eH2O nanofluidin a solar collector as an operational medium, the exergy efficiencycan be improved. This makes it a better alternative thanwater as anabsorbing medium. The experimental results show that by addingin a small amount of nanoparticles (up to 0.1 vol%); the exergy ef-ficiency is improved by 16.9%, compared to a pure water coolant. As

Fig. 12. The energy efficiency (No Fill) and exergy efficiency (solid Fill) of a FPSC atdifferent mass flow rates and different volume fractions for TiO2eH2O nanofluid.

Fig. 13. Effect of mass flow rate and volume fraction on entropy generation (No fill)exergy destruction (solid fill) for TiO2eH2O nanofluid.

Fig. 15. Overall energetic and exergetic efficiencies over the testing period for theexperimental work.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353 351

it is shown in Fig. 12, reducing the flow rate to approximately0.5 kg/min increases the collector exergy. The exergy loss at highmass flow rates is primarily due to heat transfer from the absorberplate to theworking fluid. Themain cause of exergy loss in collectoris heat transfer due to the difference between the temperature ofthe absorber plate and the solar radiation. The increase in theabsorber plate temperature leads to an increase in the differencewhich consequently decreases collector exergy loss. It can be seenthat the minimum exergy efficiency belongs to the higher massflow rate.

4.8. Entropy generation and exergy destruction

Entropy is produced in irreversible processes. Therefore, for theenergy optimization analysis, it is necessary to assess entropygeneration or exergy destruction due to heat transfer and viscousfriction as a function of the design variables selected (Kreuzer,1981) (Onsager, 1931a, b). Fig. 13 presents the entropy generationand exergy destruction with regard to mass flow rate and differentvolume concentrations.

Fig. 13 shows that increasing the volume fraction of the nano-fluid reduces entropy generation. The thermal conductivity im-proves as the volume concentration of the nanoparticles increases.The subsequent higher heat transfer reduces the irreversibilitygenerated in the system. This has a much greater effect than thelosses associated with viscous flow. The lowest entropy generationis observed for 0.1 vol%, and 0.3 vol% of the TiO2eH2O nanofluid,whereas the highest was observed for water, the values are 33.43 J/K, 40.95 J/K and 43.54 J/K, respectively for a flow rate of 0.5 kg/min.

The TiO2eH2O nanofluid shows reduced exergy destructionwith the increasing flow rate (compared to water), resulting in

Fig. 14. Output temperature (Solid Fill) and input temperature (No Fill) with respect tochanging mass flow rate for TiO2eH2O nanofluid.

lower exergy destruction. Entropy generation is also less thanwateras a working fluid. Similar patterns were observed for the exergydestruction. Therefore, the TiO2eH2O nanofluid performs betterthan using water as a working fluid.

4.9. Output temperature

Fig. 14 shows the effect of mass flow rate on the output tem-perature. The output temperature has a strong and direct propor-tional effect on the energy efficiency of a FPSC.

Solar collectors that use nanofluids are more efficient thanconventional collectors, due to their higher output temperature.Specific heat is defined as, “The heat required to raise the tem-perature of a unit mass of a substance by one unit of temperature.”It is clear from the definition that any substance, which has a lowerspecific heat, should provide greater temperature for equal heatflow.

4.10. Overall energetic and exergetic efficiencies using TiO2eH2Onanofluid

The overall energetic and exergetic efficiencies over the testingperiod for our experiment is presented in Fig. 15. The improvedenergetic and exergetic efficiencies are witnessed for the studiednanofluid. It can also be observed from Figs. 11 and 15 that themaximum irreversibility occurs at noon, when the solar radiation ismaximum and decreases as the solar energy decreases.

The temperature difference between the collector and theenvironment has an ideal point for the exergy efficiency, and alarger difference can result in lower exergy efficiency. However, anincrease in temperature difference reduces the energy efficiency,because more heat is lost to the environment. According to theresults from the experiments, the TiO2eH2O nanofluid is found tobe more appropriate as a working medium for the flat plate solarwater heater than water. The energy efficiency increased by 76.6%for 0.1 vol% and 0.5 kg/min, whereas the highest second law effi-ciency achieved is 16.9% for 0.1 vol% and 0.5 kg/min.

4.11. Pumping power and pressure drop for TiO2 nanofluid

For the nanofluid to be used in the solar collector appropriately,it is important to investigate the flow resistance of nanofluids inorder to increase the heat transfer. The pressure drop of theTiO2eH2O nanofluids in a FPSC is examined, while consideringlaminar flow.

Fig. 16. Pumping power (dotted line) and pressure drop (solid line) for TiO2/water nanofluid versus volume fraction and mass flow rate.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353352

Fig. 16 shows the effect of the nanofluid on the pumping powerand the pressure drop as a function of volume concentration andvolume flow rate (for laminar flow), respectively. These two pa-rameters are calculated using Eqs. (12)e(13) and Tables 1 and 2.

The results show that the friction factors of the nanofluids arevery similar to water. This implies that the pumping powerrequired for nanofluids (with low volume fractions) are the same aswater (Gherasim et al., 2011).

This analysis shows that the addition of nanoparticles in a basefluid (water) enhances its thermal conductivity and viscosity,which increases its friction factor. When the nanofluid propertiesare properly characterized, the friction factors of the TiO2 nanofluidare largely in agreement with the classical friction factor theory forsingle-phase flow (Tang et al., 2013). Therefore, the friction factorrelationship for the single phase flow can be applied for nanofluids.However, there was no evidence that use of nanofluids would costany significant addition of pumping power.

5. Conclusion

TiO2eH2O nanofluid sustained stable for a period of more thanone month. The thermal conductivity improvement is directlyrelated to the volume fraction and enhances up to 6% with 0.3 vol%of TiO2. Viscosity increases with particle loading and reduces withrising temperature. The energy efficiency increased by 76.6% for 0.1vol% and 0.5 kg/min, whereas the highest exergy efficiency ach-ieved is 16.9% for 0.1 vol% and 0.5 kg/min, using the nanofluids incomparison to the water. The solar collector efficiency using theTiO2eH2O nanofluid has higher energy and exergy efficiencies thanwater.

It can be concluded that the investigation reported here willdeliver the solar investigators with knowledge about enhancing thesolar collector systems using nanofluids. This knowledge is alsorequired for recognizing energy and exergy conservation oppor-tunities of these systems.

Acknowledgement

This research is supported by UM High Impact Research GrantUMRG Project: RP015B-13AET from the Ministry of Higher Educa-tion Malaysia.

References

Alphonse, P., Bleta, R., Soules, R., 2009. Effect of PEG on rheology and stability ofnanocrystalline titania hydrosols. J. Colloid Interf. Sci. 337, 81e87.

Bayat, J., Nikseresht, A.H., 2012. Thermal performance and pressure drop analysis ofnanofluids in turbulent forced convective flows. Int. J. Therm. Sci. 60, 236e243.

Bejan, A., 1996. Entropy generation minimization: the new thermodynamics offiniteesize devices and finiteetime processes. J. Appl. Phys. 79, 1191e1218.

Bejan, A., Kearney, D., Kreith, F., 1981. Second law analysis and synthesis of solarcollector systems. J. Sol. Energy Eng 103, 23e28.

Bobbo, S., Fedele, L., Benetti, A., Colla, L., Fabrizio, M., Pagura, C., Barison, S., 2012.Viscosity of water based SWCNH and TiO2 nanofluids. Exp. Therm. Fluid Sci. 36,65e71.

Chen, H., Ding, Y., 2009. Heat transfer and rheological behaviour of nanofluidseareview. In: Advances in Transport Phenomena. Springer, pp. 135e177.

Chon, C., Kihm, K., 2005. Thermal conductivity enhancement of nanofluids byBrownian motion. J. Heat. Transf. 127, 810.

Das, S.K., Putra, N., Thiesen, P., Roetzel, W., 2003. Temperature dependence ofthermal conductivity enhancement for nanofluids. J. Heat. Transf. 125, 567e574.

Das, S.K., Choi, S.U., Yu, W., Pradeep, T., 2008. Nanofluids: Science and Technology.Wiley-Interscience, Hoboken, NJ.

Davis, W., Martín, M., 2014. Optimal year-round operation for methane productionfrom CO2 and water using wind and/or solar energy. J. Clean. Prod. 80, 252e261.

Demir, H., Dalkilic, A., Kürekci, N., Duangthongsuk, W., Wongwises, S., 2011. Nu-merical investigation on the single phase forced convection heat transfercharacteristics of TiO2 nanofluids in a double-tube counter flow heat exchanger.Int. Commun. Heat. Mass Transf. 38, 218e228.

Duangthongsuk, W., Wongwises, S., 2009. Measurement of temperature-dependentthermal conductivity and viscosity of TiO2-water nanofluids. Exp. Therm. FluidSci. 33, 706e714.

Esen, H., 2008. Experimental energy and exergy analysis of a double-flow solar airheater having different obstacles on absorber plates. Build. Environ. 43,1046e1054.

ESTIF, 2011. Solar Heating and Cooling for a Sustainable Energy Future in Europe.Report was prepared by the European Solar Thermal Technology Platform(ESTTP). http://www.estif.org/fileadmin/estif/content/projects/downloads/ESTTP_SRA_RevisedVersion.pdf.

Fedele, L., Colla, L., Bobbo, S., 2012. Viscosity and thermal conductivity measure-ments of water-based nanofluids containing titanium oxide nanoparticles. Int. J.Refrig. 35, 1359e1366.

Fotukian, S., Nasr Esfahany, M., 2010. Experimental investigation of turbulentconvective heat transfer of dilute g-Al2O3/water nanofluid inside a circular tube.Int. J. Heat. Fluid Fl 31, 606e612.

Fu, Y., Liu, X., Yuan, Z., 2015. Life-cycle assessment of multi-crystalline photovoltaic(PV) systems in China. J. Clean. Prod. 86, 180e190.

Gangadevi, R., Senthilraja, S., Imam, S.A., 2013. Efficiency analysis of flat plate solarcollector using Al2O3/water nanofluid. Image 2, 3.

Gherasim, I., Roy, G., Nguyen, C.T., Vo-Ngoc, D., 2011. Heat transfer enhancementand pumping power in confined radial flows using nanoparticle suspensions(nanofluids). Int. J. Therm. Sci. 50, 369e377.

Gibbs, J.W., 1873. A Method of Geometrical Representation of the ThermodynamicProperties of Substances by Means of Surfaces. Connecticut Academy of Artsand Sciences.

He, W., Chow, T.-T., Ji, J., Lu, J., Pei, G., Chan, L.-s., 2006. Hybrid photovoltaic andthermal solar-collector designed for natural circulation of water. Appl. Energy83, 199e210.

Z. Said et al. / Journal of Cleaner Production 92 (2015) 343e353 353

Kamyar, A., Saidur, R., Hasanuzzaman, M., 2012. Application of computational fluiddynamics (CFD) for nanofluids. Int. J. Heat Mass Transf. 55, 4104e4115.

Koroneos, C.J., Nanaki, E.A., 2012. Life cycle environmental impact assessment of asolar water heater. J. Clean. Prod. 37, 154e161.

Kreuzer, H.J., 1981. Nonequilibrium Thermodynamics and its Statistical Foundations.Clarendon Press, Oxford and New York, 455 p. 1.

Li, C.H., Peterson, G., 2006. Experimental investigation of temperature and volumefraction variations on the effective thermal conductivity of nanoparticle sus-pensions (nanofluids). J. Appl. Phys. 99, 084314.

Luminosu, I., Fara, L., 2005. Determination of the optimal operation mode of a flatsolar collector by exergetic analysis and numerical simulation. Energy 30,731e747.

Masuda, H., Ebata, A., Teramae, K., Hishinuma, N., 1993. Alteration of thermalconductivity and viscosity of liquid by dispersing ultra-fine particles. NetsuBussei 7, 227e233.

Mintsa, H.A., Roy, G., Nguyen, C.T., Doucet, D., 2009. New temperature dependentthermal conductivity data for water-based nanofluids. Int. J. Therm. Sci. 48,363e371.

Natarajan, E., Sathish, R., 2009. Role of nanofluids in solar water heater. Int. J. Adv.Manuf. Tech. 1e5.

Nguyen, M., Ziemski, M., Vink, S., 2014. Application of an exergy approach to un-derstand energy demand of mine water management options. J. Clean. Prod. 84,639e648.

Onsager, L., 1931a. Reciprocal relations in irreversible processes. I. Phys. Rev. 37, 405.Onsager, L., 1931b. Reciprocal relations in irreversible processes. II. Phys. Rev. 38,

2265.Otanicar, T.P., Golden, J.S., 2009. Comparative environmental and economic analysis

of conventional and nanofluid solar hot water technologies. Environ. Sci.Technol. 43, 6082e6087.

Otanicar, T.P., Phelan, P.E., Prasher, R.S., Rosengarten, G., Taylor, R.A., 2010. Nano-fluid-based direct absorption solar collector. J. Renew. Sust. Energ. 2, 033102.

Pak, B.C., Cho, Y.I., 1998. Hydrodynamic and heat transfer study of dispersed fluidswith submicron metallic oxide particles. Exp. Heat Transf. Int. J. 11, 151e170.

Reddy, M., Rao, V.V., Reddy, B., Sarada, S.N., Ramesh, L., 2012. Thermal conductivitymeasurements of ethylene glycol water based TiO2 nanofluids. Nanosci.Nanotechnol. Lett. 4, 105e109.

Said, Z., Kamyar, A., Saidur, R., 2013a. Experimental investigation on the stabilityand density of TiO2, Al2O3. In: SiO2 and TiSiO4, IOP Conference Series: Earth andEnvironmental Science. IOP Publishing, p. 012002.

Said, Z., Sajid, M., Alim, M., Saidur, R., Rahim, N., 2013b. Experimental investigationof the thermophysical properties of Al2O3-nanofluid and its effect on a flat platesolar collector. Int. Commun. Heat. Mass Transf. 48, 99e107.

Said, Z., Sajid, M., Saidur, R., Kamalisarvestani, M., Rahim, N., 2013c. Radiativeproperties of nanofluids. Int. Commun. Heat. Mass Transf. 46, 74e84.

Sridhara, V., Satapathy, L.N., 2011. Al2O3 based nanofluids: a review. Nanoscale Res.Lett. 6, 1e16.

Standard, A., 1977. Methods of Testing to Determine the Thermal Performance ofSolar Collectors, 93e77. American Society of Heating, Refrigeration and AirConditioning Engineers, New York (USA).

Tang, C.C., Tiwari, S., Cox, M.W., 2013. Viscosity and friction factor of aluminumoxideewater nanofluid flow in circular tubes. J. Nanotech. Eng. Med. 4, 021004.

Tian, Y., Zhao, C.-Y., 2013. A review of solar collectors and thermal energy storage insolar thermal applications. Appl. Energy 104, 538e553.

Tiwari, A.K., Ghosh, P., Sarkar, J., 2013. Solar water heating using nanofluids-acomprehensive overview and environmental impact analysis. Int. J. Emerg.Tech. Adv. Eng. 3, 221e224.

Turgut, A., Tavman, I., Chirtoc, M., Schuchmann, H., Sauter, C., Tavman, S., 2009.Thermal conductivity and viscosity measurements of water-based TiO2 nano-fluids. Int. J. Thermophys. 30, 1213e1226.

Ucar, A., Inallı, M., 2006. Thermal and exergy analysis of solar air collectors withpassive augmentation techniques. Int. Commun. Heat. Mass Transf. 33,1281e1290.

Vasseur, V., Kamp, L.M., Negro, S.O., 2013. A comparative analysis of photovoltaictechnological innovation systems including international dimensions: the casesof Japan and The Netherlands. J. Clean. Prod. 48, 200e210.

Vatanpour, V., Madaeni, S.S., Moradian, R., Zinadini, S., Astinchap, B., 2011. Fabri-cation and characterization of novel antifouling nanofiltration membrane pre-pared from oxidized multiwalled carbon nanotube/polyethersulfonenanocomposite. J. Membr. Sci. 375, 284e294.

Wamkam, C.T., Opoku, M.K., Hong, H., Smith, P., 2011. Effects of pH on heat transfernanofluids containing ZrO2 and TiO2 nanoparticles. J. Appl. Phys. 109, 024305.

Wang, X.-Q., Mujumdar, A.S., 2007. Heat transfer characteristics of nanofluids: areview. Int. J. Therm. Sci. 46, 1e19.

Xie, H., Yu, W., Chen, W., 2010. MgO nanofluids: higher thermal conductivity andlower viscosity among ethylene glycol-based nanofluids containing oxidenanoparticles. J. Exp. Nanosci. 5, 463e472.

Xydis, G., 2013. On the exergetic capacity factor of a windesolar power generationsystem. J. Clean. Prod. 47, 437e445.

Yousefi, T., Shojaeizadeh, E., Veysi, F., Zinadini, S., 2012a. An experimental investi-gation on the effect of pH variation of MWCNTeH2O nanofluid on the efficiencyof a flat-plate solar collector. Sol. Energy 86, 771e779.

Yousefi, T., Veysi, F., Shojaeizadeh, E., Zinadini, S., 2012b. An experimental investi-gation on the effect of Al2O3eH2O nanofluid on the efficiency of flat-plate solarcollectors. Renew. Energ. 39, 293e298.