Investigation of a Hybrid System of Nocturnal Radiative Cooling

lable at ScienceDirect

Building and Environment 45 (2010) 1521–1528

Contents lists avai

Building and Environment

journal homepage: www.elsevier .com/locate/bui ldenv

Investigation of a hybrid system of nocturnal radiative cooling and directevaporative cooling

Ghassem Heidarinejad a,*, Moien Farmahini Farahani a, Shahram Delfani b

a Department of Mechanical Engineering, Tarbiat Modares University, PO Box 14115-143, Tehran 14117, Iranb Building and Housing Research Center (BHRC), PO Box 13145-1696, Tehran, Iran

a r t i c l e i n f o

Article history:Received 5 August 2009Received in revised form24 December 2009Accepted 4 January 2010

Keywords:Nocturnal radiative coolingCooling coilDirect evaporative coolingHybrid cooling system

* Corresponding author. Tel.: þ98 21 82883361; faxE-mail address: [email protected] (G. Heidar

0360-1323/$ – see front matter � 2010 Elsevier Ltd.doi:10.1016/j.buildenv.2010.01.003

a b s t r a c t

In this paper, the results of a study on a hybrid system of nocturnal radiative cooling, cooling coil, anddirect evaporative cooling in Tehran have been discussed. During a night, the nocturnal radiative coolingprovides required chilled water for a cooling coil unit. The cold water is stored in a storage tank. Duringeight working hours of the next day, hot outdoor air is pre-cooled by means of the cooling coil unit andthen it enters a direct evaporative cooling unit. In this period, temperature variation of the conditionedair is investigated. This hybrid system complements direct evaporative cooling as if it consumes lowenergy to provide cold water and is able to fulfill the comfort condition whereas direct evaporative aloneis not able to provide summer comfort condition. The results obtained demonstrate that overall effec-tiveness of hybrid system is more than 100%. Thus, this environmentally clean and energy efficientsystem can be considered as an alternative to the mechanical vapor compression systems.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Cooling is an essential issue in air conditioning of most buildingsin warm and humid climates. In fact, due to great consumption ofenergy in buildings, there are increasing demands to designbuilding heating, ventilation, and air conditioning (HVAC) equip-ments and systems energy efficiently. Among the HVAC compo-nents and systems, cooling systems consume the largest amount ofelectrical energy. The issues of climatic change caused by globalwarming, the consumption of fossil fuels, the resources depletion,and demand for reducing pollutant particles have led to a growthuse of natural resources instead of conventional energy resourcesor partly replacement of active cooling system. The usage of passivecooling has been considered to drive cooling cycles to providecomfort cooling. In addition, evaporative cooling system can be aneconomical alternative, or as a pre-cooler in the conventionalsystems. Also, it is known due to its zero pollution, easy mainte-nance, low energy consumption, simplicity, and good indoor airquality [1–8].

Passive cooling resources are the natural heat sinks of the planetin order that understanding their parameters is worthwhile for allvarieties of cooling methods. Three heat sinks of nature are the sky,atmosphere, and the earth. Energy transfer to sky is entirely doneby radiation in the wave-length interval from approximately

: þ98 21 88005040.inejad).

All rights reserved.

8–14 mm. In fact, the only means which the earth can lose heat isradiative cooling [1,3].

Significant thermal comfort can be achieved during summer bypassive cooling in buildings with a great reduction of cooling loads.A black object at ambient temperature interacts with all tempera-ture range of atmospheric layers and causes cool down beneath ofambient temperature in optimum situations. Heat dissipationtechniques are based on the transfer of excess heat to a lowertemperature natural sinks. Regarding sky, heat dissipation iscarried out by long-wave radiation from a building to the sky that iscalled radiative cooling. The sky equivalent temperature is usuallylower than the temperature of the most bodies on the earth,therefore, any ordinary surface that interact with the sky has a netlong-wave radiant loss [2,3].

Direct evaporative cooling (DEC) is the oldest, and the mostwidespread form of air conditioning. The underlying principle ofDEC is the conversion of sensible heat to latent heat. Througha direct evaporative cooling system, hot outside air passes a porouswetted medium. Heat is absorbed by the water as it evaporatesfrom the porous wetting medium, so the air leaves the system ata lower temperature. In fact, this is an adiabatic saturation processin which dry bulb temperature of the air reduces as its humidityincrease (constant enthalpy). Some of the sensible heat of the air istransferred to the water and become latent heat by evaporatingsome of the water. The latent heat follows the water vapor anddiffuses into the air. The minimum temperature that can beobtained is the wet bulb temperature of the entering air [8–11].

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–15281522

In HVAC systems, cooling coils unit (CCU) perform an essentialfunction by exchanging the cooling load from the hot air to thechilled water loop by pushing air flow through the coil. Also, CCUcan be utilized as pre-cooler systems to decrease temperature ofhot air. Totally, utilization of cooling coils affects performance ofHVAC systems increasingly [12,13].

Several research papers were dedicated to explore issues aboutnocturnal cooling such as, Berdhal and Fromberg (1982) [14],Argiriou et al. (1994) [15], Ali et al. (1995) [16], Mihalakakou et al.(1998) [17], Al-Nimr et al. (1998, 1999) [18], Spronken-smith (1999)[19], Erell and Etzion (1999, 2000) [20–23], Meir et al. (2003) [3],Bagiorgas and Mihalakakou (2007) [2], Bassindowa et al. (2007)[24], Salim Shirazy et al. (2008) [25,26], and Farmahini Farahaniet al. (2009) [27]. Aforementioned research studied experimentaland theoretical investigation of long-wave radiance, nocturnalradiative cooling and its potential in different conditions, andeffects of different parameters on it. Regarding direct evaporativecooling, Lueng (1995) [28], Halaz (1998) [29], Camargo et al.(2000–2005) [8,30], Dai and Sumathy (2002) [31], Liao and Chiu(2002) [32], and Al-Sulaiman (2002) [33] have proposed mathe-matical modeling and done experimental in order to analyzeefficiency or simulate direct evaporative cooling. Furthermore,Scoldfield and DesChamps (1984) [34], Al-Juwayhel et al. (1997)[35], El-Dessouky et al. (2004) [36], and Heidarinejad et al. (2009)[37] have studied two-stage evaporative cooling to examine itsefficacy on performance.

To the best knowledge of the authors of this paper, no significantinvestigation has been performed on combining nocturnal coolingand evaporative cooling. Thus, lack of information about feasibilityof this new combination is the motivation of this study.

In this research, the water in a storage tank is cooled by means ofcirculating the water through a flat-plate radiator throughouta night (nocturnal radiative cooling). During the next day, the coldwater in the storage tank is used in a cooling coil unit as chilledwater to decrease temperature of outdoor air (pre-cooling). Then,the pre-cooled air with lower wet bulb temperature passes througha direct evaporative pad (See Fig. 1). By this way, the hot outdoor airis pre-cooled through the CCU which augments efficacy of wholecooling system. The chilled water is obtained from a renewable andpollutant-free process which consumes low energy in comparisonwith conventional mechanical vapor compression systems. Theperformance and feasibility of such cooling system have beenanalyzed in this paper.

2. Modeling and formulations

This system consists of four parts 1 – Radiator, 2 – Storage tank,3 – Cooling coil, and 4 – Direct evaporative cooling. Formulationsand modeling of each part have been described in the followingsubsections.

Fig. 1. A Schematic diagram of the hybrid system of radiati

2.1. Formulation of the flat-plate radiator

The flat-plate collector as heat exchanger is studied andtemperature distribution in any desirable point along flat-plate isproposed by equation (1) [4,20]. If time interval is kept reasonablysmall, this steady state expression predicts accurate outlettemperature.

Tf � Ta þ S=UL

Tfi � Ta þ S=UL¼ exp

��ULnwFy

_mCp

�(1)

Where, Tf is the outlet fluid temperature, Ta is the ambient airtemperature, Tfi is the fluid temperature at the collector inlet, S isthe emitted radiative energy to sky from surface of the collector, UL

is the overall heat loss coefficient of the collector, n is the number ofparallel tubes in the collector structure, w is distance between thetubes, y is the tubes length, F is the collector efficiency factor, _m isthe mass flow rate through the collector, and Cp is the specific heatof the fluid.

The flat-plate collector efficiency factor denotes the proportionof the factual real gain rate per tube per unit length to the gain whichwould occur if the collector absorber plate were at the temperatureTf [38]. The flat-plate collector efficiency factor depends on structureof the collector, character and thermal conductivity of the adhesivematerial between tubes and absorber plate, and heat transfercoefficient of heat carrier inside of tube. The flat-plate collectorefficiency factor value varies between 0.85 and 0.95.

The overall heat loss coefficient UL is sum of heat losses aroundthe collector such as, convection on top of the collector andconduction under and on sides of the collector. The heat loss due toconduction beneath of the collector is the proportion of thermalconductivity of the insulation to thickness of the insulation. Becausethe sides’ loss is less than 5%, it is negligible. The top loss coefficientUt is assessed by considering convection losses from the upwardsurface of the collector. The highest amount of heat loss takes placeat the top of the collector. Equation (2) estimates top loss.

Ut ¼ 1:8þ 3:8v 1:35 < v < 4:5 (2)

where v is wind velocity. If the wind velocity is less than 1.35 m/s,this expression overestimates the heat loss [39], but it does notinterfere calculation.

2.2. Modeling of sky equivalent temperature

Ambient temperature and sky equivalent temperature are twoeffective measures of the surrounding conditions. These twomeasures affect the cooling performance and outlet temperature ofthe collector exposed to the night sky. The difference betweenambient temperature and sky equivalent temperature demon-strates the potential of the nocturnal cooling [25]. The wind speed

ve cooling, cooling coil, and direct evaporative cooling.

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–1528 1523

is another surrounding condition which is accounted for top heatloss of the collector as described previously. The sky temperature isdefined as the temperature of a black body radiator emitting thesame amount of radiative power as the sky [20]. Equation (3)relates the effective sky temperature to the ambient temperature,and dew point temperature [26,39].

TSky ¼ Ta

h0:741þ 0:0062Tdp

i1=4(3)

Where Ta and Tsky are the ambient air temperature and sky equiv-alent temperature, respectively, Tdp is dew point temperature. Inequation (3), both Ta and Tsky are calculated in Kelvin, but Tdp is inCelsius.

The emitted radiative energy to sky from surface of the collectorS can be calculated from following linearized expression [1].

Snet ¼ 43rsT3air

�Tcol � Tsky

�(4)

where 3r is the hemispherical infrared emissivity of the collector,s is the Stefan–Boltzmann constant (5.67 � 10�8), and Tcol is thecollector temperature. In this investigation the collector tempera-ture is an average of inlet and outlet water temperature of thecollector.

2.3. Stratification water tank

As a functional matter, many tanks show some degree of strat-ification. In a thermally stratified situation, the temperature of thecontained liquid varies from the bottom to the top, being less at thebottom and more at the top. This situation is in contrast to well-mixed tank in which the liquid temperature is uniform throughout.According to Fig. 2, a tank can be modeled as being divided into Nsections, with energy balance for each section of the tank. Theresult is a set of N differential equations that can be solved for thetemperatures of the N sections as functions of time [38,40]. Strat-ification improves with increasing tank height/diameter ratio,temperature difference, and inlet as well as outlet port diameters,while stratification decreases with increasing flow rate [41]. Theschematic pattern of stratified storage tank is shown in Fig. 2.

Utilizing low mass flow rate causes an increase in temperaturedifference of outlet and inlet from the radiator and consequentlyresults in higher degree of thermal stratification in the water tank.In addition, the temperature at the bottom of the tank will be

Fig. 2. A schematic pattern of stratified storage tank.

cooler. Thus, auxiliary energy consumption will reduce. Moreover,in highly stratified tank the inlet temperature to the collector willbe higher, which derives more absorbed energy. Related work insolar systems shows that energy storage efficiency increases up to6% [40,42,43]. Due to the importance of stratification in storagetanks, in this investigation mass flow rate is chosen in a way thatsaves water stratification in the tank.

In stratified tanks, an energy balance on each section can beexpressed as:

miCpdTs;i

dt ¼ a _mcCp�Tro�Ts;i

�þbi _mLCp

�TL�Ts;i

��ðUAÞs;i

�Ts;i�Ta

�

þ(

giCp�Ts;iþ1�Ts;i

�; if gi > 0

giCp�Ts;i�Ts;i�1

�; if gi < 0 ð5Þ

Where,

ai ¼

8<:

1; if i ¼ N and Tro < Ts;i1; if Ts;iþ1 � Tro < Ts;i0; otherwise

bi ¼

8<:

1; if i ¼ 1 and TL > Ts;i1; if Ts;i � TL < Ts;i�10; otherwise

gi ¼

8>>>>>>>>><>>>>>>>>>:

� _mL

XN�1

j¼1

bi; if i ¼ N

_mc

XN

j¼2

ai; if i ¼ 1

_mc

Xiþ1

j¼N

ai � _mL

Xi�1

j¼1

bi; if i ¼ 2;.;N � 1

(6)

In equation (6), _mc and _mLare mass flow rate from the collector andmass flow rate of the load respectively, Tro is the temperature of heatcarried from the collector, and TL is the temperature of the load [4,44].

2.4. Modeling of cooling coil

Because input temperatures of both fluids are given and outputtemperatures are required to be found out, the 3–NTU method ischosen. A sensible cooling process only exists when the outersurface temperature of the coil is equal to or higher than the dewpoint of the entering air. A sensible cooling process is indicated bya horizontal line towards the saturation curve on the psychrometricchart. In other words, the humidity ratio is always constant [45].

Outside surface area of the coil (Ao) multiply its overall heattransfer coefficient based on outside surface area of the coil (Uo) canbe calculated as [46]:

UoAo ¼�

1hshaAo

þ�

Dout

2AiKtubeln�

Dout

Din

��þ 1

hiAi

�1

(7)

where, hs is finned surface efficiency, which depends on area of fins,outside surface area, and fin efficiency, Dout, Din are outer and innertube diameter, Ktube is the thermal conductivity of metal of thetubes, Ai is inner surface area of tubes, and ha, hi are heat transfercoefficient of air side of the coil and tube side of the coil, respectively.

2.4.1. Air side heat transfer coefficientOn the air side, hot air meets chilled water tubes rows that lead

to reduction in temperature of the air flow. Several heat transfercoefficients have been proposed. For the staggered arrays, thefollowing approximates Nu within �15% [47].

Nu ¼ hadk¼ CRen

d;maxPr0:36 (8)

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–15281524

Where, C, n depend on geometrical arrangement of the coil, Red, max

is Reynolds number based on the velocity through the narrowestcross section, and Pr is Prandtl number.

For corrugated fins, ho should be multiplied by a factor Fcor of1.1–1.25 to account for the increase in turbulence and rate of heattransfer.

2.4.2. Water side heat transfer coefficientFor chilled water at turbulent flow inside the tubes, the inner

surface heat transfer coefficient hi can be calculated by the Dittus–Boelter equation [48]:

NuD ¼hiDi

kw¼ 0:023Re0:8Pr0:3 (9)

where, kw is thermal conductivity of water in tubes. This formula isthe most popular correlation.

2.5. Modeling of direct evaporative cooling

In a direct evaporative cooler, the transformation of the heat andmass between air and water causes decrease in the air-dry bulbtemperature (DBT) and increase in its humidity, while the enthalpyis basically constant in a perfect process. The minimum tempera-ture that can be attained is the wet bulb temperature (WBT) of theincoming air. Wet pads or porous materials equip a water surface inwhich the air is humidified and the pad is wetted by dripping water.

Assuming the hot air flow near to a wet surface, according toFig. 3, heat transfer occurs due to the difference in surfacetemperature Ts and the flow of air temperature Tair. Mass transferalso occurs, because the absolute humidity (concentration) of theair close to the surface uair is different from the humidity of the wetsurface us.

The total differential heat flow is

dQ ¼ ½hcðTs � TaÞ þ hmiysðus � uaÞ�dA (10)

where, Ta is air temperature, Ts is surface temperature, ua absolutehumidity of air, and us is absolute humidity of wet surface. Usingthe specific enthalpy of the mixture as the sum of the individualenthalpies and assuming that air and vapor are perfect gases,equation (10) can be rewritten as:

dQ ¼ hc

Cpu

�ðis � iaÞ þ

ðus � uaÞLe

ðivs � ivLeÞ

dA (11)

where, hc is convective heat transfer coefficient, hm is mass transfercoefficient, ia is enthalpy of air, is is enthalpy of surface, ivs isenthalpy of vaporization of the water at surface temperature, iv isspecific enthalpy of the vapor at surface temperature, Cpu is specificheat of the humid air, and Le ¼ (hc)/hmCpu is Lewis number.

By considering Le ¼ 1 the second term in bracket in equation(11) is negligible in presence of the first term this mainly because,the order of both us and ua are low and obviously, the order of theirdifference is much lower than order of difference of is and ia. Hence,

Fig. 3. Schematic element of direct evaporative cooling.

the second term can be eliminated. Therefore, by combiningequation (10) with equation (11) and integrating, the temperatureof output air will be:

Ta;out � Ts

Ta;in � Ts¼ exp

�� hcA

_maCpu

�(12)

It is assumed that the makeup water entering the sump toreplace evaporated water is at the same adiabatic saturationtemperature of the incoming air. Dowdy and Karabash introduceda correlation to establish the convective heat transfer coefficient ina rigid cellulose paper evaporative media [49]:

Nu ¼ 0:1�

lel

�0:12

Re0:8Pr1=3 (13)

where, le ¼ V/A is characteristic length in which V is volume ofevaporative pad and A is total wetted surface area (area of the heattransfer surface), l is pad thickness.

Reynolds and Nusselt numbers are calculated by characteristiclength.

2.6. Saturation effectiveness

The cooling effectiveness of the CCU and DEC can be calculatedby the following equation:

3 ¼Ta;in � Ta;out

Ta;in � Ts;in(14)

where 3 is the cooling effectiveness, Ta,in and Ta,out are the inlet andoutlet dry bulb temperatures of the air stream respectively, and Ts,in

is the inlet wet bulb temperature of the air stream. According toequation (12), cooling effectiveness of the stand-alone DEC unit islower than unity, but for the combined system it may be greaterthan unity. This is because the outlet dry bulb temperature of theair stream can be lower than the inlet wet bulb temperature. Asexpressed in equation (14), the definition of saturation effective-ness is only based on the temperature difference. Energyconsumption of transportation of fluids is not considered in thisdefinition.

3. Results and discussion

The theoretical investigation of the hybrid system of nocturnalcooling, cooling coil and direct evaporative cooling has beenstudied in Tehran, capital city of Iran. Tehran (Longitude 51.4�,Latitude 35.7�) which is the most populated city in Iran has thehighest electrical consumption during summer. Finding a viablealternative with higher effectiveness and capability of providingcomfort condition during summer to convenient cooling systemswas the motivation of this research. The study has been done onAugust 5th and 6th, 2008. The ambient temperatures, dew pointtemperatures, and wind speed are derived from Iran meteorolog-ical organization internet site.

An uncovered (unglazed) flat-plate collector is used fornocturnal radiative cooling. The dimensions of the flat-platecollector are 2.054 � 1.11 m2, and it consists of eight copper tubes.Each tube has a diameter of 10 mm and center-to-center distance of0.147 m. The 200 L water tank used here is isolated with a 20-cmthick glass wool, k¼ 0.05 W/m �C. Volume of the water storage tankis big enough to adequately store required cold water for use in thenext day. Therefore, the ratio of the radiative surface area to thewater storage tank is A/V ¼ 0.0114 m2/l. This ratio impacts outletwater temperature and water storage tank. As far as mass flow rateis concerned, it is worth mentioning that mass flow rate does not

Fig. 4. Validation of the radiative cooling part with Meir et al. [3] at May 19–20,A/V ¼ 0.019 m2/l.

Fig. 5. Verification of cooling coil modeling, air volume flow rate is 530 CFM and watermass flow rate is 0.2 kg/s.

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–1528 1525

significantly influence cooling process in close circulations.Decreasing mass flow rate causes to lower outlet water tempera-ture. Since volume of water with lower temperature is less thanbefore, total amount of heat which is added to the water tankremains constant. Nevertheless, from stratification in storage tankpoint of view, high mass flow rate impairs it. However, in an opencirculation, lower mass flow rate can cause cooler heat carrier [18].Totally, a stratified water tank has much higher performance thana fully mixed water tank, so a proper mass flow rate should beconsidered to save energy in the water storage tank. In order to gainbest results and stratification be maintained, the water mass flowrate of _mw ¼ 0:01667 kg=s ð _Vw ¼ 60 L=hÞ is chosen. Moreover, allthermodynamic properties of water (in the radiative part) and air(in the CCU and DEC) vary by temperature.

The stratification water tank modeling is used to analyzetemperature changes of liquid in the water tank. Therefore, thewater tank is divided into five sections. At night for the radiativecooling in which the water in the storage tank is intended to becooled, the output water, which is warmer, exits from top of thetank and input water enters from the radiator at the bottom of thewater tank. However, during the next day, the colder water atthe bottom of the tank flows towards the cooling coil and turnsback at the top of the water tank. This approach makes the coolingprocess more efficient.

The radiative cooling part has been validated by results of Meiret al. experimental investigation [3]. Fig. 4 compares this modelingand their experimental results.

As far as the cooling coil is concerned, it has corrugatedaluminum fins with staggered copper tubes. The water mass flowrate is similar to the radiative cooling part and air volume flow rateof 1000 CFM (air face velocity in SI is 2.27 m/s) is chosen. This flow

Table 1Operating parameters of cooling coil.

Height � width 46 � 46 mm2

Number of rows of tubes 6Number of tubes per row 12Fin pitch 394 per meterFin efficiency 0.935Fin thickness 0.254 mmVertical tube spacing 38.1 mmHorizontal tube spacing 30.5 mmTube outside diameter 15.875 mmTube wall thickness 0.508 mm

rate is sufficient for a room to be conditioned. Table 1 lists othergeometrical specifications.

The theoretical modeling of coil has been verified by usingexperimental setup test located in Building and Housing ResearchCenter (BHRC) as shown in Fig. 5.

The air flow after pre-cooling process enters evaporative padwith same velocity. An evaporative pad with dimensions of0.46 � 0.46 � 0.2 m3 is considered. It is worth mentioning thatwidth and height of the pad are equal to dimensions of the coolingcoil. In other words, it is assumed that air passes through the samechannel.

As shown in Fig. 6, modeling of DEC is validated according toexperimental results of Camargo et al. [30] in which efficiency isillustrated versus Reynolds number based on characteristic length.

As stated in Figs. 4–6, all parts of the hybrid system are modeledaccurately. Consequently, combination of them can preciselypredict results.

The first segment of this system is nocturnal radiative coolingwhich provides cold water for cooling coil unit. Fig. 7 shows averagetemperature of the water in the storage tank from August 5th at9:00 PM to August 6th at 6:00 AM. The average temperature iscalculated based on average temperature of five sections in thewater tank. All the initial temperatures, including water in tank andinitial inlet water temperature are assumed to be 28 �C. Laterresults prove that these initial values were selected precisely. As it

Fig. 6. Validation of direct evaporative cooling modeling.

Fig. 7. Average temperature of water storage tank and temperature of ambient.Fig. 9. Two-stage cooling process on psychrometric chart.

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–15281526

is shown in Fig. 7, temperature of the water reduces due to radiationtowards sky.

The cold water obtained at night is used during 8 h of the nextday (August 6th) as the chilled water in the cooling coil to reducetemperature of outdoor air. Usually offices begin at 9:00AM andcontinue to work up to 5:00PM. Fig. 8 depicts temperature differ-ences after each process. Relative humidity of entrance (outdoor)air is considered to be 29%.

Fig. 8 shows that hot outdoor air is pre-cooled by means ofcooling coil unit and the average temperature difference betweenentering and leaving air is 8 �C. Then, pre-cooled air flows throughevaporative pad. Through the evaporative cooling process relativehumidity of air increases, but its sensible temperature dropsdown. The average temperature difference between entering andleaving air in direct evaporative cooling is 5.5 �C. Thus, tempera-ture of hot outdoor air approximately decreases 13.5 �C at anytime step.

Fig. 9 illustrates two last points of Fig. 8 on psychrometric chart.As shown in Fig. 9, the hybrid system can provide comfort condi-tion. However, stand-alone direct evaporative cooling cannot meetthe comfort conditions as it is demonstrated. Note that the directevaporative cooling satisfies the comfort conditions critically inTehran’s ambient conditions.

Regarding stratified water tank, during cooling process from9:00 PM to 6:00 AM temperature of the layers varies. The warmer

Fig. 8. Temperature differences after each process.

layer is always at top of the tank while the coldest one is at bottomof the tank. This situation is shown in Fig. 10. After cooling processby the radiator, from 6:00AM to 9:00AM, water in the tankcontinues its stratification and due to warmer ambient, tempera-ture of all layers increase insignificantly. During use of the coldwater in the tank from 9:00AM to 5:00PM, increase in temperatureof higher layers at this period is slower than lower layers. This isbecause, on the one hand, water exits from the bottom of the tankand the upper layer takes place of the lower layer. So after a whilethe lowest layer will gain temperature of its upper layer. On theother hand, temperature of entering water at top of the tank doesnot remarkably differ from the top of the tank. From 5:00PM to9:00PM, temperature of layers of water in the tank increasewithout losing their stratification. Indeed, demand for more accu-rate calculation necessitate stratification modeling of the watertank. As shown in Fig. 10, average temperature of water in thestorage tank at the last point (9:00 PM) is near to 28 �C and itevinces that assumed initial temperature is admissible. Pressuredrop in tubes is insignificant during day time and night time,because during night time water is circulated between the waterstorage tank and the radiator. Tubes between these two – the watertank and the radiator – are assumed to be short, and also, the lengthof tubes through the radiator is short (approximately 2 m). Thuspressure drop is not important. The same process happens during

Fig. 10. Temperature of different stratified layers of water tank.

Fig. 11. Effectiveness comparison of three cooling methods, DEC, CCU/DEC, and IEC/DEC.

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–1528 1527

day time, but the cooling coil is used instead of the radiator.However, pumping power may insignificantly influence watertemperature approximately 1 �C.

Fig. 11 compares effectiveness of three cooling methods, directevaporative cooling, two-stage system of cooling coil unit/directevaporative cooling, and two-stage direct/indirect evaporativecooling. It shows that effectiveness of conventional direct evapo-rative cooling can remarkably increase by adding a cooling coil unit.In addition, according to Fig. 9, leaving air from the hybrid systemcan reach the comfort conditions. However, direct evaporativecooling is not able to provide this condition lonely. Furthermore,although two-stage direct/indirect evaporative cooling has a higheffectiveness, it consumes more water than the hybrid system ofcooling coil unit/direct evaporative cooling. It is worth mentioning,the effectiveness is calculated only based on temperature differenceof air through the CCU and DEC. However, considering efficiency ofall devices, including pumps, a fan, and fluid flow, needs a newformula and with the new formula efficiency might not be morethan unit.

As far as economical point of view is concerned, the price of thenew system (price of the radiator, CCU, evaporative pad, and otherdevices) can be the same as the price of a mechanical vaporcompression system (it depends on the brand, cooling power, andother factors of system). However, electrical energy consumption ofthe new system is much lower than mechanical vapor compressionsystems. The reason is that although, two small pumps, whichpump water during night and day time to radiator and CCU, anda fan, that blows air through the evaporative pad, consume energyin this new system, a compressor in mechanical vapor compressionsystem consumes much more energy in comparison with afore-mentioned devices. Furthermore, according to the governmentalnew policy, all energy subsidies will be eliminated within threeyears. Thus, electrical energy increases remarkably, necessitatinga low electric consuming system in Tehran. Besides economicalaspects, this system adds no heat to environment.

4. Conclusion

The behavior of the hybrid system of nocturnal radiative cooling,cooling coil, and direct evaporative cooling has been investigatedduring 8 h in Tehran. The chilled water for the cooling coil unit isprovided by the nocturnal radiative cooling at previous night. Theresults show that Tehran has the capability of providing cold waterat night during summer. Also, whereas direct evaporative coolerslonely cannot provide comfort conditions, the hybrid system has

high potential to provide comfort conditions. In addition, theeffectiveness of the hybrid system is considerably higher thanstand-alone direct evaporative cooling. Taking advantage of sky asa renewable source of the passive cooling, the hybrid coolingsystem can be considered as an environmentally clean and energyefficient system. Thus, this system can be used as a replacement formechanical vapor compression systems, leading to decrease elec-trical energy consumption.

References

[1] Cook J. Passive cooling. Cambridge: MIT Press; 1989.[2] Bagioras HS, Mihalakakou G. Experimental and theoretical investigation of

a nocturnal radiator for space cooling. Renewable Energy 2008;33:1220–7.[3] Meir MG, Rekstad JB, Lovvik OM. A study of a polymer-based radiative cooling

system. Solar Energy 2003;73(6):403–17.[4] Duffie JA, Beckman WA. Solar engineering of thermal processes. John Wiley &

Sons; 1980.[5] Santamouris M, Asimakopoulos DN. Passive cooling of buildings. James &

James; 1996.[6] Wu JM, Huang X, Zhang H. Numerical investigation on the heat and mass

transfer in a direct evaporative cooler. Applied Thermal Engineering2009;29:195–201.

[7] Yu Xin, Wen Jin, Smith TF. A model for the dynamic response of a cooling coil.Energy and Building 2005;35:1278–89.

[8] Camargo JR, Ebinuma CD, Silveira JL. Experimental performance of a directevaporative cooler operating during summer in a Brazilian city. InternationalJournal of Refrigeration 2005;28:1124–32.

[9] Watt JR, Brown WK. Evaporative air conditioning handbook. Fairmont Press; 1997.[10] Sodha MS, Ashutosh S, Kumar A, Sharma AK. Thermal performance of an

evaporatively cooled multi-story building. Building and Environment1986;21:571–82.

[11] Heidarinejad G, Bozorgmehr M. Heat and mass transfer modeling of two stageindirect/direct evaporative air coolers. ASHRAE; 2007–2008 [Thailand ChapterJournal].

[12] Wang Yao-Wen, Cai Wen-Jian, Soh Yeng-Chai, Li Shu-Jiang, Lu Lu, Xie Lihua. Asimplified modeling of cooling coils for control and optimization of HVACsystems. Energy Conversion and Management 2004;45:2915–30.

[13] Jin Guang-Yu, Cai Wen-Jian, Wang Yao-Wen, Yao Ye. A simple dynamic modelof cooling coil unit. Energy Conversion and Management 2006;47(15–16):2659–72.

[14] Berdhal P, Fromberg R. The thermal radiance of clear skies. Solar Energy1982;29:229–314.

[15] Argiriou A, Santamouris M, Assimakopoulos DN. Assessment of the radiativecooling potential of a collector using hourly weather data. Energy1994;19:879–88.

[16] Ali AHH, Taha IMS, Ismail IM. Cooling of water flowing through a night skyradiator. Solar Energy 1995;55(4):235–53.

[17] Mihalakakou G, Ferrante A, Lewis JO. The cooling potential of a metallicnocturnal radiator. Energy and Building 1998;28(3):251–6.

[18] Al-Nimr MA, Kodah Z, Nassar B. A theoretical and experimental investigationof a radiative cooling system. Solar Energy 1998;63(6):367–73.

[19] Spronken-smith RA, Oke TR. Scale modeling of nocturnal cooling in urbanparks. Boundary-Layer Meteorology 1999;93:287–312.

[20] Erell E, Etzion Y. Radiative cooling of buildings with flat-plate solar collectors.Building and Environment 2000;35(4):297–305.

[21] Erell E, Etzion Y. Analysis and experimental verification of an improvedcooling radiator. Renewable Energy 1999;16(1–4):700–3.

[22] Erell E, Etzion Y. A radiative cooling system using water as heat transfermedium. Architectural Science Review 1992;35:35–49.

[23] Erell E, Etzion Y. Thermal storage mass in radiative cooling systems. Buildingand Environment 1991;26(4):389–94.

[24] Bassindowa H, et al. An experimental investigation on night radiative cooling.In: Proceeding of the seventh Suadi engineering conference; 2007.

[25] Salim Shirazy M, Heidarinejad G, Delfani S. Investigation of potential of usageof radiative cooling in some cities of Iran to use passive cooling so as to reduceenergy consumption. Kerman, Iran: ISME; 2008 [In Farsi].

[26] Salim Shirazy M, Heidarinejad G, Delfani S. Numerical and experimentalsimulation of a nocturnal radiative cooling with solar collectors. Kerman, Iran:ISME; 2008 [In Farsi].

[27] Farmahini Farahini M, Heidarinejad G, Delfani S. Theoretical investigation ofpotential of nocturnal cooling in Iran. Tehran, Iran: ISME; 2009.

[28] Leung CW. Forced convection inside horizontal isosceles triangular ducts.IMechE conference transactions. Mechanical Engineering Publications Ltd;1995. 117–121.

[29] Halasz B. A general mathematical model of evaporative cooling devices. RevueGenerale de Thermique 1998;37:245–55.

[30] Camargo JR, Ebinuma CD. A mathematical model for direct and indirectevaporative cooling air conditioning systems. In: Proceedings of the ninthBrazilian congress of thermal engineering and sciences, CONEM; 2002.

G. Heidarinejad et al. / Building and Environment 45 (2010) 1521–15281528

[31] Dai YJ, Sumathy K. Theoretical study on a cross-flow direct evaporative coolerusing honeycomb paper as packing material. Applied Thermal Engineering2002;22(13):1417–30.

[32] Liao C, Chiu K. Wind tunnel modeling the system performance of alternativeevaporative cooling pads in Taiwan region. Building and Environment2002;37(2):177–87.

[33] Al-Sulaiman F. Evaluation of the performance of local fibers in evaporativecooling. Energy Conversion and Management 2002;43(16):2267–73.

[34] Scofield CM, DesChamps NH. Indirect evaporative cooling using plate typeheat exchangers. ASHRAE Transactions 1984;90:148–53.

[35] Al-Juwayhel FI, Al-Haddad AA, Shaban HI, El-Dessouky HTA. Experimentalinvestigation of the performance of two-stage evaporative cooler. HeatTransfer Engineering 1997;18:21–33.

[36] El-Dessouky H, Ettouney H, Al-Zeefari A. Performance analysis oftwo-stage evaporative coolers. Chemical Engineering Journal 2004;102(3):255–66.

[37] Heidarinejad G, Bozorgmehr M, Delfani S, Esmaeelian J. Experimentalinvestigation of two-stage indirect/direct evaporative cooling system invarious climatic conditions. Building and Environment 2009;44(10):2073–9.

[38] Sukhatme SP. Solar energy principles of thermal collection and storage.McGraw Hill; 1990.

[39] Clark E, Berdahl P. Radiative cooling: resource and applications. Berkeley; 1980.[40] Han YM, Wang RZ, Dai YJ. Thermal stratification within the water tank.

Renewable and Sustainable Energy Reviews 2009;13(5):1014–26.[41] Lavan Z, Thompson J. Experimental study of thermally stratified hot water

storage tanks. Solar Energy 1977;19(5):519–24.[42] Shah LJ. Investigation and modeling of thermal conditions in low flow SDHW

systems. Department of Building and Energy Technical University of Denmark;1999 [Report R-034].

[43] Furbo S. Optimum design of small DHW low flow rate solar heating systems.ISES Solar World Congress Budapest Report 93-24; 1993.

[44] Cristofari C, Notton G, Poggi P, Louche A. Influence of the flow rate and thetank stratification degree on the performances of a solar flat-plate collector.International Journal of Thermal Sciences 2003;42:455–69.

[45] Wang SK. Handbook of air conditioning and refrigeration. 2nd ed. McGraw-Hill; 2001.

[46] Jones WP. Air conditioning engineering. 5th ed. Edward Arnold; 2001.[47] Zukauskas AA. Heat transfer from tubes in cross flow. Advanced Heat Transfer

1972;8:93–160.[48] Dittus PW, Boelter LMK. Heat transfer in automobile radiators of the tubular

type. Heat Mass Transfer 1985;12:3–22.

[49] Dowdy JA, Karabash NS. Experimental determination of heat and masstransfer co-efficients in rigid impregnated cellulose evaporative media. ASH-RAE Transactions 1987;93(Part 2):382–95.

Nomenclature

A: total wetted surface area (m2)Ai: inner surface area of tubes (m2)Cp: specific heat of the fluid (J/kg �C)Cpu: specific heat of the humid air (J/kg K)Dout, Din: outer and inner tube diameter (m)F: collector efficiency factorha,hi: heat transfer coefficient of air side of the coil and tube side of the coil,

respectively (W/m2 �C)hc: convective heat transfer coefficient (W/m2 �C)hm: mass transfer coefficient (kg/m2 s)ia: is enthalpy of airivs, iv: specific enthalpy of vaporization of the water and vapor at surface

temperature (J/Kg)K: thermal conductivity (W/m K)l: pad thickness (m)le: characteristic length (m)Le: Lewis number_m: mass flow rate (kg/s)

n: number of parallel tubes in the collector structureS: emitted radiative energy to sky (W/m2)Ta: ambient air temperature (�C)Tcol: collector temperature (K)Tdp: dew point temperature (�C)Tf: outlet fluid temperature (�C)Tfi: fluid temperature at collector inlet (�C)Tsky: sky equivalent temperature (K)UL: overall heat loss coefficient (W/m2 �C)v: wind velocity (m/s)V: volume of evaporative pad (m3)w: distance between the tubes (m)y: tubes length (m)3r: hemispherical infrared emissivity3: saturation effectivenesshs: finned surface efficiencys: Stefan–Boltzmann constant (W/m2 �C)