performance analysis of masdar city’s concentrated...

Performance Analysis of Masdar City’sConcentrated Solar Beam-Down Optical

Experiment

Marwan MokhtarResearch Engineer

Laboratory for Energy and Nano Science (LENS)Department of Mechanical Engineering

Masdar Institute of Science and TechnologyAbu Dhabi, UAE

Email: [email protected]

Steven A. MeyersResearch Engineer



Peter R. ArmstrongAssociate Professor



Matteo ChiesaAssociate Professor



An analysis of the Beam Down Optical Experiment (BDOE)performance with full concentration is presented. The anal-ysis is based on radiation flux distribution data taken on 21stMarch 2011 using an optical-thermal flux measurement sys-tem. A hypothetical thermal receiver design is used in con-junction with the experimental data to determine the opti-mal receiver aperture size as a function of receiver lossesand flux distribution. The overall output of the plant is cal-culated for various operating temperatures and three differ-ent control strategies namely, constant mass flow of the HeatTransfer Fluid (HTF), constant outlet fluid temperature and

real-time optimal outlet fluid temperature. It was found thatthe optimal receiver aperture size of the receiver ranged be-tween (1.06 and 1.27 m2) depending on temperature. Theoptical efficiency of the BDOE ranged between 32% to 37%as a daily average. Depending on the control parameters andassuming an open receiver with solar absorptivity of 0.95and longwave emissivity of 0.10, the average receiver effi-ciency varied from 71% at 300oC down to 68% at 600oC.The overall daily average thermal efficiency of the plant wasbetween 28% and 24% respectively for the aforementionedtemperatures. The peak of useful power collected in the HTF

was around 105 kWth at 300oC mean fluid temperature andpeaked around 89 kWth at 600oC.

Nomenclatureα Receiver absorptivity in solar spectrumβ Volumetric thermal expansion coefficient (K−1)γ Receiver intercept factor∆T Temperature difference between mean plate tempera-

ture and ambient air temperature (K)ε Receiver longwave emissivityν Kinematic viscosity of air (m2/s)σ Stefan-Boltzmann constant (5.670373x10−8Wm−2K−4)A Receiver aperture area (m2)Cp Specific heat of the HTF at constant pressure

(kJkg−1K−1)Di Receiver inner tube diameter (m)Do Receiver outer tube diameter (m)F ′ Receiver efficiency factorFR Heat removal factorg Gravitational acceleration (m/s2)Gr Grashof numberG(r,θ) Radiation flux map in polar coordinates (kW/m2)hc Receiver convection heat transfer coefficient

(kWm−2K−1)h f Convection heat transfer coefficient between HTF and

tube wall (kWm−2K−1)hr Receiver linearized radiative heat transfer coefficient

(kWm−2K−1)k Thermal conductivity of air (Wm−1K−1)L Receiver characteristic length (m)m HTF mass flow rate (kg/s)Nu Nusselt numberPr Prandtl numberqin Solar flux intercepted by the receiver (kW/m2)Qrad Radiative heat loss (kW )Qu Useful thermal power (kW )R Receiver aperture radius (m)Re Reynolds numberTe Ambient air temperature (oC)Tf i Inlet mean fluid temperature (oC)Tf ilm Mean film temperature (oC)Tf o Outlet mean fluid temperature (oC)Topt Real-time optimized fluid outlet mean temperature

(oC)Tmp Mean plate temperature (oC)Tsky Effective sky temperature (oC)UL Overall heat loss coefficient (kWm−2K−1)

AbbreviationsBDOE Beam-Down Optical ExperimentCCD Charge-Coupled DeviceCR Central ReflectorCSP Concentrated Solar PowerDNI Direct Normal IrradiationHFS Heat Flux SensorHTF Heat Transfer FluidLEC Levelized Energy Cost

1 IntroductionCentral receiver Concentrated Solar Power (CSP)

plants promise higher energy conversion efficiency andconsequently lower levelized energy cost (LEC) [1]. It isexpected that 48% of the cost reduction in tower plants costswill come from technology research and development [2].The beam down concept has the potential of producingenergy at lower costs due to the ability of locating bulkyreceivers on ground level in addition to the possibility ofusing smaller receiver apertures and secondary concentratorswhich is made possible by the narrower view angle of con-centrated radiation [3]. In order to demonstrate the potentialof such configuration and investigate the possibility of ascale-up, a pilot 100kWth Beam-Down Optical Experiment(BDOE) was constructed in 2009 in Masdar City Abu Dhabi(24.44oN 54.61oE), see Fig. 1.The heliostat field consists of 33 ganged-type heliostats oftotaling 280 m2 of reflective area ( see [4] for more details).In this paper we present optical and thermal performanceanalysis of BDOE based on experimental data collected inMarch 2011.

Optical performance is quantified by the optical effi-ciency, which includes the effects of cosine loss, reflectivityloss, beam attenuation loss, blocking and shading loss inaddition to light spillage around central reflector (CR) mir-rors and around the receiver aperture (intercept factor).Thecombined effect of the aforementioned (except for the finalintercept factor) is evaluated experimentally using the fluxmeasurement system. The intercept factor is a functionof receiver aperture size/geometry and flux distribution.Consequently, the receiver aperture size must be specified inorder to evaluate the optical efficiency.

Thermal performance depends on the thermal efficiencyof the receiver, which, similar to optical efficiency, is also afunction of receiver aperture size/geometry. Thermal perfor-mance is also clearly dependent on the operating tempera-ture.Because of the inverse effects of receiver size on optical andthermal efficiency, an optimization is necessary to determinesize and operating temperature to maximize a certain objec-tive function. The objective function used in our case is totaldaily energy collection.

2 Experimental SetupTo study energy flux distribution and its magnitude on

the focal plane of the BDOE, concentrated solar radiation isintercepted by a 5m x 5m ceramic tile target, located 2.3mabove ground level (see Fig. 1). These white tiles can with-stand the high flux levels and are highly reflective in a dif-fuse manner approximating a Lambertian target. This allowsa CCD camera located at the top of the tower to measurethe distribution of luminous intensity. Embedded within thetiles at eight locations are heat flux sensors (HFS) to mea-sure the concentrated solar flux. The HFSs only provide dis-crete thermal flux data and the CCD camera provides data of

5

1 2

3 4

6

7

8

100 200 300 400 500 600 700 800 900

100

200

300

400

500

600

700

800

900 500

1000

1500

2000

2500

3000

3500

4000

4500

5000

(a)

Central Reflector

Ganged-Type Heliostat

Lambertian Target

CCD Camera Aperture

(b)

(c)

Fig. 1. BDOE Overview. (a) Ganged-Type heliostat field and central reflector mirrors. CCD camera aperture is in the center of the centralreflector taking shots of the Lambertian target below it. (b) Embedded within the Lambertian target are water-cooled heat flux sensors at eightlocations to calibrate the CCD camera images. (c) A typical luminance map taken by the CCD camera.

the luminous intensity on the target. A correlation is devel-oped between the flux sensor data and the CCD camera dataso that an accurate measurement of flux distribution can beobtained. In this way enough information of flux distribu-tion on the focal plane can be acquired at each moment oftime which allows the study of the plants performance undervarious conditions. [5] The CCD camera temperature is con-trolled by a combination of heating and cooling devices. APID controller regulates the air pressure supplied to a vortexcooler in order to keep the CCD camera temperature constantat 23oC [5]. The heat flux sensors used in the measurementsystem also require cooling. For accurate heat flux measure-ment, it is more important to prevent transients in water tem-perature than to keep a precise set point. Cooling water issupplied from a tank of sufficient thermal mass to preventany abrupt change in the water temperature and to operate forseveral hours with less than 20K temperature rise. The watertemperature going into the sensors and the outlet temperatureof each sensor are recorded and then used in correcting theHFS for thermal losses using a correction model developedin [6, 7]. The foregoing combination of optical and thermalradiometry provide the required flux mapping. While heatflux measurement is accurate and reliable, it is only feasi-ble to be implemented at discrete points on the target. Opti-

cal methods on the other hand are less accurate but providevery high spatial resolution. Heat flux measurement is usedas a reference for calibrating the optical system and hencecombining the advantage of the two methods. Similar hybridmeasurement systems have been used to measure heat fluxon the targets of parabolic dishes [8] and tower plants [9,10].

3 Performance AnalysisIn this section we present the method used for analyzing

the performance of the BDOE. The optical performance ofthe plant (not including the intercept factor) is evaluated fromthe experimental results acquired by the flux measurementsystem. Then the intercept factor and the receiver thermalperformance are evaluated simultaneously for varying oper-ating temperatures using the optimization algorithm which isdescribed later in section 3.2.

3.1 Receiver Thermal PerformanceHeat loss from solar receivers is a subject of much lit-

erature. McDonald [11] presented a comprehensive investi-gation of heat loss from cavity receivers including convec-tion, radiation and conduction. Receiver convection loss has

been studied in great detail by Clausing [12–14] and oth-ers [15–18]. These studies included analytical and numer-ical models supported by experimental studies for variousgeometries and receiver orientations. Most of these stud-ies however only considered losses from downward facingreceivers except for Leibfried and Ortjohann [15] who alsoconsidered upward facing receivers. Although some stud-ies considered forced convection [19], most studies neglectedforced convection loss from cavity receivers [12, 15, 16, 18].In this paper we will base our analysis on the performance ofan upward facing open receiver and we will neglect forcedconvection effects because the mean wind speed during thetest was small (2.6 m/s).Here we present the equations and procedure used in assess-ing the performance of a hypothetical receiver. The methodis based on Hottel-Whillier equation [20] as presented inDuffie and Beckman [21]. The Heat Transfer Fluid (HTF)selected for the analysis is liquid Sodium. Under steady stateconditions the net useful power output of the receiver is thedifference between the absorbed solar radiation and the ther-mal loss from the receiver surface.

Qu = A(αQin−UL(Tmp−Te)) (1)

Where Qu is the useful heat output, α is the absorptivityof the receiver in the solar spectrum (taken as 95%), Qin isthe solar power intercepted by the receiver aperture, UL is theoverall heat loss coefficient, Tmp is the mean plate tempera-ture and Te is the air temperature both in K.Eqn. (1) is reformulated in terms of the inlet fluid tempera-ture (Tf i) and the heat removal factor (FR). FR is introducedto account for the underestimated losses calculated based onthe inlet fluid temperature (Tf i) instead of mean plate tem-perature.

Qu = AFR(αQin−UL(Tf i−Te)) (2)

The heat removal factor FR reduces the useful gain fromwhat it would have been had the whole absorber plate beenat the inlet fluid temperature to what actually occurs [21]. FRis given by:

FR =mcp

AUL

(1− e−

AULF ′mCp

)(3)

Where m is the mass flow rate, Cp is the specific heatof HTF and F ′ is the receiver efficiency factor [21] that ac-counts for the thermal resistance between the plate and thefluid inside the tubes:

F ′ =1

ULDO

(1

ULDi+ 1

πDih f

) (4)

Where Do is the outer tube diameter, Di is the inner tubediameter and h f is the convection heat transfer coefficientbetween the inner wall and the fluid. For liquid metals likeliquid sodium, convection heat transfer correlation betweenthe fluid and the tube walls is given by [22]:

Nu f = 6.3+0.0167R0.85e Pr0.93 Turbulent

Nu f = 4.36 Laminar(5)

The overall heat loss coefficient UL accounts for heatlosses in the receiver. There are five main loss mechanismsassociated with the receiver of a concentrated solar plant:convective loss, radiative loss, conductive loss, loss due toreflection, and spillage. Conductive heat losses are assumedto be negligible conpared to other loss mechanisms, whilelosses due to spillage are accounted for in the calculation ofintercepted power by the reicever (section 3.2), and reflectivelosses and spillage around the CR are accounted for by thefact that radiation is measured at receiver aperture level.According to [23], in a well designed receiver, thermal lossesaccount for 5%-15% of total available energy, hence a rea-sonable estimate using Tmp is adequate to determine receiverperformance. We assume, as mentioned earlier, that the dom-inant convection mode is free convection and that forced con-vection is negligible. Grashof number for free convection isgiven by:( [22])

GrL =β∆T gL3

ν2 (6)

Where β is the volumetric thermal expansion coefficientin K−1,∆T is the temperature difference between mean platetemperature (Tmp) and ambient air temperature (Te) in K, g isthe gravitational acceleration in m/s2, L is the characteristiclength of the receiver equal to it diameter in m, and ν is thekinematic viscosity of air in m2/s. The Nusselt number forlaminar and turbulent flows are given by:( [22])

Nunatural = 0.54(GrLPr)1/4 ,TurbulentNunatural = 0.14(GrLPr)1/3 ,Laminar

(7)

Therefore, the convection heat transfer coefficient (inW/m2K) is given by

hc =kNu

L(8)

Where k is the is the conductivity of air in W/mK. Fluidproperties are evaluated at the mean film temperature calcu-lated as Tf ilm = (Tmp +Te)/2.The radiative heat loss is given by:

Qrad = Aεσ(T 4mp−T 4

sky) (9)

0 0.5 1 1.5 2 2.5 30

0.2

0.4

0.6

0.8

1

Radius (R) [m]

Inte

rce

pt

Fa

cto

r (γ

)

Fig. 2. Day-average intercept factor (γ) as a function of receiveraperture radius (R).

Where Qrad is the radiative heat loss in W , εis the longwaveemissivity (taken as 10%),Tmp is the mean plate temperatureand Tsky is the effective sky temperature both in K. Hencethe linearized radiative heat transfer coefficient relative to airtemperature may be defined as:

hr = AεσT 4

mp−T 4sky

Tmp−Te(10)

The mean plate temperature Tmp required for solving theprevious equations can be calculated by:

Tmp = Tf i +Qu

AFRUL(1−FR) (11)

Finally,the overall heat loss coefficient is given by:

UL = hr + hc (12)

3.2 Receiver Intercept FactorDue to various optical and mechanical error sources in

the concentration system in addition to sunshape. The powerintercepted by the receiver will depend on the receiver aper-ture size. Fig. 2 depicts the day-average intercept factor vari-ation with receiver aperture radius. The receiver interceptfactor γ is calculated as follows:

γ =

∫ R0∫ 2π

0 G(r,θ)dθdr∫∞

0∫ 2π

0 G(r,θ)dθdr(13)

where γ is the receiver intercept factor, R is the receiveraperture radius, G(r,θ) is the flux map defined in polar coor-dinates.

6 8 10 12 14 16 180

5

10

15

20

25

30

35

40

45

50

Time [hr]

Eff

icie

ncy [

%]

Optical EfficiencyAverage Eff = 37%, 34%, 33%, 32%,

Tfo

= 300oC

Tfo

= 400oC

Tfo

= 500oC

Tfo

= 600oC

Fig. 3. Optical efficiency with optimal aperture

3.3 Receiver Aperture SizingOptimal receiver aperture size is determined by maxi-

mizing the day average useful power in Eqn. 2. This is cal-culated based on the required outlet mean fluid temperature.An iterative algorithm is used for evaluating the receiver per-formance and finding the optimal receiver size. Dependingon the temperature/flow control strategy used, the algorithmis slightly modified. Below is an outline of the major calcu-lation steps used:

1. Calculation of cumulative flux maps which give therelation between receiver size (R) and interceptedpower(Qin).

2. Specification of control strategy and setpoint of controlvariable : mass flow, outlet temperature or optimal outlettemperature.

3. Selection of a receiver radius (R).4. Calculation of receiver performance based on input

paramters : Tf i,Tf o,R,Qin and ambient conditions.

(a) Radiation and convection heat transfer coefficients(hr, hc) are calculated using an initial value for themass flow rate (m) and mean plate temperature(Tmp).

(b) Receiver efficiency factor (F ′) is calculated.(c) m is updated using the calculated (F ′) and overall

heat loss coefficient (UL).(d) Net useful power (Qu) is updated using the new m

where Qu = mcp(Tf o−Tf i) .(e) Heat removal factor (FR)is updated with the new

Qu.(f) Tmp is updated.(g) Steps a-f are repeated until Tmp is accurate enough.

5. The calculation is repeated for different R and Tf o.

4 ResultsIn this section we present the results obtained from the

BDOE performance under full concentration using the exper-imental data collected and the receiver model we presentedin previous sections.

6 8 10 12 14 16 18

0.4

0.5

0.6

0.7

0.8

0.9

1

Time [hr]

Inte

rce

pt

Fa

cto

r (γ

)

1.0 [m]

1.5 [m]

2.0 [m]

Fig. 4. Intercept factor variation during the day for different receiveraperture sizes

4.1 Optical EfficiencyFig. 3 depicts the optical efficiency calculated based on

the optimal receiver size chosen for each fluid outlet tem-perature. Optical efficiency represents the amount of energythat is intercepted by the receiver aperture normalized by theproduct of incident DNI and heliostat Area. Optical effi-ciency accounts for all the factors that reduces the amountof concentrated radiation until it reaches the receiver, but be-fore it gets converted into thermal power.Optical efficiency includes cosine factor, reflectivity of bothheliostats and the CR mirrors, beam attenuation, CR inter-cept factor (spillage), incident angle modifier, blocking andshading of heliostats,and the intercept factor of the receiver.In Fig. 3, it can be seen that the average efficiency of the re-ceiver is varying with outlet temperature from 32% to 37%;this is because the optimal receiver size changes based onthe required outlet temperature. Since higher temperaturescall for smaller receiver apertures, the optical efficiency isdegraded. This is directly caused by the lower intercept fac-tor for smaller receiver radii.Fig. 4 depicts the daily intercept factor variation for differentreceiver sizes. The daily intercept factor curve is almost flatin the middle region when sun elevation angle is highest butthen drops rapidly in early morning and late afternoon. Thisis mainly caused by optical aberrations (astigmatism in par-ticular) that causes an elongated flux distribution, and there-fore results in radiation to be spill outside the receiver aper-ture. This can be seen more clearly in Fig. 5 which illustratesflux maps taken at three different times of the test day.

4.2 Receiver Thermal EfficiencyReceiver thermal efficiency is a function of its mean

temperature and aperture size. Fig. 6 illustrates as a functionof aperture radius, the incident power on the receiver, con-vection losses, radiation losses, and net useful output. Thecurves are calculated for an outlet mean fluid temperature of400 oC. It can be seen that a maxima for net power occursat a certain receiver radius. Fig. 7 depicts the daily variation

0 0.5 1 1.5 2 2.50

10

20

30

40

50

60

70

80

90

100

Radius [m]

Po

we

r [k

W]

Net Useful Power (Day Average), Tfo

= 673[K]

Net Useful Power

Incident Power

Radiation Loss

Convection Loss

Fig. 6. Net power collected as a function of receiver radius. Con-vection and radiation losses are also shown as a function of receiversize

6 8 10 12 14 16 180

10

20

30

40

50

60

70

80

90

Time [hr]

Eff

icie

ncy [

%]

Receiver EfficiencyAverage Eff = 71%, 73%, 71%, 68%,

Tfo

= 300oC

Tfo

= 400oC

Tfo

= 500oC

Tfo

= 600oC

8LIVQEP�)�GMIRG]8LIVQEP�)�GMIRG]6IGIMZIV�8LIVQEP�)�GMIRG]

Fig. 7. Receiver thermal efficiency

of receiver thermal efficiency for several outlet fluid temper-atures. It can be seen from Fig. 7 that the receiver thermalefficiency at 400oC (R=1.27 m) is less than that at 500oC(R=1.15 m) or 600oC (R=1.06 m).This is because the opti-mal receiver aperture size at 400oC is larger, which resultedin a higher thermal losses despite the lower operating tem-perature. At higher temperatures the effect of thermal lossesstarts to dominate over the aperture size and we notice thatthermal efficiency becomes mainly a function of outlet tem-perature. It is important to note that since receiver aperturesize is optimized for overall efficiency, the overall efficiencyat 400oC is still higher than those for higher temperatures.

4.3 Overall Efficiency and Power OutputThe overall efficiency (Fig. 8) is the product of optical

and thermal efficiencies, as expected it is inversely propor-tional to the outlet temperature and varying on average from22% to 26%. Fig. 9 depicts the thermal power output (kW )collected in the HTF for different temperatures. The total en-ergy varies from 623.1 kWh at 400oC to 529.2 kWh at 700oC.

Fig. 10 depicts the maximum possible mechanical power

Time 08:00

100 200 300 400 500 600

100

200

300

400

500

600

0.5

1

1.5

2

2.5

x 104

Time 16:00

100 200 300 400 500 600

100

200

300

400

500

600

1

2

3

4

5

6

x 104

Time 12:00

100 200 300 400 500 600

100

200

300

400

500

600

1

2

3

4

5

6

7

8

9

10

x 104

Fig. 5. Flux maps at different times of the day. Aberration is evident in early and late parts of the day which correspond to reduced interceptfactor

6 8 10 12 14 16 180

5

10

15

20

25

30

35

40

45

Time [hr]

Eff

icie

ncy [

%]

Overall Thermal EfficiencyAverage Eff = 28%, 26%, 25%, 24%,

Tfo

= 300oC

Tfo

= 400oC

Tfo

= 500oC

Tfo

= 600oC

Fig. 8. Overall efficiency

6 8 10 12 14 16 180

10

20

30

40

50

60

70

80

90

100

110

Time [hr]

Th

erm

al P

ow

er

Co

llecte

d in

HT

F [kW

]

Tfo

= 300oC

Tfo

= 400oC

Tfo

= 500oC

Tfo

= 600oC

Fig. 9. Thermal output of receiver as function of time and outlettemperature

(exergy rate) of the fluid for varying temperatures, calculatedbased on Carnot efficiency assuming cold reservoir tempera-ture of 80 oC. Maximum possible mechanical power is alsocalculated for real-time optimized fluid outlet temperature(Tf o = Topt ).

6 8 10 12 14 16 180

10

20

30

40

50

60

Time [hr]

Maxim

um

Mechancia

l P

ow

er

[kW

]

Tfo

= 300oC

Tfo

= 400oC

Tfo

= 500oC

Tfo

= 600oC

Tfo

= Topt

Fig. 10. Maximum Mechanical Power

4.4 Comparison of Control StrategiesThe selection of the control strategy is often dictated

by the consumer process, most processes require a constanttemperature supply from the heat source. This obvious chal-lenge for solar systems maybe overcome by some kind ofthermal inertia (storage) or hybridization with fossil fuelbased sources. Therefore the solar engineer might have thechoice between constant-temperature-variable-flow controlstrategy, constant-flow-variable-temperature control strategyor a variable-temperature-variable-flow control strategy thatmaximizes exergy (or any other objective function).While constant flow control is the easiest to implement prac-tically because it requires less hardware and also simplercontrol the merits of the other control strategies might justifythe change, this has to be assessed on a system level. Theobjective function must be modified to include the consumerprocess operation parameters.

Fig. 11 shows simulation results for the three afore-mentioned control strategies. It can be seen in (a) that thedifferences between the three control strategies is negligible.This is because the constant temperature and constant flowrate were also selected by maximizing the objective function.

Fig. 11. Comparison of Control Strategies. (a) Daily variation of Maximum Mechanical Power, (b) Mean Fluid Output Temperature, (c) MassFlow Rate

If the temperature deviates from the optimized constanttemperature then the Mechanical Power will be lower (seeFig. 10). It should be also noted that the although theoptimal mass flow and outlet temperature are constant inthis example day, they are expected to change during theyear. Moreover, the small differences between the controlstrategies are affected by receiver design which have lowthermal losses, when thermal losses increase, the realtime optimized temperature control strategy may have asignificant advantage.

5 Discussion and ConclusionsThe overall efficiency of the BDOE varied from 24% to

28% in the test day. The main losses in the system resultfrom optical losses : cosine loss, reflectivity loss, beam at-tenuation loss, blocking and shading loss in addition to lightspillage around Central Reflector (CR) mirrors and aroundthe receiver aperture (intercept factor). The cost incurred bythe high optical losses can be justified by the higher temper-atures that could be achieved.

References[1] Kaltschmitt, M., Streicher, W., and Wiese, A., 2007.

Renewable energy : technology, economics and envi-ronment. Springer, New York.

[2] Group, S. . L. L. C. G. . L. L. C., 2003. Assessmentof parabolic trough and power tower solar technologycost and performance forecasts. Tech. Rep. NREL/SR-550-34440, Chicago, Illinois, Oct.

[3] Tamaura, Y., Utamura, M., Kaneko, H., Hasuike, H.,Domingo, M., and Relloso, S., 2006. “A novel beam-down system for solar power generation with multi-ringcentral reflectors and molten salt thermal storage”. InProceedings of the 13th International Symposium onConcentrating Solar Power and Chemical Energy Tech-nologies.

[4] Hasuike, H., Yuasa, M., Wada, H., Ezawa, K., Oku,K., Kawaguchi, T., Mori, N., W.Hamakawa, Kaneko,H., and Tamaura, Y., 2009. “Demonstration of tokyotech beam-down solar concentration power system in

100kw pilot plant”. In Proceedings of 15th Interna-tional Symposium on Concentrated Solar Power andChemical EnergyTechnologies.

[5] Meyers, S. A., 2011. “Flux mapping of the beam downsolar thermal concentrator at masdar city, uae”. Mas-ter’s thesis, Masdar Institute of Science and Technol-ogy, Abu Dhabi,UAE.

[6] Mokhtar, M. M., 2011. “The beam-down solar thermalconcentrator experimental characterization and model-ing”. Master’s thesis, Masdar Institute of Science andTechnology, Abu Dhabi,UAE.

[7] Mokhtar, M. M., Meyers, S. A., Rubalcaba, I., Chiesa,M., and Armstrong, P. R., 2012. “A model for improvedsolar irradiation measurement at low flux”. Solar En-ergy, 86(3), pp. 837–844.

[8] Ulmer, S., Reinalter, W., Heller, P., Lupfert, E., andMartinez, D., 2002. “Beam characterization and im-provement with a flux mapping system for dish concen-trators”. Journal of Solar Energy Engineering, 124(2),pp. 182–188.

[9] Ballestrın, J., and Monterreal, R., 2004. “Hybrid heatflux measurement system for solar central receiver eval-uation”. Energy, 29(5–6), pp. 915–924.

[10] Ulmer, S., Lupfert, E., Pfander, M., and Buck, R.,2004. “Calibration corrections of solar tower flux den-sity measurements”. Energy, 29(5–6), pp. 925–933.

[11] McDonald, C. G., 1995. Heat loss from an open cavity.Tech. Rep. SAND95-2939, Sandia National Labrato-ries.

[12] Clausing, A., 1981. “An analysis of convective lossesfrom cavity solar central receivers”. Solar Energy,27(4), pp. 295–300.

[13] Clausing, A. M., 1983. “Natural convection correla-tions for vertical surfaces including influences of vari-able properties”. Journal of Heat Transfer, 105(1),pp. 138–143.

[14] Clausing, A. M., Waldvogel, J. M., and Lister, L. D.,1987. “Natural convection from isothermal cubical cav-ities with a variety of side-facing apertures”. Journal ofHeat Transfer, 109(2), pp. 407–412.

[15] Leibfried, U., and Ortjohann, J., 1995. “Convectiveheat loss from upward and downward-facing cavity so-

lar receivers: Measurements and calculations”. Journalof Solar Energy Engineering, 117(2), pp. 75–84.

[16] Taumoefolau, T., Paitoonsurikarn, S., Hughes, G., andLovegrove, K., 2004. “Experimental investigation ofnatural convection heat loss from a model solar con-centrator cavity receiver”. Journal of Solar Energy En-gineering, 126(2), pp. 801–807.

[17] Prakash, M., Kedare, S., and Nayak, J., 2009. “Inves-tigations on heat losses from a solar cavity receiver”.Solar Energy, 83(2), pp. 157–170.

[18] Paitoonsurikarn, S., Lovegrove, K., Hughes, G., andPye, J., 2011. “Numerical investigation of natural con-vection loss from cavity receivers in solar dish applica-tions”. Journal of Solar Energy Engineering, 133(2),p. 021004.

[19] Ma, R. Y., 1993. Wind effects on convective heat lossfrom a cavity receiver for a parabolic concentrating so-lar collector. Tech. Rep. SAND92-7293, Sandia Na-tional Labratories.

[20] Hottel, H., and Whillier, A., 1958. “Evaluation of flat-plate solar collector performance”. In Transactions ofthe Conference on Use of Solar Energy (E. F.Carpenter,ed.), University of Arizona Press, p. 74.

[21] Duffie, J. A., and Beckman, W. A., 2006. Solar engi-neering of thermal processes, 3rd edition ed. Wiley.

[22] Mills, A. F., 1998. Heat Transfer, 2nd edition ed. Pren-tice Hall.

[23] Winter, C.-J., Sizmann, R. L., and Vant-Hull, L. L.,1991. Solar Power Plants: Fundamentals, Technology,Systems, Economics, first ed. Springer.

performance analysis of masdar city’s concentrated...

Documents