1 s2.0 0038092X79901701 Mai

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

SMar F~t,~y, Vol. 22, I~. 413-4.~ 003g-O92X/79~J01-0413/$02.

00/0
© Persataon Press Ltd., 1979. Printed in Great Britain

ANALYSIS, DEVELOPMENT AND TESTING OF A FIXED


TILT SOLAR COLLECTOR EMPLOYING REVERSIBLE
VEE-TROUGH REFLECTORS AND VACUUM TUBE RECEIVERS

M. KUDRETSEL(~UK
Jet Propulsion Laboratory, CaliforniaInstitute of Technology,Pasadena, California,U.S.A.

(Received 8 .lane 1978" revision accepted 24 October 1978)

Abstract--The Vee-Tmugh/Vacuum Tube Collector(VTVTC) aimed to improve the efficiencyand reduce the cost
of collectors assembledfrom evacuated tube receivers.
The VTVTC was analyzed rigorously and a mathematical model was developed to calculate the optical
performance of the vee-tmugh concentrator and the thermal performance of the evacuated tube receiver. A test
bed was constructed to verify the mathematical analyses and compare reflectors made out of glass. AIzak and
aluminizedFEP Teflon. Tests were run at temperaturesrangingfrom 95 to 180°Cduring the months of April, May,
June, July and August 1977. Vee-trough collector efficienciesof 35-40 per cent were observed at an operating
temperature of about 1750C.Test results compared well with the calculatedvalues. Test data coveringa complete
day are presented for selected dates throughout the test season.
Predicted daily useful heat collection and efficiencyvalues are presented for a year's duration at operation
temperatures rangingfrom 65 to 230°C. Estimated collectorcosts and resultingthermal energy costs are presented.
Analyticaland experimentalresults are discussed along with an economicevaluation.

1. lIeri~C'l'lON centrator is also being considered for use with a flat plate
This paper discusses the analyses and test experiments collector to enhance its output at high temperatures[8].
conducted on vet-trough concentrators to prove their In addition to the use of a selective coating on the
usefulness in improving the efficiency and reducing the absorber with high a and low ~ to reduce radiation
cost of collectors assembled from evacuated tube losses, which is widely applied now, honeycomb cell
receivers. convection suppressors[9] or reduction of convective
An asymmetmal-reversible vee-trough reflector main- losses by partial evacuation of the space between the
taining a year-round concentration factor of about 2 has absorber plate and a transparent cover[10, 11] have also
been studied. Figure I illustrates the principle of opera- been attempted to assure high efficiencies. The former
tion of such a collector having an optical concentration reduces the incoming flux by absorption and increases
ratio of about 3. the backward conduction. Moreover, potential material
These collectors are applicable for heating/cooling, as problems exist with plastic honeycombs, and the glass
well as total energy systems and small-scale rural power honeycomb is expensive.
supplies (especially in combination with an organic fluid Evacuated flat plate collectors have numerous prob-
turbine), such as solar pumping stations. The perfor- lems. Among them are stresses on the glass plates and
mance of a solar Rankine, mechanical compression, air difficulties of maintaining vacuum during lifetime (which
conditioning system would also be enhanced by using the requires either expensive vacuum seals or continuous
proposed collector assembly[l]. operation of a vacuum pump). Plastic covers for evacu-
ated flat plate collectors offer some advantages over
7..LITE~TUItEREVIEW glass from a stress standpoint: however, operational
The conventional flat plate collector has been studied problems such as scratching, distortion and even melting
and built in various forms for almost a hundred years. under static conditions and degassing under vacuum
Present cost projections for these types of collectors will must be considered. Recently, the design originally pro-
probably not reduce significantly since material posed by Speyer et al. in the 1960s[12], using evacuated
requirements are substantially the same regardless of tube collectors mgde of borosilicate glass tubes with a
variation in structural design. Among attempts made to fiat plate absorber, has been tested as a non-tracking
improve the fixed collector performance and to reduce solar heat collector[3]. Evacuated tubes of the thermos
its cost are the use of mirror boosters in the early bottle type are also being offered[4]. The latter design
1960s[2], the recent introduction of vacuum tube employs a diffusely reflecting rear surface to boost the
collectors[3,4], and the use of vee-trough reflectors. collector output. The effect is more pronounced at the
Vee-trough reflectors to improve solar cell performance, off-noon periods, during which time the ratio of the heat
as proposed by Hollands[5] and by Durand[6], have collected to the daily total insolation available is not
recently been used in a box-type flat plate collector by significant[13]. Although the performance of a collector
Bannerot and Howell[7]. The compound parabolic con- consisting of a closely spaced array of vacuum tubes is

413
414 M. KUDRET SELCUK

DIMENSIONS OF W= 44.4 me ~ .t UL reaches to 3.SW/mZ°C thus the stagnation tem-


v,cuu,, nm 6. me / ; perature for a solar flux level of 1000W/m2 with
~CEIVER A,- 3.8me /A t -
(COURTESY CC~U'41~ D,, fiX)me jr/-- 8 I ... reflectors is about 500°C. Without reflectors the stag-
G U ~ WORKS t - 3.8 me ir w r - - - w-..- i nation temperature is about 250°C. Black chrome coating
comb, x;. N.y.) d : 8~me ,~ / , ~ I i ..P.,-d,I
- 7.) me IV" ; B D ! F 71 is reported to survive at these temperatures for several
L - 2.13. ~\ .... : , ~ 1 ~ / I
r - 0.lime /[ ",,,~,'T'-'~ I days[20]. Larger concentration ratios are unsafe for dry
(ELECTRON / I ~ J run operations unless special measures are taken for
IEAM WELD) --'L/IONO~ O =I
protection in case the fluid circulation stops. The General
THICKNESS
Electric Company has recently introduced a back-
FLUID GLASS 7VtF,.~U~
OUTLET ENV1ELOPE- ~ reflecting concentrator (a low concentrating parabolic
FLUID ~ --~-" .. ,, '. cylinder) having an optical concentration ratio on the
,
order of 2. It has a thermos bottle type of evacuated
GLASS- TO- R SPRIN TRANSFER glass envelope (which is hermetically sealed) and a con-
METAL SEAL- - f PLATE "--/ CLIP - - ' TUIE centric cylindrical plane receiver[21].
COLLECTION AREA -I= COLLECTIONAREA =~
3. ANALYSISOF THE VEE-TROUGH CONCENTRATOR AND VACUUM
W WI/'4fER r ' - SUMMER I ~ REVERSIBLE
TUBE RECEIVER
\ / I OELTA I '~ / i The analysis used in this project was based on a
\ / I SecT~N I \ / i
/ I REFLECTORtit' \ / L mathematical model of the vee-trough vacuum tube col-
\ / ~MOOULe- < , L "l
'~.
i
HOSE lector incorporating an optical model of the vee-trough
,i X ', I -'-.<\ ',cLAM,. concentrator and a thermal model of the vaccuum tube
" \' ,' "X , /
receiver. Varying solar flux and ambient conditions were
considered.
TU " IY__ ;I " -
cLAm, I ~ ~I---VACUUM I ~ [ - ~ R
AND I ~ ~ I R E C E I V E e I/ L~ J.l 3.1 Analysis of the optical model
REFLECTOR ! q l ~ f T l l l /TUBE ~1 ql~Ttl!
SUPPORT~'IIIlililIIII i II tlirilliilllilJ The following assumptions were used in the for-
"G ' )l
(1/3 COLLECTION mulation:
(1) The solar beam is speculafly reflected from the
mirror surface having a reflectivity of p. Since the target
Fig. I. Reversible asymmetric vee-troughcollectorwith vacuum size is about 1/3 of the aperture size and, for a practical
tube. design, the flap width is on the order of 30cm, the
divergence of the reflected beam due to the parallax of
superior to conventional flat plate designs, its cost is solar rays and surface errors and roughness were
expected to be well above the simple flat plate. ignored.
Recently the evacuated tube receivers have been (2) The diffuse radiation intensity at the bottom of the
examined both analytically and experimentally. Among vee-trough is assumed to be about 80 per cent of the
these investigations are the original studies undertaken at diffuse radiation incident on the aperture plane. This
Coming Glass Works by Dr. U. Ortaba~l et al.[14, 1S] assumption, previously based on data in Ref.[5], was
and more recently by Dr. S. Karaki et al. at the Colorado later confirmed when the flux intensity on the aperture
State University, Fort Collins, Colorado[16]. Both in- plane was compared to the flux intensity at vee-trough
vestigators have examined vacuum tube collector tubes concentrator bottom during an overcast day.
and modules without a concentrator. The latter, (3) Surface reflectivity is taken to be dependent upon
however, has examined reflections from a back sheet. the angle of incidence as given in Ref.[22]. However,
Use of vacuum tube receivers with moderately con- change of reflectance with the wavelength was neglected.
centrating systems has also been considered by some (4) Secondary reflection of the beam radiation was
investigators and research teams. Argonne National considered.
Laboratories[17-19] has developed a compound (5) End effects, namely changes of the position of the
parabolic concentrator (CPC) in connection with a reflected beam along the tube axis and effects on the
vacuum tube receiver for concentration ratios above 3. total energy incident on the absorber tube, were taken
The tilt of a CPC with a concentration of 3 must be into consideration.
adjusted twice a year, unlike the asymmetrical vee- (6) The glass envelope (Pyrex tube) has a sharp cut-off
trough suggested in this project which requires only of the transmissi'vity curve for the wavelengths ,~ > 4~.
reversal of the reflector twice a year. Under stagnation The final optical model yields the flux intensity dis-
conditions temperatures may reach to levels which are tribution over the absorber plate which is placed inside
detrimental to the selective coatings. The stagnation the glass envelope. Figure 2 illustrates the mirror and
temperature is a function of the overall heat loss tube geometry and incoming, reflected and transmitted
coefficient of the tube (UL) and the concentration ratio. beams traced. Both sides of the vee-trough are divided
The Vee Trough concentrator--Evacuated tube into strips as shown in Fig. 2. The angle of incidence 0b)
receiver combination suggested has no transparent cover is determined for each ray. The point of intersection with
unlike the CPC designs developed. Therefore the UL is the absorber plane as well as the attenuated solar flux
lower than the 3X CPC design. At stagnation calculated intensity are determined.
Analysis, development and testing of a fixed tilt solar collector 415

taken, The thermal model of the vacuum tube receiver


wiih~iit the concentrator is almost identical to those
developed by Karaki and Ortaba~L However, the
configuration used in the models developed by these
investigators has included the effect of neighboring
vacuum tubes on the tube under study. These effects of
shadowing and reflection as well as reflections from a
rear plate do not apply in the configuration studied in this
project. Since the centerlines of the tubes are about 3
diameters apart, and the space between the tubes is filled
with the vee-trough reflectors, the tubes themselves are
assumed to have no effect on each other.
Analysis of vacuum tube without re]lectors. The fol-
lowing assumptions were used in the formulation of the
mathematical model of a single vacuum tube:
(1) Convection inside the tube is completely eliminated
since the pressure is P < 5 . 3 3 x 1 0 -3Pa ( < 0 . 4 ×
Vlt-Tl~ith CllCIltrlttil with Circullir GllSl Envelooe I~liIIcil
llio~e 1 (VTCG[1) 10-" torr). Studies in Ref.[15] reveal that under a vacuum
level of P < 1.33 x 10-2 Pa (P < 10-4 torr) convection
(a) i+ :
losses are reduced to a level so that effects inside the
tube can be disregarded.
(2) Conduction to the wall through the clips, attached
to the absorber plate to center it in the glass tube, is
neglected.
ADI
(3) Conduction through the U tube and manifolding is
STRIP (i
on the order of 5-10 per cent for temperatures around
sTn,P.~ _ _ j .7"..//"////'-/.
1 ) ~ 100 and 150°C, respectively. A correction factor was
applied, as will be explained later.
Dimensions of the vacuum tube under study are given
in Fig. 1, as supplied by the manufacturer, Ref.[24].
ZR , z Additional data are: Absorber plate length, L = 2.13 m:
_ ~ 1, ;-"
- ~ + , ; ,~-,
. ~~_
- fxili.~i
f?*~ Selective coating absorptivity a = 0.935: Selective coat-
zRi+ ing emissivity, ~p = 0.08; Emissivity of the underside of
( S t r i p s ) Ilodel fop the i e i - l l ~ i l h (VTFR1) - (VTFI2) the plate, eph = 0.12t; Glass surface emissivity, e, = 0.88;
(b) Glass index of refraction, n = 1.472; Glass extinction
Fig. 2. Vee-trough concentration with circular glass envelope coefficient, K = 0.078 c m - ' : Thermal conductivity of the
optical model. plate, k = 385 W/m°C.
Energy balance of the vacuum tube receiver. First Law
The total integrated insolation incident on the absorber of Thermodynamics can be applied to the vacuum tube
plate can be calculated from: receiver. The energy balance equation for the absorber
plate considering thermal storage effects is
Q, = ~ r(~)(ley(dz)Lih + r(~)(IRr(dz)L,):
i=m i=n l,(ra),Ap = Q, + Qt + Q,. (6)
+ IBAp~B + laAp~a. Since the thermal capacity of the working fluid and the
tubes is low, Q~ may be neglected. The total useful
Symbols are given in the Nomenclature. energy gain of the collector Q, for quasi-steady-state
Once the integrated solar flux over the absorber plate operation can be expressed as
is determined, then the flux concentration ratio (CR) can
Q, = mCpA T = ApFRIl,('ra), - U,.( Tt.,- 7",)]. (7)
be calculated from:

CR = Total energy incident on the absorber plate with vee-trough (5)


Total energy incident on the absorber plate without vee-trough'

3.2 Thermal analysis of the vacuum tube receiver Derivations of equations giving FR, (ra)e and Ut~ are
Thermal analyses of the vacuum tube receiver with given in Ref. [23].
and without the vee-trough concentrator were under- The vacuum tube efficiency is obtained as

";'epb is greater than ep due to the fact that in the coating


process the coating material adheres to parts of the underside of ~ - ~~. (8)
the absorber plate.
416 M. KUDRETSELCUK

where the incident solar heat input is The optical model, which considers the secondary
reflections, yield a concentration ratio, on the average,
Q~ = I,Ap. between 5-10 per cent above the results of the first
reflections model.
Formulation of the vacuum tube thermal model with The insertion of a circular glass envelope, having an
re/lectors. The useful heat and efficiency equations for average transmissivity of 0.92, reduces the yearly
tubes with concentrator can be derived in steps similar to average concentration ratio of 1.723. This corresponds to
that described for plain tubes. The useful heat, the an equivalent transmission factor of about 1.723/2.013 =
incident solar heat input based on the aperture area and 0.855, if first and second reflections are considered.
the hourly efficiency of the collector, respectively, are Since a transmission loss of 92 per cent had to be
given below: considered even with a clear pyrex tube, the curvature of
the tube results in losses with a factor of only
Q. = F R A , [ C R I , ( ' f a ) . - UL( TI., - To)] (9) 0.92/0.855 = 1.076.
Changing mirror surface reflectivity values of 0.9, 0.85
Qi. = l, Ac (I0)
and 0.8 reduces the concentration ratio. The reduction of
the concentration ratio, however, is not as large as the
~ = F" A-t-[ CR(I"°t)" --~, ( T" - T~)] (11)
ratio of reflectivities. For example, for 8 July at solar
noon, the calculated concentration ratio absorptivity
4. RESULTSOF THE MATHEMATICALANALYSF_~ product is CR ra = i.775 for a specular reflectivity of
Results of computer codes that were generated to p = 0.9, and CR ~'a = 1.667 for p = 0.8. The reduction of
solve mathematical models described in Section 3 are the reflectivity is 0.9/0.8= 1.125, whereas the concen-
discussed in this section. tration ratio is reduced only by a factor of i.77.5/1.667 =
4.1 Optical model results 1.064.
Results of the optical model without a glass envelope
are given for selected days of the year in Fig. 3. Day-long 4.2 Thermal models results
variation of the concentration ratio for a surface having a Solutions to the thermal model of the vacuum tube
reflectivity of p = 0.9, a collector plane tilt of 34.10°, and receiver with and without vee-trough concentrators were
flap angles of 0, = 55° and 02 = 850 are given. Flap widths obtained. The procedures were as follows.
are 33.6cm and 28.5 cm for wide and narrow flaps, 4.2.1 Solution without reflectors. First UL was cal-
respectively. culated using the glass wall temperature for various
Near the solstices the concentration ratio varies from a ambient temperatures and wind velocities under the
figure of about 1.15 to a peak of about 2.3, whereas previously mentioned assumption that the sky and am-
during equinoxes it is constant around 1.4. bient temperatures. Then the heat removal factor Fa was
The daily average concentration ratio (which is defined calculated for various UL, T,, and m values. Once the UL
as the ratio of the total incident concentrated flux on the and FR values were obtained, the useful heat O,, and the
absorber plate during the period from 8:00am to collector efficiency were calculated for given solar heat
4:00 pm to the incident flux on the receiver without any fluxes by means of eqns (9) and (!1), respectively.
concentrators) is given in Fig. 4 for a period of one year. Details are presented in Ref. [23].

2.61-
2.4 t .....-- -~. r--Aug 17

~~ I= 1-~-- _":'~z__:::.--~.-_'~:----~,--:~-- ~oc,


o o I.---.-. . . . . . . . . . . oc,8
¢...... ,:. -. . . . . . . . . . ----,<-;,: ..... s,p, 4
re ,.8 ............. ---- ................
I - / /,;7 X,, Fo ,,
1.6 I- June 22 ~ .

,. . . . . . . . . . . . . . ::::::::::::::::: . . . . . . . . . . . ,,

i 1.2 ' f . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ S e p t 22

1.0 lst REFLECTIONS ONLY


0.8 REFLECTIVITY =0.9
TILT=LATITUDE • 34.10"
0.6
O4
0.2
0.0 J = L • I I I I
8 9 I0 II 12 13 14 15 16 TIME OF THE DAY
NOON (hours)

Fig. 3. Daylong variation of the concentration ratio for a vee-trough concentrator.


Analysis, development and testing of a fixed tilt solar collector 417
I AVERAGE
3.o REFLECTIVITY P" 0.Q (AVG CONC • 2.013)-~ TRANSMISSIONTHROUGH
2.8 ~_,~ THE GLASS ENVELOPE:
2.6 (AVG CONC l.~ll ~ 1.723fl.931 • 0.$q
2.4 CAVGCONC 1.723)~ \ \ , \ \
0 2.2-
\\
o~ 2.O1

~ 1.2
N Iol - o~
02
• 85 02 • 55" w2~.,~..
o,8 FLAP LENGTHS: WI = 33.5cm Wl . . Z'
06 W 2 • 28.5crn ~ ( ~
0.4 TUBE DIAMETER • i00rnm
0.2 ABSORBERWIDTH • 89 rnm
0 I l I l I [ l l i ~ [ 1 i I l I l I

,o ,o ,o ,~ .o

Fig. 4. Variation of daily average calculated concentration ratio.

4.2.2 Solution with retlectors. The same procedure as Therminol 44 is preheated by means of two electrical
in 4.2.1 has been used in order to calculate the values of immersion heaters and the inlet temperature is regulated
Tg, Ut., FR and consequently the values of Q, and ~7 by at the desired operation temperature. Both the tank and
using eqns (9) and (10), taking into consideration the the flexible piping connecting the pumping station to the
concentration ratio CR resulting from the optical model collector stand are insulated against heat losses.
calculations for different type reflectors. Collector Test Stand--Vacuum tubes are installed on a
Figure 5 gives results of the thermal model of the stand facing south tilted 35°C which is the latitude of
vacuum tube receiver with and without reflectors. The JPL, and instrumented for thermal performance evalua-
tube efficiency ~ is plotted against ATf.i, fluid inlet tem- tion.
perature minus the ambient temperature. The top set of Figures 7 and 8 show the test bed and the vee-trough
curves gives the efficiency of the receiver tube based on vacuum tube test stand.
the flux incident on the absorber plate. At most fluxes, up
to l l00W/m 2 is expected without a vee-trough concen-
trator. 5.2 Instrumentation
The purpose of the vee-trough concentrator is to in- The flow rate of Therminol 44 is measured by turbine-
crease the flux on the absorber to levels around type flow meters. Both the absolute fluid output tem-
2400 W/m2. perature and the differential temperature between the
The net efficiency of the vee-trough/vacuum tube col- fluid inlet and outlet are measured on four vacuum tubes.
lector, given as a function of the incident flux intensity Additional thermocouples are attached to the surface of
on the aperture plane, is based on the aperture area. each vacuum tube. These thermocouples are used as an
Therefore, it is lower than the receiver efficiency based indicator of the thermal insulation status (i.e. condition
on the absorber area. However, the cost of the collector of the vacuum and selective coating). Pressure drops
based on the aperture area is also low. As a result, the through the tubes (four tubes may be connected in series
cost per GJ (Btu) is low and results in a cost-effective or parallel by manipulating valves) and absolute pressure
design. are determined using pressure transducers. Tests were
run in the series configuration of receiver tubes.
S. DESIGN OF THE TEST BED AND INSTRUMENTATION
5.1 Test bed design
A test bed was designed and constructed for experi- 6. TESTING ANDEVALUATION
mental evaluation of the vee-trough collector consisting The test bed was first tested against any fluid leakage
of Coming Glass Works vacuum tube receivers and and other mechanical problems. The air entrapped in the
vee-trough reflectors. The test bed piping lay-out and manifold and lines was bled by opening the valve in the
instrumentation are shown in Fig. 6. air bleed line. After the flow meter reading was stabil-
The test bed consisted of the following components. ized, the valve was closed. The procedure was
Pumping Station--The working fluid, Therminol 44, occasionally repeated to eliminate dissolved air or gas
was circulated through the evacuated tubes by means of released due to decomposition of Tberminol 44 (modified
a gear type pump. The pumping station has features such Therpbenyl, a trademark of Monsanto Co.), which might
as a pressure relief valve, a bypass loop used to regulate have caused flow instabilities to invalidate the experi-
the flow, a drain line and an expansion tank. ments.
418 M. KUDRETSELCUK
90
I t
INCIDENT SOLAR FLUX
RECEIVER ON THE COLLECTOR PLANE
ONLY (BASED ON ABSORBER AREA)
BO

7O

6O
p-
Z
ILl
¢O
¢ INC

W
_u VE
'. 4o -VA

30 -

/
d
h
C
A
I0--
V ' 5 m/see ( l l . 2 m p h )
r n - 2 4 5 k g / h r m = (491b/hr f l =)

o I
de~; F 5O 100 15O ZOO 25O
I I f I I
de~l C 25 50 75 I00 125 150
(FLUID INLET TEMP-T,. e) AT; = Tf,;-To

Fig. 5. Results of the thermal model for the receiver and collector.

6.1 Calibration of instruments significant. These data were used in calculations either
The useful heat calculations and efficiency deter- using linear interpolation techniques or by curve fitting.
mination require the following basic data, which has to
be known within permissible error limits: (I) Mass flow 6.2 Performance of tests
rate of the working fluid n~ which consists of d (density) Tests were run mainly under clear day conditions for
and V (volumetric flow) rate terms; (2) Specific heat of daytime efficiency determinations and at night for heat
the working fluid Cp; (3) Temperature rise of the working loss experiments. The latter was conducted to determine
fluid in the evacuated tube AT: (4) Solar flux intensity at actual (UL) values for use in theoretical calculations of
the tilted collector plane/,. the collector efficiency.
Items (I), (2), and (4) were determined using calibrated 6.2.1 Daytime tests. Prior to test initiation, auxiliary
instrumentation. Specific heat of Therminol 44 was taken electric heaters were switched on and the fluid in the
from the manufacturers data. storage tank was heated to the temperature selected for
These figures were also verified by tests performed at the day which ranged from 95° to about 1600C. All
IPL's Chemistry Laboratory. Property change due to temperatures (absolute and differential), flow meter
slight coloring of Therminol after several runs, was not readings, pressures, solar flux intensity, wind speed, a•d
Analysis, development and testing of a fixed tilt solar collector 419
~A

AIR llLEEDLINE

I ANEMOMETER

HOT

KENDALLRADIOMETER
(NORMALINCIDENCE)
i \\
% I
I

,I
I
:~,v' : TL~
VACUUMTUllE
REFLECTORSFIXED

/AL,,,~® / J S PYRANOMETER
(TOTAL
INCIDENT
!
i

I
I
i ~;~</,,,®
RADIATION)
I ~'- ~ ,,~,,,~uu~,~o)
/
I
I
REFLECTOR"~ I

BY-PASSVALVE
I
I OVERFLOW
TC.~ FEEDERLINE

PRESSURERELIEFVALVE
] FLOWMETER
ELEC~,C /~
X(~C THERMOCOUPLE PREHEATER~ ""~-~
\ REFLECTIONSURFACES TEMPERATUREJ
\~TC DIFFERENTIALTHERMOCOUPLE CONTROLLER

ABSOLUTEPRESSURE
DIFFERENTIALPRESSURE

Fig. 6. Test arrangement for the vacuum tube receivers.

FIGURE 7
Fig. 8. Vacuum tube receivers with and without vee-trough
Fig. 7. Test bed and pumping station. reflectors.

SE VoL 22, No. 5--B


420 M. KUDRETSELCUK

wind direction were recorded on photodensitive paper at transfer coefficient UL between the working fluid and the
selected intervals, normally I0 rain. sky, heat losses have been measured without any heat
The temperature of the working fluid gradually rose gain during the night. UL has been obtained from
during the day since heat gain from the sun was more
than the loss through lines and tank insulation. This has UL = tnceAT
enabled the obtaining of test data around the set point.
A,,(T,.,-To) (12)
The present study does not examine a system incor- where AT is the temperature decrease of the working
porating a storage and a load. Instead, its purpose is to
fluid. Since the glass tube surface temperatures were
determine quasi-steady-state performance of the evacu- measured, the Uc value could be calculated by the
ated tube with and without the vee-trough concentrators.
equation
Therefore, it is quite satisfactory for a quasi-steady-state
evaluation to have input temperature varying at a rate of UL = Fcero(T,'- T / ) / ( T t . , - T~) (13)
14°C (27°F)/hr since the thermal capacity (response time
of a tube; i.e. time required for a temperature rise of and compared to the above value. The plate temperature
14°C) is only 1/3 of a minute. Furthermore, the average has been taken as the average fluid temperature
fluid transit time through a glass tube is 35 sec, which is
much less than the interval of observation. 1
6.2.2 Night tests. In order to evaluate the overall heat
T. = ~(Tt, + Tfo) (14)

Table 1. Comparisonof the results of the tests run on 6 July 1977at 11:50 with the calculated values

Tube No.
Parameter '1 2 3 4
Reflector ~ p e Glass Alzak EP Teflm None
TO Fluid outlet temperature °C 127.8 116.7 106. S 97.3

AT Fluid temperature increme °C 10.9 11.1 11.0 7.6

Tm Fluld average temperature oC 122.4 111.1 101.3 93.3

Mms Flow rote & - Vd kg/hr 28.4


L.,
Cp Spe©if;c heat (for Tm) KJ/kg °C 2.21 ( 2.18 2.16 2.13

Qu
Useful heat
Qu = mCpAT gJ/hr 684
JI 687 676 459

it Total solar Flux W/,"2 912

Ac Collection (aperture) area ,.2 0.56 0.65 0.645 0.19


~beorbed
Total solor input
Qin Qin = ItAc W 510 593 588 173
Overall collection efficiency Qu
based on total ~lar flux ~1= 0.378 0.325 0.320 0.74
and aperture area (measured)
i.
To Ambient temperature oC 31

AT; ATiEXCess
temperature=
T; - Tg °C 85"9 I 74"4 I 64"7 I 58"5

V Wind v e l o c i t y kz/hr 5.8


Calculated Performance (for the c o n d i t i o n s
Cc Manifold heat l o s s coeffJ- _ 1.03 1.03 1.025 1.025
c~ent
Overall heat t r a n s f e r
UL coefficient (calculated) W/m2Oc 2.09 2.04 1.87 1.64
at Ta ~ T~
TQ)e Total effective trans- 0.855
mLttanee

CR Concentration r a t i o * ~ 1.6 1.6 1.6 I

FR Heat r e ~ v a l factor 0.95 0.95 0.95 0.96

Qu Useful heat KJ/hr 693 703 715 448

Efficiency (calculated) - 0.39 0.33 0.34 0.72

"The concentmtian ratio given is a cl~|ved figure which accounts fas the surface
distortions, micro oncl macro irregular|ties of the reflector surface, and dust
and dirt effects on the mirrm$ and glass tube.
Analysis, development and testing of a fixed tilt solar collector 421

The average plate temperature differs from the average ?. OIq~ffZATIONSTUDI~


fluid temperature due to the thermal resistance between 7.I Optimization of the vee-trough design for maximum
the fluid and plate. Since the tubes are spaced only 2 in. thermal energy collection
apart and are bonded to the copper absorber plate by Both the concentrated solar flux intensity and the
electron beam welding, the temperature difference of T evacuated tube efficiency must be maximized for the best
plate from T fluid is less than 3 per cent. thermal energy collection for year-round operation.
Major factors influencing the vee-trough and evacu-
6.3 Evaluation of test data ated receiver performance are discussed below.
Test data obtained were used for evaluation of the Optimization of the ~lap angles. The aperture angle
hourly useful heat and the efficiency of the collectors and flap angles 01 and 02 may be varied to obtain the
with or without the vee-trough reflectors. A typical combination which yields the maximum year-round
evaluation of the test data is given in Table 1. Additional averaged concentration ratio for the case in which there
results are in Ref. [23]. is a demand for both heating and cooling. Figure 10
Table 1 also gives the theoretical efficiency calculation indicates that either good summer and good winter per-
for four tubes tested under the conditions on 6 July, formance or good year-round performance can be
1977. Test data and theoretical predictions compare well. obtained by choosing 01, 02 and ~ properly.
Results of various tests run during July 1977 are Figure 11 gives the results of these runs in which the
summarized in Fig. 9. Measured efficiencies of the bare concentration ratio is plotted vs 01, ,1 being a
tube receiver and receivers with various reflectors are parameter. ? = 40° was selected for the test hod design.
plotted vs AT~ = (T~, - To). Although the concentration ratio improves slightly for

CORNING 788 W/m2 (250 Btu/l~r It 2)


GLASS
WORKS 951 W/m2 (300 Btu/~r if2)
DATA
1085 W/m2 (350 Btu/l~r ff~

7o

VACUUM TUBEWITHOUT
J CONCENTRATORSCAL-
60 C~I CULATEDPERFORMANCE
i -- 9o8 W / m 2 , = C

VEE TROUGH
VACUUM TUBE
COLLECTOR

JPL TESTS
"
20--
870 < I < 912 W/m2 I-
• GLASS1rEFLECTOR (UL= 1.98 W/m2C F 100 C)
• ALZAK (UL- 1.71 W/m2C FOR TF1= 100 C)
10 -- • FEPTEFLON (UL= 2.,55 W/m2C FOR-TFI = 100 C)
• NO REFLECTOR (UL= 1.58 W/m2C FOR Tf1= 100 C
I I I At,. tt. - t.,b
050 lOO 150 2(]0 250 300 350 4OO
deg F
I I I I I I I I I
60 80 100 120 140 160 180 200 220
deg C
Fig. 9. Test data for collectorefficiencyvs temperature.
422 M. KUDRETSELCUK
3.0 I I I I [ I I I
OPTICAL CONCENTRATION FACTOR
GOOD PERFORMANCE DURING 35 r ~ ~
2.5 m WINTER AND SUMMER ONLY (6o/85) -
r ~, -

~ .o
"rt, .30°-\@, //% E

< ,/,, a0 ° /// \\


u. T/== 3 0 ° ~ / ",,,,~
~ i .5 1 7 / . 30°
p-
," GOOD YEAR
g T/g30° ROUND PERFORMANCE
,,Z,1.0 _~-'NORMAL ~ REVERSED NORMAL :_...
01 02 - [ - 01 02 fl
Z ° ° ~o _._~o
O 60° 90°
u0.5 65° 85° 850 650 65° 85° ----30 °
- - 70 ° 80 ° , 80° 70° i 70° 8o~ -.-3o~
60° 85° 85° 70° EQUI 60o 85° ~ 35v
I I E~u~I I I Ii I AIs Ioi I I I
DEC 21 J F M A M J J N D

Fig. 10. Variationof the daily averaged concentrationratio with flap side tilt and aperture angles.

~4 Design and material properties o[ the evacuated tube


3.3 receiver. Performance of the evacuated tube receiver can
3.2 be improved by:
w,
3.1 (1) Improving the transmissivity of the glass envelope
3.0 which can thus be increased up to 98 per cent, using new
29 techniques of antireflective coatings.
W ~ T MIRROR REFLECTIVITYp=O.9
2,0 (2) Improving the absorptivity of the selective coating.
O
217 UDE An absorptivity value improvement to 94.4 per cent is
tY 2.6 ,r,~ " A.NGLE
expected for black chrome coatings.
2.5
LEN.G OF ''PS I (3) Improving the emissivity of the absorber surface.
2.4 W, 27.8 cm [ Refinements to lower the emissivity e to about 0.066 are
n.,
I - 2,3 Wz-24.1 cm I expected.
z
2,2 A: RECEIVER Optimization o[ the collector plane tilt. All previous
OPENING I0.1 cm
2.1 oz_~9o. analyses and test data refer to a collector tilted to the
KEY~
20 latitude. If 6, the collector plane tilt, is changed, then the
L9 ~.o -n
34"- •
daily average concentration ratio throughout a year is
1.8 36" - 0 affected. If the tilt is more than the latitude, for example,
1.71 38"-0
40"-0
40°, then the winter performance is better than the per-
1.6 4Z~- I I formance of a vee-trough tilted to the latitude (34.10° in
1.5 ~5 I I I this example). Similarly if the tilt is less than the latitude,
5O 55 60 65 ?0
O~wdegrees for example 30°, then the summer performance is better
Fig. I I. The effect of vee-trough aperture angle tilt on the than the performance of the collector tilted to the lati-
average year-round concentration ratio. tude, 34.10°. This feature of the vee-trough would be
very useful for those applications requiring winter heat-
< 40°, there are sharp peaks and valleys on the curves ing or summer cooling only. In such instances the col-
of concentration ratio as seen in Fig. 10. lector tilt may be about 6 - ( 1 0 to 150) for summer
Optimization o[ the [lap widths. Increasing the flap operation and 6 + (10 to 15°) for winter operation, respec-
length for a fixed aperture angle would increase the tively. Its exact value must be determined by a simulation
geometric concentration ratio almost linearly. The actual model which would consider the climatic variables, load
year-round concentration ratio increases at a much and storage relations for the system studies.
slower rate. An optical concentration ratio increase
beyond 3 is not usually justified, since the total mirror 7.2 Optimization of the use[ul heat collection
length almost doubles for an actual concentration ratio Since the thermal output (net useful heat) of the vee-
change of 5 per cent. trough collector is dependent upon the operation tem-
Re[lectivity optimization. The reflectivity of the mirror perature, various sets of operating conditions and vee-
surfaces has a significant effect on the concentration trough configurations were tested using the thermal
ratio without a glass envelope. The choice of the model of the vee-trough/evacuated tube collector. Hour-
reflecting surface must be made on a cost-effective basis by-hour radiation and ambient temperature data for
rather than on the basis of the highest concentration Burbank, California, during 1962 was utilized. Computer-
ratio. plotted curves of day long efficiencies at 65° and 175°C
Analysis, development and testing of a fixed tilt solar collector 423

ioo
YEARLY AVERAGE EFFICIENCIES YEARLY AVERAGE EFFICIENCES
90 p,0.94 (SILVERED TEFLON) AT 65°C ¢/-0.40
rl AT 175°C ~- 0.237
"O'944'~,iMPROVED BLACK CHROME
,0.066J MIRROR REFLECTIVITY p-0.8 (ALZAC)
AT 6.5~ Q-0.48 VEE-TROUGH ~ =55°
AT 175% ~ ,0.367 02,~ o
t.- 70 W, =0.3m
z W2=0.26 m
W
(.3 a=0.93
• 0.09

o7 50
W

~ii 4o
w
3o

20

I0

0
0
DAYS

Fig. 12. Variation of daily average efliciencies at operation temperatures of 65° and 175°C.

are presented in Fig. 12 for one year. Yearly averages are The annual heat collection QT in kJIm2 is obtained by
also summarized for two design conditions. One of the integrating the daily useful heat given in Fig. 12. Other
sets (p = 0.8, etc.) describes the performance of the assumptions in generating Fig. 13 were:
vee-trough collector as designed and tested at JPL. The
latter (p = 0.94 employing silvered Teflon reflectors, etc.) Atube = 0.3 m21m2 aperture,
gives the output of an advanced idealized collector which ARch = 2.0 m2/m2 aperture, and
can be designed soon. Cost of frame + assembly = 20 $1mz

8. DISCUSSIONAND CONCLUSIONS
7.3 Optimization of the vee-trough collector /or the The vee-troughlvacuum tube collector proposed can
lowest energy cost be used in solar heating and cooling applications, or,
It usually turns out that neither maximum heat collec- owing to its high-temperature capabilities 150--200~C, it
tion nor the lowest collector cost figures are the answers could also be utilized for power generation purposes in
to the cost-effective design. combination with an organic Rankine conversion system.
An attempt was made to predict the cost of the It is especially recommended for those unattended
evacuated tube and vee-trough reflectors. Details of the pumping stations since the reflectors only require rever-
evacuated tube receiver and collector module cost are sal once every 6 months.
given in Ref. [24]. Cost estimates are summarized in Test results reported represent the performance of the
Table 2 for glass tube diameters ranging from 51 to veeqrougb/vacuum tube collector combination based on
178 mm and a reflector cost of 5.38 $1m' of the reflector the aperture area. The data are defined for the total
surface area. The module suggested measures approx. incident flux on the collector plane tilted 35° to the south.
1.9x5.1 for small tubes, or 2.1x5.1m for 176ram All instruments used for the measurement of tem-
diameter tubes. From the data in Table 2, the optimum perature, flow rate and solar radiation were calibrated.
(minimum cos0 appears to correspond to 127-152mm. Differential thermocouple readings were accurate to _+
Since cost data are not yet commercially verified, a 0.1°C; the absolute temperatures were to +-0.4°(:.
parametric study of the energy cost has been made. The Volumetric flow of the working fluid, Therminol 44, was
energy cost has been predicted for a set of receiver tube measured within_+3 per cent. Total solar radiation
and reflector costs and is presented in graphical form in measurements were-made using a Spectran precision
Fig. 13 for easy visualization. Present cost estimates pyranometer (_+ I per cent). Combination accuracy
yield energy costs of about $5.2/GJ ($5.5/MBtu) and measurements were within _+6 per cent.
$5.97/GJ ($6.3/MBtu) at 65 and 121°C, respectively. Efliciencies shown in Fig. 9 and those tabulated in
The energy cost is calculated from Table 1 are the values obtained from the test bed without
any corrections due to differences in UL values of tubes.
C = 0.196 CJQr. (25) As will be noted, UL varies from 1.58W/m~C to
2.55 W/m2°C at 100°C for ~he tubes tested. A fair com-
The collector module cost is parison of reflector surfaces requires using vacuum tubes
having the same UL values. Since such tubes were not
Cc = Ctube X Atube + CRESS× Aframe+assembly. available, actual tests could not be run. However, it is
424 M. KUDRET SELCUK

Table 2. Summary of the evacuated tube receiver and collector module design features and cost estimates

Unit Tube Dlm~eter

.~ 50 76 100 127 152 I 178


q

Width x Length m 1.82x5.18 1.82x5.1E 1.82x5.18 1.9x5.18 1.82x5.18 2.14x5.18


o f Module
Aperture Area m2 7.8 7.8 7.8 8.12 7.8 9.1

Tube Cost $ 5.9 11.2 20.1 24.1 30.9 47.1


(2.49 m long
each)
Tube Cost With $ 22.3 28.4 40.2 46 55.7 74.8
Absorber and
Seals
Number of 24 16 12 i0 8 8
Tubes Per
Module
Total Tube $ 535 454 482 460 445 598
Cost

Tube and $ 1279 950 855 770 693 846


Frames Cost
Vee-Trough $ 189.5 157.5 142 140 131.5 156
Cost*
5.38 $/=2
Reflector

Total Module $ 1468.5 1107.5 997 910 824.5 1002


Cost

C o l l e c t o r Cost ;/m2 744 511 438 373 346 356


Per U n i t Ab-
sorber Area

Collector Cost $/m 2 188.3 142 126 112 " 105 ii0
Per Unit
Aperture Area

*Vee-trough c o n c e n t r a t o r cost i s based on a r e f l e c t o r (0.5 mm e l e c t r o p o l i s h e d a l u -


minum sheet area o f 2 =2/m2 a p e r t u r e plus 1 ~ t h i c k o r d i n a r y aluminum bottom
sheet r t v t t e d t o r e f l e c t o r sheet to form the A s e c t i o n .

8 /I 1 1 l I I I/ I '
/ VEE-THROUGH VACUUM t /
7
TUBE COLLECTOR ~ / -
.=7
O0
6

_3' ~' ~ PLATE COLLJ~'~,~' IRFL nOB/m"


5

/ ~ ~_~'~ TUBE AT?50 */m' 4 "-)


~"
>- u J : ~// :~'/ 9 / NEAR TERM COST -
/~" ' / IRn. ~*/rn
bJ
//' ,IRFL 55/m z ,
3

.J TUBE ;T ,~0 ,)m ~ 2


<:E 2 (ATTAINABLE)
re
i,I
RFL IO$/m z / T U B E AT IOO $/rn z-
RFL 5$/rn z | F U T U R E COST
I I I I I I I I
o 50 ~O 150 200 z5o 3OO 350 400 450
COLLECTOR COST $/m z ABSORBER AREA

Fig. 13. Thermal energy cost vs collector cost based on the absorber area.
Analysis, development and testing of a fixed tilt solar collector 425

obvious from the test data that the aluminized FEP NOMENCLATURE

Teflon reflector output would have been improved if UL A area, m2


collector area, m2
were i.71 instead of 2.55 Wlm~C at 100°C over the A~
absorber plate area, m:
Alzak reflectors. Similarly, the tube with a glass reflector A~C cost of collector, dollars
would yield a higher efficiency if UL were 1.71 instead of c energy cost, dollars
1.98 W/m~C at 100°C. CR flux concentration ratio, dimensionless
Results of the thermal performance analysis given in C~ specific heat, kJ/kg°C
d density, kg/m3
Fig. 5 were based on a mathematical model without d, width of the image of the strip, m
reflector end effects and without losses due to copper FR heat removal factor, dimensionless
tube axial conduction and supporting clip conduc- F~ correction factor for manifold losses, dimensionless
tion. Is beam radiation intensity, W/m:
diffuse radiation intensity, W/m2
The pressure drop through the collector module was IR reflected radiation intensity, W/mz
on the order of 0.02 MPa (3 psi) for 4 tubes in series. I, total radiation intensity, Wirez
The suggested module design may use at most 6 tubes in i angle of incidence, deg.
series for 100mm tubes or 4 tubes in series for 151 mm K glass extinction coefficient,Ilcm
tubes. In both cases, the pressure drop is within reason- k thermal conductivity, Wlm°C
L length, m
able limits and the pumping power is small. L~ length of the image, m
Energy cost predictions are given in terms of the mass flow rate, kg/hr
collector cost based on the unit absorber area. Such a N day number, dimensionless
presentation enables one to visualize the effect of corn- n index of refraction for glass, dimensionless
ponent costs. Since there are not firm cost figures avail- Qt
Q,. total heat collected, Lllhr
incident heat, kJ/hr
able for the evacuated tube receiver, a definite energy Q, useful heat, Id/hr
cost cannot be quoted. A cost estimate was, however, Qt lost heat, kJ/hr
presented using suggested net prices for off-the-shelf Q, stored heat, kJIhr
pyrex tube to give an idea about the near future LT temperature, °C
ambient temperature, °C
costs. T~.~ fluid inlet temperature, °C
It is predicted that long-range mass production costs AT temperature difference, °C
could be as low as $150/m2 for the evacuated tube AT/.~ fluid inlet temperature--ambient temperature, °C
receiver and $5/m2 for the reflector. UL heat loss coefficient,Wlm2°C
The vee-trough/vacuum tube collector competer with Wv wind speed, m/sec
flap side width, m
conventional flat plate collectors costing $80/m 2 and x. Y,Z coordinates of flap corner points and their
operating at 120°C. At 65°C, the veo-trough collector has reflections, m
to be assembled with tubes which cost less than $150/m 2
and reflectors costing less than $51m2 in order to be Greek Symbols
a absorptivity, dimensionless
competitive.
6 declination, deg.
The merit of the collector concept is in combining the emissivity, dimensionless
relatively expensive vacuum tube with an inexpensive "~ vee trough aperture angle, deg.
concentrator, which enhances the tube performance by efficiency,dimensionless
increasing the incident flux and reducing its cost due to 0 flap side tilt, deg.
,~ wavelength, ~tm
the low cost feature of the vee-trough concentrator. The p reflectivity,dimensionless
present study is considered a confirmation of the mag- (~-a), effective transmissivity absorptivity product,
nitude of the efficiency predictions with the test data for dimensionless
quasi-steady state conditions. Further tests and analyses, transmissivity, dimensionless
especially simulations of systems with storage and vari- ~(~) transmissivity of the glass envelope, dimensionless
is average transmissivity of the glass envelope to the
able load features, are needed for more accurate predic- beam radiation not reflected from either mirror,
tions. dimensionless
average transmissivity of the glass envelope to the
diffuse radiation not reflected from either mirror,
dimensionless
Acknowledgements--This paper presents the results of one phase oo Stephan-Boltzmann constant
of research carried out at the Jet Propulsion Laboratory, Call- d~ latitude, deg.
fornia Institute of Technology, for the Department of
Energy/Office of Conservation & Solar Applications, Solar Subscripts
Heating and Cooling R & D Branch, under lnteragency a ambient
Agreement No. E(49--26)--1024by agreement with the National b bottom
Aeronautics and Space Administration(Contract NAST-100). B beam
Assistance given by several individuals from JPL, especially c collector
Mr. Eugene W. Noller, Mr. James A. Bryant, Mr. Donald C. d diffuse
Schneider, Dr. Burton Zeldin, Mr. Nabil El Galabawi and Mr. f fluid
Edward (Larry) Noon, is gratefully acknowledged, g glass
Contributions of Dr. Kemal Onat, Professor of the Mechanical i inlet
Engineering Faculty of istanbul Technical University, Turkey i index for the strip analysis only (eqn 4)
and Dr. U~ur Ortabael and Dr. Francis P. Fehlner of Coming m first strip which reflects on the absorber plate, Ist
Glass Works (CGW) are highly appreciated. flap (eqn 4)
426 M. KUDRETSELCUK

n first strip which reflects on the absorber plate, 2nd II. P. A. Kittle and S. L. Cope, Outside performance of
flap (eqn 4) moderate vacuum solar collectors. 1SES Conf.. Paper No.
k last strip which reflects on the absorber plate, Ist 32/8, Los Angeles (1975).
flap (eqn 4) 12. E. Speyer, Solar energy collection with evacuted tubes.
I last strip which reflects on the absorber plate, 2nd Trans. ASME (Power), pp. 270.-276 (1965).
flap (eqn 4) 13. F. F. Simon. Solar Collector Performance Evaluation with
o outlet the NASA Lewis Simulator--Results for an Overall Glass
p plate Evacuated Tubular Selectivity Coated Collector with a
pb plate bottom Diffuse Reflector. NASA TM-X-71695, Lewis Research
Center, Cleveland, Ohio, Apr. 1975.
~CF_.S 14. U. Ortaba,~l and W. Buehl, Analysis and performance of an
I. M. K. Sel~;uk, A fixed collector employing reversible vee- evacuated tubular collector. Paper presented at ISES Conf.
trough concentrator and a vacuum tube for high temperature at Los Angeles, California (1976).
solar energy systems. Proc. ilth lntersoc, Energy Cony. 15. U. Ortaba§l, Indoor test methods to determine the effect of
Engng Conf., State Line, Nevada, Paper No. 769222 (1976). vacuum on the performance of a tubular flat plate collector.
2. H. Tabor, Mirror boosters for solar collectors. Solar Energy ASME Paper No. 76.--WA/SOL-24.
10 (3), I I I-118 (1966). 16. S. Karaki and D. M. Frick, Performance of an evacuated
3. Coming Tubular Collector, Coming Glass Works, New York, tube solar collector. Paper presented at ISES Con/., Win-
(23 Jan. 1975). nipeg, Manitoba, Canada (1976).
4. D. C. Beekley and G. R. Mather, Jr., Analysis and experi- 17. W. Schertz, Argonne National Laboratories, private com-
mental tests of high performance tubular solar collectors. munication.
ISES Int. Conf., Paper No. 32110, Los Angeles, California 18. A. Rabl. Collectors with cusplike compound parabolic con-
(1975). centrator and selective absorber. Proc. ISES Vo[. 2, pp.
5. K. G. T. Hollands, A concentrator for thin film solar cells. 327-350 (1976).
Solar Energy, 13, 149 (1971). 19. G. Thodos, Predicted heat transfer performance of an
6. H. Durand et aL, Periodically adjustable concentrators adap- evacuated glass-jacketed CPC receiver, countercurrent flow
ted to solar cell panels. COMPLES Int. Meeting, University design. Argonne National Laboratories, Report No. ANL-
of Petroleum and Minerals, Dhahran, Saudi Arabia (1975). 76--67, Argonne, Illinois (1976).
7. R. B. Bannerot and J. R. Howell, Moderately concentrating 20. B. Gupta, Honeywell Inc., private communications (1977).
flat plate solar collectors, ASME Paper No. 75 HT-54. 21. W. F. Moore, General Electric Company, Valley Force
8. R. Winston and H. Hinterberger, Principles of cylindrical Space Center, private communication (1977).
concentrators for solar energy. Solar Energy 17 (4), 255-258 22. J. A. Duffie and W. A. Beckman, Solar Energy Thermal
(1975). Processes. Wiley-Interscience, New York (1974).
9. K. G. T. Hollands, Honeycomb devices in fiat plate solar 23. M. K. Selcuk, Final Report, A Fixed Tilt Solar Collector
collectors. Solar Energy 9 O), 159-164 (1965). Employing Reversible Vee-Trough Receivers for Solar
10. H. A. Blum et a/., Design and feasibility of flat plate solar Heating and Cooling Systems. JPL Publication No. 77-78
collectors to operate at 10(Y'-I50°C. Proc. UNESCO Conf., (DOE/JPL-1024--77/I), December 1977.
Sun in the Service of Mankind, Paris, Paper No. E 118 (1973). 24. U. Ortaba~l, Coming Glass works, private communication.

You might also like