9

AlllDD TTbEn
NATL INST OF STANDARDS & TECH R.I.C.

A1 11 0099621
Hill, James Edward/Experimental verlfica
TA435 .U58 V117;1979 C.I NBS-PUB-C 1979

i|35

.U58
NO. 117
1979
C.2

NBS BUILDING SCIENCE SERIES 117

Experimental Verification of a
Standard Test Procedure
for Solar Collectors

U.S. DEPARTMENT OF COMMERCE • NATIONAL BUREAU OF STANDARDS * »•

rrrTTJT.T. TT tttt 1 m~ v I
 NATIONAL BUREAU OF STANDARDS

The National Bureau of Standards' was established by an act of Congress March 3, 1901. The
Bureau's overall goal is and advance the Nation's science and technology and
to strengthen
facilitate their effective application for public benefit. To this end, the Bureau conducts
research and provides: (1) a basis for the Nation's physical measurement system, (2) scientific
and technological services for industry and government, (3) a technical basis for equity in
trade, and (4) technical services to promote public safety. The Bureau's technical work is
performed by the National Measurement Laboratory, the National Engineering Laboratory,
and the Institute for Computer Sciences and Technology.

THE NATIONAL MEASUREMENT LABORATORY provides the national system of

physical and chemical and materials measurement; coordinates the system with measurement
systems of other nations and furnishes essential services leading to accurate and uniform
physical and chemical measurement throughout the Nation's scientific community, industry,
and commerce; conducts materials research leading to improved methods of measurement,
standards, and data on the properties of materials needed by industry, commerce, educational
institutions, and Government; provides advisory and research services to other Government
Agencies; develops, produces, and distributes Standard Reference Materials; and provides
calibration services. The Laboratory consists of the following centers:

Absolute Physical Quantities' — Radiation Research — Thermodynamics and

Molecular Science — Analytical Chemistry — Materials Science.
THE NATIONAL ENGINEERING LABORATORY provides technology and technical
services to users in the public and private sectors to address national needs and to solve
national problems in the public interest; conducts research in engineering and applied science
in support of objectives in these and maintains competence in the necessary
efforts; builds

and technical service; develops engineering data

disciplines required to carry out this research
and measurement capabilities; provides engineering measurement traceability services;
develops test methods and proposes engineering standards and code changes; develops and
proposes new engineering practices; and develops and improves mechanisms to transfer
results of its research to the utlimate user. The Laboratory consists of the following centers:

Applied Mathematics — Electronics andElectrical Engineering' — Mechanical

Engineering and Process Technology' — Building Technology — Fire Research —
Consumer Product Technology — Field Methods.

THE INSTITUTE FOR COMPUTER SCIENCES AND TECHNOLOGY conducts

research and provides scientific and technical services to aid Federal Agencies in the selection,
acquisition, application, and use of computer technology to improve effectiveness and
economy in Government operations in accordance with Public Law 89-306 (40 U.S.C. 759),
relevant Executive Orders, and other directives; carries out this mission by managing the
Federal Information Processing Standards Program, developing Federal ADP standards
guidelines, and managing Federal participation in ADP voluntary standardization activities;
provides scientific and technological advisory services and assistance to Federal Agencies; and
provides the technical foundation for computer-related policies of the Federal Government.
The Institute consists of the following divisions:

Systems and Software — Computer Systems Engineering — Information Technology.

'Headquarters and Laboratories at Gaithersburg, Maryland, unless otherwise noted;

mailing address Washington, D.C. 20234.
'Some divisions within the center are located at Boulder, Colorado, 80303.

The National Bureau of Standards was reorganized, effective April 9, 1978.

 JAN Z W3
NBS BUILDING SCIENCE SERIES
1^1
117

Experimental Verification of a Standard Test Procedure

for Solar Collectors

James E. Hill
John P. Jenkins and
Dennis E. Jones

Center for Building Technology

National Engineering Laboratory
National Bureau of Standards
Washington, D.C. 20234

Prepared for
The Department of Energy
Research and Development Branch
for Solar Heating and Cooling
Office of the Assistant Secretary for
Conservation and Solar Applications
Washington, D.C. 20545

U.S. DEPARTMENT OF COMMERCE, Juanita M. Kreps, Secretary

Jordan J. Baruch, Assistant Secretary for Science and Technology

NATIONAL BUREAU OF STANDARDS, Ernest Ambler, Director

Issued January 1979

 Library of Congress Catalog Card Number: 78-600138

National Bureau of Standards Building Science Series 117

Nat. Bur. Stand. (U.S.), Bldg. Sci. Ser. 117, 127 pages (Jan. 1979)
CODEN: BSSNBV

U.S. GOVERNMENT PRINTING OFFICE

WASHINGTON: 1978

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
Stock No. 003-003-02008-7 Price $3.00
(Add 25 percent additional for other than U.S. mailing).
 CONTENTS

Page

Abstract iv

1. Introduction 1

2. ASHRAE Standard 93-77 3

3. Description of NBS Test Facility 11

4. Test Results 25

5. Summary and Conclusions 41

6. Recommendations for Future Work 45

7. References 47

Tables 51

Appendix A - Calculation of Air Flow Rate Using a Nozzle

and Pressure Difference Measurements 61

Appendix B - Description of a Radiometer to Determine Sky

Temperature 64

Figures 67

iii
 Experimental Verification of a Standard Test Procedure
for Solar Collectors

by

James E, Hill, John P. Jenkins and Dennis E. Jones

Abstract

A proposed procedure for testing and rating solar collectors based

on thermal performance was published by the National Bureau of Standards
(NBS) in 1974, Subsequently, the American Society of Heating, Refrig-
erating, and Air Conditioning (ASHRAE) developed a modified version of
the NBS procedure which was adopted in early 1977 as ASHRAE Standard
93-77. A test facility for water-heating and air-heating collectors has
been built at NBS and was used to support the development of Standard
93-77, The purpose of this report is to describe the recently adopted
test procedure, the NBS test facility, and the tests that were conducted
to support the development of the procedure.

Keywords: Measurement; solar collector; solar energy; solar radiation;

standards; standard test; testing.

In discussing this experimental program, certain commercial components

were used and are identified in order to provide a descriptive characteri-
zation of their features. Inclusion of a given component is this report
in no case implies a recommendation or endorsement by the National Bureau
of Standards, and the presentation should not be construed as a certifica-
tion that any component would provide the indicated performance. Similarly,
the omission of a component does not imply that its capabilities are less
than those of the included components. This report is intended to be
informative and instructive and not an evaluation of any commercially
available components.

iv
1. INTRODUCTION

A proposal for testing and rating solar collectors based on thermal per-
formance was published by the National Bureau of Standards (NBS) in 1974
[1-3]," The procedure prescribed that a series of outdoor steady-state
tests be conducted to determine the near-normal-incidence efficiency of
the collector over a range of temperature conditions. The American
Society of Heating, Refrigerating, and Air Conditioning Engineers subse-
quently developed a modified version of the NBS procedure which was
adopted in February, 1977, as ASHRAE Standard 93-77 [4]. It is similar
to the NBS procedure but calls for additional tests to determine the
collector time constant as well as an incident angle correction factor
that can be applied to the near-normal-incidence efficiency to determine
collector performance both early in the morning and late in the day.

Three test loops have been built at NBS in accordance with ASHRAE Stan-
dard 93-77, two for modular water-heating collectors and the other for

Figures in brackets indicate literature references at end of the report,

1
air heaters. A major part of the NBS collector testing program since
mid-197 6 has been devoted to using these facilities to support the devel-
opment of Standard 93-77, The purpose of this report is to describe
the recently adopted test procedure, the NBS test loops, and the tests
that were conducted during 1976 and 1977 to support the adoption of
the procedure.

A second major part of the NBS collector testing work has been to con-
duct a round robin program in which two flat-plate liquid-heating
collectors were tested by 21 organizations around the United States
during 1976 and early 1977. The purpose of the program was to have a
a variety of testing laboratories attempt to utilize the NBS test pro-
cedure* and then determine the extent to which the results differed
or were comparable. The test data and subsequent analysis for this
program are available in separate publications [5, 6],

*ASHRAE Standard 93-77 was not adopted at the time this program was
initiated.
2
 93

BASBDO^^^'>'^'^OE DEVICES

2. ASHRAE STANDARD 93-77

The testing procedure recommended by NBS in reference [1-3] called for

the determination of the thermal efficiency of the solar collector by
passing the heat transfer fluid through it at a steady rate with the
collector mounted outdoors under clear sunny conditions. Measurements
were required to be made during the middle of the day and consisted
primarily of determining the fluid flow rate, temperature rise in the
fluid as it passed through the collector, and the incident solar radia-
tion (also called irradiance). These data could then be used to
compute collector efficiency. The tests were to be made so that at
least 16 "steady-state" efficiency values could be determined over a
range of temperature differences between collector fluid and ambient
air in order to draw an efficiency curve for the collector. A
significant amount of detail was included on the recommended
apparatuses, the instrumentation, and the procedure to follow. The
testing procedure specified in ASHRAE Standard 93-77 consists of the

3
same series of efficiency tests (with some modifications to be noted
below) plus additional tests which allow one to determine the transient
response of the collector as well as how efficiency changes with
increasing incident angle (between the direct solar beam and outward
drawn normal to the plane of the collector aperture).

The major changes in the conduct of the efficiency tests as specified in

Standard 93-77 compared to that specified in [1, 2] are as follows:

1, The testing apparatus for water-heating collectors has been modified

to include a storage tank for damping out thermal transients and a
bypass to permit periodic calibration of the flow meter in place.
Figure 1 is a schematic drawing of the recommended closed-loop from
[4], In addition, two other alternate open-loop configurations are
also given (see Figure 2 and 3) and considered acceptable provided
the test requirements and specifications can be met,

2, The testing apparatus for air heaters has been rearranged so that
air is "pulled" through the collector instead of being blown through
it (slight negative gauge pressure in the collector). Alternately,
an open-loop test configuration similar in concept to Figure 3 can
be used,

3, More stringent requirements have been included for the measurement

of incident solar radiation. Pyranometers are required to meet or
exceed the characteristics of a first class pyranometer as classi-
fied by the World Meterological Organization [7],

4, In conducting the test, data must be taken when the solar incident
angle is less than 30° (compared to 45° in [1, 2]),

5, The time period required for the integration of energy quantities

to compute one efficiency value has been decreased from 15 minutes
to either 5 minutes or one time constant, whichever is larger,

6, In computing efficiency, the gross frontal area of the collector

is used instead of aperture area.

7, The efficiency curve is drawn by plotting efficiency as a function

of the difference between inlet fluid temperature and ambient tem-
perature divided by the incident solar radiation. (Average fluid
temperature was used in [1, 2].) Inlet fluid temperature was chosen
to be used in the plot because the primary characteristics of the
collector required for the system design procedures of [8, 9] can be
determined directly from the slope and the intercept of the curve.

The major completely new features of Standard 93-77 compared to [1, 2]

are:

1. The collector is required to undergo a preconditioning test prior

to the start of the thermal tests. The collector must be exposed

4
 for three cumulative days with no fluid passing through it and
with the mean incident solar radiation measured in the plane of
the collector aperture exceeding 17,000 k.J/(m day) (1500
*

Btu /(ft^ • day)).

2. Prior to conducting the efficiency tests, the time constant is

determined (see below).

3. After completing the efficiency tests, a series of tests is con-

ducted to determine the collector's incident angle modifier (see
below)

4. The entire group of tests may be made Indoors using a solar simulator
if desired. The specifications for the simulator are included and
follow closely those of references [10-12],

2.1 COLLECTOR TIME CONSTANT

When a solar collector is operating under steady-state conditions, the

relationship that governs its thermal performance is:

the rate the rate the rate of the rate

of useful of energy energy loss of energy
energy = absorbed from the carried
extracted in the collector by away by the
from the collector conduction, transfer
collector by the convection fluid
absorber and radiation
In equation form:

mc.
^ = FRl(Ta)e - FRUL(tf,i-ta) = ^(tf,e " tf,i) (1)

where

*^u = rate of useful energy extracted from the collector, W

A = cross-sectional area of the collector, m 2

collector heat removal factor

I* = total solar energy incident upon the plane of the collector per unit
time per unit area, W/m

( ta)^ = effective transmittance-absorptance product for the collector

Uj^ = heat transfer loss coefficient for the collector, W/ (m 2 .

•c)

- temperature of the transfer fluid entering the collector, **C

I

*In the technical literature, frequently the symbol E, irradiance or G,

global irradiance will be used instead of I [38].
t„ = ambient air temperature, °C

m = mass flow rate of the transfer fluid through the collector, kg/(s»m^)

Cp = specific heat of the transfer fluid, J/(kg»s)

tc = temperature of the transfer fluid leaving the collector, °C

Whenever transient conditions exist, the above equations are not valid
since part of the energy being absorbed is used for heating-up of the
collector and its components. The corresponding relationship for tran-
sient conditions is:

the rate the rate the rate the rate

of change of energy of energy of energy
of energy absorbed loss from carried
stored in in the the collector away by
the collector collector by conduction, the
and its by the convection transfer
compo nents absorber and radiation fluid

In equation form [13,14] :

C
-A_J.
A
cl

de
= FI(Ta)
R e
- F U (t -t )
R L f,,i a
- ^
*

A
(t
f,e
-t
f,i
)
^2)

where

= effective heat capacity of the collector, its components, and the

transfer fluid in the collector, J/°C

6 = time, s

Equation (2) can be solved for the exit temperature of the transfer fluid,
tf eJ.
as a function of time, 9, after making the following assumptions:
,

The exit temperature, of the transfer fluid is related to the

average fluid temperature, t^, by

^ de
= K^hjL±
de
(3)

6
where [ 13]

K=!;P_ (II -1) (4)

and F' = collector efficiency factor

2. I, ( Ta)g, Uj^, t^, m, Cp, and t^ ^ are all constant for the period
covered by the transient solution.

The solution to equation (2) is then:

mc
F^Kra)^ - Wtf,i - ^a)
S ('f,e " ^f,i)
= e
"
^ KC^^
®
(5)

F^Kxa)^ - W^f.i - ^a) (tf,e, initiart ,i>

A

It is conventional practice in the physical sciences that whenever an

equation of the form of (5) describes the response of a particular
system, the quantity KC^/mCp is called the system's time constant.
The physical interpretation of the time constant is the time required
for the quantity on the left side of equation (5) to change from 1.0
to 0.368 where 0,368 = 1/e, It was felt that this concept should be
maintained in order to be consistent with the accepted interpretation
of time constant. Unfortunately, this has not always been done in
the solar collector field, Simon [15] has published the value of "time
constant" for nine different water-heating collectors using the 0 to
99% change time, Wijeysundera [16] has computed the "response time"
of typical one, two, and three cover air-heaters using the 0 to 90%
change time. Fortunately, the conventional time constant can be
computed from the published values by multiplying by an appropriate
constant in each case.

The time constant for a collector could be calculated theoretically

using equation (5), However, there is a large uncertainty in the value
of the effective heat capacity, C^. Duffie and Beckman [14] have shown
that the upper limit for this heat capacity is the sum of the products
of mass times specific heat for each of the components that make up the
collector (absorber, glass, insulation, etc), plus that of the transfer
fluid required to fill the collector. However, if this were used for C^,
the time constant would be overestimated. This is due to the fact that
the temperature of some materials in the collector only change a fraction
of the amount that the fluid temperature changes in such a transient

7
process. Consequently, the time constant is required to be determined
experimentally in ASHRAE Standard 93-77.

The actual test can be carried out in one of two ways. The most straight-
forward technique is to expose the collector to the solar radiation and
after the entering and exiting fluid temperatures have stablized, suddenly
shield the collector from the sun and record the exit fluid temperature
on a strip chart recorder. The incident radiation must be above 790 W/m^
(250 Btu/(h • ft^)). The entering fluid should be within ±1°C C+Z^F) of
the ambient temperature for the duration of the test. This latter require-
ment simplifies the data reduction process using the left side of equation
(5).

A second technique that can be used is to shield the collector from the
sun altogether (conduct the test inside for example). The inlet fluid
temperature is adjusted to 30°C (54°F) above the ambient and after the
exit temperature has stabilized, the inlet temperaure is suddenly
decreased to within ±1°C (+2°F) of the ambient, and the exit fluid tem-
perature is again recorded as a function of time. It should be noted
that the time constant determined experimentally according to either of
the above procedures is valid only for the range of ambient temperatures
used in the test. Its value is temperature dependent as can be seen
from equation (5) (U^^ is temperature dependent) and from Figure 5 of
reference [16]. However, since the main value of such a test is to
allow a relative comparison of collectors, it was felt that the one
test would be adequate.

2.2 COLLECTOR INCIDENT ANGLE MODIFIER

Simon and Buyco [17] have shown that the effective transmittance-
absorptance product, (Ta)^, of a solar collector can be satisfactorily
]

described by:

(xa), = ^6)
K.^^-)e,n
where K^^ = incident angle modifier

(Ta)g ^ = effective transmi ttance-absorptance product for the

'
collector at normal incidence

The incident angle modifier, K^^, is nothing more than a correction

factor which is a function of the incident angle, the angle between the
direct solar beam and the outward drawn normal to the plane of the col-
lector aperture. It describes how the optical efficiency of the collector
changes as the incident angle changes. It is an essential factor for
predicting all-day efficiency for a stationary solar collector.

Figure 4 shows how the effective transmi ttance-absorptance product, (Ta)^

varies as a function of incident angle for two types of flat-plate
collectors exposed to 100% direct radiation [18]. Using the relationship
for the incident angle modifier as given in equation (6), the variation

8
of incident angle modifier with incident angle for these same collectors
is shown in Figure 5, Based on the optical characteristics of flat-plate
collectors, data from Figure 5 can be replotted as in Figure 6 and will
result in linear plots as shown. This same linear relationship may or may
not exist for non flat-plate collectors depending on their optical charac-
teristics.

As with the time constant test, there are two ways in which the incident
angle modifier can be determined. The first technique is applicable when
the incident angle can be arbitrarily adjusted such as with a solar simu-
lator or with an outdoor movable test rack. In this case, a thermal effi-
ciency test is conducted in accordance with all the requirements of the
procedure used to get the near-normal-incidence efficiency curve except
that:

1, The inlet fluid temperaure is held within +1''C (±2°F) of the ambient
temperature.

2, The test is made at incident angles of 0, 30, 45, and 60°.

By comparing the above test results with those obtained in establishing

the efficiency curve at incident angles of less than 30° , values of
K can be computed as a function of incident angle.

The second technique is applicable for outdoor testing with a permanent

test rack where the collector orientation cannot be arbitrarily adjusted
with respect to the direction of the incident solar radiation (except for
perhaps adjustments in tilt ), The collector is tested for a complete day
with the inlet fluid temperature held constant as above. The efficiency
values are computed continuously and "pairs" of values are selected, one
from the morning and one from the afternoon, which correspond to values
of 30, 45, and 60° in incident angle. The two efficiency values in each
pair are averaged to compensate for transient effects and then used as
above to compute the incident angle modifier.

9
!

3. DESCRIPTION OF ^fBS TEST FACILITY

The collector testing loops designed, built, and used in the experiments
described in this report are housed at an abandoned NIKI Missile Site
adjacent to the NBS Laboratories in Gaithersburg, Maryland. One of the
underground bunkers built originally for storage of missiles was used
for fabrication of the test loops and they were mounted on the elevator
used originally for transporting the missiles up to ground level. This
unique facility allows NBS personnel to expose the test equipment to the
outdoor environment only on days when tests are run. In addition, the
same test equipment can then be used for conducting indoor tests of
collector heat loss characteristics.

Figure 7 shows the elevator door closed over the underground bunker occu-
pied by the NBS collector testing group*. Figure 8 shows the equipment

*Two additional underground bunkers are currently occupied and being used
by the NBS Fire Research Center.

11
 .

mounted on the elevator just after the doors have opened. Figures 9
and 10 show the equipment being raised and stopped in a convenient posi-
tion to clean the collector cover plate assembly on an air collector
array prior to the start of a test day. Figure 11 shows the equipment
in the testing configuration with the two water loops in the foreground,
the room holding the data acquisition system and associated instrumenta-
tion in the center, and the air loop on the far side. Some collector
arrays that have been tested were too large to mount on the test loop
structure and had to be kept permanently above ground and connected
to the test loops through flexible hoses. An example is the evacuated
tubular collector array in the foreground.

3.1 TEST LOOPS FOR WATER-HEATING COLLECTORS

Two separate test loops were built for the water-heating collectors in
contrast to a larger single loop which could accommodate 2-4 collectors.
This was done to be able to test more than one collector at a time and
still have the flexibility to make adjustments to individual collectors
during testing and yet not affect the other collectors being tested.
Both test loops are essentially identical and consist of an integral test
unit capable of supporting a typical flat-plate collector at a chosen
orientation while containing the flow loop within the enclosed base. The
tested collector can be adjusted over a wide range of tilt angles (0-70")
and orientations (0-360**) and easily added to or removed from the struc-
ture. As already indicated, a separate adjustable collector support
frame is used in cases where an exceptionally large collector is being
tested

Figure 12 shows a close-up of one of the test loops with a flat-plate

collector (collector No. 2) mounted to the top of the frame. Figure 13
is a schematic drawing of the test loop. As can be seen, it is a closed
loop similar in principle to that of Figure 1 and was designed to be
able to control and stabilize the collector fluid inlet temperature
to within tO.S^C (±1.0*F) and the fluid flow rate to within ±1%.
Pure water has been used as the working fluid in all collectors whose
characteristics are reported herein. This has eliminated uncertainties
and possible errors associated with knowing the working fluid properties.
Figure 14 shows the loops from the north and, as can be seen, all
equipment is sufficiently protected from the environment by having
it enclosed within the housing and also insulated to minimize heat
loss. Plumbing and hardware within the test stand base are encased
with 1.3 cm (0.5 in.) foamed rubber insulation and all exposed extension
plumbing is wrapped with 3.8 cm (1.5 in.) foamed rubber insulation.
The following paragraphs describe the details of the equipment used
in the flow loops starting with the collector and moving clockwise
in Figure 13. The specifications for the equipment and sensors are
included in Table 1.

The temperature measurement sections are located immediately upstream

and downstream of the collector allowing temperature sensors to be
inserted for measurement purposes. Figure 15 is a schematic drawing
of the temperature measurment section at the outlet of the collector

12
and Figure 16 is a photograph taken prior to insulating the pipe. These
sections are located as close as possible to the connection of the col-
lector and are insulated with 3.8 cm (1.5 in.) foamed rubber insulation
in order to minimize thermal heat losses and to insure a temperature
change of less than 0.05°C (0.1°F) between the sections and the collector.
Proper fluid mixing is provided by allowing the fluid to pass through
at least one right angle bend immediately before entering the measurement
section. The sections are constructed so that two temperature sensors
can be inserted simultaneously. One temperature sensor can be a sheathed
thermocouple or resistance thermometer inserted through a compression
fitting while at the opposite end a set of thermopile junctions can be
inserted into a thin-wall copper oil-filled well. To provide proper
bleeding of trapped air from the entire system, air bleed valves are
located immediately adjacent to the wells. An alternate technique for
insuring a well-mixed fluid stream at the temperature measuring station
is shown in Figure 17. This technique should work well but was not used
in either of the two loops described here.

Three types of temperature sensors have been used to monitor the absolute
temperature and the temperature difference between the inlet and outlet
of the collector. Only two types of temperature sensors are used simul-
taneously. Calibrated platinum resistance thermometers (PRT) or sheathed
type-T thermocouples are inserted in one end of the well for measuring
absolute temperatures while a six junction thermopile or PRT is used at
the other end to sense the temperature difference across the collector.

The choice between platinum resistance thermometers or thermocouples for

measuring absolute temperatures is primarily a matter of cost. Typically,
commercially available platinum resistance thermometers are accurate to
within ±0.1°C (±0.2°F). NBS calibrated 0.47 cm (3/16 in.), sheathed
platinum precision resistance thermometers with related electronic
bridges used in this study are accurate to within +0.05°C (+0.1°F).
In comparison, calibrated 0.64 cm (1/4 in.) sheathed type-T thermocouples
that were also used in this study are accurate to within ±0.1°C (lO.Z^F).
Therefore, both types of sensors meet the required accuracy for absolute
temperature measurements of within ±0.5°C (+1.0°F) as given in [4]. In
either case, caution should be taken to insure that the temperature
sensor immersion length into the temperature well agrees with the
immersion depth at calibration.

A type-T six-junction thermopile is used to measure the temperature

difference across the solar collector. At least six junctions were
determined to be necessary in order to determine small temperature
differences to an accuracy of within +0,1 "C (+0,2''F) (assuming the
thermopile readout instrumentation accuracy is within +0,01 mv ),
Currently the test procedure in [4] allows either resistance thermometers
or a thermopile to be used for determining the temperature differences
across the collectors. For reasons of accuracy and reliability, the
thermopile proved to be superior. Because of the large number of
junctions at either end of the thermopile, the wire gauge should be
as small as practically possible in order to minimize thermal conduction

13
 . .

losses along the thermopile leads away from the temperature measuring
junctions. For this installation, 30 or 36 gauge thermocouple wire
was found to be adequate. To further eliminate error due to thermal
conduction losses, the thermopile leads were doubled back along side
of the oil-filled temperature well as shown in Figure 15. The thermopile
is inserted into the well to at least a depth of 10 cm (4 in.). The
entire thermopile was constructed from the same spool of calibrated
wire

A water-to-air heat exchanger is used in each flow loop and an addition-

al water-to-water heat exchanger is used in one of the loops in order to
allow for a wider range of heat dump for some of the higher performance
collectors being tested. The water-to-air heat exchanger is a 36 cm
(14 in,) square by 2,5 cm (1,0 in,) deep fluid radiator (see Figure 18)
while the other is a counter-flow single-pass water-to-water exchanger.
For adjustment purposes, the water-to-air heat exchanger uses ganged ball
valves to serve as a bypass and modulate the heat dump. The ganged by-
pass valves produce a constant back pressure and consequently a steady
flow rate over the full range of adjustment. The heat dump of the water-
to-water heat exchanger is adjusted by controlling the secondary side
fluid flow rate supplied from a local water main.

Although not shown in the schematic drawing of Figure 13, the flow loop
can be opened in order to provide a means of periodically calibrating the
flow meters in place against a weigh tank and stopwatch. In addition,
the flow meters can also be easily removed for cleaning, checking, and
additional calibration.

Further downstream are the storage and recovery tanks including a pressure
relief valve. The 38 liter (10 gal.) hot water storage tank can be seen
in Figure 10 and is incorporated within the flow loop to act as a buffer
and eliminate thermal cycling which tends to occur within a closed system.
In addition, the storage tank contains a 1.5 kW immersion heater which is
periodically used to increase the overall flow loop temperature. In order
to further eliminate any air within the closed system, a fluid recovery
tank was substituted for the recommended expansion tank. Whenever the
flow loop pressure exceeds the 210 kPa (30 psi) relief valve setting,
the fluid is dumped into a 3 liter (0.8 gal.) recovery tank and later
recovered as the pressure in the loop drops,

Mercury-in-glass thermometers calibrated from 0-100°C (32-212°F) and

accurate to within +1°C (+2''F) are inserted into oil filled 0,93 cm
(3/8 in,) copper wells in line with the inlet and outlet to the storage
tank in order to monitor the respective temperatures. The inlet and
outlet temperatures are required for adjusting the heat exchanger bypass
in order to properly maintain steady conditions in the flow loop. To
obtain a steady-state condition, the heat exchanger bypass is adjusted
so that the storage tank inlet and outlet temperatures are equal to
within 2°C (4°F)

14
A provision for make-up fluid In the flow loop serves several purposes.
It can be used for filling or draining the loop or for pressurizing the
entire system. Make-up water is provided from a local water main.

A 0-0.4 l/s (0-6 gal./min.) 5-raicron particulate fluid filter is incor

porated within the loop in order to protect the system from foreign
particles. The filter is located upstream of the pump and flow meter
since they are both very susceptible to damage by particulate matter.

The controlled 500-watt heater downstream of the filter serves to sta-

bilize the inlet fluid tempeature to the collector. The input power to
the heater is adjusted so that the fluid temperature downstream of the
pump is held constant to within ±0.1°C (±0.2°F). The heater power is
controlled by a proportional temperature controller which senses the
temperature downstream of the heater immediately after the pump with
a type-T thermocouple and then proportionally controls the heater power
in response to the sensors* deviation from a set point temperature.
The proportionality of power output to set point deviation is adjustable
in order to minimize temperature cycling within the flow loop.

The flow rates encountered with the flow loop are very low (0-0.63 l/s
(0-1 gal./min.)) requiring a stability of +1% while periodically oper-
ating against high back pressures. For such circumstances, a low-flow
0-0.32 Vs (0-5 gal./min.) positive-displacement eccentric-disc pump
capable of pressures up to 340 kPa (50 psi) was selected over several
centrifugal pumps. Most typical centrifugal pumps are limited to higher
flow rates and lower working pressures. This pump is further capable
of coarsely adjusting the flow rate to within +0.0013 ii/s (+0.02 gal./
min.) while a 1.3 cm (0.5 in.) needle valve downstream of the pump
allows for fine adjustment.

Different types of flow meters are used in the two different flow loops.
The first is a turbine type flow meter with a passive magnetic transduce
and the second, a positive-displacement oval-gear flow meter with an
active transducer. Each flow meter produces a digital signal output pro
portional to the monitored flow rate. The low-flow omnidirectional
turbine meter and electronic display were initially calibrated over a
flow and temperature range of 0-0.063 l/s (0-1 gal./min.) and 20 - 100°C
(68-212*'F) respectively with stated accuracy and linearly of within
±0.9%. Likewise, the low-flow oval-gear positive-displacement meter
was calibrated over a flow range of 0.01-0.07 Vs (0.17-1.1 gal./min.)
and temperature from 20-100°C (68-212''F) with a similar accuracy. When
calibrated on a monthly basis, each flow meter can retain the specified
accuracy of within +1%.

During this test program, each flow meter was always calibrated in-situ
with a typical open-loop weight tank arrangement and in conjunction with
any related electronic displays or readouts. Over a period of a year,
the accuracy of the turbine flow meter would have changed by + 2.5%
if it were not for frequent calibration. It is believed that the 2.5%

15
drift was primarily due to wear and scale building up within the turbine
ball bearings. In comparison, the oval gear flow meter was not subject
to any similar drift in calibration over the same time interval. However,
the oval gear meter did experience a gradual increasing internal flow
resistance due to bearing wear. Consequently with time, the pressure
drop across the flow meter became unstable resulting in a pulsating flow
and the use of the meter was discontinued.

A visual flow meter is located immediately downstream of the electronic

flow meter in each loop. Functioning as a course flow indicator, the
visual flow meter also serves as a sight glass for determining whether
air is trapped within the system. Because the flow meter is used for
adjusting the coarse flow rate, the accuracy and resolution could be quite
low. The flow meter used, a rotometer, is capable of measuring flows of
0-0.07 Vs (0-1.1 gal./min.) with an accuracy of within +2% of full scale.

For safety purposes, a 0-210 kPa (0-30 psi) static pressure gauge is con-
nected at the solar collector inlet. The purpose is to visually indicate
the system pressure and guard against collector and system overpressuri-
zation.

3.2 TEST LOOP FOR AIR-HEATING COLLECTORS

The test loop for air-heating collectors is shown in Figure 20 and sche-
matically in Figure 21, It is divided into two major parts, the air
handling module and the collector stand. The air handling module is
mounted on a portable cart for ease of movement. The blower delivers
air to the air reconditioning apparatus where it is conditioned to the
desired temperature before entering the secondary flow measuring device.
The air then flows through flexible ducting to the collector inlet
measuring section which is attached directly to the collector array.
Leaving the collector array, the air passes through the collector outlet
measuring section and into another section of flexible ducting which is
connected to the primary flow measuring device. The air then passes to
the blower, thus completing the loop.

The primary details to note in the layout of the loop are the placement
of flow measuring devices on both the collector array inlet and outlet
with the primary flow measurement being on the collector outlet. Col-
lector temperature and pressure measurements are made as close to the
collector array as possible in duct sections which approximate actual
installed duct sections. The blower is located on the downstream side
of the collector array with creates a negative gauge pressure in the
collector. This is generally the way most solar air-heating systems
are configured. The details of the equipment in the flow loop will be
described in the following paragraphs starting with the collector and
moving clockwise in Figure 21. The specifications for the equipment and
sensors are included in Table 2.

16
The collector stand is shown in Figure 22 from the side and is designed
to accommodate a 2,5 m (8 ft) by 4.5 m (15 ft) collector array. The
collector array can be tilted from 0 to 65 degrees from the horizontal.
Casters on the base provide for rotation and mobility as shovm in
Figure 23. The stand is constructed of 7.6 x 7.6 cm x 0,64 cm (3 in.
X 3 in, X 1/4 in.) angle iron except for the pivot beams which are
10 cm X 10 cm X 0.95 cm (4 in. x 4 in. x 3/8 in.) angle iron. The base
is a solid welded unit mounted on locking casters. Two hand winches
provide a means for setting the collector tilt.

A 2.5 m (8 ft) by 4,5 m (15 ft) stud and plywood platform insulated with
glass fiber batts is mounted on top of the stand. The collectors are
mounted on this platform as normally installed on a roof. The outside
edges of the collector array are also insulated to simulate the effect
of adjacent collectors. The collector measurement sections are attached
directly to the collector manifold and are suspended from the bottom of
the platform as can be seen in Figure 22, The collector measuring sec-
tions were constructed in accordance with reference [4] and are located
at both the collector inlet and exit. Collector inlet and exit tempera-
ture difference, and pressure drop are measured at these locations.
Reference [4] requires a length of 2,5 V a x b at the inlet and 6,5
Va X b at the exit between the temperature measuring station and the
collector manifolds where a and b are the cross-section dimensions of
the duct, A second requirement is that the air inlet and air outlet
ducts shall be insulated in such a manner that the heat loss to the
ambient air would not cause a temperature change for any test of more
than 0.3''C (0,5°F) between the temperature measuring locations and the
collector. In order to satisfy both of the above requirements, highly
insulated measuring sections were required.
2
The collector stand is designed to test collector arrays of up to 10 m
(100 ft^) in area. Assuming that the maximum flow rate used with a 10 m
(100 ft^) collector would be 0,20 m^/(s,m^) (4 ft^/min,ft^) for a total
of 2 m /s(400 ft^/min,), a 20 cm (8 in,) diameter circular duct was
selected. This size provides duct velocities of approximately 5.9 m/s
(1150 ft/min.) for the maximum size collector array. The amount of insul
tion required was then calculated assuming a minimum flow rate of 0.03
m"^/s (64 ft^/min.)* and a maximum temperature difference between the air
stream and ambient of 70°C (126°F). Based on the properties of glass
fiber insulation and a length of 1,5 m (5 ft) between the collector
exit and the measuring section, it was found that 7 ft (2,1 m) of insu-
lation was needed to meet the requirements of [4], A more reasonable
thickness of 20 cm (8 in,) was chosen and based on actual test conditions
for the collector tested in this study, it was calculated that the max-
imum error possible in collector efficiency due to duct heat loss between
the collector and measuring sections was less than 1%,

*This would correspond to a flow rate of approximately 0,01 m /(s'm )

(2 f t^/(min,f t^) for a two module array and is the minimum measurable
flow rate through the minimum size nozzle in the present test loop,
17
The duct sections were constructed with 20 cm (8 in.) inner diameter
thin-walled aluminum tubing. The "L" shape seen in Figures 22 and 23
was necessary due to physical dimension limitations imposed by the
collector stand. The flow duct and surrounding insulation are all
contained in the 61 cm (24 in.) square sheet metal box.

The pressure measuring stations consist of four 0.48 cm (3/16 in.)

nipples soldered to the duct and centered over 1 mm (0.04 in.) diameter
holes. The four static pressure taps located symmetrically around the
duct at each station are manifolded to a single tygon tubing pressure
line. The pressure taps inside the duct were smoothed using emery cloth
to remove any possible burrs. The pressure drop across the collector is
measured by using a 0-0.25 kPa (0-1.0 in. H2O) inclined manometer.

The temperature measuring stations consist of thermocouples mounted in

probes which are inserted into the air stream through compression fittings.
The probes were made using 0.3 cm (1/8 in.) o.d, brass tubing inserted
into 0.32 cm (1/8 in.) compression fittings which were threaded into the
duct wall. The thermocouples were made from 24 gauge type-T thermocouple
wire. Thermocouples are permanently positioned at three different points
in the duct in order to indicate any variation in temperature across the
duct. In addition, a thermocouple is mounted on a probe which can be
moved across the duct when desired to determine the temperature distri-
bution. Temperature difference between the inlet and outlet of the
collector is determined using a six-junction type-T thermopile constructed
from a single spool of 24 gauge thermocouple wire. The thermopile junc-
tions are located at the centers of equal cross-sectional areas as shown
in Figure 24.

Flow mixers were not installed in the duct. Flow conditions at the temper-
ature measuring sections were found to be uniform (temperature traverses
were made). Flow mixers should only be used when really necessary. They
will cause a larger pressure difference between the collector and ambient
than normally experienced in an actual system installation, resulting in
increased collector air leakage and thus different performance.

The primary air flow measuring apparatus consists basically of a receiving

chamber, a discharge chamber, and an air-flow measuring nozzle as shown in
Figure 7 of reference [4]. The two chambers were formed separately using
24-gauge sheet metal to construct two boxes, one 100 x 40 x 40 cm (39 x 16
X 16 in.) and the other 70 x 40 x 40 cm (27 x 16 x 16 in.). The nozzle
mounting plate is a 0.5 cm (3/16 in.) steel plate with outside dimensions
of 50 cm (20 in.) square framing a center hole of 32 cm (12.5 in.) square.
The plate surface was ground flat and smooth where the nozzle is mounted.
Diffusion baffles which are decorative grills having approximately 40%
free area were purchased from a local hardware store and installed in the
two chambers at the locations specified in [4]. The two chambers and the
nozzle mounting plate are joined with gaskets and bolted together. An
air-tight window was formed in the duct to allow access to the nozzle.
Holes were cut in the back walls of the two chambers to allow connections
through duct work with the rest of the loop.

18
Nozzles were constructed in accordance with reference [4] and are fixed
to 36 cm (14 in.) squares with magnetic mounting strips. The magnetic
mounting strips and pressure of the air hold the nozzle in place and
form a tight seal. Five interchangeable nozzles were purchased, in sizes
of; 5, 6.5, 7.5, 10, and 12.5 cm (2, 2.5, 3, 4, and 5 in.) throat
diameters, which provide a range of flow rates from 0.03 to 0.41 m /sec
(70 to 950 ft /min,). Pressure taps are incorporated into the duct
walls as specified in [4] by centering and soldering 0.5 cm (3/16 in.)
nipples over 0.013 cm (0.04 in.) holes located on the four duct walls.
The two sides of the pressure sensing devices for the nozzle pressure
difference and the nozzle discharge gauge pressure each are connected
to four externally manifolded pressure taps. In addition, a pi tot tube
is used to sense velocity pressure at the nozzle throat. The pressure
difference across the nozzle is determined using a 0-1.2 kPa (0-5 in.
H2O) inclined manometer in parallel with an electronic pressure trans-
ducer. The pressure transducer is an elastic diaphragm type and produces
a 0 to 5 volt signal over the 0-1,2 kPa (0-5 in. H2O) pressure range.
The output signal is reduced to a 0 to 500 mv range using a voltage
divider network. The nozzle discharge gauge pressure is determined using
a 0-2,5 kPa (0-10 in. H2O) vertical manometer. The pi tot tube is
connected to a 0-1.2 kPa (0-5 in. H2O) inclined manometer. The procedure
for calculating air flow rate using these measurements is given in
Appendix A.

The blower is a centrifugal industrial model with a 3 phase 220 volt 3

horsepower motor capable of 0.47 m"^/s (1000 ft'^/min.) at standard condi-
tions and a static head of 2,5 kPa (10 in, H2O). This is a great deal
more capacity than is needed for most test size collector arrays. How-
ever, the air handling module was designed to also be used for testing
full-sized thermal storage units. The blower must be sized to provide
the required flow rate over the pressure losses in the test loop. For
the present system, the major pressure loss occurs at the flow measuring
nozzle and is 0,25 to 0,75 kPa (1 to 3 in, H2O), The amount of air
delivered by the blower is controlled by the bayonet-type damper
on the blower outlet (coarse control) and stovepipe-type damper on the
inlet (fine control). The blower assembly is mounted on rubber vibration
isolators and duct connections to the blower are made using flexible
rubber connections which results in nearly complete vibration isolation
of the blower.

The air reconditioning apparatus was designed to deliver air at a set

temperature at its exit. Through the use of two hand-set dampers, part
of the hot air from the blower is allowed to pass on to the collector
while the rest of the hot air is exhausted to the atmosphere. Further
downstream, a third handset damper is used to bring in ambient make-up
air. Through proper positioning of the three dampers, the resulting
air stream temperature can be adjusted to a value slightly below the
desired temperature. The final temperature is attained and finely
controlled using 6 kW electric resistance heaters coupled to a propor-
tional power controller. The heater provides the correct amount of

19
energy needed to bring the air-stream up to the desired temperature.
It is important that the temperature sensor be placed far enough away
from the heaters to sense the true temperature of the air stream. In
addition to the proportionally controlled heaters, two 18 kW industrial
duct heaters are also built into the loop. These duct heaters are con-
trolled in 3 kW increments by hand set switches and are used primarily to
provide high temperature step inputs for thermal storage device testing.

Based on analysis and experience gained during this test program, it is

felt that steady-state conditions are easier to maintain if the tempera-
ture difference between inlet fluid and ambient air is held constant
rather than holding the inlet fluid temperature constant alone. Thus,
electric duct heaters with no controls (other than manual) can be used
when operating the test equipment in an open-loop configuration. Since
ambient temperature is fairly constant over typical test periods, the
test requirements in the current Standard 93-77 can still be met.

The dry bulb and wet bulb temperatures of the air stream are measured
just downstream of the air reconditioning apparatus in order to be able
to determine the specific volume and humidity ratio of the air. Both are
measured using thermocouples, the latter with a saturated thermocouple
exposed to the air stream. The humidity ratio is used in the calculation
of air flow rate as explained in Appendix A and needs to be measured at
only one point in the loop.

The secondary flow-measuring apparatus is a commercial unit consisting of

air flow straighteners upstream of an averaging pi tot tube arrangement.
An aluminum honeycomb flow straightener precedes a pitot tube station.
The pitot tube station consists of several total and static pressure
taps carefully spaced to produce an average duct velocity reading. The
static and total pressure taps are connected to common static and
total pressure manifolds. The pressure difference between the two mani-
folds is read by a 0-0,25 kPa (0-1 in, H^O) slant gauge manometer and a
"cfm meter". The "cfm meter" is a bourdon-tube pressure gauge with a
cfm scale calibrated to match the specific design of the unit. This
dial reading was only used to adjust the blower and its dampers to the
approximate flow rate desired for a particular test. This secondary
flow-measuring apparatus was found to yield flow rates approximately
15% high when calibrated against the nozzle apparatus,

3,3 METEORLOGICAL MEASUREMENTS

The specifications for the instrumentation is given in Table 3. Two

types of ambient air sensors are being used in conjunction with a stan-
dard calibrated total immersion, ASTM liquid-filled thermometer. One
is a type-T calibrated thermocouple and the other a precision platinum
resistance thermometer. The liquid-filled and resistance thermometers
are accurate to within ±0,1°C (±0,2*'F) while the calibrated thermocouple
uncertainity is +0,1''C (±0,2''F), All are housed within a well ventilated
small instrument shelter located 1,25 m (4,1 ft) above the ground with
its door facing north.

20
The vd.nd speed is measured by a 3-cup wind anemometer delivering a dc
voltage proportional to the wind velocity. Being mounted upon a por-
table adjustable base, the wind anemometer can be located close to
any tested collector and adjusted in height. As the result of a wind
tunnel calibration, the resulting uncertainty in wind velocity measure-
ment is +0.35 m/sec (+ 0.8 mi/h).

The wind direction was recorded by visual observation of a directional

wind vane during each test interval for the first part of the test pro-
gram. Later in the testing program, a weather vane producing an analog
output proportional to wind direction was installed to continuously
monitor wind direction.

For measuring total solar radiation incident on a solar collector two

types of pyranometers were used. Initially, a "black-and-white" model
8-48* pyranometer was used and was later replaced with an "all-black"
model PSP*. Both pyranometers have a thermopile sensing element and
similar protective dome glazings but differ in the receiver optical
coatings and patterns. An intercomparison was made between two of the
model 8-48 instruments and a model PSP when all were tilted at several
angles from the horizontal up to 56**, typical of collector tilt angles
to be used in testing. The results are shown in Table 4. There was as
much as a 7.4% difference even though the instruments agreed within 1.5%
in the horizontal position. Similar experiences by other researchers
in the field lead to the adoption of the requirement for using a first
class pyranometer as classified by WMO [5]. The model PSP pyranometer
is capable of an uncertainity of + 3.0% and has been shown to have
typical tilt errors of less than 0.9% for angles ranging from 0° to
50°. These data were obtained during an intercomparison against an
Angstrom pyreheliometer .**

The pyranometer is mounted on an adjacent surface parallel to the

collector in such a manner that it does not cast any shadow onto the
collector. Care is taken to insure that the pyranometer is at the same
tilt as the solar collector and to minimize reflected and reradiated
energy from the solar collector onto the pyranometer. The typical mount-
ing scheme is shown in Figure 25.

The diffuse component of the incident solar radiation is determined for

each efficiency test point by shading the pyranometer following the
recommended technique of reference [4] (see Figure 26).

*Eppley Laboratories, Newport, Rhode Island.

**Personal communications with the Solar Energy Research and Educational

Foundation, Washington, D.C.

21
In addition to monitoring the total solar radiation, the direct normal
radiation is measured using a pyrheliometer , The pyrheliometer tracks
the sun and thereby measures the direct normal incident solar radiation
throughout the day.

The sky temperature is being determined for some tests by the

use of either of two instruments. One is a commercially-available
pj^geometer and the other is a net radiometer designed and built by
the French Building Research Center for NBS and described in Appendix B,
Both the pyr geometer and net radiometer are located within the collector
plane. Based upon a calibration against a blackbody source, the
pyrgeometer is capable of determining the absolute incoming long wave
radiation to an uncertainity of + 3% over the spectrum of 0,35 to 50^m.
The net radiometer has a greater degree of uncertainty (typically + 8%
in net radiation) when compared against the pyrgeometer. The principal
cause of the large uncertainty is suspected to be the variable effective
optical properties of the net radiometer in response to exposure and
solar spectral distribution.

Barometric pressure is determined periodically throughout the test

period using a precision barometer,

A special incident angle meter was designed and built for determining the
angle between the sun's direct beam and the outward drawn normal to plane
of the collector aperture, A schematic drawing of the meter is shown in
Figure 27, The principle of operation is very similar to that of a sun
dial. It consists of two flat 0,64 cm (1/4 in,) thick clear plastic
10 cm X 10 cm (4 in, x 4 in,) sections joined and braced orthogonally.
The vertical clear plastic member contains a quarter-circular graduated
scale in angular degrees identical to a protractor, and a shadow pointer
located at the center of curvature of the angular scale. In operation,
the meter base is placed on the surface of the collector and rotated
until the shadow cast by the pointer is located within the plane of the
vertical member. The incident angle is then determined by the inter-
section of the shadow and the graduated scale. The incident angle meter
has consistently agreed to within 1° to 2° when compared against the
analytically-predicted incident angle upon a tilted surface for typical
test conditions,

3,4 DATA ACQUISITION

A variety of data are monitored and recorded by the data acquisition

system and other recorders. Data such as flow rates, temperatures,
and pressures in the air and liquid collectors test loops as well as
meteorological data are read into one central instrumentation room.
The room is environmentally controlled and is located between the air
and liquid collector test loops as shown in Figure 11.

The input signals from the various test loop transducers consist of
analog voltage, digital voltage, or variable resistance. Analog signals
are produced by thermopiles, pyranometers, anemometers, or pressure
22
transducers; digital signals by liquid flow meters; and variable resis-
tance from resistance thermometers. The digital and variable resistance
signals are converted into analog signals before input to the data
acquisition system. A flow rate monitor interprets and converts the
flow transducer digital signals into analog form while separate individ-
ually matched and calibrated bridge amplifiers interpret and convert
the resistance thermometer signals. Type-T thermocouples are provided
with an automatic electronic reference junction or an ice bath.

After conditioning the input data into either analog or digital signals,
the information is fed into the equipment shown in Figure 28. The equip-
ment is comprised of a combination of strip chart recorders and electronic
integrators connected in parallel with a data logger. The purpose of the
strip chart recorders is to monitor pertinent specific information on a
continuous basis while the data logger scans and records all the input
data on a periodic basis. Information such as insolation, flow rates,
and inlet-outlet temperature differences, are continuously recorded on
strip chart recorders in order to readily observe any transients. Elec-
tronic integrators are periodically used to integrate quantities such
as incident solar radiation or fluid temperature rise across a solar
collector. The data logger scan interval should be as small as possible
and has most often been one minute. Of course, the rate of data scanning
should depend on the type and intensity of transients being monitored.
The data logger is capable of calculating and recording the arithmetric
average of up to 16 inputs over a specific time interval; thus the scan
rate can be once per minute, whereas the average of the inputs can be
automatically calculated and printed every five minutes. Both the scan
rate and the averager time interval are independent and adjustable.
Typically input quantities such as wind speed, air temperature, and
collector fluid flow rate are averaged using the built-in averager and
printed on paper tape for later analysis. Approximately half-way through
the experimental study described in this report, a magnetic tape drive
recording system was interfaced with the data logger system to expedite
the data reduction on the NBS central computer facility.

Table 5 includes the specifications for the strip chart recorders, inte-
grators, and data logger used.

23
i
4. TEST RESULTS

During the period covered by this report (1976 and 1977), six different
collectors were tested. They are described in Table 6. Five were
liquid-heating collectors, the other an air heater. The following
sections describe the tests conducted and results obtained.

4.1 LIQUID-HEATING COLLECTORS

Time Constant : The collector time constants were determined following

the procedures described in Section 2,1 of this report and reference [4],
To provide a radiation step-change, the collectors were either suddenly
shaded or unshaded with a white opaque cover suspended above the collector.
The opaque cover is shown in place above collector no. 4 in Figure 29.
In the case of the linear Fresnel lens tracking collector (collector
no. 5), the step-change was accomplished by rotating the collector off
the track, of the beam radiation toward the diffuse sky.

25
A graph of inlet and outlet temperature for collector no, 1 during a
time-constant test when the collector was suddenly shaded is shown
in Figure 30. The time constant of each of the five liquid-heating
collectors is given in Table 7. The tabulated time constant was calcu-
lated by averaging the results from numerous tests where the collector
was both shaded and unshaded. On the average, the difference in the
measured time constant between shading and unshading was less than 6%,
In order of decreasing magnitude, the time constant was largest for
collector no. 4 then collectors no. 3, 2, 1, and shortest for collector
no. 5. Collectors no. 3, 2, and 1, being of the flat-plate design,
exhibited very similar time constants (on the order of 100 s) . The
short time constant (55 s) for collector no. 5 was due to a smaller
absorber mass and high flow rate compared to the flat-plate collectors.
The very large time constant (20 min.) for collector no, 4 was primarily
the result of a large collector fluid capacity (34 liter (9 gal,)) and
a very low flow rate (0.018 liter/s(0.3 gal. /min.)).

The alternate technique of conducting the test indoors and suddenly

changing the inlet fluid temperature was used for one collector and
found to give comparable results. However, this technique is much more
difficult to use experimentally.

During the deliberations that led to the adoption of ASHRAE Standard

93-77, there was controversy over what time constant is actually being
determined by this test. In reality, there can be more than one time
constant for a given collector since the various components of the
collector respond over a different time period to a sudden change in
the solar insolation conditions. Although not shown in Figure 30,
several strip chart recordings of the exit temperature from the flat-
plate collectors during a time constant test indicated another noticeable
but more gradual response after 6 or 7 min. which was an indication
that some parts of the collector were just beginning to respond. The
time constants indicated in Table 7 and determined in accordance with
Standard 93-77, are primarily a function of the fluid dwell time and
the absorber plate response.

Near-Normal-Incidence Efficiency: Collector no. 1 was the first col-

lector tested. It was encased within a wooden frame on the edge in
order to simulate the effect of adjacent collectors that would exist
in an actual field installation. This was felt to be particularly
important for this collector due to a design that put the absorber
plate in intimate contact with the edge of the collector case.
Throughout the testing of this collector, the "black-and-white" model
8-48 pyranometer was used for measuring the incident solar radiation
in the plane of the collector aperture. As already mentioned in Section
3.3 of this report, possible errors exist in the efficiency values due
to the collector being tested at tilt angles (from the horizontal) of
20-40°. Correction factors were not introduced to account for the
pyranometer error in analyzing the data for this collector since no
reproducible correction factors have yet been determined for this

26
pyranometer. For example, the data from Table 4 indicates the correction
factor could be anywhere from 3.9% to 7.4% for the 45° tilt position
depending on the insolation level and the particular instrument used.*

A more comprehensive laboratory study has recently been completed [19]

in which the error is indicated to be only between 1.5 and 2.0%. This
pyranometer was not used for any further tests following the tests on
collector no, 1,

The results of the efficiency tests are shown in Figures 31 and 32. The
efficiency values are based on the aperture area of the collector, 1.61 m^
(17,3 ft^). In Figure 31, all data were taken at solar noon on a number
of different test days whereas in Figure 32, the data were taken + 3/4 h
symmetrical with solar noon, A linear regression** curve fit of the data
in both figures resulted in identical curves. Therefore, for this type
of collector, the data scattering caused by collector thermal capacitance
can be averaged out if care is taken to make the measurements symmetrical
with solar noon as required by [4],

A solar screen was placed over the collector as shown in Figure 33 to

reduce the solar radiation incident upon the collector. The screen was
a flat-weave opaque glass fiber type and allowed efficiency data to be
taken at larger values of At/I, Normally, test data are obtained over
the entire range of At/ I values primarily by changing the operating
temperature of the collector, tj ^, This is because the ambient temper-
ature and incident solar radiation are generally fixed (within limits)
for the location and time of year in which the test is being conducted.
However, Figure 32 shows that is was possible to establish part of
the efficiency curve by using the screen. This technique is not generally
recommended since in reality using lower insolation levels is not
equivalent to operating the collector at elevated temperatures. However,
the error is small for flat-plate collectors which operate at relatively
low temperatures (<80°C (175°F)),

Figures 34 and 35 show the results of the efficiency tests for collectors
no, 2 and no. 3 (again based on aperture area). The edges of these col-
lectors were not encased in a wooden frame nor insulated on the edge
due to a good thermal design of the modules themselves.

Collector no. 4 was a single-glazed evacuated tubular collector with a

back diffuse reflector. The results of the efficiency tests are shown

*This is assuming the calibration factor for the "all-black" model PSP
does not change with the tilt angle.

**It is felt that a linear plot is adequate for most collectors. Even
though collectors don't have a constant heat loss coefficient as is
indicated by a linear plot, data scatter caused by variable ambient
conditions and experimental uncertainties prevent the determination of
the "true" 2nd-order response of the collector.

27
in Figure 36 and a photograph of the collector under test is shown in
in Figure 25. The collector array consisted of two modules connected
in parallel. The large thermal time constant resulting from the large
collector fluid capacity and low operating flow rate coupled with a
large temperature rise across the collector contributed to problems of
stabilization during testing. Extreme care had to be taken to eliminate
any perturbations within the collector to insure steady-state conditions.
At least one hour was allowed for temperature stabilization after reaching
operating conditions. The time interval over which the experimental
data are collected and integrated should be at least one time constant,
according to [4], amounting to a period of 20 min, for this collector.
Initially, a period of one hour was used for integration and it was
found the 20 min, period resulted in efficiency values only 1,2%
different for data collected across solar noon.

The aperture area was used in computing the efficiency values in

Figure 36 and was interpreted as the area of the collector tubes and
the interstitial spacing between the tubes excluding the end brackets
and center manifold.

Collector no. 5 was a tracking, concentrating collector utilizing

a linear Fresnel lens and a selective-black-chrome-coated absorber.
The efficiency curve is shown in Figure 37. The tested collector
consisted of two modules connected in parallel, whose tilt angle could
be periodically adjusted and which automatically tracked the diurnal
movements of the sun. The tested modules are shown in Figure 38.
In Figure 37 two curves are shown, one representing a 0° incident angle
efficiency and the other a 23° incident angle efficiency. The efficiency
values in Figure 37 were calculated, only using the direct component of
the solar radiation and based on the aperture area of 1.8 m^ (19.4 ft^).
During three of the five days of testing, the collector tilt angle was'
such that the incident angle across solar noon was 23". On two other
days, the collector had an incident angle of 0 and 10° across solar noon.
Based on results obtained by Pendleton [20], the incident angle has little
or no effect for angles up to 10 degrees. Therefore, a single curve
was used for data collected at 0 and 10 degrees. Note that the efficiency
decreases by 8% for an incident angle of 23° at solar noon. This is
equivalent to the yearly declination variation for a collector tilt
angle equal to the latitude.

Figure 39 shows the results of the near-normal-incidence efficiency tests

for all five of the water-heating collectors. The efficiency values are
based on the aperture area as given in Table 7. The data are replotted
in Figure 40 with the efficiency values based on gross collector area as
required by ASHRAE Standard 93-77. The change in efficiency between
these two figures is small for collectors 1 through 4 since there is
only a relatively small difference in aperture and gross area for these
collectors (<15%). For collector no. 4, the gross area was assumed to
to include the end brackets and center manifold. There is a large
change in the efficiency of the concentrating collector (collector no, 5)
because of the authors' interpretation of gross area for this collector.
The aperture area of collector no. 5 was assumed to be the frontal area

28
of the linear Fresnel lens; the gross area was calculated based on
the center to center spacing between individual collectors.

During all tests, the pressure drop across the collector was measured
and recorded. Figures 41 and 42 show how the pressure drop varied as a
function of flow rate for one of the flat-plate collectors (Figure 41)
and the evacuated tubular collector (Figure 42).

Incident Angle Modifier: Two different techniques were used to determine

the incident angle modifiers of the water-heating collectors. The first
technique involved keeping the collector test stand stationary with the
collector facing south, as was done when determining the near-normal-
incidence efficiency, and allowing the incident angle to change with
respect to the diurnal movement of the sun. In this procedure, the
instantaneous collector efficiency was determined throughout the day as
a function of incident angle, and morning and afternoon values (for the
same incident angle) were averaged to eliminate the thermal lag effect.
In this way, the incident angle modifier was determined for a wide range
of incident angles in one day. The second technique involved orienting
the test stand so that the collector faced north but tilted to a near-
horizontal position with the tilt angle adjusted in order to obtain a
specific incident angle across solar noon. With this procedure, it was
only possible to obtain one value of incident angle modifier per test day.

The results for collectors no. 1, 2, and 3 are shown in Figures 43, 44,
and 45, respectively. The second technique described above was used to
obtain the curves shown in Figures 43 and 45. Both techniques were
used for collector no. 2 and, as can be seen in Figure 44, gave comparable
results. The second technique or orienting the test stand to the north
was used to determine the solid line in the figure. The first technique
was completed in one day, the data points indicated by circles taken
in the morning and those by triangles in the afternoon. The dashed line
is a curve fit through the average of these points. As can be seen by
the fact that the circles and triangles are relatively close together,
there was only a small thermal lag with this collector. This is consis-
tent with its small time constant of 1,6 minutes. As can also be seen
in Figure 10, there was essentially a linear relationship between K^^
and the term (l/cos9) - 1, which is consistent with the theoretical
model for optical efficiency of a flat-plate collector [17],

Figure 46 shows test results for collector no, 4, the evacuated-tubular

collector. Note that the thermal lag effect is more pronounced in
Figure 46 than in Figure 44 which is consistent with the fact that
collector no. 4 has a time constant of 20.3 minutes. In addition, the
incident angle modifier increases with increasing incident angle, the
opposite from that of the flat-plate collector in Figures 43-45. Since
the tubes were not used with a concentrating reflector, this tubular
collector has its minimum efficiency at solar noon and larger values
at other times of the day. This is explained by the fact that the
tubes intercept a larger fraction of of the incoming solar radiation
at the large incident angles and also that the transmittance of solar

29
radiation through each glass cover tube does not change with time of
day since the sun is practically always normal to the glass surface.

With simple glazed flat-plate collectors, the incident angle modifier

varies in a smooth manner with incident angle and is basically indepen-
dent of orientation (except for perhaps secondary effects due to changes
in tilt and hence convection heat loss or different shading of the
absorber by the collector frame at different orientations). The testing
procedure in Standard 93-77 for determining the incident angle modifier
was based on these characteristics and as a result is only really valid
for simple flat-plate collectors. The way in which the test should
be conducted for other types of collectors depends in general on the
specific collector being tested. For example, the tubular collector
tested and described in this report is sensitive to changes in incident
angle due to changes in both the sun's altitude angle and hour angle.
Therefore, determination of two incident angle modifiers would be appro-
priate; one describing how the collector's efficiency changes as the
angle between the outward drawn normal and direct beam increases in a
direction parallel to the axis of the tubes and a second modifier for
the direction perpendicular to the tube's axis. The data in Figure 46
were determined by keeping the collector facing south in a fixed position
with the tilt adjusted so that the incident angle was zero at solar noon.
Therefore, the data is representative of changes due primarily to a
change in the diurnal hour angle.

A testing procedure to get the two incident angle modifier curves for a
non-movable trough-type concentrating collector with an east-west axis
using a south-facing outdoor test stand has recently been described by
Thomas [21]. In a similar fashion, Johnson [22] has demonstrated the
difficulty of this type of testing for a single-glazed flat-plate
collector. The collector had a non-selective coating on a copper roll-
bond absorber. Mylar strips running the width of the collector were
installed between the absorber and glazing to decrease convection and
reradiation losses. The strips were approximately 10 cm (4 in.) deep
and 0.95 cm (0.37 in.) apart. The optical characteristics of this col-
lector were similar to that of a trough-type concentrator with the trough
axis running east-west.

Figure 47 shows the results of tests conducted by Johnson using an indoor

simulator at the NASA Lewis Research Center. With the collector tilted
at 57** and oriented normal to the simxilator source, the collector was then
rotated about a vertical axis and data taken to establish the upper dashed
curve. Then the tests were repeated for other tilt angles as indicated.
Unfortunately, by rotating the collector about a vertical axis, the angle
between the "sun's" beam and both principal axes in the plane of the
collector aperture were changing simultaneously. As a result, the effect
of incident angle changes perpendicular to each principal axis could not
be isolated. Notice in the figure that with each of the lower three
dashed curves, one of the data points lies far from the indicated curve.

30
If the tests were to be repeated, the collector orientation could be
changed in such a way as to establish two incident angle modifiers for
the collector. To show both effects, a family of curves should be
plotted for K^^ as a function of the incident angle projected into an
east-west plane with Q-^ of Figure 48 as a parameter. The test itself
would be relatively easy to conduct using a simulator.

The experimental incident-angle-modifier curves for collectors no, 1-4

are shown collectively in Figure 49 and in Figure 50 in comparison with
the theoretically-predicted curves for single and double-glazed flat-
plate collectors with non-selective coated absorbers exposed to direct
beam radiation. The experimental results are in approximate agreement
with the theoretical curves. The differences are due to several reasons.
Collector no, 2 was a double-glazed collector but its incident angle
modifier was not as dependent on incident angle as was expected from
the theoretical curve. The experiment was conducted under 18-40% dif-
fuse radiation which should cause the collector's efficiency to decrease
less with incident angle than if it were exposed to 100% direct beam
radiation. In contrast, the efficiency of collector no. 3 decreased
more rapidly than would have been though since it was a single-glazed
collector and tested under 11-24% diffuse radiation. However, the
absorptance of the selective surface on the absorber could have been
a function of incident angle and hence caused the indicated response.
The angular response for collector no. 1 may be in error due to (1)
the use of the "black-and-white" pyranometer during the tests or (2)
as a result of 15-25% diffuse radiation during the test.

An incident angle modifier was not determined for collector no. 5, the
tracking, concentrating collector. As with collector no. 4, two incident
angle modifiers would be in order. The value of one of them, that which
describes the change in efficiency as the angle between the direct beam
and the outward drawn normal increases in a direction perpendicular to the
axis of the concentrator, is identically 1 since this collector tracks
the sun in the east-west direction to a tolerance of 0.2°. However, the
collector tilt angle is not normally changed during the year to reflect
changes in the sun's declination. Therefore, an incident angle modifier
should be used to account for the resulting change in performance. As
already indicated in Figure 37, efficiency data v^re taken on this col-
lector for constant incident angles of both 0° and 23**. The incident
angle modifier was computed, based on these data, and is shown in Figure
51. Additional tests could be conducted at other angles to establish
a complete incident-angle-modifier curve for this collector.

4.2 AIR-HEATING COLLECTORS

Time Constant : Figure 52 shows results of the time constant test on

collector no. 6, the air heater. The test was conducted outside under
a clear, sunny sky by shading the collector after steady-state conditions
had been reached. Initially the fluid entered the collector at 28.5''C
(83.3'*F) and exited at approximately 63°C (145°F) with a flow rate of

3L
0.01 m /(s.m ) (2 ft /(min. ft ). As can be seen, the time constant
was found to be approximately 12.7 minutes. This is much longer than
for flat-plate collectors which use water as the transfer fluid, but
less than that for the water-heating ev a cu a ted- tubular collector
(collector no. 4).

Near-Normal-Incidence Efficiency :Figure 53 shows the results of the

near-normal-incidence efficiency tests for the air heater. As can be
seen, curves are shown for two different flow rates. A recommendation
is given in Standard 93-77 that two different curves be established
because, unlike water-heating collectors, the efficiency of air heaters
is significantly affected by flow rate. The curves in Figure 53 are
based on the gross area of 7.25 m^ (78.0 ft^) as recommended in the
Standard. Figure 54 shows how the curve for the test with the lower
flow rate changes if the aperture area were used to compute efficiency.

The efficiency values in Figure 53 are lower than expected, based on

theoretical predictions for this collector. It was found during the
test that air was leaking into the four-module array. Air leakage in
air heaters is a troublesome problem that can affect test results,
as well as the actual performance of an installed system. During test-
ing, the test loop can be sealed sufficiently well with duct tape,
caulking, etc. Leakage in and around the collector may occur as a
result of the collector design and recommended installation practice.
This type of leakage of course should not be eliminated. However,
measurements should be made in such a way as to determine the true out-
put of the collector array. In addition, the array should be tested
in a manner such that the air leakage occuring during the tests will
be indicative of the leakage that will occur in an actual installation.

Close and Yasoff [2 3] have recently published the results of an analysis

of air leakage in air-heating collectors. They assumed a constant leakage
rate along the length of the collector and determined its effect on the
efficiency measurements for all combinations of:

a. operating the collector under negative pressure (air leaking in) and
positive pressure (air leaking out), and

b. measuring the air flow rate before (upstream) the collector or after
(downstream) the collector.

The results are shown in Figure 55. The abscissa of the plots is the
leakage rate divided by the measured flow rate and the ordinate is the
ratio of the actual collector efficiency to the measured efficiency.
As can be seen, for three of the testing configurations, the discrepancy
between measured and actual efficiency is a direct function of the
leakage rate. In addition, for the case where the collector is operat-
ing under negative pressure and air is leaking in, the difference in
efficiency depends on the difference in temperature between the ambient
air and the entering air stream to the collector. Therefore, the error
in efficiency for this case is larger for the data points at the higher

32-
inlet fluid temperatures relative to ambient. Also note that when the
air leaks in, the actual efficiency is larger than indicated by the
measurements whereas just the opposite is true when air leaks out. This
is consistent with what one would intuitively expect. When air leaks
into the collector at a cooler temperature than is measured at the collec-
tor inlet, the collector is heating up the air over a larger temperature
difference than is indicated by the measurements. In addition, if the
air flow measurement is upstream of the collector, the collector is
heating a larger quantity of air than is indicated by the measurements.
When air leaks out of the collector, and the flow rate is measured up-
stream, the quantity of useful heated air is less than indicated by the
measurements. However, if air leaks out and the air flow measurement
is made downstream of the collector, the quantity of useful heated air
is precisely what is measured and as a result, there is no difference
between actual and measured collector efficiency. Data for this case
are not shown in Figure 55 since they would produce a horizontal
straight line with an ordinate value of 1.

The results of the above analysis have a direct implication for the
testing of air-heating collectors:

1. If the collector is normally operated under positive pressure, it

should be tested while operating under positive pressure and the air
flow rate measured downstream of the collector.

2, If the collector is normally operated under negative pressure, it

should be tested while operating under negative pressure and the air
flow rate measured both upstream and downstream of the collector in
order to quantify the leakage rate. In this manner, an estimate of
the actual collector efficiency could be made by the user of the
collector with the data of Figure 55 even though the "correction"
might not be made and published as part of the test results.

The data of Figure 53 were obtained with collector no, 6 operating under
negative pressure (normal installation practice for this collector) and
the flow rate measured downstream of the collector. Use of "smoke-bombs"
around the collector indicated air leakage into the collector. Therefore,
the actual collector efficiency was higher than that indicated in Figures
53 and 54.

As indicated by equation (1), the useful heat output of the collector is

directly proportional to the collector heat removal factor, F^. If the
collector test results are to be used without correction to predict the
performance of the collector in an actual field installation, then the
value of which occurs during the test should be approximately the same
as that which will occur in the field. The expression for F^ is [14]:

(7)

33
Therefore, the conditions during testing and in actual field operation
should be such that m/A, Cp, Uj^, and F' are approximately the same in
both cases. For liquid-heating collectors, this is easy to accomplish.
The flow rate per unit collector area, m/A, the type of fluid and
hence specific heat, Cp, and the collector tilt angle, temperature,
incident solar radiation, and ambient conditions are such that the
heat loss coefficient, Uj^, is approximately the same. In addition, the
collector efficiency factor, F' is primarily a function of the geometry
,

of the absorber (material, thickness, distance between tubes, etc.)

and of course is the same for the tested and installed collectors at
any flow rate.

The situation is slightly different for air heaters. Although m/A, Cp,
and U-^ can be made the same in a similar fashion as with the liquid-
heating collectors, assuring the same value of F' is more difficult.
For air heaters, F' is primarily a function of the convection heat
transfer coefficient between the absorber and the air stream. Since
the air flow is nearly always in the turbulent flow range (to maximize
the heat transfer), the heat transfer coefficient is determined pri-
marily by the value of the Reynolds number in the collector. For a
collector designed with a simple rectangular channel under the absorber
such as in collector no. 6, the Reynolds number is given by:

where V = velocity of the air in the collector, m/s

= hydraulic diameter of the channel through which the air flows, m

u = kinematic viscosity of the air, m/s

Since the hydraulic diameter of the channel is equal to twice its width,
equation (8) can be rearranged to give:

Re = 2mL (9)
y A
where

y = dynamic viscosity, Pa's

L = flow path length of the collector array, m

Consequently, since the mass flow rate per unit area is being maintained
the same between the tested collector and the value to be used in the
field installation, the length of the collector or flow path length should
be maintained to insure the same Reynolds number, convection coefficient,
and hence F '

The particular collector tested in this program is designed to be

installed in the field with two modules in series and a flow path

34
length of approximately 4 m (13 ft). Therefore, four modules (two in
series and in parallel with two others) were used in obtaining the data
of Figure 53.

It would be useful to be able to test only one collector module and then
apply a correction technique to the data in order to predict the perfor-
mance of an array consisting of two or more collectors in series. This
would simplify the testing of air heaters. Two techniques are suggested
here for doing this. In both techniques, the single module must be
tested at a flow rate that will result in the same flow velocity which
will occur in the array of collector modules in series. This will insure
the same Reynolds number, same convective heat transfer coefficient, and
hence same F'. As an example, if one collector module were being tested
and two are normally installed in series, by equation (9), the flow rate
per unit collector area to be used in the test would be twice the value
to be used in the actual installation in order to have the same Reynolds
number and the same flow velocity.

Technique No. 1*:

After the test on the single module is completed, the y intercept

of the efficiency curve is determined and set equal to Fj^(Ta)g.
By separate measurement or calculation, (xa)^ is determined for the
cover plate assembly. is then calculated. In a similar fashion,
the slope of the efficiency curve is set equal to Fg^Uj^ and then
Uj^ is calculated from knowing F^, From the above information and
knowing mc for the test, equation (7) is used to compute F'.
Assuming F*^ and Uj^ to be the same for the collector modules in
series, the new value of mass flow rate per unit collector area is
used to calculate the value of Fj^ applicable for the modules in
series from equation (7). The new efficiency curve is then con-
structed from the value of Fj^, (Ta)^, and Uj^.

Technique No. 2:

A correction factor is applied to the test data on the single module.

The correction factor is derived as follows. Equation (1) governs
the performance of the single collector. For two collectors in series

#z [F^Kxa)
R e
F U (t,
R L
,
f ,i

(10)

*Suggested by Bruce Cole-Appel, Solaron Corporation, Denver, Colorado.

35
 where

~ the exit fluid temperature of the first collector

Q l
* '
equal to the inlet fluid temperature of the second, °C

Solving equation (1) for the exit fluid temperature of the first
collector,

'f,e,l = ^f,i +-A_ [F Kta)

' - F U (t^ - tj] ^^^^
mCp R ' ' e R f ,i

Substituting equation (11) into equation (10) and rearranging.

1 AFrUt
q^ = A[FRl(Ta)e " W^f.i " ta)] [1 " ^ ^-^]
mCp ^"-^^

The single module is tested and its efficiency curve determined. Then a
new efficiency curve is established by correcting the collector output
for the single module by the multiplying factor

[1
i
- i *
2
mCp

In a similar fashion, the correction factor for applying the single

module test results to three modules in series can be shown to be,

AFrUt AFrUt 2
1 - _JLJi -f
1
1 (
R L)
[ ]

mcp i^Cp

and to four modules.

^ AFrUt AFrUt t AFrUt

mc mCnp mc
""-p
P

The second technique above was verified by conducting an experiment on

one-half of the four module array of collector no, 6, Two of the
modules were connected in parallel and efficiency data were gathered
for a flow rate per unit collector area of 0,02 va.^/s,Ta.^ (4 ft /min, ft^)
Therefore, the flow path length was only one-half of that used when
gathering the data shown in Figure 53 but the flow rate per unit area
was twice the value of 0,01 m /s.m (2 ft /min, ft^) used for those

36
tests. The results of these second tests and comparisons were as
follows
/

Corrected to
Two Module Four Module Four Module
Test Performance Test, Figure 53

0.52 0.46 0.46

(W/m2.°C) 5.61 4.95 4.69

The data from the two tests are shown together in Figure 56 along with
additional data for the two collector modules in parallel tested at a
flow rate of only 0.01 m /s*m^ (2 ft /min'ft ), indicated by the data
in triangles and the long-short dash curve.

The discrepancy noted above between the heat loss factor, Fj^U^, for the
four module array and the corrected heat loss factor from the test on
the two modules in parallel at twice the flow rate per unit area could
have been due to the different leakage rates which occured during the
two different tests.

Incident Angle Modifier Figure 57 shows the results of the incident

:

angle modifier test being run on the air heater using both of the tech-
niques that were used with the liquid-heating collectors. As can be
seen, there was a large difference in the results using the two dif-
ferent procedures. The most reasonable result was obtained by keeping
the collector stationary, facing south and using the efficiency data
all through one test day to obtain the incident angle modifier values
indicated by the triangular data points in Figure 57. Notice the large
difference between the morning and afternoon values. This is consistent
with the large time constant of the collector and is the same character-
istic that was observed for collector no. 4 and shown in Figure 46.
The second technique of orienting the test stand to the north and tilting
the collector so that a specific incident angle occurs across solar noon
gave completely erroneous results as indicated by the circular data
points in Figure 57. This is not considered an acceptable procedure
for collectors with large time constants.

4.3 ALL-DAY SOLAR COLLECTOR EFFICIENCY

Based on the tests run according to ASHRAE Standard 93-77, it is possible

to predict the all-day performance of the solar collector. In addition
to having the two curves of near-normal-incidence efficiency and incident
angle modifier, the following must be known as a function of time through-
out the day:

1. inlet fluid temperature to the collector, t^ .

37
/ 2. ambient temperature, t„, and

3. incident solar radiation in the plane of the collector, I

The step by step procedure in Table 8 can then be carried out, A simi-
lar procedure has been used and reported by Zerlaut, Dokos, and Heiskell
[25].

Three separate all-day tests were conducted on collectors no. 2 and no. 3
in which the all-day efficiency was measured. The test days were com-
pletely clear with steady insolation. The above procedure was used to
compute the all-day efficiency using the same test conditions ( and I)
and the incident angle modifier and near-normal-incidence efficiency curves
previously determined. The results are given in Table 9 and a comparison
of instantaneous efficiency throughout one of the test days for collector
no. 3 is shown in Figure 58. As can be seen, the agreement is very good.

The all-day collector performance can of course be determined by experi-

ment as was done above. Many researchers over the past several years have
proposed that this would be a preferable way to collect data and report
performance instead of performing separate tests to determine the near-
normal-incidence efficiency and the incident angle modifier. Indeed, if
a large amount of daily efficiency data were collected for all types of
weather conditions and incident angle variations over perhaps a complete
year, then a plot of daily efficiency versus daily At/I would give a good
estimate of the expected long-term performance of the collector in
actual operation. This concept has been proposed by Tleimat, Howe, and
Buckland [26] and has been successfully used by then to characterize the
performance of solar stills. However, in solar collector testing, data
are usually collected over as short a period as possible and only a few
daily data points could lead to erroneous performance predictions.

It should be recognized that on a plot of daily efficiency versus daily

At/ 1, the y intercept of the plot is:

and not just

FR(^«)e,n

where day
f^T ^dT (13)
day
^
J ^dx

Consequently, the resulting curve is only valid for days in which the
incident angle variation, percent diffuse radiation, and hence K ^ are
approximately the same as on the test days, Simon and Miller [2^^ 28]

38
have used a technique for adjusting all-day collector efficiency obtained
outdoors to an effective efficiency value at normal incidence, in order
to compare the test data with data obtained on the same collector indoors
using a solar simulator. In essence, the all-day efficiency values and
th e da ily average value s of At/ I were both divided by a calculated value
of K^^ before plotting. K^^^ was calculated by knowing the incident
angle variations throughout the test days as well as the curve of
versus 0 from previous tests*. The resulting comparison with the inaoor
near-normal-incidence data was excellent.

Figure 59 shows data taken over several days for collector no. 6, the
air heater, and shows the kind of scatter that can exist when data other
than those collected on clear sunny days is used. The maximum integration
period used on any of the data in Figure 59 was four hours. Since this
was a flat-plate collector tilted so that the incident angle was approx-
imately 0 at solar noon and the data were all taken symmetrically with
solar noon, the maximum incident angle for any of the data was approx-
imately 30° and hence K^^^ ~ 1. Even though the scatter is relatively
large, a mean curve through a large amount of such data should result
in a reasonable performance curve for the collector in actual operation.

*The actual technique was slightly more complicated since Simon and Miller
also accounted for the fact that the curve of K^^ was determined indoors
using the simulator and hence under 100% direct radiation and the outdoor
data was taken under condition of up to 50% diffuse radiation.
39
5. SUMMARY AND CONCLUSIONS

This report has briefly summarized the test procedure adopted In

February 1977 by the American Society of Heating, Refrigerating, and
Air Conditioning Engineers and known as ASHRAE Standard 93-77, "Method
of Testing to Determine the Thermal Performance of Solar Collectors."
A testing facility has been constructed at the National Bureau of Stan-
dards In Galthersbur g, Maryland, in accordance with this Standard. It
consists of two test loops for the liquid-heating collectors and one
for air-heating collectors. During 1976 and 1977, six different col-
lectors were tested outdoors in accordance with the Standard or slight
variations of the procedures in the Standard. The main conclusions
of the study are:

1, The three primary thermal tests in the Standard, the time constant,
near-normal-incidence efficiency, and incident angle modifier tests

41
 I

can be carried out with little difficulty on flat-plate liquid-

heating and air-heating collectors. The three day pre-conditioning
test was not evaluated during this study for any collectors,

2. The time constant and near-normal-incidence efficiency tests can be

carried out on an evacuated tubular water-heating collector and on a
tracking, concentrating collector having a low concentration ratio
with little difficulty. For collectors with large time constants
(on the order of 20-30 min.) such as the evacuated tubular collector,
the near-normal-incidence efficiency curve is more difficult to
obtain because of the fewer test points that can be obtained during
a single test day due to problems of stability and larger test
intervals

3. The incident angle modifier test as currently specified in Sec-

tion 8.3.3 of the Standard is only applicable for flat-plate
collectors. An entirely different procedure is required for other
types of collectors and it will vary depending upon the optical
characteristics of the collector. For example, a single-axis tracking
collector should have the incident angle modifier determined as a
function of the angle between the outward drawn normal to the plane
of the collector aperture and the direct solar beam in a direction
parallel to the axis of the collector. In contrast, a stationary
evacuated tubular collector should have two incident angle modifier
curves determined; one as a function of the incident angle in a
direction parallel to the tubes and one perpendicular to the tubes.

4. The procedure for conducting the time-constant test in Section 8.3.1,

Method (1) of the Standard, which involves suddenly shading the col-
lector during an outdoor test, is much easier to complete and hence
preferable to Method (2), which involves suddenly changing the enter-
ing fluid temperature during an indoor test.

5. The procedure for conducting the incident angle modifier test in

Section 8.3.3, Method (2) of the Standard, which involves keeping
the test stand stationary facing due south and gathering all the
data in one all-day test, is preferable to Method (1), which requires
the test stand to be rotated and data collected across solar noon.
Method (1) requires several test days to complete and can produce
erroneous results for collectors with large time constants (10-30
minutes)

6. A class 1 pyr anome ter> as classified by the World Me terological Organ-

ization [7]jis required to insure reasonably accurate test results.

7. It is necessary to collect thermal efficiency data symme triccilly with

respect to solar noon in order to eliminate collector-heat-capacity
effects and to insure an accurate near-normal-incidence efficiency
curve

42
Since the near-no rmal-lncldence efficiency data is always collected
on clear sunny days with the incident solar radiation in the plane
of the collector near 945 W/m^ (300 Btu/h'ft^), it is not necessary
to allow data to be taken at levels as low as 630 W/m^ (200 Btu/h'ft^)
as currently specified in Section 5.1.4 of the Standard. Such low
levels may result in additional unnecessary scatter in the data.

The definition of gross collector area and aperture area in the Stan-
dard allows too much interpretation to be made by the testing organi-
zation and consequently the possibility for error and misuse. In
addition, the aperture or absorber area should be used in the compu-
tation of collector efficiency since it is more closely related to
the performance of the collector than gross area. When gross area is
used, the y intercept and slope of the near-normal-incidence efficiency
A , . A
curve are _a Fj^(Ta)^ and aF^^U^, respectively. The subscripts a and g

refer to aperture and gross, respectively. This fact must be realized

when relating the experimental data to the theoretical characteristics
of the collector as described in [14]. This possible confusion would
be eliminated if aperture area were used.

In the testing of air heaters to obtain efficiency values, care should

be taken to eliminate air leakage in the test loop and to properly
handle air leakage in the collector. If the collector is tested under
positive pressure, the air flow rate should be measured downstream of
the collector. If the collector is tested under negative pressure,
the air flow rate should be measured on both sides of the collector.

If a single module of an air heater is tested and the collector is

normally installed with two or more modules in series, it is possible
to adjust the test results to predict the performance of the field-
installed array by techniques outlined in this report.

The all-day collector performance of flat-plate collectors on clear

sunny days can be predicted with reasonable accuracy using the
test results obtained in accordance with the Standard.

The turbine flow meter used in one of the test loops for water-heating
collectors was found to drift in calibration by 2.5% over a period
of one year.

If any significant amount of collector testing is to be done, an

automatic data logger with a magnetic tape drive recording system
is preferable to the use of strip chart recorders and hand recording
of data.

An accurate measurement of ambient air temperature is just as impor-

tant as an accurate measurement of the transfer fluid entering the
collector. Both affect the abscissa value on the plot of near-normal-
incidence efficiency in the same way.

43
16. An open loop is preferable to a closed loop in testing air heaters,
since it's easier to maintain steady inlet temperature to the
collector.

17. The technique of determining the fraction of the incident solar radi-
ation which is diffuse by periodically shading the pyranometer gave
results comparable vd.thin 2% to using a pyheliometer and obtaining
the diffuse fraction by deduction.

18. For flat-plate collectors having a time constant of less than five
minutes, the three primary thermal tests can be completed in six
complete test days using a stationary south-facing test stand. One
each for the time constant and incident angle modifier tests and four
to obtain the near-normal-incident efficiency data. It may be
possible to reduce the time to four or five days if the time con-
stant and incident angle modifier data are taken on the same day.

19. The use of a solar simulator as specified in Section 7.3 of the

Standard was not evaluated during this study.

20. The use of integrators is not essential to obtain time averages of

variables such as fluid flow rate, t across the collector, and
incident solar radiation when determining the near-no rraal-incidence
efficiency. Across solar noon, each of the variables typically
changes by less than 2% over a 1-minute interval and because
of linearity of variation, the arithmetic average of sampled data
agrees to within 1% with continuously monitored and integrated
data. This does not imply that the variable should not be contin-
uously monitored by strip chart recorders.

44
6. RECOMMENDATIONS FOR FUTURE WORK

Based on the results of this study and consultation with researchers

around the world doing solar collector testing, the following recom-
mendations are made for future work to improve the state-of-the-art
in collector testing:

1. A general procedure for establishing the incident angle modifier

curve(s) for a variety of non-flat-plate collectors should be
established

2. A method or procedure should be developed for using the time con-

stant as experimentally determined in accordance with the Standard
for design purposes.

3. The feasibility of measuring air flow rate both upstream and down-
stream of an air collector during test to characterize its leakage
rate should be determined,

4. The feasibility of adopting the Standard to a wide range of concen-

trating collectors typically used in building applications should
be examined. A major part of this would be accomplished in 1. above,

5. The relationship between long-term collector performance exposed to

a wide range of weather conditions and the performance indicated
by the tests of the Standard should be determined,

6. It would be desirable to obtain the same performance parameters for

the collector as is presently obtained by using a combination of
indoor and outdoor tests (without the use of a solar simulator).
This would reduce the time and costs of performing the tests, partic-
ularly in the less sunny parts of the U.S. Three different organiza-
tions have proposed the use of an indoor test and pumping hot water
through the collector to determine its heat loss characteristics
[29,30,31]. This would be an alternate way of establishing the slope
of the near-normal-incidence efficiency curve, thus requiring only
a short outdoor test to determine the ordinate intercept. This
technique should be evaluated to determine if an accurate efficiency
curve can be established. Work has already begun as a result of the
procedure in [31] being used in a round robin testing program being
conducted by the International Energy Agency (lEA) in 1978,

7. Because of the allowable range of environmental conditions in the

Standard during the test to establish the near-normal-incidence
efficiency curve, considerable scatter is possible in the data, A
correction technique has recently been demonstrated [5,6] whereby each
data point can be corrected to "standard environmental conditions"
using a mathematical model of the collector. This is a potentially
powerful tool for outdoor testing of collectors and should be experi-
mentally verified and then "packaged" as a set of nomographs or com-
puter program.

45
 8, Several alternate testing configurations have recently been proposed
[32,33] that might enable the solar collector thermal output to be
determined more accurately than is now possible by the separate
measurements of m, Cp, and ^ - ^ , They basically involve
inserting an electrical heater in the test loop with a measurement
of At across the heater and eliminating the flow meter and a need
for knowing the specific heat of the transfer fluid. The accuracy
of this approach should be determined.

9. The liquid-heating collectors are frequently tested with one fluid

and one flow rate and used in actual installations with another fluid
and/ or flow rate. A correction technique for predicting performance
under other than test conditions needs to be developed from first
principles and verified for use in the field.

10. The potential for establishing a more accurate near-normal-incidence

efficiency curve by testing the collector side by side with a "stan-
dard collector" whose characteristics are well known should be inves-
tigated.

11. Determination and evaluation of a standard instrument and test procedure

for determination of effective sky temperature should be made.

12. For air-heating collectors, an additional test should be developed

which will result in the determination of how collector performance
varies with transfer fluid flow rate. This would enable a designer
to more easily select the optimum flow rate for his system.

46
 7. REFERENCES

1. Hill, J.E. and T. Kusuda, "Methods of Testing for Rating Solar

Collectors Based on Thermal Performance," NBSIR 74-635 December,
,

1974.

2. Hill, J.E. Streed, E.R. Kelly, G.E.

, , ,Geist, J.C., and T. Kusuda,
"Development of Proposed Standards for Testing Solar Collectors and
Thermal Storage Devices," NBS Technical Note 899 February, 1976.
,

3. Hill, J.E., and E.R. Streed, "A Method of Testing for Rating Solar
Collectors Based on Thermal Performance," Solar Energy, Vol. 18,
No. 5, 1976.

4. "Methods of Testing to Determine the Thermal Performance of Solar

Collectors," ASHRAE STANDARD 93-77, ASHRAE, 345 East 47th Street,
New York, New York, 10017, 1977.

5. Streed, E.R. Thomas, W.C., Dawson, A.G.

, Wood, B.D.
, and J.E. Hill,
,

"Results and Analysis of a Round Robin Test Program for Liquid-Heating

Flat-Plate Solar Collectors," NBS Technical Note 975 Augustjl978.
,

6. Streed, E.R., Thomas, W.C., Dawson, A.G., Wood, B.D., and J.E. Hill,
"Results and Analysis of a Round Robin Test Program for Liquid Heating
Flat-Plate Solar Collectors," Solar Energy , accepted for publication,
1978.

7. Guide to Meteorological Instrumentation and Observing Practices ,

Secretariet of the World Meteorological Organization, Geneva,
Switzerland, 4th Edition, 1971.

8. Klein, S.A, , Beckman, W.A. , and J. A. Duffie, "A Design Procedure for
Solar Heating Systems," Solar Energy , Vol. 18, No. 2, 1976.

9. Klein, S.A. , Beckman, W.A,, and J. A. Duffie, "A Design Procedure for
Solar Air Heating Systems," Solar Energy , Vol. 19, No. 5, 197 7.

10, Yass, K. and H. B. Curtis, "Low-Cost Air Mass 2 Solar Simulator,"

,

NASA TM X-3509 presented at the United States Section Meeting of the

,

International Solar Energy Society, held in Cleveland, Ohio, October

3-4, 1973.

11, Simon, F,F., and P. Harlamert, "Flat-Plate Collector Performance

Evaluation, the Case for a Solar Simulator Approach," NASA TM X-71427 ,

presented at the United States Section Meeting of the International

Solar Energy Society, held in Cleveland, Ohio, October 3-4, 1973.

12, Ramsey, J.W., Borzoni, J.T. , and T.H. Holland, "Development of Flat-
Plate Collectors for Heating and Cooling of Buildings," NASA CR-13804
June, 1975.

47
13. Klein, S.A. , Duffle, J. A., and W.A. Beckman, "Transient Considerations
of Flat-Plate Solar Collectors," ASME Journal of Engineering Power ,
Vol. 96A, p. 109, 1974.

14. Duffie, J. A., and W.A. Beckman, Solar Energy Thermal Processes , John
Wiley and Sons, 1974.

15. Simon, F.F., "Flat-Plate Solar-Collector Performance Evaluation with

a Solar Simulator as a Basis for Collector Selection and Performance
Prediction," NASA TMX-71793 , presented at the 1975 International Solar
Energy Society Meeting, held in Los Angeles, California, July 29-
August 3, 1975.

16. Wi jeysundera, N.W., "Response Time of Solar Collectors," Solar Energy ,

Vol. 18, pp. 65-68, 1976.

17. Simon, F.F., and E.H. Buyco, "Outdoor Flat-Plate Collector Performance
Prediction from Solar Simulator Test Data," NASA TM-71707 presented
,

at the 10th AIAA Thermal Physics Conference, held in Denver, Colorado,

May 27-29, 1976.

18. "Solar Thermal Electric Power Systems-First Quarter Progress Report,"

Report No. NSF/RANN/SE/GI - 37815/PR/74/1, Colorado State University,
April, 1974.

19. Flowers, E. "Test and Evaluation of the Performance of Solar Radia-

,

tion Sensors at Inclinations from the Horizontal Under Laboratory

and Field Conditions," Department of Energy Report DSE/1041-1,
September, 1977.

20. Pendleton, R.L. "Evaluating a Solar Energy Concentrator," ASHRAE

,

Journal pp. 47-50, November, 1976.

,

21. Thomas, W.C., "Thermal Efficiency Test Procedures and Results for a
Cyclindrical Concentrating Collector," Proceedings of the ERDA
Concentrating Solar Collector Conference held in Atlanta, Georgia,
,

September 26-28, 1977.

22. Johnson, S.M., "Standardized Performance Tests of Collectors of Solar

Thermal Energy-A Flat-Plate Copper Collector with Parallel Mylar
Stripping," NASA TM X-73553 November, 1976.
,

23. Close, D.J., and M.B. Yosoff "The Effects of Air Leaks on Solar
,

Air Collector Behaviour," Solar Energy Vol. 20, No. 6, pp. 459-
,

463, 1978.

24. Farber, E.A. and C.A. Morrison, "Clear Day Design Values,"
Chapter IV of Applications of Solar Energy for Heating and Cooling
of Buildings, ASHRAE Publication GRP 170, 1977.

48
25. Zerlaut, G.A., Dokos, W.T. and R.F. Heiskell, "The Use of ASHRAE
,

Standard 92>-ll in Predicting All-Day Performance of Flat-Plate

Collectors," Proceedings of the 1977 Flat-Plate Solar Collector
Conference held in Orlando, Florida, February 28-March 2, 1977,
,

26. Tleimat, B.N., Howe, E.D., and R.E. Buckland, "A Proposed Method
of Rating the Thermal Performance of Solar Collectors," presented
at the Joint Conference of the International Solar Energy Society
and the Solar Energy Society of Canada, held in Winnepeg, Canada,
August 15-20, 1976.

27. Simon, F.F., and Miller, D.M., "A Generalized Correlation of Experi-
mental Flat-Plate Collector Performance," NASA TM X-71832 presented
,

at the Conference on Standards for Solar Heating and Cooling, held

in Philadelphia, Pennsylvania, October 14-15, 1975.

28. Miller, R. "Performance Correlations of Five Solar Collectors Tested

,

Simultaneously Outdoors," NASA TM X-73546 presented at the Joint

,

Conference of the International Solar Energy Society and the Solar

Energy Society of Canada, held in Winnipeg, Canada, August 15-20, 1976.

29. Smith, C.C., and T. Weiss, "Design Application of the Hottel-Whiller-

Bliss Equation," presented at the 1975 International Congress, "Solar
Use Now-A Resource for People," held in Los Angeles, California,
July 28 - August 1, 1975.

30. Symons, J.G. , "The Direct Measurement of Heat Loss from Flat-Plate
Solar Collectors on an Indoor Testing Facility," CSIRO, Division
of Mechanical Engineering, Technical Report No. TR 7, 1976.

31. "Guidelines and Directions for Determining the Feasibility of Utilizing

Solar Collectors, A. Efficiency Tests of Solar Collectors," A Proposed
Method of BSE, The German Solar Energy Industries Association,
September, 1976.

32. Carroll, J. and D. Wald, "Improved Instrumentation for Performance

Testing of Solar Collectors," unpublished note, Santa Clara Solar
Research Institute, Santa Clara, California, 1978.

33. Reed, K. A., "Interim Solar Collector Test Plan," Proceedings of the
ERDA Concentrating Solar Collector Conference held in Atlanta,
,

Georgia, September 26-28, 1977.

34. Kusuda T. ,
"NBSLD, The Computer Program for Heating and Cooling Loads
in Buildings," NBS Building Science Series 69 ,
July, 197 6.

35. Buchberg, H. and J.L. Cairns, "A Convective Conductance Meter:

, Design
and Evaluation," ISA Transactions, Vol. 10, No. 3, pp. 325-330, 1971.

49
36. Ito, N. Kumura K., and J. Oko, "A Field Experiment Study on the
Convection Heat Transfer Coefficient on Exterior Surface of a
Building," ASHRAE Transactions Vol. 78, Part I, 1972.
,

37. Aagard R.L., "Convection Free Instrument for Measuring Infrared

Radiation in the Atmosphere," The Review of Scientific Instruments ,
Vol. 29, No. 11, pp. 1011-1016, November, 1958.

38. Beckman, W.A., et. al,, "Units and Symbols in Solar Energy",
Solar Energy , Vol. 21, pp. 65-68, 1978.

50
Table 1. Specifications for the Equipment and Sensors Used in the
Liquid Collector Test Loops

Equipment/ Sensor Specifications

Temperature measurement Outer shell assembly is constructed from

section/ temperature 1.27 cm (0.5 in.) i.d., 0.317 cm (0.125
well in.) wall hard-drawn copper tubing and
fittings; 0.952 cm (0.375 in.) o.d.,
0.081 cm (0.032 in.), wall capped brass
tube provides for a thermopile well
while a 0.474 cm (0.187 in.) compres-
sion fitting with teflon ferrule allows
for insertion of sheathed temperature
probes; flow rate and flow configuration
produces turbulent flow and fluid swirl-
ing after passing through one right angle
bend (See Figure 15.)

Water-to-air heat Vega automative radiator; effective

exchanger surface area of 1.85 m^ (20 ft^);
value of 181.5 W/(m^*°C)
(32 Btu/(h*ft^»°F))

Water-to-water heat Single-pass counter-flow heat exchanger

exchanger

Storage tank 37.81 (10 gal.) glass-lined domestic

hot water tank; 1500 W (5118 Btu/h)
immersion heater

Pressure relief valve 2.06 X 10-^ Pa (30 psi) spring-loaded

valve

Filter 5 micron cartridge filter for hot

water applications

Electric Heater 500 W (1705 Btu/h), rod resistance

immersion heater

Temperature Controller Adjustable proportional set point

controller; 1000 W (3410 Btu/h) load
capability using a triac assembly
and a type-T thermocouple sensor

51
Table 1 (Cont.)

Equipment/ Sensor Specifications

Pump Self-priming, eccentric-disc positive-

displacement pump; variable pumping
capacity of 0-3.45 x 10"'^ m^/s (0-5 gal./
min.), forward or reverse and a static
pressure head of 3.44 x 10 Pa (50 psi)

Flow meters Turbine meter; Paddle-wheel type incor-

porating an orifice upstream and down-
stream providing a linear span of
0-6.3 X 10"^ m^/sec (0-1 gal. /min.)
using ball bearings and a magnetic pick-
up; specially linerized to within + 0.9%
over the intended ranges of temperature
and flow rates

Positive Displacement meter; oval-gear

positive-displacement incorporating an
active RF pickup functional over the flow
range of 1.07 x 10"^ to 6.30 x 10"^ m^/s
(0,17 to 1 gal. /min.)

Temperature sensors 100 ohm, 3 wire platinum resistance

thermometers interfaced with linear
bridge amplifiers for a 1 mv/'C output

Thermocouples, type-T, 24 gauge incor-

porating stainless steel or copper
sheathing with and without grounded
thermocouple junctions

Temperature difference Platinum resistance thermometers - same

sensors as for absolute temperature measurement

Thermopile, type-T, 6 junctions formed

from 30 gauge wire with each junction
individually insulated with shrink
tubing

Pressure drop sensor Inverted U-tube manometer; 0.635 cm

(0.25 in.) i.d. and 25.4 cm (10 in.)
long filled with water and allowing
the upper U to be trapped with air

52
Table 2. Specifications for the Equipment and Sensors Used
in the Air Collector Test Loop

Equipment/ Sensor Specification

Nozzle Aluminum ASME long-radius nozzle;

factory-calibrated and traceable to
NBS standards; nozzle diameters of
5.1, 6.4, 7.6, and 10.2 cm (2.0,
2.5, 3.0, 4.0 in.)

Blower Industrial centrifugal blower; rated

at 0.47 m-^ (1000 ft^/min.) at 2500 Pa
(10 in. H2O); 746 W 2500 ( 3 hp)
electric motor

Duct heaters On/off; 39 kW in 3 kW increments

Proportionally-controlled; 6 kW

Temperature sensors Thermocouples laboratory-fabricated

;

from premium-grade type-T 24 gauge

thermocouple wire

Temperature difference Six junction thermopile; laboratory

sensors fabricated from premium-grade type-T
24 gauge thermocouple wire

Pressure sensors Nozzle pressure drop; 0 to 1000 Pa

(0 to 4 in. H2O) inclined manometer;
5 P^ (0.02 in.H20) smallest scale
division; also 0 to 1250 Pa (0 to
5 in. H2O) elastic diaphram pressure
transducer

Nozzle gauge pressure; 0 to 2500 Pa

(0 to 10 in. H2O) vertical manometer;
25 Pa (0.10 in. H2O) smallest scale
division

Pitot tube pressure; 0 to 1250 Pa

(0 to 5 in. H2O) vertical manometer;
5 Pa (0.02 in. H2O) smallest scale
division

53
Table 3. Specifications for the Instrumentation Used
to Make the Meteorlogical Measurements

Measurement Specification

Ambient air Calibrated type-T, 24 gauge thermo-

temperature couple or precision platinum resis-
tance thermometer housed within a
vented weather shelter

Wind speed Standard 3-cup wind anemometer incor-

porating a d.c. generator; output
of 0.0581 volts/m/s (0.026 volts/raph)
directly proportional to wind velocity

Wind direction Direction vane with a two-wiper

potentiometer (0-540°)

Total solar radiation Precision spectral pyranometer

incident on the utilizing an all black thermopile
collector tilted detector and temperature compensation;
surface class 1 pyranometer as classified by
the World Meteorological Organization

Diffuse component of 10 cm (3.93 in.) shadow disc attached

the solar radiation to a 1 m (3.28 ft) long support rod
incident on the
collector tilted
surf ace

Direct beam solar Normal incident pyrheliometer with

radiation a collimated view of 5.7° and an
automatic tracker

Sky temperature Pyrgeoraeter; precision infrared

radiometer capable of undirectional
measurement of either incoming or
outgoing long wave terrestrial
radiation; a modification of the
precision spectral pyranometer
using a silicon hemisphere with a
transraittance of 0.50 from 0.35
to 50 ym

Incident angle between Shadow protractor; see Figure 27

sun's direct beam and accompanying description
and the outward
drawn normal to
the plane of the
collector aperture

54
 o
03 3 00

c c
o o
•H •H
U 00 4J
rt <f
C •H
H
1

00 >
1-1 o
CJ CM Q
c CN
h-l

U 00
OJ c
4-1 •H
<u T3
6
o a)
c c 03
3
oo
n)
a.
?^
eu 0-.
e
a
3

r-i O
00 >
+ +

(U

55
Table 5. Specifications for Strip-Chart Recorders, Integrators,
and the Data Logger

Strip-Chart Recorders

2 Pen - Multi Range Input Span

Uncertainty: ±0,25% of span
Input Impedance: 0 - 5 V, 30 megohms; 5 V - 10 V, 2.5 megohms
Time Constant: 0.5s
Input Spans: 0.1 mV - 10 V, multiple range
Type Inputs Monitored: thermopiles, pyranometers flow rates
,

2 Pen - Fix Input Span

Uncertainty: +0.05% of reading
Input Impedance: 5 megohms
Time Constant: 0.75s
Input Span: 0 - 5 mV
Type Inputs Monitored: referenced thermocouples

3 Pen - Fixed Input Spans

Uncertainty: +0.5% of span
Input Impedance: 5 megohms
Time Constant: Is
Input Spans: 0 - 5 mV, 0 - 100 mV, 0 - 5 V
Type Inputs Monitored: thermopiles, flow rates, resistance
thermometers

Integrators

Uncertainty: +0.5% of reading, +2 digits/h

Input Impedance: 1 megohm
Input Span: 0 - 30 mV
Type Inputs Monitored: thermopiles, pyranometers

Data Logger

Uncertainty: +0.01 mV, Max Error: 0.005% FS, ±0.2 °C

Input Imedance: 100 megohms
Input Ranges: 0 - 400 mV type-T thermocouples
;

Type Inputs Monitored: All input data thermocouples, 0 - 100 "C;

analog signals, 0 - 400 mV
Special Features: 16 channel arithmetic averager with separate
averaging time base
Magnetic Tape Drive: Incremental tape drive, 800 bpi, 9 track with
a DTL/TTL plus voltage true interface

56
 4J
X) tn a II d ••> 1 X) a
01 01 a* 0) M
-d 4J X u-i u 0) 0 CJ 0 01 0
0)
4-1
c
3 a
o 3
+J
01
X
0 (U
4-1
0
to 0) 0 C
0 0
4-1 cn
l-l
a
to
a 4J
0 tn a
c O •H 3 rH tu 0. X 0 0) 0 cn H 1-1 <u
a
1-1 X 00 u a in 4-1 H 4-1 1^ CU
to 00 tH 0 0 a
II

to
cx <u
o
0)
X T3
01
II

X
1-1
-d
•-
a
c
01
II c 0
•H 0
CJ
01
tH
4-1

CJ II 0 H
00
c:
?X X
tH 4-1
l-l

(U
-C3

a
H
OJ
01 to a in u 0 <u in r-l a; >4-l to >-l 0 12
4-1 IH o 0) O 00 CU 3 CU U 00 0 tH OJ CJ 01 0 C
o X3 CI4 c X X3 c 0 C
•H l-l rH 4-1 to rH to l-l cn sO to
X
X X a X
3 X 0) 3 • 0) M -H 0 3 >, iH <U
3 cn tH 0) rH 4-1 in X 4-1 H 4-1 CN 0 4-1 to to X CO iH CJ

T3 M cn u (U
(U T3
O4) to 0) c <u d x tn
J3 •H to
00 O
T3 o >, ^
H o 3 to
c •H OO •r-l 1-1 4J -H 4-1 to
o «-l •H XI t4-4 in 4-1 •H
> a
tH 4-1 ^
•H a 0 l-l a a to CO T3 00
0 to O 1 u u V) 0 l4 rH to tn iH C b
o fl 05 •H <u CO OJ 3 0 tn 3 •H a 01
rH
3
CO a jQ o <9 tNl J3 0) to tn x; 0 X
PQ tH (U •H 1-1 •H C c r-< C o; iH
Ul 00 00 (A >4-l 00 ro 14-1 1-1 1-1 00 1-1 X U-l

MC
1

1-1 (C
4-1 (I) C tn
c > 0
n) •H rH
to a
01
0)
tj
1 1
4-1 00 00 (U 1 1 ij to f-l 1

d (U 0) tj c tj 0) <u to a. l-l 01 0)
•H iH m .H Ul <U •H OJ to .H to 3 to to to c CO
N XI Ul XI U5 4-1 (U <4-l OO tfl 00 m o 0) tn tn (0
« 3 n) 3 to U-l to l-l u c to c to to l-l C 01 C 3 cd
O iH o 7-1 01 o 3 •H •H > •H •H l-l 01 O iH
O T3 00 T3 OO V4 o 4-1 to I/l 00 tfl 00 0) to r-l H T3 &0

^O I

CJ 0)
0) Vi CO OJ CJ
•> c 4-) 0) tH tH 10

HO. X
1-1 Q< CU tM
§Co to
CO
in IJ

X a
O. O - 3 01
c CJ 00 01 tn - I c « 4J CJ l-l I iJ CJ
iH ^ H C 0 rH ^ OJ CO to CU CJ CO
X X
I

a tH o OJ iH O OJ CU CJ 01 CO 01 UH 13, C 01 t4-l

3 tH (0 OJ 4-1 l-l > 01 to CO H l-l Q. CO -H l-l 0) -H

tH O tH 4-1 1-1 3 X 1-1 •U rH CU H OJ 3 O rH
X q; 3 4J (d
to >-l X CO 3 4-1 CJ 4J cn X in 00 cn CO c; cn cn (1) O.

org
CO 4J 0
in
0 CU
rH CO
- 00 vO
ON
00
rH
m
01
0 rH 0) •
0 M rH rH rH cn
CJ <

OJ
1-KSI
3 a
4-1 rH
U CO vD
CU 01
a u rH
< <

tn
C T3
CO -H
U 3
H -H iH
(d
u
to
0)
H

.H ZO
o

57
Table 7. Time Constants for the Liquid-Heating Collectors Tested

Transfer Fluid
Flow Rate Time Constant
Collector No, m /s min.

1 3.2 X 10-5 1.7

2 2.8 X 10-5 1.6

3 3.5 X 10-5 1.8

4 3.8 X 10-5 20.3

5 5.0 X 10-5 0.9

58
 Table 8. Computation of All-Day Solar Collector Efficiency

Calculation Steps Hour of the Day, Solar Time"'-

Daily
6-7 7-8 4-5 5-6 Total

1. Inlet fluid temperature to the

collector, t^ °C
. ,

2. Ambient air temperature, t^, °C

3. Incident solar radiation of a horizontal

surface, 1^^, W/m

4. Ratio of total incident solar radiation on the

tilted surface (normal to the
collector plane) to that on a
horizontal surface, R

5. Incident solar radiation on the collector

plane, I, W/m line 3 x line 4
,

6. Collector thermal efficiency at nor-

mal incidence determined in accor-
dance with Sections 8.3.2 and 8.5
of ASHRAE Standard 93-77 and using
data from lines 1, 2, and 5

7. Incident angle between the direct

solar beam and outward dr^wn normal
to the collector plane, 9

8. Incident angle modifier, determined

in accordance with Sections 8.3.3
and 8.6 of ASHRAE Standard 93-77
and using the value of 6 from line 7

9. Energy output from the collector, W/m 2 ,

line 5 x [line 6 + Fj^CTa)^ ^ x (K^^ - 1)]

10. Collector thermal efficiency,

line 9/line 5

to convert standard time to solar time,

see reference [14], pp. 18-19.

'to compute R, see reference [14], pp.

48-55.

'can be determined by direct measurement

using data from reference [24] or other
accepted methods.

^to compute 6, see reference [14], pp.

14-18.

59
Table 9. A Comparison of Measured and Predicted All-Day Collector Efficiency

Measured Predicted
Daily Daily Measured Predicted
tf,i-ta dayjid x Collected Collected Daily Daily
Energy Energy Efficiency Efficiency
Collector ''C kJ kJ kJ % %

2 21 33,400 18,500 19,200 55.4 57.5

2 52 34,800 15,800 15,840 45.4 45.5

2 64 34,100 15,800 14,900 46.3 43.7

3 19 41,700 28,300 27,000 67.8 64.7

3 52 42,600 22,100 21 , 300 51.9 50.0

3 64 44,100 18,800 19,800 42.6 44.9

60
APPENDIX A

Calculation of Air Flow Rate Using a Nozzle

and Pressure Difference Measurements

The air flow rate through the nozzle apparatus is calculated byi

Qmi = 1-^1 (^Pn V^-^ (1)

where

= measured air flow rate, m /s

Qjjjj^

= nozzle discharge coefficient

2
Aj^ = nozzle throat cross-sectional area, m

Apjj = static pressure difference across the nozzle. Pa

f -5

= specific volume of the air at the nozzle, m-^/ kg dry air

The nozzle discharge coefficient C^, is determined from the following

table

Reynolds Number Discharge Coefficient

%e
20,000 0.96

50,000 0.97

100,000 0.98

150,000 0.98

200,000 and above 0.99

Alternately, a value of C^^ can be calculated from the following equation

for 20,000 £Nj^g ^200, 000:
'Re
= 0.95 + 3.7 X 10-7 N^^ (1 - (2)
^-^fj^)

61
The nozzle Reynolds Number is calculated from:

NRe = ft ^a % (3)

where

V- = velocity of the air at the nozzle throat, m/s

Dj^ = nozzle throat diameter, m

= temperature factor for equation ( 2 ) above

determined from the following table:

Temperature °C Temperature Factor

-6.7 78275

+4.4 72075

+15.6 67425

+26.7 62775

+37.8 58125

+48.9 55025

+60.0 51925

+71.1 48825

can be determined from the velocity pressure measured at the nozzle

throat with a pitot tube or calculated from the following equation:

= 1.40 (Ap^ (4)

The nozzle throat cross-sectional area, Aj^, is determined by measuring

its diameter in four places approximately 45 degrees apart around the
nozzle in each of two planes; one at the exit and the other in the
straight section up in the nozzle. The nozzle pressure difference /^jg,
is measured as specified in the body of the report; with both an inclined
manometer as well as an electronic pressure transducer.

The specific volume of the air at the nozzle, u ^, is calculated from:

^ = 10.1 X 10^ un/[PN (1 + W^)] (5)

62
where

= specific volume of the air at the wet bulb and dry

bulb temperature existing at the nozzle but at
standard barometric pressure, m'^/kg dry air

= absolute pressure at the nozzle throat, Pa

Wj^ = humidity ratio of the air at the nozzle,

kg H20/kg dry air

Alternately, can be calculated for:

.
_ Ra % (1 + 1»608 W^ )
^
(6)

where

R„ = gas constant for dry air, (P„*m )/(kg'''K)

= absolute temperature of the air at the nozzle, °K

Both and W^^ are determined from a psychroraetic chart or simplified

psychrometic computer routines such as those on pp. 150a - 158a of
reference [34]. The wet bulb and dry bulb temperature are measured
together at only one point in the test loop; Wj^ is computed based
on those measurements, and it is assumed to be constant throughout
the loop. The absolute pressure at the nozzle throat is calculated
as the sum of the barometric pressure and the nozzle discharge gauge
pressure, p^. is measured as indicated in the body of the report
using a vertical manometer.

Once the air flow rate, Qj„-j^ is calculated, the following equation issued
>

to calculate the flow rate, Q in m /s at standard temperature and

mi s^ ^
.

,
pressure

Q . = Q ./(I. 2 u') (7)'

^mi , s ^mi' * n'^ ^

63
APPEtlDIX B

Description of a Radiometer to Determine Sky Temperature

If an electronically-heated plate mounted in a frame with the upper

surface exposed, is placed outdoors and allowed to reach steady-state
conditions, the following energy balance applies:

the rate the rate of

of electrical + incident solar
energy input radiation
to the heater absorbed by
the heater

the rate of the net rate of the rate of heat

heat transfer + heat transfer to + transfer through
to the ambient the ambient the edges and back
air by convection surroundings by of the mounting
radiation frame by conduction

In equation form:

^ + as I = h^(Tn - T;,) + eadn - To ) + % (1)

where
p =
— rate of electrical energy input to the heater per unit of surface
^ area, W/m^

cxg = absorptance of the electrical heater surface to the solar

radiation

I = incident solar radiation on the heater surface, W/m

h„C = convection heat transfer coefficient between the surface of the

9
heater and the ambient air, W/(m C)

Tp = surface temperature of the electrical heater, °K

T^ = ambient air temperature, °K

e = emittance of the surface of the electrical heater

a = Stef an-Boltzmann Constant, 5.6697 x 10"^ W/(m^ -"K)

Tg = effective sky temperature, °K

Qk = rate of heat transfer through the edges and back of the mounting
A frame by conduction, W/m^

64
Using the above principle, a number of researchers have built instruments
to determine the value of one or more of the variables in equation (1)
with the rest predetemined or eliminated, or measured.
, In some cases,
two identical electrical heaters have been built with all but one design
characteristic the same and by using the heaters or sensors side-by-side
exposed to the same environment, a simple relative measurement is used
to determine the value of an unknown variable. Buchberg and Cairns [3 5]
and Ito, Kimura, and Oka [36] have used the principle for determining
an experimental value of the convection coefficient, h, for air flow
over a surface of specific geometry. Aagard [37] adapted the principle
to the determination of effective sky temperature at night with I = 0.

The radiometer designed and built by the French Building Research Center
and one of the instruments being used by NBS to determine the effective
sky temperature during a collector test also makes use of the principle
of the balance expressed by equation (1). Two flat receivers are mounted
side by side on the top of an insulating slab. Each receiver is a
0.093 (1.0 ft^) square surface with a central area of 0.01 m^
(0.11 ft^) which is backed by a resistance heater. As can be seen in
Figure 60, the first receiver has it outer perimeter area covered with
a white surface of low absorptance and high emittance and the central
heater covered with a shiney surface of low absorptance and high emit-
tance. The second receiver is treated in just the opposite manner. A
star-shaped thermopile is mounted on each receiver to detect any temper-
ature difference between the central heated area and the perimeter surface
area. A special electronics package was designed for and is used with
the receivers. The electronics supplies electrical energy to the central
heater of either receiver in order to keep its surface temperature equal
to the surface temperature of the surrounding perimeter area. The elec-
tronics package also integrates the input electrical energy over either
a 10 minute or 30 minute period. Only one of the receivers is used at
a time and the particular one is selected automatically.

Writing equation (1) for both the central heated area as well as the
perimeter area of a given receiver, two equations in several unknowns
result. If, however, it can be assumed that the convection coefficient,
h, and the back losses are the same for both areas and recognizing
A~
that the surface temperatures of two areas are controlled to the same
value, then the two equations can be solved simultaneously for the
effective sky temperature, T^ as a function of the following variables
,

which are either known or are measured during the test:

asl» as2'

65
 CD
Z
X

67
 0)

c
01

s
}-(

o
4-t
o
0)

o
CO

o —
•H

•rH

C 1-J
o

00 CO
C -H
•H
M 13
M -iH
Q) 3
H rH
fa
o.
O i-i

0 <u

1 CO
C C
0) CO
a v-i

o H

CNI

fa

68
 (U
x:
+j •

c ^

O 3
*J O
a 3

c
o
CJ
TJ
0)
•1-1

rH
a-
a
3
cn
O
M-l CO
•H
c
o 73
•H C!
U
cd
i-i -a
•H
S) 3
•H cr
•H

ao CO
C -r-l

•H
4-1
CO •H
OJ 3
H -H
ft.
a,
o >-i

o <U
M-l
CO
c d
0) CO

o H

CO

a
•H
fa

69
 0.9 T
1 Glass cover plate

0.8 2 Glass
cover plates

0.7

0.6

0) 0.5
u
c
(B

o
« 0.4

u
c

0.3
I
(B

09

J 0.2

Glass extinction coefficient = 0.08/cm

Glass refractive index = 1.53

Absorber plate absorptance = 0.9

15 30 45 60

Angle of incidence.G, degrees

Figure 4 Effective Transmittance-Absorptance Product for Three Flat-Plate
Solar Collectors with Non-Selective Coatings on the Absorber [18]

70
 1.2

1.1

1 Glass cover plate

1.0

2 Glass
1^ 0.9 cover plates

0.8
E

CO

^ 0.7

'5

0.6

0.5

Glass extinction coefficient = 0.08/cm

0.4 -
Glass refractive index = 1.53

Absorber plate absorptance = 0.9

0.3 I \

15 30 45 60

Angle of incidence, 6, degrees

Figure 5 Incident Angle Modifier for Three Flat-Plate Solar Collectors with
Non-Selective Coatings on the Absorber.

71
 1.2

1.1

1 Glass cover plate

2 Glass cover plates

:! 0.7
03

0.6

0.5

Glass extinction coefficient = 0.08/cm

0.4
Glass refractive index = 1.53

Absorber plate absorptance = 0.9

0.3

0.2 0.4 0.6 0.8 1.0 1.2

1
-1
COS 6
Figure 6 Incident Angle Modifier for Three Flat-Plate Solar Collectors
with Non-Selective Coatings on the Absorber.

72
 i—

4—

1—1

PQ

GO
n3

O
4-)
c/^

•n't

o •
XI
c
tT3

.H
>,
>-i

»—
Ma
{2

GO
&0 u
C 3
•W
S-i

0) S-i

> 0)
o J=
o 4-1
•H
i-l

o o
o
Q CO
S-i pa
o
u
>
j-i

.H •H
CO

S-I

73
Solar Collector Testing Equipment Mounted on the Elevator and
Lowered to the Floor of the Underground Storage Bunker at the
NIKI Site, NBS, Gaithersbur g, Maryland.

74
 >
0)

C
3
O
o

13 •
0) t3
W C
•H CO
m rH
02 >^
U
00 cfl

C S
•H
0) »>

PQ 00
u

0) tfl

g
CX, 0)

a- 'H
W CO

00
O
C •>

CO
PQ
w 2
0)

0)
4J
o
4-1
a

-H l-l
O 2
CJ
OJ
U
<T)
X
4-1
-H
O J-)
CO cC

75
 0)

4-1

W
QJ
H
O
4-t

o
•H
5-1
P-i

c •

•SI
PQ

O u
CO
O
O CO
CO PQ

c •>

•H 0)
U J-l
CO -H
0) CO

•H

76
77
79
80
 TEMPERATURE MEASUREMENT SECTION

Figure 15 Schematic Diagram of the Temperature Measuring Stations in the

NBS Test Stand for Liquid-Heating Solar Collectors.

81
 o

Fid
88
CO i-(
l-i O
a,
B ^
CU CO

o
cd c/3

^1 £10
o c
4-1 -H
•u
d CO
O 0)
•H ffi
J-J I

CO 13
V-i -H

cr
H -H
M-l hJ
c
UO
>-i

O
<+-(

U (X
CO O
p: o
0)
4.) -u

<: 0)

CO CO

4-1 d
O -rH

s c
COO
t>0 4-1
CO CO
•H 4-1
Q CO

O 60
•H C
•M -H
CO >-l

S 3
(U CO
nC CO

CO s

(U

1^
=o

131 N

83
84
85
PQ
2;

01

•H

M
0)

CO
5-1

O
4-1
O
0)

OO

O
an
bO
d
•H
4-1

CO

5-1

o "
M-l M

4-) !-J

CO 0)

4J w
cn -H
0) nJ

o
CNI

0)

•H
 3
QNiiDna anaixaid a3ivnnsNi

w
u
o
o

00

CO
0)

u
•H
<C
(-1

o
M-i

c
(T3

e
CO

00
CO
•H
o

J-J

to
e
0)

o
CO

(U
(J

QNiiona 3"iaix3nj QBivmsNi 00

•H

87
•H
CO

B
o

£3
QJ
0)
W
CO
CO
c
CO CO
!-i

O >^
4-1 U
CJ
0)

UO
W)
M
3
!-i ^
CD
cO
rH S-l

O OJ
C/3 ^
4-1
00 -H
C CC
•H O
CO
01

u
o
14-1

c
CO

x:
+J 4-1

CO
01 4-1

H CO

•r-t
89
 13
C
CO

C/D

m
H
c/:i

pa

0)

c
•H

CO

O
CO
c
0)

CO
3 u
•u o
CO 4-1
CJ
(U <u
ex tH
B -H
(U o
H a
}-i u
O cO
4-1 >—
o

Pi
•H
4-1

CO CO

90
 00
c
•H
•u
CO
0)
EC
I

'O
•H
3
cr
•H
hJ

c
4-)
CO
4-)

CO
01
H
CO
pa
2;

G
cd

C
o
)-l

a)
•u
0)

o
d
CO
u •
p>~i cn

O
4-1
C a
CO <u

M rH
o O
O
4-1
o
OJ i-i

rH CO
rH rH
O O
CJ) CO

0)

•H

91
92
 3

a
t3
rrt

C2 m
OJ
d) f-i
»>
i> cO
•U
cu O
n 1

o
l-i CD
(V
•H
•H I—]

>-< dJ
hH

o
n 1
TO

00 n\
U>
r 1

""^ cd

d)
r!
M
4-1 CO

•H
O
3
CO M-l
CO
cu

o o

•U
C
CU m
u
i a
•u
cn }-i

MC CT3

CNl

ure

Fig

93
 CO
TJ
c!
ti

OT
4J
CO
(U
H
o
o
1)
,—1
r—
Q
u
•
*u •X3
rH C
o CO
CO

Ma

4-) 60
•H V-i

3
13 CO
(U (-1

CO

4-)
4-t •H
CO
O
•H CO
3 CQ
cr z
w
d)
o 4-)
•H •H
4J C«
•H
CD M
•H
3 M
cr z
o
<3 CU
JC
CO 4J
4-)
03
Q CO

<»
CN

ure

Fig

94
 O 00
u (-1

3
•UJ

CO cfi

OJ S-i

H 0)

CO 4-J

•H

d
•H
O
S-I

3 CO
Q PQ

0) •>

CO 4-1
43 -H

00 M
•H M
OJ Z
pa

O 4J
•u

0) CO

o c
O CO
4-1
S-I cn
CO c
rH O
O CJ
no
<u
^ E
CO -H
-H H
3
43 OJ

H ^
3
u
X) (U •

•H 13
CO e CO
3 V-i

O 01 >^
(-1

OJ CO
W Q S

OJ

0)
S-I

3
&0
•H
fi-

gs
 •H
1
—— 1
10

B
C 00
H M
6 CO
(U
H
I

CO
Ml iH
o u
o
d
^ — o
CO

iH <m C XI
cn CNI
0) a»
•

o 1
•H
O
4J
CJ
•H 0)
cn
00 1 ro
iH
QJ
4J04 ^ .H
QJ
<4H iH
1

d W
e un 11
II e
O rn O OO
QJ c rH 0 0)
<r c U CO a CO
o 3 0 C 4-1
o
1

II •H 4-1 LO
v-( 1

QJ CO
rH
xt
+J CO <r a C3^^sl X) Q
CJ >-i 1 •H
X 0) (U QJ 1 00 II O ,-1

CNI
01
•H
X)
CL,
B
QJ
o
CN
-d-
1

00
iH
(NJ QJ
CO
<r
1^ o
-a XI
4-1

II rH U-l
3
rH
o d (VJ
II

II c c U QJ <4-t
CO iH
•H •H •H QJ X 0 •H LTl
II

e
QJ 12 CO .H O O X 4-1
u •
00 e •H
1

a o o
4-1
CO
U
d)
bD
QJ
(30
4J
C
C
CO >>
4-1

CO
4-1

C
II QJ
a
o
o
2 z
I

CO CO QJ QJ iH QJ QJ u J-l

S-i •H rH
4-1 O CJ Cu QJ GO CO
CO
rH
o CU
>
rH Cu
C
•H Cu
H w M
cn

Ph
!-i

Q)
O
rH
4-1

C
M
o z
OJ

o fll C/D
0)
Pi o
p-l CO
4-J
Bl
5-1 1 c s-l

O 4-1 o o
U CO •H
O 4-1

0)
.H
fin CO
iH
CM
e O u Q
CO
c/b
.H
O VI
3
cn 00
vO
d
be In c
(U M rH
4-1 O ^ CNI
CO
CO u e II
iH OJ
rH rO CO o CO nc
Pi <c P3
ea No c
1
o
4-t
CO
13
C
>-i

QJ .H CO o QJ (U O
Z
rH o
pa •H
43
II
or
CO

ol
d B
•H
Jh
-u
CO
IS 1-1
1

XI .H CO 4-1 a) CO
0) -H cn Q) O &, s-l rH
N O tfi U 0) 4-) ><: o o
CO
rH
Pi CO
rH
CO tH
H
CO W C4H C/^

E o CU O CO
1 3 Vj U 4-1
(M
<U C e 3
rH •H o 4-1 CO
CO
Q o CO
rCl
3 3
E
0) O
S-l tfi

rH d u
CU

o tH U rH
Q <C <: o <:
3,
•H
PC4

o o o o
o ro CJ

% *A0N3l3IJd3 dO103niO0

97
 6

iH
CO
•H O
e
5 •H
V-i
C
•H 1—
00
6 •M u
CO
<U a
H ym
CO o a
o
>^ CO
o
1—1 C3
lO r^j
0) cu
(VJ •H 4-)
~6 O O
O 1

00 1 ro er 6 •H
<4-l
CU
iH
1-H 4-1 CM >+-l rH
CO
1

II
0)
6
B W O
CJ)
B u-i II en (U
no
OJ rH 0 (J CO
LTl <r ur c 4-1
on ra
o
o
1

•H 0^
1
(U CO
O
1
II

-o O
4-1

cfl -cr a 0^<^4 o 'O

•H
J-l 1
o rH
X OJ
<u
0)
S-i
QJ 1 00
<r iH
II

<r 6 CJ -In TV
Csl a. •H
CO X)
e
OJ
o
Csl
1 CM
00 cr>
QJ

ro u p
rH
O
X) .H CO
II c c u
•H •H •H
II

OJ
OJ
TD C
II

M-l
•H LO
II
o
o
e
>-(•
OJ to rH O O T3 4-1 o O
a
1

4-1 00 e •H 00
0) U c 4-1 4-1 QJ
00 60 C to CO
.H
C
II

o o 1

J-l

o
o
ct)

u
CO
S-I
OJ
•H 4-1 iH
rO iH
QJ
O
CO
QJ
CJ
Q)

o
s^
QJ d CO
<U 4-t

> >
0)
e •H
S-I 4-J
2: o
H w Mc c
QJ .H
< <: Ph c o c 0
O o o o
<u
rC .H
i-H

CJ o o
UO
4-J
CO 2: 2;
iH 5-1

PQ S-I CO 5-1

u o u
o
1

4-1 o
CO
rH
.H
o
CO
.H o U-l CO
4-1

o
CO
rH
•H
4-1

CO
O
CO
o
CO 6 CO
4-1
rH
o
QJ QJ
rH rH B 4-) S-I 5-1 .H
1 3 CO O QJ 3 bO
o S-I 03 oc l-M 4-J CO C •
CJ QJ c vO to QJ (U •H a
XI M 4-1 P3 <: Pi 4J o
QJ 5-1 iH CO
CO o
4-) O Q U)
iH z
CO
.H X3 CO
C/3 a B II

rH
S-I

3 3
5-1
o cd
0)

1
<c PC -H CO
0)
CO
i-l
O O
rc CO
d 4-t

C u
1

cu
>-i

CO
iH
4-1 S-I S-4 c 4-1
O
to c QJ -H CO QJ <r -3- CO e 4-J

rH O XI B o •H
U
CO
:s
C/3
fj-l PQ •H II S-i •H rn CO
1 o S-I 1 1 C Qi x:
-Td .H CO 4-1 QJ rH rH QJ CU u 4J
0) rH QJ u D- QJ o •H
N o Cfl S-I QJ X 4-1 4-1 5-1 CM W 5
CO
iH
CO
rH
CO rH
tH
W CO to CJ
CO o
O e O QJ o e to CO
d CM
1 3 S-I CJ o 4-1 HI c fO
QJ C B to CO 3
iH •H u 4-1 p-l P o CO
CU
XI e S-I
J-l
QJ
no lu
& ro

o < < o s,
•H
pC4

o o O O O O o O O
o CD in ro CVJ

% AON3IOIdd3 d0103in00

98
 o
2
u
o

0) w
-( QJ
o a
^ c
0)

CO
.—IO
O c
C/3
I

00
c
•H
U
CT3 o
I

I n
01 OJ
00
n3 u
0) 3
0)
X
4-1
J3
W
> S-i

o 01 CD
XI
<:
e •^^
S-i 03

U
(U OJ
J-l
O
CO)"
D Q CT)
O «
S O 2
j-i
C
OJ W 4-)
!-J OJ T-l
U H
CO
CO t-H
V4 :^
CO 00 I—
-H
O
C 2
-rH
CO S-i 01
^ 3 -C
<: Q 4J

on

0)

00
•H

99
 (0
d

4-J

5 CO
0)

6 H
t>^
O
§
•H
U
•H
U-(
M-l

d W
(U
O
ti
0)

•H
O
o In

d CVl
1

r-l
CNJ
CO
E B
u r\
o
o
o 1
\J

J-l

00 CO o
o (U
2:
1
1

O
i-H
d QJ ,
—
•H (

e 1
+J o
rH >-i
CTJ u O
&0 0 U-f CO
cn iH
<r T-H CO O
<r o-|
Q 4-) W
o Z 1
o rH
3 W)
1 O CO
0) •H
C3

CN OJ e Pi
m 1 3 4-1

01 00
4-J
CO
II

e II e in u iH
PS
0)
0) o 0 CO
IT) CO C !-i c i-H 4-1 1

O
1

o 3 O
o
1

II H JJ iH o 0)
»-i

(U
-H
T3
4J
a
n3
)-i
CM
1
t-HCN
d B
•H
4-t

CO
X OJ 0) OJ 1 II 12
q; a Ph a^
00 D- •H e T-H 0) r-- »-i
cn X) 0) CNI Cl, 00 CO
O
CNI 4-J 3 O
XI t3 II i-H II M-l v£)
C C !-i 0) U-l
II

•H •H •H 01 T3 c •H m II

CVI

•w
OJ CC

w
T-t O O
E •H
X) 1 4-1
o
u
0) OJ
60 C
c
>.
4-1
CD C
4-1

o
II cu
d
nj OJ .H 0) 0) S-4 u
5-1 (-1 •H 4-1 O u a 0)
o OJ OJ XI CO u o 4-1

> > •H a C 0) rH c •H
< <: H w t— D-. 1^

o o o O
o
0IJd3 d01D3inOD

100
 1
— — I

\
to

G
in
^C rsi
4J
CO

•H
e
I
6
to
o
00 M 2 oU C3
<u
e 1
•H
iTl 00 O
CN •H
O tH
CVJ
tU
IW
OT
-H 1
1

<} (U B
d
1 4-1 (U
(U O
o ou
II

CN II e
cu o rsl
0)
(N u c
o
1 o 3
•H 4-1
o to O
<-\
1
T3
•H
Sj
U
o
II

iH u CO <f >^
C
-v o >^ P. 1

X
u-1 a- •H
(U (U
(X
B VO
1

P-1
ON
II

OJ
00
o
00
6 .H •
M XI (U cn m E CO ro
ro
-a
4-1

II .H U-(
ro o B
Vj •
II

II c C
•H •H •H
^-1

QJ
0)
T3 C
IH
•H
II
o
o
O o
z
<U to i-H O O -a 4-1 •H
u
1
(U 1

(30 e •H <x a u
to 01 0) 4-) c 4J 4-> II 0) la
to
00 CO o
(30
to
00
to
C
(U
td >.
(U
CO
tH
C
Q) QJ
o
5-1
OJ
IH
)-l
o 2
(U 4J
CJ

o
u •H 4J tH
<-i Q.
O a D. (U
CO O
S-i 4-1
,o
B
3 6 0)
cu
iH
•H D. a OJ iH C
1 H W M CO M CO
Q) 4-1
rH
O
H> U
V-i
4-1
CJ c o U
o CH
(U o t\3

iH
tH •H
4-1

a
<u
0)
C/1
4-1
to
o CO
4-1
O
CO
tH tH B C) tH
tH (U 3 00
O B tn vO CO c
0) •H
ro In
PS 4.)
cu tH CO
4-1 u ^rsi rH ai
to 1 U B II c to
.H CO o 4-1 1
CL,
1
a
CO
PQ CJN CO
CD
o o c
<u
5-1

<U
u
4J
to
rH
tH
CQ
!-i

OJ tH
•

CO >-l

CO
6 e
•H
4-1

CO
|ij •H II U .H
O O OJ
T3 (U CO 4-1 to a.
(U XI tfl OJ u O
N u
CO o
tn
CO
u
CO tH
ai 4-1

CO
w 4-1

.H tn iH H CM
O CJ OJ o CO
o in
1 < ^-1 CJ 4J
Q)
rH tH
(30 OJ
B
o 4-1 to
CO
Q
to
d (U

•H
(U
4-1
o 0)
cx
o
u
iH
tH
in CO < o <i3

Pt4

o o o O
o ro CO

% *ADN3ioidd3 doioannoo

101
 d

5 CO

6 0)
H
>^
o
CO
00 CO 0)
•H
O
d •H
MH
MH
13 0) CNl d W
w B 5-S
OJ
QJ 00 CJ
e 1 e 3 to 0 O
CO O c
•H LO (-0 c i-l c <f 1 QJ
O o 3 CO T3
o
1

II •H
u
u (-1 CJ^ lOrg •rH
•H rH CO O
rH
rH X
Xl
0)
u u
01 QJ
a 1

LO
o
M
II

01
00
Q)

CO
!-i

•rl
XI
Cu 0
e
Q)
LO
<r c/:i in
Csl Ph 00
QJ
CO O
<! d •

C ro •IH 3 ON CD
<u XI -a II rH II MH m U •
C c Q) M-l
o o
a
11

•H •H •H QJ X) C •H o II

z
te
CO rH
mo
o XI 1
4J
00 •H a. 1

»-< CO 0) QJ JH c •U 4-1 II 01 $H }H
o U 00 00 C CO CO C u GO CO o
u
o
CO
QJ
CO CO
U
QJ
•H u
Q)
rH
rH
O
QJ 0)
a QJ
o OJ 4-)
o
QJ
rH 3 rH
o
i-l

0)
>
QJ rH
•H
a. CO
c
CJ
5-1

QJ
o
rH C
4-1
d OJ
OJ
rH
rH H < < H w 1— cn 1— rH
o •u o
u t3
•H c O u
o MH
rH CN o rH
3 pa 6
OJ
u
o CO
4-1
o
C/J

H or
e
o
00
ON
la
d rH
3 ao
1

X)
OJ
CO o
1— CN1
Ln o CD
CU
c
•H
4-1 <! e II PS 4-»
nJ a; CO
3 a» o CO iH (U
CJ >
•H
o QJ
)-i
•H CO PC
4-) 1

w> o
•IH CO tl
C!
o D rH <U

XI
(U
rH
QJ
tn

X)
II

CO
O a
•u•H
CO
d s
•H
QJ 00 0) <U u !-i

N J-i tU U 0)
CO a c CO T— 0) V4
r-l •H •H rH e o
O CO (U o ym MH
1
4H Pm S-i o
(U c 1 C/3
u
rH
00o
QJ
U u
H
CD
03
O
CO o
c C
•H o
cn u
0)
a.
<:
1-1

o Q
4J
CO d (U

•H

ooooooooo
J \ L

% *AON3IOIdJ3 d01D3"inOO

102
 (0

6
=)

•H
e
o CO
2 CM 5[ (U

I
6 H
>»
c ^ o
^ B
cn
U
o
§
•r-<

CO B o
o bO •H
>M
c» . I

d M
c 3
•H CD
e O rH
^ I -d-
CUCM <U
rH
•U iH O
cC
U >
0)
ro
I 13 4J
CU
B
\B :s g
4-J •H o
c U II
B !-i C CM •H
cu O D o CO <1-
O
a OJ
C rH
CO 4-1 ^ o c
CU CO <r a> CM
o 0)
u
!-i o
I

T3 u CX B
d
I

c/:) CO CU (U I
CVJ
CU
60 (U (X CO LO
c & CO Q) P-i <X> B
u •
•H o CU
o o
O
U
!-i

4=
CU
15
CO
C
•H
CU o
o
1 o o >-( 1-1

H c CU E -H GO CO o
O B o
CO
CO
U
O
o
iH
CU
iH &0
(U
txO
CU
bO >^
4-1
CO
4-1

c
II

o CU
z
4-1
a
tc
C
iH 00
pq
rH
u
^—
&0
c
CO CO CU
>H
r-l
O
(U CU
d CU
CU

cu 0) CU 0-, CD >-J o
CO 4-1
> > &.c CU rH o
iH
u
0)
II (-1

o O
-H 4-1 C
CU
<: <; M H (Xi c« u
(U cti u cn CU 13
u
c (U o T3 •H O V4
CO
cn o U CU 4-1 •H CJ M-i
0) cn cfl iH CO O c r-i

u iH C M o CO
4J
O
CU O CO h-

^1 3
O 4-1

CO 0
0
00 d 60
cd cu Cfi Q O CM
<u ex u CD •r-l

C Ch O iH
•H o a J-i >H
1
CO
i-J <: o <3
1
.H CU
CO 33
4-) I

C H
o CU <ll

d e
•H
4-1

CO

0)
a. vj
X o

o CO

d u

•H
PC4

o o o O o o
o 00 fO

%*AON3IOIdd3 doi33nnoo

103
 CO
CO
5

u •
O TJ
J-) c
O C«
OJ rH
rH

cS^
CO 00
rH U
O D
CO
00 u
C 0)
•H X
U *->

CO -H
QJ CO

(U cyj

CO S
00 Q)
C J-*

•H -H
4J CO
CO

C M
CD 2
O
C 01
O X
U W
•> 4-1
00 CO
C
•H T3
^O <U
4-1

CO CO
U <U
H H
C30

•H
fa

104
 u
(0
14-1

V4
a
CO

>
•H
U
O
0)
.H CO
tn 0) 0)
(U u
o 0)
cd 0) MH
<4-l e U
(-1 o 3
3 >-( cyi
C/1 * *

U)
<D
> o
4-i «l •H L
C (fl o 4-1 C
< u
o
•H
n)
CM
u
0
u
rH
O
PQ
J-i

O
4-1
O
11
rH
o ie

o u 0) O tu
iH
QC
O
t—
H
O
(U
r-l o
rH
<u

0 H
0)
(-1 bO rH

•H o
rH
Q) CO

Q)
6
CO Co

CO
U pa u o CO
c 3
0) U
1

0) o OJ
S-i

x; o
4-1 n) l-l 4-1 CJ u •
(fl 1-1 CD 1

— o CO

o
o
iH
PM

J-)
1
li,

13
•«
pj •H
1

4-1
c
iJ
nj
a
&
o
0) rH a:
P-.
1

4-1
o
rH
(

c o
u
o 4-1
cfl C nJ O u pa u
iH O rH u
CO
CO
[n PQ
rH
o 0)
0)
i-H

o 0)
N o N
1

1—
rH
T3
0)
>
•H Xi
u N
4-1
\>
d)

o o CO o
T3
0)
l-l

<u
,Q
l-l
CsJ
.-H

O
O

CD o
oi

e
n)
.H
O
0)
.H
cn rH
O
U5 o E CJ
U
1 D 1
1+4 <; 1 < *
C la
c a) <u 0)
05
f-i
e •H
•l-l S-i .H rH rH
60 0) o OJ
•H
So
3 d
O
4-J

o C
0)
OJ
4-1
c
•H
(U
4-1
o o O
•H oo
Q < Q C/5 C/3
f c
1

m1
oo f •H
00 iH CM
o w
<
CO
o
CJ
4-1
cc
<U

CO d
CO o
r
»— 0) u
1

UJ
> o 1—
•H
CU

cc o CO
CO c
ID o 4-' l-H
:3

CJ CD .H
1
ed

>- u
in g o
CD (U

o No el
1

C3 u
tfl

o Ne fo

CO
CC o
o o
o
CSI

•r-l

o o
fclH

o
_L
o o o
oo
o
CO
o o
CO
o
CVJ

% 'A3N3l3ldd3 a0i33l103

105
 <u
u
ca

u
3
CO

>
•H

m 01 CU o <u
0) C/^ u u 4-1
u CO o
CO
CH
CP M 0)
E S-l C
!-i o 3 o •H
3 S-J W 4J 4-1 cu
CO j:; CJ CO
CJ (U CU S-l

CD > 4-1

^
1

4-1 CO •H rH C
c CO o 4-1 O CU
fo
u •H rH CO s-l o U o
o CO o CL5 iH o cu d
4-1 4J W 4-1 rH S-l o

3 1
—
a
CD ^O
CO
a
QJ
iH
I—
0)
QJ
)J
C
o
cu
CU

cu
CO

3 s-l

cu
CJ

M
c
CO
0^
(-1

< u m
o

(U
1

4-)
0
(J)

0)
H
C
O
•H

H
3
o e 3
CJl

OJ
o
!-i ^
-3
-Q •H
Sj
O a
cn CO
<3

cu

4-1 CO 4-1 4-1 (U s-l

rH
CO .H CO
iH
S-l

0)
CO
rH
1 4J < ]
t
CO
•H CX P-, O 3 cu 1 I

1 1 4J a 1 CO u > CD
<u
4-1 XI 4J CO o 4-1 tH CO •H C
C O u <&I
W>
n) CO CO CQ 4J CU
O CJ> iH O hJ
pa CU
Q- CU (J tJ .H i-H

<
1

-d T3 > cu 13 T)
0)
N o N
i-H
rH OJ •H
4-1
<U
N
(U
CU
N
CU
CO
CU
C
CD
:= E o
(fl CO O o CO o CO a CU

O o e
3
o
OJ
.H
y-i
cn

<:
rH
o
<
cn rH •H
O s-l

4-1
s-l

fa o
o
cu
C/3

O
1 1 1 1

dJ C cu 0) 0) CU C S-l
CO
.-H •H ^1 rH rH cu CO

LU J3 E ,£3 H cy CU 00 o cu
3 3 4-1 0) c 0) C c c
CO O O c 4-1 •H 4J •H o •H

<
QQ
1
<: Q
1

Csl
<; CO
1

c^
C/^ cn CJ iJ
1 1
o
c
cu
CO
•

CD •H o
CO o
LU
>
QC
<+-(

4
W
1
cu

o
U CU
o
CJ)

>- c
o cu

•H
CO
iH
O
o CW

o M
1
60
t-*
iH •H
CO 4-1
CO
g CU
O 33
O 1 U
1

o >-l

CO
CU
(U
•u
CO
2 IS

o
o o
CU
u

•H
tL4

% 'A3N3l3ldd3 aOi33n03

106
 Double-glazed flat-plate collector (Lennox)
0.106
Steel absorber with copper tubing
Aperture area = 1.40m2 (15.1 ft^;

0.9 Gross collector area = 1.66m2(17.9 ft^]

Inlet fluid temperature = 33°C

0.104

0.8

0.102
0.7

0.6 0.1 Recommended

flow rate
CM
re
a.
c

J 0.5
0.08

CO

0.4

0.06

0.3

0.04

0.2

0.02
0.1

OL
0.1 0.2 0.3 0.4 0.5 0.6

Flow rate, gal./min.

I L I I

1.0 2.0 3.0 4.0

Flow rate, 10^ m3 /s

Figure i+1 Pressure Drop Versus Flow Rate for Water-Heating Solar Collector
No. 2.

107
 Single-glazed evacuated tubular collector (Owens-Illinois)
14.0 r Concentric selective absorber
Aperture area = 5.04m2 (54.3 ft^)

90.0 - Gross collector area = 5.98 m2(64.4 ft^)

Inlet fluid temperature = 25°C

0.2 0.4 0.6 0.8 1.0

Flow rate, gal./min.

0 1.0 2.0 3.0 4.0 5.0 6.0

Flow rate. 10 5 /s

Figure 42 Pressure Drop Versus Flow Rate for Water-Heating Solar Collector
No. 4.

108
 4J

•H

t-l
0)
o
a)
1—
DQ
1
CO
o 4J
4-i
o
n!
,-H
o
(U
1—
I-l 4J

uo 0)
CO

(U
H
w o
(fl CO h
1-1 <U

I o
o
u
nJ c
H o
ON
PQ

-o I-l
in i
(U rH
(U
N O
n)
tH 00
a e
3
1

(U c o
iH •H VD O CM
.o S
3 3 c
O 1-1
a < •H
e (J
CO
o
o
iH
nJ —
^ ^o
u^
m in o
60
CO
0 r~
vO
H 2:

I-l tp^ o

o
m •
CNl

in
r-.
o 1^
w 1
w o
'4-1
c vD 1 in 0)
o — rH
o
c
o\
on
•> (0
1
CM ^
cu 3 rH
1-1

11 1-1
w o
\o 6 II

cu
00
C!
0) o
n) •>in CM c u CO OS
4J iH 0 o 3
o
1

>-l o in II •H 4J U
O 10 <r iH 4J to C
c 13 CJ cu
o
4J II X <U 0) cu XI ~*
c
bO n) (U D, -H tN
ti (0 CM (X -H e o e cu
•H •O
O
(U
iH CO
• to X) cu
u
c (0 O e
(U 3 •H c
(0 4J
o c:
00 14-1 II U
y-i C II u >*-) ro 0)
0) <o •H •H •H cu c •H
u 1-1
o iH •U
<u
4J
to iH O X)
00 -H cu
O CO
4J O C to <U
o O <u u 00
0)
00 to
C iJ
CO
4J
c
O o
Q) X)
iH nJ •H IS u
to to cu
•H 4J
iH
O o
tu
V y
1

m
rH 4J o o (U iH
II
CO
(U CO ^1
o nJ C iH -a-
•HC cu o
u Q M lii H M
(U

3,
•H
o

oo CO 1X3
o o
1 D

109
 al,
oo

4-) o-
O
o o

1
<f
w) CO cn I
3 g u
m II
3^
V4 vC e u-i II <u
01
X 4J 1 in a o
4-1 o 3 u-i
3 -O o o
1

II -H 4-1 00
O 0) 0^ 4-1 to tNl
to 4-1 T3 O 1 VD
o 0) 0) QJ - CO
00 0)
II

(U (-4 o II

o
C .H CO 00 fx -H e m OJ
•H ,-H
U O
01 ui -a 0) Csl [0 o
CM 3
n) o
1—1
00 c a t3
4-1
tH II II II
u
o d d u
II in t4-(
CO
to o
oo •H -H •iH 01 d •H dco
u u c c OJ o a 0)
O
4-1
to
Q
4J •H u 4-1
(0 r-i
00 H
C c 01 to 01 QJ 4J c 4-1 4-1
<u
o
•
^ H
o 0) 4-1
>-l M 00 00 d to to d 8 1

^< T3 o
•H
a
to to OJ .—1
< IS UOJ U •H 4-1 o tj
o m
OJ H
^ II
cu
(U J3 1-1 •H
•H d
0) »W
o •H
« TJ

3
CO

60

CM
4-1
vD a\ 00
^ (J^ iH c
0)
CD
o C X)
O o 1-1 •H
H o CJ
o
c
1
4-1
o Mc •
c c« u 3 CNI
(U nj to
0)
ij
as
ea MC o •

u .i!
u-1 in iH o
o
u
O ac O
HC nJ >-l
tn
o a> CO <t --^ o >-i
0)
tH
(LI

^4
j:>
3 •V o •
g <4-l o
.-1 H u O 4-1
o
u
as
O
— tjN

00
• CO o
C <U m . m
4-1 (U
0) o D. o . I
iH iH
D. u Oro iH
n) 00 O g
c O
<u vo u-i
CO O
&^ •H tH -*iri ^ 0) CJ
1

4-1
4-1

(0
J2
4-1
P-i

U 4J
o
OO
I

11
d pa
<0 o •H T3 O to 00
i-i

•rl
d •a o vj OJ
rH CO
oc II X OJ OJ -a Cd iH
•H U
•o > <u Pi 00 -H
a. -H CM
e o g
4-1 O
01
N
•H
4-1 u
1

•H
CD
0)
.H CNI
•
.

•o OJ d ^ OJ
u)
3
o C!
(U
CO
nl u o B 00
01 0) •H U •
t4-l
t4-(
II

r
B C»0
O fl Ab
CO •H OJ d -H d' •H c
OJ to rH O 13 "r »-l •H
<u E 00 -H
>J
•H <u
u
4-1

d
OJ
4-1
to
^4
OJ
oc d
d
to
-U
to
4-1 ^ OJ
c
(U
a-
4J
CO
3 4-1 01 u-i OJ to a to OJ rH 01 H ><! 01
Do An St 00 •Hg
U O
^4
0)
t-1

OJ XI rH
4-1 O O
CO
W 32
Md M (ju
>
<; H M PL, pt, J3
o

oo CO
O
I

•H

ID

110
 ^O r~.
CO
CM

o
vD <r
^ in
ro
iH

u
o
U-l

4-1
B
in 0 CO
rH rH
— o
m CO 0)
01 ^
00 w o iH
iH
H
1 o i-i
in "b 0)
o o •H
O O— ^
0^ o
U-l
c 00 1 •H
o (0 X)
o o 0)
O
c O ro 1

II 1-1
B II 00
Vj (U c
cfl LTl a u CO 0)
1-1 0 o V
C
1

)-l o o
in H
II •H 4J 4-1
1-H
o to 4-1 CO c
c •v O (U <} C
4-) II X <u 01 <u -a ro
<a
00 n) 0) U a. -HcN CO
c CD in 6 o e 01
•H •a
O
<U
H00 CO m •a (U to O
OJ 4J 3
CU 4-1 X) -a 14-4 II
0)
o c II C C >-l M-l
0) •H •H •H 0) c •H C CN XI
u iH (U CO 1-1 o T3 «CX) •H
C
o iH iJ
4J O c CO
4-1 00 -H
^O
0) iH O
(U 0) 4-1 C 4-> 4-1
c
•H O a (U u c c S
(0
<U
o T3
60 00 CO CO
iH H
1
M •
r-l
U
at
01
iH
1-1
cfl

u O o
1-1
CO
)-l

<U
CO
u •H 4J
1-1
O O
CO
0)

J-l
^ II CO
01 u o CSC iH •H C QJ a; o CO
S V3
CJ Q M H M Oh o
u 4-1 o
•

CO
£
o
<u
> o ^
<4-lo
>-i

o (u
CJ
u
CO o
U U 4J 4-1 <u
<U C/3 (d 1-1 rH
i-l 3 4)
3 1-H

0 B w CO o
CJ o c (D CJ
1-1 M Pi
0)
u u ^
£ u
r-H CO
ca r

1-1 ^o « CO <-{
0^ m 4-1 o
U
1 CD
i-l M c C/3
CO CQ <U 0)
^ X>
- -H
e QO
Pu •r-l c
•a <u
J-l •H
0)
N
^
»4
m
(0 a
<D 4J
CO
cd o cd X 0)
i-H
0 ^ o
to .H
w
1<
0) a
1-1
00
M c;
(U
C
•H
0)
iJ
o•
C/2 C/2 r~
ure

Fig

03 oo CO
d
ID
>l

111
 o B
0) I
•o • 00
— o u
o ^
1
o
<u
ca in 4-1
u to
SW
tNI B
0 :i
u >* m
o
B ^ II
II

<U o
CO
l-l

JS ro c u rH o
O
1

4-1 bO o 3 M-l
o
1

CO 3 o 1-1 4-1 rH
•H o O H II

4J to C7^
O cn T3 tj u in
CO
0) OJ OJ
•H 60 4-1
II

01 )-l 0,0 O II

<t (U
CO 00 a. •H B o 01 H
rH
h-l
ci ed le ro
cn -a
te
tsi
US
O
1
4-1 CbO T3 "a 14H II II II

0) <^-lu C O UH
II <u
C 01 to oo o •H •H •H 0) c •H c rH •H
01 U d to o -o in (4-(
o •rH
4-1U 4J 60 T-l 01 CM
•H
a •u o
'J u 0)
a c OJ
4-1 V4
01
1-1
01 c 4-1
tcJ

60 60 C cd to c 3 O
4-1 4-1
•T3
0> T3 o to COt4-(
0) rH 0)
iH to •H <: s U •iH u o tJ II

O t-H o
4-1 - o OJ 01 J3 rH CO u
u O to > > •H c 01 PS o
a U Q M < < H M CL, x>

o
CM
C
cu
o u
01
•H
00 O
o
o Mc •
0)
u <
•
*J o
u
o
-a 01
M-l O
4-1

N CO O
ca o 4-1 <U
.-H -H tH
-u
3 -H
I

0)c CA O
rH 0) (U
00O piS
c c
•H O
U
iH
U m CO
CO
r-t
o JJ
c
O
zr>
00
o cu
e bO
•H C
•H
CU 4-1
CO
o a CO
X OJ
in w
o
v£5

CO
o CU
U
o
•H
[X4

'-: O OO CO un CO CM '-r 05
CM CM o
ID

112
 1.2

INCIDENT ANGLE AT SOLAR NOON

O 0° 12° A 27° O 42°

1.1

1.0 1 GLASS
COVER PLATE

5 0.9

2 GLASS
COYER
0.8 PLATES

CO
0.7

e 0.6 \0
\

0.5
Legend :

THEORETICAL PREDICTIONS, CONVENTIONAL FLAT-PLATE

COLLECTORS
0.4 EXPERIMENTAL RESULTS, FLAT-PLATE COLLECTOR,
SINGLE-GLAZED, COPPER ABSORBER WITH FLAT-BLACK
PAINT, PARALLEL MYLAR STRIPPING (Sunsav)
0.3 I I
\ . \

15 30 45 60 75

ANGLE OF INCIDENCE, O, DEGREES

Figure 47 Experimental Results for a Incident Angle Modifier Test for a Water-
Heating Flat-Plate Solar Collector with Parallel Mylar Stripping [22].

113
114
 o
to
14-1

u
3
CO

0)
>

U) 0) 0)
4-J <U CO 0
c u n)
•H CD 0) ii-i

U-J e >-<

l-i 0 3 0
3 CO 4J
ly: u
0 u 01 01
> rH
^ 1

•H rH
OQ n) 0 4J 0
1 U tfl >-i U t_)
U 0 0 0 0)
nJ 4-1 CO 4J tH l-l

iH 01 0 0) cfl

CD <u 01 CO rH
00 rH 3 U
x: rH 0) X) 0)
0 •H e 3 J3
<D U u 0 H )H
c 3 S-i 0
(J (U 0 H 01 XI U)

CO 0 iJ 4-1 u 0)
(0 (T) 00U 1
4-1 <
cc < a.
c 0)
•l-l D.
rH
0-1 CJ
n)
3
4-1 a 1 u >
13 4J « 0 4-1 rH •H
C 0 u —
w>
(T) US ca 4J
0 :-H u CJ
m 0)
1 )-i ^ rH
rH T3 > 0) T3 01 TD 0)
rH <u •H XI OJ -D 01 CO
0 N 4-1M N (J N
OS 0 0 0 nj 0 tfl

01 w iH W5 rH •r4
e r-H
1 < 1 < 1 4-1

c 0) 01c
CD u
0)
rH rH rH 01
e Xi •H 0) 00 01 00 u
< H3
< Q <
3
0 c
1
4-1 a c
4-1 •rH
CO CO CO
a c
•H 0
0
60
O
1 1

Csl ro <3-

GC
O

ID
>l

115
 COS e

Figure 50 Incident Angle Modifiers for Selected Water-Heating Solar Collectors

in Comparison with Theoretical Predictions,

116
 1.2

1.1

1 GLASS COVER PLATE

1.0

5 0.9

2 GLASS COVER PLATES

0.8

0.7

» 0.6

0.5
_ Legend :

O LINEAR FRESNEL LENS

TRACKING CONCENTRATING COLLECTOR
0.4
COPPER ABSORBER
BLACK-CHROME SELECTIVE SORFACE

0.3

0 15 30 45 60

ANGLE OF INCIDENCE, O. DEGREES

Figure 51 Incident Angle Modifier for Water-Heating Solar Collector No. 5 at
One Angle of Incidence.

117
 CO
CD

IXi
o
O
•H e
£ 4J
0)
(U
H
I
CO o
O c
~cr o
I

B CD CN
o
in o
CO M
ON 1 MEN
a>
1 00 CU E o
1

II
et c
a>
B
II

<u o o 00
O •H
c !-i c
o D o o u
o II •H 4J
o
CO
1-1 I—
<r o
tJ O U
CO
CN
?^
O. o
m
1

o < <
or UJ LU
cu
(U a
OJ
!j
QJ
Pi o
I

CM CO
aH E o o
1

00 <r oc oc
(/)
05 £ e
s < «^
05 ro cu
o o 0
o c c
Xl
4-i

II rH II o o
•

h- 00 to
1-1 CU 1 o
o o o •H •H •H
S CO
QJ
iH
c
00 CO O
oc oc
CD CD
C>C
C
OUI OT
o 2
CU 0)
01
00
OJ 4-1

c CO
4-1
CO
tH
o >H
CO
u
QJ
•H u
cu
O a 0) O
,— cu
0) O. cn
a a
> > e
< < < H w
•H c
1— o
I

(U
o >-i

O
CO
r-l
y-i o
CO
CO
o CO

o 3 -H
CO iJ
in 1) CO

p
o
CSS

rH
(U

CO U
^••^
o B (U
•H
J-l
X
*J
0)
CO
o X! O

CO
CSJ
o cu
u
3)
•H

% 'A3N3IOIdd3 y0133n03

119
 CD
O
Ln
o

o u
o
o
CO
o
o
u
CO

o .H
o
CSJ CM
E E c
LO •H
in
CN 4-1
CO
cd 0)

<
UJ
—\
CSI u
I

OC
<
o ^
< <
05
o
CO
CO o o
0)

o
o
1-4
oo o
CO o CO

u
CM c
o <u
•H cn
.

O
QC
< b o •H
<4-l
CO
0)
U
< CSJ CO 14-1 <
UJ CN
p 4-t

t—
QQ E o <u
c/> to
<, < CO p
d
<U

•H IW
(U

O -H
CC UU
C Q
UJ
CO
QQ
^ o I o

< O 00
CO CO a
cd

g c
CD ca <r o CO o o
^ < o p
CO U (U
=) E
o 2 o CO CO

o
a CO cn
CM 0)
2 CO

u- Li- p
o
pa

-a-
m
p
o
(U

•H

o
o o o
oo
o o
CD
O
in
o o
CO CM >—

% 'A3N3l3ldd3 H0i331103

120
 2.0

AIR LEAKING IN - AIR

FLOW
MEASURED BEFORE COLLECTOR

>-
>-u
U z
z UJ
UJ
U
U
UJ
o
O
I—
I—
u IN - AIR FLOW
AIR LEAKING
U MEASURED AFTER COLLECTOR

O
O u
u
UJ
<
I— to
U <
<

AIR LEAKING OUT - AIR

FLOW
0.95 - MEASURED BEFORE COLLECTOR

0.90

AIR LEAKAGE RATE

MEASURED FLOW RATE
Figure 55 Relationship Between Actual Efficiency and Measured Efficiency as a
Function of Air Leakage Rate for Air-Heating Solar Collectors [23].

121
 00
O

O u
o
u
o

z
o to oo
IT
O u
< ««

-J o
o w
<
q: q: c
UJ o O
•H
4-*
I- ««
o (U

is Ul UJ
to
UJ E
3J

u
•H
O < (T <:
< ^ O cr. CM UJ CM
o
(A <
a.
to «•—
o
(U

4-)
V)
O E z >v
E
<^
OC.
O to
U to u fo
UJ
o z _l
3
(VJ
o CO
•

< o 3 c G
o a CM CM ro O
q: (T UJ
o E
MOD
E O •H •H
o 4-J
UJ (/)
f=! OD m < \in a: fi ra
q: 00
q:
o o m a
< o u_
>
f-
Ef
(O to
0 m to o
h-
o E o E •H
14-1

< z o c
1

UJ
< UJ
Ul
-Io
CO
O
LLE
(VI c o
_i _j
_l
O o _j
o 6 O d O 01
T>
CJ
CD UJ
3 UJ o o II
•H 4J
O c
o
a
O
t7)
U-
UJ CVJ
II

•> (VJ •> Mc u

1 0)
tH
O CO
<4-l
U-l
s •H
Q
o
2: O
M
td
(U c
Z •H

v£>
o O o m
(VJ O OD
a>
u
S>
%'AON3IOIdd3 d0133in03

122
 in
c6
(U

u
o
(4-1

o 4-)
CO
CO 0)
£-1

M
<U
•H
<4-(

•H
•o
o
<U
I—
Ml

XJ
c
1)
CM CO 'O
o
u •H
o
1—

a)

o
4-J

o (V
mh iH
i-H
CO O
4-1

i-H
3
CO CO
(U l-l
OCJ o
CO

CO
4-1
c •H
CU iJ
e CO
•H <U
S-i

(U
X
1

Lf)
a u
X
w

in

ure

Fig

M 'UBIdlQOIAI mm INHQIONi

123
 3
O

O
J-i

H
O
(U
•H
O
•H
M-l
M-l
W
CO
3
O •
CD CO
•
o
c
cd
M i-i

CO o
C +J
t— o
0)
rH
(U iH
-l-l
O
O CJ
•H
!-i

<U CO
U iH
P-i O
CO
T3
CJ
cC
•H
4J
CO CO
3 <U
j-i
(J
X
1

<: >-l

o CO
IS
d
o
CO o
•r-t

ca >^
& CO
e Q
o
CO

00
in

ure

Fig

(s Ln ^ {«)

124
 CO
3
O
•H
M
to
u
CO
>
U
O 0)
in >
o
bO
C
•H 01
4.) 00
CO to
<U u
0)
I >
u <:

-a
c
CO

CO
C!
o
•H
6 U
•r-t

X)
o C
oo
u
Qi

to

w
3
O
•H
)-l

CO
>
<u
XI •

C W
to
O
c
e
•H
>-i

a e
O -H
w

in

•H
fa

%*AON3IOIJd3 dO103inO0

12.5
126
NBS.114A (REV. 7.73>

U.S. DEPT. OF COMM. 1. PUBLICATION OR REPORT NO. 2. Gov't Accession 3. Recipient's Accession No.
BIBLIOGRAPHIC DATA No.
SHEET NBS BSS 117
4. TITLE AND SUBTITLE 5. Publication Date

January 1979
EXPERIMENTAL VERIFICATION OF A STANDARD TEST
6. Performing Organization Code
PROCEDURE FOR SOLAR COLLECTORS

8. Performing Organ. Report No.

James E. Hill, John P. Jenkins, and Dennis E. Jones
9. PERFOllMING ORGANIZATION NAME AND ADDRESS 10. Project/Task/Worlc Unit No.

NATIONAL BUREAU OF STANDARDS 742 4417

DEPARTMENT OF COMMERCE 11. Contract/Grant No.
WASHINGTON, D.C. 20234

12. Sponsoring Organization Name and Complete Address (Street, City, State, ZIP) 13. Type of Report & Period
Department of Energy Covered
Research and Development Branch for Solar Heating and Cooling
Office of the Assistant Secretary for Conservation and 14. Sponsoring Agency Code
Solar Applications, Washington, D.C. 20545
15. SUPPLEMENTARY NOTES

Library of Congress Catalog Card Number 76-6OOI3S

16. ABSTRACT (A 200-word or less factual summary of most significant information. If document includes a significant
bibliography or literature survey, mention it here.)

A proposed procedure for testing and rating solar collectors based on thermal
performance was published by the National Bureau of Standards (NBS) in 1974.
Subsequently, the American Society of Heating, Refrigerating, and Air Conditioning
(ASHRAE) developed a modified version of the NBS procedure which was adopted in
early 1977 as ASHRAE Standard 93-77. A test facility for water-heating and air-heatingj
collectors has been built at NBS and was used to support the development of
Standard 93-77. The purpose of this report is to describe the recently adopted
test procedure, the NBS test facility, and the tests that were conducted to support
the development of the procedure.

17. KEY WORDS (six to twelve entries; alphabetical order; capitalize only the first letter of the first key word unless a proper
name; separated by semicolons) Measurement; solar collector; solar energy; solar radiation;
standards; standard test; testing.

18. AVAILABILITY Unlimited 19. SECURITY CLASS 21. NO. OF PAGES.

(THIS REPORT)

I I
For Official Distribution. Do Not Release to NTIS 127
UNCLASSIFIED
nn Order From Sup. of Doc, U.S. Government Printing Office 20. SECURITY CLASS 22. Price
Washington, D.C. 20402, SD Stock No. SN003-003 (THIS PAGE)
$3.00
I I Order From National Technical Information Service (NTIS)
Springfield, Virginia 22131 UNCLASSIFIED

ir U.S. GOVERNMENT PRINTING OFFICE: 1978 O— 281-067 (339)

 NBS TECHNICAL PUBLICATIONS
PERIODICALS NOTE: At present the principal publication outlet for these
data Journal of Physical and Chemical Reference
the
JOURNAL OF RESEARCH—The Journal of Research
is

Data (JPCRD) published quarterly for NBS by the Ameri-

of the National Bureau of Standards reports NBS research
can Chemical Society (ACS) and the American Institute of
and development in those disciplines of the physical and
Physics (AIP). Subscriptions, reprints, and supplements
engineering sciences in which the Bureau is active. These
available from ACS, 1155 Sixteenth St. N.W., Wash., D.C.
include physics, chemistry, engineering, mathematics, and
20056.
computer sciences. Papers cover a broad range of subjects,
with major emphasis on measurement methodology, and Building Science Series —
Disseminates technical information
the basic technology underlying standardization. Also in- developed at the Bureau on building materials, components,
cluded from time to time are survey articles on topics closely systems, and whole structures. The series presents research
results, test methods, and performance criteria related to the
related to the Bureau's technical and scientific programs. As
structural and environmental functions and the durability
a special service to subscribers each issue contains complete
citations to all recent NBS publications in NBS and non-
and safety characteristics of building elements and systems.
NBS media. Issued six times a year. Annual subscription: —
Technical Notes Studies or reports which are complete in
domestic $17.00; foreign $21.25. Single copy, $3.00 domestic; themselves but restrictive in their treatment of a subject.
$3.75 foreign. Analogous to monographs but not so comprehensive in
scope or definitive in treatment of the subject area. Often
Note: The Journal was formerly published in two sections:
serve as a vehicle for final reports of work performed at
Section A "Physics and Chemistry" and Section B "Mathe-
matical Sciences."
NBS under the sponsorship of other government agencies.

DIMENSIONS/NBS —
Voluntary Product Standards Developed under procedures
published by the Department of Commerce in Part 10,
This monthly magazine is published to inform scientists,
Title 15, of the Code of Federal Regulations. The purpose
engineers, businessmen, industry, teachers, students, and
of the standards is to establish nationally recognized require-
consumers of the latest advances in science and technology,
ments for products, and to provide all concerned interests
with primary emphasis on the work at NBS. The magazine
with a basis for common understanding of the characteristics
highlights and reviews such issues as energy research, fire
of the products. NBS administers this program as a supple-
protection, building technology, metric conversion, pollution
ment to the activities of the private sector standardizing
abatement, health and safety, and consumer product per-
organizations.
formance. In addition, it reports the results of Bureau pro-
grams in measurement standards and techniques, properties Consumer Information Series —
Practical information, based

of matter and materials, engineering standards and services,

on NBSresearch and experience, covering areas of interest
to the consumer. Easily understandable language and
instrumentation, and automatic data processing.
illustrations provide useful background knowledge for shop-
Annual subscription: Domestic, $11.00; Foreign $13.75 ping in today's technological marketplace.
Order above NBS publications from: Superintendent of
NONPERIODICALS
Documents, Government Printing Office, Washington, D.C.
—
Monographs Major contributions to the technical liter- 20402.
ature on various subjects related to the Bureau's scientific
and technical activities.
Order following NBS publications —NBSlR's and FIPS from
the National Technical Information Services, Springfield,
Handbooks —Recommended codes of engineering and indus- Va. 22161.
trial practice (including safety codes) developed in coopera-
Federal Information Processing Standards Publications
tion with interested industries, professional
and regulatory bodies.
organizations,
(FIPS PUB) —Publications in this series collectively consti-
tute the Federal Information Processing Standards Register.
Special Publications—Include proceedings of conferences Register serves as the official source of information in the
sponsored by NBS, NBS annual reports, and other special Federal Government regarding standards issued by NBS
publications appropriate to this grouping such as wall charts, pursuant to the Federal Property and Administrative Serv-
pocket cards, and bibliographies. icesAct of 1949 as amended, Public Law 89-306 (79 Stat.
—
Applied Mathematics Series Mathematical tables, man- 1127), and as implemented by Executive Order 11717
uals, and studies of special interest to physicists, engineers, (38 FR 12315, dated May 11, 1973) and Part 6 of TiUe 15
chemists, biologists, mathematicians, computer programmers, CFR (Code of Federal Regulations).
and others engaged in scientific and technical work. NBS Interagency Reports (NBSIR)— A
special series of
—
National Standard Reference Data Series Provides quanti- interim or final reports on work performed by NBS for
tative data on the physical and chemical properties of outside sponsors (both government and non-government).
materials, compiled from the world's literature and critically In general, initial distribution is handled by the sponsor;
evaluated. Developed under a world-wide program co- public distribution is by the National Technical Information
ordinated by NBS. Program under authority of National Services (Springfield, Va. 22161) in paper copy or microfiche
Standard Data Act (Public Law 90-396). form.

BIBLIOGRAPHIC SUBSCRIPTION SERVICES

The following current-awareness and literature-survey bibli- Superconducting Devices and Materials. A literature survey
ographies are issued periodically by the Bureau:
issued quarterly. Annual subscription: $30.00. Send subscrip-
Cryogenic Data Center Current Awareness Service. A litera-
ture survey issued biweekly. Annual subscription: Domes- tion orders and remittances for the preceding bibliographic
tic,$25.00; Foreign, $30.00.
services to National Bureau of Standards, Cryogenic Data
Liquified Natural Gas. A
literature survey issued quarterly.
Annual subscription: $20.00. Center (275.02) Boulder, Colorado 80302.
U.S. DEPARTMENT OF COMMERCE
National Bureau of Standards
Washington, D.C. 20234
POSTAGE AND FEES PAID
U.S. DEPARTMENT OF COMMERCE
OFFICIAL BUSINESS
COM-215
Penalty for Private Use, $300

SPECIAL FOURTH-CLASS RATE

BOOK