A new method for the measurement of solar collector

time constant
H.J. Hou
a,b
, Z.F. Wang
c
, R.Z. Wang
a,b,
*
, P.M. Wang
c
a
Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University, Shanghai 200030, China
b
Engineering Center for Solar Power and Refrigeration, MOE China, Shanghai 200030, China
c
Solar Energy Lab, Institute of Engineering Thermophysics, Chinese Academy of Sciences,
Beijing 100080, China
Received 24 April 2004; accepted 13 August 2004
Available online 30 November 2004
Abstract
A new test method about the time constant of the solar collector has been presented in this paper. It
is simple and has been validated through experiments. With the new method it is not necessary to
adjust the inlet temperature of the transfer uid as closely as possible to the ambient air temperature.
Also, it is not necessary to know the characteristic parameters of the collector in advance. The model
used in the paper is a rst order system model, as in most cases. The experimental data obtained from
the test of solar collector time constant shows that the solar collector is not a strictly rst order
system. A criterion is proposed to decide whether the system is a rst order system or not and the
resemblance of the system to the rst order system.
q 2004 Elsevier Ltd. All rights reserved.
Keywords: Solar collector; Time constant; First order system
1. Introduction
As one kind of solar energy absorbers, solar collectors usually work outdoors, so they
are subjected to unsteady weather conditions. For example, when a cloud shades the sun
abruptly, the incident solar energy on the collectors abruptly reduces to zero, thus a
0960-1481/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2004.08.005
Renewable Energy 30 (2005) 855865
www.elsevier.com/locate/renene
* Corresponding author. Address: Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University,
Shanghai 200030, China. Tel.: C86 21 6293 3838; fax: C86 21 6293 3250.
E-mail address: rzwang@sjtu.edu.cn (R.Z. Wang).
______________________________________________________ http://www.paper.edu.cn
transient driving force is generated on the solar collectors. The collectors then begin to
operate under a transient condition. Thus besides to steady state or quasi-steady state,
transient process also needs to be studied. Time constant is one of the most important
parameters of solar collectors thermal performance under transient condition. It indicates
the response characteristics of solar collector under transient forces. Therefore the solar
collector time constant test is one important part of the thermal performance tests.
Although the time constant of the solar collector can be obtained through the two
established standards: ISO 9806-1 and ASHRAE 93-86 standard, many limitations are
imposed by them. In the two standards, we need to adjust the temperature of the transfer
uid at the inlet T
f,i
as closely as possible (preferably within G1 8C) to the ambient air
temperature T
a
otherwise ISO 9806-1standard is invalid or another parameter of the solar
collector F
R
U
L
is needed to meet the requirements of ASHRAE 93-86 standard.
2. Test methods about solar collector time constant
Relevant solar collector time constant tests procedures are different in different
standards [1,2], which deal with thermal performance test of solar collectors.
In ISO9806-1 standard [1] the test procedure for collector time constant is described by:
1. Testing shall be performed either outdoors or in a solar irradiance simulator.
2. During the test the solar irradiance on the plane of the collector aperture shall be greater
than 800 W m
K2
.
Nomenclature
A system amplication coefcient ()
A
c
collector aperture (m
2
)
C
p
specic heat of the heat transfer uid (J/(kg 8C))
F
R
solar collector heat removal factor ()
G solar irradiance (W/m
2
)
_ m mass ow rate of heat transfer uid (kg/s)
(mc)
e
effective heat capacity of the solar collector (J/(m
2
8C))
T time constant (s)
T
a
ambient air temperature (8C)
T
f,i
collector inlet temperature (8C)
T
f,o
collector outlet temperature (8C)
U
L
solar collector heat loss coefcient (W/(m
2
8C))
T
f,o,initial
collector outlet temperature at the beginning of time constant test period
(8C)
T
f,o,N
collector outlet temperature at time tZN(8C)
Greek
(ta) effective transmittance absorptance product ()
t time (s)
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 856

______________________________________________________ http://www.paper.edu.cn
3. The heat transfer uid shall be circulated through the collector at the same ow rate as
that used during collector thermal efciency tests.
4. The temperature T
f,i
shall be set approximately equal to the ambient air temperature T
a
.
5. Initially, the aperture of the collector shall be shielded from the solar radiation by
means of a solar-reecting cover. When a steady state has been reached, the cover shall
be removed and measurements continued until steady state conditions have been
achieved again. The time constant of collector T is dened as the time taken for the
collector outlet temperature to rise by 0.632 of the total increase from T
f;o;t
KT
a
j
tZ0
to
T
f;o;t
KT
a
j
tZN
:
In ASHRAE 93-86 [2], the rst three steps are the same as that ISO9806-1 standard.
While step 4 for ISO9806-1 is not necessarily for ASHRAE 93-86.
Then the step 5 states:
The collector is opposed to an incident solar ux of greater than 800 W m
K2
. When a
steady state has been reached, the incident solar energy is then abruptly reduced to zero by
either shielding the collector from the sun or shutting off the solar simulator.
Simultaneously, the temperature T
f,i
and the temperature T
f,o
(t) are continuously
monitored. If the temperature T
f,i
is set approximately equal to the ambient air temperature
T
a
during the test period the time constant of collector T is the time required for the
quantity T
f;o
tKT
f;i
=T
f;o;initial
KT
f;i
to change from 1.0 to 0.368. Otherwise, T
fi
sT
a
during the test period, the time constant T will be the time required for the quantity
A
c
F
R
U
L
T
f;i
KT
a
C _ mc
p
T
f;o
t KT
f;i

A
c
F
R
U
L
T
f;i
KT
a
C _ mc
p
T
f;o;initial
KT
f;i

to change from 1.0 to 0.368. Here as we can see the parameter F

R
U
L
is introduced and
needs to be known in advance.
3. Theoretical basis
When the collectors work under the condition in which the heat transfer uid moves
through the collector and carries the useful energy gained away. The energy balance
equation is:
mc
e
dT
f;m
dt
ZA
c
F
0
S KU
L
T
f;m
KT
a
K _ mc
p
T
f;o
KT
f;i

ZA
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o
KT
f;i
(1)
where SZGta: The model is based on the Hottel Whillier Bliss equation [3] with
correction term of thermal capacitance.
When a steady state for the collector has been reached, if the solar irradiation G is
suddenly changed and then held constant, the solar collector behavior will change from
one steady state to another. During the transient process, assuming that the temperature
T
f,i
, T
a
, wind speed and mass ow rate of the heat transfer uid _ m are kept constant, then
the rate of change of the transfer uid exit temperature with time is related to the rate of
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 857

______________________________________________________ http://www.paper.edu.cn
change of transfer uid average temperature with time by [4]:
dT
f;m
dt
ZK
dT
f;o
dt
(2)
Thus, the energy balance equation of transient process is changed as
mc
e
K
dT
f;o
dt
ZA
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o
KT
f;i

T
f;o
ZT
f;o;initial
t Z0
_
_
_
(3)
Eq. (3) can be rearranged into the following:
mc
e
K
_ mc
p
dT
f;o
KT
f;i

dt
CT
f;o
KT
f;i
Z
A
c
F
R
_ mc
p
S KU
L
T
f;i
KT
a

T
f;o
ZT
f;o;initial
t Z0
_
_
_
(4)
The solution of Eq. (4) is:
A
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o
t KT
f;i

A
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o;initial
KT
f;i

Zexp K
_ mc
p
t
Kmc
e
_ _
(5)
In ISO9806-1 standard, by controlling the temperature T
f,i
to be almost equal to T
a
, thus
the Eq. (5) becomes
A
c
F
R
S K _ mc
p
T
f;o
t KT
a

A
c
F
R
S K _ mc
p
T
f;o;initial
KT
a

Zexp K
_ mc
p
t
Kmc
e
_ _
(6)
Under the former steady state condition, the solar irradiance on the plane of the collector
aperture was adjusted to be equal to zero. Thus, the equation T
f;o
ZT
f;i
ZT
a
is established
during the period. Accordingly the initial condition of the transient process between the two
steady state conditions is: T
f;o;t
j
tZ0
ZT
f;o;initial
ZT
f;i
ZT
a
: Then Eq. (6) becomes
T
f;o
t KT
a
Z
A
c
F
R
S
_ mc
p
1 Kexp K
_ mc
p
t
Kmc
e
_ _ _ _
(7)
When t/Nthen T
f;o;N
KT
a
ZA
c
F
R
S= _ mc
p
; thus
T
f;o
t KT
a
T
f;o;N
KT
a
Z 1 Kexp K
_ mc
p
t
Kmc
e
_ _ _ _
(8)
The time constant of the collector is the time required for the quantity
T
f;o
tKT
a
=T
f;o;N
KT
a
to change from 0 to 0.632. Whereas T
f;o
sT
a
during the former
steady state condition that means the solar irradiance on the collector is not zero at this time.
The time constant of the collector should be the time required for the quantity
T
f ;o
KT
a
j
t
KT
f ;o
KT
a
j
t0
T
f ;o
KT
a
j
tN
KT
f ;o
KT
a
j
t0
to change from 0 to 0.632. This is the test method of collector time constant described in
ISO9806-1 standard.
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 858

______________________________________________________ http://www.paper.edu.cn
In ASHRAE 93-86 standard, under the former steady state condition, the incident solar
energy is controlled greater than 790 W m
K2
. Then it is abruptly reduced to zero and kept
constant during the following period of the transient process. Thus the Eq. (5) becomes
A
c
F
R
U
L
T
f;i
KT
a
C _ mc
p
T
f;o
t KT
f;i

A
c
F
R
U
L
T
f;i
KT
a
C _ mc
p
T
f;o;initial
KT
f;i

Zexp K
_ mc
p
t
Kmc
e
_ _
(9)
If T
f;i
ZT
a
is valid, Eq. (9) becomes
T
f;o;t
KT
f;i
T
f;o;initial
KT
f;i
Zexp K
_ mc
p
t
Kmc
e
_ _
(10)
The two Eqs. (10) and (9) are the test methods of collector time constant described in
ASHRAE 93-86. The time constant of the collector is the time required for the left hand
terms of the Eqs. (10) and (9) to change from 1 to 0.368. Under the two conditions the
temperature T
f,i
is either adjusted to be close to the ambient air temperature T
a
or not
respectively.
By observing Eq. (4), we nd it has the same form as that of a rst-order system has.
T
dy
dt
Cy ZAx (11)
Eq. (11) is the typical differential equation of a rst-order system [5]. Where y is the
system output, x is the system input, T is the time constant of the object, A is the system
amplication coefcient. They are expressed as
y ZT
f;o
KT
f;i
x ZS KU
L
T
f;i
KT
a
T Z
mc
e
K
_ mc
p
A Z
A
c
F
R
_ mc
p
(12)
where T is the time constant of the collector, which is dened in the two standards
mentioned above.
4. New test method
In many cases, it is difcult to adjust the temperature T
f,i
to be almost equal to the
ambient air temperature T
a
. For example during winter in Beijing, the water temperature is
always higher than T
a
. In order to obtain the time constant under this condition, besides the
method described by the ASHRAE 93-86 standard, a new method is presented.
The method is simpler than the one described by ASHRAE 93-86 standard. For the
new method, it is not required to have T
f,i
equal to T
a
. The detailed formula derivation is
given below.
By adding TZmc
e
K= _ mc
p
to Eq. (5) and making rearrangement of it, it is then
changed as
_ mc
p
T
f;o
t KT
f;i
KT
f;o;initial
KT
f;i

A
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o;initial
KT
f;i

Z1 Kexp K
t
T
_ _
(13)
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 859

______________________________________________________ http://www.paper.edu.cn
Thus,
T
f;o
t KT
f;o;initial
Z
A
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o;initial
KT
f;i

_ mc
p
1 Kexp K
t
T
_ _ _ _
(14)
If tZN;
T
f;o;N
KT
f;o;initial
Z
A
c
F
R
S KU
L
T
f;i
KT
a
K _ mc
p
T
f;o;initial
KT
f;i

_ mc
p
(15)
Then
T
f;o
t KT
f;o;initial
T
f;o;N
KT
f;o;initial
Z 1 Kexp K
t
T
_ _ _ _
(16)
Here as we can see, the time constant of the collector is the time required for the
quantity T
f;o
tKT
f;o;initial
=T
f;o;N
KT
f;o;initial
to change from 0 to 0.632.
In this method, we used the Eq. (5) directly so we need to meet the requirement that
described by the paragraph before Eq. (2). According to the requirement, during the test
period, the solar irradiation G can be changed into the two modes described in the two
established standards and the temperature T
f,i
, T
a
and the mass ow rate _ m keep constant.
It is not necessary to adjust the temperature T
f,i
to be equal to T
a
as it is needed in
ISO9806-1. Also it is not necessary to know the F
R
U
L
in advance while it is necessary in
ASHRAE 93-86.
5. Experimental verication
The tests were conducted outdoors in Beijing, China. During the experiment the
weather was clear. The measurements were made on a glazed at plate collector and the
heat transfer uid was water. The aperture area of the at plate collector was 1.2 m
2
.
Various transducers about _ m; G, T
f,i
, T
f,o
and T
a
are connected to a data acquisition
instrument interfaced to a computer. The sampling rate is set as 4 s for each data. During
the test period the ambient air temperature was around 30 8C. The ow rate of heat transfer
uid was adjusted to 1 l/min.
In order to validate the new method, three experiments were conducted. During
Experiment 1 and 2, the temperature T

was regulated at 38 8C, the solar irradiation G was

changed into the two modes described by ISO9806-1 and ASHRAE 93-86, respectively.
As T
fi
sT
a
and F
R
U
L
is unknown, the time constant of the solar collector cannot be
obtained by the two standards mentioned above. But it can be obtained by the new method.
Experiment 1. When the solar irradiance G was changed as was required in ISO9806-1
standard, the ow rate _ m; the ambient air temperature T
a
, the temperature T

and T
f,o
of the
collector were continuously monitored at the same time until the next steady state
established. Here, the equation T
f;o
ZT
f;o;N
is established. For the purpose of this test, a
steady state condition is assumed to exist when the temperature T
f;o
varies by less than
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 860

______________________________________________________ http://www.paper.edu.cn
0.05 8C per minute. By plotting
R
1
Z
T
f;o
t KT
f;o;initial
T
f;o;N
KT
f;o;initial
as a function of time in Fig. 1, the time constant of the collector is 76 s. The temperature
T
f,o
as a function of time is plotted in Fig. 2.
Experiment 2. The solar irradiance G was changed as required by ASHRAE 93-86
standard and the ow rate _ m; the ambient air temperature T
a
, the temperature T

and T
f,o
of
the collector were monitored continuously. By plotting
R
1
Z
T
f;o
t KT
f;o;initial
T
f;o;N
KT
f;o;initial
as a function of time in Fig. 3. The time constant of the collector is 73 s. The temperature
T
f,o
as a function of time is plotted in Fig. 4.
Experiment 3. The time constant of the collector was tested according to the method
described by ISO9806-1. During the test period the temperature T

was adjusted
Fig. 1. R
1
as a function of time for Experiment 1.
Fig. 2. T
f,o
as a function of time.
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 861
______________________________________________________ http://www.paper.edu.cn
approximately equal to the ambient air temperature T
a
. The time constant of the collector
is 78 s. The temperature T
f,o
as a function of time is plotted in Fig. 5.
6. Results and discussion
The experiment results showed that the three time constants obtained by the new
method are approximately equal to each other.
By observing Eqs. (8) and (16), taking natural logarithm for both sides, the two
equations become Eqs. (17) and (18), respectively
ln 1 K
T
f;o
t KT
a
T
f;o;N
KT
a
_ _
ZK
_ mc
p
t
Kmc
e
(17)
ln 1 K
T
f;o
t KT
f;o;initial
T
f;o;N
KT
f;o;initial
_ _
ZK
_ mc
p
t
Kmc
e
(18)
Fig. 3. R
1
as a function of time for Experiment 2.
Fig. 4. T
f,o
as a function of time.
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 862
______________________________________________________ http://www.paper.edu.cn
If
T Z
Kmc
e
_ mc
p
;
Y Zln 1 K
T
f;o
t KT
f;o;initial
T
f;o;N
KT
f;o;initial
_ _
;
and
Z Zln 1 K
T
f;o
t KT
a
T
f;o;N
KT
a
_ _
are dened, then the Eqs. (17) and (18) becomes:
Y ZK
t
T
(19)
Z ZK
t
T
(20)
It indicates that the variables Y, Z and the time t are of linear relationship. By
using the experiment data of Experiments 1, 2, and 3, Figs. 68 could be plotted.
Fig. 5. T
f,o
as a function of time.
Fig. 6. Yt relationship corresponding to Experiment 1.
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 863
______________________________________________________ http://www.paper.edu.cn
Figs. 68 have shown the variables Y, Z and the time t are not of linear relationship
apparently.
The reason is that the solar collector model used here is a rst-order model as in most
cases [1,2,68]. But in fact it is not a strictly rst order system, a solar collector includes
many energy storage elements, such as working uid in absorber tubes, absorber tubes,
etc. it is a multi-capacity object. An n-order model is better than a rst order model. A rst
order system model is just a simplied model. The deviation of the experiment data to its
linear regression curve reversely indicates the resemblance of the system to the linear
system. In other words, the less deviation, the more likely to be a rst order system.
7. Conclusions
1. A new method for testing the time constant of the solar collector was validated. During
the tests it was not necessary to adjust the temperature T

to be equal to T
a
as it was
needed in ISO9806-1 and it was also not necessary to know the F
R
U
L
in advance like in
ASHRAE 93-86.
Fig. 8. Zt relationship corresponding to Experiment 3.
Fig. 7. Yt relationship corresponding to Experiment 2.
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 864

______________________________________________________ http://www.paper.edu.cn
2. In fact a solar collector is an n-order system. To consider it as a rst order system is just
a simplied model.
3. A criterion is presented to decide whether a system is a rst order system or not and the
resemblance of the system to the rst order system.
References
[1] ISO 9806-1. Test methods for solar collectors. Part 1: thermal performance of glazed liquid heating collectors
including pressure drop, 1994.
[2] ANSI/ASHRAE 93-method of testing to determine the thermal performance of solar collectors. New York:
ASHRAE Inc.; 1986.
[3] Dufe J, Beckman W. Solar engineering of thermal processes, 2nd ed. New York: Wiley; 1991.
[4] Simon F. Flat-plate solar-collector performance evaluation with a solar simulator as a basis for collector
selection and performance prediction. Solar Energy 1976;18:45166.
[5] Zhang FW, et al. Thermal system process control, analysis, design and debugging. Nanjing: Southeast
University Press; 1999 [in Chinese].
[6] Muschaweck J, Spirkl W. Dynamic solar collector performance testing. Solar Energy Mater Solar Cells 1993;
30:95105.
[7] Amer EH, Nayak JK, Sharma GK. Transient method for testing at-plate solar collectors. Energy Convers
Manage 1998;39(7):54958.
[8] Amer EH, Nayak JK, Sharma GK. Anewdynamic method for testing solar at-plate collectors under variable
weather. Energy Convers Manage 1999;40:80323.
H.J. Hou et al. / Renewable Energy 30 (2005) 855865 865

______________________________________________________ http://www.paper.edu.cn