Труды Всемирного конгресса Международного общества солнечной энергии - 2007. Том 2
Подождите немного. Документ загружается.
Proceedings of ISES Solar World Congress 2007: Solar Energy and Human Settlement
692
heat transfer coefficient of the closed cavity increased
linearly with input thermal load but had no effect by
inclinaton angle.
Fig. 5: Variation of the thermal resistances(open cavity).
Fig. 6: Variation of the thermal resistances(closed cavity).
Fig. 7: Variation of evaporation heat transfer coefficient
(open cavity).
Fig. 8: Variation of evaporation heat transfer coefficient
(closed cavity).
Fig. 9: Variation of condensation heat transfer
coefficient(open cavity).
Fig. 10: Variation of condensation heat transfer
coefficient(closed cavity).
3 SOLAR COLLECTOR TECHNOLOGIES AND SYSTEMS
693
4. CONCLUSION
The hybrid solar receiver was designed and manufactured
to supply the thermal energy to the stirling engine. The
hybrid receiver has produced thermal test by gas-fired heat
to investigate heat transfer characteristics for input thermal
load between 15 kW to 45 kW, inclination angle from 0 deg
to 30 deg and the aperture to cavity diameter ratios of 0 and
1.0. Comparison of the results with an open and closed
cavity has been constructed to measure temperature
variation in cavity surface and to analyze thermal
resistances, and the evaporation and condensation heat
transfer coefficient.
The hybrid solar power generation system, which could
utilize solar and combustion heat simultaneously, is
expected to help a maximum and optimum operation of the
system by realizing the constant operation using solar,
combustion, or a hybrid thermal energy depending on solar
conditions. It could have an important role in increasing the
economical efficiency of the system for the final
commercialization of the solar power system.
5. REFERENCES
(1) Laing D. and Palsson M., “Hybrid Dish/Stirling
System: Combustor and Heat Pipe Receiver
Development”, Journal of Solar Energy Engineering,
2002, Vol. 124, pp. 176~181.
(2) Laing D. and Reusch M., “Design and Test Results of
First and Second Generation Hybrid Sodium Heat Pipe
Receivers for Dish/Stirling Systems”, Proc. Int. Solar
Energy Conf., 1998, pp. 403~408.
(3) Duaainger P.M., “Design, Fabrication, and Test of a
Heat Pipe Receiver for the Cummins Power Generation
5 kWe Dish/Stirling System”, Proc. of the 29
th
Intersociety Energy Conversion Engineering
Conference, 1991, pp. 171~176.
(4) Hartenstine J.R. and Dussinger P.M., “Development of
a Solar and Gas-Fired Heat Pipe Receiver for the
Cummins Power Generation 7.5 kWe Dish/Stirling
System”, Proc. of the 29
th
Intersociety Energy
Conversion Engineering Conference, 1994, pp.
1935~1941.
(5) Moreno J.B., et al., “Recent Progress in Heat-Pipe Solar
Receiver”, Proc. of the 36
th
Intersociety Energy
Conversion Engineering Conference, 2001, pp.
565~572.
(6) Moreno J.B., et al., “Dish/Stirling Hybrid-
Heat-Pipe-Receiver Design and Test Results”, Proc. of
the 37
th
Intersociety Energy Conversion Engineering
Conference, 2002, pp. 556~564.
(7) Diver R.B., et al., “Trends in Dish-Stirling Solar
Receiver Designs”, Proc. of the 25
th
Intersociety
Energy Conversion Engineering Conference, 1990, pp.
303~310.
(8) Andraka C., et al., “Felt-Metal-Wick Heat Pipe
Receiver”, Proc. of the Int. Solar Energy Conference,
1995, pp. 559~564.
(9) Mancini T., et al., “Dish-Stirling Systems: An
Overview of Development and Status”, Journal of Solar
Energy Engineering, 2003, Vol. 125, pp. 135~151.
(10) Kang M.C. et al., “Thermal Characteristics of Hybrid
Solar Receiver using a Solar and Combustion heating”,
Korean Society for New and Renewable Energy, 2006,
Vol. 2, pp. 33~38.
A NEW DYNAMIC TEST METHOD OF THERMAL PERFORMANCE FOR
EVACUATED TUBE SOLAR COLLECTOR
Bowei Wang, Zhifeng Wang
Institute of Electrical Engineering, CAS
Beijing 100080, China
Wangbowei_2000@tom.com
Xing Li, Yi Ruan
Himin Solar Energy Group Co. Ltd
Dezhou City, China 253092
Lixing_1976@163.com
ABSTRACT
In the present paper, the thermal performance of evacuated
tube solar collector is investigated. A new dynamic model
for testing solar collector is analyzed. The model is derived
from thermal energy balance and solved by transfer
function. In the model the collector is like as a system, the
outlet temperature ( T
fo
) is the system output, the other
variables: the inlet temperature ( T
fi
), the ambient
temperature ( T
a
) and the solar irradiance ( I ) on aperture
plane per unit area are the system inputs. By the
measurements of T
fo
, T
fi
, T
a
and I the characteristic
parameters of the transfer function can be obtained. And
then the system output ( T
fo
) can be calculated by the
transfer function and the system inputs. The results on the
basis of the new method are very close to actual measured
values when the system inputs are highly variable.
1. INTRODUCTION
The knowledge of the thermal performance of a solar
thermal collector is essential for the prediction of the
energy output of any solar thermal system. There are many
test methods for the thermal performance characterization
of solar thermal collectors. Most well-known test methods
are under steady-state conditions such as ISO9806-1 and
EN12975-2. These steady-state test methods limit the
available outdoor testing time very much in variable
climates and make an outdoor test very expensive.
The new dynamic test method offers a much wider range of
collectors compared to the steady-state test method. Less
restrictions in the test requirement make it easier to find
periods for outdoor testing: this leads to easier and cheaper
tests, especially for places with varying climate conditions.
2. MODEL
The model is based on all glass evacuated tube solar
collector with U-tube. One evacuated tube collector is
shown in Fig.1
7
7
7
7
Fig.1: Schematic of one evacuated tube collector.
The model makes the following assumptions:
(1) All heat transfer processes are considered as one
dimensional.
(2) The glass is in the same temperature T
b
.
(3) Specific heat of copper is neglected.
(4) The water in U-tube is in the same temperature T
s
.
(5) The fluid flow rate in the collector is constant.
We can obtain energy balance equations from the glass and
water in U-tube [1].
3 SOLAR COLLECTOR TECHNOLOGIES AND SYSTEMS
695
() ()
dT
mC AI K T T K T T
dτ
=− −− −
(1)
()
()
dT
mC K T T mC T T
dτ
=−− −
(2)
The model assumes that T
s
is equal to the average of inlet
temperature and outlet temperature. Put T
s
=(T
fi
+T
fo
)/2 into
Eq.(1) then make Eq.(1) and Eq.(2) together with deleting
T
b
. We can get Eq.(3)
dT dT dT dT
A
BT A BT CI DT
dd ddττ ττ
++=−+++−
(3)
Where
2mC K
K
K
A
mC mC
+
+
=+
22()KKK mCKK
B
mmCC
++ +
=−
2
bs b
bsbs
KK
C
mmCC
−
=
2KK
D
mmCC
=
Here we make an assumption that A, B, C and D are
constants for they are not change much with temperature in
the process.
3. ESTIMATION OF PARAMETERS
Outdoor tests are done in Beijing. Throughout the course of
testing, only the fluid rate in collector is constant and mass
flow rate value is 0.02kg/m
2
s. Changing the inlet
temperature and irradiance, we record the inlet temperature,
outlet temperature, ambient temperature, solar irradiance on
aperture plane per unit area and flow rate. Test data are
shown in Fig. 2 and Fig. 3.
The Eq.(3) can be changed as follow:
()()
fo fi fi fo fo fi a
TT ATT BTT CIDT
′′ ′′ ′ ′
+= − − − +−
(4)
22
24
26
28
30
32
34
36
38
time (h)
temperature (
ć
)
720
740
760
780
800
820
840
solar radiation (w/m )
DPELHQWWHPSHUDWXUH LQOHWWHPSHUDWXUH
RXWOHWWHPSHUDWXUH VRODUUDGLDWLRQ
Fig. 2: Experimental measurements under changing inlet
temperature.
25
27
29
31
33
35
37
time (min)
temperature (
ć
)
0
100
200
300
400
500
600
700
800
solar radiation (w/m )
DPELHQWWHPSHUDWXUH LQOHWWHPSHUDWXUH
RXWOHWWHPSHUDWXUH V RODUUDGLDWLRQ
Fig. 3: Experimental measurements under changing solar
radiation.
The Eq.(4) can be calculated with central difference
quotient method , which is expressed as follow :
[]
(1)2() (1)yyn ynyn t
′′
=+− +−Δ
[]
(1) (1)2yyn yn t
′
=+−−Δ
The above experimental data are regressed with least squares
method, we can get the parameters values from Table1:
TABLE 1: PARAMETERS
PARAMETERS A B C D
VALUE 4.21 1.47 0.28 −0.07
Proceedings of ISES Solar World Congress 2007: Solar Energy and Human Settlement
696
We get the transfer function as follow:
*
4.21( ) 1.47( ) 0.28 0.07
fo fi fi fo fo fi a
TT TT TT I T
′′ ′′ ′ ′
+= − − − + +
where:
*
30
I
I=
4. COMPARISON WITH ACTUAL MEASURED VALUES
To certificate the results of new testing method, we can
make comparisons with actual measured values under
different conditions. Figs. 4-6 shows the comparisons.
20
21
22
23
24
25
26
27
28
29
30
time (min)
temperature (
ć
)
754
755
756
757
758
759
760
761
762
solar radiation (w/m )
DPELHQWWHPSHU DWXUH
LQOHWWHPSHUDW XUH
RXWOHWWHPSHUD WXUH
SUHGLFWHGRXWO HWWHPSHUDWXU HZLWKQHZ PHWKRG
VRODUUDGLDWLR Q
Fig. 4: Experimental measurements under steady-state
conditions and predicted outlet temperature with
new method.
5. CONCLUSION
According to the method, only the fluid flow rate needs to
be kept constant. System inputs such as the irradiance, inlet
temperature and ambient temperature can be variable
arbitrary. Compared with the steady state test of solar
collector, the new test method makes the cost of testing
much lower, it also makes the fit time for testing longer
20
22
24
26
28
30
32
34
time (h)
temperature (
ć
)
700
720
740
760
780
800
820
solar radiation (w/m )
DPELHQWWHPSHUDWXUH
LQOHWWHPSHUDWXUH
RXWOHWWHPSHUDWXUH
SUHGLFWHGRXWOHWWHPSHUDWXUHZLWKQHZPHWKRG
VRODUUDGLDWLRQ
Fig. 5: Experimental measurements under changing inlet
temperature and predicted outlet temperature with
new method.
20
22
24
26
28
30
32
time (min)
temperature (
ć
)
0
100
200
300
400
500
600
700
800
900
solar radiation (w/m )
DPELHQWWHPSHUDWXUH
LQOHWWHPSHUDWXUH
RXWOHWWHPSHUDWXUH
SUHGLFWHGRXWOHWWHPSHUDWXUHZLWKQHZPHWKRG
VRODUUDGLDWLRQ
Fig. 6: Experimental measurements under changing solar
radiation and predicted outlet temperature with new
method.
3 SOLAR COLLECTOR TECHNOLOGIES AND SYSTEMS
697
during one year. The results on the basis of the new method
are very close to actual measured values when the system
inputs are highly variable.
6. NOMENCLATURE
m
b
Effective mass of evacuated collector tube
m
s
Effective mass of fluid in the collector
C
b
Specific heat of evacuated collector tube
C
s
Specific heat of fluid in the collector
T
b
Evacuated collector tube temperature
T
s
Fluid temperature
A
b
Aperture area of the collector
I Solar radiation on collector per unite area
m
Mass flow rate
K
b-s
Heat-transfer coefficient between evacuated
collector tube and fluid
K
b-a
Heat-transfer coefficient between evacuated
collector tube and atmosphere
T
fo
Fluid outlet temperature
T
fi
Fluid inlet temperature
7. REFERENCES
(1) Hongjuan Hou, Research on dynamic thermal
performance test method of solar collectors” , 2005.
(2) J.A.Duffie, W.A.Beckman., “Solar engineering of
thermal processes”, 1991, Wiley, New York.
(3) ASHRAE 93-86, 1986. Methods of testing to determine
the thermal performance of solar collectors.
(4) EN 12975-2-2001. Thermal solar systems and
components solar collectors Part 2: Test methods.
(5) GB/T 4271-2000. Test methods for the thermal
performance of flat plate solar collectors.
EXPERIMENTAL STUDIES ON HIGHLY CONCENTRATED SOLAR
RADIATION BY USING FRESNEL LENS GROUP
Y.M. Han, R.Z. Wang, Y.J. Dai, A.H. Xiong
Institute of Refrigeration and Cryogenics, SJTU
Shanghai 200240, China
ymhan@sjtu.edu.cn, rzwang@sjtu.edu.cn, yjdai@sjtu.edu.cn
ABSTRACT
With advantages of lower volume, lower weight, smaller
focal length and lower cost, Fresnel lenses are suitable for
solar radiation concentration, energy control, and system
modularization. In this study, a modular device composed
of two-stage Fresnel lens is present and its performance is
experimentally investigated for different conditions.
Facular point image quality of Fresnel lens is investigated
by simulation for lens battery design. As for the Fresnel
lens with 100mm in diameter, 220mm in focal length, when
incident angle is kept within 1°, the focal point will not
move out of the receiver domain with the dimension of
10mm*10mm. when coupling with another Fresnel lens of
30mm in diameter, it has found that 50~60% of the
collected solar radiation can transmit through the Fresnel
lens group under condition of 1000 suns.
1. INTRODUCTION
The key point of popular applications of solar energy lies in
economic cost, which is inherently related to the
characteristics of solar radiation. The lower density of solar
radiation needs more investments on solar collection
devices, while concentration optics results in system
volume smaller and then the manufacturing cost lower on
solar concentrator and receiver, material saving and easy
modularization. Considering these advantages, Fresnel
lenses are adapt for solar radiation concentration, energy
control, and system modularization, especially, optical
efficiency of transmitted solar radiation is also improved
compared to the thick ordinary lenses.
Among the literatures on Fresnel lenses study, we find that
Fresnel lenses are designed to variable forms according to
the requirements when being used as solar concentrators,
such as convex linear Fresnel lens [1], flat linear Fresnel
lens [2], point focus Fresnel lens [3] etc. Study contents
include new exploitations on application such as
Photovoltaic and thermal storage [4], greenhouse [5],
daylight [6] and so on. Lately, for the advantages of light
weight and tolerance of shape errors, inflatable Fresnel
lenses are being developed for optical concentrators in solar
application systems.
In order to adapt to the requirement on the modular design
of highly concentrated solar energy, the optics systems are
characterized by easier reproduction, lower weight,
portable and moveable, convenient in combination and
disassembly, and these characteristics are very important
for cost saving.
The purpose of this study is to discuss possible ways to
further increase of concentration ratio and to easy
modularize the optical system. the module is designed into
two stage refractive optical system with composite Fresnel
lens group. In particular, optical properties, lens efficiency,
diameter of facular spot and its moving characters with the
sun, tracking precision are carried out under the condition
of concentration ratio is over than 1000 suns.
3 SOLAR COLLECTOR TECHNOLOGIES AND SYSTEMS
699
2. EXPERIMENTAL SETUP BUILDING
As shown in Fig.1a, the setup of the highly concentrated
solar radiation is designed into two main parts: racks of
Fresnel lens for one or tow stage, adjustment used for
control the distance between the two stages lens. The Rack
of lens module includes two layer solar panels, each with
four lens housing, Diameter of first stage lens rack has four
types of householders with 100mm, 100mm, 75mm, and 50
mm, diameter of second stage lens holder is 30mm. the
plate of top panel is made by aluminum alloy material,
adjustment penetrating the top panel acts as the executive
component of lens distance variation. The minimum
modular unit as designed with 2×2 optical collectors,
measuring 400×400mm2 can be distributed in a field at
balcony, hidden on a roof, or easily applied to a historical
building.
In this study, the Fresnel lenses used are all made of
polymethyl-methacylate (PMMA), the refraction index of
which is 1.5. Fresnel lens with 50 grooves per inch is
100mm in diameter, 220mm in focal length, which is
installed as the first stage in this experimental setup, as is
shown in Fig.1b. These following parameters are under
variation: focal distance, lens aperture area, and receiver
diameter. Modular design of whole system will benefit
from compact volume which can be placed on the any
sides exposed to sunlight, and partial load supplied for
user can be obtained by individual control of the required
modules. The secondary concentrator is also Fresnel lens
(in Fig. 1(c)) with 30mm in diameter, 30mm in focal
length, which embedded in the rack used for lens position
adjustment.
Fig. 1: Racks utilized for allocating the radiation (a) Lens
racks prototype; (b) first stage Fresnel lens; (c)
secondary stage Fresnel lens.
The adjustable rack of the secondary stage concentration
offers three dimensional fine-turning by micrometer screw.
As shown in Fig. 2(a), the two screws control X and Y
dimensional moving respectively, while the screw installed
at the bottom of the rack, shown in Fig. 2(b) fulfill the
micrometer adjustment in Z dimension. The precision is
kept at 0.1mm with central initial adjustment, and here
screws can move the lens hold in 0.01mm level. Fig. 2(c)
shows that the fiber and the focus regulation are
incorporated with the lens holder. The centering
mechanism is composed of a clamper holding the fiber, and
it is pushed by a leaf spring and in the opposite side by two
screws, the socket screw within a fiber clamper is
employed to fix the fiber in the correction position. With
the development of coupling technology, the alignment
between fiber and collecting optics cannot be critical if the
mechanical working is made with accuracy.
Fig. 2: Adjustable rack of the secondary concentration
connected with coupling optical fiber. (a) adjustable
lens holder; (b) socket screw with optical fiber
clamper; (c) coupling beam path (4 in 1).
The optical fiber material consists of 4×3m ends of silica
core clad fiber with a connector, and the four fibers,
coming out form the bottom panel, have a connector and
one line output at the other side, As Fig.2c shown, NA
(numerical aperture) of the fiber material is checked to be
0.36.
3. TESTS AND COMPARISONS
The experiments are performed with Fresnel lens to
examine the distribution of the radiation on the focal plane
of the lens, and this device is tested outdoors to examine
the illumination and thermal energy. The effects of solar
moving on the deviation of the focal position are measured.
In order to estimate thermal energy and luminance on the
receiver, we use power meter and illuminometer to record
Proceedings of ISES Solar World Congress 2007: Solar Energy and Human Settlement
700
the corresponding values. The Fresnel lens group(100mm
in diameter as the first stage, and 30mm in diameter as the
secondary stage), which is mounted on the movable solar
panel as mentioned before, is used to study the distribution
of converged solar rays on focal plane as a function of
incident angle and azimuth. At the central of the focal plane,
a black spot marked as the base point of the initial position
of converged solar rays.
3.1 Variation of the Facular Spot Position
As shown in Fig. 3, diameter of the observation window is
40mm, which is use to ensure the focus to fall into this
range. As the incidence angle increasing form 0 to 1°, the
position of the focus stars to deviate form the central mark,
Fig. 3: Flux distribution and deviation range of Fresnel lens
group, with f/2.2 for various incidence angle. (a) 0,
(b) 0.2, (c) 0.4, (d) 0.6, (e) 0.8, and (f) 1.0°, Black
spot represent the center mark of initial position.
and irradiance distribution and shape become asymmetric.
For incident angle 0, which correspond to solar solstices of
a location where the surface faces south and has inclination
equal to the latitude. The changes of the radiation profile
with the moving of the sun in about four and a half minutes,
and each step of 0.2° are shown respectively shown from (a)
to (e). When the incidence angle becomes 1° relative the
receiver plane, the azimuth range departures about 0.25mm,
and the outer edge of the facular spot begin to align to the
line left of the central line 0.5mm, that is, if the allowable
tracking precision is controlled within 1°, the maximum
area of the receiver is limited to 10×10mm
2
.
3.2 Optical Efficiency of Optical Components
For the whole spectrum of the solar radiation, the optical
efficiency is defined as the ratio of the emergency radiance
of the lens group and the incident rays radiance. The
performance of the Fresnel lens and optical fiber are study
in forms of decomposing and integrations.
TABLE 1: RESULTS OF EXPERIMENTAL TESTS
Specification
Testing Item
Diameter Focal length
Optical
efficiency
First stage 100mm 220mm 69.19%
Secondary stage 30mm 30mm 79.84ˁ
Two stage lens
group
30mm 199mm 55.24%
Infrared light
filter
Diameter: 30mm
Length :3000mm
49.15%
Laser power meters labeled FL250A-AH-V1 and
PD300-UV are used as the experiment instruments,
measure ranges are 200mW~250W and 10pW~300mW
respectively. Measurement accuracies are all limited within
2.0%. Tab.1 shows the experimental results of the Fresnel
lens group. According to the measurement, the optical
efficiency for the first stage Fresnel lens shows 69.19%, the
main reason lies in the accuracy of actual manufacture, the
simulation results are based on the presume of paraxial rays
incidence, energy loss from marginal rays occupies very
little of the total collected energy, while spherical
aberration and actual error lead to more diffuse effect
during rays transmission. When combined the secondary
3 SOLAR COLLECTOR TECHNOLOGIES AND SYSTEMS
701
Fresnel lens with optical efficiency 79.84%, the finally
obtained efficiency is 55.24% for the two stage lens group.
If this lens group is used for day-lighting, infrared rays
including heat effect should be blocked, visible rays are
required to pass through the interior of the buildings. If the
infrared light filter used, the whole efficiency falls to
49.15%, half of the incidental energy are lost due to the
diffusion and absorption.
Efficiency of the optical fiber is tested on the He-Ne laser
generator, rays focal length at 633nm are generated and
sent into the fiber entry surface from the one-end, four-end
outputs are measured respectively, and the optical
efficiency of each output end shows different with 5.64%,
16.92%, 11.65% and 13.10%. The efficiency of the four
outputs amounts to 47.31ˁ, that is, nearly half of the
collected energy are lost during the transmission in optical
fiber.
4. DISCUSSIONS AND CONCLUSIONS
In this study, modular Fresnel lens group for the high
concentration solar energy experimental platform is devised
and tested. The focal spot deviation and optical efficiency
are in consideration for two conditions: first stage only or
combined with the secondary stage. Concentration ratios
are about 500 suns or 2000 suns respectively.
Optical efficiency measurements have bee carried out on
actual Fresnel lens group. According to experimentally
study, positions of focal spot varying with the incident
angle of solar radiation are recorded continuously, and it
has found that nearly 50~60% of the collected solar
radiation can transmit through the Fresnel lens group under
condition of over 1000 suns. Test on optical fiber reports
that about half of collected solar energy is lost because of
absorption and diffusion.
Technologies on high concentration solar energy
application are expected to provide the modular production
of the concentration optics, which plays an important role
in cost saving in system exploitation. The key factors of
energy lost lies in Fresnel lens used and optical fiber
coupler, the modular design of Fresnel lens group user for
high concentration indicates feasible, optical efficiency can
be ensured within the incidental angle and the receiver area.
5. ACKNOWLEDGMENTS
We are deeply grateful to the financial support of the
Shanghai Expo special project No.2005BA908B07 and also
the key project for Fundamental Research of Shanghai
under the contract NO.06JC14040.
6. REFERENCES
(1) Nelson, D.T., Evans, D.L., Bansal, R.K., Linear Fresnel
lens concentrators. Solar Energy 17, 1975, p.285–289.
(2) Khalil, E.J. Al-Jumaily, Munadhil, A.K.A. Al-Kaysi,
The study of the performance and efficiency of flat
linear Fresnel lens collector with sun tracking system in
Iraq. Renewable Energy 14, 1998, p.41–48.
(3) G. R. Whitfield, R. W. Bentley, C. K. Weatherby, A. C.
Hunt, H. -D. Mohring, F. H. Klotz, P. Keuber, J. C.
Miñano and E. Alarte-Garvi , The development and
testing of small concentrating PV systems Solar Energy
67, 1999, p.23-34.
(4) Ralf Leutz, Akio Suzuki, Atsushi Akisawa, Takao
Kashiwagi, Shaped non-imaging Fresnel lenses.
Journal of Optics A: Pure Applied Optics, 2000,2,
p.112–116
(5) Tsangrassoulis, L. Doulos, M. Santamouris, M.
Fontoynont, F. Maamari, M. Wilson, A. Jacobs, J.
Solomon, A. Zimmerman, W. Pohl, et al., On the
energy efficiency of a prototype hybrid daylighting
system Solar Energ 79, 2005, p.56-64.
(6) Tripanagnostopoulos, Y., Siabekou, Ch., Tonui, J.K..
The Fresnel lens concept for solar control of buildings.
In: Proceedings of the International Conference
PALENC2005, Santorini, 2005, 19–21 May, Greece, pp.
977–982.