See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228900053

Modelling of Cavity Receiver Heat Transfer for the Compact Linear Fresnel
Reflector

Conference Paper · January 2003

CITATIONS READS

8 450

4 authors:

John D. Pye Graham L. Morrison

Australian National University UNSW Sydney
126 PUBLICATIONS   660 CITATIONS    227 PUBLICATIONS   3,446 CITATIONS   

SEE PROFILE SEE PROFILE

M. Behnia David R Mills

Macquarie University The University of Sydney
342 PUBLICATIONS   5,210 CITATIONS    98 PUBLICATIONS   2,356 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Heliostat Development View project

Solar fuels View project

All content following this page was uploaded by John D. Pye on 05 March 2014.

The user has requested enhancement of the downloaded file.

 MODELLING OF CAVITY RECEIVER HEAT TRANSFER
FOR THE
COMPACT LINEAR FRESNEL REFLECTOR

John D Pye, Graham L Morrison and Masud Behnia

School of Mechanical and Manufacturing Engineering, University of New South Wales, Sydney, NSW 2052, Australia
Phone +61 2 9385 4093, Fax +61 2 9663 1222, [email protected]

David R Mills
Department of Applied Physics, University of Sydney, Sydney, NSW 2006, Australia
Phone +61 2 9351 3311, Fax +61 2 9351 7725, [email protected]

Abstract – Development of a cavity receiver for the Compact Linear Fresnel Reflector is presented.
Response to changes in ambient temperature and external convection coefficient, as well as changes in
cavity depth and width have been studied, for the purpose of determining the design parameters for an
initial prototype plant. New heat loss correlation equations are provided.

1. BACKGROUND through the tubes. Tubes may be configured to pass

up and down the absorber several times if desired.
The Compact Linear Fresnel Reflector (CLFR), shown • Exit quality of steam planned to be 0.8 to avoid
in Figure 1 is a concept for a large-scale solar thermal instability problems associated with superheated
power plant. Its goal is to provide high-pressure steam steam at outlet and the thermal stresses induced by
from a system configuration that is cheaper than other short-cycle movements in the dry-steam point in the
designs. Key features are: tubes.
• Mirrors arranged in Fresnel reflector configuration to
obtain a concentration ratio of 23 at the absorber The CLFR is being financed by an Australian power
surface. Mirrors up to 200m long can be tracked by a station, who will use the generated steam to displace coal
single motor. consumption. This approach allows an affordable entry
• Rows of Fresnel reflector mirrors point at alternate into renewable energy for existing coal-power producers,
(parallel) linear absorbers, achieving higher ground and allows them to meet the mandatory renewable
area efficiency. energy targets set by the government of New South
• Receiver is an inverted, trapezoidal, linear cavity Wales . (Hu et al, 2003)
receiver with a window at the base of the cavity. The
cavity is at ambient pressure; the upper surfaces are The Direct Solar Steam (DISS) project at the
insulated. The material proposed for use as the Plataforma Solar de Almería has already produced
'window' has a transmissivity of 0.95 at solar experimental work, and has performed transient and
wavelengths. steady-state analysis of system performance.
• The initial prototype will have an absorber surface Experimental results published by DLR (Rheinländer
with an emissivity of approximately 0.49 at operating and Eck, 2002) can be used to validate the computation
temperature. The absorptivity is 0.9 to 0.95 at solar work on the two phase flow model in this case.
wavelengths.
• The absorber will consist of a series of parallel steel 2. PROPOSED PROTOTYPE
pipes, in which water will be boiled in a single pass
At present, the initial prototype receiver is under
construction in New South Wales, Australia. The design
of the prototype is as follows:

• 60 m long absorber
• Absorber surface width 500 mm
• Cavity depth 150 mm
• 12 rows of mirrors 1.84 m wide
• Absorber positioned at a height of 9.8 m
• Mirrors pivoted 1.4 m above ground

Figure 1: Compact Linear Fresnel Reflector (CLFR) sketch, Mirrors will be made of standard quality glass and will
showing a single absorber 'module' of what will ultimately be have a reflectivity of only 0.8 approximately. The
a many-module system with mirror rows interleaved in- mirrors are elastically formed to the required focal
between modules.
length. Mirrors are mounted on a sandwich support Global radiation data for the prototype site is
structure, in turn mounted on curved C-sections. A 60 m summarised in Figure 2. Clearness at the power station
length constructed in this manner will be of sufficient is high and maximum beam radiation reaches up to 1100
torsional stiffness to be driven from one end only. It is W/m² peak.
anticipated that up to 200m may be driven by a single
motor. A first mirror panel has been constructed and The present work has been required to assist in the
development is continuing. design of this prototype system. Computational
The absorber cover will be of a cheap material and will modelling for the heat losses from the cavity receiver
have the desired effect of allowing radiation to pass previously performed was for a wider absorber than that
through it while shielding the hot absorber from wind- currently planned for construction. New CFD models
forced convective currents. have been created and heat transfer correlations created
for the latest design.
The prototype is proposed to be configured as shown in
Figure 3. 3. HEAT LOSS MODEL
This minimal system will allow full boiling and full
condensation in the tubes safely through the use of a The purpose of this modelling was to specifically
pressure relief valve as well as sufficient quantity of simulate the conditions expected in the prototype system
water beneath the lower level switch in the steam drum under construction, as well as to examine the effects of
to fill the absorber tubes when the solar input drops. It ambient temperature changes on the heat loss.
will be possible to measure the performance of the We used computational fluid dynamics (CFD) to obtain
system by knowing the feedwater pump flowrate and the the steady-state performance of the system at a range of
vessel pressure. assumed absorber temperatures and ambient
temperatures, for a variety of cavity designs.
If the prototype performs satisfactorily, the project will By modelling the heat losses for a fixed absorber
be scaled-up by extending absorber lengths to 600 m and temperature, this heat loss model is made suitable for
adding increasing numbers of absorber rows. integration into a larger system model incorporating
insolation as well as heat transfer to the absorber tubes.
Heat loss mechanisms modelled are shown in diagram
Figure 4.
30
Month-averaged Daily Global

25 Convection from the upper surface was ignored because

Radiation (MJ/m²/day)

in the numerical modelling this surface was assumed to

20 be a fixed temperature boundary. Such losses should be
15 allowed for when this model is integrated with the two-
phase flow model for the flow in the tubes.
10

5 3.1 Experimental Model

An experimental model of the heat loss from the cavity
0
1 2 3 4 5 6 7 8 9 10 11 12
was performed by Reynolds and Jance at the University
of New South Wales (Jance et al, 2000). A full-scale
Month
cavity of width 1.2m was constructed (see Figure 5) in
Figure 2: Solar radiation profile for prototype site, showing which heating elements were placed at the top of the
month-averaged daily global radiation (MJ/m²/day) plotted cavity to simulate the heating effect of concentrated solar
by month from January to December. Data derived from radiation. Internal convection was studied using flow
satellite data from the Australian Bureau of Meteorology visualisation inside the cavity (Reynolds et al, 2002).
Forgan (2001)

Pressure Relief
Psat at 260°C
Absorber tubes (120L capacity)

Steam Level switches Inlet water tank

drum
100 MPa
450L

Circulating
Feedwater
pump
pump
Figure 3: Schematic of prototype CLFR system
 Tenv
Absorber surface at set temperature Tenv
External convection
Table 1: Fixed parameters used in the simulations
External convection
εa
Constant Parameter Symbol Value
Internal conduction
hw Tc Internal radiation hw
Absorber emissivity in cavity a 0.49
εw Internal convection εw
εg
Wall emissivity in cavity w 0.1
εg
Emissivity of the window (internal and
external)
g 0.9

External convection External radiation Operating temperature of the cavity

Tenv Tsky (for Boussinesq approximation)
T op 370 K

Figure 4: Diagram showing heat transfer modes in the Overall external heat loss coefficient on
computation model for cavity heat loss cavity walls
hw 0.5 W/m²·K

3.2 Model Parameters Table 2: Varied parameters for the computational model
Table 1 shows the values of the fixed parameters that Varied Parameter Symbol Values
were used in the modelling. The data in Table 2 were Cavity depth 100, 200 and 300 mm
D
used to generate the 72 different model cases studied.
Notation for the cavity dimensions used is shown in Cavity width (at top) W 500, 1200 mm
Figure 6. Absorber temperature Tc 530, 570, 610 K

3.3 Solution method Ambient temperature Te 290, 305 K

Mesh: The computational model uses a mesh size of
Convection coefficient on outside
2mm, following the mess sensitivity carried out by of 'glass' cavity window
hg 2.6,10 W/m²·K
Reynolds et al. (2002). The structure of the grid used
here was rectangular between the parallel surfaces, and
triangular 'paved' under the sloping surfaces. The side symmetry, only one side of the cavity needed to be
walls are lined with a single row of 2mm rectangular simulated.
elements for improved convection modelling. By
Computation: The solution was found using the
segregated steady-state two-dimensional solver, with the
Boussinesq approximation used for convection effects.
The Discrete Transfer Model (DTRM) was used to
model radiation transfer within the cavity. Simple
average convection coefficients and constant-value
emissivities were used to model external losses. The
cavity cover was assumed to be opaque, for radiation
from the absorber because the transmissivity at the
radiated wavelengths will be low. The solution was
found using Fluent 6.0 software.

The radiation 'sky' temperature was taken to be 5 K

above ambient temperature to allow for the fact that the
cavity will be facing mostly mirrors which will be
opaque at the wavelengths emitted, and these mirrors
Figure 5: View from underneath the experimental rig. The top will have heated somewhat in the sun.
surface was electricall heated and convective and radiative Heat loss through the insulated side-walls was modelled
losses were observed using thermocouples and photographing simply as if it were external convection at a very low
of convected smoke in the cavity value of convection coefficient.

W For each of these models, 20,000 iterations were run.

N 3.4 Results
θ =30° D The overall heat loss from the absorber is shown in in
Figure 7. In this diagram, points are coloured by
absorber temperature, Ta. For each absorber depth and
B width at a given absorber temperature there are four
Figure 6: Cavity geometry
points, corresponding to a pair of values of external
 W=1.2 0.5 m dominant because, at T g ­ T e  ≈ 55 K and

0.2 0.3 m T g ­ T sky ≈ 50 K , then hg T    a T  . 4

D=0.1
For the cavity depths studied, the internal heat loss by
radiation easily dominates over the convection and
hg=10 conduction heat loss, with Q conv at most reaching 8%
of Q tot . This maximum occurs in in the deepest
hg=2.6 (D=300 mm) of the narrow (W=0.5 m) cavities.
D=0.3 m Radiation losses were shown to closely follow the form
0.1 0.2

Qrad = F rad   a  T 4a ­ T 4g 
K

Figure 7: Total absorber heat loss per unit of absorber area where Frad is constant. This value can be seen in Figure
(W/m²) for both the W=1.2m and W=0.5 m absorbers. 9 to be approximately 0.90± 0.03 . A simple
convection coefficient (hg=10 W/m² are the upper two calculation of the two-dimensional view factor between
points) and a pair of values for ambient temperature. the absorber and the cavity cover for the geometry
studied gives values in the range of 0.925 to 0.985.
3.5 Observations These values are higher than the values of Frad observed
The process of heat loss from the absorber occurs as because of the non-uniform temperature profile on the
follows. From the hot absorber temperature, which has cavity cover (see Figure 10) and the interaction with
been heated by solar radiation heat is transferred to the convection processes inside and outside the cavity.
other internal surfaces of the cavity by a combination of
conduction, convection and radiation. Upon being Convection cells in the absorber cavity (see Figures 13,
conducted through to the external surfaces of the cavity, 11 and 12) are seen to occupy approximately the lower
heat escapes to the atmosphere via a second half of the cavity, with the upper half of the cavity
combination of radiation and convection. remaining as a relatively stratified conducting region,
The process of internal heat transfer is seen to be with only a small amount of sideways flow resulting
dominated by radiation effects. This is because the from convection down the cooler neighbouring side-
absorber temperature is much greater than that of the walls. This convection pattern results in the lower half of
other surfaces in the cavity, and radiative heat transfer is
proportional to differences in the fourth power of surface
temperatures. Externally, the process of convection is W=1.2 m W=0.5 m

W=1.2m
hg=10W/m²
D= hg=2.6W/m²
0.1 0.2 0.3m D= 0.3m
0.1 0.2

W=0.5m
Figure 9: Graph of radiation view factor, Frad, for all of the
cases studied. Points are coloured by cavity depth (as shown
on the legend, in metres), and are divided as indicated by
K
cavity width and external convection coefficient on the
Figure 8: Graph of the ratio (%) of convective heat loss at the window. The difference in Frad values for the two sets of
absorber to the total heat loss at the absorber, which shows external convection coefficient values clearly indicate the
the consistently low fraction of heat transferred from the interaction between convective and radiative processes in this
absorber by convection.The left group of points is for the wide system. Frad is therefore only a 'quasi' view factor because
absorber; points are coloured according to absorber radiative heat transfer is occuring between surfaces in a way
temperature. Scatter is due to differing ambient temperatures that depends on the temperature of the side-walls, in turn
and external convection coefficient on the window. dependent on convection parameters.
the cavity being at a relatively uniform temperature, and induce a more even temperature profile over the window,
a thermal gradient occurring in the upper half where but as the transmissivity of the proposed material is very
flow is low. The volume-weighted average cavity low at solar wavelengths, this effect is not likely to be
temperature is therefore less than the ½  T a T g  great. Any cover material must be chosen to withstand
the peak hot-spot temperature found from the
assumed in the correlation equations.
simulations.
Locations where fluid motion is fastest are down along
The fact that the radiation losses dominate the thermal
the side walls, then in along the window. The next-
losses inside the cavity suggest that the choice of sky
fastest fluid motion appears as a layer above the window,
temperature might be an important one. The current
with the fluid rising up from the window at the
work assumes a sky temperature of 5 K above ambient.
symmetry plane and moving back outwards towards the
This choice was made on the basis of mirrors being of
side-walls. This fluid mostly flows back to the side walls
low quality glass that will absorber significant heat in
and recirculates although some appears to enter into the
the sun. Also, the mirrors are opaque at the re-radiated
more stable stratified zone, replacing that taken by
wavelengths, and the design of the CLFR means the
convection from the top of the side-walls in that zone.
majority of the 'sky' as viewed from the cavity window
will consist of mirrors slightly off-focus. Incoming
The temperature of the side walls is consistently about
radiation has already been allowed for through the
30 K higher than the window temperature.
setting of the absorber temperature, so the hot sun does
This temperature might be expected to be higher,
not need to be allowed for in the sky temperature
however the view factors from the side walls to the
calculation, even though optical modelling shows a solar
absorber surface are low, so heating of this surface is
concentration ratio of between 8 and 20 occurring in the
primarily by conduction from the mostly stratified cavity
plane of the window (Buie, 2003). However, looking at
air. Heat loss through this surface is low and can almost
the external heat losses from the lower surface of the
be ignored but for its effect in initiating downward
collector, it is seen that most external heat loss is by
convection and setting up the symmetric convection
convection.
cells.
For the cavity with depth 200 mm, absorber width
0.5m, at Te=290 K, Ta=610 K, hg=10 W/m²K, modelling
At greater cavity depths (Figures 11 and 12), some
shows external heat transfer from the cavity cover to be
instability in the convection process arises. This occurs
1900 W/m due to convection and 1560 W/m due to
due to a localised hot-spot on the window surface about
radiation. In general, for the cases modelled, external
the symmetry plane. At this point, the local view factor
from the window to the absorber is highest and there is
also stagnation in the convection flow. Higher thermal
gradients therefore occur here and generate less stable
flow. It might be expected that the effect of concentrated
solar radiation passing through the cover material will Figure 13: Contours of stream function for W=0.5 m, D=100
mm cavity, Ta=610 K, hg=10 W/m²K, Te=290 K.

Figure 11: Contours of stream function for W=0.5 m, D=300

mm, Ta=570 K, Te=305 and hg=2.6 W/m²K. Note the irregular
flow at the symmetry plane.

Figure 10: Window temperature (K) versus distance (m) from

symmetry plane for W=0.5m, D=0.3m cavity. Notice localised
maxiumum in window temperature at the symmetry plane
where convection is low and the local view factor between
window and absorber are maximal. This variation also makes Figure 12: Contours of stream function for W=1.2 m, D=300
problematic the modelling of external radiation and mm, Ta=610, Te=305, hw=10 W/m²K
convection effects.
heat losses due to radiation and convection are of similar simplicity, the interaction effects noted there are ignored,
order of magnitude, unlike internal heat loss processes. because this would involve combining internal and
external effects in one equation. Hence, we use:
The results of earlier simulations (Reynolds et al, F rad = 0.90 (6)
2002) have shown that the optimum cavity depth was
around 100 – 150 mm. The present simulations do not Q rad = F rad   a  T 4a ­ T 4g  (7)
cover a wide enough range of data confirm that finding, The overall heat loss from the absorber is therefore the
however it is anticipated that the optimum depth will be sum of these radiative losses and convective/conductive
different for the narrower W=0.5 m cavity. More losses.
simulations are required to determine this. Q tot = Q rad  Q conv (8)

3.6 Correlation equations The above equations adequately model the processes
When considering how to formulate suitable internal to the cavity. In combination with simple
correlations, it was considered desirable, as far as equations describing the external effects, Eqs. (9), (10)
possible, to separate the internal and external heat and (11), an over model for the cavity heat losses is
transfer processes. Interaction effects will be limited to formed. The temperature of the cavity cover and the total
those which can be modelled by an equilibrium of heat transfer through the cover are the 'boundary
surface temperatures; correlations for internal cavity variables' between the internal and external models.
behaviours will not include any reference to external
parameters such as convection coefficient or ambient
temperature.
Q w , conv = N hw T c ­ T e  (9)

In order to calculate non-dimensional quantities, cavity Q g , conv = B hg T g ­T e  (10)

conditions must be evaluated at a suitable fluid
Qg , rad = B  g   T 4g ­T 4e 
temperature Tc. Previously a simple average of Tg and Ta
(11)
has been used (Reynolds et al, 2002), and this value is
again used here: Note that in order to avoid having to correlate for
1 another surface temperature, that of the cavity side-walls,
T c=
2
 T a T g  (1) the expression for the convective losses through the side
walls, Eq. (9) for Qw,conv assumes that the wall is
Eq. (2) below estimates the convective and conductive effectively at the temperature Tc. This is a false
heat loss from the absorber. assumption. However, the low convection coefficient due

 
0.6432
D to the insulated walls ensure that external heat losses
Nu= 1.1917Gr 0.10363 (2) from these surfaces are a small fraction of the total.
W
The above correlation gives total heat loss at the
The Nusselt number is defined by Eq. (3). Because Qconv
absorber to within 4.6% for all of the simulated cases.
takes the units of power per length of absorber (W/m),
Accuracy of the correlation is best for the wide cavity,
there are two length quantities in the equation.
for which the Nusselt correlation is more accurate.
Q conv /W
Nu=

 Dkc
T g ­ T a 
(3)

The Grashof number used in the correlation is that

given by the Eq. (4). As with kc in Eq. (3), βc and νc
must be evaluated at the cavity temperature used to W=0.5 m
generate the correlation given by Eq. (1).
9.8  c T a ­ T g  D3 W=1.2 m
Gr = (4)
2c
The above Nusselt and Grashof number correlation is
drawn from the data shown in Figure 14. The above
correlation does not adequately account for the scattering
in the W=0.5 m results (see Figure 15), so further work
on this correlation will be performed.
Figure 14: Nu versus Gr for the cases studied - see equations
The final correlation equation estimates the radiative (2) and (4). The upper groups of points are the the wider
heat loss. This heat loss was found to be close to constant cavity. The grouping of points in the horizontal direction is of
in all of the cases modelled (see Figure 9). For increasing absorber depth, D=0.1, 0.2, and 0.3 m.
 To improve this correlation, a simple surface-to-surface IAPWS is incorporated into the library to validate the
radiation transfer model assuming uniform surface code.
temperatures is under development to replace Eq (7).
Further, the systematic variation in Frad will be examined 5. FUTURE WORK
and a correlation developed, based on cavity-internal The outcome of the continuing work planned will
parameters if possible, and the correlation for Nusselt include:
number will be improved. • Higher accuracy will be sought for the absorber heat
The correlations here do no incorporate any of the loss correlations.
dimensionless parameters which could be formed using • Simulations at higher values of the window
the fixed-valued parameters from the simulations (Table convection coefficient, hg are required.
1). More model cases will be run to allow these • Correlated equations for the transient performance of
parameters to be incorporated into a more elaborate the system will be required in order to perform
correlation. TRNSYS modelling of the system with the revised
correlation for absorber heat loss.
4. HYDRODYNAMIC MODEL • Experimental results using the prototype system to
A model for the heat transfer from the outer surface of confirm heat loss, exit quality and pressure drop
the top of the absorber to water flowing through parallel results. The prototype is expected to be completed
tubes mounted on that surface is under development, later this year.
continuing the work previously done at UNSW
(Reynolds et al, 2002; Jance et al, 2000; Odeh, 1999). 6. CONCLUDING REMARKS
This model uses the above prediction of heat loss A project is underway for a Compact Linear Fresnel
combined with the predicted insolation from the Reflector (CLFR) prototype at a power station in New
concentrating Fresnel reflector and models the forced- South Wales. The work presented here relates to the
convection boiling process occurring in the tubes. It is modelling of heat loss processes in the cavity receiver to
able to predict the pressure drop through the system be used in the prototype. Based on previously obtained
based on the Martinelli-Nelson correlation for two-phase experimental data for a trapezoidal cavity receiver,
pressure drop, and also the exit quality of the steam at a computational fluid dynamics (CFD) models were
given flow rate. A simple validation of the model against developed for the purpose of predicting these heat losses,
results of the DISS project has been performed, and which are a combination of convective, radiative and
further validation will be possible from results of the conductive losses.
CLFR prototype. A number of simulations covering a reasonable range
At present, the model is a steady-state model. The need of design and environmental parameters were carried
for a full transient model is recognised, following the out.
work on flow stability by PSA and DLR for their DISS The heat loss processes are complex due to non uniform
system. (Rheinländer and Eck, 2002) temperatures on cavity surfaces and interaction between
Computer software has been developed for this work convection and radiation processes. An attempt made to
which may be of use to others, namely a set of object- develop design correlations for the prediction of heat
oriented C++ classes implementing the IAPWS 97 losses using suitable Nusselt and Grashof numbers for
industry-standard steam tables (Wagner and Pruss, 1995, convection, and an effective radiation view factor met
IAPWS, 1997, 1998; Spang, 2002). Test data from with reasonable success.
Continuing work will improve these correlations and
incorporate them in the hydrodynamic model for the
forced convection boiling in the absorber tubes.
TRNSYS modelling of the overall system will be carried
out, and the transient response of the overall system to
changes in solar radiation intensity will be studied.

NOMENCLATURE
Subscript Meaning
a Absorber surface at top of cavity
g Glazing on cavity cover at bottom of cavity
W Cavity side-walls
e Environment, ambient
sky Sky conditions for radiative heat transfer from
Figure 15: Correlation of Nusselt number. The correlation
the cavity cover
coefficient was 0.977. Maxiumum relative error was 15%.
Most scattering is apparent in the W=0.5m points. conv Convective plus conductive
 Subscript Meaning 10. IAPWS (1997), Release on the Viscosity of Ordinary
Radiative Water Substance,
rad
http://www.iapws.org/relguide/visc.pdf
tot Total 11. Jance M.J., Morrison G.L and Behnia M. (2000) ”Natural
Convection and Radiation within an Enclosed Inverted
Symbol Meaning Absorber Cavity”, Proceedings of ANZSES Annual
Qw,conv Convection heat transfer from the exterior of Conference - From Fossils to Photons, Brisbane, pp563-
the cavity side-walls per length of absorber 569.
12. Mills D.R. and Dey C.J., (1999) Transition strategies for
Qg,rad Radiation heat transfer from the exterior of the
solar thermal power generation. Proceedings of
cover glazing of the cavity, per length of
International Solar Energy Society Congress, Israel.
absorber
13. Mills D.R. and Morrison G.L. (1999) Modelling study for
Qg,conv Convection heat transfer from the exterior of compact Fresnel reflector powerplant. Solarpaces, Font-
the cover glazing of the cavity, per length of Romeu France, Journal de Physique IV, 9,159-165.
absorber 14. Mills D.R. and Morrison G.L. (1999) Compact linear
Fresnel reflector solar thermal powerplants. Solar Energy.
 Coefficient of volumetric expansion
68, pp. 263-283.
 Kinematic viscosity 15. Morrison G.L. and Mills D.R. (1999) Solar Thermal Power
Systems – Stanwell Power Station Project” Proceedings of
k Conductivity (W/mK) ANZSES Annual Conference, Geelong.
h External convection coefficient (W/m²K) 16. Odeh S. (1999), Direct Steam Generation Collectors for
Solar Electric Generation Systems, Ph. D Thesis,
T Temperature University of New South Wales.
Nu, Nu2 Nusselt number 17. Reynolds D.J., Behnia M. and Morrison G.L. (2002), A
Grashof number Hydrodynamic Model for a Line-Focus Direct Steam
Gr
Generation Solar Collector. Proceedings of Solar 2002,
Q Heat transfer be length of absorber (W/m) Australian and New Zealand Solar Energy Society.
Quasi view factor for radiation transfer inside 18. Reynolds D., Jance M., Behnia M. and Morrison G.L.
F rad
the cavity. (2001). An Experimental and Computational Study of the
Heat Loss Characteristics of a Trapezoidal Cavity
W,D,B,N Cavity geometry, see Figure 6. Absorber, Proceedings of International Solar Energy
Society 2001 Solar World Congress, Adelaide.
REFERENCES 19. Rheinländer J., and Eck M. (2002), Direct Solar Steam
(DISS): Numerical Modelling of Pressure Losses, DLR
1. Buie D., A solar and vector class for the optical modelling Internal Report.
of solar concentrating systems, Solar Energy, In Press. 20. Spang B. (2002), IAPWS Equations for Transport
2. Dey C.J., Mills D.R. and Morrison G.L. (2000). Operation Properties and Surface Tension of Water and Steam, The
of a CLFR research apparatus. ANZSES Annual Chemical Engineers’ Resource Page,
Conference 'From Fossils to Photons', Brisbane. http://www.cheresources.com/iapwsif972.pdf.
3. Duffie J.A. and Beckman W.A. (1991) Solar Engineering 21. Wagner W. and Pruss A. (1995), The IAPWS Formulation
of Thermal Processes, 2nd edn. Wiley Interscience, New 1995 for the Thermodynamic Properties of Ordinary Water
York. Substance for General and Scientific Use, J. Phys. Chem.
4. Eck M., Zarza E., Eickhoff M., Rheinländer J., Valenzuela Ref. Data, 31, 387-535.
L., Applied research concerning the direct steam
generation in parabolic troughs, Solar Energy. In Press.
5. Eck M. and Steinmann W. D. (2002) Direct steam
generation in parabolic troughs: First results of the DISS
project. ASME J. Sol. Energy Eng. 124, pp134-139.
6. Forgan, B (2002), GMS Global Solar Exposure Data
1990-2001 and Australian Radiation Network Data 1947-
2001, version NCCSOL 2.209, Australian Bureau of
Meteorology.
7. Hu E.J., Mills D.R., Morrison G.L., Le Lievre P., Solar
Power Boosting of Fossil Fuelled Power Plants,
Proceedings of ISES World Congress, Gothenburg, 2003.
In Press.
8. IAPWS (1997), Release on the IAPWS Industrial
Formulation 1997 for the Thermodynamic Properties of
Water and Steam,
http://www.iapws.org/relguide/IF97.pdf
9. IAPWS (1998), Release on the Thermal Conductivity of
Ordinary Water Substance,
http://www.iapws.org/relguide/thcond.pdf

View publication stats