info:eu-repo/semantics/doctoralThesis
Beam splitting filters aplied to thermally de-coupled solar photovoltaic/thermal (PV/T) collectors
Autor
Crisostomo Muñoz, José Felipe
Institución
Resumen
Effcient utilisation of solar energy has been intensively addressed by solar researchers
in recent decades. Many efforts have been made to improve PV cell efficiency, but they
are ultimately limited by their bandgaps in terms of utilising the solar spectrum. In
contrast, solar thermal collectors can achieve efficient utilisation over the full spectrum,
but thermal energy is of relatively low value compared to electricity.
As a combination of these systems, hybrid photovoltaic/thermal (PV/T) collectors have
the potential to use the whole spectrum and - if thermally de-coupled - they can also
provide a reasonably high value output. Most of the PV/T studies in the literature
propose coupled systems where the heat transfer fuid is in contact with the PV cells.
In the present research, however, spectral beam splitting is explored as a means to
thermally de-couple the thermal receiver (to achieve > 100°C to meet the demands of
the industrial heat market) and to operate the PV cells at low temperatures (to prevent
efficiency deterioration).
To efficiently utilise the entire spectrum, a general methodology for the optimal par-
tition of the solar spectrum in hybrid PV/T collectors was developed herein. This
methodology proposes an objective function which corresponds to the ratio between
the weighted sum of the electrical and heat outputs in the hybrid configuration to the
electrical output from a stand-alone concentrated photovoltaic system (of the same PV
cell type). The results reveal that Si PV cells are the best candidate; with around 45%
more theoretical weighted power delivery in comparison with a Si concentrated PV sys-
tem in the ideal case with perfect optics. Furthermore, the methodology indicates that
the optimal partition corresponds to a band stop/pass window rather that a cut-o
method to utilize both high and low energy photons optimally.
To experimentally validate this concept, two different approaches have been investi-
gated: (1) dichroic mirrors as reflective band-stop filters and, (2) nanofluids as band-
pass filters. The design of (1) was addressed by using two different multilayers struc-
tures: SiNx/SiO2 and TiO2/SiO2. For (2), Ag/Si core/shell nano-discs were suspended
in water creating a spectrally selective absorbing fluid. These methods were chosen be-
cause they have potential to ultimately be mass produced at low fabrication cost.
SiNx/SiO2 filters were tested in an indoor experiment revealing that the PV cells illu-
minated with the reflected light from the filters operate on average at 9.2% absolute
higher efficiency than the same cells without the filter. Furthermore, for the best filter,
the measured hybrid output is about 9% higher than the electrical output of a PV cell
exposed to the same light source without beam splitting. TiO2/SiO2 filters were tested
under concentrated sunlight with the main goal to investigate them as an add-in for
a commercial solar thermal rooftop concentrating collector. The results indicate that
only for low temperatures (<50°C) does the hybrid configuration deliver more power
and perform better than the original collector.
Due to the fact that thin film filters are hard to fabricate at large scale, an optical/heat
transfer model was developed to estimate the hybrid output using selective absorbing
fluids in a hybrid PV/T configuration. Also, an experimental set-up was designed
and built to validate the model and demonstrate the feasibility of such fluids. Four
Ag/Si nanofluid solutions were fabricated and characterised. The measured spectral
transmissivity of the most concentrated solution was found to be very close to the
ideal target for Si PV cells. Furthermore, the nanofluids were tested in real operation
conditions where the thermal and electrical outputs agreed very well with the model.
The experimental results revealed that up to 12% more power can be obtained from
the hybrid system compared to the measured output from a PV cell system under the
same concentration ratio.
Finally, a comparative economic model was developed to quantify the savings that could
be achieved from PV/T collectors using real operation data. The model was run for the
proposed collector configurations revealing that the hybridisation of a concentrated PV
system using nanofluids as selective absorbing fluids could be an economically viable
option. The results for the best nanofluid solution indicate that an extra savings of
up to 90 AUD/m2 can be achieved by the hybrid collector compared to the PV only
configuration. This corresponds to the maximum extra cost that should be paid for
hybridisation of this particular system.
Overall, this work details a general methodology for the solar spectrum partition in
concentrated PV/T collectors using beam splitting. Additionally, it provides a com-
prehensive optical and heat transfer model which together with the experimental and
economic analysis will enable further development of thermally de-coupled PV/T tech-
nology.