The Global Community for Sustainable Energy Professionals
1 hour / Central European Time (Brussels, Paris, Berlin)
JC Miñano, P. Benitez
During the last decade, there has been an impressive advance in the efficiency records of PV solar cells due to the continuous improvement of III-V multi-junction solar cells, whose efficiency has surpassed 40%, and with a slope of 10% absolute per decade. However, this great potential is achieved at a high cost per unit of cell area and, consequently, the future of these cells in the terrestrial market is linked to High Concentration Photovoltaics (HCPV).
The development of commercial products in HCPV systems has progressed rapidly in the last few years, and such development involves multidisciplinary science and engineering to be combined and globally optimized. One of the key components is the Optics, whose capability to provide the necessary tolerances and efficiencies is crucial (and, surprisingly, is often underestimated), particularly at high concentration levels.
In this webinar, a description of recent, state-of-the-art advances and the future of concentrating optical technology was presented, and the comparison between HCPV and its main competitor at present (flat PV tracking systems) was discussed.
Comments
Please find here below the summary of questions and answers.
TECHNOLOGY ISSUES
Q: Is HCPV useful for heating or only for electricity generation?
A: There are some developers working on this concept (for instance, Menova Energy in Canada has a combined parabolic trough (thermal) + CPV design, see http://www.power-spar.com/Power-Spar/index.php). Since PV cell efficiency tends to be lower at higher temperatures, this co-generation of electricity and heat is not well adapted for most of the industrial processes (which need heat at high temperature). Even though the temperature coefficient of the III-V cells efficiency is small compared to other technology options (as silicon), high temperature operation is not recommended for reliability issues. The additional complexity added by the heating system can be justified in PV roof installations where there may be a close need for hot water at medium-low temperature.
Q: I worked on CPVs in the 1980s. Are heat dissipation and CTE mismatch still problems?
A: Heat dissipation is still a big issue. The bottleneck in the heat flow is usually the contact between the cell receiver and the heat sink. The point is to conciliate good thermal contact and electrical insulation. That’s why this interface uses to be the most critical from the thermal point of view. To improve the thermal contact between cell and heat spreader it is necessary a good mechanical contact on this interface and this implies that materials at both sides of the interface must have similar Coefficient of Thermal Expansion (CTE).
Q : Who is developing a system with dish and CPV in the dense array method? Is it suitable with additional thermal use?
A: Solar Systems (http://www.solarsystems.com.au/) in Australia has developed such a system. They have used Spectrolab cells (http://www.spectrolab.com/) and active cooling.
Q: What materials are used to make the freeform secondary optics?
A: Glass is preferred in general. Plastic is a secondary candidate, because of possible degradation due to high radiation. Hollow secondaries can also be used made of conformed reflective metallic foils.
Q: What are the major technical barriers in making a quantum leap towards the thermodynamic limit of 86% and the current efficiencies of 27%?
A: It has to be considered the fact that 86% limit assumes a perfect optic system (100% efficient) and infinite cell junctions. In practice, we can increase the number of junctions, but the spectrum sensitivity increases too following daily and annual fluctuations if the cells are series connected. So it makes no sense going further 4 to 5 junctions. Reaching 50% efficiency cell is expected for the coming years but just with 1 or 2 additional junctions, no more. Spectral sensitivity can be improved in the future by making one of the junctions use the principle of Intermediate Band Solar cells (Luque et al), which in principle can behave as three equivalent junctions, one in parallel with the other two in series connection.
Q: What are the biggest technical hurdles that you see for CPV right now?
A: We think that CPV has now a unique opportunity to succeed. Concentration cells have reached a very high efficiency and a sufficient maturity level. Optics have also reached a top level with the new high-concentration, high-tolerance designs that we have done. Trackers have also reached a satisfactory level of development.
The biggest technical hurdle now is to prove that CPV is suitable for mass-production. For this goal, a wide concentrator tolerance angle will be the key to success.
Q: From the point of view of the power converter (inverter) does CPV have any particularities to be considered?
A: I don’t see a particular effect of CPV for inverters.
Q: What are the temperatures encountered at the focal point at X1000 concentration levels?
A: The thermal design of the concentrator must be such that the junction to ambient temperature drop is not above 50-80 deg Celsius. This optimum temperature is somehow independent of the concentration level. It mainly depends on the heat sink cost and on the efficiency vs temperature curve.
These temperature drops at 1000x can be achieved with cost-effective passive cooling if the cells are small enough (<1cm2 approx)
MARKET AND COST REDUCTION
Q: How much the Optics costs in the total cost of the module? What the portion will be?
A: Depends basically on concentration level, between 15% and 30%. Note that an increase of efficiency of global system thanks to better optical system allows levelized cost of energy reduction.
Q: What are the prospects for reducing the cost of Multiple Junction cells?
A: In the range of 10–20 $/cm2 with the prospect of cost reduction to around 10 $/cm2. Industry is still young and scaling is coming now, so there is room for cost reduction, but it remains difficult to guess the future price level. Besides that, it seems logical that further improvements in the cell technology to increase the convesion efficiency will probably be done at the expense of a higher cell cost.
Q: How fast do you think the market for CPV will develop?
A: CPV has had low level of development for long time (20 years). Now we are seeing a fast development at private companies level. The coming 3-4 years will allow more maturity and there will be more visibility on the market share and expectations for CPV.
Q: When do you think the first utility-scale CPV application will hit the market?
A: There are already CPV power plants in operation, such as those of Guascor Fotón company in Spain (about 12 MW and planned 10 MW additional). ISFOC has installed 3 MW and other companies such as EMCORE expect to install more than 200 MW from 2009. A dedicated feed-in tariff for CPV would help much. As for bigger power plants (>50 MW), it is difficult to predict when this technology will be ready and mature. Have a look to http://www.slideshare.net/sustenergy/bulk-solar-power-generation-sp-and-cpv-technologies?src=embed
Q: Do you think that the complexity in manufacturing of the modules could be a bottleneck for their market success?
A: This becomes a bottleneck unless you have a good optical design. That is why we have stressed the importance of the tolerance angle. This angle has to be wide enough to make room for many aspects related with module manufacturing (optical surface errors, concentrator unit assembling, array assembling, tracking errors, tracker stiffness etc.) For instance, a design with a tolerance angle of ±0.5 degrees leaves no room for manufacturing tolerances: ±0.27 degrees are used by the sun’s angular extension and consequently there is a mere ±0.23 degrees for optical surface errors, concentrator unit assembling, array assembling, tracking errors, tracker stiffness, radiation scattering by dust particles, etc. This tight tolerance angle makes the manufacturing prohibitively expensive due to the necessary high precision. Such ±0.5 degrees design is only good for a prototype.
We at LPI (http://www.lpi-llc.com/, for more detailed info please contact to info@lpi-europe.com ) have a unique set of design and manufacturing tools to achieve the greatest tolerances, very close to the thermodynamic limit (in particular our SMS design tool). That is why we achieve the highest tolerances even with the highest concentration levels.
Q: Would you please contrast HCPV vs Si and thin film in terms of land area required (hectares or acres) per MW?
A: Have a look to http://www.slideshare.net/sustenergy/bulk-solar-power-generation-sp-and-cpv-technologies?src=embed. This presentation states how a CPV system thanks to its increased efficiency can reduce land area requirement by 30% (comparing a flat PV system 14% efficient and a CPV system 25% efficient). However, comparison is not easy as the reference for installed power is not the same in flat PV and CPV. Expected power density in the tracker area is 250 W/m2 for CPV, 170 W/m2 for the most efficient flat PV and 50W/m2 for amorphous silicon thin film.
Q: What is the state of art of HCPV in terms of your figures of merit ($/KWH, KWH/m2/yr)?
A: Solfocus announced the installation of 10 MW at a price of $10/W. However, ISFOC in Spain made a call to bid and price of the various companies (Solfocus, Isofotón, Concentrix…) was around 6 – 6,5 €/W (complete PV system). In South Spain this allows a levelized cost of energy around 30 c€/kWh. The expectation is a cost reduction to 4 €/W in 2010 and 3 €/W in 2013, being in this case under the price expectation of flat PV. Have a look to : http://www.leonardo-energy.org/drupal/taxonomy/term/664
Q: Have you looked at rooftop sized panels being proposed by Soliant or EnFocus. As these compete w non-tracking flat panel, do you expect these to be competitive in terms of cost of electricity produced?
A: CPV for rooftop has been proposed by several companies at present and in the past (SEA, Sol3G). Competitiveness related with flat PV is not clear for the time being, even though they compete with the retail electricity price. Integrating tracking equipment in buildings is usually considered difficult at this moment.
Q: Morgan Solar has developed Light-guide Solar Optic. Have you included that in your ECT space?
A: We have not, but it can be included and the fundamental limitations still apply. The thermodynamic limit, as stated in the ETC space, only assume that there is no frequency shift between the incoming radiation and the radiation leaving the concentrator towards the cell and, apparently, the LSO is not doing frequency shift. Consequently the limitations also apply.