Not to be confused with Concentrating Solar Thermal Power (CSP), Concentrating Photovoltaics (CPV) systems use mirrors or lenses to focus the sun’s light onto a small area of photovoltaic material. By focusing the sun’s light (usually by several hundred times, but potentially by up to 1000 times) in order to reduce the amount of expensive semiconductor material that is needed to produce a usable quantity of energy, and so reduce the overall costs of the system.
While CPV has had its supporters for many years, it has so far made little impact on the global photovoltaic market. Nonetheless, despite this relatively slow progress, a series of recent developments and the possibility of rapidly falling costs mean that the sector is once again attracting attention.
The basic principle of CPV is to focus sunlight onto a small area of photovoltaic material. This is done through the use of lenses (often a Fresnel-type lens is used), or mirrors arranged in parabolic dishes or troughs. The concentration ratio can vary: if the concentrated sunlight falls onto a well designed CPV cell, the cell will produce at least 10 times, or 100 times, the electricity. In fact the conversion efficiency of solar cells increases under concentrated light, so the correlation can be greater than one-to-one, depending on the design of the solar cell and the material used to make it.
At present CPV systems are divided into three broad categories – low (<10x), medium (up to around 150x) and high (>200x) concentration. Most utility-scale commercial systems operate at these higher concentrations. In the future, even higher concentrations are likely to become increasingly common. Importantly, since CPV can only operate efficiently in direct sunlight with a relatively low angle of incidence they require tracking systems to make sure that they stay focused on the sun throughout the day.
At low concentrations, usually with single-axis trackers, CPV generally makes use of silicon cells. However, concentrating PV also offers the option of shifting away from crystalline silicon to use the very high-efficiency, non-silicon cells. Such cells have mostly been developed primarily for space applications. These multi-junction III-V cells (which use elements from columns three and five of the periodic table, typically gallium and arsenide) are prohibitively expensive for extensive use in large flat panel arrays.
Concentrator systems, however, because they require far smaller and fewer cells, can afford the higher cost of multi-junction cells and yet still be manufactured at an acceptable dollar-per-watt cost. Indeed, combined with dual-axis tracking, the high efficiencies of these cells (typically 35%-40%) can enable the CPV units to operate at high module efficiencies, up to 27% (compared to a typical module efficiency of 20% for crystalline silicon or 12% for thin-film cadmium-telluride). This can greatly help reduce the cost per kWh produced.
At present there are several main manufacturers of III-V solar cells for use in CPV installations. Perhaps the most widely used is Spectrolab, a division of Boeing and a company that regularly sets records for cell efficiency. Other companies include Emcore, Azur Space and JDSU. Concentrix also manufactures its own cells. Aside from efficiency, there are a number of other factors which module manufacturers might take into account when sourcing cells for their systems. CPV is still a relatively new technology, and as such there is an element of higher risk perception associated with its deployment. Companies may be wary of using cells from smaller companies for fear that they might not be able to honour warranties in the event of a major programme. Bigger companies, with large backers, therefore have an advantage in terms of ‘bankability’ or risk reduction.
Nonetheless there is some evidence that the sector is finally gaining momentum with a number of utility-scale projects under development and a series of corporate tie-ups that could greatly improve the fortunes of the industry.
An important feature of CPV is its relatively low level of water use in operation. While it is important to cool CPV systems (the efficiency of a solar cell decreases with rising temperature) this is usually achieved by backing the cell onto a highly conductive metal, such as copper. Providing that this layer is sufficiently large it is usually sufficient to keep the system cool; additional cooling systems can also be employed. Given the pressure on water in many arid areas, and the cooling demands of concentrating solar thermal power (CSP) systems in particular, the ability to dry cool as standard is a clear additonal advantage in the future fortunes of CPV.
Land use is also something that many CPV manufacturers are particularly keen to emphasize. Per GWh, high-concentrating CPV uses 2.2 acres (0.9 Ha) of land compared to 4 acres (1.6 Ha) for cadmium telluride thin-film and 4.2 acres (1.7 Ha) for a solar thermal power tower. While this could provide an advantage in terms of permitting an environmental clearance, so far it appears to have little impact on the economics of these systems, since land costs typically represent a small part of the overall system costs. However, while this is certainly true in the western US, it may not be quite so inconsequential in southern Europe where land costs are higher and there is far less space available. In India too, with its massive overcrowding and demands on space, this is likely a positive advantage in developing large-scale solar systems.
STATE OF THE MARKET
To date the CPV market has been something of a fringe concern, accounting for less than 0.1% of installed photovoltaic capacity worldwide. According to GTM Research, total installed capacity at the beginning of 2011 amounted to just 28 MW, out of more than 33,000 MW of installed photovoltaic capacity worldwide, with only modest progress in 2010. According to GTM, total installations in 2010 amounted to just a few MW (although according to another consultancy – Strategy Analytics – this may have been up to 16.3 MW). Nonethless there are reasons for optimism, and at the time of writing there are 689 MW of CPV capacity under construction and in various stages development around the world – overwhelmingly in the USA, but also in Spain, Portugal, Mexico and Australia (of this 689 MW, some 414 MW has agreed a power purchase deal or been signed up to a feed-in tariff, so these are perhaps more definite). There are also a number of potential large scale projects in India and South Africa, although these are still uncertain.
Currently three companies – Soitec (which recently acquired Concentrix Solar), Amonix and SolFocus – dominate the module market, accounting for 96 percent of total projects in operation or in the pipeline. As the sector expands, however, it is likely that new players will begin to emerge, and there may be consolidation among existing companies as major investors move into the market.
SOITEC / CONCENTRIX
Of this ‘big three’, and despite installing just 2 MW in 2010, Soitec has been dominating the CPV headlines, starting in late 2009 with its acquisition of the German company Concentrix Solar. With a 25-MW manufacturing facility in Freiburg, Concentrix is one of the pioneers of CPV, using Fresnel lens reflectors to focus the sun’s energy up to 500 times onto III-V multijunction silicon solar cells and, achieving fully installed system efficiencies of up to 25 percent. In early 2011 Soitec commissioned its first utility-scale CPV power plant, a 1-MW facility at a Chevron mining facility in New Mexico, constructed in 2010. Covering 20 acres and made up of 173 units, it was the first in a number of important developments for Soitec in the North American market. For more on this project see our Large Scale Solar supplement published together with our May-June edition.
Next, in early 2011 Soitec announced that it had struck a deal to supply the 150-MW Imperial Solar Energy Centre CPV installation for Tenaska Solar Ventures, and that a subsidiary of Tenaska has reportedly already signed a power purchase agreement with San Diego Gas and Electric (SDG&E). Shortly after the Tenaska deal, Soitec signed another series of agreements directly with SDG&E. The first of these was to provide systems for three more solar power plants, totalling 30 MW, followed by two more deals with a total capacity of 125 MW. As with the 150-MW installation, construction work is due to begin in 2013, with the completion estimated in 2014-2015. To support the development of the market, Soitec has announced plans to construct a new 200-MW per year manufacturing facility in San Diego, as well as developing another 50 MW manufacturing plant in its home town of Freiburg.
While this brings the potential pipeline for Soitec to some 305 MW in North America alone, there appears to be some doubt over some of these plans, and the original 150-MW Imperial Solar Energy Centre agreement in particular. The uncertainty surrounds Tenaska’s application for a loan guarantee from the US Department of Energy, a programme that appears to be under financial pressure, with projects being warned that they may not be successful. If these loan guarantees are not made available, it is possible that Soitec will not establish its 200 MW manufacturing plant, placing the larger projects in doubt. Nonetheless, Soitec appears to have a number of solid projects under development and by 2015 it seems likely that it will have well over 100 MW of commercial CPV up and running.
Although the US is the main market in the immediate future, the potential for CPV extents far beyond the deserts of New Mexico and California. The Middle East and North Africa (MENA) region is a prime candidate for CPV development and Concentrix was among the first CPV companies to join the Desertec Initiative – a consortium of financial and industrial interests seeking to develop solar energy on a vast scale in the deserts and sunny regions of North Africa, the Middle East and Mediterranean Europe. Indeed, Soitec already has a number of small-scale test projects underway in the Middle East. In Egypt, five projects totalling 30 kW are up and running to the west of Cairo in the Wadi Natrun region (where essential minerals for the preservation of ancient Egyptian mummies were mined). In Jordan, a 6-kW facility is operational while there are also two small projects in Oman and Abu Dhabi.
While Soitec may have made the news, one of its key rivals, Amonix, has installed the bulk of the systems. Altogether Amonix has around 19 MW of systems operational worlwide, and has already installed several multi-megawatt CPV facilities, notably 7.8-MW, 1.5-MW and 2-MW plants in Spain that were completed in 2008 and 2009, and several smaller utility-scale projects in the United States. In 2010 its total installations amounted to around 15 MW.
Like Soitec, California based Amonix uses high-efficiency III-V multi-junction cells, this time manufactured by Spectrolab, a division of Boeing. Under concentrating conditions, Amonix states that these cells can achieve conversion efficiencies of nearly 40 percent, and AC system efficiencies of up to 29 percent.
In early 2010, Amonix teamed up with Flextronics and began construction of a 150-MW manufacturing facility in Las Vegas, Nevada, which was completed in early 2011. Just prior to this it also announced the completion of a 2-MW CPV facility at the Solar Zone of the University of Arizona’s Science and Technology Park, constuction of which began in 2010.
As well as completing a number of projects and opening a new manufacturing plant, Amonix has also been taking important steps to try and reduce the risk to investors of opting for CPV.
Finally, in early 2011 Amonix announced that it had been selected to supply the 30-MW Alamosa project in Colorado, under development by Cogentrix (not to be confused with Concentrix, now Soitec). The project managed to secure a $90 million loan guarantee from the Department of Energy, and the power will be sold to utility group Xcel Energy.
Looking to the future, of the 41 MW of CPV actually under construction, GTM Research estimates that 85 percent are being constructed by Amonix, with another 75 MW in the pipeline.
Like its two rivals, SolFocus also installed some megawatt-scale installations in 2010, at Victorville and Hanford in California. Along with smaller projects in the USA, Australia and Mexico, this brought its installations for the 12-15 months up until May 2011 to around 2.7 MW. With additional operational projects in Spain, Italy, Hawaii and the western US, SolFocus has installed well over 3 MW of operational CPV.
Reducing risk for investors has also been a feature of SolFocus’ strategy and they have managed to secure technology performance insurance from Munich Re, one of the world’s largest reinsurance companies. This is the first insurance of its kind to be issued to a CPV manufacturer and will provide an additional level of protection for SolFocus, allowing it to reduce its technical risk. It should also help to reassure potential customers of the robustness of its product, since the insurance was only awarded after a thorough inspection of its manufacturing process and technologies.
Following on from this move, SolFocus has been awarded a number of high profile contracts, including a 1-MW facility with Bechtel to provide energy to pistachio growers in California (completed in April 2011), a 1.5-MW project on the Greek island of Crete and a potential 5-MW system in Portugal.
Finally, the company has announced that it is to provide modules for Saudi Arabia’s largest CPV installation to date, under development by the Vision Electro Mechanical Company. Altogether SolFocus has a development pipeline of at least 39 MW, the vast majority in North America.
ECONOMICS OF CPV
As with all forms of renewable energy the first question anybody usually asks is ‘What does it cost?’. With an emerging sector like CPV there is currently no simple answer. The big hope for CPV is that by using smaller amounts of photovoltaic material at high efficiencies it will be able to drive down costs and compete with fossil fuels – a hope shared by thin-film PV and concentrating solar thermal.
At present, CPV still has some way to go, although it does have some factors in its favor. Certainly, in terms of installation costs per kW, CPV is far from the cheapest. According to GTM, the pre-profit cost for a high-concentrating multi-junction system is roughly $3.35/W installed, compared to $2.04/W for thin-film CdTe and $2.52/W for polysilicon.
What really matters, though, is the cost per kWh produced, and here things are a little better for CPV – it can operate at capacity factors of up to 26 percent, compared with 20 percent for CdTe. Once that is taken into account, CPV is considered to be broadly competitive with non-concentrating PV.
Looking to the power purchase agreements that have been signed, it is clear that CPV companies are being ambitious, with Amonix, for example, signing power purchase agreements with utilities for less than $0.10/kWh, due to be delivered in the next couple of years. At this level it seems unlikely that the CPV companies will generate much profit, but by getting their technology out there and proving it they should be able to increase their market share and so improve their margin.
Research and development and increasing economies of scale are also likely to play an important role in the coming years. Indeed, GTM Research’s predictions on the size of the market for 2015 are predicated on there being a 30 percent reduction in total installed cost by then. More efficient manufacture, transport and handling have been identified as key areas of development, along with increases in cell efficiency, which will allow the module manufacturers to reduce the amount of photovoltaic material required for a given output.
PREDICTIONS – MARKETS AND COSTS
The coming years are likely to see stellar growth for the CPV sector, with some estimates suggesting that the sector could be installing up to 1 GW per year by 2015 – a growth of almost 200 percent per year. It is worth bearing in mind, though, that even if this comes to pass, CPV will still represent just a small proportion of the more than 25-42 GW of annual PV capacity that the European Photovoltaic Industry Association predicts will be installed in 2015.
It also seems likely that, for the next few years at least, the United States will remain the largest market for CPV technology thanks to its high insolation, long-term power purchase agreements and demand for clean power. According to GTM, if California is to meet its Renewable Portfolio Standard then it may need to install 26 GW of solar. So far power purchase agreements for just 13 GW have been signed across the whole of the Southwest US (of which around half is CSP), leaving plenty of room for growth. The real question is whether CPV can lower its costs and gain enough customers to compete in the increasingly crowded solar marketplace.