Solar power is expected to receive another boost this week, following a hectic couple of months during which the advocates of different solar technologies have been vying for the limelight – as well as trying to steal a march on their rivals. The drama has been unfolding in Australia where there are three technologies competing for commercial credibility. Two are based on solar thermal systems, which recover heat from the Sun to drive steam turbines. One uses parabolic troughs to focus sunlight onto tubes containing fluid, which heat up and create the steam; the other uses enormous dishes. A third technology relies on photovoltaic cells that use semiconductors to turn sunlight directly into electricity. Though most solar research has focused on photovoltaic cells, the high manufacturing cost of the cells has proved a commercial stumbling block.
Steam challenge
In May, however, the team that holds the record for designing the most efficient photovoltaic cells announced that it had found a way of making the cells more economically viable. Martin Green and his colleagues at the University of New South Wales in Sydney said that they could cut cost of a cell by 80 per cent by using poorer quality silicon. But these cells will need to be developed quickly if they are to keep pace with the advances being claimed for solar thermal systems.
Advertisement
Last month, such a system used an enormous hexagonal dish to turn sunlight into steam efficiently enough to challenge the performance of photovoltaic cells. The Big Dish, which stands on the campus of the Australian National University in Canberra, is the largest of its type in the world with an area of 400 square metres. Within 12 months, the ANU and a consortium of power companies plan to install up to 28 of the dishes at Tennant Creek in the Northern Territory, where they will supply up to 2 megawatts of electricity to the mining town.
Now it’s the turn of a rival solar thermal system, the parabolic trough, to take the stage. Solel, an Israeli-Belgian company that supplies power plants built from arrays of troughs, is expected to announce this week the deals it has struck with Australian manufacturers and power companies. Solel wants to start making solar equipment in Australia and to build a plant capable of generating 70 megawatts. Another of its plans involves setting up a research centre in Sydney in partnership with David Mills, an applied physicist from the University of Sydney, who has developed alloys that promise to extend the range of climatic conditions in which solar thermal plants can operate. Solel and its partners will also reveal details of their scheme for supplying solar thermal power for the Olympic Games in Sydney in 2000.
Solel claims to be the only company in the world capable of supplying large solar power plants, up to 300 megawatts. In 1992, it took over the technology developed by Luz, an Israeli company that went bankrupt the previous year. Luz built nine power plants in the Mojave Desert of southern California which, between them, can generate 354 megawatts of electricity – about 2 per cent of the capacity of the local grid. These plants account for about 50 per cent of the world production of solar power. The rest is largely supplied by hundreds of small photovoltaic systems, most supplying less than 1 megawatt.
Despite its financial difficulties, Luz showed that solar power on this scale could work on a large power grid. Its plants in California consist of long parallel lines of troughs, each about 100 metres in length and computer-controlled to rotate and follow the Sun during the course of the day. Each trough is made up of 224 glass mirrors and has a collecting area of 545 square metres. The parabolic shape of the mirror ensures that the light from all the mirrors reflects onto a receiver – a horizontal pipe, containing oil, that runs above the trough. The solar energy heats the oil to about 390 °C, and pumps then force the oil through a series of heat exchangers that use the recovered heat to convert water into superheated steam for driving a turbine. The largest of the nine plants built in California, an 80-megawatt facility, needs an array of about 900 troughs.
Solar ambitions
Solel now plans to start building power plants based on the Luz technology. The company has already announced that it will build a 200-megawatt plant in Israel, but wants to expand into Australia. By this weekend, Solel’s executives should have finalised plans to set up a factory in partnership with several Australian companies. In the first stages, this will supply solar thermal plants for specialist uses, such as generating steam to sterilise equipment in hospitals. Later, Solel expects to set up a factory capable of building fully-operating power plants. Asia will be a key market, especially as trough technology may prove well-suited to the tropics. Solel is also negotiating the establishment of three other power plants, each of between 200 and 300 megawatts, in North America and at sites on two other continents.
Another item on the agenda during Solel’s visit to Australia has been the 2000 Olympics. Sydney’s bid for the Games was supported by environmental groups, especially Greenpeace, in return for billing the event as the Green Olympics. Solel and Mills are setting up a consortium to promote solar thermal energy as the most environmentally friendly technology for supplying the estimated 10 megawatts of power that the Olympic village, hotels and main sporting venues will need.
Solel’s plans to build a 70-megawatt power station, which would be sited in either Queensland or the Northern Territory, rely on taking the technology that Luz established in California to the next stage. The Luz plant, explains Mills, is a hybrid, with about 25 per cent of the electricity supplied by natural gas. The gas superheats the steam to about 500 °C – a temperature better suited for use in conventional steam turbines – and provides a back-up in poor weather. This makes it difficult to put a price on the solar component of the electricity but, according to Mills, it is usually estimated to be 8 American cents per kilowatt-hour; competitors suggest the figure is closer to 14 cents per kWh. Before its demise, Luz predicted that the next generation of technology would reduce the cost to about 5.5 cents per kWh, putting it on a par with conventional coal and nuclear plants. However, several improvements are necessary to achieve this.
The first is to replace the oil in the pipes with water. Mills reckons this will conserve about 8 per cent of the electrical output because there will be no need to pump the oil through pipes to the spot where it superheats steam. Moreover, because water requires less elaborate heat exchangers this would save about another 2 per cent. The cost of the collectors could also be reduced by using metal or metal foils instead of glass to reflect sunlight. In the longer term, improvements to the design will make it possible for the collectors to capture more of the Sun’s radiation.
The Luz troughs focus sunlight onto a receiver consisting of an inner stainless-steel tube surrounded by a glass cylinder. A vacuum between the steel and the glass reduces loss of heat by convection, and coating the tube with a so-called selective surface, such as aluminium oxide and molybdenum, promotes the absorption of solar radiation.
The system’s performance depends crucially on the ratio of the amount of light absorbed by the tube and the amount reradiated as heat at infrared wavelengths. The goal is high absorption and low reradiation. Energy loss is proportional to the size of the absorbing surface, and the Luz engineers minimised this loss by concentrating the incoming sunlight on a comparatively small receiver. The penalty was that they could capture the most intense radiation – that coming directly from the solar disc – but very little of the light scattered by the atmosphere (circumsolar light) or by water vapour, clouds and dust particles (diffuse light).
The Luz system is designed to work well for the desert regions of California, where the sun shines bright and there is little water vapour, but most cities are not located in deserts. By implementing two developments, new selective coatings for the tubes and revised alignments for the troughs, Mills believes it will be possible to increase the efficiency of solar thermal plants and to extend the geographical range of the technology.
Working with Qi-Chu Zhang, a fellow physicist at the University of Sydney, Mills has created a selective coating using insulating materials, ceramic-metal alloys known as cermets, and metals. An outer layer, made from a transparent material such as silicon oxide, is an antireflection layer. Next come two cermet layers. Mills and Zhang have experimented with various cermets, and say that composites of metals, such as copper, gold and nickel, and insulators, such as silicon dioxide, work well. Finally, a reflector of copper or another metal conducts the trapped heat to the oil or water.
More efficient
Sandwiched together, the layers are effective at absorbing solar radiation, says Mills, while the heat loss is three to five times less than from the surface on the Luz tube. This means it converts more than 90 per cent of the sunlight that strikes it into heat. Mills expects to have the double-cermet layer ready for commercial use within 12 months. When this happens, he says, it will no longer be necessary to focus the light so precisely. It will also be possible to increase the size of the tubes relative to the collector, and he is proposing to reduce the size of the troughs to as little as 2 square metres compared with the 545 square metres of the Luz troughs.
Another criticism levelled at the Luz technology in California is that its performance in winter is poor. Each collector is horizontal and aligned on a north-south axis which maximises the amount of sunlight that is captured in summer, the time of the year when southern California requires electricity to run air conditioners in homes and offices. In mid-winter, however, it supplies very little, sometimes nothing, to the grid.
Mills’s theory is that if the trough is tilted at the angle of latitude – in Darwin, for example, at an angle of 12 degrees and in Sydney at 34 degrees – the focal axis of the trough would match the direction of the Sun’s rays for much of the year. This would increase the amount of energy gathered in winter at high latitudes. Tilting Luz’s original 100-metre trough would present problems, but not so his smaller troughs. Mills developed computer models to examine the effect of tilting, and they predict that in temperate latitudes inclining the troughs would boost the amount of solar radiation collected by at least 11 per cent per square metre.
Within three years, says Mills, improvements in the selective surfaces and reflective materials will reduce the cost of generating electricity to about 7 cents per kWh. About two years after that, with more advanced selective surfaces and other improvements, the price will be close to 5.5 cents. The new selective surfaces will be capable to converting an average of 20 per cent of solar radiation to electricity throughout the year, compared with less than 16 per cent for the Luz technology. Under peak conditions – full sunlight in the summer months – conversion efficiency would be about 28 per cent.
International competitor
By the year 2000, Mills is confident that solar thermal electricity will be competitive with other energy sources over a much greater geographical range, including two-thirds of the US, all of continental Australia, southern Europe, and in the developing countries of Africa, South America and the tropics. ‘Its time has come,’ says Mills.
Not everyone agrees. ‘As far as most power utilities are concerned, solar is still a largely untried, unproven technology,’ says Bruce Godfrey, head of Australia’s Energy Research and Development Corporation. Nevertheless the corporation, which manages the Australian government’s investment into energy research, is by no means ignoring solar power and has directed A $300 000 ( £150 000) to a feasibility study of the commercial use of the Big Dish.
This technology gained the upper hand in Australia in 1992, when the power companies had to make a decision about which solar thermal technology to back; at that time the dish was further advanced than the trough. Designed by Stephen Kaneff of the Energy Research Centre at ANU, the dish works much like an optical or radio telescope. Light falling on its glass mirrors is reflected to a receiver, a coil of tubing above the centre of the dish. As the heat is absorbed, water inside the tube turns to steam and is sent through insulated piping to the ground, where it drives a small engine. A collection of the dishes could drive a steam turbine.
In desert locations, according to calculations by Kaneff, the dish can generate power for as little as 4 cents per kWh. It can also focus so much light onto a small area that it produces very high temperatures – up to 1500 °C compared with about 390 °C for troughs. This, says Kaneff, makes the Big Dish more suitable for generating superheated steam and, unlike the Luz troughs, it does not require a gas supplement to raise the temperature of the oil or water to achieve high temperatures.
Another advantage is that the dish can point to any part of the sky because it can tilt in two directions. This allows it to track the sun in the azimuth direction along a vertical axis and to move along a horizontal axis to face the horizon, making it more efficient in winter and at higher latitudes. Also, because of the intense heat, dishes will be able to generate thermochemical reactions. The heat will separate fuels such as ammonia into gases for power generation or industrial uses. Troughs cannot generate enough heat to be able to do this.
Now that the demonstration unit has generated steam, Kaneff is hard at work with a consortium of seven power companies and energy authorities to install the technology as part of the grid at Tennant Creek. The consortium plans to erect up to 28 dishes (three will be in reserve), all the same size as the one at ANU, to produce up to 2 megawatts of power. This will be the first commercial demonstration of this technology, says Kaneff.
The project, likely to cost around A $15 million, is regarded as a feasibility study on the way to a larger system, according to Wesley Stein from Pacific Power, the business arm of the Electricity Commission of NSW and one of the partners in the proposed development. At present, Tennant Creek has three small gas turbines and four diesel units, but these are old and polluting. A combined solar/natural gas plant will gradually replace them. The solar system, says Kaneff will be engineered to produce electricity to meet peak demand. The natural gas component will be capable of generating the full 4 megawatts that the town and mine needs, but normally will supply only 2 megawatts. If the plant is successful, more collectors can be added, but there will still be a need for natural gas to produce power at night and on cloudy days.
Solar thermal energy may be the way to go, but it will be another five or six years before Mills and Kaneff can be certain that their technologies are economic. And before either can be widely used in areas such as in the tropics, where cloudy and moist conditions prevail, they will have to prove themselves in a much greater range of conditions. In the meantime, another potential supplier of solar power – photovoltaic (PV) cells – promises to become more competitive.
For the past 17 years, Green’s group at the Centre for Photovoltaic Devices and Systems at the University of NSW has been steadily increasing the efficiency of the conversion of solar radiation to electricity; recently he obtained an efficiency of 23.5 per cent, a world record for PV cells. But the cells, usually between 300 and 400 micrometres thick, consist of two layers of high-quality, purified silicon. Their purity makes them good conductors of electricity – and expensive to make.
Then in May, Green’s team claimed that it could produce efficient PV cells using silicon that is up to a thousand times less pure. In a bold statement, Green said: ‘We have developed a new design strategy which should eventually meet all nations’ needs for cheap and environmentally-sound energy supplies.’ He claims that in about ten years the cost of solar electricity, using these cells, could be cut from today’s 30 to 40 cents per kWh to about 5 to 8 cents.
Layered cells
The university has taken out patents on a series of technologies that Green has developed over the years. It is a combination of these technologies that has enabled Green to redesign the PV cell using cheaper silicon. Conventional cells consist of two layers of silicon, one negative and one positive, with a junction to collect the charge. To reach the junction, the charge often has to travel the thickness of the cell, which can be up to 400 micrometres. With poorer quality silicon, the electrons would normally be absorbed back into the silicon before reaching the junction.
The new design uses a laser to cut a groove through multiple layers of lower grade silicon. ‘We refer to it as the Sara Lee solar cell,’ says Alistair Sproul, a member of Green’s team, after the Sara Lee range of layered cakes found in every Australian supermarket. The total thickness of the layers is only 20 to 30 micrometres, about half the diameter of a human hair. A metal conductor in the groove makes contact with each of the layers, which have alternating positive and negative polarities. As each layer has a junction, the distance that electrons have to travel to reach a junction after sunlight has excited them is only around 2.5 micrometres.
In tests, Green’s group has achieved an efficiency of 15 per cent for a 5-layer cell, but it is confident that this figure will improve as the technology is developed, and cells are designed with many more layers. So will PV energy become a major supplier for the electricity grid market? Green says that generating power for the grid is a long-term aim. Before that, he believes, PV energy will be widely used for residences. ‘Not all countries will want solar power for the grid. Japan, for example, does not have the land for large solar plants. But it wants solar power for houses and PV is best suited to supply this.’ The power will come from PV cells on the roof. The idea is to connect thousands of homes, with any excess power being sent back to the grid. A demonstration of this technology is already being tested in Germany, where 2500 homes have been connected. Green predicts that his layered cells will be commercially available within ten years and that within twenty years they will be manufactured in sufficient volume for large-scale electricity generation. But will that be too long?
‘Solar thermal electricity will not stand still in the meantime,’ says Mills. Even if the price of PV cells drops by five times in ten years, he says, the cost of generating electricity using them will still only match what solar thermal is capable of today.
Despite the enthusiasm of Mills, Kaneff and Green for their respective technologies, solar power still has to prove that it is commercially viable. ‘It’s still expectations and promise as much as anything else,’ says Kaneff. ‘We haven’t built a station yet that is commercial in the sense of providing a continual source of power.’ Kaneff believes that this is likely to happen in India, which is ‘one of the few countries that is serious about solar’. India has established a ministry of nonconventional energy resources which has called for five solar thermal power stations, each of 10 megawatts. ‘I believe it will be demand like this – from a country with terrible pollution problems and, at the same time, a desperate need for industry – that will drive solar power.’
But others, like Mills, see the push coming from the need to curb greenhouse gases. Whatever the impetus, solar power may finally have found its place in the Sun.