MARTIN GREEN HAS BEEN GOING to solar conferences for more than 30 years. But Solar Power 2006, held in October of that year in San José, at the southern end of California's Silicon Valley, was something new.
"This year there were about 6,000 attendees, compared with 1,000 last year," says Green who, as director of the Photovoltaics Centre of Excellence at the University of New South Wales (UNSW), in Australia, was a keynote speaker. "It was all investment bankers and the like, the kind of people we haven't seen at these conferences before."
The money men were drawn by solar power's recent stellar performance. From 2000 through 2005, sales of solar panels soared at a compound annual growth rate of 41 per cent, according to Paula Mints, who tracks the solar industry for California-based Navigant Consulting. "That's amazing growth for any industry," Mints adds.
In 2005 alone, the solar power sector grew 44 per cent in volume, 50 per cent in revenue and a whopping 149 per cent in profit. In that year, three of the largest stockmarket floats of technology stocks were for solar energy companies, raising over US$800 million (A$1 billion) in total.
American venture capitalists invested a record US$739 million in renewable energy start-ups, solar specialists prominent among them. This year, the industry will be worth around US$20 billion and, at the current rates of growth, by 2011, US$100 billion.
No doubt about it: solar power is hot. But photovoltaics have been around for ages, so why the sudden interest? In a word, subsidies. These have propelled solar from its traditional, off-grid niches — powering remote homes, long-distance phone lines and water pumps — and plugged it into the mainstream: the electricity grid. Today, the grid-connected market accounts for 83 per cent of solar sales.
During the 1990s, the government of resource-poor Japan poured US$200 million a year into rebates for homeowners who installed solar panels on their roofs. The incentives have since been phased out, but they have had the desired effect: close to 300,000 Japanese homes are now solar-powered; that's more than one-in-100 households. The goal is to increase the ratio to one-in-25 by 2010 and one-in-four by 2020.
In addition to a thriving domestic market for photovoltaics, the policy has also spawned a booming export industry. Japanese manufacturers such as Sharp, Sanyo and Kyocera have seized large slices of the global solar market. Sharp alone is the world's largest producer of panels with a share of around 25 per cent.
In 2000, the German Bundestag passed the Renewable Energy Act, which legislated innovative 'feed-in' tariffs. These require utilities to buy surplus electricity at premium prices from owners of grid-connected photovoltaic systems; in effect, presenting purchasers of solar panels with a 20-year annuity. Householders and farmers who take out a low-interest loan to stick a solar system on the roof of their home or barn receive rebates that exceed repayments.
Thanks to this generous program, environment-friendly Germany has rapidly become the largest market for solar energy, accounting for 45 per cent of worldwide demand in 2006. As in Japan, a whole new industry has sprung up, employing tens of thousands of workers.
AT AROUND NOON on 26 July 2006, a watershed event occurred in Germany. For the first time ever, the cost of electricity generated from conventional sources such as nuclear exceeded that for energy generated from the Sun during a period of peak demand.
Now, as policy makers wake up to the implications of global warming, other European countries — Spain, Italy, Portugal and Greece as well as the United States, in particular California — are emulating the German example. As manufacturing volumes increase, so the price of solar panels will drop, by 20 per cent for each doubling in production.
Meanwhile, the introduction of taxes on emissions and expensive palliative measures such as 'clean coal' technologies and carbon sequestration means that the cost of fossil fuels will inevitably rise.
Today, the solar power industry is basking in the prospect of double-digit growth for years to come. In fact, demand is so high that supplies of silicon feedstock are drying up. Crystalline silicon is the material of choice for more than 90 per cent of photovoltaic panels. The solar boom is sucking up production, and the price of the raw material has jumped from US$65 a kilogram to as high as US$400.
Boosting production of purified silicon to meet demand is not trivial, as it involves massive capital outlays. Such investments are underway, but it may take silicon producers several years to catch up with demand.
The unhappy result of the silicon shortage is that, instead of falling, prices of solar panels have actually been rising. This has opened a window of opportunity for innovative photovoltaic technologies that use less silicon — or none at all. As it happens, Australian researchers and firms are pioneering both approaches.
IT HAS TAKEN SOLAR POWER over 30 years to work its way toward the mainstream. The photovoltaic cell was invented at Bell Laboratories in the U.S. in 1954, and solar's first applications were extraterrestial, powering the likes of communications satellites and lunar modules. Space panels were hand-crafted like jewellery; price was no object.
The origin of the modern photovoltaic industry dates back to 1975, when Bill Yerkes founded Solar Technology International (which became Arco Solar) to address terrestrial markets, such as charging batteries. Yerkes was responsible for two major innovations: he stuck silicon wafers on the back of glass, protecting the cells and making his panels almost indestructible. Second, he developed an inexpensive screen-printing technique for depositing metal contacts.
Since then, developments in good old flatplate solar technology have mostly been a matter of eking out incremental improvements in efficiency. Today's best panels are capable of converting around 22 per cent of sunlight into energy.
Over the years, it has been repeatedly demonstrated that, in solar cells, there are no magic bullets. "It's very hard to make this stuff work," Mints says. "It has to sit out in the Sun for 20, 25 years. I mean, how many other products do you know that can do that?" The fact of the matter is that commercialising solar products takes money and time. As one long-suffering solar industry survivor puts it, "Reality is a bitch."
In 2000, Andrew Blakers and Klaus Weber of the Centre for Sustainable Energy at the Australian National University in Canberra came up with an ingenious new idea for reducing the amount of silicon used, and hence the cost of solar cells.
Instead of using silicon wafers sawn from the ingot, they took the process a step further — into the third dimension — and used micromachining to slice up wafers into a thousand narrow rectangular slivers, each just 0.05 mm thick. Lifting the slivers out to form a cell is, Blakers insists, a simple matter. A single 15 cm wafer yields enough slivers to cover a square metre.
The university licensed the technology to Australia's Origin Energy, which invested over A$30 million in the construction of a pilot plant in South Australia to manufacture sliver cell panels. In August 2006, amid much hoopla, the then Australian Prime Minister John Howard announced the launch of a A$75 million initiative to build 'solar cities'. In the first of these, in North Adelaide, 1,700 homes will be equipped with solar panels made in Australia by BP and Origin Energy.
From the outside, the sliver cells story would seem to be a model of successful university-industry technology transfer and commercialisation. From the inside, however, the story is very different.
In 2003, following transfer of the technology, funding from Origin for research at the Australian National University was cut. Since then, Blakers' group has developed an improved second generation of sliver technology, but government support has almost evaporated.
In his office, little more than a stone's throw from Canberra's Parliament House, Blakers complains that for the last couple of years, he has been forced to let one member of his staff go every few months. Meanwhile, Origin is now looking for an offshore partner to manufacture the cells on a large scale.
While Australia's dwindling handful of solar energy researchers struggles to survive on tiny grants, their German counterparts are flush with cash. Centres like the Fraunhofer Institute for Solar Energy Systems in Freiburg and the Institute for Solar Energy Research in Stuttgart muster hundreds of staffers and receive dependable annual budgets of tens of millions of dollars.
"Here within Australia, we've got a firm foothold in the industry, we've got the ability to grow rapidly but, unfortunately, we do not have the financial resources to match our competitors — in Europe in particular," a frustrated Blakers explains. "We are simply falling behind our competitors, and will 'drop off the tree' in the next five years or so if nothing dramatic happens."
Even Green's pioneering group at UNSW, which for more than 20 years has been setting world records for efficiency in silicon wafer solar cells, is not immune to pruning. Continued funding for its photovoltaics research is subject to annual review by a committee of sceptical senior scientists. At a recent meeting a senior member, to Green's dismay, questioned whether there was any evidence of a cost to the environment from carbon emissions.
Green's approach to reducing the need for silicon is CSG: crystalline silicon on glass. In this technology, a thin film of amorphous silicon, less than two micrometres thick, is deposited onto a glass sheet. The silicon is heat-treated in a furnace to turn it crystalline. Lasers and ink-jet printing technology are used to build the electrical contacts. CSG uses 99 per cent less silicon than the conventional wafer technology.
Commercial development of CSG technology began in 1995 with the formation of Pacific Solar, a joint venture between UNSW and Pacific Power, the state-owned electricity utility. This rather cosy relationship ended when, as a result of restructuring and privatisation, Pacific Power ceased to exist. CSG chief executive David Hogg quickly realised that, if his company was ever going to become a serious business, then Sydney was certainly not the place to be.
In 2004, Hogg and a handful of colleagues decamped to Thalheim, Germany. There, with new European investors, and a new name — CSG Solar — the firm has recently begun commercial production of thin-film silicon solar modules. As Hogg admits, however, after more than a decade of effort, module efficiencies are still down around seven per cent — one third that of the best wafer-based panels.
Long-term reliability may also be an issue. "Thin films actually do fairly well in Germany, where it's overcast much of the time," Mints says, "but I'm really cautious about how they'll do in Spain and Portugal with the Sun beating down on them."
In addition to developing technologies, UNSW's other output is graduates: its Photovoltaics Centre of Excellence runs the world's only undergraduate photovoltaic program. It produces 20 to 30 graduates a year, supplying the solar industry with talent. A few, like Blakers, have found positions locally. But lacking a domestic Australian industry to employ them, many of Green's graduates have to head overseas.
China in particular is happy to welcome them: three former UNSW post-doctoral graduates head Chinese solar panel makers. Among them, Zhengrong Shi, chief executive of Suntech Power, stands out. A physicist by training, Shi joined Green's group in 1989. During the 1990s, he worked part-time at Pacific Solar as the company's deputy research director, helping develop thin-film silicon technology.
In 2000, Shi was lured back to China to set up a solar manufacturing operation in Wuxi, a city two hours' drive west of Shanghai. Using second-hand equipment, Suntech quickly began production of conventional flatplate solar panels. Presciently, Shi locked in a long-term contract for silicon feedstock, thus immunising his company from the effects of the silicon shortage.
By 2005, thanks largely to insatiable demand from the German market, Suntech had broken into the world's top 10 solar panel manufacturers. That December, Shi took his company public on the New York Stock Exchange. It was one of the most successful floats of the year, making the Australian-Chinese scientist, at least on paper, a billionaire and Australia's fourth-richest person.
Suntech is expanding fast: sales in 2006 are expected to top A$700 million. That, as Green has pointedly noted, is far more than the total Australian exports of uranium. The company already runs two factories and is building a third, showpiece facility, due to open in May 2007. Shi is investing heavily in research and development, both focussing on incremental improvements to conventional panels and on next-generation thin-film technology.
He maintains close links with UNSW: Green's colleague Stuart Wenham splits his time between Sydney and Wuxi, where he works as Suntech's chief technology officer. The collaboration works both ways: Suntech staff train at the centre, the centre's students get experience using the company's facilities.
Shi is a generous donor to the centre, sponsoring joint research as well as undergraduate and postgraduate scholarships: research funding is about A$500,000 this year, and is set to double in 2007. "He's the best thing that ever happened to us," says Richard Corkish, who heads the school of photovoltaic and renewable energy engineering.
Corkish recently toured China on a trade mission with Australia's environment minister at the time, Ian Campbell. "There's immense interest in solar energy there," he comments. Indeed, China is investing heavily under ambitious new renewable energy legislation that calls for an enormous expansion in the domestic market, from 20 MW in 2005, to 10 GW in 2020 — a 500-fold increase. "There's a huge opportunity for technology transfer now," Corkish enthuses, "to give China a leg-up and also do well for ourselves at the same time."
THE CONVENTIONAL IMAGE of solar power is multiple photovoltaic panels stuck on a roof. But there is also another approach, one that is much closer to the conventional, centralised-utility model of electric power generation. Concentrator technology stems from the recognition that sunlight is a dilute resource.
Using mirrors that track the Sun to concentrate solar light, much as kids use a magnifying glass to burn paper, you can focus the Sun's rays into a beam that is powerful enough to melt steel. Concentrators can be either photovoltaic, generating electricity directly, or solar thermal, generating heat to drive steam turbines [see 'Ways to catch rays', p75].
As with every other solar technology, concentrators are not new. Invented in the late 1970s, they have undergone a long, painful apprenticeship as developers learned how to deal with problems like eroding mirrors and tracking devices that became jammed with dirt.
Now, at the height of the silicon shortage, concentrator firms have unexpectedly become the darlings of venture capitalists. At Solar Power 2006, the concentrator session was standing room only. "We were all amazed," recalls Mints, "because that's usually the session you go into if you want to take a nap."
Interest is particularly keen in Australia following an ambitious announcement in October 2006 by Melbourne-based concentrator specialist Solar Systems. The firm plans to build in northwest Victoria what will be — by an order of magnitude — the largest photovoltaic power plant in the world. State and federal governments are providing A$125 million for the 154 MW plant. The private sector is expected to provide up to a further A$295 million.
Solar Systems has already demonstrated the effectiveness of its concentrator technology with four community-sized stations, providing power to remote Aboriginal communities around Alice Springs. The stations consist of a cluster of 14 metre-wide tracking dishes, each mounted with 112 curved mirrors.
The mirrors reflect the light — concentrated 500 times — onto an array of solar cells mounted at the focal point of the dish. High-performance proprietary heat sinks, integrated on the back of the cells, keep them cool enough to touch.
During daylight hours, the solar stations replace diesel-driven generators, saving A$1,000-worth of fuel a day and eliminating polluting greenhouse emissions. Each station generates enough electricity to power a thousand local homes.
To achieve such high performance, Solar Systems uses the most efficient photovoltaic cells available. These are made by Spectrolab, a Los Angeles-based subsidiary of Boeing; and they are not made from silicon but from multiple layers of compound semiconductors like gallium arsenide. Normally used to power satellites, Spectrolab's cells offer efficiencies of up to 35 per cent, and are expected to top 40 per cent by the end of the decade.
Of course, such cells are much more expensive than their silicon counterparts but, as Solar Systems' technical director John Lasich explains, you don't need many of them. A module the size of a mobile phone can produce enough juice to run a house, putting out 1,500 times more power than a standard photovoltaic panel of the same size.
"So when you come to manufacturing, we only have to produce one fifteen hundredth of the stuff," Lasich explains. "It's a massive advantage because it costs you roughly three to five dollars per watt of annual production capacity to build a photovoltaic factory.
So if you wanted to build a facility that could produce 100 MWs a year, it would cost you around A$300 million for solar panels, whereas it would only cost us a fraction of that price."
High-performance solar cells driven by intense beams work wonderfully well, converting sunlight into AC power at very high efficiency — about 25 per cent. "That's the same as they get at Loy Yang [a large Australian coal-fired power station]," he says. "But we can turn light into electricity instantly, whereas with coal you've got to have something that's been worked on for 20 million years."
Lasich has been developing photovoltaic systems ever since the mid-1970s, first as a student, then in his own backyard, and finally at Solar Systems, which he founded in 1990. An electrical engineer-turned-physicist by training, Lasich also has extensive experience of building large-scale thermal and control systems for the petrochemical and power industries.
The mirrors and tracking devices that will constitute 85 per cent of the new plant are made from laminated windscreen glass, steel and concrete — all relatively cheap materials. In place of dishes, the plant will consist of arrays of 250 heliostats — individual mirrors that track the Sun and reflect it at solar cell modules housed on 40 metre-high towers.
It will be built over seven years at several sites, covering 800 hectares in total. They are located in the Mallee, an inland area that is the hottest and driest part of Victoria.
Ultimately, the plant will be connected to the national power grid, functioning as a 'peak lopper'. That is, it will help to deal with peaks in demand that occur on the hottest days of summer, when everyone simultaneously turns on their air-conditioners.
Hotter weather and lower prices have led to a boom in air-conditioning, making Australia one of the world's 'peakiest' countries. Happily solar power, unlike wind, can be relied upon to match daytime supply with demand since, when demand for air-conditioning is greatest, the Sun is also shining.
With the economies of scale developed in building this first large-scale plant and the technology of non-silicon solar cells continuing to improve, Lasich is bullish about the future.
"I think we're at a watershed," he says. "Solar power is about 10 bucks a watt now, our stations that we've built out there are now about eight bucks, and the cost per unit in a big plant is vastly less. If we can get below the three-dollar-a-watt mark, it's just going to explode."
What percentage of the demand for electric power could solar be realistically expected to provide? "You could put 10 per cent into the grid without disturbing it too much," he suggests, "because the daytime peak is about 10 per cent above night-time.
With current electricity generation capacity in Australia almost 50 gigawatts, 10 per cent would be five gigawatts; it would be around A$15 billion worth of photovoltaic. That's enough business to establish a whole new industry."
TODAY, SOLAR ENERGY SUPPLIES just 1.5 GW — less than one per cent of the world's total energy needs. Estimates of how big solar will become, and what percentage of demand it can meet, vary enormously. "Five per cent of the renewable energy pie," predicts a cautious Mints. "Fifty per cent of the world's electricity in 2050," reckons an optimistic Blakers.
But the truth is, no one knows what will happen. Much will depend on the fast-changing politics of climate change. Even in Canberra, doors that have long remained closed to solar researchers have lately started to swing open. What can be said with certainty is that, at long last, the solar industry has momentum and the money to maintain it.
Throughout its history, the photovoltaic business has been the poor cousin, forced to feed off scraps from the semiconductor industry. But while the chip makers don't care how much they pay for silicon, for the panel makers, the price of their raw material is the key metric.
Now, flush with venture capital and profits, the photovoltaic folks can afford to develop processes and equipment specific to their needs. For example, Yerkes, the solar panel pioneer, is today chief technology officer for California-based Solaicx. The company's goal is to change the economics of silicon wafer production, attacking the biggest obstacle to the widescale adoption of solar power: cost.
The conventional method of making solar-grade silicon is to grow individual ingots, then saw them up into wafers. Solaicx has developed a continuous crystal grower that Yerkes says can crank out ingots like sausages. One machine is capable of growing five megawatts' worth of 15 cm-diameter ingots every year.
The company's first grower is about to go into continuous 24-hour operation, making Solaicx one of the few firms in Silicon Valley that actually produces silicon. More machines will enter service next year at a new facility in Oregon. Yerkes claims that the technology has the potential to reduce the cost of silicon wafers — which account for about half the price of a solar panel — by 75 per cent.
Some people wonder why it has taken solar power such a long time to reach critical mass. As he looks back on three and a half decades of hard-won progress in solar cells, Yerkes, now 72, disagrees. "I don't think it has taken very long," he says. "Compared to my genes and your genes and that sort of thing, it's happening almost instantaneously."
Bob Johnstone is a science and technology writer based in Melbourne.
