The Solar Industry Gains Ground

At a time of economic pain and planetary peril, a renewable global powerhouse takes shape. Just when we need it most.

The Solar Industry Gains Ground
Empowered: Detail of a solar panel built with crystalline PV cells from Q-Cells, the world’s largest manufacturer. | photograph by Charles Masters Empowered: Detail of a solar panel built with crystalline PV cells from Q-Cells, the world's largest manufacturer. | photograph by Charles Masters

To get a sense of just how bright and sunshiny the future looks to the solar-energy industry, consider The Graph: It’s a standard affair, projecting solar’s share of global energy production over the coming century. The Graph was created by a scientific organization that counsels the German government, but it has since become a prized piece of propaganda, embedded in glossy brochures and PowerPoint presentations by solar companies from California to gray-skied Saxony. At the left-hand, present-tense end of the scale, solar power is a microscopic pencil line of gold against the thick, dark bands of oil and natural gas and coal, an accurate representation of the 0.04% of the world’s electricity produced by solar power as of 2006. The band grows slowly thicker for 20 years or so, and then around 2040 a dramatic inversion occurs. The mountain-peak lines indicating the various fossil fuels all fall steeply away, leaving a widening maw of golden light as solar power expands to fill the space. By 2060, solar power is the largest single band, and by 2100 it is by far the majority share.


This has always been solar energy’s tantalizing promise, since the first photovoltaic (PV) cells emerged out of Bell Labs in the 1950s to power space probes and ignite the dreams of a generation of giddy utopian dreamers. Solar energy is as plentiful as daylight, as limitless as organic life itself, a fuel that comes free of charge and replenishes itself every time the earth rotates on its axis. Almost all energy, after all, is ultimately stored solar power: Oil, gas, and coal were born of the ancient sunlight that fed prehistoric animals and plants, the wind is set howling by the sun’s unequal heating of the atmosphere, and even a campfire draws its warmth from solar power trapped long ago through photosynthesis. Enough radiation from the massive fusion reactor at the center of our solar system hits the earth every hour to fill all of its energy needs for a year.

Fifty years on, the PV cell remains the most effective engine yet devised for the conversion of sunlight into electricity. The core of every PV cell is a semiconductor, traditionally a highly purified wafer of crystalline polysilicon, virtually identical to the “chips” upon which computer circuits are built. When sunlight strikes the semiconductor, its atoms get excited; if the light beam’s photons pack a sufficient punch, they knock the semiconductor’s electrons loose for collection by the PV cell’s circuits, creating an electric current. Assemble a handful of these cells in a glass frame and wire it to a battery or a power grid, and you have got a solar panel (or module, as it’s sometimes called in the industry). A small, pollution-free power plant. Compared to splitting an atom or sucking liquefied phytoplankton from 300-million-year-old bedrock, it’s practically child’s play.

Solar has nevertheless been stuck for decades in the future tense. PV cells have been too inefficient and too expensive, or too reliant on a fickle sun. But the solar industry has recently made a dramatic leap from the feel-good margins to the mainstream. An unprecedented production boom began around 2004, well before the rise and current fall in crude-oil prices; that boom continues unabated and has led to plunging costs, vastly more rational supply chains, and record-setting gains in the efficiency of traditional crystalline silicon cells (the best now conduct electricity at efficiency rates almost 30% better than the lab record of 2003). In the past 50 years, about 10 gigawatts of solar power — roughly the output of 10 standard nuclear reactors — have been installed worldwide. But current estimates, which have been modified only upward in recent years, are that 10 gigawatts more will be brought online in 2010 alone.


A new global industry is taking shape before our eyes. A journey through this energy revolution suggests that the age of truly ubiquitous solar may at last have begun. Solar’s emerging titans are scattered across three continents and three technological generations — from established crystalline PV manufacturers in California to newer “thin film” cells now reaching mass-production scale in Germany and to even third-generation compounds being developed in Australia that can be integrated into building materials to deliver power in the darkest shade. Even in this time of enormous financial uncertainty (not to mention a deepening concern, if not panic, about the health of the planet), the sense of boundless potential, the promise of The Graph, is palpable. Erik Straser, who oversees the clean-energy portfolio at Mohr Davidow Ventures in Silicon Valley, puts it this way: “Sometimes I ask myself, ‘If this company was successful, would people name libraries and public high schools after it?’ Who made the steam engine? Who made the lightbulb? Who will those people be for the 21st century? Who’s the person that made mass-market solar affordable?”

A global industry is taking shape, with the hottest spots in Silicon Valley, Germany, and Australia. The age of truly ubiquitous solar may at last have begun.

Back in the present tense, in the piercing glare of a July morning in Silicon Valley, I’m shielded by the smoked glass of a standard-issue corner office. Tom Werner, CEO of SunPower, America’s largest PV manufacturer by revenue, sits across from me at a glass-topped conference table, essentially making the case that he’s that guy, the godfather of cheap solar. One of them, anyway.

“The hypothesis of SunPower,” Werner tells me, his argument bottom-line blunt, “was take a high-technology, high-efficiency solar cell and mass-produce it at low cost. And it worked.” He slides a small pane of glass out of a file folder. It’s about the size of a household floor tile and inlaid with a blue-black hexagonal pattern. This is SunPower’s PV cell, which, at 22% efficiency, holds the world record for a commercial product. (The industry average is about 16%.) He holds it up for my inspection, and I notice the hexagons are identical to the ones on the tabletop between us, which turns out to be a large SunPower panel mounted on four legs. “As you create this market for solar,” Werner says, “you create the opportunity to scale. And so what happens is, you innovate your way down the cost curve.”


Werner strikes me as exactly the type of hypercompetitive, profit-obsessed executive the solar industry had long been lacking. At 48, he’s trim and athletically wiry, with tidy side-parted hair and a goatee. He slurs a bit through gritted teeth as we talk because the night before, in a rec-league basketball game, he caught an elbow that drove his incisors through his lower lip. “I’ll get that guy,” he tells me. I’d bet on it.

Werner’s spiel might sound like textbook, first-year MBA stuff, but the second part of it — the cost-reducing, mass-producing part — was mostly absent from the solar industry for its first half-century. Incremental efficiency gains were the industry’s core focus, and solar companies tended to be small and mission-driven, with a university lab’s sense of priorities. SunPower was, until recently, no exception: Founded in 1985 by a Stanford engineering professor, it spent its first 15 years building solar-powered aircraft prototypes for NASA and sci-fi solar concentrators shaped like satellite dishes.

“What’s the purpose of solar?” Werner asks, switching to rhetorical mode. “It’s to get energy, right? So then the question is, Will the cost of solar energy converge on the cost of the way you get energy otherwise? And in the ’70s, the answer was definitively no.” The answer was still a pretty strong no when Werner took the helm of SunPower in 2003. He’d been transferred from Cypress Semiconductor, which had recently bought the company, and was given the monumental task of bringing the price down to mass-market rates.


“Today,” he says, “the cost of the grid’s gone like this.” He slices his hand sharply upward, indicating the skyrocketing price of conventional energy. “And the cost of solar is coming down. So that crossover point is imminent.”

Werner calls that point the “levelized cost of energy,” but in most of the industry the preferred term is “grid parity” — that magic moment, which may arrive by 2012 or even sooner in heavily subsidized energy markets such as California, Germany, Italy, and Japan, when the price of a kilowatt-hour of solar energy is about the same as one generated by any other fuel source. Grid parity: It obsesses solar executives like a grail, rolls off their lips like a forceful boast or a solemn promise. Grid parity: not if, no longer if. Only when.

Under Werner, SunPower has rapidly reinvented itself, catapulting from the lab to the center of an exploding global market. At every turn, he and his team uncovered the weaknesses and irrationalities of an immature industry, particularly at the postproduction end, where installation costs often represent 50% of the total price of a PV system. In response, the company simply expanded into those markets. SunPower is unique in the business today in that it manages essentially its entire supply chain, from silicon ingot to installation. “We’ve industrialized this industry,” says vice president Julie Blunden.


The company’s timing has been impeccable. Every year since Werner’s arrival, the global solar industry has grown by at least 40%. It has jumped from humble residential roofs to the wide, flat expanses of big-box stores and office parks, and graduated from rooftop array to greenfield power plant. Solar power was the planet’s fastest growing energy source in 2007, and in recent years, demand has outpaced supply and given rise to the industrial-scale solar business’s enviable lament: No company could seem to manufacture PV panels fast enough to meet demand. Today, solar is a $13 billion global industry, and it’ll hit $40 billion by 2012 (unless it exceeds forecasts for the next five years as it has for the past five). Gigawatt-scale “fabs” — that is, single production lines capable of producing enough panels every year to add 1,000 megawatts of installed capacity to the global energy mix — are in the works in Germany, China, and Japan, while several American companies are poised to soon reach similar levels of production. SunPower’s 2007 revenue came in at $775 million, triple the 2006 figure. The company estimates it will clear $1 billion — with $90 million in profit — in 2008.

Solar is a $13 billion industry today. It will hit $40 billion by 2012, unless it continues to exceed forecasts, as it has over the past five years.

Critics of crystalline PV — particularly the heads of newfangled thin-film solar companies eyeing crystalline’s 86% market share — continue to suggest that it might never be produced cheaply or quickly enough to compete with other power sources in the long term. But when I present this critique to Werner, he responds with a sotto voce dismissal — “Well, it’s ridiculous” — and then bullet-points me through SunPower’s bona fides: that record 22% efficiency (against thin film’s paltry 8% to 12%), its reductions in hardware and installation costs, its rapid expansion to 400-plus megawatts of production capacity. The thin-film startups can wave around all the jaw-dropping cost-per-watt estimates they want (some claiming production costs as low as $1 per watt, versus crystalline PV’s average of $3 or more per watt), but Werner just points to SunPower’s completed projects. Its recent installation at Nellis Air Force Base in Nevada was the largest PV plant in North America when it was completed last April: 70,000 panels mounted on 5,300 of SunPower’s new tracking racks, which use GPS to adjust the panels’ positions minute by minute, improving efficiency up to 30% all by themselves. (At least six much larger projects, including two by SunPower, are now in the works.) The tracking system, Werner tells me, is just one of the 185 patented technologies by which SunPower intends to stay well ahead of the thin-film arrivistes. As for nuke, with its 10-year-minimum lead times? “One nice way to get out of a nuclear argument is to say, ‘Well, I’m worried about the next decade.’ ” And what about coal, trying to reverse the tide as more than 60 proposed plants are denied permits in the United States in the last year alone? “Fighting gravity.”

Of course, there are caveats. Disclaimers and indemnities and asterisks. As Neal Dikeman, a partner at clean-tech investment firm Jane Capital Partners, argues, the variables built into the cost of a kilowatt-hour of electricity are so numerous and byzantine that grid parity itself may be an illusory near-term goal. The sunniest estimates of that blessed event, Dikeman notes, are based on the cost of generating more power when the demand is highest. But this peak-demand power — the kind required by millions of air-conditioners at midday in California — is only a single-digit percentage of the total generating capacity on most electricity grids, and the cost of producing the juice is just one of many line items on the average power bill, alongside transmission, distribution, and maintenance costs. And in solar’s case, there’s also the cost of reconfiguring the grid to account for hundreds of thousands of small-scale installations. So what does the version of grid parity touted by solar boosters amount to? “A legitimate sales tactic,” Dikeman suggests, which “takes the best case to justify a subsidy to get down the cost curve.” He says true grid parity is “still close to a decade down that curve.”


Dikeman’s guess is probably as good as any on that score, and he’s certainly right that the solar industry’s rise has depended on subsidies. Solar’s growth has been largely driven by legislators, goosed along by various tariffs and tax incentives. But the energy business has long been a lonely place for free marketeers: According to British investment firm Ambrian Capital, the global renewable-energy industry receives about $11 billion in subsidies each year, versus $200 billion for fossil fuels, already a wildly profitable industry. So what’s a clean, limitless power source worth? And what scale could solar reach if there were a similar national investment behind it? If U.S. capacity ramped up to equal the 10 gigawatts expected to come online worldwide in 2010, that would be enough to power 3 million homes and reduce greenhouse-gas emissions equivalent to taking 22 million cars off the road. As BP likes to say, it’s a start.

The global renewable-energy industry receives about $11 billion in subsidies each year, versus more than $200 billion for fossil fuels.

Oil companies might greet such numbers with an eye roll, but solar execs are quick to point out how far the industry has come. They like to quote solar’s emerging corollary to Moore’s Law, that digital-age observation that a semiconductor’s processing power doubles every two years, even as prices plummet: The cost of solar, they’ll tell you, drops by 20% every time volume doubles, and the market of the past 10 years has borne this out. The idea of industrial-scale solar power seems even less naive and futuristic when you look at California, which has become the main production hub and primary market in the United States. True, the state is riding a $3.3 billion earmark package from Arnold Schwarzenegger’s administration, but the impact of those incentives is impressive. One installer, Solar City, has grown from nothing to $29 million in annual sales in just two years. Another, Akeena Solar, has moved out of its founder’s garage to 12 offices nationwide.

Then there’s Applied Materials, a $10 billion Silicon Valley giant that built its fortune by producing manufacturing equipment for computer chips and flat-panel displays. Back in 2004, while it was struggling to recover from the dotcom bust, Applied went looking for new markets and quickly discovered that the cost of a solar panel was dropping with almost semiconductor quickness and that its manufacture looked quite a bit like making computer chips or giant LCD screens. That was more than enough to spur the company to a substantial expansion into the solar market. “When I went to the board in 2005, we said, ‘This looks like it’ll be a $500 million business in 2010,’ ” says Mark Pinto, Applied’s chief technology officer. “And people said, ‘Wow, you know, that’s pretty good.’ Now we’re talking about $3 billion — plus.” A single Chinese startup, Best Solar, accounts for $1.9 billion of those sales; it plans to have Applied’s thin-film SunFab machinery driving its production line by early 2009 — and to reach 1 gigawatt of annual capacity by 2011. “Thin film is more utility scale,” says Chris Beitel, who oversees Applied’s solar operations. “It’s not about the residential rooftop. It’s about larger commercial rooftop areas, it’s about office parks — those are the areas where we’re going to succeed.”


But try telling that to Tom Werner. Under one recent contract, SunPower blanketed the roofs of 28 California Macy’s stores in PV, the majority of them under a “power purchase agreement” that is quickly becoming the commercial-roofing standard. Under the deal, Macy’s leases the rooftop space to SunPower for free and agrees to buy the panels’ output at a fixed rate for the next 10 years. Macy’s gets a competitive and stable electricity price in a volatile market, and SunPower simply gets a market. (Maryland-based SunEdison, which closed a $131 million round of venture funding in May, has solar-tiled the big-box tops of hundreds of Staples, Kohl’s, U-Store-It, and Whole Foods stores in California and beyond under the same kind of arrangement.) In mid-August, SunPower announced the biggest deal in its history, a contract to supply California utility giant PG&E with 250 megawatts of PV. This comes as part of a commitment by PG&E to construct two mammoth solar farms by 2011, capable of generating 800 megawatts of power — significantly more than the entire existing PV capacity nationwide. “This pair of deals actually changes the landscape of the solar PV industry,” says Roy Kuga, vice president of energy supply at PG&E. SunPower’s Blunden is even more ecstatic: “This is historic, monumental, tectonic — those are superlatives that are fair to apply to the announcement.”

It just might be, in other words, the solar industry’s very own Hoover Dam — the birth of the utility-scale solar industry in the United States. Which, by the way, is nowhere near the global industry’s epicenter.

Solar power’s rise has been fueled by sporadic bursts of political vision and courage. The ascent begins in Japan, where in 1994 the government introduced an incentive package in which it agreed to pick up fully half the cost of every installed panel for 10 years, spurring a handful of old-guard electronics firms to go industrial — Sharp, in particular, still the No. 2 global producer. The recent and much more robust solar boom, however, began with Germany’s Renewable Energy Sources Act. The German law, passed in 2000 (and since copied from China to California), is a “feed-in tariff” that obliges electricity retailers to buy power from renewable sources at above-market rates. The rates decline by a certain percentage each year for 20 years, depending on the source, at which point grid parity is presumed to carry on the work. An overhaul in 2004 placed particular emphasis on solar: Small, rooftop installations, for example, sell electricity back to the German grid at about six times baseline prices.


The German solar business promptly went supernova, precipitating a global polysilicon shortage from which the industry has only just recovered. (Silicon is the second-most-abundant element in the earth’s crust and a chief ingredient in much of the world’s sand, but silicon-processing companies were slow to realize they had an entirely new class of customer.) More than 100 solar companies have since set up shop in Germany — more than half of them in the former East Germany (GDR) — and now employ 57,000 workers and generate $7.3 billion in annual revenues. Seven of these firms have already vaulted onto the TecDAX 30, the technology index on Germany’s stock exchange, now sometimes jokingly referred to as the “solar DAX.”

The geographic center of the rapidly expanding German solar business is a decaying industrial belt south of Berlin, an East German reliquary that had been trapped in a seemingly terminal decline since the fall of the Berlin Wall. The world’s first solar heartland has emerged literally in the shadow of communism’s ruin, a place kissed by about as much sunlight each year as southern Alaska. A place with a name that sounds like a bad-weather curse: Bitterfeld.

Back in the days before the Wall tumbled, there was a German saying that went something like this: If we don’t meet in this world, we’ll meet in Bitterfeld. A once-proud little industrial burgh — Agfa developed the first color photographic film in the area in the years before World War II — Bitterfeld had been buried under an industrial blight so savage it seemed otherworldly. It had been turned into the Eastern Bloc’s chemical and pharmaceutical workhorse after the war, dooming it to half a century of sloppy Soviet-style industrialization and a flagrant disregard for the environment. When the Wall finally fell, 55,000 of the 60,000 jobs in Bitterfeld’s outmoded factories promptly evaporated.


The scars are still plainly visible in the boarded windows fronting the sturdy old brick buildings downtown and the industrial detritus beneath the rail-yard overpass, but I sweep past them in a smart new Mercedes taxi, bound for the site of Bitterfeld’s startling renaissance. Out beyond the befouled chemical plants, on the edge of a village called Thalheim, I find a gentle hollow rapidly filling up with low warehouses; construction cranes and wind turbines stand against the gray horizon. This out-of-the-way industrial park has come to be known as Solar Valley, and the sprawling complex at its center houses the operations of Q-Cells, the biggest company by market cap on the TecDAX and, since mid-2007, the world’s largest manufacturer of crystalline PV cells.

“We liken it to the car industry and say the status we’re now at — the cells you see out there — that’s kind of Tin Lizzy,” Stefan Dietrich, the company’s head of public relations, tells me. “That’s the T-Model Ford. That’s where we’re standing.”

Dietrich means it metaphorically — that the whole industry stands at that birth-of-an-industrial-age spot — but he could just as accurately make the case that solar’s industrial age was born here, in this cafeteria, with the stern chemical industry vets in coveralls on one side, chain-smoking through their lunch hour, and the business-casual crowd from Berlin and beyond munching salads on the other.


One of the curious things about the GDR’s chemical business, Dietrich explains, is that it bred workers skilled at applying chemical sealants to glass, an essential step in the production of a solar cell. So a trio of Berlin scientists and their partner, drawn by this skilled workforce and the ample subsidies available to any enterprise willing to relocate to the former GDR, began production in Thalheim in 2001. Having started in the cottage-scale solar business, they knew how hard it was to find a reliable supplier of PV cells, so they focused exclusively on making those. (This is like making engines instead of cars.) A staff of 19, including the founders, ran that first Q-Cells production line out of a quaint little wooden building. The company’s payroll now numbers more than 2,000; annual revenue topped $1 billion in 2007 and will verge on $2 billion for 2008. The old wooden factory is all but smothered by the sleek, mirrored-glass facility in which Dietrich and I sit.

Dietrich mentions sort of offhandedly that Q-Cells might soon, by virtue of raw revenue alone, become the first TecDAX company ever to jump to the main stock exchange’s DAX 30 — alongside BMW and Deutsche Bank and ThyssenKrupp. “That really feels a bit strange,” he says in his lightly accented English, his tone almost self-deprecating, like it can’t quite be true.

After lunch, Dietrich and I don lab coats and booties to inspect the driver of this miraculous performance — one of the five industrial-scale Q-Cells production lines up and running in Thalheim (two more are under construction). Dietrich leads me to a wide space filled with labyrinthine snakes of gleaming white metal and smoked glass linked together by robotic appendages and conveyor belts. “It looks like some ’70s science-fiction movie,” he tells me, hollering a bit over the machinery’s hum as we watch flat squares of gray silicon cycle briskly through the system. These wafers, Dietrich notes, are barely half as thick as the ones used in 2003, and they come from REC, a Norwegian company that set up the first wafer plant dedicated solely to the solar industry. We follow along as the wafers pass through furnaces and chemical baths, get smeared with silver-aluminum paste and “printed” with electrical contacts. They turn deep blue and develop a mirror’s sheen. In the final leg of the labyrinth, they are tested for efficiency (generally between 15% and 16.6%), stacked in cubes of 100, and packed in logoed boxes for worldwide shipment. The lines run three shifts a day, 24/7, churning out about 150 million cells per year — 585 megawatts’ worth this year, scaling up to a full gigawatt by the end of 2009. The foundation of a new German industrial powerhouse.


As the solar industry reaches maturity and expands into new markets, specialization has fast become the preferred business strategy. Highly efficient but relatively expensive crystalline PV cells such as Q-Cells’ or SunPower’s make sense if you’re trying to power a home with limited roof area, but their efficiency comes at a lofty price. If you have an acre of rooftop (or a 100-acre field), then thin-film cells, which use nano-size layers of silicon or futuristic metal alloys — either cadmium telluride (CdTe) or copper-indium-gallium-diselenide (CIGS) — promise to do the job for a fraction of the cost. The thin-film game is dominated by ambitious startups, but Q-Cells, unique among first-generation solar companies, has jumped into the fray as well, launching a spin-off CdTe producer called Calyxo.

Unfortunately for the Germans, second-generation solar already has a superstar: Arizona-based First Solar, which has developed a CdTe panel for a production cost of an astounding $1.14 per watt, less than half the cost of its nearest rival at its debut. And by the end of 2009, the company plans to manufacture its revolutionary panels at gigawatt scale. CNBC’s Jim Cramer has dubbed First Solar the “Intel of solar,” helping to inflate its stock 1,000% over the course of 2007, peaking at 253 times earnings. The company’s name now rolls off the tongues of even its most combative competitors with naked admiration.

Thanks largely to First Solar, second-generation thin-film technologies are now expected to grow even faster than the crystalline industry and to move from about a 14% market share to as much as 28% by 2010. That kind of growth attracts a crowd, of course, and already a would-be usurper claims to have bested First Solar’s vertiginous dive down the cost curve with a rival technology. Nanosolar, based in San Jose, produced new CIGS panels for a test project last December that it claims will sell for 99 cents per watt — 80% below the average for crystalline PV and more than 10% less than the production cost of First Solar’s thin-film panels. That’s low enough to flirt with grid parity in many markets even without a feed-in tariff. Nanosolar has secured $500 million in venture funding, $300 million of which it claims is still in the bank. That includes a $50 million deal with EDF of France, one of the world’s largest utility companies — just the sort of partnership that lends instant credibility to a risky new technology like CIGS.


Nanosolar’s numbers have attracted such skepticism that its CEO, Martin Roscheisen, felt compelled to post a video clip to his blog demonstrating his company’s new production tool in action. The video (which has drawn nearly 100,000 viewers since it was posted to YouTube in June) is a minute long and completely silent, with home-movie production values. Its sole subject is a large white chamber, which houses a machine capable of spreading Nanosolar’s patented “nanoparticle ink” (CIGS semiconductor material in liquid form) onto a roll of flexible backing at the rate of a gigawatt per year.

According to Roscheisen’s blog, the tool set the company back all of $1.65 million, which in this capital-intensive business is the equivalent of buying a license to print money. In fact, Roscheisen’s revolutionary PV machine has more in common with a printer than it does with anything you’d find on the factory floor at Q-Cells. And if all goes according to plan, by early 2009 the machine will be the centerpiece of two up-and-running production lines, including a 500-megawatt facility located in — naturlich — a down-at-heel East German industrial town not far from Berlin.

In Nanosolar’s repurposed beer-crate factory on the outskirts of Luckenwalde, I find Erik Oldekop, head of German operations. Oldekop offers me a tour of the half-empty factory floor, which is quickly filling with a range of conveyor-belted equipment and white robot arms to swish the cells from stage to stage in the production process. Oldekop points out a laminating machine that stacks 16 of the industry’s standard laminators on top of each other — a sixteenfold amplification, in other words, of a standard production line’s throughput. “There’s no reason why you couldn’t have 100% of electricity generated by renewables,” he tells me. “I’m not saying by solar, but by renewables, and solar is going to make a large contribution.”

Anticipating my skepticism, Oldekop then draws an analogy between the energy business and the ’80s-era conventional wisdom about mainframe computers. Nanosolar, he tells me, intends to begin by selling to Europe’s myriad municipal utilities, small operations that can wire a 1-to-10-megawatt solar farm directly into the local grid, bypassing the costly national-grid transmission apparatus entirely. “Isn’t the electricity company that actually uses central power plants — isn’t that the mainframe?” he suggests. “And we have very small power plants, 1 to 10 megawatts. Isn’t that the PC?”

Radically distributed power production. That’s the kind of talk that gets pulses racing, and I have to admit mine’s doing exactly that as I follow Oldekop across the factory floor. Is this a Gatesian figure I’ve found here in the sleepy Saxon countryside? Could the solar skeptics be akin to the experts who reckoned, just 30 years ago, that computers would always be great whirring beasts that could never become more numerous than, say, power plants? Just 10 years ago — maybe as recently as five — I’m sure I could’ve filled Nanosolar’s conference room with energy wonks who would have sworn I wouldn’t be touring the nearly operational factory floor of a thin-film PV plant in eastern Germany in 2008. Of course, not even the Oldekops of the solar business think little sun-fueled power plants will be as ubiquitous as laptops a decade from now. But beyond that? The sky’s the limit.

As persuasive as I find Oldekop’s analogy, I still haven’t laid eyes on a thin-film solar panel actually pumping out power by the kilowatt-hour. Fortunately, although First Solar is headquartered in Tempe, it too has a major manufacturing facility in eastern Germany. And so on one uncharacteristically bright morning, I take two trains and three buses out of Leipzig, then hike the last mile or so down a country road until I come, finally, to the site of First Solar’s — and the world’s — first industrial-scale thin-film power plant.

Maybe it’s the Kremlinesque veil of silence around First Solar — the unanswered emails and voice mails, the word from its local installation partner that there was “no chance” of a tour — but as I trek up a dusty side lane, I half expect a Stasi jeep to come reeling around the corner. I slip past a low concrete restraining wall to find a high chain-link fence stretching away in either direction to points far over the horizon. Through the crosshatch I can see aluminum frames lined up row on row over an undulating pasture, filled to varying degrees with First Solar’s thin-film panels.

So it exists. And it’s growing: I can hear tools pounding on metal in the distance, and my six-month-old press release tells me at least 12.7 megawatts of thin-film solar produced here is powering German homes as I watch. When it’s done, it will be, at 40 megawatts, Germany’s largest solar installation, marking the commercial debut, after decades of lab-rat tinkering, of a second generation of solar power.

I have no direct experience with the wilds of Silicon Valley circa 1980, but I wonder if it felt about the same as this desultory industrial park in suburban Canberra, Australia. There’s one of these, at any rate, on the outskirts of every city in the industrialized world — a bland agglomeration of welding shops and fencing suppliers and landscaping companies housed in aluminum sheds in a dozen shades of beige. On one corner of this one, there’s a small, single-story brick building with a couple of loading bays. Inside, I find an engineer hunched over a standard screen-printing machine, more or less identical to the ones they use to print Your Name Here on T-shirts down at the mall. He’s using it to make solar cells.

The engineer’s name is Graeme Evans, and he works for a small startup called Dyesol. He’s dressed in a slightly ratty golf shirt, and if not for his blue surgical gloves, he could be working the drill press in his suburban garage. He’s using the screen printer to spread a thin layer of yellowy goo on the surface of a rectangle of glass the size of a postcard, smearing it through a sort of stencil that divides it into six smaller strips in two rows.

The goo is titania — titanium dioxide, more precisely, a plentiful, electricity-conducting material commonly used in toothpaste and paint — and once it has dried, it leaves a porous coating of nanoparticles with an extraordinarily large surface area for such small strips. The glass panes will then be dipped in a rust-hued dye consisting most notably of a little-known metal called ruthenium and then fused to a second piece of glass coated in electrolyte. And that’s how Dyesol makes its photovoltaics — “dye solar cells” by name, a technology based on a breakthrough that emerged from a Swiss university in the late 1980s. The ruthenium dye absorbs available solar energy the way chlorophyll does, taking in electrons and transferring them to the titania layer to create electricity.

“The principle is just like a leaf,” explains Sylvia Tulloch, cofounder of Dyesol. We’re around the corner from the R&D building in a little green structure that looks like it should belong to a screen-door wholesaler. She points to a coaster on the conference-room table and to a patch of red berries in the festive scene it depicts. “As long as you can see that that’s red, you know that it’s absorbing light. And so it’s not dependent on how much light is hitting it.”

To her left, a handful of Dyesol cells have been mounted in a trade-show display stand. There are six small fans affixed to the sides of the stand, and they whir quietly as we talk. It’s only when I stare down at the red berries on the coaster in the shadowy light that I get what’s odd about that: We’re indoors, and the room’s windows are half-shaded by Venetian blinds. Dyesol’s cells are spinning those fans with essentially no direct sunlight at all.

“That’s one of the key advantages of dye-solar-cell technology,” Tulloch explains. “It accepts light from all directions; it accepts light in all light conditions. And the other key advantage is its manufacturing process. You need very sophisticated equipment for either the first or the second generation, but for dye solar cells, there are kits sold for children. My son, when he was 9, made one and did a demonstration at school.”

Dyesol is one of a rapidly expanding roster of firms worldwide experimenting with this third generation of solar technology — a subsection of the lab usually referred to as “organic PV.” Dyesol’s key distinction, though, is its startling proximity to market readiness and the name of the business partner intending to bring it there: Corus, the industrial behemoth formerly known as British Steel. In early 2007, Dyesol signed a $1 million contract with Corus to assess the feasibility of incorporating dye solar cells into its prefinished-steel-roofing materials. Corus churns out 100 million square meters of this Colorcoat roofing for use in factories and warehouses each year — more than enough to reroof every Wal-Mart in America — and the process already involves applying layers of paint. Replace some random decorator color with the ability to generate clean power and the appeal would be obvious and enormous — particularly in Europe, where makers and buyers of building materials are increasingly required to account for the emissions involved in producing them.

By the end of 2007, the project had cleared what Tulloch calls “the area of highest technical risk,” in which it was determined that Dyesol’s cells could in fact feasibly be printed on a massive unspooling roll of steel as it zooms down a production line at 3 to 5 meters per second. The government of Wales has since invested in the project, and Corus has converted one of its Welsh production lines into a demonstration facility for solar-coated steel roofing. The test phase continues through 2009, and there’s little chance the product could be on the open market before 2011 — which is likely why Corus is declining comment on the technology’s potential for now “as a way of managing expectations.”

Fair enough. After all, the prospect of Europe’s second-largest steel producer integrating solar cells into 100 million square meters of roofing per year might set certain fevered minds racing.

“Can you imagine metal roofs all around the world that are power generators?” Dyesol COO Ross MacDiarmid asks me.

The truth is, I can. It doesn’t even seem like an act of imagination anymore.

Chris Turner is the author of Geography of Hope, a global survey of sustainable technology.