Solar energy's mainstay is its ability to deliver electricity at prevailing local price. But, solar's own popularity may be working against its price. The reason? Supply of the raw material, crystalline silicon, hasn't kept up with skyrocketing PV demand.
The brass ring for solar energy has long been "grid parity," or delivering electricity to the grid at the prevailing local price. But the solar photovoltaic (PV) march toward parity has been delayed, ironically, by the sector's own surging popularity.
The main reason? Supply of the raw material, crystalline silicon, hasn't kept <span style="font-size: x-small">up with skyrocketing PV demand,</span> resulting in severe shortages and high prices. But with new factories and new technology on the horizon, <span style="font-size: x-small">the supply shortage may end soon, which means the prices could drop like a rock</span>. Or maybe not.
Confused? Read on...
It Starts with Silicon
The raw material for solar PV cells, which turn sunlight into electricity, is crystalline silicon. The semi-conductor is better known for making electronic chips used in computers, cell phones, cars and countless other products.
<span style="font-size: x-small">For both electronics and solar cells</span>, the silicon is sliced into very thin wafers before other steps lay down the needed circuitry. Since electronics are worth a lot more than solar cells, those chip foundries can easily outbid solar manufacturers for the silicon. This has left solar cells to be the metaphorical frankfurter of the semiconductor world: mostly made with scraps, rejects and leftovers.
And that makes perfect sense for technical reasons, as well for the economics. A solar cell simply doesn’t demand the same quality as a computer CPU chip. A small flaw that would make a CPU inoperable could still make a perfect solar cell, or at worst, one that is less efficient but to a trivial degree.
Growth Leads to Shortage
Demand for solar cells has gone up at an incredible rate. Grid connected PV increased worldwide last year by 83% to 8.7 gigawatts. About half of that increase was in Germany, which now leads the world in installed PV. This demand has been met by new manufacturers entering the business and existing producers doubling their capacity or growing even more. The result? Solar cell production has doubled in the last two years.
But shortages have remained, keeping prices high. The bottleneck hasn't been the solar cell fabricators, but the raw material, the crystalline silicon.
Last year, the demand for silicon rose to 75,000 tons, but production was only 40,000 tons. And demand is still increasing 40% each year. Naturally, the price for solar silicon has been driven up, too. It's now at $65 to $120 per kilogram for long-term contracts, and it's much higher in the spot market, as much as $300 per kilogram.
The silicon shortage has been so severe that stock prices for PV manufacturers have soared when supply deals have been announced, typically a non-event. That's unusual stock market behavior. Normally a company’s stock jumps when it announces a big deal to sell, not to buy.
Such is the perception of the business on Wall Street these days. The constraint on revenue isn’t selling the finished product; it’s acquiring the raw material. Anyone can sell the stuff, but finding new silicon is like finding gold.
Shortage Leads to New Supplies
But lest you worry, relief from the silicon shortage and the high prices is coming. Some new plants are on line and more are coming soon. Even more interesting, two different companies have developed new PV technologies that could take over the market and eventually drive the prices down far enough for grid parity.
News on conventional plants first...
The standard fabrication for silicon is called the Siemens process. One of the largest silicon makers, Michigan-based Hemlock Semiconductor, has nearly doubled capacity this year to 19,000 tons and will double again by 2011. Siltronic and Samsung started production at a new plant last month. Other large producers are also planning expansions.
New Technology #1
Renewable Energy Corporation (REC) says it is the world’s largest provider of solar silicon and PV wafers. The company sold 3,600 tons to PV plants last year and 2,400 tons to electronics fabricators. Rather than the Siemens process, the firm developed a fluidized bed reactor.
REC expects to more than double capacity in 2008. Its plans are to focus on solar moving forward, so as current contracts to deliver electronic grade silicon expire, it will move the output to solar.
New Technology #2
Reaction Sciences Silicon Products (RSI) won the 2007 MIT Ignite Clean Energy business plan contest for an entirely new way to make silicon crystal at one-third the cost. Travis Bradford, president of the Prometheus Institute had this to say about RSI:
This new breakthrough process could have a dramatic impact on the cost of future solar cells, accelerating market growth and future grid parity.
RSI makes silicon at $7 to $9 per kilogram versus $25 to $30 for a Siemens plant. (Sorry, couldn't locate a price for a fluidized bed reactor.) Remember: This stuff is selling for $65 to $120 per kilogram, so getting the price down to $7 plus a reasonable profit is going to make a big difference to the price of finished solar cells. The cost of silicon in current PV cells is about $1.50 per watt. RSI claims it can reduce it to 25 cents per watt in five years.
RSI’s advanced solution was invented by the company's chief technology officer, Steven Amendola. He bolted together several standard chemical industry processes that use fairly standard equipment. Nothing too sexy, nothing too risky, just some reliable well-proven workhorses that happen to be a lot cheaper.
And the main reason they're cheaper is that they don’t produce electronic grade silicon. Electronic grade silicon is eight 9’s pure (that is, 99.999999% pure), but solar grade silicon -- the "hot dog" grade, if you will -- is merely six 9’s pure (or, 99.9999%). Both the Siemens process and REC's reactor make silicon that is eight 9's pure. Not only can RSI use simpler, cheaper processing but they can start with a lower cost raw material instead of metallurgical quality silicon.
A simpler process means a simpler plant, too, one that can be built in half the time at a much lower cost. RSI expects that a typical 5,000-ton plant can be built in 15 months. Its first plant -- a 1,000 ton per year pilot plant in Pennsylvania -- has slipped one quarter, as is typical, but should start running this quarter.
Meanwhile, a full-sized, 24,000-ton plant is scheduled to open in Alabama in 2009. The company plans to add 5,000 tons of more capacity in each of the following two years.
Supply Up, Demand Up, Prices...?
Travis Bradford of the Prometheus Institute expects supply to quadruple by 2012 to 125,000 tons, so how much will prices drop?
He's predicting that between 2007 and 2010, the cost of silicon PV will go from $3.66 per watt to $2.14 per watt. A bold prediction. The grid parity price depends on the price on your grid, but $1 per watt for PV is the price most often mentioned as reaching parity. In the US, kilowatt per hour prices range from about 5 cents in Idaho to about 20 cents in Hawaii.
The price will continue to be set by supply and demand, and therein lies the mystery. The supply piece is pretty easy to get a handle on. The money and time needed to build a plant are large enough that new capacity is not going to be a secret.
Demand, however, is the wild card. It's affected by the final PV panel price, the local price of electricity, the local sunshine and perhaps most importantly, government incentives.
The reason Germany now leads the world is due to a program that guarantees a high feed-in tariff, that is, the price the PV owner gets for the electricity. California, the US leader, has a program called the Million Roofs Initiative to encourage installations by sharing the cost with homeowners. Elsewhere in the US and the world there is a mishmash of policies, which are all subject to change. Some cities and utilities have their own programs. Spain has a very successful program, but it's considering cutting it by nearly 70%. There are a lot of moving parts here, so predicting demand and ultimately prices is going to be tough.
So: The new supply should lead to big price cuts, unless demand rises to match the supply and the shortage continues. The good news is that production will keep growing as long as these very high profits continue, and eventually, when supply and demand get close, prices will come down substantially.
Welcome grid parity!
Thin Film
So how does all this affect those thin film PV producers like First Solar and Nanosolar? Short answer: Not much.
They print very thin films on aluminum and use about 1% of the silicon used in the traditional wafers, so they don't use enough to be bothered by the current high prices, nor are they helped by future reductions.
The biggest effect will be a loss of the competitive advantage, which thin film presently enjoys against wafer PV technology by being so frugal with silicon.
</span>
<span style="font-size: x-small; color: #0000ff">
<span style="font-size: x-small">
</span>
</span>