I had the recent opportunity to listen to a discussion led by Robert Ford, CEO of Solaicx, on the state of the silicon supply, speaking at a luncheon arranged by the Antenna Group.
With c. $20 mm invested, Solaicx is preparing to build a commercial plant for the manufacture of single crystal silicon ingots (and wafers) to the photovoltaic solar module industry, and has been providing product from its prototype plant to GE Solar for some months.
From their website: “The Solaicx system includes a continuous Czochralski (CZ) crystal grower that implements several Solaicx breakthrough technologies, allowing the manufacture of low cost, high quality single crystal silicon ingots at high volume for conversion into solar wafers.”
When I asked about his take on the silicon supply / demand curves for the next couple of years, Bob’s answer paraphrased: “There are announced capacity increases from c. 35,000 metric tons to c. 100,000 metric tons over the next 4 years, and maybe half a dozen additional unannounced capacity additions. They also see an up to 65% reduction in silicon usage per unit over that time frame (coming, if I understood him correctly, from a combination of reduced amount of silicon per unit of area and increased cell efficiency). The upshot of a 3x increase in supply and 3x improvement in silicon utilization equates to an effective capacity increase of 9x by 2010 or so, making for a very short lived supply shortage.”
Solaicx’s bet is that their process, which they claim is up to 80% more efficient than conventional crystal growing, will give them a low cost advantage as the silicon supply situation eases.
The other comment he made that I found particularly catchy came in his discussion of silicon manufacturing for the photovoltaic industry:
“Solar is a dumb, dirty diode”, and basically a materials handling business when compared to semiconductor sector, “not high precision manufacturing”. The implication was that when the industry focused on making solar grade silicon cost effectively (as Solaicx has targeted) instead of semiconductor grade product, that the results would be significant improvement in cost and supply in the solar silicon market as manufacturers optimize for the PV industry.