The Future of Polysilicon

A recent report (Polysilicon Industry Faces Shakeout) notes:

"...fluidized bed reactor technology has not delivered on its promise of lower manufacturing costs."

The funny thing is, I've heard the exact opposite from insiders as recently as a few months back. Is FBR going to take over the poly space overnight? No, but we should start to see a shift in what kind of plants get built in the medium term - i.e. over the next 5 years.

Comparing Then to Now

THEN (2004-2006): Polysilicon was bottlenecking and prices were rising. You could sell any poly you made for a handsome profit. What do you do when prices are high? Add production capacity so you can sell more... Given a choice between Siemens and FBR plants, which are the refiners going to build? One could reason that the shovel ready refinery projects with short lead times (real or perceived) had a clear advantage. That means Siemens refineries.

NOW (2009-2010): Spot market and contract prices have been trending down continuously for two years. There is excess poly supply with still more coming down the pipe. The expectation is for poly prices to trend down in the future. We're projecting more demand for PV which means more demand for poly but the confidence level of these projections isn't there yet. The only course for a refiner to take in this sort of environment is to methodically plan capacity expansions. In this sort of environment the advantage shifts from short lead times to low production costs. That means FBR refineries.

If REC or MEMC announce yet another 10,000 MT FBR plant in the next year or so (which I expect) we'll have a strong indication that FBR delivers a cost competitive product.