Thesolarindustry can potentially reduce costs 40 percent over the next five years as the silicon shortage ends, according to Graham Stevens, an associate director at Navigant Consulting.

At the Intersolar North America conference in San Francisco this week, panelists discussed different ways to reduce those costs.

For example, Roy Johnson, CEO of Calisolar, said that the cost of producing upgraded metallurgical-grade silicon, also known as UMG silicon, can potentially be one-sixth that of making polysilicon.

The company plans to make cells from 100 percent UMG silicon, which Johnson said is three orders of magnitude less pure than polysilicon, and is aiming for efficiencies of 16 percent to 17 percent (see a Q&A with Johnson here).

While previously, many in the industry thought that upgraded metallurgical silicon would cost too much and deliver efficiencies too low to be worthwhile, higher polysilicon prices have opened up a new opportunity for the material, he said.

“Nobody paid attention when silicon was at $55 a kilogram,” he said. “It’s the same way that $150-a-barrel oil is good for most people in this room.”

Now, he said, virtually all the leading metallurgical silicon producers around the world are moving into UMG silicon, and even Q-Cells, the world’s largest solar-cell manufacturer, plans to make cells from the material.

Making cells from UMG silicon requires new technology, he said. “You can’t just take the same technology and apply it to the UMG material – you will be disappointed,” he said.

But Calisolar thinks it’s worth the trouble, even if the polysilicon shortage ends soon. A number of analysts have predicted that polysilicon prices will drop as companies build additional polysilicon capacity more quickly than the industry’s projected demand.

While Johnson said there’s no question companies are growing their capacity, he said the prices won’t necessarily decrease.

“It depends entirely on demand,” which could increase more than expected, he said. “If oil reaches $200 or $300 a barrel, how much more attractive will solar be then?”

And even if silicon prices do decrease because capacity increases more than demand, he said, Johnson thinks UMG still makes sense.

“The UMG guys still win because the UMG guys can ramp much more quickly,” he said.

Johnson cited Timminco’s subsidiary Becancour Silicon as an example. Becancour has said it expects to build 1,000 tons of UMG silicon capacity for $15 million in six months, while polysilicon plants cost about $100 million for 1,000 tons of capacity and take about three years to build.

The technology also has plenty of skeptics, who have questioned whether the costs really will be any lower, relative to the lower efficiency, than making polysilicon (see Timminco Shares Down Despite Deal with Canadian Solar, Timminco Sues for Libel, Scientific Debate Gets Complicated and Solar: Doing the Dirty).

It also has its advocates.

Eicke Webber, a cofounder of Calisolar and chairman of the Intersolar North America conference, said he believes upgraded metallurgical silicon could potentially reach efficiencies of only 0.5 percent less than polysilicon for a far lower cost (see Incentives, Tech to Spark Debate at Intersolar: Metallurgical Silicon).

Thin-film solar is another technology with the potential to reduce costs.

First Solar, for example, reached costs of $1.10 per watt in the fourth quarter.

But aside from First Solar, thin films haven’t yet been “a stunning commercial success,” said William Mulligan, vice president of technology at SunPower Corp.

“It’s certainly been a long time coming,” he suggested, adding that most thin-film technologies have been around for 30 years. “What has changed aside from a lot of money in the space?”

Still, he said, the levelized cost per watt will be the determining factor for success, and First Solar’s future cost-per-watt projections are similar to SunPower’s.

“We don’t discount their technology,” said Mulligan, “but there’s still a lot of head room in silicon. It’s not as much a mature technology as [people] often believe.”

According to Stephens, the silicon shortage has changed the economics, making technologies such as thin films, which use little or no silicon, more attractive.

Today, he said, some customers will take any panel they can get to make money by taking advantage of the feed-in tariff, such as in Germany. That dynamic could change if more panels become available.

Stevens sees the end of the silicon shortage coming, with a tipping point at the end of this year or the beginning of the next.

“This will be an interesting race going forward,” he said.