Practical Chip Design

SEMI, always quick with the attractive luncheon topic, came up with a house-packing one today: how electronics companies can make money from the current explosion of investment in solar energy generation. In fact the speakers addressed this question only indirectly, by describing the marketing and fabrication of photovoltaic cells, but none the less the house was packed, and no one wandered away in mid-lecture.

Two speakers kept the audience in their seats. First, Paul Basore, VP technology for the Scandinavian-based Renewable Energy Group, discussed the basic economics and marketing of photovoltaic panels. Then Mark Pinto, CTO, senior VP and GM at Applied Materials, talked about the technology and economics of fabrication. While neither of the talks directly addressed the question of how this would pay off for the electronics industry, both left plenty of room for inferences—and some pretty interesting ones at that.

Basore led off by distinguishing between the two primary kinds of photovoltaic cells used to make up panels: multi-crystalline silicon cells built from polycrystalline silicon wafers, and thin-film cells build by depositing semiconductor films on glass to form PIN diodes. The former, he said, are typically larger and are currently more efficient in energy conversion, with a 156 mm-squared polysilicon cell generating a maximum of about 3.5W. In contrast, a 6 mm by 1.2 meter-long thin-film cell generates about half a Watt.

Basore then launched into the market discussion. The growth of the market for photovoltaics, he said, has been determined by two major factors: sunshine and subsidy. This has led to their deployment first in Japan, then in parts of Europe—especially Germany—and now, in early stages, in California, USA. He defined a single figure of merit, Grid Parity, that will control much of this growth in the future. Grid Parity is the point at which the cost of electric energy in useful form is equal between a bank of solar panels and the electric grid in the area. Basore quickly pointed out that the computation of this figure is complex, depending on everything from weather and politics—the two great discussables—to local power rates, maintenance costs, and the opportunity cost of covering a given space with panels.

But a major factor in the calculation is initial price. This is driven by all the usual suspects, but particularly materials—the cost of which is enhanced by limited capacity at the moment—manufacturing efficiency, and conversion efficiency. Of these, the first two appear to be more dynamic than the third, which seems to be remarkably resistant to major advances.

Looking deeper into the manufacturing question, Pinto took over the podium, talking about how the needs of the display market had created a manufacturing infrastructure—the ability to manufacture huge sheets of highly uniform glass and large, very thin poly wafers—that could be applied to panels. He said that plants coming on line today had capacities in the range of 50 MW/year, and that we would soon be seeing plants capable of producing a GigaWatt in panel capacity per year. The increase in volume is primarily a result of the huge size thin-film panels will soon reach: 5.7 meters-squared. Not only are the huge panels more efficient to build, he said, but they required fewer connections and control boxes when installed, and could be used as architectural elements, not merely as big ugly slabs on the roof.

Pinto said—as many have over the last few years—that photovoltaic panels are an industry on the brink. “We are within factors of two of Grid Parity in Europe today,” he claimed. At this point, he suggested, there were fortunes to be made by finding innovations in the structure of thin-film panels, where there still seem to be great ideas to be found. Basore added that there were also opportunities to innovate in polysilicon cells—finding an alternative for the aluminum film used to create an alloy passivation layer on the back of the wafer, for example.

But for the electronics industry, except for a low-margin business in controllers and power converters, the opportunity may lie elsewhere: not in selling stuff into the solar panel market, but in exploiting the impact this market will have on electronics.