by Paula Mints, Navigant Consulting
March 25, 2009 – Selling equipment to the photovoltaic industry has traditionally followed a machine-by-machine paradigm: A tabber/stringer here, a furnace there, deposition equipment, etc. Manufacturers developed individual manufacturing lines based on the needs of their technology and their production. This “not invented here” (NIH) philosophy of manufacturing, along with over 30 years of industry unprofitability, made a turnkey PV manufacturing strategy less than attractive for equipment manufacturers.
Then in 2004 came a solar boom followed by a silicon shortage right behind. Accelerated demand for solar products smacked a significant raw material constraint for crystalline technologies, giving thin films an opportunity to enter. Unfortunately, at the time there was very little manufacturing capacity for thin films — simply, the technology had not needed it. Crystalline technologies, higher in efficiency and perceived as more reliable for years by installers of solar electric systems, held a 90% or higher share of the market for many years. The figure below highlights the long road that thin films have traveled over time to achieve a 14% share of technology shipments in 2008, showing the contribution of thin-film and crystalline technologies over time.
Thin-film and crystalline technology contribution to sales, 1998-2008. (Source: Navigant Consulting)
Though the conversion efficiency of thin-film technologies is lower — often significantly so — than that of crystalline, advantages are potentially lower manufacturing cost (more cushion in the margin when setting price) and better operation in low light conditions. When demand for photovoltaic technologies exploded in 2004, investors flocked to thin films believing that low cost was the answer, and that more-expensive-to-manufacture crystalline technology was on the wane. Startups were funded on the basis of being thin-film (particularly CIGS, amorphous silicon, and micromorph) or being perceived as new (again, CIGS, amorphous silicon, and micromorph) — despite the fact that most thin-film technologies have been in development or commercial production for ~20 years.
The other reason for investment in thin films was perceived low manufacturing cost. It is important to note that cost and price are driven differently — cost is raw material and economy of scale driven, whereas price is a market function. Low manufacturing cost will not necessarily lead to lower prices, nor will it naturally translate to sales. Thin-film technologies are lower in efficiency than crystalline, and lower efficiency leads to higher system installation cost because of the area penalty. In short, the installer needs more of everything down to the nuts and bolts than for higher-efficiency crystalline systems.
Enter semiconductor equipment manufacturers, eager and certain that they could change the paradigm from one machine at a time, to ready-to-go turnkey facilities. Initially these manufacturers believed that the synergies gained from similar industries (e.g., flat-panel displays) would lead to faster development timelines. They also assumed that the lower cost to manufacture thin films would lead to market leadership.
Not so fast. Most sales of turnkey lines were to new entrants, manufacturers from other industries (primarily semiconductor) who believed that low cost was the key to success and were experienced with turnkey solutions to manufacturing. It has not proven easy for these new entrants, and very little commercial product has come off of their turnkey lines. Moreover, in most cases, production on the lines has become more customized as the new manufacturers begin a development process closer to that of the traditional photovoltaic industry. Quite simply, it is not easy to make this stuff work, and it has to sit in the sun, in all weather, and produce electricity for 25 years — reliably.
That does not mean that the turnkey philosophy has failed. It is, however, becoming less static and more customized. Sales of customized turnkey lines will take a larger share of the equipment market in time, just as some sales will be true turnkey. Most equipment manufacturers will need to adjust to the one-machine-at-a-time paradigm, and all will try and control the consumable contracts that typically make up a significant percentage of after-sales revenues.
The photovoltaic industry is a >30-year-old startup. It is maturing and will evolve. The nontrivial aspect of technology development is unlikely to change, though it can and likely will adapt.
Paula Mints is principal analyst, PV Services Program, and associate director in the energy practice at Navigant Consulting. E-mail: [email protected].