Forecasting thin-film investments and capacity expansions has never been an exact science. However, with the exception today of First Solar, there are few market-based arguments to support the immediate capacity expansion of most thin-film fabs presently struggling to match the downward crystalline silicon cost and price curve.

More often than not, it has been thin-film equipment suppliers that have been at the forefront of marketing the viability of thin-film technologies: not only their customers, who ultimately are the ones directly exposed to volatility in the rate of demand growth, but also an increasing glut of (readily financeable and aggressively marketed) cost-competitive c-Si modules from China.

Nonetheless, there remains an insatiable desire to invest in thin-film start-ups. However, recently publicized investments in the range of US$50-100 million pale somewhat into insignificance compared to the plans being unveiled by China’s Hanergy and the continued vote of confidence in one of their equipment suppliers Fujian Apollo.

The Hanergy/Apollo story is not new, but by virtue of being a wholly Chinese affair, has been less widely publicized outside Asia. In fact, the deals have been tracked most closely by equipment and materials suppliers that stand to benefit considerably from the capacity expansion announcements that come to fruition.

In Apollo’s production equipment design, large manually-loaded, batch PECVD chambers hold 72 thin-film panels vertically in a single deposition architecture, and employ a range of sequential absorber deposition steps to form multi-junction combinations of a-Si and SiGe layers.

Market demand and PV manufacturing heritage aside, the challenge to transition such a technology approach to a multi-GW-level production scale cannot be underestimated.

The first order from Hanergy – worth approximately US$3 billion – was signed in 2010 and called for various phases of turnkey multi-junction a-Si-based production tooling to be delivered to different Hanergy sites across China. However, by the time Apollo reported 1H’11 results during August 2011, the euphoric announcements of 2010 had been replaced by a somewhat more pragmatic assessment of the PV industry as it had evolved by then:

“With the effect of shrinking demand in solar PV modules and declining [sic] in module price, the cost competitiveness of thin-film solar PV module is declining. The potential investors in building thin-film production line are becoming more hesitated [sic]. The tightening in bank loans have imposed stress and negative impacts on our customers… and resulted in deferred and delayed investment in thin-film production lines. The pace of the delivery of production lines to our existing customers has been invariably affected.”

Given the softening of PV demand through 2011 and the caution being exerted into capacity expansion investments by tier 1 module suppliers, this summary was not entirely out of context.

But with less than 20% of the first contract delivered to Hanergy – and (proven) mass production still classed as work-in-progress – there was minimal evidence to anticipate the latest news from Apollo.

Apollo has just indicated that Hanergy has now followed up their initial purchase order (valued at US$3 billion) with another thin-film equipment contract for Apollo. This time, the contract is valued at approximately US$6 billion (split between equipment and sales and service support). The precise phasing of capacity is not identified yet, nor the exact connection to the 2010 requirements for equipment deliveries.

It is worth just putting this new figure in context to get a feel for the numbers involved here:

  • Solar Frontier’s Miyazaki Fab 3 is reported to have involved an investment at the US$1 billion level.
  • The new Hanergy/Apollo contract is greater than the collective sum of all a-Si turnkey thin-film fabs delivered to the industry until today. This includes all lines supplied by Applied Materials, Oerlikon, ULVAC, Jusung and a host of lower-throughput a-Si production lines whose technology – similar to Apollo – is based closely upon a deposition chamber design pioneered by EPV over 10 years ago.
  • Finally, the latest Hanergy/Apollo contract is greater than the total thin-film equipment spending across all a-Si, CIGS, and CdTe expansions covering the whole of 2010 and 2011.

Indeed, the new contract calls out for production equipment comprising over 1,000 PECVD chambers and more than 160 PVD sputtering tools. Based on estimates of MWs-per-PECVD chamber, the total nameplate capacity for this contract would translate to over 6GW of a-Si thin-film panel output annually.

Even downgrading the nameplate specification towards an effective manufacturing capacity figure closer to 5GW still represents an incredible amount of capacity addition to the industry at a time when a small group of 15 tier 1 PV producers has the capability of supplying a large proportion of end-market demand to 2015.

There has been no shortage of GW aspirations announced by tier 2 and tier 3 participants during the past couple of years, including many new entrants to the PV industry. Often, such announcements are merely accompanied by images of ribbon-cutting ceremonies and artists’ impressions of what these GW fabs may look like in the future.

During 1H’11, as the stark reality of supply and demand became clear, many of these plans were shelved or significantly cut back. Many companies, however, were already committed to equipment deliveries through 2011 and despite the best efforts from lower-tier equipment suppliers to cancel or push out 2H’11 shipments, the legacy of Q4’11 may simply be one of equipment crates awaiting to be opened at the first signs of a market pick-up.

By the end of 2011, Hanergy will have taken delivery of turnkey a-Si production lines from two different equipment suppliers with a nameplate capacity approaching 500MW, and at a cost just below 1 US$/W. What differentiates many of the thin-film announcements, however, is the timescale to mass production (simply by equipment delivery times, fab construction, or ramping to production) which is typically longer than for c-Si capacity expansions. As such, investments into thin-film expansions (with the exception of First Solar) are not directly tied to short-term volatility in PV demand, but more to longer-term, external strategic investment goals.

Investments into thin-film technologies continue to surprise many, but only serve to emphasize the aggressive ambitions of industry ‘outsiders’, keen to have a significant impact on industry production capacity in coming years.

Thin-film technologies (especially a-Si and CIGS variants) are still considered as leading candidates in any investment that seeks to be differentiated from the (currently dominant) China-driven tier 1 c-Si approach. Indeed, with limited options to compete directly on price/branding/marketing with the leading c-Si players in the foreseeable future, it is entirely feasible that the thin-film segment will continue to see strong investments – almost regardless of any close examination of the commercial and technical track-record of its participants to date.

In the short term, the equipment supply-chain to Apollo may be the greatest beneficiaries, but it will be down to Hanergy to meet industry cost-driven requirements in manufacturing to merit subsequent fab build-outs planned. However, the success of the Hanergy/Apollo contracts may ultimately be judged upon the ability of Hanergy to subsequently finance downstream project deployment at a level not previously seen within the industry.