Why Thin-Film PV Is Struggling, But Not Done Yet

Equipment supply to the solar photovoltaic (PV) industry has changed significantly since Oerlikon’s entry as a turn-key supplier of a-Si based production lines in 2006. While not exclusively focused on entire turn-key lines, success ultimately would be based upon the market-adoption of amorphous/microcrystalline (a-Si/uc-Si) silicon tandem technology, market-traction of early adopters, and — crucially — repeat orders for turn-key fabs.

Similar to Applied Materials (AMAT), initial bookings came from the European thin-film investment phase of 2006-2009. Oerlikon benefited from new a-Si thin-film entrants in Europe such as Inventux, HelioSphera, Gadir and Pramac. This was followed by traction in the Asia-Pacific region, and in particular from Tianwei, Astronergy and Auria.

By the middle of 2008, Oerlikon’s backlog stood at an impressive $676 million, and the basis of optimism was clear to see. Indeed, adding in AMAT and ULVAC’s thin-film order intake at that time provided a total a-Si turn-key backlog that exceeded $1.5 billion.

Alarm bells started to sound, however, as a-Si based suppliers struggled to gain significant market share, were exposed with panel efficiencies below 10%, and had manufacturing costs that were not aligned with how the industry had been shaped by the c-Si expansions across the APAC region. Fab utilization rates for many a-Si manufacturers ran well below (modeled) break-even status, investments into a-Si capacity dried up, and the entire thin-film segment was required to adjust strategies to compete with China-based c-Si panel dominance.

While AMAT and ULVAC pulled back or officially exited from the turn-key a-Si equipment market, Oerlikon’s response was to redesign the turn-key technology. Targets were now being set by First Solar’s efficiency and cost roadmaps, not necessarily by Tier 1 c-Si market-leaders. Capex, throughput, efficiency, and operating costs were all reduced significantly. The result was the launch of the ThinFab line. Indeed, Oerlikon had effectively become the marketing voice of their existing customer base — but had few (competitive) allies to assist in championing this message to the PV industry as a whole.

By this time, investments into a-Si capacity had declined to very low levels, and what new investments were coming into this segment were being driven from China or from Sharp/3Sun — where equipment-supplier preferences represented a somewhat different challenge to European-based Oerlikon. Jusung and Apollo had also entered the turn-key a-Si equipment supply-chain, providing a new competitive threat within a significantly different PV supply-chain management environment. Hanergy unveiled its plans to invest up to US$6 billion into Apollo and populate a-Si fabs across China with a competitor’s production line that was being manufactured within China. Clearly, new forces were now driving equipment spending in the a-Si segment.

Oerlikon did succeed in capturing several new customers within China, but the gap between new thin-film entrants and the leading Tier 1 c-Si market leaders was by this time significant. During 2010 and 2011, new order intake had become extremely lumpy — at a time when all PV equipment suppliers were confronted with an industry that had overcommitted to capital equipment, and was consequently set for a long and severe downturn.

Oerlikon’s decision to divest its solar segment was an easier decision to make. Many other PV equipment suppliers have no adjacent market segments to defer to until PV capacity expansions return.

While the prospects for turn-key thin-film equipment suppliers are currently going through a severe downturn, they are not alone. New order intake has declined considerably for every PV equipment supplier during 2012, with 2013 showing very few signs of offering anything more than moderate pick-up for a select group serving the Tier 1 expansion plans. And it is way too early to write off any PV technology based purely on the events of the past 12-18 months — especially not a technology like a-Si that has seen significant development since the mid 1970’s, and has accumulated over two decades of manufacturing know-how and panel deployment.