During the past couple of weeks, two of the leading custodians of thin-film solar PV technology have restated or amended long-term industry plans. The two companies in question are First Solar and Hanergy Solar (shortened to Hanergy in this blog).

Given there is only a handful of thin-film companies whose long-term strategies hold much weight these days, this provides a good chance to see what the thin-film segment is thinking at the start of 2014.

Where these two companies have come from could hardly be more different. What they are doing today in the PV industry also has few parallels.

But they share one common dream that, for the past few years, has appeared to be somewhat off limits to many: the long-term competitiveness of thin-film PV when compared to c-Si, and how thin-film will still come good on efficiency and cost.

Before we look at some of the details, it is worth reviewing where thin-film stands in the industry now, and what is driving the manufacturing plans of the companies in question.

Market-share by fab expansion

When the solar PV industry was a 5-20GW annual market, the prevailing wisdom was that adding capacity was the answer to expanding market share. It was thought then that revenues could only increase if shipments increased, assuming (of course) that ASPs would not decline too much.

Therefore, to grow a business meant adding capacity: simple enough. Everyone was at it until 2010, some more so than others. Banks, investors and governments queued up to champion even the most ambitious thin-film business plans or technologies (often pitched as a means of generating home-grown cleantech jobs). Between 2006 and 2012, more than US$17 billion of capex was allocated to new thin-film PV fabs.

But when overcapacity hit the PV industry in 2012, c-Si and thin-film makers were impacted differently and the fallout of this can still be seen today for the thin-film segment in particular.

Thin-film producers (by virtue of product line and process differentiation) had to think about what their existing capacity levels meant, and whether they could afford to run them. Or, was there an option other than ‘make-and-ship’ to third-party downstream customers?

Conversely, overcapacity meant that leading c-Si suppliers were in a stronger competitive position, with flexibility to outsource production to allow shipment levels to increase quickly as the market demanded. The fact that there was so much c-Si capacity at lower tier makers (especially in China) made this an even greater benefit to c-Si leaders. The cross-fertilisation of know-how in China further enhanced the options available to Chinese c-Si producers. The concept of fab-lite operations quickly came into being.

Indeed, at the start of 2014, if the likes of Yingli Green or Trina Solar need an extra gigawatt of capacity (from ingot to module), this can appear within six months at a cost of US$250-300 million. Halve this number if distressed companies are used for capacity; reduce it even more if secondary tooling is used or leasing arrangements are put in place. But better still, let someone else produce the product and rebrand it, and have no new capacity-related costs at all.

Thin-film makers can’t do this obviously from a module sales perspective. But they would have to if they wanted to grow their downstream business more than their in-house production capacity allowed. Otherwise, downstream business (for new projects realised within a pipeline in any given year) is capped by in-house thin-film capacity utilisation rates.

For example, if a leading thin-film maker that has downstream projects business were to run into any unforeseen manufacturing problems – or wanted to expand its projects business very quickly – then the only route would be to buy in c-Si modules.

After all, project developers are in the business of selling energy (or selling assets) and most retain a technology-agnostic approach: so why should vertically integrated thin-film producers/project developers not give their downstream teams the flexibility to source modules in-house or from c-Si makers, if there are short-term business opportunities to grab?

New thin-film capacity: the strategy of last resort?

All thin-film manufacturers (active or dormant) are currently still feeling the impact of their capacity expansion strategies between 2006 and 2012. This includes First Solar that shifted its strategy a few years back from ‘new-fab-capacity-supply-pushed’ to ‘new-project-pipeline-demand-pulled’.

Similarly, (the other notable thin-film maker), Solar Frontier, quickly retreated to domestic safety under the umbrella of Showa, after its bold move to build the 900MW plant in Miyazaki, and once it became clear that the industry cost curve had shifted massively from fab ground-breaking to initial fab operation. (And of course, presented with a domestic market opportunity that few had predicted at the time of fab ground-breaking.)

Basically, every thin-film manufacturer put the brakes on new fab builds during 2012 and 2013. That is, except for Hanergy, but more on this shortly. (And with the exception of some small activity such as Solar Frontier’s planned 2015 line addition in Japan.)

For many manufacturing companies in the PV industry that expanded to include the prospects of more fruitful downstream revenues, decisions on whether to ‘buy-and-flip’ or ‘buy-and-hold’ on a project portfolio have now replaced the historic debates on where and when to add new manufacturing capacity for their upstream operations.

Contrasting thin-film strategies today

On the surface, First Solar’s recent outage on long-term technology aspirations marks the end of what has been a dramatic re-org from a few years ago; from being aggressive on adding new capacity to largely keeping factory footprint static (or reduced), and doing everything else possible to move the bar forward.

Instead of adding 2GW of new capacity (similar to existing CdTe lines) to get to 4GW, the alternative is to take the existing 2GW and transform that into 4GW within five years. Within the scope of the PV industry – and when comparing to the upstream strategies of First Solar’s thin-film and c-Si peers – that’s one of the key takeaways from First Solar’s investor day output.

Of course, one can argue whether the efficiency roadmap is too ambitious or sandbagged, whether splitting the front and back end of thin-film lines is going to be cost effective (Uni-Solar and Nanosolar were proponents of this concept also), or how much idled tooling First Solar can rework to save even more on capex.

But as a concept, strategy and roadmap, it is way out there – miles ahead of what anyone else is advocating. And as an added benefit, the route to get there is shown with the milestones to be achieved each year.

There is almost no precedent for this level of guidance across other tier one suppliers in the PV industry today, c-Si and thin-film alike. It provides a benchmark for First Solar’s executive team to assess progress against industry averages; and it provides the industry with a benchmark to assess First Solar’s executive team against the company’s new roadmap.

Now, let’s turn to Hanergy’s thin-film strategy update, as outlined in its Annual Report released last week.

Over the past couple of years, Hanergy has also re-organised its thin-film play in the PV industry. Previously, multi gigawatt fab builds in China for a-Si based production (including the different absorber flavours) were the sole focus. Apollo Solar was the acquired vehicle for this bold move. In the four years to the end of 2013, investments into Apollo Solar for turn-key a-Si based thin-film deliveries reached the US$1.5 billion level.

In its recent Annual Report, Hanergy commented that: “The Group is now expeditiously fine-tuning the manufacturing lines.” (To its thin-film competitors, the idea of fine-tuning 1-2GW of a-Si based capacity across five to six factories may appear as a somewhat alien concept.)

Hanergy’s most recent shift to CIGS has been well documented, including its plans to add 5.25GW of CIGS capacity across its acquired portfolio of overseas CIGS operations in the coming years. Whether CIGS is a parallel or tangential (or cannibalistic) thin-film development is not clear today, largely because the investments into a-Si based production still appear to be work-in-progress and there is not enough field data on offer. What’s more, Hanergy is planning to build even more a-Si based capacity.

However, it is perhaps more insightful to examine Hanergy’s steadfast loyalty to all things thin-film. What is driving Hanergy to be so confident in thin-film when its investments until now have yet to come to fruition and when so many thin-film manufacturers have shut shop in the interim?

Part of the answer can be found in Hanergy’s Annual Report, and the fact that Hanergy is still holding firm to what appears to be the old European thin-film roadmap. Namely, that thin-film (as a generic technology) will grab large market-share gains over c-Si out to 2020.

The European PV cavalcade was awash with these roadmaps five to 10 years back. Thin-film would eat away at c-Si market-share, as the self-appointed ‘Gen 2’ industry technology, with a ‘Gen 3’ offering gaining market share from 2015 onwards.

This ‘Gen 3’ grouping was to be driven by the Konarkas and Dyesols of this world, with dye-sensitised and organic-based solar cells being the next big thing. Within these old roadmaps, Gen 2 and Gen 3 technologies would ultimately consign c-Si to the history books by 2030. Back in the day, some people considered this rather plausible.

The figure attached to this blog summarises some of these themes. The top graph is based upon the graphic shown in Hanergy’s Annual Report released last week (page 37). The bottom graph is the equivalent graphic from NPD Solarbuzz’s forthcoming PV Equipment Quarterly report in April 2014.

The NPD Solarbuzz data is derived bottom-up across (the now reduced count of) 160-170 cell/thin-film manufacturers of 2013/2014 that are still in production, with forecasts for each out to 2018 based upon NPD Solarbuzz’s most-likely end-market requirements; and examining in intricate detail what is happening within all the production lines globally for these 160-170 firms.

Comparing the two graphics in the figure, the differences in thin-film market-share out to 2018 are striking. Hanergy’s Annual Report graphic is predicting that the share of c-Si in the PV industry will fall below 60% in 2018. Furthermore, Hanergy’s Annual Report graphic is predicting that ‘other’ (Gen 3) technologies will grab some 8% of the PV market in 2018.

In Hanergy’s own words: “The chart … shows the projected fast growing market share of global thin-film solar technology from 2012 to 2020, pointing to a bright future for thin-film solar modules.”

If the PV industry produces 100GW in 2018 (using some round numbers here just for simplicity), then this certainly begs the following questions: is it likely there will be 32GW of thin-film panels produced in 2018? And, are the DSSC/OPV companies going to ship 8GW of panels to the industry in 2018? Each of these concepts seems very hard to visualise right now, in March 2014.

Technology market-share can only be done bottom-up

Hanergy is not alone, of course, in forecasting PV technology top-down, based on market-share percentage estimates. While Hanergy appears to be doing this for thin-film, there are plenty of analogies in the PV industry where c-Si variants are forecast in this way.

Forecasting ‘n-type versus p-type’ and ‘mono versus multi’ are two c-Si examples that often follow this route. Similarly, the same mantra holds as we get into the details of c-Si variants: how much copper plating will be adopted, how many c-Si cells will have interdigitated rear structures, etc.

Today, the PV industry still has no roadmap that is being followed. So doing top-down market-share forecasting is of limited value to anyone in the near term. Technology market share is being shaped by what the market leaders decide to do in their frame of reference, be this First Solar or Hanergy or the Yingli’s, Trina’s and Canadian Solar’s of the industry; there are still many ways to skin a cat.

Therefore, it is perhaps more meaningful to get back to the basics. Having a pipeline to produce against (whether in-house or third-party), hitting customer-required metrics on cost, price and efficiency and having a business model that makes sense in the long run.

Thereafter, what capacity is in place and needed, and what technology is used, simply comes out in the wash. And the role of thin-film within this overall mix will be a consequence of what is produced relative to c-Si technologies: so, bottom-up, not top-down.

Whether this means that thin-film has a market-share of 5% or 15% by 2018 then becomes no more than an Excel spreadsheet calculation, and should probably not be used even then as a metric upon which to judge thin-film against c-Si.

The level to which Hanergy contributes to this equation by 2018 is a much harder question to answer than the level to which First Solar ends up meeting (or adapting) its five-year production targets. Either way, the role of thin-film as the competing technology to c-Si is certain to continue for some time yet; something that can only push the c-Si leaders to increase investments into R&D.

And in this respect, the aggressive forecast from First Solar recently may well have the greatest impact on the Chinese c-Si leaders and could even become a catalyst in finally seeing c-Si technology (and R&D) in China get prioritised as a key ingredient needed to maintain or grow market share in the PV industry over the next five years. It has been a long time since c-Si players paid attention to the thin-film world, but this could all be about to change soon.