Disparity over a potential global shortage of primarily multicystalline solar wafers has emerged after two market research firms have recently issued contrasting views. Image: Meyer Burger
Disparity over a potential global shortage of primarily multicystalline solar wafers has emerged after two market research firms have recently issued contrasting views.
IHS Technology had previously warned in October, 2015 of the potential shortage that could emerge in multi c-Si wafers in 2016, while more recently, EnergyTrend, a division of TrendForce based in Taiwan said it expected the tight global supply should begin to ease later in the first quarter as around 5GW of new production capacity starts to come on stream.
However, IHS has reconfirmed its view that due to current wafer capacity levels and end-market growth demand in 2016, wafer supply would continue to remain ‘tight in 2016’.
The market research firm noted that total wafer (mono and multi) reached 61.9GW in 2015, up from 47.6GW in 2014, and increase of around 14.3 million.
The two market research firms do agree on the fact that multi c-Si wafer ASPs have increased on tight supply issues.
Major wafer producers
IHS noted that it had estimated that all wafer suppliers were operating at 83% utilization rates, while major suppliers such as GCL-Poly and Green Energy Technology (GET) were operating at utilisation rates of 88%, the highest utilisation levels since 2010. However, with further capacity expansions average utilisation rates are expected to remain high at around 85% on average.
GET has regularly noted in monthly financial filings that it had been operating at 95% utilisation rates throughout most of 2015 and expected to fully utilise in-house capacity through the second half of last year. The company produced its highest sales for 2015 in December, after sales has increased for six consecutive months.
As first reported by PV Tech, the largest merchant multi c-Si wafer producer, GCL-Poly said late last year that its polysilicon and wafer operations were at full capacity in the third quarter of 2015, and struggling to keep up with demand.
GCL-Poly reported that wafer production in the first nine months of 2015 had been 10,928MW, 14.0% higher than the first nine months of 2014, while third quarter production topped 3,826MW, up only 3.9% on capacity constraints of adding 1GW of extra production in 2014 to bring total nameplate capacity to 14GW.
IHS noted that the top three independent wafer producers (GCL-Poly, Xi’an Longi Silicon (mono only), and Green Energy Technology) held one third of total wafer market share in the last two years, including internal wafer capacity from vertically integrated players.
The three wafer producers were said to have an even stronger position in the merchant wafer market, which was said to be close to 60% market share in 2015.
Major integrated producers
According to PV Tech’s ongoing PV capacity expansion announcement analysis, major integrated PV manufacturers, such as the Silicon Module Super league members (Trina Solar, Canadian Solar, JinkoSolar, JA Solar and Yingli Green) made no meaningful in-house ingot/wafer expansions in 2014, while several finally announcing expansions in November, 2015.
Key ingot/wafer expansions were announced by Canadian Solar (600MW), JinkoSolar (300MW) and Trina Solar (100MW), respectively.
However, it should be noted that these SMSL members have all adopted imbalanced capacity levels with in-house ingot/wafer production being significantly lower than solar cell and module assembly capacity, relying on key suppliers such as GCL-Poly as in the case with Trina Solar and Canadian Solar for example.
Other integrated PV manufacturers such as REC Solar had already initiated expansions of in-house ingot/wafer production through upgrades, yield improvements and debottlenecking, an approach many have adopted in recent years to limited capital expenditures and boost production productivity, while lowering manufacturing costs.
According to IHS, monocrystalline wafers are expected to gain market share in 2016, due to the increasing share of end-market rooftop installations.
Mono c-Si wafer production is expected to increase to 26% of total wafer production in 2016, up from 24% in 2015, according to IHS.
Like EnergyTrend, IHS noted that mono wafer ASPs have been declining for the last 12 months. As a result, the pricing gap between mono and multi wafers has narrowed significantly. However, EnergyTrend has previously noted that overcapacity existed in mono wafer production, helping to push prices lower towards multi wafer pricing levels.
IHS said that mono wafer ASPs were expected to stop declining faster than multi wafer ASPs in 2016. Overall, IHS forecasts a blended wafer ASP of US$0.20/W in 2016, a decline of nearly 1% over 2015.
Strong multi wafer ASP issues could cause PV manufacturers to raise module prices or risk taken a hit on margins to retain market share, according to IHS.
EnergyTrend expects multi wafer ASPs to peak in February, 2016.
The analyst's view - Finlay Colville, Solar Intelligence:
Market research firms have been struggling to fully understand the real wafer capacity in the solar industry in the past few years. Indeed, this confusion is not just confined to market research firms, but also many of the leading equipment suppliers of casting furnaces, pullers and wire saws.
There are several reasons why the effective or available capacity issue has been hard to understand, due in large part to the glut of capacity of equipment shipped during the period 2010 to 2012. Indeed, the wafer segment was the worst affected by China’s solar manufacturing expansions at the time, with much of the equipment being the same across the industry.
This issue continues to plague wire saw equipment suppliers, with only glimpses of ordering having taken place in the past 18 months. It is likely that no-one actually knows how much ingot and wafer capacity is really out there, and how much can still be consumed quickly to meet any demand upside.
While this is clearly a big deal when forecasting ingot or wafer market tightness, it is only an issue if the demand is really coming back upstream at a faster rate than wafer supply can accommodate.
Let’s look at some reference points though, that are potentially the key ones to use when assessing if there really will be a wafer problem in 2016.
Our latest analysis of wafer production (not nameplate capacity, but real production) for the big-six SMSL members reveals that this group produced 8.6GW of wafers during 2015, and is forecast to produce an additional 2GW of wafers during 2016.
Now, if we factor in just one more company making wafers – GCL – due to in-house debottlenecking done in 2015, and expected continuous improvements in 2016, this one company is forecast to produce an extra 2GW without adding any equipment capacity (which it can do very quickly anyway).
So, without even going any further, and with no other changes to their plans, we have 4GW extra wafer production, and we’ve not had to look at the rest of the industry. But the main thing that many third-party onlookers fail to realise is that ingot and wafer capacity can be added very quickly, even if companies do have to resort to buying from equipment suppliers.
When looking at wafer availability, it is an altogether different issue to, for example, polysilicon capacity or thin-film panel manufacturing. Indeed, with GCL effectively having a much greater degree of in-house tool supply control, multi-GW capacity can be added very quickly at with capex levels that were not thought possible a few years back.
The supply of multi-GW wafers, cell and modules within China, Taiwan and Southeast Asia also has a strong behind-closed-doors unwritten agreement when it comes to who is adding polysilicon, wafers, cells and modules. This is especially true in China, and it is incorrect to assume that the SMSL is going to be working off a wafer supply demand forecast that is not fully in alignment with GCL, for example. While short-term pricing is a cyclic issue – and will continue to be – the reality of annual supply and demand in 2016 may see a far more controlled addition of capacity as and when needed.
The solar industry is still back-end weighted in nature and if wafer supply was not fundamentally a gating factor in market demand during the past couple of quarters, it is hard to see how there would be any fundamental change in the first half of 2016. Come the second half of this year, there is then ample time for a few gigawatts of capacity to be added without any major panic buttons being pressed across the supply chain.