GCL Doubles Down Downstream

According to its 2012 annual report, GCL-Poly Energy Holdings lost $450 million in 2012. Chairman Zhu Gong Shan cited the primary reason for the large losses was crashing polysilicon and wafer prices. He said prices fell 56% and 54% for polysilicon and wafers respectively in 2012 relative to 2011. Meanwhile, GCL’s costs only decreased 6% and 42% for polysilicon and wafers respectively. The company quoted polysilicon costs reaching $19.7/kg and wafer production cost at $0.25/W by the end of the year, compared to average prices of $20.8/kg and $0.25/W. The company also increased production 26% to 37,000 MT, and sold 12,600 MT of polysilicon and 5.6 GW of wafers in 2012. As a result of the losses, GCL has tasked its managers to cut salaries by 30% to 50%.

GCL’s downstream project development business, GCL Solar Energy, sold 140 MW of projects in the U.S. and earned a $16 million profit before tax. GCL signed a cooperation framework agreement with China Merchants New Energy to develop 1 GW of rooftop solar projects in China by the end of 2015. The company also has two 75 MW projects under development in South Africa.

GCL’s performance summarizes the PV industry in 2012. Huge overcapacity, driven by Chinese manufacturers such as GCL, has resulted in losses for most upstream manufacturers. With the overcapacity, manufacturers have the choice to either idle lines and incur underutilization costs, or continue production and sell below cost to prevent inventory buildups; GCL chose the latter selling 5.6 GW of wafers and 12,600 MT of polysilicon in 2012. Meanwhile, the downstream business was profitable with relatively low project volumes. Numerous upstream manufacturers have moved downstream as a result (client registration required), such as MEMC, which is rebranding to its downstream arm, SunEdison. To drive short-term profits, GCL will increase downstream operations and take advantage of the booming Chinese solar market as the government targets 40 GW of installations before the end of 2015. Interestingly, the company targets distributed generation, which the State Grid is working to promote (client registration required). Upstream, the company aims to manufacture quasi-mono crystalline silicon (qc-Si) wafers – which are 1% absolute more efficient than the company’s multicrystalline silicon wafer – in an effort to gain back some margins with the higher efficiency product. However, little can be done for polycrystalline silicon with the exception of convincing the government to penalize imported polysilicon (client registration required).

GCL has announced a non-binding agreement to cooperate with Yingli on research and development, and along the supply chain. This could be an early indicator of what may take place as the industry consolidates. GCL-Yingli would be the most vertically integrated player in the industry ranging from polysilicon through project development and would have the greatest potential to profit throughout the value chain as the industry returns to equilibrium after 2015 (client registration required). This is merely speculation; mergers between operations-heavy companies are often messy and fraught with peril, but this early relationship could ease the transition if the two companies decide to take the risk over the next three years.

Hangzhou First PV Material puts incumbent encapsulant suppliers under pressure

Hangzhou First PV Material produces ethylene vinyl acetate (EVA) films and flouropolymer back sheets and is located in Hangzhou, Zhejiang province, China. Hangzhou First PV Material has more than 500 employees and sales revenue of $67 million. According to the Hangzhou First PV Material’s prospectus, before 2008, STR Solar, Mitsui Chemicals, Bridgestone, and Solutia (Etimex) were the dominant companies producing EVA film – taking 60% of the global market share. Hangzhou First PV Material exceeded Bridgestone and Solutia, and became one of the top three suppliers in 2008. In 2010, the company claimed 25% of the EVA market share; its primary customers are Suntech, Yingli Green Energy, Trina, and Jinko Solar, some of the largest module manufacturers in the world. The company experienced a net profit growth rate of 252.34%, 346%, and 10.76% in 2009, 2010, and 2011 respectively, according to the company’s annual report. This decelerating profit growth, reaching $94 million in 2011, is due to slower growth in the broader solar market, according to the company. Hangzhou First PV Material priced their EVA film at an average of $2.41/m2 in 2011, including a 37.26% gross profit margin (GM).This price is between $0.4/m2 and $1/m2 cheaper than EVA made in Europe or the U.S. (see the report “Module Cost Structure Update: Path to Profitability” — client registration required). Although market conditions are less than ideal for the greater solar industry, the tight-lipped Hangzhou First PV Material has been able to swim against the current. The company hopes to issue approximately 58.1 million shares on the Shanghai Stock Exchange to raise $179 million to ramp an EVA film production line with an annual output of 180 million m2, a backsheet production line with annual output of 2 million m2, and a PV material research center.

Historically, module manufacturers have chosen encapsulants based on the lowest cost, rather than performance, as long as modules pass IEC and UL certification tests; as a result, EVA has dominated the encapsulant market. Still, silicones, thermoplastics, and polyolefin encapsulants continue to compete with EVA. Dow Chemical started production of its ENLIGHT polyolefin encapsulant (client registration required) in Thailand in August aiming to replace EVA. Similarly, Wacker Chemie has rolled out a silicone-based thermoplastic film that claims better transparency and faster lamination times at a similar price to incumbent EVA suppliers (i.e., $3/m2 to 3.50/m2). While encapsulant suppliers with alternatives to EVA claim better performance and/or faster lamination times, success will ultimately come down to cost and how easy it is for module manufacturers to transition from EVA to the new encapsulant. Polyolefins have potential to be less expensive than EVA and Wacker’s silicone-based film can increase efficiency – which can reduce the overall cost-per-watt of the module – but capacity needs to ramp up in locations near module manufacturers to compete with players like Hangzhou First PV Material.

Solar Module Prices Drop 36% in 2011

After analyzing quarterly numbers from Yingli, Trina, Canadian Solar, and Suntech, we have determined that module prices declined 36% over the course of 2011. The average selling price (ASP) in Q1 of 2011 was $1.76/W. By Q4 2011, the ASP was $1.11/W for these tier 1 manufacturers. The ASP over the entire year of 2011 was $1.44/W.

The precipitous fall in module prices was caused primarily by declining polysilicon costs. In March 2011, the polysilicon spot price was around $80/kg, which fell below $27/kg in December. The polysilicon spot price rebounded to $29/kg in February 2012, but fell back below $27/kg in late March 2012. At the same time, the intense competition caused by a global oversupply of PV modules has eaten away margins for manufacturers, further reducing prices.

The margins are so low that, even if polysilicon prices continue to drop, we expect module prices will stay between $0.90/W and $1.00/W over the next few years as manufacturers restore margins. Make sure to note that these prices are for tier-one, crystalline silicon manufacturers, not smaller, less reputable manufacturers nor thin-film manufacturers who will sell at lower prices.

Falling panel prices and Chinese competition creating solar woes

The precipitous fall of module prices has led to a few casualties in solar. Evergreen Solar filed for Chapter 11 bankruptcy, and is currently trying to sell its assets and core “String Ribbon Technology.” With falling polysilicon prices (currently between $55/kg and $60/kg), the company couldn’t compete with standard crystalline silicon technology which, as we’ve mentioned previously*, has been made substantially cheaper by Chinese manufacturing firms. Chinese manufacturers have enjoyed continuing support from the Chinese government through inexpensive capital, low prices for electricity, and low labor costs. This is clear evidence that government subsidies and support are extremely critical to the growth of the solar industry.

Meanwhile, Solon announced that it has decided to shut down its Tucson facility given its inability to be cost competitive with the low-cost Chinese module manufacturers and instead focus on its project development and power plant business.

In addition, high-cost American sweetheart Solyndra was forced to shut its doors and file for bankruptcy. The start-up attracted high scrutiny for its inability to compete due to significant price drops in polysilicon, even after receiving a loan guarantee from the U.S. government. And Ascent Solar*, a thin-film CIGS manufacturer that was likely heading the Evergreen route was rescued by TFG Radiant Group of China, by signing a royalty and strategic partnership agreement.

Non-cost-competitive technologies and companies with poor strategy and balance sheets will likely go out of business faster given the shift in demand dynamics worldwide for PV that have significantly impacted module prices. This news bodes well for all the low-cost Chinese manufacturers such as Yingli, Trina Solar, and Suntech, all of which are better able to withstand the low-price environment. This news should make smaller thin-film solar companies wary of the competition in the industry.

* Client registration required.

Italy’s solar market reaches 3 GW in 2010 amid growing uncertainty

Following months of questions about the exact size of the Italian solar market, the Gestore Servizi Energetici (GSE), Italy’s grid interconnect agency released official installation numbers through March 2011: In total, the country has a little over 4.18 GW of cumulative solar installations. Yet, despite the official announcement, uncertainties remain.

Not surprisingly, planned reductions to Italy’s feed-in-tariff (FIT) led installations to soar to nearly 1 GW in December, bringing the total for 2010 to 2.33 GW. In February, announcements from the GSE and a flurry of news from installers and other research firms suggested that number reached 6 GW based on the number of applications received. But, given that 713 MW had been connected to the grid between January 1 and March 30 by the GSE’s count, much uncertainty remains about whether the missing 3 GW of applications were ever actually installed.

As of this writing, all eyes are on The Fourth Conto Energia and what reductions or caps it will introduce to the FIT policy at the end of April. An FIT reduction is certain, but the exact percentage remains to be seen. Reuters reported in mid-April morning that the current draft decree pegged it at an immediate 25% cut, with another 8% planned for January 2012.

A bigger concern is that the GSE will introduce a cap on new installations, which would dramatically hurt the solar market. Reuters further reported it could actually fall below 2010 numbers, at 1.55 GE to 1.8 GW for 2011 and 2.8 GW for 2012. Further, the article suggests the cap will be based on volume, and not total subsidy burden. That means price decreases will not enable higher installation figures.

The greatest concern to the growth of the Italian solar market, however, has been wild announcements of installations that are supposedly in the ground but not yet grid connected. If these 3 GW of phantom systems actually exist, the Italian market may have already exceeded the 2011 cap – effectively closing the market in 2011.

Due to these remaining uncertainties, we are hearing that module players and project developers are already rerouting inventories and supply to the U.S. as the short window between now and the June decree makes it unlikely they will be able to install systems before any possible cap. From our previous conversations, the module vendors with the most exposure to the Italian market include First Solar, Uni-Solar, SunPower, and the major Chinese players like Trina, Yingli and Suntech Power. In general, Italy’s scenario signifies the globe’s most severe case of reductions by a bankrupt government, but we expect to see more of the same elsewhere in the near future.

Lux Q2 2010 Solar Supply Tracker: Growing demand stretches upstream supply

Last week, Lux Research released its Q2 2010 Solar Supply Tracker (client registration required), which updates our figures on production and capacity throughout the value chain through 2013. Regular readers of the Tracker will find the most important update since Q1 is in the recent version’s demand forecast, which has been increased to 15 GW of market demand in 2010, driven by a surge in Germany and Italy ahead of 2011 feed-in-tariff reductions. We discussed this following PVSEC in Valencia a few weeks ago (see the September 16, 2010 LRSJ – client registration required).

On the supply side, Tier 1 and 2 module manufacturers are poised to produce 14.8 GW in 2010. Figuring in Tier 3 players for the “high case” scenario shows a potential for 18.0 GW of production in 2010. However, even with this greater-than-expected demand, Tier 1 and 2 module manufacturers maintain oversupply with 20 GW of capacity.

In Q2, crystalline silicon (x-Si) increased its technology dominance, reaching 79% of module production. Thin-film silicon (TF-Si), led by large manufacturers like Sharp and Bosch, and cadmium telluride (CdTe), dominated by First Solar, each account for about 9% of total production. Together, these three technologies make up nearly the whole module market. Copper indium gallium diselenide (CIGS) modules are still being produced in small quantities, but will quickly grow more than 1 GW if Solibro, Miasolé, and Solar Frontier execute plans for large capacity expansions over the next several quarters.

Geographically, Asia increased its dominance in module manufacturing, accounting for 75% of production in Q2 as more European capacity became unviable and low-cost Asian manufacturers like Trina, Yingli, and LDK added capacity in the downturn. On the polysilicon front, Wacker added European capacity, increasing the EU’s production share to 20% from 17% the previous quarter. However, Asia still holds the largest share at 43%.

A number of key market shifts occurred in Q2 throughout the value chain. Chinese polysilicon producers were a mixed bag as several low-cost manufacturers, including Jiangsu Shunda and Sichuan Xinguang, shut down for retooling or recapitalization while new entrants Yongxiang and Fine Silicon (owned by Yingli Green Energy) were still ramping production. Meanwhile, demand in 2010 will reach 100,000 MT and exceed the 75,000 MT production of the six top polysilicon producers – Hemlock, Wacker, GCL, OCI, REC, and MEMC – providing room for new entrants like Daqo New Energy. Further downstream, wafer supply is increasingly constrained in the face of strong demand and will hit 78% capacity utilization in Q3. Low-cost Asian manufacturers continue to dominate this segment, with GCL Silicon, Kyocera, and Green Energy Technology moving into the top 11. Strong demand has spurred several companies to make aggressive expansions in cell and module production. JA Solar led the charge in cell production, rising into the top 3 in Q2 and positioning itself to overtake the entire field by capacity in Q3. It has also integrated downstream into module production, along with China Sunergy, through acquisitions. Although these companies are emerging as new leaders, the top five module manufacturers remained unchanged from Q1, with the exception of Suntech Power, which overtook First Solar at the top spot in Q2 with 1.4 GW of capacity.

Lux’s Q1 2010 Solar Supply Tracker sees growing production amidst slight oversupply

Last month, we released the Q1 2010 version of the Lux Research Solar Supply Tracker (see Solar Supply Tracker, Q1 2010 – client registration required). It includes figures on production and capacity data throughout the value chain through 2013.

Notably, the Tracker revealed that total module production for 2010 will be 12.6 GW, an increase of about 4.7 GW from 2009 production. We’ve also updated the Lux Research demand forecast, which predicts 12.1 GW of market demand in 2010, signifying a slight oversupply this year.

Crystalline silicon (x-Si) will account for 76% of total new module production in 2010. Most of the remaining share will be split between inorganic thin-film PV – particularly thin-film silicon (TF-Si) fueled by a slew of entrants – and cadmium telluride (CdTe), overwhelmingly provided by First Solar. Each will each account for 11% of 2010 module production. Companies like Avancis, Würth Solar, and Solibro will each produce a handful of Copper indium gallium diselenide (CIGS) modules in 2010, to round out the balance of new module production.

In terms of geography, Asia continues to dominate the manufacturing scene, accounting for 45% of polysilicon production, 78% of wafer production, and 71% of module production in 2010. Though Asia dominates in absolute production, several companies are adding capacity in North America, hoping to capitalize on promising demand in the U.S. and Canada, including Canadian Solar, SunPower, and Yingli.

A number of companies made notable changes to production and plans in Q1. Upgraded metallurgical silicon (UMG-Si) producers Dow Corning and Timminco stopped production at their Brazilian and Canadian facilities, respectively. Both companies cited decreased market demand, and will leave capacity idle with plans to reevaluate demand in a few years.

While UMG-Si players are hurting, top-tier polysilicon suppliers are thriving. The top six polysilicon producers – Hemlock, Wacker, GCL, OCI, REC, and MEMC – will supply 75% of the total polysilicon to the market in 2010. Further downstream, several companies beat expectations and are accelerating ramp schedules. Taiwanese wafer player Green Energy Technology, cellmaker Neo Solar Power, and Chinese module manufacturer Solarfun all increased or accelerated capacity addition plans, citing increasing customer demand. Although Solarfun garnered more market share with its increasing capacity, it could not crack the top five module manufacturers. First Solar remained in the top spot, followed by Suntech Power, Sharp, Canadian Solar, and Trina Solar.

Looking out several years, supply remains slightly above demand throughout the value chain – except at polysilicon, where a significant supply overhang remains. As we witnessed this quarter, this supply overhang forced more expensive producers to shut down production lines, as their processes are no longer economically viable. Expect more consolidation and additional polysilicon players shutting down production facilities, as well as significant shuffling of market share as new technologies gain  traction, the vertical integration trend continues, and delayed subsidy cuts in Europe keep demand high.

Rating x-Si module makers on the Lux Innovation Grid

lig-c-siCrystalline silicon (x-Si) PV modules comprise the largest and most established portion of the photovoltaic (PV) module market, holding roughly 81% of the global PV market in 2008. These x-Si modules also have significant penetration in all sizes of grid-tied applications – from residential to large-scale utility installations.

A handful of large, top-tier manufacturers dominate the market, but smaller start-ups with differentiated technologies are still entering. As the module oversupply rolls through 2009 and 2010, some crystalline silicon module manufacturers will be at the heart of the shakeout.

Examining the performance of companies in this technology area, we find that:

  • Large corporations with differentiated technologies are among the strongest performers.Many of the highest ranking companies are large corporations that stand out due to top-level high-efficiency products and large corporate backing. Their backing provides support for module warranties, capacity expansions, pricing battles, and technology development.
  • New competition from low-cost manufacturers is driving down the value of European leaders. European module manufacturers with high-quality x-Si module technologies are beginning to struggle as module production becomes increasingly commoditized. Their quality advantage is beginning to slide as new low-cost manufacturers gain access to higher-quality materials, dropping their scores on technical value scale.
  • Even with promising technologies, start-ups face formidable barriers to growth. The most successful pure-play solar firms got an early start in the market, and offer either differentiated technologies, sharp business execution, or both. New entrants to the solar market need more than a novel design or slight technical advantage to succeed. Companies building capacity, especially those based on a novel technology, score lower than those with existing capacity because they must play catch-up with more traditional and established manufacturers. The outlook is increasingly bleak for start-ups with unique technologies that are yet to build production capacity.