By Matthew Holzmann

The photovoltaic solar energy sector is undergoing rapid change in both manufacturing methodologies as well as business models. 2011 has been a challenging year for manufacturers due to a combination of increased manufacturing volumes and reduced demand. This has created an inventory imbalance of up to 25% of total manufacturing capacity.

In addition to the surplus inventory and excess manufacturing capacity problems, the pressure to improve profit margins, reduce costs, and increase market share has led to greater vertical integration and economies of scale. This is especially notable in China, which has become the world’s largest market for photovoltaic manufacturing equipment. There is a repetition of the trend to the gigantism seen in other industries that have come to be dominated by that country’s manufacturing sector.

This dynamic is changing the entire cost and profit structure of the industry on a fundamental level. With extremely high capitalization requirements and struggles of thin-film manufacturers to produce product profitably, the Very Large Scale (VLS) monocrystalline/polycrystalline silicon-based manufacturers are challenging the viability of the advanced technology sector, notably thin film. 2011 has seen the market share of crystalline silicon (c-Si)-based modules rise above 80%.

A major gap has been identified in the long-term reliability methodologies and data. Current test methodologies employed utilize up to 1,000 hours testing, which is the equivalent of only a few years of operational use. However, end-user expectations and economic models are typically based upon 20 to 25-year operational lifetimes.

The existing database for solar module and system performance is highly limited and the results are mixed. Thus, the establishment of long-term quality assurance procedures, test methodologies, and databases is critical to bankability, insurance, repair & replacement, and economic modeling. While efficiency warranties typically cover 25 years, workmanship warranties are normally between 5 and 10 years. With the data indicating significant module and systemic failures within 10 years, manufacturing quality has become a major issue. Industry organizations are working cooperatively to resolve long-term reliability testing and modeling issues.

Economic Overview

Industry Expansion and Contraction

Figure 1 shows the global solar cell production for c-Si and Thin-Film (TF) solar technologies over the past ten years. Production has nearly doubled every year since 2008, as indicated by the exponential growth shown in Figure 1.

Projected annual production volumes, and the share division of the solar market between competing c-Si and TF technologies are shown in Figure 2. It should be noted that these projections were published prior to the recent fluctuations in demand, and do not consider the slowdown the industry is experiencing in 2011.

2010 Economic Indicators

The top ten providers of solar energy modules supplied nearly 40% of the world’s demand. Listed by market share in Table 1, the largest suppliers each provided over 5% of 2010’s 17 Billion-Watt global market.

The vast majority of the solar module manufacturing equipment was purchased in Asia, over 70% (Figure 3). European manufacturers only purchased 9% of the manufacturing equipment sold, and North America only 4%. The growth of the manufacturing industry is squarely in China and Taiwan, largely by Chinese-owned companies (Figure 4).

2011 Economic Indicators

The optimism of 2010 did not carry through in 2011, when demand began steadily declining. Book-to-bill ratios sank to their lowest levels since the solar boom took root in 2008. Table 2 shows the rapid swing market conditions took early in 2011, which precipitated some of the year’s business model modifications.

PV Business Models

The business models in the photovoltaic energy market are changing rapidly. After the initial wave of investment with over 650 module manufacturers worldwide, the inefficient and/or undercapitalized are being swept away by vertically integrated VLS manufacturers.

Europe

The European PV manufacturing industry is being especially hard hit. Global market leaders such as Q-Cells, SolarWorld and others are increasingly under pressure to reduce prices in the face of a supply overhang of 25+% in Q3/2011 and intense competition from China and elsewhere, combined with higher manufacturing costs. Reduced feed-in tariffs in Italy, Germany and other countries that are a result of the ongoing economic crisis have also affected demand negatively. Si-based module prices in Germany in May 2011 ranged from E0.86 - E1.73/W. However, it was noted that in July 2011 that certain manufacturers were offering prices as low as E0.80/W in order to reduce inventories. Despite these low prices, sales and orders have not increased as they would have in the past when purchasers would buy inventory based on their expectations of increased demand.

Because of intense cost pressure, outsourcing has increasingly become the preferred business model for European solutions providers. While there may be a brand name on the box, the likelihood that the product inside was manufactured by a subcontractor grows on a regular basis.

North America

The business model in North America has also undergone substantial change. With the difficulties faced from competition from China, the North American industry has become more reliant upon government incentives than economics. While the global industry is experiencing greater vertical integration with huge capital investments, the North American PV manufacturing industry has rapidly become a boutique business. The average manufacturing plant is 25-50 MW compared with the 200 MW-1 GW factories of the large global leaders.

In several cases, manufacturing plants have been established in North America by Chinese operators specifically to exploit local content rules and subsidies. In some situations, laminated modules are being imported as subassemblies, and only junction box attachment, framing, and final test are done in North America. Other facilities may manufacture complete modules with imported materials. Several Chinese manufacturers have established such plants and purchase almost all the materials in China─at the prices negotiated by the mother plant─but take advantage of the local, U.S. subsidies.

China

This same cost pressure has equally begun to impact the manufacturing sector in China, where many small companies jumped into the business in 2008/2009. In 2010, China was reported to have over 530 companies manufacturing modules. Tracking business failures in China is very difficult, but the number of module manufacturers exhibiting at the SNEC show in Shanghai earlier in 2011 was down substantially from the previous year.

The drive to massive vertical integration has been fostered by favorable government financing and support. Integration of silicon foundries with wafer production, cell manufacturing, and module manufacturing is the most significant trend of 2011. This follows the same trend seen in other manufacturing industries in China, where huge factories are the rule.

At the same time, with the recent economic slowdown and deteriorating prices, many companies are caught in a profit squeeze. Materials and cell suppliers have seen shipments fall as much as 25% in 2011 with the market expected to continue to fall until 2012. As in other parts of the world, end users who in the past would take advantage of down cycles do not seem to be stepping in at present, leading to some heightened concern in the industry.

Technology Trends

Long-Term Reliability and Quality

Solar manufacturing is a very high volume, repetitive industry where cost is critical. The cost of the Si cell represents approximately 70% of the overall module cost, and manufacturers are desperate to simultaneously reduce costs and improve manufacturing efficiency. The continued pressure to maintain profitability has resulted in efforts to utilize new and less expensive materials, in many cases without long-term reliability data. Uniform design practices and standards do not exist, and company standards are often developed internally and considered proprietary.

2011 production of solar modules is expected to exceed 7.8 GW, or approximately 39,000,000 modules (200 W average output). Solar installations are multi-generational. Economic and insurance models for photovoltaic installations have been based upon an expected 25-year operational lifetime. In Europe, feed-in tariffs and most megawatt-level projects require a 25-year lifetime. But neither the databases nor the test methodologies for these assumptions and requirements yet exist.

Data on long-term module reliability has been limited primarily to c-Si products that were installed beginning in the 1970’s. Despite over three decades of elapsed time, reliability data is still limited, and many failure mechanisms are still being identified. Today’s purchasers of modules are asking more and more questions on module quality. Investors and lenders are now reevaluating project risks due to module quality and performance assumptions that are increasingly being deemed inaccurate, as well as their previous experiences with financing wind energy projects.

The lack of reliability data has also resulted in the commoditization of the final product, when in fact there are substantial differences in both materials performance and construction methods that can directly affect long term reliability.

The Advanced Industrial Science & Technology Institute (AIST) of Japan recently reported that in their 7-year test protocol, failure rates of between 0.5% and 6% were recorded, dependent upon module manufacturer. One of the world’s largest solar investment firms reported a 10% failure rate for inverters within the first 7 years of operation of a 50 MW solar farm.³⁾

Solar modules and installations are power generation systems and must be tested as such. They generate DC electricity, which is fed through a junction box to an inverter that generates AC electricity, which is then distributed locally or across a grid. Currently, the most widely accepted test methodologies are primarily the fire safety standards established by UL, TUV and other test laboratories, and are the equivalent of only a few years duration. In addition, most tests are static. Many module manufacturers benchmark their products against those of their competition, but these test results are not released.

Failure Modes and Mechanisms

Field failures for solar power installations include:

-Cell cracking,

-Junction box de-lamination,

-Module de-lamination,

-Diode failure,

-Glass breakage,

-Junction box & gasket failure, and

-Soldering defects leading to cell or module failure.

The Pareto chart in  Figure 5 was developed by SunPower and is based upon data from 21 solar module manufactures. It traces the location within the module of specific component breakdown.⁴⁾

Failure rates from manufacturing and materials defects have been reported at up to 10% of modules, inverters, and other components within the first 10 years of operation. Cleanpath Ventures reported 6 fires in a 50 MW installation in its first seven years of operation. The majority of field failures can be attributed to inadequate quality assurance during the manufacturing process and/or inadequate testing of new materials prior to implementation.

New failure mechanisms are still being identified. The range of operating environments is equal to the range of climate conditions on Earth. Extremes of temperature, humidity, ultraviolet light exposure, chemical exposure, potential-induced degradation, and harsh storm factors such as load bearing under snowpack, hail, or other acts of nature have been only partially evaluated in the establishment of long-term reliability standards.

Situational factors such as shading, reverse bias, and dynamic load also affect long-term module performance. The impact of non-linear and non-uniform module and system degradation on array performance must also be considered when predicting the practical lifetime of solar panels.

Initial Quality Testing

The testing of the module prior to shipment is extensive:

-Individual cells are inspected prior to soldering and immediately afterwards;

-Alignment of strings and tabbing with glass, films, and backsheet is visually inspected;

-Electroluminescence testing is used to inspect for microcracks and other invisible defects in modules;

-Junction boxes and framing are manually/visually inspected, and

-Finally, every module is tested for output using a solar simulator.

While these tests are typically robust enough to insure short-term performance, they do not necessarily predict long-term reliability, which can only be assessed using accelerated test methods that adequately model the end-use environment.

Long-Term Reliability Testing

A number of organizations such as NREL in the U.S., AIST in Japan and EC DG-JRC in Europe, are working together on the development of long term models and test methodologies. A working group has been formed by these organizations in cooperation with manufacturers, suppliers, test laboratories, and end users, with the goal of developing robust standards.

Working group focus topics include:

-Manufacturing quality,

-Thermal & mechanical failure,

-Humidity, temperature and voltage test methodologies,

-Diode, shading, and reverse bias, and

-Ultra violet, temperature, and humidity.

In addition, a current issue of communications among all the cooperating organizations and individuals is being addressed.

There is one set of methodologies that exists that can be of significant assistance. Military (MIL) standards have long had to address every climate and operating condition from undersea to outer space. Several participants in the working group are now examining these and other existing standards to determine their utility and applicability to the current projects.

The photovoltaic industry, while huge, does not have efficient means of communication. There are many stakeholders and subsets, most of who do not regularly communicate with each other. Insurance and investment companies, as well as the operators, are often far removed from the manufacturing floors and engineering departments. Seemingly small decisions that are made in supply chain management can have large long-term effects on product performance.

An industry that is shipping millions of modules per year with a CAGR of 25+% must act decisively to install the tools necessary to meet safety, reliability, and cost performance expectations.

The photovoltaic energy industry is evolving rapidly and change is the only constant in both business models and technology. Robust operating systems, quality standards and long- term performance models are required to maintain profitability and to achieve the potential of PV as a robust, reliable source of power.

Matthew Holzmann is President of Christopher Associates Inc. (www.christopherweb.com), a supplier to the photovoltaic manufacturing industry as well as the president of the Solar Engineering & Manufacturing Association. He has over 30-year experience in high technology engineering, quality assurance, and manufacturing.  Holzmann is the author of numerous technical papers and research articles.

REFERENCES

1) Photon Magazine, March 2011

2) VLSI Research

3) Clearpath Ventures presentation, QA Task Force Forum on PV Modules, San Francisco, July 2011

4) SunPower presentation, QA Task Force Forum on PV Modules, San Francisco, July 2011