After four years of boom times in the solar industry, a significant softening of demand along with lower module prices has led to anxious times -- fewer sales, at lower selling prices and so lower revenues and, significantly, lower profits.
Even without the expected decrease in demand in 2009, technology revenues would be lower than the US $20.4 billion (€14 billion) of 2008 as cell and module prices are around 40% below 2008 levels. Figure 1 (shown below) provides technology revenues from the manufacturer to the first point of sale in the market from 2003 through 2013. For 2009 and 2010, an estimate of revenues for the recession forecast has been provided. With technology prices at the current level, even growth in sales volume, which is highly unlikely, would result in lower revenues in 2009.
Figure 1. Worldwide module revenue volume for recession, conservative and accelerated growth models 2003-2013.
Accelerated growth in the photovoltaic industry continued in 2008, with 79% market growth over the previous year to 5.5 GW. Unfortunately, the market was significantly oversold in 2008, stranding around 2 GW of product in supply side inventory at the beginning of 2009.
Most of the overselling was into Spain, which with a market volume coming in at 2.3 GW in 2008, represented 42% of total photovoltaic system sales worldwide. Along with high prices for modules and PV systems, quite a few instances of poor module product and poorly constructed systems, and permit speculation, the oversold market led the Spanish government to alter its support programme. The new decree capped the market, lowered the feed-in tariff and effectively closed Spain to new product sales for perhaps two years, or more.
Other than Germany, the PV industry currently has no other global market capable of accepting a volume of sales remotely similar to Spain. Moreover, the global recession and financial crises have further hobbled an industry that had been enjoying accelerated growth since 2004. For these reasons, the PV industry is set to experience its first decrease in demand in more than 30 years — and not just flat growth, but a decrease in sales volume of perhaps 30%, or even more.
Figure 2, (below), provides data for 35 years of PV industry growth, from 1974 through 2008, while Figure 3, (below), reveals three forecast scenarios for 2009, which are on based on assumptions related to recession, conservative and accelerated growth for the sector.
Though the PV industry enjoyed accelerated growth from 2004 through 2008, this rate will not continue in 2009, and accelerated growth is unlikely into 2010. In 2009, lending from the international debt markets continues to be depressed. Meanwhile, the loss of a major market — Spain — is having a deleterious effect on growth, inventories remain high, and global economies remain in recession.
Figure 2. Photovoltaic industry history 1974-2008 (CAGR = compound annual growth rate).
Furthermore, although market development is underway, Germany remains the only market capable of consuming more than a gigawatt of product, and other markets, such as Italy, are underperforming. Japan, South Korea, the United States and others continue to experience slow growth. The good news is that module prices in the soft market seen in 2009 continue to decrease significantly.
As previously noted, while the PV industry has experienced slow or flat growth so far this year — to July 2009 — it is an industry that has not experienced negative growth in 35 years or more. Conversely, years of significant strong growth of more than 70% include: 1975 at 150%, 1976 at 141%,1977 at 87%, 1978 at 112%, 1980 at 128%, 1983 at 88%, and 2008 at 79%.
Since 1974, the PV industry has only experienced three years of soft growth, defined here as demand growth of less than 10% in a given year: 1986 at 8%, 1993 at 3% and, 1994 at 10%.
The Incentive Driver
Historically, the PV industry has enjoyed strong growth, though at much lower volumes than today. The strong growth that the PV industry enjoyed since 2004 was driven by incentives, in particular, the feed-in tariff laws in Europe, and even more specifically, Spain’s generous programme. Though for countries in Europe (in general) there is no reason to assume that feed-in tariff programmes will stop altogether, the problems experienced in Spain (overselling, fraud and poor quality products among them) are having a sobering effect on government incentive planning in other EU countries. The support programmes of the future will need to include mechanisms that manage growth along with stimulating it.
The incentives that the industry relies on come with downward price pressure, which is a significant constraint. However, given the goal of grid parity, there is literally nowhere for price to go but down. Grid parity, nonetheless, is a complex subject, differing in most global markets. Moreover, grid parity provides a level competitive playing field for solar (a worthwhile goal on its own), but does not ensure success.
The industry also needs an increasing number of highly trained installers, sales personnel, engineers and such like, and this comes at a cost. Lower costs and prices are necessary for the continuation of incentives and, therefore, demand. For accelerated growth to continue, and for the eventual slowing of demand to happen gradually, unlike the expected steep decline in 2009, the PV industry must learn to manage its demand. It must develop incentives with triggers to control demand when it accelerates too quickly.
The industry must also control its supply chain from expensive raw material, to consumables, and through to the end user, and must participate with balance of systems (BOS) manufacturers to innovate and develop inexpensive and robust BOS. All raw materials, consumables and machinery are more expensive at this point because of the higher price of oil, which is necessary for transportation.
Other caveats to limitless growth are the high price of PV systems, and the availability of less expensive alternatives, including conventional energy sources such as natural gas and coal. In recent years, the current high volume of industry demand, coupled with raw material shortages, threw the industry into a panic. Instead of the technology standard, ‘if we build it they will come,’ the new mantra became, ‘they are coming and we can’t build it.’
The industry reacted by buying silicon feedstock and cell futures, and by raising component (module) and system average prices, globally. These long-term contracts for raw material, wafers and cells are proving unsupportable and in many cases, are being rewritten or ignored.
Figure 3. Recession, conservative and accelerated forecast scenarios for grid-connected PV, 2008-2013.
The Past Can Inform the Future
It is useful to study specific periods in the PV industry’s history, in terms of growth and drivers for growth, to see what can be learned from these periods which can be useful in understanding the direction of this still young industry. Figure 2, offers compound annual growth rates for the PV industry for specific periods, 1974–1984, 1984–1994, 1994–2004 and 2004–2008.
During 1974–1984, strong compound annual growth of 84% was due to utility and government-backed grid-connected demonstration projects. During this period, the grid-connected application was 30%–50% of total demand, though from annual totals less than 20 MW. Following this decade-long period of significant growth, lower compound annual growth of 13% for 1984–1994 was due to an almost complete cessation of these projects. During this period, grid-connected applications (primarily unsubsidized or incentivized) was less than 10% of annual demand.
Stronger compound annual growth of 33% during 1994–2004 reflects the beginning and continuation of the strong incentive programmes that continue to drive PV industry growth. Specifically in Europe, the feed-in tariff model has proven to be the most successful incentive model. Japan’s residential rooftop programme in the late 1990s, a capacity subsidy, built a sustainable market for solar roofs in that country. In the US, incentives in California created the most significant market in that country.
The 2004–2008 period also managed to encompass two significant events for the sector: the PV industry’s greatest raw material (silicon feedstock) shortage and its strongest period of sustained accelerated growth.
During this period, demand for large field grid-connected applications in Europe, largely driven by the feed-in tariff model of incentives, created the largest global market (79% in 2008) for solar systems. However, the solar-grade silicon raw material shortage that had pushed up prices for crystalline silicon modules also created an entry point for thin-film technologies, which had previously been viewed as risky. The industry’s compound annual growth for this period was 51%.
Grid-connected Growth Drivers
Like it or not ... strong growth in the PV industry comes with strong growth in grid-connected applications. Off-grid (remote) applications show slow, steady growth over time, but have not driven the industry into gigawatt sales. It is the grid-connected applications (residential, small, medium and large commercial, large field commercial and utility) that dominate the market for photovoltaic modules. Indeed, at 94% of total sales in 2008, the volume of grid-connected installation leaves very little module product available for off-grid applications.
The grid-connected application remains driven by government subsidy/support programmes (Europe’s feed-in tariffs, US rebates, for example). Without such programmes the market for grid-connected PV products would decrease dramatically. The significant decrease in demand in 2009 is a lesson to the industry about the significant changes that could take place in demand, revenues and profitability when markets are abused, and when so-called ‘black swan’ events, such as the global recession, alter the playing field and force reactive market and price setting.
Figure 3 (shown above) offers an aggregate five-year forecast for grid-connected applications. The recession forecast is presented in Figure 3, but is considered a two-year anomaly. Meanwhile, Figure 2 excludes off-grid applications. However, at more than 90% of the total market demand, the volume of grid-connected applications effectively represents the total industry volume.
All is not doom and gloom, however, with encouraging current market developments in the US and some other countries. There is continued progress in lowering manufacturing costs so that a reasonable margin can be maintained along with lower system prices. We see progress in increasing efficiencies for all technologies, and business model innovations, meaning that accelerated growth will resume for the PV industry. Certainly, at this stage in PV industry development (which could be likened to its preadolescence) there is room to grow and much to learn before a stable, sustainable level of annual growth settles in. Until then, exciting, and sometimes painful times remain ahead.
Sidebar: A Brief History of the PV Industry
1980s: The market for PV products changed from ‘demand-limited’ to ‘supply-limited,’ largely due to a restriction in silicon wafer supply.
1992: The market situation changed to ‘demand-limited’ with the addition of significant increments of manufacturing capacity, a recession in the semiconductor device market that brought a surplus of silicon wafers to the photovoltaic industry, and the impact of a worldwide recession.
1995: The photovoltaic market again experienced a restriction in the supply of wafer and silicon starting material. This did not bring about a complete switch to a ‘supply-limited’ market but did restrict the output of some manufacturers and served to maintain a general flat pricing condition in the general trend of photovoltaic price reduction.
1999: The photovoltaic industry encountered a period of high growth and supply-limited market conditions that was sustained for three years. Grid-connected markets, the majority of which were funded by federal programmes in Japan and Germany, primarily drove the demand portion of this imbalance.
2001–2002: Having experienced more than two years of shipment delays, a number of distributors and installers over-ordered at the end of the year. This resulted in a filled distribution pipeline, just as market demand paused in early 2002. In addition, the cell manufacturers that had been racing to meet market demand had succeeded in bringing on a record 58% in run-rate capacity during 2001. During 2002, market conditions remained oversupplied, with heavy end-userz demand returning during the latter half of the year. Extreme price reductions in 2002 drove down revenues.
2004: the PV industry again experienced a supply-limited situation, driven by strong demand in Germany, Japan, and California, and by an upsurge in demand for grid-connected product in other areas of Europe and the US. Demand for silicon wafer material also picked up from the semiconductor market, further straining the PV industry’s access to the raw material supply. Supply constraints contributed to flat or moderate increases in module pricing.
2005: A year of full-blown crisis in terms of raw material supply for the PV industry. Prices for silicon feedstock (influenced by demand and raw material scarcity) increased from $65/kg (€45/kg) by the end of 2004 to close to $90/kg (€63/kg) on the spot market by the end of 2005. Demand continued unabated, but shortages kept demand participants at a disadvantage in terms of controlling module sales prices.
2006–2007: Reports of spot silicon feedstock prices in the $200–$400/kg (€140–€280/kg) range. Silicon capacity expansions planned, but needing at least 18 months to realize. Much of the silicon feedstock that would become available had already been acquired. In 2007, raw material constrains continued to limit shipments of crystalline product, but provided a boon to thin-film start-ups and sales. Higher transportation costs favoured locating manufacturing close to the market. Though silicon raw material supplies increased, most of the silicon was sold under long-term contracts, limiting the ability of new crystalline manufacturers to enter the market. Extremely strong demand in Spain drove up module prices globally. At more than 70% of demand for solar products, Europe drove the market for PV systems.
2008: Module and system prices remained high, with sales into Spain dominating the market. Anxiety over changes in Spain’s feed-in tariff and the announced cap led to overselling the market. By the end of 2008, inventories were at an extremely high level. During the last quarter of 2008, the new cap in Spain (given inventory levels in that country) effectively ended the market in that country. With the market in Spain essentially closed, a global economic crisis – on the banking side, debt markets shut down, leaving investment for large systems dry – led to a sudden drop in demand. Module prices began to fall.
2009: A veritable crash in demand and cell/module prices for the following reasons: the global economic meltdown and on-going recession, loss of Spain as the primary market with no near-term replacement, and continued dysfunction in the debt markets.