Official German PV installation figures for November have recently been released, giving a clear picture of expected record installations in the country for 2009. However, there is a major mismatch between expectations for December and total installed system numbers from Photon, which has claimed 4GW and beyond. In an exclusive interview with Andreas Hänel, founder and CEO of major PV systems integrator Phoenix Solar, we discuss his new forecast for 2009 and 2010, which add to the conclusion that Photon needs to drastically change its forecast downwards to retain any credibility in the future.
Official German PV installation figures for November
The recently released official PV installation figures for November highlight that the German market did not overheat last year and indeed is nowhere near the boiling-over point. Although record installations were logged, with 2.34GW installed through the first 11 months of the year, any forecast that takes installs higher than 3GW looks highly unlikely.
The official Bundesnetzagentur (German Federal Network Agency) figures reveal that there was a sequential monthly increase in installations after the prolonged cold winter had thawed. Only August showed a slight dip in installs, a decrease caused by the annual holiday season slowdown.
The August dip is also reassuring because it points to a normal event having an impact on installations. Had the figures shown a continued monthly increase, then we may have seen a market starting to overheat in the third quarter, and most definitely in the fourth quarter.
Nevertheless, it looks likely that the November installation figure of 496.8MW will prove to be the peak month in 2009. There are several factors that support this view.
First, numerous end-users and PV inverter and other component suppliers claimed that supply was maxed out in November, making it very difficult for increased supply to be shipped and installed before the end of the year.
Had installers had not secured the required inventory in November or even earlier to meet demand in December, then an effective cap caused by supply shortages would have taken place. There is some evidence that rather than shipping components over ground, some reverted to air freight – even for modules, to meet demand in the fourth quarter. However, the scale of such practices does not seem to have been widespread.
Second, the natural elements must be considered. It was not a mild end to the year in Europe, with heavy snow and sub-zero temperatures pervading across the continent. As with the slowdown in August installations, the bad weather and Christmas holidays would be expected to have curtailed installs.
Although hard to measure, these two conditions would have simply hampered available days for installations, especially residential rooftop, which still makes up the largest portion of the German market.
But despite these obstacles, it should be noted that there was a good level of activity in the large-scale utility market at the end of the year, and some major projects were being fast-tracked to beat the expected changes in the feed-in tariff. These projects would have had components tied down and locked in to their construction deadlines, which meant that shortages in the supply chain did not have a major impact.
Phoenix Solar German Market Forecast 2009-2010
Back in mid-November (when official figures through September were released), Phoenix Solar had projected a realistic installed figure of 2.5GW in total for the year. This was also close to my extrapolated figure of 2.452GW, which I had projected in news and blog items earlier.
With an eye on Phoenix Solar’s presence and expertise in the marketplace, I recently got in touch with Andreas Hänel, Phoenix Solar’s founder and CEO, to get his latest perspective on installations following the release of the November figures.
Not surprisingly, he had made an upward revision to his previous 2.5GW projection, based not only on the November official figures but on his take on the installations, which is usually pretty flawless, especially in regard to those of a large-scale and commercial variety.
He confirmed that installation constraints, caused by both component supply and bad weather, had characterized market conditions at the end of the year. PV installations in his view had reached full capacity in the fourth quarter. Snow levels of 10cm and sub-zero temperatures also had an effect on system deployments before and after Christmas.
Hänel also pointed to the record installation levels for the year, and calculated that the December installation figure should reach between 350 and 400MW (see chart below). Taking these figures into account brings the total 2009 figure to about 2.74GW.
However, he noted that some major power plant projects that were close to completion before year-end could be included in the yet-to-be-released official figures, which could take the installation number closer to 3GW. These types of systems take many months to finish, so as a result were less impacted by deteriorating weather conditions. With the push to complete, he preferred to therefore give a guidance range of between 2.7 and 3.0GW for the year.
With the expected FiT changes being delayed until June 1st, Hänel expects the German market to continue its ‘hot’ status until then. He believes that installations could end up at between 3 and 3.5GW in 2010.
He noted that in order to enable an approximately 6% return on investment (ROI) for PV installations, system price declines should continue to keep demand in motion, an element that proved successful in 2009. He sees no reason for the market to come to a halt after the FiT changes, and that the second half of the year should be typically strong. Hänel is therefore remaining bullish in regard to the German market prospects for 2010, but doesn’t predict overheating, especially considering his latest forecast.
Photons Figures Don’t Make Sense
Which brings me nicely to the first problem I have with other forecasts issued for German installations in 2009. We reported back in November 2009 that Photon Consulting had projected installs to reach 3.9GW in 2009, a report that was issued within the company’s ‘Solar Annual 2009: Total Eclipse’ market publication.
Based on that figure, projected PV installations in Germany are forecasted to reach 10.3GW in 2011 and 19.2GW in 2012, resulting in market saturation and an effective ‘total eclipse’ with the market screeching to a halt.
Photon magazine then published their survey in early December 2009, based on a sample of 119 grid operators, as well as data from inverter manufacturers that put the installs at between 3GW and 4GW for the year.
Granted, the 3GW guidance now seems realistic, especially since the survey took place in the fourth quarter, but the very high 4GW potential still seems as implausible then as it does now. However, with almost a full year of installation figures already logged, the range should have been much closer. Putting to one side the inherent issues attached to any type of survey that is conducted, I simply cannot reconcile such figures.
A good example of industry sector surveys comes from SEMI. Their survey on the polysilicon market for the semiconductor industry often reflects both the sentiment and therefore the expectations of polysilicon producers for that year. In a down year, such as 2009, the forecast was quite pessimistic and so revenue expectations were very low. Although it wasn’t as bad as producers had feared, the difference in the survey and actual figures follow each other very well.
As noted in the coverage of the Photon survey, Photon International editor-in-chief Anne Kreutzmann was using the survey figures to back up an ideology that FiT rates should be cut more aggressively than the normal regression built into the system.
The problem I have with Photon’s figures worsened when Kreutzmann noted in the German language version of the magazine that PV installations in Germany could actually reach 5GW in 2009!
Photon Missing the Point
There has been a heated debate and demonstrations over the new German government’s plans to make further cuts to the FiT for mid-2010. Too much PV at too high an expense was the main thrust of the argument presented by the government.
Photon’s editor argued that the projected high installs was proof that PV was good for the environment, and ventured to say that even more installs should be made now that production costs have been significantly reduced, thus allowing a lower tariff. Keeping solar artificially high would actually restrict further adoption.
Space constraints prevent me from debating the issue of lower costs due to high-volume production. But it seems strange that this point was raised and not the one with potentially the most influential impact on crystalline production costs: polysilicon prices. Ironically, this is something Photon and others had been warning about since 2008 when they projected price declines from US$450 per kg down to below US$50 per kg for sometime in 2010.
On that front, Photon would seem to have made a good projection. Our recent report noted that JA Solar is paying slightly below US$50 per kg in the fourth quarter of 2009.
Furthermore, as reported here, a study by Fraunhofer ISE argued that only a small change in the feed-in tariff represents the ‘real’ reductions that have taken place in regard to production costs. The market was oversupplied and price drops a result of this factor, not the inferred cost reductions.
The Photons’ latest view of the German market is fundamentally flawed. How can a forecast and a survey calculate installations of approximately 4GW, when there was enough real data to show that the 4GW figures were significantly overoptimistic?
The Photons’ Credibility Concerns
The Photons needs to seriously look at their forecasting models and survey methodologies going forward. Claims of 4GW being installed in Germany in 2009, still out there even after the November official figures have been released, make the companies’ data look increasingly suspect and verging on the ridiculous. To meet this projection figures, well over 1GW would need to have been installed in December.
Credibility must be a core aspect of any market research firm; once this is lost, only fools will continue to follow. As many people know, at PV-Tech/Photovoltaics International, we do not conduct our own market research. As a media publishing house, we rely on the leading research organizations to provide their outlooks and interpretations and we happily report and publish these data. We also report, as news, what the key research firms are saying and do not have any problem in highlighting suspect projections.
I am increasingly concerned about the reports issued by the Photon organization. We do not cover every research firm or market report as there are clearly ‘credibility’ issues with some of them, but this has not previously been a particular cause for concern with the data coming from Photon. Granted, their forecasts tend to be on the ‘high’ side of forecasts, but this alone isn’t a problem.
It should be noted that it is rare for any analyst firm to be consistently accurate with its forecasting. Everyone can have a bad day, but normally the errors are noted, revised forecasts issued, and a certain amount of credibility that was lost is reset in motion to be regained.
An example of such a reset came from iSuppli, a well-respected research firm involved in both the microelectronics and PV sectors. Concerns were raised about iSuppli’s very low installation forecasts for 2009 since real install data were showing a different pattern. Soon enough, the company accepted the criticism and adjusted its figures upwards.
Being a hundred megawatts off in a projection at the beginning of a year is no big deal, but being a gigawatt and more off—as Photon’s data is—near the end of the forecast period, is something completely different.
Businesses need independent analysis of markets to help them plan their operations. Those with the financial capability will obviously perform in-house market analysis but will always need to make checks and balances against figures from research firms as part of their due diligence.
When a major market research player and publishing house is so out of step with the real data coming through, the result is a serious disservice to the industry.
There is, however, a small window of opportunity for Photon to revise their forecasts, especially before the final figures for German installs are released. Given Photon’s unique and powerful position in the global PV marketplace, there is an understandable urgency in this situation.
Along with a change in the market projection, surely the calls by Photon magazine for aggressive FiT cuts should also be reversed?
Since 2.7-3GW of installations looks like a far more realistic figure for the German market in 2009, any call for aggressive cuts is out of line with the reality of the market.
If, for some unknown reason, Photon does not soon revise its forecasts and reverse its call for aggressive German FiT cuts, this curious inaction could be interpreted as manipulative and arrogant, characteristics that neither a market research firm nor a trade publishing house should abide if they wish to maintain a reputation for accuracy, excellence, and industry leadership.