Much debate and head scratching followed the announcement from Italy’s energy manager GSE over PV installation figures that seemed significantly higher than industry observers had previously forecasted, not least amongst the PV-Tech editorial team.
My colleague Emma Hughes has posted a new story regarding what GSE is reporting while my original piece appears here in the editor’s blog section for reference, scrutiny and comment etc…
Lost in translation
Having had time to use several different translation services, though I can't be sure which one, if any, are translating better than others, my calculations from GSE’s figures stand up, but the timing/date to which those installations can be assigned to, come into question and doubt.
It seems that total accumulated power at the end of 2010 since reporting began in 2006 could have reached up to 3,000MW. That meant that currently connected installations in 2010 equated to 1,850MW, not the 5,800MW figure I had used in my story.
I also noticed that GSE used a 2009 installation figure of 711MW, but I have seen from GSE a figure of 583MW in the past and 720MW, which is in my chart.
According to the latest GSE press release, connected installations in 2010 alone equalled 1,850MW, a figure very much in line with market forecasts and representing growth of 160% over the previous year.
However, GSE also reported that another 55,000 installations had been filed for connection by June 2011 that equated to a further 4,000MW of installations, which importantly would apply under the rules for the 2010 feed-in tariff rates.
That’s where it would seem I went wrong?
According to GSE, should all these requested system connections actually get connected, then the total cumulative installed figures would have reached 7,000MW.
Importantly, in my defence, my translations clearly point to GSE stating that the 7,000MW figure would be expressed as “to end 2010.”
That seems to support my 5.8GW installed headline for 2010
Indeed, based on the FiT regulations (translation issues abound), this is provided that system connection requests have notice of the completion of the work by December 31, 2010.
The jury is therefore out at the moment as to whether the market research firms would calculate all of these connections as installations in 2010 or only those “officially” connected by the end of 2010, as stated by GSE at 1,850MW.
The final conundrum that emerged from the GSE press release was the mention of 8,000MW “already” achieved, with the assumption (by myself at the moment) that this referred to a further 1,000MW of new system connection applications already received in 2011. Thus on a cumulative basis Italy can expect to have reached its 2020 target of 8,000MW of PV installations in 2011, significantly ahead of schedule.
Below is my original article
Official photovoltaic installations in Italy reached an accumulated 7,000MW at the end of 2010, according to new figures released by GSE in Italy at an informal hearing before the Italian Senate Commission. Installations had reached 1,142MW at the end of 2009. This means that a massive 5,858MW was installed in the country in 2010, significantly higher than any previous projection from industry observers and market research firms. Significantly, the GSE noted that it expected January 2011 installations to top 1,000MW.
Market researchers had only recently made revisions to forecasts for the country, increasing installation figures to around 3,500MW, which is still a significant increase from installations achieved in 2009.
The Italian National Action Plan on PV had originally targeted achieving cumulative installations of 8,000MW by the end of 2020; that figure is expected to be reached by the end of January 2011.
According to IMS Research’s PV Research Director Ash Sharma, the GSE figures for installations also include partially completed PV systems. Based on its own renewed checks, the market research firm believes actual installations for 2010 are closer to around 4.0GW, based on the announcement from the GSE.
IMS Research noted that it is unlikely that sufficient quantities of modules and inverters were available in Italy to reach 5.8GW of new installations, amid global demand booming in 2010
They believe that project developers simply applied for grid-connection on partly constructed projects (that may not even have modules installed), in order to benefit from the higher tariff.
Based on this latest information, the market research firm believes 3.5-4.2GW of fully installed PV capacity was added in Italy in 2010.
Sharma also noted that IMS Research’s global forecast for 2010 had installations reaching 17.5 GW, somewhat higher than most analysts, banks, and major suppliers estimated; however, the discrepancy had been attributed to Italy; where many others had expected only around 1-2 GW to have been installed versus IMS Research’s 3.5GW.
Sharma said in response to the GSE announcement that, “These massive numbers from GSE will undoubtedly shock the industry and will have serious repercussions for 2011. PV capacity growth in Italy is now well ahead of the government’s intended roadmap and may now lead to an early intervention to curb growth in this overheating market. This may then lead to the unwanted consequence of a further boom in demand as developers seek to complete projects in time, as they did in Germany in early 2010.”
It should be noted that forecasting PV installations, especially in Italy has proved problematic for all industry observers since there are numerous bureaucratic systems and amendments before final figures are established.
What the analysts are saying
Barclays Capital analyst, Vishal Shah has also bothered to read the Italian small print on how PV connection rules work for systems that GSE uses to calculate total installed MW in the country.
Shah noted that investor heads had also been turning over the huge installation figures in Italy.
The Italian news, if true, increases the risk of additional subsidy cuts/annual caps in the Italian solar market. Since the press release is from GSE, we believe the probability of such a big increase in installations is very high. Applications to the GSE require proof of project completion and we doubt if many of these project applications were incomplete. Even if some of these applications were incomplete because of lack of panel availability in Q410, we believe the probability of most of these projects getting constructed before the April/August 2011 deadline is very high. In such a scenario, we believe the Italian market would likely reach the 8GW target set by the government by 1H11.
In an another investor note he noted the following;
Yesterday, GSE, the Italian grid operator, announced that the cumulative installed solar capacity could have reached 7GW at the end of 2010, compared to 1.14GW at the end of 2009. Although only 1.67GW of new systems were connected to the grid in 2010 (according to data tracked by the GSE), the grid operator has received applications for 55,000 additional systems representing 4GW worth of capacity, ahead of the 2010 year end feed in tariff cut.
If this number is accurate, it implies that Italy installed ~6GW of solar capacity in 2010. Furthermore, GSE believes that the country may have already installed another 1GW of systems in January of 2011, reaching the 2020 target of 8GW, almost a decade early.
All of which supports my original story. Thank you, Vishal I owe you a pint at Intersolar, OK?
However, in a phone conversation with market analyst, Ash Sharma from IMS Research, betting on GSE’s figures is going to be a dangerous thing. He followed-up via email.
Although the GSE knows how many applications they have in backlog, they DON’T know what the total MW value is until they process them. They have estimated an average plant size and multiplied by the number of applications to come out with 7GW, the actual number could be way higher or lower than that if they’ve gotten the average size wrong…
The general consensus is that it’s doubtful enough modules and inverters were available for such a high installation figure, so the final installed figure has more chance of falling than rising, though no one is saying that will happen with any strong conviction - just yet!
That was the title of an investor note issued by Mark Bachman at Auriga USA.
He is more sceptical over the GSE figures and provides this rationale.
Applications vs. Installations
We believe the GSE report refers more to applications than actual installations. As recently as a month ago, GSE estimated cumulative installations for 2010 would amount to ~2GW. In a change since then, 2009 installs were revised upward to 1.2GW from 800MW and expected 2010 installs to 1.8GW. Such a change at year-end is suspect and more likely representative of increased applications from investors worried about incentive declines impacting IRRs.
Channel checks missed the elephant moving into the living room
Unlike email, there is lead-time associated with the movement of modules from the manufacturer to the final installation. With investor estimates for 2010 Italian demand near 1GW, a full 4.8GW below that implied by GSE, it is highly unlikely that an entire industry of channel checks missed this massive influx of modules.
Shipments don't add up
Increasing our 2010 industry demand assumption to account for GSE's estimation, total industry demand in 2010 would come to 17.3GW. Summing reported and estimated quarterly shipments from the 38 module suppliers providing more than 90% of industry module supply, we optimistically estimate 15GW shipments for 2010 installs. In a best case scenario, this suggests shipments to Italy in 2010 could be no more than 3.5GW.
In what Mark says and indeed what market analysts said to me as they track such shipments and installs in the same way, makes acceptance of the GSE figures hard to swallow.
Mark also noted that the figures for 2009 were revised upwards to 1.2GW but that was news to me!
Another part of the conundrum is what would the GSE figures do to all the market forecasts?
Mark Bachman is basically sticking with his own methodologies and I suspect others will do. Of course that’s until the final count from GSE, which is expected in mid-year. That’s six months of FUD…
Contributions on what the installation figures for Italy, as stated by GSE, actually are going to be, would be more than welcome!