The failure of German wind power stations to live up to their performance guarantees in recent years has given rise to a deal of bad press as well as anguish among investors who have not earned the returns they were promised (Windpower Monthly, September 2005). Part of the blame is being placed on the inadequacy of "wind energy indices" as a measuring stick for wind energy content. The wind has not blown as strongly as the wind indices had indicated, suggesting they are a faulty indicator on which to base the future earnings of any wind power plant.
The wind index system on the European continent, provided by a handful of specialist companies, operates only in Germany, Denmark, Sweden and the Netherlands (box). It was designed as a mechanism by which wind plant operators can establish if productivity above or below guaranteed performance is related to higher or lower winds. If a wind plant has not produced as expected in given winds, the owner can make a claim for "operative losses" against the manufacturer, or, if the wind turbines are out of their guarantee period, against the company with which the turbines are insured. In Germany and Denmark, insurance companies each month rely on the wind indices as the basis for making settlements for missing production.
The wind indices, however, are also used to establish the likely long term energy yield of proposed wind power projects. To save on expensive site-specific wind speed measurements, energy prediction is made for a selection of nearby wind turbines based on historic local met office data. The wind speed data is supplemented by measured production data from the turbines over a specific period, adjusted for machine availability and for wind energy content using the national wind index. In this way the project developer establishes if the wind in that period was above or below the long term average. The two sets of data are used to calibrate the long term wind energy prediction for the proposed project site.
Thousands of wind turbines in northern mainland Europe have been financed based on predictions of wind energy content using the wind indices. The big question is, what certainty can there be for investors that wind patterns of the recent past will be repeated in the future?
Reason for concern
It has recently been established that both the Danish and German indexes were developed during a 20 year period of exceptionally strong winds, seen in a 140 year perspective. The providers of the wind indices are now admitting that energy yield projections for the next 20 years based on the indices are likely to be way off mark.
"After ten years with constantly less wind from 1995 to 2005 relative to the previous ten years from 1985 to 1994, there have been good reasons to be worried," says Per Nielsen from Denmark's EMD International, who provides the Danish wind index. EMD, a commercial company, also sells wind plant siting software.
The oldest and most widely used wind index in Germany is provided by a group of independent engineers associated in Ingenieurwerkstatt für Energietechnik (IWET). Developed by Helmut Häuser and colleagues, it uses a baseline wind period of 1989 to 2002. Carsten Albrecht of German wind assessment company Al-pro agrees with Nielsen that the period covered by the indices is mainly one of relatively strong winds. The result is a high wind index, leading to overly optimistic energy yield projections for wind projects installed after about 2000, which have generally disappointed investors.
"In general there is broad agreement that the German index is calibrated with a high wind period -- and this gives reasons to expect less from projects in the way of returns over the coming years," says Nielsen.
Fears that energy yield predictions have been based on a particularly windy period, rather than a typical period, are supported by long term weather patterns. Just over a year ago, Deutsche WindGuard, a company supplying wind power output production services, pointed to the so-called North Atlantic Oscillation (NAO) as an indicator of a period of unusually high winds at the end of the century (graph). The NAO is a measure of the air pressure difference between Iceland and the Azores and illustrates the zonal circulation in the Atlantic European sector, says Deutsche Windguard. A high zonal circulation is linked to mild winters and higher winds, as in the period 1900-1939 and around 1990-2000, while the opposite was the case in the winters of the 1960s.
Long term wind conditions show large variations with the NAO revealing a tendency towards cycles of about 35 years. "But the periodicity of the wind conditions is not distinct enough to be used for predictions," says Deutsche WindGuard's Axel Albers. Nielsen also points to the NAO as an indication that the period covered by the wind indexes was one of atypically strong winds, but stresses it is only an indication.
In efforts to establish the baseline for a "normal" windy year, adjustments have been made in recent years to both the Danish index and the German IWET index. Nielsen fears, however, the changes have not gone near far enough. Based on long term historical data he argues that a significant adjustment of the Danish index is necessary. Nielsen proposes that 13.5% is knocked off energy yield predictions based on the current Danish index, including 5% taken off to account for the uncertainty factor. "For northern and western Germany this might be even more due to additional wind index problems," he says, even though changes to the German index two years ago already reduced expected energy yield by 10-15%.
For the time being, says Nielsen, "until something better is available," when conducting energy yield projections for clients in Germany his company uses a recalibrated version of the German index. "But this is perhaps not conservative enough so we recommend financial analysts to add an uncertainty factor of 5% for the longer term."
The lack of accuracy of northern Europe's wind indices has been the subject of a series of workshops organised by the Bundesverband Windenergie (BWE), an organisation representing the interests of German wind plant operators. The most recent gathering was at a workshop entitled "Wind Index and Long Term Correlation," held in the German town of Kassel in November. Albrecht heads the BWE's expert committee looking into the accuracy of long term wind assessments.
Much discussion at the workshop was devoted to how calibration can be improved. "While some participants suggested the index baseline should be derived from a longer period of 30-35 years, others considered ten years to be more appropriate to reflect the effects of recent and ongoing climate change on wind characteristics," Albrecht reports. In Nielsen's view, either a 45 year period or the years 1993-2004 would be suitable as a reference in a recalibration of both the German and Danish indices. "Use of the last 12 years or the last 45 years yield approximately the same results. It's the last 20-25 years that are atypical," he says.
Indices must also be fine tuned to take account of other dynamic developments especially in Germany "where the situation is more complicated," says Nielsen. Data for the German wind index is based largely on performance of older turbines in north-west Germany. Many of these are stall controlled, producing large volumes of power in high wind periods and far less in low winds. Their performance characteristics are different from newer turbines on high towers in the south and east of Germany, which show smaller swings in output with changes in wind speed.
Big differences in wind turbine hub height in Germany also distort the indices. "Ideally, there needs to be a wind index for turbines with 50 metres hub height and another for 100 metre towers," says Nielsen.
A major problem with all the indices is continuity. Since they are based on measurements of actual production of reference turbines, each time a reference turbine is taken out of operation, or its performance is affected by the addition of new turbines upwind or a change in topography, the measured wind energy content is influenced.
Nielsen still believes the wind index system can form the basis for accurate energy yield predictions on which to base wind plant financing. First, however, a set of common standards needs to be agreed on by all the wind index providers, he says. The Danish and German indices should at least match at the border between the two countries, which is not the case today.
Just over a year ago, BWE's wind assessment committee agreed on a standard for the preparation and documentation of energy assessments and of wind measurements for the over 30 German companies active in the sector. At the same time, the Fördergesellschaft Windenergie (FGW), an association for the promotion of wind energy, has developed a technical regulation on assessing wind station performance. "The two are to be combined by 2006. Worldwide there is no comparable system for quality control and continuing development in wind assessment," says Albrecht.
Company compliance with the standard and the technical regulation play a role in getting accreditation from the German accreditation council, the Deutsche Akkreditierungsrat (DAR). The DAR mark of approval is important for wind measurement companies bidding for work elsewhere in Europe," says Albrecht. DAR is part of the federal institute for materials research and testing.