Accuracy of Power Curve Measurements

Carl Jørgen Christensen (Editor), J.B. Dragt (Editor)

    Research output: Book/ReportReport

    476 Downloads (Pure)

    Abstract

    This report represents a start on the problem of reaching comparability and common acceptability of power curves measured at the different test stations. The problem is essential for establishing a future common set of certification rules. Such common rules in turn will serve the need for opening the internal market for windmills within the EEC. A few non-EEC members of the international IMTS cooperation between test stations have taken part in the project. This has been welcomed, as it will help opening the global market too.

    Through the examination of a large number of error source, the test site specific errors were found to be the most important source of uncertainty. This has to do with how well an anemometer in a mast separated from the windmill represents the wind speed that drives the windmill. Three factors create this importance. First, when calculating the progression of errors from a power curve error into an uncertainty on the yearly production prediction, a multiplier of 2-*J must be applied to the velocity uncertainty. Secondly, turbulence induces errors. Thirdly, the terrain around the windmill is almost never homogeneous enough to avoid a
    flow distortion, that induces errors in the wind speed chosen to represent the driving speed.

    A host of measures to be taken by the test station are recommended, aiming at reducing the uncertainties. Without an effort in this direction, the power prediction uncertainty can well be up to 25-30$. But it ought to be possible to reduce this uncertainty to a level of 5-10$. It is recommended, that the power prediction error is calculated and quoted for a Rayleigh distribution with 7 m/s average wind speed.
    Original languageEnglish
    Place of PublicationRoskilde, Denmark
    PublisherRisø National Laboratory
    Number of pages133
    ISBN (Print)87-550-1307-4
    Publication statusPublished - 1987
    SeriesRisø-M
    Number2632
    ISSN0418-6435

    Keywords

    • Risø-M-2632

    Fingerprint Dive into the research topics of 'Accuracy of Power Curve Measurements'. Together they form a unique fingerprint.

    Cite this