ANEMOS: Estudio sobre prediccion eolica en la Union Europea (2007)

Report with an overview of the project and its results prepared by partner IDAE.

IN SPANISH.

Download the document

Last update: 12.04.2006

Short-term forecasting - ANEMOS project
Workshops Track / Session Code: BW4
Tuesday, 28 February 2006, 16:00 - 18:00

Chair(s):
Thierry Langlois d'Estaintot, DG RTD, European Commission, Belgium
Philip O'Donnell, ESB National Grid, Ireland

download papers & presentations

			paper	present.
BW4.1	Next Generation Short-term Forecasting of Wind Power - Overview of the Anemos Project.	George KARINIOTAKIS, Ecole des Mines de Paris, France.
BW4.2	Short-term Forecasting Using Advanced Physical Modelling	Gregor GIEBEL, RISOE, Denmark.
BW4.3	Advanced Statistical Modelling and Uncertainty Assessment for Wind Power Forecasting	Torben Skov NIELSEN,Technical University of Denmark, Denmark.
BW4.4	Short-term Forecasting of Offshore Wind Farms Production	Jens TAMBKE,Univ. Oldenburg, Germany.
BW4.5	Evaluation of Advanced Wind Power Forecasting Models	Ignacio MARTI,CENER, Spain.
BW4.6	The ANEMOS Wind Power Forecasting Platform.	Igor WALDL, OVERSPEED, Germany.

Views of the Anemos Stand and the Exhibition Hall.

On 14-15 November, the European Commission’s Directorate-General for Research hosted the second “Communicating European Research” Conference. Based on the very positive experiences from the 2004 edition, the event focused on the manifold aspects of science communication and provided an excellent forum and meeting place for scientists, communication professionals and journalists. The conference took stock of the way towards the Seventh Framework Programme.

Download the document

This brochure provides an overview of research in the field of Wind Power, describing the current state of the art and the results achieved in EU-funded research and demonstration projects under the Thematic Programme ‘Energy, Environment and Sustainable Development’ of the Fifth Framework Programme (1998-2002). The projects, which have been compiled into six research areas, from turbine technologies to demonstration of wind power applications, are summarised giving the scientific and technical objectives and achievements of each, plus contact details for the participating organisations.

Objectives:

This Work-Package aims to develop forecasting models for offshore wind farms. In a given area of interest, offshore wind parks are currently located in the 5 to 10 km zones from coast. Considering the surface covered by wind farms, the resolution of the wind grid needed is about 1 km. The data available for offshore applications will be measurements at the level of wind farm as well as Numerical Weather Predictions (NWP) provided by a meteorological system. The downscaling of NWPs to the level of wind farm involves several difficulties; for example due to the variable roughness of sea surface that depends on wind speed. This Work-Package focuses on the specificities of the meteorological conditions offshore. The roughness is very low, and the thermal state of the atmosphere, the stability, is for long periods very different from the near neutral case expected onshore. Additionally, the low roughness increases the influence of stability on the wind speed profile dramatically. It is of interest to investigate the parameters that influence the wind speed offshore most. In a first step purely NWPs are evaluated against measurements.

Of special interest is the evaluation of meteorological forecasts by estimation of wind speed on a kilometric grid from Satellite Radar data. Radar images from open waters contain information about wind speed and wind direction. In contrast to mast measurements, the absolute accuracy is quite poor (rmse ~2 m/s), but spatial information for a large region can be delivered. This spatial information about wind speed and direction can be used e.g. to validate shadowing within the windfarm. Patterns of wind field around the wind farm can be defined from Satellite images. It will be examined how this high-resolution information can be integrated in either physical or statistical prediction models.

The ultimate objective is to develop statistical (i.e. artificial intelligence based ones) and physical models able to perform accurately for the offshore case. The physical models should be able to describe the vertical wind profile for vertical refinement of the wind speed. Emphasis is given in modelling spatio-temporal characteristics in large offshore wind farms. Offshore installed capacity will be very concentrated. Therefore large power fluctuations will occur much more often then onshore.

Description of Work:

Project flow-chart

Back to Task Overview

Objectives:

Short-term prediction is used operationally in a number of countries. The number one complaint by the utilities that use the predictions operationally is that the predictions are not of the same quality as they are used to from load predictions. The short-term prediction models themselves (defined as the "translator" of the NWP model results into wind power forecasts) are quite accurate if fed with measured data. The models also reach a significantly better result than pure meteorological forecasts. However, the main error source is the NWP model. WP-4 tries to especially help out in cases of bad predictions, where the performance of the NWP model for a particular site is not good enough.

The reasons for this less-than-adequate performance can be insufficient resolution of the local flows, not good enough parameterisation of physical phenomena on a scale below the grid resolution, a non-representative data source feeding the model in the vicinity of the site, and a number of other (often large-scale) effects. The main push in this WP is to use meso-scale and CFD models to capture the sub-scale effects not resolved by the main NWP model. The main idea here is that better forecasts of the worst forecasted sites leads to a higher quality forecast for the areas, thus increasing the usability of wind power in a large-scale grid and reducing the "ecological rucksack" of wind power.

Description of Work:

The aim of this work package is to improve existing physical models or to elaborate more advanced ones for wind resource prediction especially in complex terrain. The developed models will consider as input high-resolution meteorological information. Due to the shortening of the modelled area, more detailed models for the terrain will be developed. The Work Package will examine the benefits from downscaling wind power using advanced CFD modelling.On the other hand, multi-dimensional MOS correction matrix will be generated based on a number of simulations with KAMM. This will increase the speed of a physical model and will account for Mesoscale effects.Finally, the long-term (up to 7 days ahead) wind resource predictability will be assessed.

Project flow-chart

Back to Task Overview

Objectives:

The objective of this Work Package is to develop accurate models for wind resource forecasting based on advanced statistical and mainly on artificial intelligence methods (i.e. fuzzy logic, neural networks). These techniques permit to combine various types of explanatory inputs like wind direction, wind speed from neighbour sites, numerical weather predictions etc. Statistical techniques are very promising when high-resolution meteorological information is used as input to predict wind production up to 48-72 hours ahead.

Emphasis will be given to the development of statistical power curve models to estimate the relationship between wind power and local forecasts for meteorological variables such as wind speed and direction. Experience so far shows that one of the main error sources in wind power prediction lies in insufficient power curves. The use of certified power curves does not guarantee that the relation between wind speed and power output is accurately described. Models based on artificial intelligence will be developed for statistical downscaling. This consists in using a non-physical model for interpolating metrological forecasts from the nodes around wind farm to the level of wind farm and height of wind turbines. Statistical modelling can be an alternative technique to explicit terrain and roughness modelling.

Tasks will be dedicated on the development of approaches for regional or national forecasting of wind power (upscaling). Robust approaches for the on-line tuning of prediction models will be developped and evaluated. Such methods will permit adaptive models to avoid producing outliers in case of erroneous data input or extreme wind conditions.

Description of Work:

Project flow-chart

Objectives:

The aim of this work package is double fold. Initially, existing models developed by the partners called hereafter as “base-line” models will be set-up for a set of case-studies defined in WP-1. Then, using sets of data, prediction results will be produced covering sufficiently long periods. The development of ANEMOS shell and database in WP-6 will provide the possibility to host the data and interface with base-line models so that these prediction results can be easily compared to models developed in the project, but also, in order to examine the performance of combined models. In a second stage, the off-line evaluation of the models developed in the project will follow. The results from base-line models will permit benchmarking and extensive evaluation of the benefits from the developments in the project. Appropriate case-studies will be defined in WP-1 to cover a wide variety of environments, flat and complex terrain, on and offshore wind farms; also different forecasting horizons, from minutes to days; and different spatial scales, from the wind farm to the region.

This work package will undertake the challenging task to compare output of meteorological models for a number of selected sites. The conclusions are expected to be very critical for selecting the appropriate NWP systems for wind forecasting applications using cost/benefit analysis.

Finally, appropriate criteria will be defined for the evaluation of the methods with emphasis to their performance in extreme weather conditions, as well as their robustness in on-line operation conditions. The evaluation of the performance of the applied techniques will be done in off-line mode, using the historical database created in WP1. The validation of the models in on-line operation will continue in WP6 after the installation of ANEMOS at the selected sites.

The results of this work package will be crucial for the development of the project. The error estimation methodology will be used throughout the life of the project in order to assess the performance of the different models. The test cases will provide results, which will be the guidelines to develop the prediction models.

Description of Work:

Project flow-chart

Back to Tasks Overview