CAPÍTULO II.2: REVISIÓN DE PROGRAMAS DE DEPORTE ESCOLAR
FASE 6 Diseño de instrumentos de seguimiento y evaluación
12. Deporte escolar en la Comunidad Foral de Navarra
Wind Energy Resource Modelling
8
Wind E nergy Resource Modell i n g
This chapter reports on the use of the wind-energy resource model, WAsP , and follows on from the wind-monitoring study (Chapter 7). Mortensen et al. (2002) described in detail, WAsP version 7.2, which was in use by up to 800 users worldwide in 2002 to model their wind energy resources. The literature indicated the WAsP model has been increasingly used in the micro siting of wind turbines and assessment of potential wind turbine and wind farm outputs since the first version was released in 1 987.
In this section, details are given of some applications of WAsP and some of the limitations inherent in the design and use of this model. Section 8.2 details the assessment of which of two reference sites would be most suitable to form the basis of the regional wind atlas model. These assessments include a correlation and regression analysis between the wind data, and an assessment of the terrain differences between sites through an analysis of the Ruggedness Index (RIX) numbers (section 8.2 . 1 ). The next stage of the modelling was the setting of the model parameters of inversion scale length, and the level of forcing needed (section 8.2.2).
The results are given in Section 8.3 which includes sections on comparisons between the observed wind climate and that modelled (section 8.3 . 1 ), regional wind atlases for wind speed and wind power density (section 8.3.2), and a comparison between the RIX number and the prediction error (section 8.3.3). These results are discussed (Section 8.4) and then summarised (Section 8.5).
8. 1 Appl ications of WAsP
Accurate predictions of wind-speed data at other sites can be made if both the predictor and predicted sites are subject to the same weather regime, prevailing weather conditions are close to being neutrally stable, the surrounding terrain is sufficiently similar, and that the reference data were reliable (Bowen & Mortensen, 1 996; Reid et a/. , 1 998; Reid, 1 997; Bajic, 1 999). If the terrain was very steep, then separating flows usually occurred and these flows were not treated correctly by the calculations used by WAsP. WAsP cannot consider any potential large-scale atmospheric stratification due to thermally driven wind flow systems (Farrugia & Scerri, 1 999), and in acknoWledging this limitation, they still considered that using WAsP saved money and avoided time-consuming monitoring programmes for site prospecting. This was confirmed by Hansen & Mortensen (1 999) who used WAsP modelling for micro-siting and wind farm layout optimisation over a five month period of measurements using calibrated site monitoring parameters and reference data.
Reid (1 997) used an early version of the WAsP model to determine the mean wind speed and direction frequencies at ten anemometer stations in high wind areas within the Manawatu region of New Zealand. Only one of the sites modelled displayed closeness to reality. The other sites were highly channelled by the surrounding terrain and WAsP under-
predicted both the mean wind-speed and the predominance of the channelled directions. However, Reid (1 997) also found that the results from a standard WAsP run were not overly affected by contours more than a few kilometres away from the grid centre of the zooming grid area. Only adjustment of the inversion and softness parameters to encourage channelling did the outer ranges have an impact on the wind flow. This point has been noted with respect to this study where localised channelling effects were found at Wind Sites 2 and 3 (Chapter 2). The adjustment of the inversion and softness parameters will be undertaken in order to encourage this channelling behaviour in the model.
8 . 2 Modelling with WAsP i n the SPiRAL Framework
The WAsP model normalises a wind-speed and direction data set by removing the effects of the roughness and site obstacles at the subject site and this data is then used to estimate the wind regime at other sites re-initiating the effects of the new site-specific roughness and obstacle inputs and assumptions (Figure 8. 1 ).
Model: Sheltering obstacles
Data input: Position and dimensions
Model: Roughness of the terrain Data Input: Terrain classification
output: Modelled wind climate predicted for a separate site
Model: Mountainous or complex terrain Data input: Terrain contour line data
Normalised regional wind climate data with sheltering obstacles, roughness, and terrain complexity removed from the data set. This data is then re-applied based on the location being modelled.
Figure 8.1 A schematic diagram of the WAsP wind energy resource modelling process. Figure 8.1 adapted from Mortensen et al. (2002).
The modelling process of WAsP was prescribed in the literature (Mortensen et al. ,
2002) and this study has adhered to these modelling precepts as required to produce suitable output from this study (Figure 8.2). The results of the modelling will be used to 'prospect' for potential wind sites using the initial wind atlas mapping outputs. This output will subsequently be validated using the results from the intermediary sites. This will produce the required short term duration data pertaining to the modelling of the wind resource in HOMER, the short-term duration mean wind-speed and Weibull 'k' values.
Wind Energy Resource Modelling
Figure 8.2 The WAsP wind energy resource modelling process within the SPiRAL decision framework outlining the interaction with resource monitoring.
Due to the sequence of progress in the study, much wind data from Wind Site already existed prior to the purchase and use of WAsP in December 2000 (Irving, 2000; Murray & Sims, 2001c; and Murray & Sims, 2002). The chronological sequence is documented in Figure 7.29. This indicates the full extent of the data set available for use in the WAsP process in SPiRAL.
The wind data from Wind Site 1 was therefore used to produce an initial wind atlas of the Totara Valley region (Figure 1 4. 1 ). From this, further potential wind monitoring sites were identified. An alternative potential reference site (Site 5) was identified. This was considered necessary due to the complexity of the terrain around the existing Wind Site 1 . An analysis between the potential reference sites would indicate which one to use for modelling. The reference site will than be used to assess the short-term duration before a wind atlas could be produced, and from this duration, the WAsP model will be used to produce the data required for the short-term duration data required by HOMER (Chapter 9).
8. 2. 1 Reference Site Analysis
In order to assess which of reference Wind Sites 1 or 5 to use to model a regional wind atlas, an assessment needs to be made on the similarity of the terrain and the data between the reference sites 1 and 5, and Sites 2, 3, and 4. WAsP uses a data set from one location to predict the wind climate of another location, and if there are data from the predicted site, this can be used to validate the WAsP predicted results.
Reference site selection can be done with a RIX analysis or comparison between wind data with regression or correlation analysis. This type of reference site analysis is only necessary due to there being two reference sites for this study. Most other analyses would only have one site and the analysis would only be required when assessing this site with the intermediary sites.
Ruggedness Index Analysis
The ruggedness index (RIX) number was defined as the percentage fraction of the terrain within a user-defined radial distance from a specific site that is steeper than some nominated critical slope (Mortensen et al. , 2002). There was no indication in the literature of what radial distance was best to use in a RIX analysis, but figures of 250 and 1 500 to 3500 metres were used in reported studies (Bowen & Mortensen, 1 996; Bajic, 1 999; Mortensen et al. , 2002). I n WAsP however, the default radius was 3500 metres (Mortensen et al. , 2002) has been used in this study.
The RIX number is to be used in the following manner: