Wind%20Resource%20Estimation%20Using%20Data%20in%20the%20Public%20Domain

1
Wind Resource EstimationUsing Data in the Public
Domain

• Group 3
• Alex Thomson
• Arnaud Eté
• Isabel Reig Montané
• Stratos Papamichales

2
Academic Supervisor - Dr Nick Kelly Contacts
at SgurrEnergy - Richard Boddington - Neil
Doherty - Jenny Longworth
3
Plan for the Presentation

1. Brief description of key background points.
2. Processing of the raw data.
3. Description of the methodology.
4. Description of the modelling software WAsP.
5. Application in Scotland / Validation of the
   results.
6. Case studies in India / Results.
7. Uncertainty of the wind resource prediction.
8. Conclusion.

4
Novelty of our Project

• Current method of wind estimation
• Site survey.
• Meteorological mast erected for at least 12
  months (between 15,000 and 22,000 in the UK).
• Correlation with long term data from a nearby
  meteorological station.
• Its not always possible to set up a mast and
  have a nearby weather station.
• Novelty of our project create a model to predict
  wind resource in any given location by using
  reanalysis data, topographic maps and Google
  Earth (all freely available on Internet),

5
Aims of our Project

• Establish a methodology to estimate the wind
  resource of any site.
• Test the compatibility between the reanalysis
  data and the modelling tool WAsP.
• Verify the accuracy of this methodology by
  applying it to well-known sites in the UK before
  using it to identify a few good sites in India.
• Determine whether the methodology is suitable to
  estimate the wind resource of a site as a
  substitute to site survey and erection of
  meterological mast.

6
The Reanalysis Project

• Joint project between the NCEP and the NCAR.
• Project created to reanalyse historical
  atmospheric data.
• Aim was to build a Climate Data Assimilation
  System.
• The observations include
• Balloon soundings,
• Surface marine data,
• Aircraft data,
• Satellite measurements.
• These are all real observations, not output from
  a numerical model.
• Data available from 1948 to the present.
• Second improved version from 1979 onwards.
• Use of 27 years of reanalysis data in this
  project.

7
The Shuttle Radar Topography Mission

• Joint project between the National Geospatial
  Intelligence and the National Aeronautics and
  Space Administration (NASA).
• Produced digital topographic data of the Earths
  land surface (between 60ºN and 56ºS latitude).
• Data points located every 3-arc-second on a
  latitude/longitude grid.
• Format incompatible with WAsP. Required some
  modifications.

8
Google Earth

• Provides high quality satellite imagery.
• Covers the entire globe.
• Used in this project to identify terrain
  characteristics.

9
Data Processing
10
Reanalysis Data Processing
11
Topographical Maps from SRTM Data to WAsP Maps

• SRTM data obtained from the NASA Website
  incompatible with WAsP and cannot be directly
  used
• coordinates transformed from latitude/longitude
  grid to UTM coordinates.
• raw data processed into a contour map.

12
Simulation with WAsP
13
How WAsP Works

• WAsP calculates a Wind Atlas (geostrophic wind)
  using the wind data of a reference site and
  considering the terrain roughness, contours and
  obstacles of the site.
• The Wind Atlas is transferred to the potential
  turbine site (considered as representative for
  both sites).
• WAsP generates the wind climate of the potential
  site taking into account the terrain conditions
  at the site.

Diagram taken from WAsP website
14
The Methodology
15
Case Sites in the UK
16
Case Sites in the UK

• 3 wind farms
• Dun Law
• Hagshaw
• Elliots Hill
• 6 meteorological masts
• Dounreay
• Beinn Tharsuinn (3)
• Coldham
• Kentish Flats

17
Correlation between Measured and Reanalysis Data
18
Example of Dun Law
19
Results
20
(No Transcript)
21
(No Transcript)
22
Additional Errors to Consider

• The turbine availability nominal loss of 3.
• Power curve density correction
• Dun Law - 2
• Hagshaw - 2
• Elliots Hill - 1
• Power curve performance nominal loss of 0.83.
• Wind hysteresis loss of 0 - 0.5.
• Blade contamination nominal loss of 0.5.

Total additional losses 7
23
Discussion

• It is clear that the introduction of the
  roughness estimate has a significant effect on
  the model.
• Roughness estimate (from Google Earth) gives a
  more accurate prediction.
• Results for the wind farms are within 15 of the
  actual power produced (within 10 using 27 years
  of reanalysis data).
• The results appear good enough to justify an
  application of the methodology in India in order
  to get a first approximation of the wind resource
  of the sites.

24
Case Sites in India
25
Gujarat
Tamil-Nadu
26
Elements to Consider in Locating the Turbines

• The power density map.

• The isoslope maps the maximum slope to build
  a turbine is 10.

• The capacity factor of the farm.

27
Results and Performance of the Wind Farms
gt 30 good gt 25 ok lt 25 poor
Site Capacity Factor in

UK
Dun Law 27.3
Hagshaw 27.5
Elliots Hill 35.2

India
Gujarat 1 25.3
Gujarat 2 21.3
Tamil Nadu 19.5
28
Estimation of the Energy Production
29
Uncertainty of the Wind Prediction
30
Sources of Error in the Wind Prediction

• Reanalysis data
• Topographical maps
• Prediction by WAsP

31
Accuracy of WAsP Prediction

• The conditions to fulfil to obtain an accurate
  predictions using WAsP are
• The whole area is clearly subject to the same
  weather regime.
• The prevailing weather conditions are close to
  being neutrally stable.
• The surrounding topography is sufficiently
  gentle and smooth to ensure that flows stay
  attached and that large-scale terrain effects
  such as channelling are minimal.
• A good quality of data.
• A proper use of the WAsP program.

32
Factors Affecting the Prediction Process

• Atmospheric Conditions
• Orography
• Weibull Frequency Distribution
• Wind Direction

33
Project Outputs

• General methodology use of publicly available
  data with computer modelling tool.
• Algorithm for transforming reanalysis data into
  WAsP format.
• Configuration of WAsP to use reanalysis data.
• Methodology validation process.
• Error estimation of the methodology.
• Site selection process.

34
The Methodology Summed Up
35
Conclusion

• After processing, reanalysis and SRTM data can
  be used with WAsP.
• The reanalysis data appears to be suitable to
  estimate the wind resource of any given site.
• The results for our case sites in the UK stay
  within 10 of the actual energy produced using 27
  years of reanalysis data.
• However, the methodology was only validated on 2
  sites.
• Further studies at different sites should be
  carried out to confirm the suitability of the
  methodology.

36
Questions ?