(LSA) Land suitability assessment for a Wind Farm location (2011)

What is a Land Suitability Assesment?:

Also known as layer-based GIS, this is the most frequent form of geographic-data space representation. Particularly, LSA (Land Suitability Assessments) is a very complex analysis in which the overlapping of thematic maps result helpful.

The analysis is based on the combination of different thematic layers such as topographic, geologic or drainage as well as in their interpretation independently from one to another.

Wind Farm location selection:

Land suitability analysis. wind farm UK

Criteria to locate a wind farm in Europe differ a little bit from one country to another.

In the case of the British Isles, the British Wind Energy Association (BWEA) estimates that the extension average of a wind farm with 20 turbines is "1" square kilometer (BWEA).

The principal objective was to locate a suitable area to construct a 90 MW Wind Farm.

The analysis was divided into two levels:

1. A national level to obtain potential areas suitable to undertake a new wind development.
2. A local level to select the final location. 
1. Two ‘asc’ files (wspeed45.asc and dominant.asc) together with roads map extracted from the 1:250000 OS map (Digimap) were used to achieve the aims of the analysis. 

Beside of wind potential, two files were used to determine the most appropriate sites.

1. Land cover data from CIS2000:
1. Factors such as woodlands, water bodies (rivers, lakes, reservoirs, etc.) and man-made built structures (cities or urban agglomerations).
2. Road map extracted from digimap (OS strategi, 1:250000). Urban and roads cause a ‘distance decay effect’ that affect the suitability for any new wind farm location.

Data gathering and management is particularly important in the selection and estimation of areas suitable for a wind farm complex.

The Mapping process:

The mapping process used two different sources.

The ‘asc’ files were converted into raster in order to map water, wood and urban agglomerations which they are variables frequently used in wind farm constraints maps, and roads which were also estimated in the assessment and extracted from the 1:250000 OS UK map.

Infrastructures such as roads are usually employed as a relevant factor that determines the final location of a wind farm. Henning (2005:82) and Baban & Parry (2000:68) recognise roads proximity as one of the most important factors in a GIS-based multi-criteria analysis.

After converting ‘asc’ files, number ‘0’ was extracted from the classification in both new raster files but a new error raised. Both maps showed differences in their screen position, making it hard the analysis. The problem was resolved by adding a shapefile of a correct UK map. 

UK Map Suitable areas wind Farms

This permitted to know which of the new maps had to be fixed. The ArcGis option ‘shift management’ enabled the analyst to resolve the differences by adding 54 in ‘x’ and 11 in ‘y’. These numbers were obtained by comparing the shapefile with the damaged new raster ‘domland’.

Euclidean distance was computed in the three extracted thematic maps from the dominant file water (water_euc_dis), wood (wood_euc_dis) and urban (urban_euc_dis)) and the roads map buffering distances to obtain the most suitable places for the final location. After buffering distances, thematic maps as well as roads and wind speed maps were reclassified from ‘0’ or non suitable to ‘1’ or suitable to develop constraint maps at the national level. Finally, the five maps were computed in the raster calculator to get the suitability national map (Output_NL.img).

As commented above, the analysis is divided into two main parts. The national level and the local level permitted to verify the suitability of the area by multiplying the national map with two local factors, slope and aspect, each of which influence in the location of the wind farm. At the local level, both, slope and aspect were derived from OS Land-Form Panorama DTM (1:50000) tiles of a Scottish region.

The tiles were processed by the function ‘mosaic to raster’ selecting the British coordinate system. Then, ‘define projection (management)’ and ‘project raster’ in which the tiles are modified from meters to kilometres, complete the process prior the slope and aspect calculation.

Slope and aspect were finally calculated and then reclassified to permit the final estimation of the most appropriate area to install a wind farm. Slope was classified following criteria found in relevant literature, selecting as unsuitable those areas in which the slope is higher than 6º. Aspect was obtained in three phases, from computing the radiant [Aspect_Sampl1*.0174] and sin [sin (aspect_rad)] to reclassifying ‘aspect_sin’ from suitable ‘1’ to unsuitable ‘0’.

Finally, the analysis was completed by computing the two derived local maps with the final output or national map [(slope_rec)*( aspect_rec)*( Output_NL.img)]. 

Conclusions:

The location and how turbines are distributed goes from 300-600 meters distance between turbines to a kilometer in Spain or Scotland (National Wind Watch).

In this case study, the location selected, North-eastern Scotland has an extension of more than 7.5 square kilometres meeting the requirements of energy production (90 MW) exposed in the beginning of the work. 

Local analysis. Scotland