Onshore Methods
The first step to determining ideal onshore locations for wind farms requires the isolation of areas that are suitable to build upon. This requires the removal of inapt land such as land reserves, areas that exhibit dangerous slope steepness, as well as areas at higher elevations where winter temperatures may be too cold for proper win turbine functioning.
In order to combine and spatially analyze all these factors using GIS, we converted all layers to a raster format under a uniform resolution and projection. Raster resolution was set to 77.09 meters by 77.09 meters and the projection used was NAD_1983_Albers. The raster resolution was defined by the Elevation layer which had the lowest, and thus limiting, resolution.
Land Use
A land use layer for BC was used to remove reserved areas, water bodies, or any other land areas where a wind farm would be inappropriate or prohibited to be built on. These land use types were reclassified to “NoData” so that they would not be included in our multi-criteria evaluation (MCE). “NoData” land types were: alpine, estuaries, areas of fresh water, glaciers and snow, mining, recreational areas, residential/agricultural mixtures, selectively logged areas, sub-alpine avalanche chutes, urban sites and wetlands. Land types that were included in our analysis were reclassified to values ranging from 0 to 1 depending on their suitability for wind farm construction. These layers and their assigned values are shown in the table below:
| Land Use Type |
Reclassified Value |
| Agriculture |
0.5 |
| Barren Surfaces |
1 |
| Old Forest/td>
| 0.4 |
| Rangelands |
0.5 |
| Recently Burned |
0.7 |
| Recently Logged |
0.7 |
| Shrubs |
0.8 |
| Young Forest |
0.4 |
| Other |
NoData |
Elevation
Areas of land at elevations greater than 1500m were taken out of our analysis via reclassification to “NoData” as well. These areas were removed because wind turbines begin to freeze at low temperatures of -20°C, and heights above 1500m can exhibit such temperatures in BC during the winter (Parker 2008). Elevations below 1500m, similar to land use types, were reclassified to values from 0 to 1 depending on their suitability. It was assumed that greater elevations were more suitable for wind turbines due to increased wind speed with altitude. In addition, birds are known to fly at lower altitudes during standard flight (not during migration) and thus areas of greater elevation would be environmentally beneficial (Lee 2004). Selected elevations and their assigned values are shown in the table below:
| Elevation (m) |
Reclassified Value |
| min-250 |
0.05 |
| 250-500 |
0.2 |
| 500-750 |
0.4 |
| 750-1000 |
0.6 |
| 1000-1250 |
0.8 |
| 1250-1500 |
1 |
| 1500-max |
NoData |
Slope
Like elevation, slope is an important factor in assessing the suitability of locations for constructing a wind farm because it affects the building and energy producing aspects of the wind farm. A wind farm should not be constructed on land with slopes greater than approximately 16 degrees as slope angles greater than 16 degrees may result in flow separation over the wind turbines. If flow separation does occur, the air will re-circulate and generate more turbulence, which in turn increases stress on the wind turbine and decreases energy producing capacity (Peterson et. al 1998). Furthermore, wind farms shouldn’t be constructed on slopes greater than 16 degrees simply because there is great difficulty associated with building on steep slopes. To generate a slope raster surface, the slope spatial analyst tool on the British Columbia DEM was used and cells with slopes greater than 15 degrees were removed by reclassifying them as “NoData”. The remaining values were reclassified according to the break values found in the table below.
| Slope (°) |
Reclassified Value |
| 0-5 |
1 |
| 5-10 |
0.75 |
| 10-12 |
0.5 |
| 12-15 |
0.3 |
| >15 |
NoData |
Wind Speed
Wind speed is the most important component of our multi criteria analysis because it determines the energy producing capacity of a proposed wind farm. Wind data used was data of BC’s annual wind speed at a height of 80 meters, obtained from the Canadian Wind Energy Atlas. According to BC Hydro, wind travelling at 4 m/s at a height of 65 meters is rated at fair wind resource quality (BCHydro 2009). Furthermore, the United States Department of Energy considers areas with wind speeds greater than about 6 m/s at a height of 80 meters to have suitable wind resource for development (US Department of Energy 2010). Therefore, we chose to use mean annual wind speed data at 80m height and base our break values on the aforementioned wind speeds. The data supplied by the Canadian Wind Energy Atlas was in a format known as MIF or Map Info files. A total of twelve files were downloaded each one containing a tile covering a portion of British Columbia. We used the “MIF to shapefile” converter in ArcCatalog to convert the MIF files to shapefiles. The wind shapefiles were then merged together and clipped using a BC regions shapefile to create a wind speed map of British Columbia. Wind speeds of less than 6 m/s were reclassified to “NoData” and the remaining values were reclassified according to the break values in the table below. The reclassification values were decided based on the findings that wind speeds less than 6 m/s at 80 meters height are not suitable for wind development. Therefore, all wind speeds below 6 m/s were ignored and any speeds above 6 m/s were given increasingly higher reclassification values.
| Wind Speed (m/s) |
Reclassified Value |
| min-6 |
NoData |
| 6-7 |
0.6 |
| 7-8 |
0.8 |
| 8-max |
1 |
Power Grid
To incorporate distance to power grid as a criteria in our multi criteria evaluation, we had to generate a measure of how close each cell of land is to the power grid of British Columbia. The Euclidean Distance tool was used on a BC power grid shapefile to generate a raster distance surface. The British Columbia regions shapefile was used to clip the Euclidean distance surface to achieve a Euclidean distance surface of just British Columbia. We then reclassified the British Columbia Euclidean distance surface to values ranging from 0 to 1according to the break values listed in the table below. Grid Cells closer to the power grid were considered to be more suitable for wind farm construction because it would be easier and less costly to connect the wind farm to the power grid. The break values were determined subjectively. Cells greater than 200km away were reclassified to a value of “NoData” because a building a transmission line that long would be too costly and might not be justified by the energy output of the wind farm.
| Distance from Power Grid (km) |
Reclassified Value |
| 0-30 |
1 |
| 30-50 |
0.8 |
| 50-80 |
0.6 |
| 80-120 |
0.4 |
| 120-200 |
0.4 |
| >200 |
0 |
Multi-Criteria Evaluation (Weighted Sum)
As a final step, we performed a multi-criteria evaluation of land use type, slope, elevation, wind speed and proximity to the BC power grid in order to determine the most ideal locations for a wind farm in our province. The weighted sum tool (under Spatial Analysis Tools: Overlay) was used to set weights to each criteria dependant on their relative importance to the generation of wind power. Weighted values were assigned as followed:
| Criteria |
Assigned Weight |
| Wind Speed |
0.55 |
| Land Use Type |
0.2 |
| Proximity to Power Grid |
0.15 |
| Land Slope |
0.07 |
| Elevation |
0.03 |
Deleted Features
In this final step, prohibited areas of construction were removed from the weighted sum layer. These areas consisted of national parks, provincial parks, special protection areas, agricultural land reserves, native reserves, and areas nearby cities and airports. Cities were buffered according to their DMTI Spatial Populated Place Name Code, which identifies a populated place as a major city, minor city, town, or community. The larger the population of the populated place, the larger the buffer distance (Major city > Minor city > Town > Community). Airports were buffered according to their degree of regulation. Military airports are likely to be more heavily regulated by the government and thus were given larger buffer distances. The features and their buffer distances can be found in the table below. All prohibited areas were then merged into a single layer and converted into raster format for deletion. Cells containing prohibited area were reclassified to “NoData” and all NoData cells of the merged layer (which were cells that were not part of the prohibited area) were reclassified to a value of 1. Finally, the Raster Math Times tool was used to multiply this reclassified deleted features value to the overall values obtained from the multi-criteria evaluation. This omitted the prohibited areas out of our analysis and left all other functional grid cells with their original MCE-assigned values.
| Feature |
Buffer Distance (km) |
| Major City |
10 |
| Minor City |
5 |
| Town |
3 |
| Community |
3 |
| Military Airport |
10 |
| Civil Airport |
8 |
| Other Airport |
5 |
Offshore Methods
Ideal offshore sites were determined by the isolation of areas in close proximity to commercial fishing zones, recreational routes, the shoreline, marine bird colonies, and finally, the BC power grid. Distances from each area were required as to avoid environmental and social issues.
Identical to the onshore analysis, all layers were converted to a raster format under a uniform resolution and projection. Raster resolution was set to 77.09 meters by 77.09 meters and the projection used was NAD_1983_Albers. The raster resolution was defined by the Elevation layer which had the lowest, and thus limiting, resolution.
Wind Speed and Commercial Zones
Wind speed carries the same importance in the offshore analysis as it does in the onshore analysis. Again, mean annual wind speed data at 80m height from Canadian Wind Energy Atlas was used. A total of four files were downloaded each containing a tile covering a portion of the British Columbia coastline. The wind data files were converted from MIF to shapefile and merged together in preparation for the deletion of areas not suitable for building an offshore wind farm.
Areas where building an offshore wind farm would have detrimental environmental and commercial affects were buffered and erased from the coastal wind map of British Columbia. These areas include commercial fishing areas, boating routes, and areas for water recreation. All the deleted features and their associated buffer distances can be found in the table below. The buffers for the boating routes were decided based on a study done on offshore wind farms in the Great Lakes. The study suggested a 1nautical mile (˜1.85km) on boating routes, which we doubled to ensure any suitable areas for wind farm development are located a safe distance away from passing boats. The 3km buffer on the coastline was selected subjectively to reduce visual and sound impacts. Buffer distances for commercial fishing areas were also set subjectively to 5 km. It must be noted that further environmental assessments must be performed when a site is selected, which could increase the buffer distances to commercial fisheries.
After the removal of all the buffered unsuitable areas, the areas with depth of 0-200m were selected from the bathymetry layer and clipped with the wind speed layer resulting in a wind speed map of areas along the coast with 0-200m depth. Although current foundation technology suggests that the maximum practical water depth for deployment of offshore wind turbines is roughly 50m, we decided to use water depths of 0-200m for two reasons: 1) we could not find bathymetry data with depth ranges less than a few hundred meters. 2) There exists technology such as deep water floating turbines that allow wind farms to be built in waters deeper than 50m.
The last step in preparing the wind speed layer for the final weighted sum was to convert it to raster and reclassify it. According to the United States Department of Energy winds speeds greater than 7 m/s at 90m height are generally considered to have suitable wind resource for offshore wind farm development (US Department of Energy 2010). Therefore, when reclassifying the wind data, any wind speed below 7 m/s was reclassified as “NoData” and any wind speeds above 7 m/s were reclassified to increasingly higher reclassification values. The break values and associated reclassification values can be found in the table below.
| Wind Speed (m/s) |
Reclassification Values |
| 0-7 |
NoData |
| 7-8 |
0.6 |
| 8-8.5/td>
| 0.75 |
| 8.5-9 |
0.9 |
| >9 |
1 |
| Feature |
Buffer Distance (km) |
| Ferry Routes |
3.7 |
| Cruise Routes |
3.7 |
| Commercial Salmon Fishing Areas/td>
| 5 |
| Commercial Shrimp Fishing Areas |
5 |
| Commercial Squid Fishing Areas |
5 |
| Commercial Octopi Fishing Areas |
5 |
| Commercial Crab Fishing Areas |
5 |
| Commercial Anchovy Fishing Areas |
5 |
| Commercial Herring Fishing Areas |
5 |
| Commercial Goose Neck Barnacle Fishing Areas |
5 |
| Commercial Geoduck Fishing Areas |
5 |
| Commercial Sea Cucumber Fishing Areas |
5 |
| Recreational Boating Routes |
2 |
| Recreational Kayaking Routes |
2 |
| Recreational Diving Sites |
2 |
| Coastline |
3 |
Bird Colonies
Offshore wind farms pose a potential risk to marine bird colonies that exist along the coasts of British Columbia due collisions, displacement, and blockage of migration and foraging routes. Therefore, it is of great importance that wind farms are placed in areas away from bird colonies (McCorquodale 2005). A Euclidean Distance surface was generated for the British Columbia coastal bird colonies and Extraction by Mask was used to attain the Euclidean distance to bird colonies in areas with depths ranging from 0-200m. The break values for the distance to bird colonies were determined subjectively. Any area less than 50km away from the bird colonies was reclassified to “NoData” because 50 km was deemed to be within the foraging distance of the bird colony. The break values and associated reclassified values can be found in the table below.
| Distance to Bird Colony (km) |
Reclassification Value |
| 0-30 |
NoData |
| 30-50 |
0.4 |
| 50-80 |
0.6 |
| 80-100 |
0.8 |
| >100 |
1 |
Power Grid
The distance from a wind farm to a power grid is very important because it has many cost and construction implications. This is especially true for offshore wind farms where little power transfer infrastructure exists. Offshore wind farms usually rely on underwater power lines as well as land based power lines to connect to the central power grid, which is costly. Therefore, it is imperative that an offshore wind farm be built a reasonable distance from the central power grid. The Euclidean Distance tool was used to quantify the shortest distance from each cell of to the British Columbia power grid. Extraction by mask was used to attain the Euclidean distance to the power grid over areas of the coast with depths ranging from 0-200m (see map below). The cells were then reclassified according to the break values below. The selection of the break values and the maximum acceptable distance was determined based on the fact that the central power grid of British Columbia was not designed to accommodate offshore wind energy production. As a result, a reclassification scheme differing from the one used in the onshore analysis was used to reclassify the data. We raised the maximum acceptable distance from 200 km for onshore to 500 km for offshore, so any cell with a distance above 500 km was reclassified to “NoData”.
| Distance from Power Grid (km) |
Reclassification Value |
| 0-100 |
1 |
| 100-200 |
0.8 |
| 200-300 |
0.6 |
| 300-400 |
0.2 |
| >400 |
NoData |
Multi-Criteria Evaluation (Weighted Sum)
We performed a multi-criteria evaluation of wind speed, proximity to bird colonies, and proximity to the BC power grid in order to determine the most ideal locations for an offshore wind farm along the British Columbia Coastline. The weighted sum tool (under Spatial Analysis Tools: Overlay) was used to set weights to each criteria dependant on their relative importance to the generation of wind power. Weighted values were assigned as followed:
| Criteria |
Assigned Weight |
| Wind Speed |
0.45 |
| Distance to Power Grid |
0.25 |
| Distance to Bird Colonies |
0.3 |
