To design appropriate multi-criteria evaluations (MCE) based on the environmental and socioeconomic constraints and criteria listed by the previous objective.
To determine the optimal site for the Sharp Hills Wind Farm, two multi-criteria evaluations (MCEs) were conducted. An MCE is a type of suitability analysis based on both spatial constraints and criteria. MCEs are used to identify optimal sites by assigning raster data cells with relative suitability scores. The first MCE conducted took into consideration only environmental constraints and criteria (Figure 3). From this MCE, the most environmentally suitable sites were located. The second MCE took into consideration only socioeconomic constraints and criteria (Figure 3). From this MCE, the most socioeconomically suitable sites were located. The overall most suitable sites were determined by overlaying the two MCE outputs (Figure 3).
Figure 3: Multi-criterion evaluations and outputs.
Constraints involve the use of the Boolean designation of pass/fail for a cell. Criteria provide the relative suitability of each cell on a continuous scale. Before incorporating the criteria into the MCE, the criteria scores were standardized on a scale from 0 to 100 using Equation [1] for beneficial factors and Equation [2] for cost factors, where X’ij represents the standardized score for a given cell, Xij represents the input value of the given cell, Xmin represents the minimum cell value of the criterion, and Xmax represents the maximum cell value of the criterion (Table 1). Weights for each criterion were allocated using a pairwise comparison method (Table 2, Table 3, Table 4). The consistency ratio was calculated to equal 0, indicating a low probability for logic errors within the conducted pairwise comparison. Slope and proximity to roads were given the highest weighting and equal weight values. Both are important in allowing for the wind farm to be economically viable, and both have important environmental implications. Agricultural land quality was given a lower weight value than those of proximity to roads and slope. While it is important to maintain high quality land for the purpose of food production, it is recognized that some crops can continue to be cultivated in areas that contain wind turbines (Rajewski, 2013).
X’ij = 100*((Xij – Xmin) / (Xmax – Xmin)) [1]
X’ij = 100*(1 – ((Xij – Xmin) / (Xmax – Xmin))) [2]
Table 1: Standardized criterion scores on a scale of 0 to 100 using a linear transformation.
| Criterion |
Standardized Score |
| Ag land class 1 |
0 |
| Ag land class 2 |
16.66667 |
| Ag land class 3 |
33.3333333 |
| Ag land class 4 |
50 |
| Ag land class 5 |
66.66666667 |
| Ag land class 6 |
83.3333333 |
| Ag land class 7 |
100 |
| Slope (0% ≤ x < 1%) |
100 |
| Slope (1% ≤ x < 2%) |
90 |
| Slope (2% ≤ x < 3%) |
80 |
| Slope (3% ≤ x < 4%) |
70 |
| Slope (4% ≤ x < 5%) |
60 |
| Slope (5% ≤ x < 6%) |
50 |
| Slope (6% ≤ x < 7%) |
40 |
| Slope (7% ≤ x < 8%) |
30 |
| Slope (8% ≤ x < 9%) |
20 |
| Slope (9% ≤ x < 10%) |
10 |
| Slope (10% ≤ x) |
0 |
| Proximity to road (150 m ≤ x < 1150 m) |
100 |
| Proximity to road (1150 m ≤ x < 2150 m) |
89.84771574 |
| Proximity to road (2150 m ≤ x < 3150 m) |
79.69543147 |
| Proximity to road (3150 m ≤ x < 4150 m) |
69.54314721 |
| Proximity to road (4150 m ≤ x < 5150 m) |
59.39086294 |
| Proximity to road (5150 m ≤ x < 6150 m) |
49.23857868 |
| Proximity to road (6150 m ≤ x < 7150 m) |
39.08629442 |
| Proximity to road (7150 m ≤ x < 8150 m) |
28.93401015 |
| Proximity to road (8150 m ≤ x < 9150 m) |
18.78172589 |
| Proximity to road (9150 m ≤ x < 10,000 m) |
8.629441624 |
| Proximity to road (10,000 m ≤ x) |
0 |
| Wind power density (0 W/m2 ≤ x < 50 W/m2) |
0 |
| Wind power density (50 W/m2 ≤ x < 100 W/m2) |
10 |
| Wind power density (100 W/m2 ≤ x < 150 W/m2) |
20 |
| Wind power density (150 W/m2 ≤ x < 200 W/m2) |
30 |
| Wind power density (200 W/m2 ≤ x < 250 W/m2) |
40 |
| Wind power density (250 W/m2 ≤ x < 300 W/m2) |
50 |
| Wind power density (300 W/m2 ≤ x < 350 W/m2) |
60 |
| Wind power density (350 W/m2 ≤ x < 400 W/m2) |
70 |
| Wind power density (400 W/m2 ≤ x < 450 W/m2) |
80 |
| Wind power density (450 W/m2 ≤ x < 500 W/m2) |
90 |
| Wind power density (550 W/m2 ≤ x < 300 W/m2) |
100 |
Table 2: Agricultural land class, slope, proximity to road, and wind power density were each assigned a pairwise rank in order to complete the pairwise comparison method.
| |
Agricultural Land Class |
Slope |
Proximity to Roads |
Wind Power Density |
| Agricultural Land Class |
1 |
1/3 |
1/3 |
1/3 |
| Slope |
3 |
1 |
1 |
1 |
| Proximity to Roads |
3 |
1 |
1 |
1 |
| Wind Power Density |
3 |
1 |
1 |
1 |
Table 3: Agricultural land class, slope, proximity to road, and wind power density were each assigned an individual weight in order to complete the pairwise comparison method.
| |
Agricultural Land Class |
Slope |
Proximity to Roads |
Wind Power Density |
| Agricultural Land Class |
0.1 |
0.1 |
0.1 |
0.1 |
| Slope |
0.3 |
0.3 |
0.3 |
0.3 |
| Proximity to Roads |
0.3 |
0.3 |
0.3 |
0.3 |
| Wind Power Density |
0.3 |
0.3 |
0.3 |
0.3 |
Table 4: Assigned weights calculated using the pairwise comparison method.
| Criterion |
Weight |
| Agricultural land class |
0.1 |
| Slope |
0.3 |
| Proximity to road |
0.3 |
| Wind power density |
0.3 |
Table 5: Constraints consisted of the feature and its associated setback distance.
| Feature |
Setback Distance (m) |
| Urban Area |
2000 |
| Hospital |
500 |
| Airport |
3000 |
| Tourist Facility |
1000 |
| School |
500 |
| Road |
150 |
| Forest |
1000 |
| Wetland |
1000 |
| River |
400 |
| Lake |
400 |
| Provincial Park |
1000 |
MCEs were conducted following the general format of Equation [3], where Cni represents constraint maps, Cri represents criteria maps, and Wi represent the weights for each criterion. Constraint maps included in the MCEs consisted of the feature and its associated setback distance (Table 5). The environmental MCE was conducted using Equation [4]. The socioeconomic MCE was conducted using Equation [5]. Due to a lack of scientific research, there is no spatial data available on the location of aquifers within the study area. As a result, the constraint regarding ground water could not be included in this study.
Suitability = Cn1*Cn2 * …*Cn3*(W1*Cr1 + W2*Cr2 + … + Wn*Crn) [3]
Environmental suitability = Forest * Wetlands * Lakes * Rivers * Provincial Parks [4]
Socioeconomic suitability = Urban Areas * Hospitals * Airports * Tourist Facility * Schools * Roads150 * (Agricultural Land Class*0.1 + Slope*0.3 + Proximity to Roads*0.3 + Wind power density*0.3) [5]
