Solar Farm Site Selection Using Spatial Multi Criteria Evaluation Method: Acase Study of Kewet Wereda, North-Center Ethiopia


 Solar energy can provide a great opportunity for alternative energy source and development of the nation of the country. It can access freely throughout earth’s surface so, solar energy is the best alternative energy source for solving limitation of power especially when there is drought. Therefore, this thesis aims to select suitable solar farm sites by using spatial multi criteria evaluation method in Kewet Wereda. Different spatial and non-spatial data such as solar radiation, slope, aspect, LULC, proximity distance to road, proximity distance to sub-station site, proximity distance to river, proximity distance to railway and proximity distance to town have been used to identify and map area. Some parameter data are taken from 30 meter resolution of DEM such as solar radiation, slope, aspect and river streams. Analytical Hierarchy process (AHP) was used for Calculation of the criteria weighted overlay to produce a suitable solar farm site map. The result shows that solar radiation has the highest weight and distance to river has the lowest weight with 36.6% and 3.4%, respectively. In addition, optimal solar farm sites were identified and highly suitable area found in the south, southeast and northeast part of the study area and covers 35988.71 hectare area. Those potential zone contributes to fill the energy gap between the demand and supply of the country when there is a shortage of rainfall through the country and shortage of water at the dam.


Introduction
Solar energy is vast, free, clean and renewable power source that is available everywhere on the planet [1].
The energy stored in the fossil fuels which we burn originated from the sun when they formed [2].
Energy is a critical part of modern life as almost all human activities are strongly connected with it [3].
Although in the current economic situation, the rational use of the available resources and the need to overcome the undesirable environmental effects and other problems associated with fossil fuels have forced many countries to enquired into and change to more environmentally friendly alternatives which are renewable in order to sustain the increase energy demand and potential availability of solar energy differs from region to region in the word because of geographical location and world climate zones [4].
GIS is a powerful tool for recognizing suitable site selection, for consulting, editing, analyzing and reduce risk costs of large solar energy investments [5]. In this era, GIS-based spatial multi criteria evaluation method had become gradually popular as a tool for different site selection studies especially in the energy planning [6].
Ethiopia is located in the horn of Africa and among one of the fastest developing nations on the continent. In terms of population, the country is home to over 110 million people, which makes it the second most populous country in Africa behind Nigeria and just ahead of Egypt. As a result of Ethiopia's rapid population over the past decade, the electricity demand has also been increasing gradually. The electricity generation capacity approximately 90% of the installed generation of electricity is from hydropower while the remaining 8% and 2% are from wind and thermal sources respectively [7]. Even if the country has bulk electric power resource potential in nonrenewable and renewable energy resources, these have insu cient access to all the Ethiopian regions, zones and other administrative areas [8]. Most remote areas such as Kewet wereda where there is drought or the amount of dam water is become decline impossible to get adequate electric energy. Therefore, the study aimed to select solar farm sites by determining solar radiation from 30 meter resolution DEM using spatial multi criteria evaluation method.  Nowadays, analysis of land-use suitability requires consideration of different criteria including not only the natural / physical capacity of a land unit but also socio-economic and environmental effects [10].
AHP has speci c application in group decision making, and was used all over the world indifferent decision situations, in elds such as, industry area ,government sector, healthcare sector, business area, shipbuilding, site suitability selection and education [11]. Somewhat recommending a "precise" judgement, AHP aids for decision makers delivering best generation for their goals and their understanding of the problem [12]. The essentials of the order be able to join to any feature of the conclusion di cult touchable or intangible, wisely digni ed or approximately expected, t or unwell assumed everything at totally that applies to the choice [13]. Land use/cover To know types of land covering through the study area.

Sub-station site
For identifying its economical distance from the sub-station site Road network For identifying its economical distance from the roads.

River
For identifying its proximity from the rivers.

Railway
For identifying its economical distance from the railway Slope For identifying the land surface of the area Aspect For identifying the direction /azimuths Town For identifying its proximity distance from the town.

Pairwise comparison matrix
It is a method which has commonly used to challenge the individual and unbiased decisions around qualitative and quantitative measures in multi-criteria decision making, especially in the Analytical Hierarchy Process (AHP) denoted as pairwise comparison matrices [14].

Evaluation of Matrix Consistency
The consistency ratio was calculated by the following formulas.
The CI stand for Constancy Index, deliver a measure of departure from constancy, to determine the goodness of CI, AHP compares with it by using the random index.
CR means constancy ration and Random index (RI) the CI of a randomly created pairwise comparison matrix of order 1to 10 obtained by approximating random indices using a sample size 0f 500, (Table 4) shows the value of RI sorted by the order of the matrix.  Figure 3 shows the distribution of each LULC map of the study area. The LULC of the study area is covered by the agriculture (45%), shrubs (16%), mixed forest (6%), built-up areas (21%), barely lands (9%) and water body (3%) ( Table 5).  Figure 4 shows the suitability map of the slope. There are high relations between slope and solar farm site selection which indicate that the slope less than 15 degree were good such that <3 degree is highly suitable from 3 up to 5 degrees is suitable, from 5 to15 degrees is less suitable but greater than 15 degrees taken as unsuitable because of its economic factors (Table 6).  Figure 5 shows suitability map of aspect of the study area. High solar radiation received in the orientation of respectively at area, south, south east and southwest, but east -face and west-face are medium potential and north, northeast and northwest taken low potential (Table 7).  Figure 6 shows the suitability distance of the road. The road proximity to the solar farm site is divided based on its economical factor, so <500 meter highly suitable, 500-1000m suitable, 1000-1500m less suitable and >1500 meter taken as unsuitable (Table 8).  Figure 7 shows suitability map of sub-station. Proximity of the sub-station was taken 0-500 highly suitable area, 500-1500 suitable, 1500-2500 less suitable and >2500 taken as unsuitable (Table 9).  Figure 8 shows suitability of distance of river. Highly suitable (>1500 meter), suitable (1000-1500), less suitable (500-1000) and less than 500 taken as unsuitable.  Figure 9 shows suitability map of railway.

Land use land cover Classi cation
Proximity distance to railway <500 meter high suitable, 500-1000 meter suitable, from 1000-1500 less suitable and >1500m unsuitable distances to railway (Table 11).  Figure 10 shows suitability proximity distance map of the town. Highly suitable (>6km), suitable (4-6km), less suitable (2-4km) and unsuitable (<2km) ( Table 12).  Figure 11 shows suitability map of solar radiation. (Table 13) indicates suitability range.   it, such as slope, aspect, and solar radiation and river streams. The estimated solar radiation depended on the resolution of DEM as well as the parameters such as time interval and sky size resolution [15]. Higher resolution with smaller time interval and bigger sky size results more accurate output, but also increases considerable calculation time.
GIS-based spatial multi criteria evolution method applied to select the ideal location for selecting a suitable solar farm site at kewet wereda. In addition to solar radiation potential, various topography, economic and environmental factors were taken into considering in the site selection process. As environmental (solar radiation, aspect, slope), economical (proximity to road, proximity to sub-station, proximity to the railway, proximity to the river and proximity to town). The accuracy assessments and the weighted of the study area were both have acceptable values.
[16]Kappa value 0.81 up to one indicts almost has perfect agreements. The study area accuracy assessment done with 351 points based on this points the accuracy, verify by eld survey, google earth and google map and we generate 87.7 % of overall accuracy and kappa value 81% which match with our study area so, it was acceptable to do further analysis. According to [17], the AHP pairwise comparison CR value less than 0.1 was acceptable. So, the study weight derived from the AHP pairwise comparison method its CR value 0.07 which was less than 0.1 and the weight was acceptable.
In the same title [18] was also used a multi-criteria approach in his GIS analysis to select suitable solar farm site throughout Ethiopia. By using six criteria and with low resolution satellite image and his study area coverage was very large. Which included all Ethiopian area so, in case of wide area coverage the researcher was veri ed the accuracy assessments with google earth and google map. Because of this, the researcher have poor accuracy classi cation and less accurate result, but in this study which differ from Teshome the coverage of the study area was limited and which have done with nine criteria and with 30 meter resolution of satellite image and the accuracy of the study area classi cation was veri ed by eld trip, high resolution satellite image, google earth and google map which have acceptable classi cation and good result. [19]was estimated the solar radiation from six stations measured data by using interpolation methods and mapped the Ethiopian solar energy distribution, in addition [20]was estimated the solar radiation from 17 stations measured data by interpolating to know the solar energy potential of Ethiopia. [21]unlike temperature and rainfall, solar radiation cannot be measured at sample meteorological stations and cannot be interpolated for the surrounding areas.
This is mostly since incoming solar radiation is extremely dependent on landscape and geographical structure [22].Tesfay and Sharew was identi ed and mapped the solar energy of Ethiopian with a few stations by interpolation with all Ethiopia's topography features from the lowest depression (-125 meter) to the highest elevation (4550 meter), because of this, their result is less necessary. But this study differ from both of them with methods, coverage area and data usage, this thesis was cover less area and done with determining solar radiation by using 30 meter resolution satellite data which contain all the topography of the study area and determine good solar radiation value with using nine criteria such as based on economic factors(road, sub-station ,railway, river ,town ,LULC) and environmental factors (solar radiation, slope, aspect) were used to do this study.

Conclusions
The result shows that solar radiation has the highest weight and distance to river has the lowest weight with 36.6% and 3.4%. Although from total area of 78580 hectares an optimal location area was selected 35988.71 hectares and highly suitable 6%, Suitable 40%, less suitable 44% and Unsuitable 10% areas were covered. The most potential of solar farm sites are found in the south, southeast and northeast parts of the study area. Those potential zone contributes to ll the energy gap between the demand and supply of the country when there is a shortage of rainfall through the country and shortage of water at the dam.