There is a current urgency to reduce fossil fuel emissions and provide economical and environmentally sustainable energy to maintain and enhance our standard of living (WBGU, 2009; Pearce, 2000; Metz et al., 2007). The current energy resource relies on burning fossil fuels such as coal, oil and natural gases, which release carbon dioxide and contributes to global warming (United Nations, 1992). As a result, there is a movement towards renewable energy resources, including solar energy, which unlike fossil fuels, is renewable and undepletable (Nguyen & Pearce 2010). Solar photovoltaic (PV) is a technology that converts sunlight into electricity (Cansia, 2011), and is considered as one of the leading energy sources of the future (Mondino et al., 2013), as it is the most technologically robust, scalable, geographically disperse and has enormous potential for providing a sustainable source of energy (Nguyen & Pearce, 2010). The earth is experiencing climate change, thus implementing technologies that provide the growing population with adequate and renewable energy is an important issue and challenge.
There is great interest in building large-scale photovoltaic solar farms and many studies have been performed in order to site the optimal locations of these farms. Studies have shown that it can be a rather complex process when evaluating a potential site because developers must take into account a number of environmental, social, and economic factors (Macknick et al., 2014). Solar insolation varies by atmospheric conditions such as cloud cover, topography, latitude and slope, as well as by ecological processes such as snow melting, evapotranspiration and human activities (Duffie & Beckman, 1991; Hofierka & Suri, 2002). It is important to understand the factors, which impact PV siting, to ensure longevity of the infrastructure and minimize environmental impacts.
Studies have demonstrated that many of these factors that influence the siting of a PV farm are categorized as a constraint, meaning a farm can either be built there or it cannot. For example, PV farms cannot be constructed on highly valuable agricultural land (Mondino, Fabrizio, & Chiabrando, 2013; Watson & Hudson, 2015). Other important areas that are considered constraints in the studies are conservation/protected lands (national and provincial parks), historically important areas, wildlife and environmentally sensitive habitats, residential areas, and water bodies (lakes, streams, etc.) (Carrion et al., 2008; Macknick et al., 2014; Mondino et al., 2013; Sanchez-Lozano, Antunes, Garcia-Cascales, & Dias, 2014; Watson & Hudson, 2015). Other studies have also included different criteria factors that influence the siting of a PV farm, meaning there are certain conditions for building a farm. These criteria factors include slope, aspect, proximity to urban areas and transmission lines, landuse and land ownership, and soil type (Carrion et al., 2008; Macknick et al., 2014; Mondino et al., 2013; Nguyen & Pearce, 2010; Sanchez-Lozano et al., 2014; Watson & Hudson, 2015). These factors have preferable conditions for siting PV farms. Macknick et al. (2014) reported that siting requires slope of the land to be less than 5% for there to be development of a PV farm. Mondino et al. (2013) stated that if farm locations were to be constructed on steeper slopes, the slopes need to be south-facing.
Applications involving the proper siting of large solar infrastructure require the ability to spatially analyze a region in order to find the optimal location (Brewer et. al, 2015). Traditionally, a model is developed based on pre-defined objectives, and sites are analyzed numerically based on these models. The advance of aerial photography, and more recently satellite-based remote sensing, has made analysis of large regions far less costly and more accurate. Therefore, the application of GIS in both public and private infrastructure development has become very beneficial. Most large companies involved in these businesses will have entire sectors dedicated to GIS Analysis, and the study of spatial trends in development has been shown to improve the efficiency of site planning for PV farms in particular (Wu et al., 2011; Sanchez-Lozano et al., 2014).
The importance of siting infrastructure properly is even more prevalent in the renewable energy sector, due to many political, social, and physical constraints involved in their development. For example, renewable energy tends to require a much larger area per MW produced than nuclear powered facilities, however it is more advantageous in terms of CO2 reduction per MW produced (Karaveli et al., 2015). Although it has the extreme benefit of producing little to no emissions, renewable power tends to gain a negative stigma in some rural areas due to socio-economic factors driving the beliefs of people in these areas. On top of this, developers are constantly constrained by physical barriers such as rough terrain, availability of land, and the presence of conservation lands. Due to the great amount of potential factors that can influence the siting of renewable energy infrastructure, the use of GIS is absolutely necessary in order to guide the developers in finding sites that will be cost-effective to develop (Baltas & Dervos, 2012).
Currently there is no standardized method for siting these solar farms using GIS, and many of the approaches in previous studies use a multi-criteria evaluation with widely varying criteria and constraints. There is an urge to standardize PV farm siting, as well as determine which factors are the most influential for PV farm locations.
This research will focus solely on the solar renewable energy sector, specifically the siting of relatively large photovoltaic energy farms. Typically, a top-down approach is used in which the developer identifies the factors that affect their cost the most, then those factors are modelled into a relationship which can be used throughout a region to quickly assess site effectiveness and determine the optimal location (Sanchez-Lozano et al., 2014). Contrarily, this research will explore a new method in which the existing PV farm sites will be used as the basis for determining which factors are the most relevant in determining the locations of these sites. A wide range of both physical, economic, and social factors will be examined using spatial regression analysis with the existing sites to determine if any significant relationships exist. If a relationship is determined, a model will be developed to predict the best locations for future PV developments.
Purpose of Research
The purpose of this research is to develop a GIS-based regression model that uses statistical analysis to determine the most spatially significant factors affecting the siting of photovoltaic solar farms in Ontario.
Objective One: To define the physical, social and spatial factors related to PV farm siting in Ontario.
Objective Two: To develop a GIS-based regression model that assesses physical and socio-economic factors which contribute to the siting of PV farms.
Objective Three: To validate the model by statistically assessing the correlation between the determined factors and existing PV sites.
Objective Four: To evaluate the strengths and limitations of the regression model.
Objective Five: To create a suitability map providing a visualization of the regression model.