**2. Multi-factor geospatial method**

As a new vehicle system, the current number of public EV charging stations is insufficient to meet the demand, in contrast to the vast number of gas stations available. The average range of EVs per charge is much less than that of conventional gasoline vehicles. This lack of infrastructure presents a challenge for the widespread adoption of electric vehicles, and also leads to range anxiety and limited access to charging stations. Range anxiety (driver's fear that EV has insufficient battery charge to cover the road distance before the destination is reached) prevents this technology from being adopted rapidly by the traveling public. EVs can provide up to 100 miles of distance on a fully charged battery [17]. According to the U.S. Department of Transportation Federal Highway Administration, 100 miles is ample for 90% of trips generated by the traveling public in the United States [18]. Longer trips would require proper planning to ensure the availability of EV charging stations to one's destination. Range anxiety is a common feeling among EV owners that they may be unable to reach their destination before running out of battery power. This phenomenon is enhanced by driving habits, excessive speed, and weather conditions where battery power is used to heat or cool the cabin of the vehicle [19].

In this research, we used a GIS suitability model to evaluate multiple factors and determine the optimal locations for charging stations. This GIS model intakes input demand factors through a series of spatial analysis procedures to generate an output demand factor grid. To run the model and perform the spatial analysis properly, all the input demand factors need to be collected and prepared or spatially interpolated to the right spatial format. Geospatial analysis of these input demand factors are an important step in modeling. A series of data sets that characterize the physical and urban features of the study area are analyzed first to identify trip attractions by the driving public. Highly desirable and most visited locations are usually identified as employment centers, shopping districts, major transportation hubs, public facilities, and recreational areas. Urban and environmental factors can have negative or positive factors to the driving public. A comprehensive analysis of these location factors is performed and a scoring system is developed to evaluate the contributing impact of each factor on the charging station. To prevent any of the input factors from dominating the scoring system, a uniform weighted system is developed to distribute weight to each input demand factor. This systematic scoring method provided the necessary input parameters that were utilized during the analysis to determine suitable and unsuitable areas for EV charging stations.

Input demand factor is evaluated through the use of a GIS suitability model. The suitability model intakes input demand factor through a series of spatial analysis

**Figure 1.** *Input demand factor model.*

procedures to generate an output demand factor grid. **Figure 1** illustrates each spatial analysis step used in the GIS suitability model.

The first step is to intersect the input demand factor layer with a walkability grid. A walkability grid indicates how far a commuter would be willing to walk to a particular destination (charging station). A 0.5 × 0.5 mile walkability grid was created in this research as a comfortable walking distance (0.25 mile) for commuters to the charging stations [20].

The second step is to perform a quantile classification to produce a master grid layer. The walkability grid created in step one was then applied to intersect each input demand factor for the study area. The ID of the walkability grid is assigned to associate each individual input demand factor with the walkability grid. A summary statistics is generated using the attribute table of the newly generated input demand factor grid. Based on the summary statistics, a new composite index score is calculated for each grid. This procedure took each individual grid square and tallied up all the input demand factor index scores to produce a composite index score for that particular grid square. For example, if a single grid square contained the following land use types: residential (−4), commercial (+5), and mixed use (+5), the summary statistics would calculate the composite output score to be −4 + 5 + 5 = 6. This composite index score of 6 would then be associated with its grid's id. Each input demand factor grid is added to the intersect tool in ArcMap to produce a master grid layer. This process allowed for each individual index demand factor layer score generated by the model to be combined into one layer for further analysis. A new attribute field called 'factor score' is added to the master grid layer to hold the initial calculated index demand factor score. This calculation is conducted using ArcMap's field calculator which allowed for an equation to be written that summarizes all attribute fields representing each input demand factor layer's composite scores.

The resulting summary statistics table is then joined back to the walkability grid to produce a final output demand factor grid. The final output demand factor grid is then exported out to a new feature class and stored in a file geodatabase. This spatial procedure is then repeated again for each individual input demand factor.

*Using Geospatial Analysis to Assist with Clean Vehicle Infrastructure DOI: http://dx.doi.org/10.5772/intechopen.110864*
