Network Analysis of Frac Sand Mines

Goals and Objectives of Network Analysis:

For this section of the frac sand mining project, we were tasked with performing network analysis on the routes between the sand mines and the rail terminals that transport the sand to frac mining locations. Network analysis is a function performed in a GIS program that analyzes the path of travel in a certain environment. This tool will assist in finding the most direct route from the sand mine to the rail terminal, which will, in turn, provide a value of the total distance from mine to terminal. Using this value, a road cost figure can be generated for each county, detailing the expenses that particular county will incur due to the increased heavy truck traffic transporting the sand to the rail terminal. The goal of this section of the project will be to produce a cost figure for each county that can be used for planning purposes for the sand mines that are already operational, and serve as an estimate for each mine that is a proposed site.

Data Sets and Sources:

In order to perform network analysis accurately, a number of data sets are required. The first was already available, as it was the geocoded sand mine feature class generated in the previous section of the project. In addition to these, rail terminal locations within the state of Wisconsin and a network dataset was necessary to complete the analysis. The rail terminal feature class was provided to us in the form of a shapefile, which was then imported into the geodatabase that had already been created to store data for this project. The network dataset, the North America streets dataset, was provided by ESRI with the purchase of ArcGIS. A network dataset is essential for the completion of network analysis, because this accurately models complex transportation networks, containing information such as street type, limitations on roads, possibilities of turns, centerlines, and other features necessary for navigation along a route. One of the goals of a network dataset is to provide the fastest possible route from getting from one point to another. Using the provided information in a network dataset results in a hierarchy of routes, such as highways being preferred to country roads, which assists in getting to the desired destination in the fastest time.

Methods:

The first step to conducting network analysis of the frac sand mines and the rail terminals was to add the necessary data to the map. This included the mining locations that were generated in the previous section of the project, the rail terminal feature class provided for this assignment, and the North American streets network dataset provided by ESRI. The next step was to use the Network Analysis tool to perform Closest Facility analysis, which determines the closest facility to an incident, and then assigns a route based on a specific cost. Inputting the correct feature for a facility and an incident is absolutely critical in this step. If the wrong feature was input, then the route would show the fastest route from the rail terminal to the closest sand mine, which is not the goal of this project. Inputting mines as facilities and rail terminals as incidents would produce this undesired result. Therefore, mines should be assigned as incidents, and rail terminals as facilities to produce the correct route.
Now that this method of network analysis was completed, ModelBuilder was then employed to perform the more advanced functions in one fluid motion. ModelBuilder is a tool available in ArcMap that allows the user to build their own parameters for a tool, and then execute it. This is much simpler than utilizing multiple tools in succession in order to get the same result. The same processes as the previous step are followed, except each function is added to the model. The desired function of creating a Closest Facility Layer is entered, as well as the North American Streets as the network dataset. From here, the mine and rail terminal feature classes were added to the model, and the closest facility function was calculated. Again, the mine feature class needed to be assigned as incidents and the rail terminals as facilities in order for the model to run in the desired way. This result was copied and exported as its own feature class, and then projected into UTM Zone 15N to convert the default units from decimal degrees to meters. A Wisconsin counties feature class was also added to be used later in the calculation of total routing miles per county. This feature class also required being converted into the UTM Zone 15N projection to have units consistent with the routes feature class. To get both the counties and route data into the same feature class, and, consequently the same table, Identity was performed to combine the two feature classes into one. Finally, Summarize Statistics was used to calculate to find the sum of the Shape_Length field in each of the counties that had a route going through them. An image of the final model can be viewed in Figure 1 below.

Figure 1: Model used to produce the combined feature class of routes and Wisconsin counties to be used to calculate the cost each county can expect to incur due to increase heavy truck traffic transporting sand to rail terminals.

The next step in calculating the cost to each county was to apply the values provided to the Shape_Length field for each county. It was estimated that each mile traveled by heavy truck would cost the county 2.2 cents, and each truck would make fifty trips from the sand mine to the rail terminal, and then another fifty trips back to the sand mine. Figure 2 demonstrates the equation used to figure this value.

Figure 2: Formula used to calculate cost per county of increased heavy truck traffic

To calculate the cost per county, the Shape_Length field was multiplied by 0.000621371 to convert the value from meters to miles. This number was then multiplied by 0.022 to determine the cost in dollars per trip of a truck going one way in the route. Finally, this number was then multiplied by 100 to factor in the total of 100 trips per road each truck will be taking from mine to rail terminal and back to the mine if 50 round trips are to be taken per year.

Results and Discussion:  

The impact of the increased heavy truck traffic on the roads per county is quite significant. Figure 3 below shows the table of results, including the road length in miles, cost per one directional trip, and the total cost of fifty round trips to and from the frac mines.

Figure 3: Table displaying cost each county can expect to incur because of heavy truck traffic transporting sand to rail terminals

The five counties that will have costs in excess of $1,000.00 are Chippewa, Eau Claire, La Crosse, Monroe, and Trempealeau Counties. At first glance, one could simply assume that these five counties have the greatest amount of frac sand mines present, and that is the reason why they will have to pay relatively more when compared to the other counties. However, this may not always be the case. Take La Crosse County, for example, Figure 4 shows a map of the routes from each frac sand mine to a rail terminal. La Crosse County has one mine operational within its borders, and yet it has the highest estimated cost of all of the counties involved. This is because of the rail terminal located in the county, and its proximity to neighboring Trempealeau, Monroe, and Jackson Counties, which have significant amounts of sand mines, but no rail terminal. That extra cost that La Crosse County has to bear is because of these counties transporting their sand to the rail terminal located there. Eau Claire County is another good example of this phenomenon. Eau Claire has four sand mines present within its borders, and yet it has the second highest cost to its roads. This is because of the large number of mines in neighboring Chippewa County, and the transportation cost associated with transporting the sand to the closes rail terminal, which is located in Eau Claire County.

Figure 4: Results of Network Analysis process showing routes from each frac sand mine to the closest rail terminal

These findings have significant impact on each county. Even if a county does not have a significant amount of frac sand mines, they still can incur a high amount of costs associated with the increased heavy truck travel throughout the county to get to the closest rail terminal. Figure 5 below shows a chloropleth map of the cost that each county that has a route going through it can expect to pay for maintenance of its roads. Clearly visible are the five counties mentioned earlier: Chippewa, Eau Claire, Trempealeau, La Crosse, and Monroe, which are represented most vividly in the chart found in Figure 6.

Figure 5: Expected cost each county will incur due to increased heavy truck traffic



Figure 6: Percentage of total cost each county can expect due to increased heavy truck traffic

 

Conclusion:

The results of this section of the frac sand mine project undoubtedly confirm that there will be costs associated with the increase in heavy truck traffic going back and forth from the sand mines to the closest rail terminal. Network analysis was used to compute these costs, and from here, an actual dollar amount can be given to each county so that they can figure that into their annual budget. There is one issue here, however. Most of the counties will be able to use increased revenue from workers on the sand mines spending their money in the county to lessen the impact of road maintenance. However, counties like Eau Claire and La Crosse, which have a relatively low amount of mines when compared to the other neighboring counties, will not have as much tax revenue to work with. The only reason they are involved in the equation at all is because of the presence of a rail terminal within the county. It seems almost unfair that these two counties have to bear the largest percentages of total cost per route, but it is a factor that needs to be accepted. There are simply not as many rail terminals located in the western part of Wisconsin, and that region is where the best sand mines are located. The costs each of these counties have to pay is an unfortunate consequence of having the only rail terminals in that part of the state, and that phenomenon is one thing that can be revealed by the use of network analysis in this exercise.