The goal of this exercise was to use network analysis to create trucking routes from sand mines in southwest Wisconsin to the railroad terminal nearest to each mine. This data was then used to predict yearly costs to maintain county roads with the additional truck traffic in mind. The objectives of this exercise were as follows:
- Use the Network Analysis extension in ArcMap to create a closest facility network.
- Build a model to calculate the same network.
- Calculate the cost of trucking travel on the roads (by county)
Methods
Using our mine data from the previous exercise and railroad terminal data provided by our professor, it was a simple task to create a routing network for the sand mine trucks. First though, it was required to query out the sand mines that were close enough to a railroad track that it could utilize a spur to transport it, instead of using trucks to move the sand to the nearest terminal. To do this, I made a Select By Attribute query to select all sand mines within 100 meters of a railroad, and then used aerial imagery to confirm that these mines were not to be included. I then made a feature class of the rest of the mines to be used in the network analysis. The closest facility route creates routes from a collection of instances (in this case, the mines), to a collection of facilities (the railroad terminals). We used ESRI street map USA as the Network Dataset, or the streets to be used in the routing of the trucks.
Once the route was created, I had to organize the segments of the route based on the counties that they go through. To do this, I simply used the Intersect tool with a Wisconsin Counties feature class, and the routesCopy feature class itself. This produces a table of every route segment that runs through each county, which was quite a lengthy list. I then summarized the total length of routing through each county, creating a much smaller table that is easier to read and interpret. This was also an opportune time to add another field to the table, to display the total cost per county. Our professor gave us the value of 2.2 cents per mile, with each mine sending 50 trucks on the route per year. The equation for this field was then ([SUM_totalLength]*.022)*50. Figure 2 below displays the model and flow of this entire process.
 |
| Figure 1: A model for the creation of the closest facility network. The blue ovals are the features that were used with the tools, the yellow boxes are the tools used, and the green ovals are the feature classes produced with the tools. It is easy to see the flow from the creation of the layer, addition of locations, and then creation and exporting of the feature class to a database. |
 |
| Figure 2: A model for the creation of the summary table. I used the Intersect tool to join the routes with the counties, then summarized the route data by county, and then created another field to display the yearly cost per county. |
Results
 |
| Figure 3: The summarized table, sorted with the most expense at the top. |
 |
| Figure 4: A mines, terminals, and truck routes in western Wisconsin. |
Conclusion
I was surprised to discover how much all these mines would affect the individual counties of Wisconsin. It is easy to see where all this additional traffic will be, and it is interesting to note the significant clustering of the mines in Wisconsin. Perhaps with this new need for more railroad terminals, another significant rise will occur in this part of the state. It is clear that Western Wisconsin has far less terminals than Eastern Wisconsin. With all of this new activity it wouldn't surprise me if we started to see a new pattern take form, to minimize the trucking traffic through each of these counties.
