Network Analysis: Cost of Sand Mining on Local Roads

Goals and Objectives

The goal of this lab was to become acquainted with the Network Analysis tool, and use it to perform analysis of the impact that sand mining has on the transportation system here in Wisconsin. The general workflow I followed is as follows.

1. Determine which rail terminal each mine will travel to in order to get the sand to a rail car.
2. Determine the most efficient route from the mine to the terminal.
3. Calculate the length of the route by county.
4. Estimate the costs each county will incur for road maintenance due to heavy sand truck traffic.

General Methods

The first step was to use a Python script to select all of the mines that were active, did not have a rail loading station included, and was not within 1.5km of any railroad. This script can be seen in my posting from November 11th. Using this new dataset of mine locations, I then had to add a streets network dataset, and all of the feature classes participating in it. I then added a rail terminals feature class to the map, and remove features that had an air terminal facility by selecting the desired features, and creating a layer from the selection. Then, I used the Network Analyst: closest facility tool to determine routes that trucks shipping sand would take to get to the nearest rail terminal. We then used model builder to visualize our workflow, and make it a reusable process. I essentially went through the same process I just explained, and exported the resulting feature class to a geodatabase. The model is shown below. The next step was to calculate the total road length covered per county. I did this by intersecting the trucking routes with the county boundaries, and summarizing (giving me the total length) based on the county field. From there, I created a new field called COST, and calculated its values based on the assumptions that each sand mine takes 50 truck trips per year to the rail terminal, and costs 2.2 cents per mile of wear and tear on the roads. The equation is as follows: [SUM_Shape_length] * 50 / 1609.344 * .022. Finally, I mapped my results. They are shown below.

Results

This image shows the process I went through in using network analysis to calculate routes
likely taken by trucks shipping frack sand.

This image shows the process taken to calculate the costs incurred by sand-moving trucks

This image shows my final results. Counties with higher impact by sand mining transportation
operations are shown in darker colors.

This image shows the table with total length of roads (in meters) affected by sand mine
transportation, and the cost associated with these values.

Discussion

After analyzing the results of this excersise, I am quite surprised at how low the annual costs on the roads are. It is important to note that the assumptions talked about in previous sections (being that each mine takes 50 truck trips per year, and that the cost per truck mile is 2.2cents) is hypothetical. More testing and analysis would be needed in order to make an accurate claim about the monetary implications that frac sand mining has on the transportation system here in WI.

Another important issue with the assumptions is that instead of calculating for 50 trips, I should have calculated for 100, accounting for the trip back from the rail terminal.

Conclusion

 

This lab was a good introduction to using network analysis, and a good way to familiarize myself with some its possible applications. I think that network analysis is one of the most practical tools that I've used so far in GIS, and I'm sure that it is a very popular area because of its use in routing. It will be a good tool to be familiar with in the future.