Post 5: Network Analysis

Background
According to a White Paper study of the transportation impacts of frac sand mining, the use of frac sand has exploded with the development and widespread use of the frac sand mining technique for oil and natural gas. With this demand, a readily supply of sand is necessary to maintain the process. This demand has centered on the west central frac sand mines of Wisconsin. This increased demand has left many communities concerned on the cost forces on their communities which contain the frac sand mines by the process. Over the course of a mines life, trucks may make hundreds of tips to and from the mine for transport and delivery. This exercise was a continuation of the ongoing frac sand project completed in this course. In it an estimation of the costs counties and communities will be forced to pay as a result of frac sand mining. What's critical is that this is not a scientific or credible case study. The true cost of transport for trucks is not known, nor is the number of truck trips to and from each mine known. Thus, an estimate will be made based on several arbitrary values.
Data

• County, railroad, railroad terminals, and mine location datasets provided by the Wisconsin DNR.
• Network Dataset provided by ESRI street map USA.

Methods

First, a python script was generated to select only the active mines within the state Wisconsin that do not have a rail loading station on site or are within 1.5 kilometers of a railway (Post 2: Script 2). These mines will be forced to utilize public roadways to transport mines to available rail terminal with loading stations. In addition, this python script created a feature class from the rail terminals feature class of only the terminals with viable loading stations for trucks. Then, the address field was removed from the mines feature class to prevent an error from happening during later analysis. From here a model was generated in order to completely format  and perform the necessary network analysis, field creation & computation,  and any other necessary procession required in order to properly analyze the network data (Figure 1). The steps in the model were as follows:

• A Closest Facility Layer was added, with the input being set to the streets. Travel was also set to facility
• An Add Locations tool was next, with the input locations being the mine selection feature class. These were set to incidents to make them the necessary start of any route
• Add Locations was added again, with the rail terminal selection being the input set as the facilities
• A Select data tool was added to select the newly generated route
• A Copy features toll was utilized to create a feature class of the route data
• A Project tool was utilized to convert the route into a projected coordinate system with a linear unit in feet. This was necessary for a later Summarize tool to properly summarize the distance trucks traveled on roads later.
• The Counties feature class was likewise projected into the same coordinate system.
• An Intersect tool was utilized between the county and routes feature classes
• A Summarize Statistics tool was used to calculate the quantitative total distance of the routes in each county.
• The Add Field tool was utilized to create a field for estimated cost in the counties feature class
• Calculate field was utilized to estimate the cost each county would be required to pay annually for frac trucks. This based on several estimates. It was estimated that trucks would take 50 to the terminal from the mine and 50 trips back to the mine annually, with each mile a truck travels costing a county $.022. It was also multiplied by 5280 to generate a cost per mile, rather than a cost per foot as a result of the summarize tool generating the distance in feet. The equation used was as follows:
• cost = (summarized distanced traveled in each each county) * .022 * 100 * 5280.
• This resulting data table was joined to the County Boundaries feature class, so the cost per county could be properly displayed.
• The resulting data and feature classes were utilized to create a map and data table which displayed the estimated cost on each county.

Figure 1: The model utilized to generate the routing from each frac sand mine to the closest available loading terminal and calculate the cost of transport each county would be forced to pay.

Results
By looking at the data (Figure 2), it can be seen that most counties accrue little to no cost as a result

Figure 2: A data table 

displaying the total length

of routes and cost accrued

by truck transport in each 

country.

of frac mining. The minimum cost is $0 and the average is only roughly $30. However, the maximum is $636 and the standard deviation is roughly $100. These highest costs are accrued by counties located in the northern portion of the collection of mines present in the western portion of the state (Figure 3). This means they have the greatest cost as a result of overlapping mines and routes within the county. In contrast, counties located outside of this clustering of mines accrue little to no cost as a result of only one or no mines present in the county.

Figure 3: A map displaying the routing for each mine to the closest viable rail terminal and the cost of transport each county will be annually forced to pay. This data is estimated arbitrarily and should not be utilized for an actual case study.

Conclusion
While this process has allowed for the creation of an estimate of costs, it is partially incomplete. As a result of the arbitrarily estimated cost per mile and number of trips each trucks take annually to and from each mine, the true cost is still unknown. However, this project allowed for the creation of a model generating the routes and estimated costs. In the future, this or a similar process could actually be utilized in for a project similar to this one.

Sources
Hart, M., Adams, T., Schwartz, A. (2013). Transportation Impacts of Frac Sand Mining in the MAFC Region: Chippewa County Case Study. In White Paper Series: 2013. Retrieved 4/20/2017, from http://midamericafreight.org/wp-content/uploads/FracSandWhitePaperDRAFT.pdf

Hupy, C. (2017). Exercise 7: Network Analysis Part 1-Data Preparation. Eau Claire, WI.

Hupy, C. (2017). Exercise 7: Network Analysis. Eau Claire, WI.

Network Dataset provided by ESRI street map USA.

County, railroad, railroad terminals, and mine location shapefiles provided by the Wisconsin DNR.