Network Analysis

Background

One negative impact that sand mining can have, is deterioration of roads from the heavy trucks carrying frac sand.  For numerous counties, this is a serious concern as it could lead to higher road maintenance costs.  The additional costs to the counties could potentially be avoided through road upgrade maintenance agreements (RUMA) between sand mining companies and county governments (Hart, Adams, and Schwartz, 2013)

Figure 1.  On the left is shown elements of a good road use agreement (Hart, Adams, and Schwartz, 2013) 

 

Part of understanding costs, is understanding the routes being used to transport sand from mines to rail terminals.  This can be done using network analysis, a set of tools that can be used to calculate the best routes in terms of distance, expense, and time.  Utilizing network analysis is often referred to as logistics, and can help companies and individuals transport more efficiently.

 

 

Goals and Objectives

In this lab, we will perform a network analysis to find the routes that would be used to transport sand from mines to rail terminals.  We will then use this estimation to produce an estimate of how much counties would have to spend on road maintenance annually.  We will be using theoretical numbers provided by our professor on how often the roads are used and what the average cost of a truck driving over a section of road is.  Because of this, our results our not intended to be an accurate representation of actual costs.  They are meant to show the process by which you could use network analysis to generate these estimates.

The general work flow for this lab is shown in Figure 2.  We prepared data by removing mines within 1.5 km of a rail road.  This was done using Pyscriptor, (code can be viewed here).  Then we calculated routes based on closest facilities using network analysis and built a model to incorporate theoretical costs given to us by our professor and calculated our final answer.

 

Methods

 

Step 1.  Prepare data using queries

Using a Python Script, mines that were within 1.5 kilometers of a rail terminal were eliminated.  This was done, because these mines would likely load the sand directly unto trains, and therefore would not likely have a great impact on the road.  The feature class produced from the python script was used in the following steps.  The mine feature class produced can be seen in Figure 3 below.

Figure 3.  Map of locations of Sand Mines and Rail Terminals in Wisconsin.  Mines within 1.5 km of rail terminals not shown.

Step 2.  Calculate routes based on closest facilities

 

After enabling Network Analysis, Esri street data was imported.  Then using the Closest Facility tool we calculated the closest routes from each sand mine to a rail terminal.  The mines were loaded as incidents and the terminals were loaded as facilities. The resulting routes can be seen in Figure 4.

Figure 4.  Map of routes from sand mines to rail terminals derived using network analysis Closest Facility tool

 



Step 3. Generate hypothetical estimates of cost per county

 

Our professor gave us the following theoretical information to use to calculate the annual cost per county that would result from hauling sand on the roads.

• Assume each sand mine takes 50 truck trips per year to the rail terminal, and that the truck has to return to the sand mine after each trip
• The hypothetical costs per truck mile is 2.2 cents

Model builder was used to organize the tools needed to perform the spatial and data analysis needed for this calculation.  The model can be seen in Figure 5.  A description of the process is written in numbered steps below:

 

1.  Used Closest Facility tool in Network Analysis to solve for routes.  Mines were loaded as incidents and rail terminals were loaded as facilities

 

2.  The routes were selected and copied into a geodatabase.

 

3.  The routes were projected in NAD 1983 Wisconsin TM (Meters) so that distance could be calculated

 

4.  The Intersect tool was used for the routes and the Wisconsin counties that were in the same projection.

 

5.  A field was added to calculate the distance of the routes in miles

 

6.  Summary statistics were used to get the sum of routes in each county

 

7.  A new field was added and calculated to account for the number of truck trips annually

 

8.  A new field was added and calculated to account for the cost of these truck trips

Figure 5.  Model used to organize tools and estimate the hypothetical annual cost per county

Results

The hypothetical annual costs of sand mining associated with road maintenance varied greatly between different counties.  Some counties with sand mines, like Buffalo and La Crosse have less than 50 dollars in associated road costs, while other counties like Barron and Chippewa have over 400 dollars in associated road costs.  The differences between counties can be seen in the map in Figure 6 and the graph in Figure 7.

Figure 6.  Map showing hypothetical annual costs of road maintenance due to hauling sand in various Wisconsin Counties

Figure 7.  Graph showing hypothetical annual costs of road maintenance due to hauling sand in various Wisconsin Counties

 

 

Discussion

 

Network analysis is incredibly useful, and can be used to make smart and efficient choices in transportation.  They can also be useful in helping determine the impact that certain businesses might have on local roadways, and can be used to make sure that they are held accountable for this.

Model building was also useful in this activity, in that it helped organize the tools used, and could be used again even if certain parameters or data was altered.  For instance, it would not be too difficult if we were to receive an updated list of mine locations, to simply adjust the model for this and run it again.  It is a great time saver when you are performing a series of operations that may need to be run again with new and updated information.

 

Conclusion

 

In this lab we prepared data using queries, we calculated routes using network analysis, and we build a model to derive hypothetical annual road maintenance costs associated with sand mining.  We went through the steps that could be used when performing research on the associated costs of sand mining.  Actual analysis on these costs are very important for counties who may need to negotiate road upgrade maintenance agreements (RUMA) with sand mining companies. 

 

Sources

 

Data

Mine Locations: Wisconsin DNR

Railroad Terminals: Federal Railroad Administration

Basemaps: Esri

Streets for Network Analysis: Esri

Wisconsin Counties Feature Class: Trempealeau county geodatabase

 

Background Information

Hart, M. V., Adams, T., & Schwartz, A. Transportation Impacts of Frac Sand Mining in the MAFC Region: , CFIRE.