Heavy Axle Study: A Statistical Assessment of Railroad Infrastructure
By Charles Leighty
Abstract
The Class 1 railroads like Norfolk-Southern and CSX use heavier cars such as the 286,000-pound or 315,000-pound
gross car weight while most short line railroad are only prepared to handle the 263,000 pound gross car weight.
The ratings for most bridge and track structure of short line railroads are not known. Short lines will accept
cars from Class 1 railroad, but typically will not weigh the cars until the cars have traveled half their line.
In an economic interest to the transportation infrastructure in Pennsylvania, The Bureau of Rail, Freight, Port,
and Waterways in PennDOT has support this study to find the cost for statewide upgrade to handle either the 286,000lb
or the 315,000 gross car weight.
Pennsylvania contains approximately 2000 short line railroad bridges and 3000 miles of short line railroad track.
It would be a costly task to rate and inspect every bridge and mile of track in the state, so a statistical study
was developed to attain reasonably accurate results with available resources. The statistical study samples approximately
25 bridges in the state and the track sections are adjacent to the bridge structures. Information on bridges and
track were collected from the short line railroads to better define the population. Different random sampling
procedures were explored with stratified random sampling being chosen as the most accurate method for the sample
to represent the population.
After the sample was determined, the bridges and track were inspected. The inspection data was used to create
models and evaluate the bridges according to AREA specifications. Of the 25 bridges, four were found to be deficient
under the 315,000-pound car loadings. Only 2 of these bridges passed under the standard 263,000-pound car loading
at full speed. 11 out of the 25 bridges could not withstand the AREA standard loading at full speed.
Proposed repairs to accommodate the 315,000-pound car were developed by adding plates or replacing ties where necessary.
The repair costs will be determined by comparing estimates from bridge contractors that the short lines themselves
use. The average cost for the bridges that needed repairs was approximately $7000. The sample cost will be extrapolated
to the population for the expected upgrade statewide cost. The confidence width for the statistical study is expected
to 610% with a 90% confidence width.