CSX oil train derailment

[Euclid]

Apparently there is a measurement of tank car safety called the “Conditional Probability of Release” (CPR).  Here are two references to it:

https://www.fra.dot.gov/eLib/Details/L02883

 

http://railtec.illinois.edu/articles/Files/Journal%20Articles/2014/Liu%20et%20al%202014%20JHM%20Multiple%20Car%20Release.pdf

 

Here is the Greenbrier “Tank Car of the Future.”  The document says that the new tank car is 8 times safer than the DOT-111 tank car, and 2 times safer than the CPC-1232 tank car which the industry has been building on good faith sans forthcoming Federal regulations. 

http://www.gbrx.com/phmsa.php

From the article:

“We are confident that the design features identified as ‘Option 2’ in the NPRM will lead to safer transport of flammable liquids at any speed," said Greg Saxton, ‎Senior Vice President & Chief Engineer at Greenbrier.”

[Euclid]

Here is how Transportation Secretary Anthony Foxx described measurement of tank car safety:

 “The American people must have confidence that when hazardous materials are transported through their communities, we’ve done everything in our power to make that train as safe as possible,” Mr. Foxx said in a written statement Thursday.

 

It sounds like the popular grade crossing solution:  “The only safe grade crossing is a closed grade crossing.”

[blue streak 1]

What would be interesting would be a relative study of derailments in general. 

1.  One would be derailments per million car miles.

2.  Then a breakdown of derailments per million tank car miles.

Using mainly item #2 the relative derailment rate of each of the RRs would be very informative.  One has to wonder if the deferred maintenance of CSX under Snow is still haunting CSX.  Can it be that even though CSX seems to be keeping their physical plant in better shape now that there are covered over hidden problems that are not apparent ?  Granted CSX runs more of its trains along river banks that can cause problems.

[dehusman]

blue streak 1

What would be interesting would be a relative study of derailments in general. 

1.  One would be derailments per million car miles.

2.  Then a breakdown of derailments per million tank car miles.

Using mainly item #2 the relative derailment rate of each of the RRs would be very informative. 

Not as much as one would think.  A comparison of each railroad to itself over time would be better, especially if you are using a "million car mile" measure.  Since more derailments occur in yards and terminals, that favors roads with longer hauls.  Doing it by tank car miles would seriously favor the western roads that handle the Gulf coast and generate thousands of long haul tank car loads a year.

If Railroad A has an average 1000 mile haul, generates 100 car loads and derails 2 cars, it has a derailment every 50,000 miles.  If railroad B averages a 250 mile haul, generates 100 car loads and derails 1 car it has a derailment every 25,000 miles.  Even though railroad A has twice the derailments, it has double the performance of railroad B purely driven by the route miles.

In a similar vein, if you look at the AAR/STB velocity stats, the Western railroads are almost always faster than the eastern roads, why?  Different route mix.

[blue streak 1]

Following link has a good report.  Look at the picture and blow it up to see the rail.  Great reconstruction.  Trains should have used it as well. 

http://www.rtands.com/index.php/track-structure/ballast-ties-rail/fra-points-finger-at-broken-rail-pushes-for-improved-rail-inspection-following-wva-derailment.html?channel=

 

[BaltACD]

blue streak 1

Following link has a good report.  Look at the picture and blow it up to see the rail.  Great reconstruction.  Trains should have used it as well. 

http://www.rtands.com/index.php/track-structure/ballast-ties-rail/fra-points-finger-at-broken-rail-pushes-for-improved-rail-inspection-following-wva-derailment.html?channel=

Since the NTSB was stating the 'break' was visible on rail scans operated by both Sperry and CSX on separate scans in the months preceeding the derailment - just how 'overt' was the evidence of the break?  

Personal question - is a scan performed by the rail manufacturers as the rails complete the manufacturing process to present a baseline?