I recently printed out a prepublication copy (uncorrected proofs) of the report Long Freight Trains: Insuring Safe Operations, Mitigating Adverse Impacts submitted by the “Committee on the Impact of Trains Longer Than 7,500 Feet.” What a name for a research group. Sounds like something created by Congress, and it was.
I took my seat at my desk, committed to the task of reading each of the 106 pages (including disclaimers, and the list of abbreviations) if it took all day/night, but 90 minutes later after finishing the page numbered “30,” I stopped, stood up, and resumed my inquiries into October baseball.
I was disappointed.
I was not disappointed by the baseball, where I happily watched the elimination of the Houston Astros, a team that cheated its way through several post-seasons not so long ago and should be prohibited from post-season play for a reasonable length of time—say 75 years.
I was disappointed in the committee’s report, which was not able to establish—per ton-mile, car-mile, train-mile or other denominator common to all manifest (mixed as opposed to unit) freight trains—a causal connection between the frequency, the rate or the severity of derailments and train length.
The committee reported that it “asked the Class I railroads, through the Association of American Railroads (AAR), to provide data on their train operations with sufficient detail to ascertain train type and length for the purpose of more granular assessments of the derailment records. However, restrictive conditions on the data’s availability and use, including a high degree of data aggregation and preapproval of the analytic methods to be used, foreclosed this option.”
As I understand it, and I looked it up myself, this study (download below) was authorized by Congressional approval. The Secretary of Transportation, charged by Congress, authorized the Federal Railroad Administration, as the DOT’s agent, to execute and sponsor this study, undertaken by the Transportation Research Board, which convened the Committee of the Impact of Trains Longer Than 7,500 Feet. What a mouthful, but the path is not difficult to follow: Congress enacts, Secretary of Transportation executes through coordination of FRA and the National Academies, which sponsor the formation of a study group or committee.
The committee then contacts AAR (among others) to gather the information necessary to making its determination about the relative operating safety of “long trains.” AAR decides that a portion of the information critical to the committee is proprietary and applies conditions so restrictive to the committee’s use and analysis as to make it, uh … problematic for the committee to determine conclusively if the frequency of the accidents involving manifest freight trains is related to, linked to, a function of and/or determined by train length.
The six Class I railroads, owners and archivists of the granular data, asserted in their 2023 presentations to the committee that “the operation of longer trains should result in safer train operations and fewer derailments overall. The six railroads maintained that longer trains result in fewer trains in total, and therefore fewer opportunities for derailments.” (See p. 16).
Apparently, none of the Big Six shared the data, or methodology used by Dr. Pangloss to arrive at this best of all possible conclusions in this best of all possible worlds. (Editor’s Note: Dr. Pangloss is a fictional character, the pedantic and unfailingly optimistic tutor of Candide, the protagonist of Voltaire’s 1759 novel Candide, a satire on philosophical optimism. The name Pangloss—from the Greek elements pan-, “all,” and glōssa, “tongue”—suggests glibness and garrulousness. A barbed caricature of the German philosopher and mathematician G.W. Leibniz and his followers, Dr. Pangloss has become a symbol of foolhardy optimism. Used as a noun, it means “one who is optimistic regardless of the circumstances. As an adjective, “panglossian,” it means “Blindly or unreasonably optimistic.” – William C. Vantuono)
“Should result in fewer derailments”? Would it be rude of me to point out that the assertion from the Big 6 assumes as fact exactly what must be proven—that derailments and train length are unrelated?I certainly hope not, but if Dr. Pangloss is offended, I can live with that.
The Big 6 also claim that “the frequency of equipment-caused and track-caused derailments should be unaffected by train length.” While these representatives didn’t provide the data necessary for the committee to make a definitive determination, let’s see what the publicly available data at FRA’s Safety Data website does indicate. Between 2017 and 2023, main line train-miles (total train-miles minus yard switching miles) operated by all freight railroads declined 15%. Did the rate of main line train-mile accidents remain the same, decrease, or increase? It increased 9% to 0.926 per million train-miles. The number of main line derailments increased 1% to 1,298.
Let’s go back to the data and sort it for the Class I’s. Because we’re just sampling, we’ll look at the four largest of the Big Six—BNSF, CSX, Norfolk Southern and Union Pacific, which accounted for approximately 66% of main line train-miles in 2023. So, for the years 2017 and 2023, here are the rates of change. It should be noted that accident rates and accident numbers fluctuate over the seven-year period, and the change between 2017 and 2023 is just that—the changes between two specific years:
In general, reportable accident numbers track closely with total train-miles, leaving accident rates relatively unchanged, except for Union Pacific. Not a bad overall record, but it tells us nothing about train lengths. Still, one might expect, if longer trains create a safer operating environment, the decline in accident rates would significantly outpace the decline in train-miles.
We cannot make any determination regarding train length impact on accidents and accident rates. Not yet. We need Dr. Pangloss to tell us the measures common to long and short trains that we can use to sort the data.
I have no idea what data the committee asked the Big 6 to provide. I do know what data I want:
- The number of trains operating longer or shorter than the critical length that defines long or short. For an industry that doesn’t seem to be certain of how many main line trains it operates daily, that may be challenging.
- The number of main line train-miles operated by short manifest trains.
- The number of main line train-miles operated by the long manifest trains.
- The total number of reportable main line accidents.
- The number of accidents involving short manifest trains.
- The number of accidents involving long manifest trains.
- The length and weight of every train involved in the accidents.
- The cause reported for each accident.
That data would be a good beginning. Then we can sort. We can get “granular.”
If I recall correctly, accurate measurement of train lengths and weights was critical input data for the development of the braking algorithm that makes PTC almost positive. And if I still recall correctly, input of accurate train weight and length is required prior to initialization. And I think I recall that the PTC braking algorithm is “predictive” rather than reactive, advising the locomotive engineer of the braking effort that must be achieved at any particular moment, and revising the braking curve as time and distance to the target change. This means that the algorithm must contain a certain ability to associate variations in train lengths and weights with different braking efforts in order to stop the train to the rear (short) of the target. In this element of safe train operations, train length and weight are critical inputs.
Remember that when FRA proposed the minimum crew consist regulation, a serious objection was raised by railroad management that there was absolutely no data demonstrating that two-person crew operations were safer than single-person operation—and if improved safety performance could not be established, the requirements for train staffing were a matter for collective bargaining, not government regulation. I think that objection was valid: “Show us the data.”
Now the question has been raised: Do trains exceeding (insert number) feet in length and/or (insert number) tons in weight constitute an elevated risk to safe train operations, and if so, what steps must be taken to mitigate that elevated risk? We can either provide and critically analyze the data to answer the first and propel our efforts at mitigation, or we can let somebody else, like Congress, do that based on speculation, special interest or ideology.
You decide.
David Schanoes is Principal of Ten90 Solutions LLC, a consulting firm he established upon retiring from MTA Metro-North Railroad in 2008. David began his railroad career in 1972 with the Chicago & North Western, as a brakeman in Chicago. He came to New York in 1977, working for Conrail’s New Jersey Division. David joined Metro-North in 1985. He has spent his entire career in operations, working his way up from brakeman to conductor, block operator, dispatcher, supervisor of train operations, trainmaster, superintendent, and deputy chief of field operations. “Better railroading is 10% planning plus 90% execution,” he says. “It’s simple math. Yet, we also know, or should know, that technology is no substitute for supervision, and supervision that doesn’t utilize technology isn’t going to do the job. That’s not so simple.”
