WASHINGTON – Long manifest trains are more likely to derail than their shorter counterparts due to excessive in-train forces – and the number of wrecks related to train makeup and handling issues has increased sharply since U.S. railroads adopted Precision Scheduled Railroading operating models that rely on ever-longer trains, according to a National Academies of Sciences report.
The 105-page report on the impacts of freight trains longer than 7,500 feet recommends several steps that the Federal Railroad Administration should take to ensure that railroads are fully mitigating the risks associated with the operation of long manifest trains.
“Railroad operations have changed a great deal over the past few decades, as have the technologies used and our understanding of the best safety management practices,” Debra Miller, a former member of the Surface Transportation Board and chair of the committee that wrote the report, said in a statement today. “So, the time is right for Congress, regulatory bodies, and the industry itself to take a closer look at railroad practices and regulations to ensure the safest operations going forward.”
The report also urges Congress to authorize the FRA to collect more data on blocked grade crossings. FRA should then work with railroads to find solutions for the most frequently blocked crossings. The safety regulator also should have the ability to levy substantial fines on railroads for blocking crossings.
Finally, the report says that railroads should be deterred from operating freights over siding length on routes that host Amtrak passenger trains. The FRA should be given the power to enforce passenger train right of preference, the report says, and be able to fine railroads when Amtrak trains are delayed because they can’t meet or pass freight trains.
Congress ordered the report as part of the Bipartisan Infrastructure Law in 2021. It specifically asked the National Academies to examine the impact of long trains on safety, grade crossings, Amtrak service, and greenhouse gas emissions.
The report noted that there’s no threshold at which a train becomes a “long train” that should be subject to greater safety concerns. As manifest train length increases, however, the number and mix of freight car types and tonnage contributes to increased forces that can stress equipment, create handling challenges for engineers, and increase the potential for derailments if in-train forces are not properly managed, the report says.
“The operational demands of long manifest trains, therefore, require a combination of responses by railroads that includes well-designed and consistently applied train makeup rules, the deployment of appropriate technology (e.g., DP units, brakes, engineer-assist programs), and assurance of crew readiness and competency,” the report says.
All six Class I railroads told the 12-member report committee that the operation of fewer but longer trains should result in a lower number of derailments overall, and that derailments caused by equipment and track should not be affected by train length. And the Association of American Railroads has insisted that railroads have safely operated long trains for decades.
But the report found a positive relationship between average through train size and derailment rates related to train makeup and handling. The report committee also took the AAR to task for not providing full access to derailment data.
“The committee 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,” the report says. “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.
“Nevertheless, a review of publicly available data on train traffic indicates that the average length of manifest trains has been increasing coincidental with an increase in the rate of derailments of interest,” the report says. “Absent more detailed data, the committee was not able to verify that the operational demands created by longer manifest trains are being fully controlled, and indeed the limited analyses that could be performed suggest that more targeted safety assurance measures may be needed.”
Jessica Kahanek, a spokeswoman for the AAR, said the information the report committee requested would have required significant data aggregation as well as disclosure of commercially sensitive information. Railroads held numerous conversations with the report committee about the data, including recommending use of a third party to analyze and make the data anonymous.
“Despite this, railroads provided massive amounts of data, time and expertise to help inform the review committee on how railroads model, build and safely operate trains no matter the length or commodity mix,” Kahanek says.
AAR said the study “does not demonstrate data supporting any claims that manifest trains exceeding 7,500 feet pose additional operational safety risks.”
The report was, however, able to show the clear trend: Derailments related to train handling and in-train forces began increasing in 2019 in line with increases in train length.
The report also found differences in the long train derailment rates of the big four U.S. systems. Derailments related to excessive in-train forces were significantly higher on Norfolk Southern and significantly lower on BNSF Railway.
“We have made significant improvements in recent years, including achieving a 35% improvement in our mainline train accident rate in 2023. There is no finish line when it comes to safety, and we are continually working to improve,” a Norfolk Southern spokeswoman says.
After derailments related to in-train forces in February and March 2023, Norfolk Southern CEO Alan Shaw ordered train lengths curtailed until the railroad could develop and implement rules on the safe building and handling of long manifest trains. NS says the changes to train makeup, train marshaling, and new software used in terminals has continued to result in improvements. The other Class I systems already had complex rules governing the building of long trains, including placement of distributed power locomotives.
In addition to having more demanding handling requirements, longer manifest trains can create challenges in maintaining radio contact over the length of the train. For example, if a train stops because of mechanical or other issues, a crew member must walk the length of the train and manually set a sufficient number of handbrakes. It is unclear whether the freight railroads have done enough to prepare their employees for these added challenges, the report states.
The FRA should survey the railroad industry for best practices in ensuring appropriate and consistently applied train-makeup procedures, effective crew training protocols, and reliable communications systems for the safe operation of longer trains. Railroads can use this information in developing their risk reduction programs, and FRA auditors can use it for assessing the quality of those programs.
Mark Wallace, first national vice president of the Brotherhood of Locomotive Engineers and Trainmen, says the report backs up what locomotive engineers have been saying about the trend toward longer trains.
“The report released today on long trains by the National Academies of Science, Engineering and Medicine … provides further evidence of what BLET members have long known: Long trains have a greater risk of derailing, have communications issues, and pose a threat to the public due to blocked crossings, among other issues,” Wallace said in a statement today.
The report did not make policy recommendations about greenhouse gas emissions and the operational fluidity of operating longer manifest trains.
Rail safety bills introduced in Congress after the NS hazardous materials derailment in East Palestine, Ohio, in 2023 propose capping train length at 7,500 feet. The wreck was caused by the catastrophic failure of a wheel bearing and was unrelated to train length.
“Any effort to impose a prescriptive train length limit would be both irresponsible and unfounded,” the AAR says. Limiting train length would snarl supply chains by increasing congestion on rail networks, AAR says, while increasing fuel use.
Note: Updated at 9:25 a.m. Central with comment from Norfolk Southern and at 10:17 a.m. with comment from BLET. Also updated at 12:34 p.m. with additional comment from AAR.
For years Big Tobacco used to say “smoking doesn’t cause cancer” and rolled out huge PR programs and manipulated statistics to prove it. They knew then that any relationship with a deadly disease would kill profit.
So here we are with the AAR saying “long trains don’t have problems” and refuses to show their own data to prove it or tell you that the analytics involved have no value. They know what everyone suspects, any relationship between long trains and unsafe operations would kill profit.
Hedge Funds want their money “now”. They don’t want to sacrifice their money on some silly thing like investments in operational safety, longer sidings and better technology to monitor train health. They want US railroads to be as stupid and as profitable as possible.
Like in the book (and subsequent movie) “Barbarians at the Gate” where reality started catching up to Big Tobacco, reality is starting to catch up to the Class 1’s.
The railroads had a tsunami of traffic just before the mega-market correction of 2008. Over the course of the next decade the railroads invested HEAVILY in their plant for when the traffic returned…which didn’t happen. Then came the PTC mandate and even more money invested. Lots-o-money has slipped through their fingers with nothing to show for it. So along comes PSR with promises of miracles. But real PSR requires real investment, the absolute LAST THING the railroads were going to do. So, like PRR & NYC being thrown together and hoping for the best, the railroads adopted ersatz PSR and hoped for the best (and hoped no one would notice). Hope in one hand and pour a milkshake in the other and see which one fills up first.
Daniel, disagree with your point on “PTC and ….nothing to show for it”
Even though they were required to create the PTC infrastructure, the Class 1’s soon found out there was a lot of metadata that could be collected on that new backbone. Data that could allow them to perform analytics to measure internal performance.
If one believes no PTC data was used to analyze how to optimize their PSR based routes, well……..
The hedge fund invasion started right after PTC was implemented. I don’t find that a coincidence.
An HO scale train can tell you all you need to know.
Passenger & freight trains ran on the same tracks for almost 100 yrs before Amtrak, the difference now is extreme greed less track, less employees they (the
RR’s) are running out of options to make more money to satisfy the stockholders. If this is the new normal then the RR’s should foot the bill to make all crossings into over/under passes or “trench” the RR tracks these long trains create numerous inconveniences & public safety hazards for the benefit ($$$) of a few!
Pax and freight ran on the same tracks…along with regulated pricing and return that would guarantee enough left over to maintain those tracks for both purposes. Then came partial deregulation and went the guaranteed income to cover costs. The flaw was not addressing how track capacity/maintenance would be funded for the public interest (passenger trains). The answer lays in stepping on the toes of individual states and no one is ready to die on that hill.
Amtrak has been using freight tracks since it’s formation, until the government railroad figures out how to get their own tracks, quit complaining, most countries with their HSR built their own passenger rail tracks, which is why they’re able to upgrade and have fairly good service, the US on the other hand isn’t able to have the kind of passenger rail system, so they use freight tracks, passenger and freight trains don’t mix.
The plan was simple: Instead of Yard A assembling multiple trains a day for multiple destinations cars are pre-blocked for those destinations and assembled into one, big land barge. Sounds great, one train, one crew. Problem is that first pre-block of cars was all empty center beam flat cars. So management yells at the yardmaster for letting that happen and the empty center beams go on the back. But that’s contrary to how the data says the train should be made up and costs time dropping that block off at its destination. So management yells at the yardmaster for not making the numbers…and so on.
Doing more with less means a very real possibility of reaping more bad consequences along with the good.
When I started reading this I wanted to yell “DUH!” This IS NOT ROCKET SCIENCE. This is common sense. The more you stretch something, the greater the chance it will snap. And while DPU’s help maintain speed and fluidity of trains, they can also create overwhelming buff forces when something ahead of them happens that immediately slows the train before the DPU’s can react. Be it a mechanical issue, a train handling issue or the poor or ill-planned placement of a light, long length car in an ill advised location in the trail, these forces will happen and that is what causes most derails in overlength trains. You can prove this with children’s wooden alphabet blocks. Line up 20-30 in a row and push them from the rear and then stop them quickly at the front before the pushing person can stop pushing and you will see the row of blocks splay left or right like a coil of a snake. Knuckles on rail cars are not designed to prevent these buff forces from acting on the mass of the car and that mass has to go somewhere, either up off the rails or to the left or right.
Congress doesn’t need a spend money on high dollar consultants to them what common sense should tell them. Reduce the length of trains to a reasonable length (8500 ft is a nice number) and many derailments currently happening will cease to be a problem. This is Congresses job. Take your NTSB investigators words for it. The railroads will fight it as long as they can because that is what railroads have always done; fight until forced by law to do what they should have been doing all along. The theory of PSR may have had merit. But the practice of PSR has only made Wall Street happy…until something goes wrong! And then you will hear “weeping, wailing and gnashing of teeth” like it was totally unexpected. Again I say, “DUH!”
Guess it will take legislation to require any data requested by FRA or STB be provided within 2 weeks. That is data requested from all 6 class 1s. No filtering any data thru AAR.
GREGG – You know, everyone reading these pages knows, every railroader knows, every fire/ police EMS first responder or 911 dispatcher knows, every motorist stopped at a grade crossing knows.
Read any history book. The fact that everyone with a brain knows the garbage that is going on, doesn’t stop it from happening.
long trains are a hazard and should be outlawed. I wonder how much that report cost, I could have told them everything for free, its just common sense
This is a compounding problem more than anything. PSR managers use DPU’s as a workaround for poorly-blocked trains regardless of length. Then when the managers try to make the poorly-blocked trains even longer, the workaround loses effectiveness.
You could haul a 10,000 ft, properly-blocked train with 2 engines up front way before PSR, and memory-holing this information unnecessarily handicaps operations.
Manifest trains contain lumber center beam cars on most occasions. Almost everyone who follows railroading knows those cars are subject to derailment on sharp curves and slack action. The AAR and railroads know this very well. And, there are other long cars as well.
The AAR is a poor place to go to collect information and this report shows it. They are more or less a lobby for railroads. Federal derailment reports are a much better source of information. But, there is one problem with them. They seem to take forever to be completed.
“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.”
Doesn’t this statement from NAS say it all about the class 1s? They restricted data access, denied NAS raw data and aggregated it themselves and limited analytical methods NAS could use. Why did the class 1s do this? BECAUSE THEY KNOW! The intent is clearly obfuscate and insure that public safety policy makers have insufficient information for decision support.
Having been a quality/reliability statistician for 35 years, I know there is no substitute for raw data. The class 1s limiting data and pre-crunching it amounts to lying.
Their lawyers are legendary at blacking out lines of necessary data from reports as “proprietary.” AAR Can not be trusted and neither can the railroads who owe allegiance to their stockholders, not the public at large…