Norfolk Southern is inspecting all of its nearly 1,000 wayside hotbox detectors in the wake of the Feb. 3 derailment in East Palestine, Ohio, that safety investigators say was caused by a wheel bearing failure.
NS also has confirmed that detectors in the East Palestine area were all operating properly when merchandise train 32N rolled past them prior to the 8:54 p.m. derailment.
“Under the supervision of the Federal Railroad Administration, the company has inspected all wayside detectors in the area of the incident and found they were operating as designed,” NS said. “Out of an abundance of caution, Norfolk Southern is now inspecting all of the nearly 1,000 wayside heat detectors on its system — on top of the regular inspection of the detectors every 30 days.”
CSX Transportation has completed a week-long inspection and calibration of its network of more than 1,000 wayside detectors as a precaution.
The Federal Railroad Administration has not issued a safety advisory regarding hotbox detectors after the NS wreck, which spilled toxic chemicals, forced the evacuation of East Palestine, and prompted a firestorm of criticism of NS and the other Class I railroads.
The National Transportation Safety Board said today in a preliminary report that the derailment was caused by a wheel bearing failure. Two hotbox detectors recorded rising temperatures on the suspect hopper car, but temperatures never reached the critical threshold that would have tripped an alarm and required the crew to stop the train until after the train passed a third hotbox detector moments before the derailment.
CSX Chief Operating Officer Jamie Boychuk ordered the systemwide testing on his railroad. “We decided to do this as an extraordinary measure,” CSX spokeswoman Cindy Schild says.
Normally, CSX’s wayside detectors are inspected every two weeks but are not necessarily calibrated during those inspections, she says.
FRA does not regulate the trackside detectors that can warn of hot axles, dragging equipment, or other defects. They’re generally spaced between 10 and 20 miles apart on main lines.
It was not immediately clear if the other big railroads have or will conduct more thorough inspections of their detector networks.
“Union Pacific is closely monitoring the NTSB investigation and reviewing its preliminary findings. We constantly evaluate our processes and the Ohio derailment offers Union Pacific and the entire industry a chance to learn and continue working together to further improve safety,” spokeswoman Robynn Tysver says.
BNSF Railway said its detectors are inspected regularly. “Our wayside detectors are inspected frequently by our signal team as part of our robust, extensive and ongoing inspection program. In fact, BNSF leads the industry in using artificial intelligence for detecting wheel defects as algorithms sort through more than 35 million readings every day taken from our 4,000 wayside sensors to identify potential issues,” spokesman Zak Andersen says.
A Canadian National spokesman said the railroad has a calibration and maintenance protocol in place for its detectors. A Canadian Pacific spokesman referred to safety-related comments CEO Keith Creel made at an investor conference this week.
NTSB Chair Jennifer Homendy, speaking at a news conference this afternoon, said the board’s investigation will put wayside detectors under a microscope. Investigators will look into the spacing of detectors, whether they are or should be monitored in real time from a central data center, and the temperature thresholds each railroad uses to trigger an alert that would require trains with defects to stop.
The spacing of detectors and how they measure equipment defects are set by railroads and their standards vary considerably, she said.
In Norfolk Southern’s case, alarm thresholds are set at three levels:
- An axle temperature between 170 and 200 degrees Fahrenheit indicates a warm bearing that the crew must stop and inspect.
- Crews must also stop and inspect when a temperature difference of 115 or more degrees is detected between bearings on the same axle.
- At temperatures above 200 degrees, the problem is considered critical and the crew must set out the affected car.
NS says its wayside detectors “trigger an alarm at a temperature threshold that is among the lowest in the rail industry.”
“Roller bearings fail. But it’s absolutely critical for problems to be identified and addressed early, so these aren’t run until failure,” Homendy said.
ProPublica, an investigative journalism website, today reported that NS allows a monitoring team to instruct crews to ignore detector alerts in some instances.
Like other railroads, Norfolk Southern has been deploying more sophisticated detectors that use high-definition cameras and artificial intelligence to identify defects.
I keep wondering if a motorist or other bystander saw the smoke, fire or glow long before the derailment but chose not to “get involved”. Wow, what a huge mistake that would be. I think the railroad industry could do a better job of educating the public in what to look for as a train passes and how to report possible problems. In addition, trains hauling dangerous products should be limited in length and have a manned caboose. Portable defect detectors should also be used on any car hauling dangerous products. That way, crews could be alerted immediately to any possible problem. How much would all of these ideas cost? Much less than the East Palestine disaster.
Note that in all of this talk, no one is addressing the issue —-> how robust is the literature on bearing failures in a railroad environment?
We know high temps area symptom of failure. But how well do know exactly when and how and where?
Technology is important. But we’re in a age of technology worship, thinking it can do everything no matter how little we truely understand about what needs to be done.
@Howard Fine Where in the preliminary report does NTSB state that the accident was ‘preventable’?
The NS line I live next to only has three announcements programmed on its defect detectors. “No defects”, “hot wheel”, and “critical alarm.” It only identifies the milepost and the axle count of the defect (if any), nothing else.
There is usually a separate discussion with Wayside Office and the dispatcher on what to do as they slow down.
Why not develop a sensor and data system for railcars? Put a heat sensor on the bearings, also provide an air pressure reading, and any other useful information. Have it connected in a wireless data bus to the locomotive or to a yard mast to the yard master.
Norfolk Southern issued the following statement on Thursday in response to the National Transportation Safety Board’s (NTSB) preliminary report on the East Palestine, Ohio, train derailment.
Statement from NS
Since the incident, Norfolk Southern has focused on supporting the Village of East Palestine and the NTSB throughout their investigation. NTSB investigators have worked tirelessly, and we appreciate everything they are doing. We share the NTSB Chair’s frustration about the significant misinformation about the incident.
Norfolk Southern will continue to support the NTSB’s investigation. Their preliminary report indicates:
• The rail crew operated the train within the company’s rules.
• The rail crew operated the train below the track speed limit.
• The wayside heat detectors were operating as designed.
• Once the rail crew was alerted by the wayside detector, they immediately began to stop the train.
The company’s wayside detectors on its network trigger an alarm at a temperature threshold that is among the lowest in the rail industry.
At Norfolk Southern, our highest priority is the safety of our people and the communities and customers we serve. Under the supervision of the Federal Railroad Administration, the company has inspected all wayside detectors in the area of the incident and found they were operating as designed. Out of an abundance of caution, Norfolk Southern is now inspecting all of the nearly 1,000 wayside heat detectors on its system – on top of the regular inspection of the detectors every 30 days.
We and the rail industry need to learn as much as we can from this event. Norfolk Southern will develop practices and invest in technologies that could help prevent an incident like this in the future. We will also work with the owners of the rail cars on the integrity and safety of the equipment we use.
The speculation and misinformation about this incident have been extremely upsetting to the citizens of East Palestine. Norfolk Southern remains committed to the people of the community and will continue its work to help them thrive. Already, we have made significant progress in cleaning the site, engaging the community, and providing financial support for families and small businesses, all to help East Palestine.
Mr. Rauch,
Watch the NTSB video press conference. The car that caused the derailment was the 23rd car in the train.
“”FRA does not regulate the trackside detectors that can warn of hot axles, dragging equipment, or other defects.””
“”The spacing of detectors and how they measure equipment defects are set by railroads and their standards vary considerably””
…..and that’s a big part of the problem when you let corporations “self-regulate.” WHY isn’t there a FRA requirement for detectors to be present every X miles on lines of 25mph or more, carrying X cars of hazardous materials annually. Why isn’t there a FRA requirement to test & calibrate at least as frequently as the manufacturer’s specifications?
Why isn’t there a requirement for acoustic bearing detectors? The RRs can afford them & it’s not new technology.
It pains me to suggest this but given that there were visible signs of trouble starting at least some twenty miles back one has to wonder if the crew could or should have been able to see them while looking back to check their train in curves and get stopped before anything serious happened.
The train which derailed is NS 32N and that day it was 151 cars, 9,300 feet long, 18,000 tons.
I’m not familiar at all with this particular territory and its curves and terrain but I am familiar with occasionally looking back at my trains over the years and 2 miles back is too far to see anything, even at night, even on fire. But, again, I don’t know how car back the derailment took place in the train. To see back one mile it would take an exceptional curve devoid of trees or anything blocking your view of the train.
Engineers and conductors looking back on their trains is recommended mostly because its easy and cheap, not because its terribly effective or reliable. A wild guess is the car with the journal on fire was not close to the head end or even remotely close because I think that fact would have come out by now.
I understand the big class one’s build defect detectors 10-20 miles apart and connect them to a network to leverage information in ways that couldn’t be any other way. I’m not a signal maintainer and have no direct knowledge or experience with detectors but I understand every car being moved along these mainlines is being scanned by defect detectors over and over and over, all day and all night, on different trains, different crews, different dispatchers territories and the technology can be comparing the data across days and weeks and months of the same car so when these bearings reach critical mass and seemingly build heat so fast that they go from appearing normal and fine to melted and on the ground in perhaps 20-30 minutes, they find most of these and stop the trains preventing a derailment. Why they didn’t find this one, is the $64,000 question.
Mr. Rauch,
Watch the NTSB video press conference. The car that caused the derailment was the 23rd car in the train.
During the Q&A session following Homendy’s description of events, at roughly the 36:45 point in the video a question was asked “did the crew have any knowledge that the temperature [of the bearing] was rising as they went over the first two hot box detectors [ie before the 3rd detector issued the critical alarm that bearing temperature exceed 200 degrees above ambient]?” Hall’s response was “the detectors as they’re designed will provide a radio announcement and so the radio in the cab would indicate that there is an issue … it would have announced the temperatures at this particular axle and told them [ie the crew] which axle it was from the front – just an automated computer voice…”
Is this correct? My impression is that the detector will only provide any over temperature audio information to the crew when the 200 degrees above ambient threshold is reached. Furthermore, the human being at the wayside desk (assuming there is one) who is monitoring detectors system wide would be the only means that the crew would find out about the rising temperature trend when passing detectors prior to getting an automatic detector alarm re: exceeding the 200 degree threshold at the 3rd detector.
As a practical matter, with a single crew operating a train of 400+ axles (ie, 800+ bearings) for 200 miles passing over 20 or so HBDs, if some random bearings were warmer than normal that could lead to data overload. Crews would have to keep track of specific axle bearing temps from prior HBD passages to determine if there was a dangerously rising temperature. That’s why I think a centralized wayside desk is used to analyze and reduce the data to something actionable for the crew (eg, stop and inspect).
Mr. Hall’s response may have been misinterpreted because following the quote that you gave, he said “above 200 degrees”. The crew would not have received a radio alarm message if the bearing temperature was below the alarm threshold of 200°F.
I can’t help but wonder whether the long trains currently run figure into it. How many won’t fit into sidings and how far are away are they if an alert goes off?
What in the heck are you going to do if you do get an indication? Is there a siding to get on to in either direction without backing a train with a questionable bearing? Do you sit there on the main?
I think the idea of an on-off threshold is a dumb idea. with the communication available with PTC, why not compare one reading with another as a train proceeds. If you have had to pay for PTC, why not use its capabillities to prevent this kind of mess?
You claim “clear warning” Howard, but the previous two readings were significantly below the threshold if I am doing my math correctly converting temps above ambient to absolute temps. Care to substantiate the accusation?
The absolute temperature, first detector, in itself did not indicate an imminent failure, it is an alert. The second reading 10 miles later is higher and suggests the first reading was accurate not an anomaly. and The next (third detector) was located 20miles (the first two were 10 miles apart) further, located at New Palestine, enough time and distance for the defective bearing to exceed NS defined CRITICAL temperature requiring set out. The train should have been brought to a controlled stop after the second hit indicates increasing temperature, and the offending bearing inspected and temperature checked. Someone made the decision not to make a timely stop, and reports are that this was not a unique situation.
Could NS’ Critical Temperature be too low?
Wonder how long it will take the Class 1s to install detectors everywhere on 10 mile separations.
Read the preliminary report Alan and listen to the NTSB this afternoon. The NTSB , responding to now wide public outcry, they went off typical non comment until final report, this accident WAS preventable, NS failed to stop the train in time, despite clear warning from the two previous detectors. Corporate Lawyers seem to have silenced NS response….