Weather forecasts for railroads

CHamilton · Dec 1, 2012

It's been mentioned several times that railroad companies subscribe to services that provide custom forecasts for their specific needs. It turns out that such a service is provided by WeatherData, which is now part of AccuWeather.

WeatherData's founder Mike Smith has written a very good book called Warnings: The True Story of How Science Tamed the Weather. The book focuses on how weather forecasting (especially for severe storms) has been improved in the last half-century. But it has a surprising amount of information on how weather affects railroads, and it starts with a harrowing description of a (luckily, fictional) Vermonter accident caused by a flood and washout.

The book is extremely well-written. Highly recommended.

Shortline · Dec 7, 2012

As an interesting note, in addition to monitoring the weather, many RR's also monitor other events-Go to Ft. Worth, Omaha, Jax, and the HQ/Dispatch centers of Class 1, 2, and 3 RR's around the country, and you'll find they're also monitoring and getting alerts from the USGS for Earthquakes, TSA for Terrorist threat levels, etc.

George Harris · Dec 7, 2012

CHamilton said:
. . . and it starts with a harrowing description of a (luckily, fictional) Vermonter accident caused by a flood and washout.

Maybe not fictional, or at the least based on a real one:

From NTSB/RAR-85/14, titled, Derailment of Amtrak Passenger Train No. 60, the Montrealer, on the Central Vermont Railway, near Essex Junction, Vermont, July 7, 1984

The abstract:

About 6:50 a.m, eastern standard time, on July 7, 1984, northbound Amtrak passenger train No. 60, the Montrealer, derailed while passing over a washed-out section of gravel embankment under the main track of the Central Vermont Railway near Essex Junction, Vermont. Two locomotive units and the forward seven cars of the train derailed and were destroyed or heavily damaged. Three passengers and an Amtrak sleeping car attendant were killed; one Central Vermont crewmember died about 3 hours after the accident as a result of injuries sustained in the accident. One Central Vermont crewmember, two Amtrak attendants, and 26 passengers were seriously injured. Damage was estimated at $6,586,312.

The National Transportation Safety Board determines that the probable cause of the accident was a flash flood that destroyed the railroad support embankment over a small stream during a prolonged period of extraordinarily heavy rainfall. The flash flood was precipitated by the heavy rains and the collapse of a series of beaver dams upstream of the embankment in heavily wooded locations that were unknown and were not reasonably detectable.

Search it out and read the whole thing. It is some interesting reading, and a good example of how the NTSB analyzes things in detail.

CHamilton · Dec 7, 2012

George,

The accident you describe is very interesting. Thanks for the information. Just FYI, the incident in the book is a different one.

There was heavy rain on October 9, 2005. That day the track’s supports washed out with nearly one hundred feet of rail left dangling in the air. And the train was operating that day—for a time. [but] The railroad sent out crews to inspect the track ahead of the train and found the washout. The train was stopped in time.

George Harris · Dec 7, 2012

I heard about the Montrealer derailment when it happened, but had never read the NTSB report. Now I have, at least skimmed it. One of the most amazing pieces in the report was,

"About 80 linear feet of the 20-foot-high embankment had washed out, but the track structure across the opening produced by the washout remained fully intact and taut. Both locomotive units and the first two cars crossed the track over the opening. According to the fireman, the locomotive dropped 3 or 4 feet and then bounced up as though it were on a springboard."

Let me say this one time real slow: Two diesel locomotives each weighing 130 tons rode across an 80 foot gap supported by the rails alone. This all but defines IMPOSSIBLE!!! The load should have far exceeded the tensile strength of the rails. The rails should have broken right after the first wheelset entered the unsupported portion.