WASHINGTON, D.C. — The Washington Post reports that a temperature change in the rail, possibly complicated by routine maintenance done several days earlier, may have been a major factor in the derailment of Amtrak’s Auto Train in Florida last week, said investigative and industry sources.
Railroad sources had said earlier that heat was apparently not a factor because the temperature was only 81 degrees at the time of the derailment near Crescent City, Fla., and there had not been any significant temperature change in the last 24 hours. However, further investigation showed the temperature of the rail was far higher than the surrounding air after a day of direct springtime Florida sun.
National Transportation Safety Board member George W. Black Jr. said several days ago that investigators were looking at the possibility that the sandy Florida soil below the track had become waterlogged. But investigators have preliminarily determined that water played no role and that, in fact, the soil was dry.
Numerous other factors also are being considered, including what effect the passage of a heavy coal train six minutes earlier may have had on the track.
Also under scrutiny is the effect of routine maintenance that was performed a few hundred yards away from the crash site several days before, as well as maintenance performed over the winter. Track maintenance can change the temperature sensitivity of a rail, making it vulnerable to heaving or kinking if crews do not take proper precautions. Investigators do not yet know what precautions were taken by CSX crews.
Sources said the safety board will look at whether last week’s accident has anything in common with the wreck of Amtrak’s California Zephyr on April 23, 1990, in which track that had undergone maintenance four months earlier suddenly moved out of alignment because of heat.
The northbound Auto Train, which hauls passengers and their cars between Sanford, Fla., and Lorton, Va., derailed as it entered a curve at 56 mph, just 44 miles after beginning its run. Four elderly people were killed and dozens were injured.
The 35-year veteran engineer told National Transportation Safety Board investigators that he saw the track just ahead of him suddenly heave out of alignment when the train was about 100 feet away.
While the safety board will take several months to determine a probable cause, sources said investigators are looking at the possibility of a “sun kink” in the rail, which happens when heat expands welded rail to the point that the pressure forces the track out of line. It is common for a sun kink to appear just ahead of a train.
CSX Transportation, the freight railroad that owns the track used by the Auto Train, uses continuous welded rail on most of its mainline track. Because of the pressures that can build up in miles-long rail lengths, maintenance crews must be careful to monitor rail temperature.
Crews are supposed to keep careful records of the temperature at which the rail is laid, setting a “neutral rail temperature” at which there is no internal pressure or stress. A large propane rail heater may be used to achieve a specific neutral temperature when rail is laid or disturbed by maintenance. If temperature rises or falls considerably from that temperature, trains must be told to slow down.
Sometimes, especially in states that have wide temperature swings from summer to winter, crews routinely relieve or increase rail pressure by cutting out a short length of rail, called a “plug,” or inserting one. That either relieves or increases pressure.
Maintenance can relieve the tension in the rail by lifting it, thereby changing the neutral rail temperature.
In the 1990 Iowa wreck, the safety board determined that maintenance crews had failed to use a rail heater to raise the neutral rail temperature to the designated level for that area — 95 degrees — after doing maintenance in the area.
Sources said much work and many interviews will be required to determine whether maintenance was a factor in the Auto Train wreck and, if so, exactly what might have been done wrong.