FALLOUT continues from the spate of spectacular train wrecks early in 1996 (RG 4.96 p177) which marred a generally excellent safety record in the USA. In a hearing on December 11, the National Transportation Safety Board spread blame widely for the February 1 1996 runaway down Cajon Pass, California. Investigators believe that somewhere near the summit a brake hose pinched shut not far behind the locomotive. With brakes on most of the 49 wagons isolated, the train accelerated down the steep grade, derailed, and burned for a day and a half, blocking a major highway.
It transpired that the battery powered end-of-train unit linked by radio to the locomotive was not fully operational, so the driver was unable to vent the train pipe from the rear end and thus halt the runaway. Incredibly, NTSB claims middle management and crew members assumed they were satisfying top management’s safety concerns merely by installing the unit on the rear coupler, even though it wasn’t activated!
Five days after the accident, the Federal Railroad Administration issued an emergency order requiring working end-of-train braking units on all BNSF trains using the Cajon Pass. Two weeks later, the FRA asked the nation’s 10 largest railways to install working units on all trains operating in mountainous terrain, and this was completed by December 1996. US operators have agreed voluntarily to install the devices on ’virtually all’ freight trains by June 30.
NTSB first proposed that rear-end application be added to end-of-train brake pressure monitors in 1989. After a Santa Fe train ran away in Cajon Pass in January 1995, NTSB asked the FRA to order early compliance, but nothing happened. According to Robert Lauby, head of NTSB’s railway division, ’it wasn’t until after the  accident that the industry acted voluntarily ... and if anyone should have known better, it was the Santa Fe.’
The FRA has recommended modifications to locomotives and coaches, including stronger and taller collision posts, to protect both passengers and crews. Although FRA’s report took three years to complete, the fact that collision posts were ’totally destroyed’ in the February 16 1996 collision at Silver Spring between Amtrak’s Capitol Limited and a MARC commuter train, in which 11 people were crushed or burned to death, has clearly influenced the outcome.
The report also recommends improved fuel tank design, better methods to control noise and temperature levels inside locomotive cabs, and the creation of ’crash refuge’ zones for train crews where they can take cover if a collision is imminent. The report is being considered by the Railroad Safety Advisory Committee made up of regulators, railway companies and unions, which will present recommendations to the FRA before mandatory rules are formulated.
On December 23 the FRA issued a final rule requiring railways to design comprehensive safety programmes to protect track workers from being struck by trains. Each operator must produce an on-track safety manual, along with clear statements of employers’ and employees’ rights and responsibilities with regard to safety. Amtrak, commuter operators and Class I railways must be in compliance by March 15. The remainder of the industry has an additional two months.
Positive train separation
Phase one field testing of positive train separation (PTS), a form of automatic train protection which enforces train movement limits, has been successfully completed by BNSF and Union Pacific. ’Initial PTS test results have been very encouraging’, reported Nick Marsh, BNSF’s Assistant Vice President responsible for technical research and development on December 11.
GE Harris Railway Electronics has been working on this PTS system in conjunction with the FRA since 1994. BNSF and UP began joint trials last August on 1300track-km in Washington and Oregon.
Three more PTS test phases are planned for 1997, including use of a satellite-based global positioning system, computerised interface with dispatching systems, and display capabilities in locomotives.
Negative train separation
No precautions can prevent sabotage by someone with enough railway knowledge to defeat fail-safe equipment. This was frighteningly demonstrated on December 4 in Nebraska after a Union Pacific freight train was uncoupled, probably while stationary straddling a level crossing, and both brakepipe cocks were closed.
The train had unmanned locomotives in the middle operated through a radio link by the driver at the head end. When the driver restarted, both sections accelerated and slowed in accordance with the head-end controls, but the front part had more power and soon outdistanced the 55 wagon rear section. Running at full speed up to 20 km behind the front part, the second section travelled 80 km passing several red signals and dozens of level crossings - including one without protection where a farmer and his pregnant wife were almost killed as the unlit wagons hurtled out of the blackness on a moonless night.
The driver of another train eventually spotted the second section and raised the alarm. The dispatcher called the driver of the errant train who made an emergency brake application; both sections came to an immediate halt. UP spokesman John Bromley explained that a brakepipe continuity test is mandatory before a remote control train starts its journey, but there is no requirement to check continuity during brief stops en route. o