RSS Feeds
L.A. Crash Could have Been Averted with Positive Train Control
PTC will transition a signaling system with near-total reliance on train crews to a digital model that compensates for human error
John Dodge, Editor-in-Chief -- Design News, October 21, 2008
Visit our in-depth investigative coverage package for more on the history and technology behind Positive Train Control.
Positive Train Control (PTC), a technology that promises to dramatically reduce rail accidents and whose origins go back almost 90 years, grabbed the spotlight following the horrific collision in Chatsworth, CA Sept. 12. The head-on collision between a Metrolink commuter train and a Union Pacific freight claimed 25 lives and injured 135.
PTC would have warned the allegedly inattentive commuter train engineer that he was to stop at the end of the track block on which his train was traveling. Failure to respond to the warning would have automatically brought the train to a halt a safe distance before reaching the end of the block. Doubtless, PTC theoretically would have averted the collision although many in the railroad business, government officials and PTC vendors hesitate to say so before the National Transportation Safety Board (NTSB) officially rules on the cause of the crash.
Modern PTC technology has been around for more than 20 years and was put on the NTSB's "Most Wanted Safety Improvement List" when it was first compiled in 1990. NTSB Chairman Mark Rosenker at an October railroad safety conference in Denver, CO called on the industry to implement it. Then again, so did his predecessor Jim Hall, who traced the technology to automatically stop trains back to 1919, in a 1996 speech.
PTC will transition a signaling system with near-total reliance on train crews to a digital model that compensates for human error by automatically enforcing speed restrictions and the right for a train to occupy a stretch of track known as a block. Except on a few small stretches of track in the U.S. today, most trains won't stop or slow down unless a human tells it to.
The present signal system principally relies on analog track circuits in the rails, signals and fixed blocks, whereas PTC moves to dynamic electronic blocks that promise to keep trains traveling in the same direction or opposing ones apart. PTC's building blocks are comprised of digital radio communications, computers, Geographical Information Systems (GIS), the Global Positioning System (GPS) and in some cases Wifi.
One early PTC system called ARES (Advanced Railroad Electronics System) implemented over 250 miles of track on the then Burlington Northern Railroad (now BNSF) worked successfully for six years until the project ended in 1993. One study says it promised to statistically reduce control system-related accidents from 50 accidents per year to 0.05 accidents per year.
"With PTC, two things cannot
occupy the same space at the same time. You have a moving bubble around each
train, maintenance vehicle, or person walking the tracks or a train," says
Steve Ditmeyer, former BN R&D director. By a person on the tracks, Ditmeyer
means railroad maintenance workers equipped with GPS receivers and radios much
like the ones soldiers have and not a happenstance individual out for a stroll.
Would the Chatsworth crash have been averted even if a commuter train's
engineer was text messaging through a red signal?
"I'm appalled that the accident occurred. Hopefully, those people did not die in vain," says Ditmeyer, also a former associate administrator for R&D at the Federal Railroad Administration (FRA), the agency that writes and enforces railroad safety rules. "My friends and I ask: ‘Is this the accident that finally makes Congress make the railroads install PTC?'"
Apparently, it is. On Oct. 1, Congress passed sweeping rail safety legislation that mandated PTC installation by Dec. 31, 2015. And on Oct. 8, Union Pacific with its Vital Train Management System announced it would implement PTC on its tracks in the congested Los Angeles Basin by 2012 with testing to proceed on 456 miles of track in Iowa, Nebraska and Wyoming. On the same day, BNSF said it would meet the 2015 federal deadline for implementing PTC across its entire system and in the Los Angeles Basin by 2012. Union Pacific, BNSF and Norfolk Southern also announced they had ironed out critical interoperability issues.
"This is all very good news," says Ditmeyer.
PTC consists of three sets of components: onboard computers, radios and displays on locomotives and maintenance vehicles; wayside radios and sensors at switches and detectors, and dispatching center radios, computers and displays. Real-time communications between the three elements is critical. PTC systems are initially installed in conjunction with pre-existing signaling systems. As they are brought on-line, they replace the existing systems and the railroads can eliminate the familiar trackside signals.
While the public focuses on how PTC can dramatically improve safety, the railroads are also looking for a return on their investment by safely running more trains faster on their same infrastructures. But raising the affordability issue with such a large loss of life in Chatsworth is touchy subject. The FRA estimates between $2.3 and $5 billion for a broad rollout on the nation's 140,000 miles of track.
"I wish we did not have to focus on cost and affordability. For a long period of time, the government has been required to do those kinds of analysis before we can put a rule into effect," FRA Administrator Joseph Boardman said in a media teleconference a few days following the accident. "We're making steady progress, but not as fast as (the public) would like. Whenever you do a rule, you do a cost benefit analysis."
Currently, the FRA lists 11 PTC projects underway across 4,000 miles of track around the nation. Those projects were jumpstarted in June 2005 when the FRA published PTC performance standards and regulations. The major technical obstacles, according to Boardman, are developing braking software algorithms to stop trains of different weights and lengths, freeing up radio bandwidth for the required communications and interoperability between railroads whose rails regularly host their rivals' locomotives.
"The technology is not ready for implementation. There are so many variables given all the technical issues to be resolved," says Tom White, a spokesman for the Association of American Railroads (AAR), the principal railroad lobbying organization. "Six steps are mandated and the furthest any railroad has gone is step two so we can't say with confidence when PTC will be implemented."
However, one knowledgeable source labeled the technical issues a "red herring," saying if they were not worked out years ago, they could have been. As for cost, the source said $3 billion for the $40-$50 billion railroad industry is "a pittance."
Of course, the accident provided the impetus for the recent developments, but TRAINS Magazine Columnist Don Phillips gives some of the credit to Boardman with trying to make the railroads make PTC a priority.
"He is actually the first FRA administrator to do something about this. The guy almost quit because the White House would not let him do anything," says Phillips, also a former transportation writer for the Washington Post. "The guy is a straight shooter and wanted to leave his mark. I've talked to a lot of railroaders who said, ‘Darn, he's actually going to do this (PTC).' Every time there was a wreck, he viewed it as a lobbying opportunity. And long ago, the railroads figured out it is actually cheaper to pay for wrecks (than PTC)." Despite several requests, Boardman declined to be interviewed.
"We do have a rollout plan, but I
would not be comfortable in stating the time frame," said Pat Hiatte, BNSF
director of corporate communications before the Oct. 8 announcement. Despite
several requests, Union Pacific elected not to participate in this article.
BNSF is launching its Electronic Train Management System (ETMS)
on 35 subdivisions and a more advanced version of 300 miles of track in Texas and Oklahoma.
"If a switch is misaligned, the system will warn the crew or brake the train to
a stop," says Hiatte.
BNSF's ETMS
relies on GPS in concert with a geographical information system to determine
train location and its VHF radio for communications. The crew in the cab is
constantly fed data about the train's and track's status on two displays in the
locomotive with respect to authority limit, speed, switch position, track
integrity to identify broken rails and signal aspect (color). Should the crew
fail to respond to a warning, ETMS
will stop the train, backstopping the train crew.
"Human factors still have a considerable effect on the industry's ability to
operate safely. ETMS
will enable us to eliminate human errors by increasing awareness for
employees," states BNSF's ETMS
literature. BNSF's PTC provider is Wabtec Corp. of Pittsburgh, PA
which is a major developer of PTC systems. Wabtec chose not to participate
in this article, but Ditmeyer speculates its PTC technology could be heavily
influenced by BN's ARES systems developed by Rockwell Collins. Wabtec bought Rockwell
Collins' Railway Electronics Division
about 10 years ago.
Two other leading providers -
Ansaldo STS-Union Switch & Signal
(US&S) and Lockheed Martin - were happy to talk about their PTC
technologies.
US&S is working the Alaska Railroad on a not-so-subtly named PTC implementation called Collision Avoidance System to be used in concert with its signaled and Direct Track Control territory. It is slated for completion late this year or in early 2009. Ironically, the two principal motivations behind PTC imply safety without mentioning it explicitly, according to Keith Szewczyk, US&S assistant vice president of engineering.
"One is to improve headway or how fast you can get a train through the system. The faster the railroads can move trains, the more people and freight they move and the more money they can bring in. The boundaries (or blocks) move with the train instead of being fixed," he says, adding PTC will squeeze the waste out of unoccupied track. "The other reason is to remove maintenance-of-way equipment that clogs up the tracks."
PTC from US&S must undergo exhaustive safety testing to statistically prove it is ready. "Safety rises to the level everybody accepts," says Szewczyk.
The two main modules include an Office Safety Server which
manages train authority limits (speed, track occupancy, etc.) and keeps tabs on
train location. The other main piece, PTC Cab, is the onboard computer and
displays in the locomotive. Interfaces to a GPS receiver, train brakes, data
radio and the locomotive's tachometer also would be included. Wayside equipment
to monitor and throw switches is optional.
Conventional signal systems use track circuits for both block occupancy
determination and for broken rail detection. PTC does not need track
circuits for determining where trains are, but railroads will have the option
of determining whether or not they want broken rail detection using track
circuits or other technology.
"Track circuits do broken rail detection. When you go to PTC, you
take that away. With overlay PTC,
you'll still see track circuits," Szewczyk says.
Lockheed
Martin is a relative newcomer to PTC which it based in part on
existing submarine navigational technology. It is designing a vital PTC system for the Australian Rail Track
Corporation (ARTC) called the Advanced
Train Management System (ATMS) on a stretch of track north of Adelaide, South
Australia.
ARTC presently uses trains' orders (authority granted to the train by paper order or voice instruction) in what is known as dark un-signaled territory of trackside signals, so ATMS will be vital when it goes live in 2010. ATMS has the three basic components of the other systems, but uses a commercial cellular network instead of a railroad-owned radio frequency.
"If the dispatcher is planning to route a train, he directs the movement of the switches from the control center. "You don't want a train moving over a switch when it is being repositioned," says Steve Osborne, rail systems chief engineer for Lockheed Martin. Still in the R&D stage, Lockheed Martin is also working to develop the system for the Norfolk Southern Railroad.
ARTC is considered a medium-density line with 20-25 trains a day, according to Osborne. Some high-density and overburdened lines in the U.S. see upward of 80-100 trains a day, which can really test PTC. BNSF, for instance, is piloting its implementation starting on lower density lines and less challenging terrain.
"We work through the operating environment on freight and passenger on multiple mains with crossovers in both signaled and un-signaled territory. Then we do mountain grades plus tunnels as we work toward more challenging terrain and entities," says BNSF's Hiatte.
PTC promises to alleviate much of that congestion which in part is caused by the present and in some spots antiquated signaling system. And it will make the railroads much safer. Still, experts hesitate to say all major collisions and over-speed accidents will be eliminated.
"One would never want to say there's absolutely no chance for a collision or over-speed accident, but with PTC you've certainly significantly reduced the probability of one occurring. A lot more devices and people must fail before an accident can occur," says Ditmeyer.
The video below, supplied by
Burlington Northern Santa Fe, shows how its Electronic Train Management System
(ETMS) works. ETMS is its version of
Positive Train Control. Watch Now
Watch more videos at Gallery »
Avoiding Collisions with Positive Train Control
