By now, the Green Line crashes have become familiar: May 2008 in Newton, May 2009 near Government Center, last month at Boylston Street Station. One dead, more than 100 injured, thousands delayed, $20 million in damage, lawsuits pending.
All three accidents could have been avoided — if the Green Line were equipped with an automated system to track and control trains to prevent train-to-train collisions and derailments when drivers speed or miss track signals that look like traffic lights.
That technology also would have prevented many of the lesser-known incidents on the Green Line, Boston’s only rapid-transit line without at least some form of computerized or mechanical accident-prevention system. Those lower-profile episodes include at least 54 of the 56 times trolley operators missed signals between 2000 and 2010, seven of which ended in derailments, according to a new study prepared for the T.
The catch? The T estimates it would cost $645 million to $721 million and require nine years to design, install, and test a modern collision avoidance-system for the Green Line. That is a staggering sum for an agency with at least $3 billion in unmet repair and replacement needs.
For roughly the same price, the T could replace nearly its entire bus fleet or complete much of the long-delayed Green Line extension to Somerville.
Investing in what is known as positive train control — an umbrella term referring not to one specific technology but different means of tracking and controlling train movements to prevent accidents — means spending resources that could go elsewhere. That money could be used to speed travel times, expand capacity, add service, or enhance reliability or comfort for the 1.3 million weekday riders who rely on an increasingly congested MBTA.
“Our customers expect safety, but they also expect service improvements as well, which is really the challenge we have,” Secretary of Transportation Richard A. Davey said.
Though most light rail lines lack the technology, the Green Line has been singled out by the National Transportation Safety Board, which investigated the 2008 and 2009 crashes.
And state officials are loath to repeat another passenger-injuring, headline-grabbing, service-disrupting crash — even as they emphasize that the Green Line completes hundreds of thousands of trips and ferries millions of riders without incident.
Davey, a daily Green Line commuter, indicated that overhauling the line with anticrash technology is a matter of when, not if, even if it means bumping other worthy projects to make room in the T’s infrastructure budget.
“We’re definitely thinking about what kind of financial bandwidth we have in the next five to 10 years and what are our priorities,” he said. “We absolutely have to get to” positive train control.
But some analysts caution that such thinking could be reactive, seeking to reassure customers at the expense of arguably more deserving improvements. Even automated systems can fail, said former state transportation secretary Fred Salvucci .
Instead, he said, the transportation agency could invest in new vehicles and the power systems needed to run more three-car trains, which could reduce breakdowns and make passenger boarding at crowded downtown stations less chaotic. That in turn could make boardings safer and “dwell times” — the time a train sits at each station — more predictable, reducing rear-end collisions.
“If someone’s got half a billion dollars to put into the Green Line, which I hope is the case, you ought to be really careful about where to put it, to improve performance,” said Salvucci, who now teaches at the Massachusetts Institute of Technology. “Performance and safety are linked.”
‘The board has articulated that safety upgrades on the Green Line have to be a priority.’
Vukan Vuchic, a University of Pennsylvania professor who has worked as a consultant for large transit agencies and written about automation and safety, said a collision prevention system can mean slower and less frequent service, and no technology yet can eliminate all accidents.
“It sounds good when you are on the side of safety, it’s true. But whatever you do, you have to see the trade-offs,” Vuchic said.
After the 2008 fatal crash in Newton, when an operator suffering from an apparent “micro-sleep” episode died after slamming into another train at 38 miles an hour, National Transportation Safety Board investigators noted the absence of technology that “would have intervened to stop the train and prevent the collision.”
And the safety board cited the Green Line and other crashes again this fall in naming positive train control to its “Most Wanted List” of national advocacy priorities. But the board does not consider a cost-benefit analysis in setting priorities to save lives, spokesman Keith Holloway acknowledged.
The Newton crash prompted the T to embark on a two-part strategy to inject technology into its traditional practice of training and testing drivers for alertness. It began a four-year, $460,000 trial of alarm systems and other technologies that fall short of full positive train control, using the 2½-mile route known as the Mattapan High-Speed Line as a laboratory.
But MBTA engineers say an alarm-style warning system would not be appropriate for the Green Line, which is why the T has also spent two years and $2.2 million on a study of what more sophisticated anticrash measures would look like for the Green Line.
Even during rush hour, Mattapan trolleys are spaced five minutes apart — rapid-fire by the standards of most light rail, the modern term for streetcar or trolley systems, but nothing compared with the Green Line.
That line — really four lines that converge — is the nation’s busiest light-rail system, moving 250,000 riders on busy weekdays and sending two- and three-car trains coursing through its underground core every 90 seconds at rush hour, and that’s just in one direction.
In that central subway section, where the four above-ground branches of the Green Line weave together, the vagaries of crowds getting on and off mean one train often catches up with another at a station before the lead train pulls away.
So an alarm meant to buzz or blink when trains get too close would activate routinely.
The project team studying safety technologies for the Green Line indicated as much last year in a presentation to the Department of Transportation board, meeting no objections. But the November crash injected a new urgency.
Some board members briefed at the Dec. 5 board meeting seemed taken aback by the project’s estimated cost and timeline, calling for cheaper, faster alternatives.
“My van has this little thing on the side mirrors that when a car comes up to me, it beeps and I pay attention,” board member Janice Loux said, visibly frustrated. “If there is something that will notify those guys that isn’t this nine-year, $750-million — put it on the things now!”
The consultant HNTB, working with MBTA engineers, surveyed anticrash technology makers and visited some of the world’s largest transit agencies, proposing three leading options. They also employed a computer model to simulate how those and other available technologies would affect station-to-station travel times as well as the waits between trains, producing two reports totaling 450 pages.
Their estimates for the cost included outfitting tracks, trains, and the T’s dispatching center with new technology but also investing in improvements to make up for the lost time if Green Line trips are spread apart for safety. Some changes to gain speed are simple — such as adding an extra employee toilet at North Station. Others are as complex as rebuilding track curves and tunnel sections.
But the team only got a few slides into their presentation to the board when they were rebuked for not proposing cheaper alternatives that could be deployed sooner.
Loux, a labor leader, said alarms should be installed even if they go off so often that drivers tune them out. “I want it on the trains, or I’ll challenge you, I’ll go out and find it myself,” she said.
Board member Ferdinand Alvaro, a corporate lawyer, told officials to put in more work. “Before this board makes any kind of a decision on a three-quarters-of-a-billion-dollar project, we ought to know what the alternatives are,” he said.
Davey said the team will explore lower-cost options as an interim measure while developing plans for a more sophisticated — albeit expensive — automated safety system, with no deadline specified.
“The board has articulated that safety upgrades on the Green Line have to be a priority,” he said. “One accident on any line, bus, or subway, is too many.”Eric Moskowitz can be reached at firstname.lastname@example.org.