Hey, there! Log in / Register

Green Line trolley derailed in Somerville because driver was going too fast, ignored stop signal, feds say

The driver of an outbound trolley that derailed just past Lechmere at the start of the afternoon rush hour on Oct. 1, sending seven passengers to the hospital with minor injuries, barreled through the equivalent of a red light at more than three times the speed limit and derailed on a switch that was still shifting into place to get the train to a new track, the National Transportation Safety Board says.

In a preliminary report on the derailment, which shut the Green Line north of North Station for two days, the NTSB writes:

Less than a minute before the derailment, the train departed Lechmere Station in Cambridge, Massachusetts, and headed east along the MBTA Green Line. The train entered a 10-mph zone at 36 mph, passed through a double red signal (which requires a stop) west of the Red Bridge interlocking, and reached a switch that was still moving to direct the train into the diverging track as required by its route. When the train passed over the switch, the lead truck of its leading railcar continued straight, while the next two trucks of that same railcar took the diverging track and then derailed.

The two-car train had 50 passengers and two T workers on board as it pulled out of Lechmere, the NTSB says.

Neighborhoods: 
Topics: 
Free tagging: 


Ad:


Like the job UHub is doing? Consider a contribution. Thanks!

Comments

If you were on that train and it was going 10mph you would be very angry. Driver just drove through a double re.

up
17

Yeah besides the point of this article, but a 10 mph speed limit along basically brand new track seems kind of ridiculous to me.

(And yes, I checked the slow zone tracker and it’s not a slow zone).

up
20

The green line has 10mph speed limits at all active switches (ones that are used as part of normal revenue service). AFAIK, the reason is twofold: (1) making operators more closely "read their rail" to make sure everything is aligned properly, and (2) keeping the Type 8 center trucks from derailing

up
23

About every 2 years a train gets destroyed and people hurt because of a leadfoot operator.

Last time was on Comm Ave and before that was Kenmore Sq. (Plus a few times in the downtown tunnels.)

up
33

More like a lead hand. All Green Line equipment uses a combo hand throttle and brake for decades now.

up
22

What would explain a train operator going 3.5x the speed limit and blowing thru a red light? People should lose their jobs over this. Yet another safety failure that indicates the T culture is still not up to par.

up
28

Yep you're right, by your logic we should also institute penalties for auto schools and the RMV workers who train and sign off on bad drivers.

up
16

What would explain a train operator going 3.5x the speed limit and blowing thru a red light?

Cocaine? Just sayin'.

up
15

watch that speed.

up
10

That would be my first guess, or else a medical episode.

You are not driving your personal car to work, buddy.

up
20

That’s the problem with Mass cops not enforcing any traffic laws, bus and train drivers forget they can actually get in trouble for running red lights and speeding (or running over pedestrians) if they do it at work.

up
27

SAFETY FIRST!

This does not feel like a “safety culture” at work.

This sounds like the opposite: the human capital version of deferred maintenance. (This we know.) Deferred maintenance is the result of a dysfunctional and abnormal state of affairs. Deferred maintenance is a red flag, a warning sign and it should be a top priority for any organization to zero out and not something to be persistent, or normalized.

Hasn’t the FTA’s been involved since the Red Line incidents of a few years ago and demanded x, y and z, so maybe that’s created a framework to allow the space to manage the backlog of deferred maintenance, but why is the MBTA not in receivership for lack of a better word? Is that a thing that can happen to a public transportation system of a historic, mid-sized city?

Why does this happen? Is there organizational pressure (expectations?) down the management chain to meet performance standards that exceed what I can imagine transportation experts would consider the empirical performance capacity for a nominal “safety culture” organization? Is this a case of a risk-embracing “go, go, go” culture illustrating a broken “Swiss cheese model” where: distraction, substances, inexperience, austerity and culture allows incidents to occur?

up
15

In 2008, a speeding Green Line operator ran into another train after experiencing a "micro sleep" (not, as reported, on their phone). The operator was killed and there were many injuries. A year later, there was another crash on the Green Line due to texting while operating. Based on this, the T and regulators decided to implement a train protection system (GLTPS).

There was an RFP in 2016. The project started in 2020. It's nowhere near being implemented.

It's hard to point fingers at any specific culprit for these. These projects are quite hard to implement since these systems are generally supposed to be fail-safe (most railroad systems are) and building a 100% fail-safe system is not easy. (For example, nothing about driving is anywhere near fail-safe.) The Green Line is, at least in the US, quite unique, with many different operating environments. Much of the line is unsignaled and just uses traffic signals, there are a bunch of random stop signs, other portions have wayside signals (the subway, the Highland Branch, and the GLX). But there is nothing fail-safe; the entire system relies on the operators obeying every signal and operating rule. Run a double-red* and nothing happens. Until something bad happens.

That said, the fact that the T has had 16 years to figure this out and is nowhere near figuring it out after nearly two decades and hundreds of millions of dollars. Without fully remembering what I've heard, the contractor has sort of given up, and there hasn't been a monthly update in more than a year, the last update was in August 2023 when the project was 41% complete. Maybe they're doing better and just stopped posting any updates. But the program really seems to have ground to a halt. And there is some scuttlebutt that instead of trying to build a safety system (basically something to interface with signals and not hit objects in front of the train), the T decided to try to built something which could replace drivers, and it didn't work. Basically the Bakerites (Poftak et al) said the project was ahead of schedule, then the adults started running the show and it turns out they were lying.

So now we have nothing, even though a GPS lookup which could sense speeding in slow zones would have prevented this crash, and basically the second car driver did just that, pushing the mushroom (the emergency stop) and braking the train. It might be good the train didn't make it through the switch at 40 mph and try to round a 20˚ curve at speed on a single-track viaduct. Anyway, Phil might have to declare the project dead, write it off, and try something new. But who knows.

One other thing: the T has a culture of safety theater on the Green Line. The line is riddled with 10 mph speed zones where they don't really matter: every street crossing (where they're probably detrimental: trains slow down at intersections but don't stop, cars think slowing down means stopping and turn in front, but the train doesn't stop, violating the first axiom of traffic: "be predictable."), adjacent to platforms, random slow zones and stop signs, anywhere street-running (even though nowhere else in the world has a 10 mph street running speed limit, including the T in the 1980s at least), etc. So then where it matters (interlockings, crossings, etc) drivers have been boy-who-cried-wolf into thinking that 10 mph speed limits don't matter and in this case they do.

Basically the entire operating procedure of the Green Line should be scrapped and redone from the ground up. Not necessarily track and signals. But all of the rules which have been layered on over the decades. And, yes, someone should figure out how to build a train protection system which doesn't take 20 years to implement.

* Red != Double Red. On the Green line, a green signal means the following block clear and the block beyond is no worse than a yellow. A yellow signal means to operate at an approach speed and expect the next signal to be red (you can often see this progression on the Boylston Street portion of the subway). A red signal means that the following block is or may be occupied, or that there is some obstruction, but you can pass a red signal after waiting for 60 seconds and then operating at a restricted speed (10 or 15 mph—I can't remember which—or able to stop within half of line of sight). I've been on a number of trains where we stop, I look at my watch, and 60 seconds later we start going again at a low rate of speed, and a couple minutes later pass a green signal and come back to speed. A double-red is a stop-and-stay signal, generally at an interlocking, which is that the switch is not aligned for your route and you'll derail or cause a collision if you pass. But there's no penalty for passing a double red unless you get caught (major violation, leads to discipline up to and including termination) or do what this operator did.

up
26

Thank you for an excellent reply!

I learned a lot and will learn more when I get to the links.

Even thought the T was built before I was born and none of my tax money built it I nevertheless take great umbrage at being called a “customer” and not a rider. I don’t relish the notion of a public utility being transferred into the hands of the private sector without significant limits, if at all.

“Privatization is a Weld Scheme” -one-time ubiquitous sticker on platform tunnel walls.

up
20

Very much appreciate you taking the time and effort to explain things in such detail!

up
11

That's been true of every law since the dawn of civilization.

up
15

It's on rails. It doesn't need to be steered. Machines can run it more reliably than human beings.

Whenever you put human intelligence in charge of a routine task that doesn't require human intelligence, you are asking for trouble. Humans get bored, distracted, and inattentive. Get these people out of the trains and assign them work that actually requires a human mind -- like driving a bus.

up
22

Do you know how many people would lose their phoney baloney union jobs? Many of them are cousins and step children of our estimable public servants. Pay up and shut up. Who cares if we lose a few lives here and there.

up
16

Can we stop this nonsense about the T being a haven for patronage.

This argument worked in 1984. It doesn't work in 2024.

It's just not true anymore. Yeah there are some lackeys left over from that time, but many have left the agency. Hence why there was such a hiring blitz over the past few years.

up
36

Why don't you volunteer to do this job for free! People like you think it is a privilege to be a slave!

Automation would cost a lot of money - and maybe you could volunteer to be among the first killed when it failed.

Admit it - you want tech to replace people because you profit from tech - even when it turns out to be a boondoggle like it always does. Get over your omniscient tech bro crap, Peter.

up
14

...the automation suggestion was mine, and I stated my reason clearly: when human intelligence is employed where it is not needed, boredom and inattentiveness ensue. Automation would therefore result in fewer accidents not more.

You need human intelligence to drive a car or bus; you don't need it to drive a train. Yes, it's a lot of money, but it's a lot of money spent once. Paying people to run the trains is forever, and the accidents will only keep on happening.

The most ridiculous idea of all, of course, would be to keep paying the human operators but build a machine to monitor how they do their jobs. If the machine knows how to do the job better than the human, why isn't the machine doing the job? But this is the MBTA we're talking about, so nothing would surprise me.

up
12

Okay, sure. Let's pretend The Powers That Be eventually agree that automation is a good idea. You say it's a lot of money spent "once", but have you thought about a) how MUCH money that "a lot" is and b) how often is that "once", given that entropy exists and trains will eventually need to be replaced and technology will improve?

Let's take the Blue Line as an example (since it's small, but slightly technically complex thanks to the combined panto/3rd rail situation). Automating that would involve: Getting a contract for the new cars AND their system (which would have to be custom-built for the T due to said complexity, age of the system, and because every system in the country is different), getting those cars built, installing the infrastructure to accommodate those cars/enable them to function (which involves not only station sensors, but also an overhaul of the switch system not to mention everything else on the tracks), testing the trains /while operating human-run trains at the same time/, and then gradually phasing in the new automated trains and scrapping the old ones. How long do you think that whole process will take? How much money will need to be spent?

Then there's the constant maintenance! Can the automated system deal with ice on the wires, or wet leaves on the tracks? What if some vandals decide to start messing with the cars in such a way that it disturbs the sensors? Will these new automated trains need software updates? If the automated trains are on the green line, what happens if a car decides to play chicken with the train at an intersection, or blows a red light?

I'm not saying "automated trains are bad" - there are plenty of systems that use them. But they're not the one-time painful magic bullet you seem to think they are. Every system has accidents and issues and technical problems. Automation won't magically make them all disappear.

up
11

A train control system for the Green line is in the works but since the entire system was originally designed for streetcar operation and line-of-sight manual operation, a combination of trackside control systems at signals (and elsewhere for control) and receiving equipment integrated with the cab controls of each train operator panel (2 per trolley, one on each end) is needed. That will of course need money. I believe the design for Type 10 cars now under design will be equipped with something. Depending on the cut-in date that will also mean rotating the existing fleet out for rebuild, which they do not want to do because the Type 10 units will be a purchase that will replace the current fleet in its entirety, retiring the Type 7 and 8 units. Type 9 would have a retro fit and about 6-8 of those are also planned to be shifted to Mattapan-Ashmont when the PCC units are retired.

up
17

Even though this happened on a new section, the signalling system on the Green Line is older than dirt. They would need to upgrade the system in order to have PTC (Positive Train Control). If they had PTC, this would have not happened because PTC would have overruled the driver and slowed them down.

However, looks like PTC and signal upgrades are part of the "Green Line Transformation" project. In order to get the headways they claim they want once the GLT is done, they will need PTC to do that.

up
22

There's the whole lawsuit and counter-lawsuit between the T and the contractor hired to install a PTC system on the green line with both sides claiming the other is at fault for why the system isn't installed and usable.

This is another situation where I blame the T (although not Eng) for dragging their feet for decades on something that could have prevented numerous crashes at this point.

EDIT: Looks I was confusing the PTC for the CR with the other projects for the green line.

up
16

That's on the commuter rail

This lawsuit? https://www.wcvb.com/article/mbta-hitachi-lawsuit-positive-train-control...

There isn't an actual program to get PTC on the Green Line at all.

up
14