Green Line Derailments Increasing

Here's the story from the Boston Globe, course some of you already knew this:

http://www.bostonglobe.com/metro/2016/1 ... b#comments

The author of that article clearly never took a statistics class though. I wouldn't be too surprised that the Green Line had more derailments than any other light rail line in the country, because it's also the busiest light rail line in the country. They really should be comparing derailments per VMT, or some similar metric. Or even not comparing it at all because you're dealing with so few data points it's impossible to establish any trends. The sample size is way too small to draw any meaningful conclusions.

You also need to consider that much of the Green Line's infrastructure is 100+ years old, while most light rail lines in this country date to within the last 50 years. And most other lines aren't nearly as complicated in terms of branches, tightly packed in yards, etc.

Just more MBTA-related fear mongering.

Half fear-mongering, half truth. Boston and San Francisco do have a derailment problem from the Bredas, and the solution to that is the wheel grinding and track grinding that they've failed to keep up on. If the Type 9s are good when we get them starting next year, the state should immediately pounce on buying enough to purge the Type 8s and then the Type 7s. The Type 8s derail a lot in revenue service; most other systems tend to have most of their derailments be minor yard accidents.

But until then, they need to keep up basic maintenance, and not settle for slapping a 25mph speed restriction every time a derailment happens. The D Branch is 10 minutes slower one-way than it was in 1959, almost entirely from speed restrictions added over the years. That's what the fear-mongering should really be about: the MBTA does not take care of - and is not allowed enough money to take care of - its tracks and vehicles. The D branch was once a 50mph line almost its whole length, and the Braintree Branch was 75mph. Both have been downgraded to save money, all at the expense of their riders.

Braintree branch never ran 75mph. When Quincy first opened, speed was 70mph and was decreased later to 50 mph. The Silverbirds originally were to be 75 mph, but Pullman-Standard convinced the T to accept 70 mph cars.

EGE nailed it!

deathtopumpkins wrote:You also need to consider that much of the Green Line's infrastructure is 100+ years old, while most light rail lines in this country date to within the last 50 years.

This is such a ridiculous statement, especially following such mind-numbingly detailed analysis. What is greater than 100 years old is a hole in the ground and supporting columns which keep it from filling back in. Every aspect of the system which contacts/interacts with the streetcars is much newer. How many years ago the system was originally built is completely irrelevant, except perhaps to say that by now we should really have this down to a science.

The age of the system is ABSOLUTELY relevant, when it means that you have curves tighter than anywhere else in the country, and when it means that you have clearances that require custom-designed equipment.

Also, no, there probably isn't any track or other hardware that dates back to the 1890s, but it's not like every so often the T rips the entire thing out and rebuilds from scratch. You're always going to be a lot more prone to things like minor derailments on a system that's been continually in use for 100+ years, even if you maintain it and replace things periodically, than on a system that was built new, from scratch, with generous clearances and curvature, intended for modern light rail vehicles instead of trolleys from 100 years ago.

What's ridiculous is you ignoring that.

What is the effect of all this repeated grinding on the service life of the rails and wheels? How much extra is this costing vs. the cost of running other types of equipment?

It actually extends the life of the rails and wheels by removing surface irregularities that create bumps/jolts/bounces/etc that cause more serious kinds of damage (like microfractures). A proper wheel-rail interface also improves the efficiency of the trains.

Grinding should be done on a periodic basis regardless of what equipment you're running, but if you have to do so more frequently to handle some types of equipment, then your extra cost is whatever the cost of a grinding job is multiplied by however many more times you have to perform grinding over the life of the wheel or rail.

deathtopumpkins wrote:The age of the system is ABSOLUTELY relevant, when it means that you have curves tighter than anywhere else in the country, and when it means that you have clearances that require custom-designed equipment.

Also, no, there probably isn't any track or other hardware that dates back to the 1890s, but it's not like every so often the T rips the entire thing out and rebuilds from scratch. You're always going to be a lot more prone to things like minor derailments on a system that's been continually in use for 100+ years, even if you maintain it and replace things periodically, than on a system that was built new, from scratch, with generous clearances and curvature, intended for modern light rail vehicles instead of trolleys from 100 years ago.

What's ridiculous is you ignoring that.

Actually, when the Bredas were suspended from service they had to do painstaking amounts of rail grinding and renewal throughout the Green Line because part of the fix for the derailment issues was resurfacing track to be just...so with a much narrower tolerance for imperfections than the cars were originally supposed to support. It's why when they did finally start getting flushed into service in halfway-decent numbers they could only be introduced on one branch at a time. The rail grinder was working overtime across the system to get them certified for service.

That was 12-15 years ago. Track that was just so in 2002 is no longer quite as so in 2016. That doesn't faze the Kinkis one bit because they're designed to tolerate a much wider and cruddier range of track conditions. And PCC's will run on godawful spaghetti rail without flinching. Track across the Green Line is in very good condition by any league-average measure in the transit world because that systemwide grinding blitz is still recent enough history on rail that on the most tangent stretches hasn't come close to a league-average replacement age. But it's a state-of-repair problem all the same because the Bredas didn't live up to the track condition tolerances the T thought it was getting upon purchase. They are requiring escalating maint intensity for the whole works--the vehicles themselves and the track structures--as the cars age. And it is not a normal aging curve. The tolerances were supposed to be a lot more generous than this.

With the whole system hurting for SGR funds spread incredibly thin and far too many unfilled job vacancies in the MOW ranks offering up too few overnight man hours for staying on top of these too-narrow tolerances, the track is teetering at not *quite* so. Is that bad? Sure...derailments inching up over a 3-year span isn't a good trend. Are they being flat-out negligent with the current rate of rail resurfacing? Not so simple an answer. Because the Kinkis are very well within current track condition tolerances...the Boeings were very well within current track condition tolerances...and the new CAF's are quite likely by virtue of a much-improved truck design to be very well within current tolerances. The 8's, by product of all the compromises that had to be made to get them safe to run, are big outliers there. This was a known worry for when the Bredas hit their teenage years, because anything less than just so becomes more a reliability unknown when moderately worn trucks have to stay on moderately worn rail. They were always going to be running up an ever-escalating treadmill. And that's why the Type 9's are not a moment too soon...and if the 24 base units and +30 options are good performers they need to immediately plunk down for the triple-digit whammy follow-up order with CAF to get this replacement marathon started. MOW can only do so much more to already pretty decent track to make it totally immaculate before even that's not going to be good enough to stay a step ahead of attrition on >20 year old trucks (and afteraffects of that attrition wearing down the trainlining reliability of both halves of the 7-8 lash-up).



Now...note who's a conspicuous #2 on that derailment list in the Globe article: MUNI, the other very dissatisfied Breda customer from the same era. They don't have the issue with the trucks design that the T does because their cars don't need to conform to nearly as narrow width. But MUNI's order were morbidly, stupendously overweight for what they thought they were ordering. So guess what...MUNI too needs more immaculately-surfaced and regularly-surfaced track than first anticipated because those cars extract a bigger pound of flesh from the infrastructure than they were ever supposed to. MUNI's behind the MOW 8-ball too. All the mass tie replacement they did to minimize the amount of bouncing those porkers did on the trackbed upon first being introduced is now starting to near re- replacement age for wood and various types of streetcar sleepers (lot of theirs also wood beneath the asphalt). So they're seeing an unfavorable derailment trend too, for similar reason of the vehicles performing below the default track tolerance spec...not the agency being negligent about staying on-spec. It's an above-and-beyond induced hardship on them too.

CRail wrote: This is such a ridiculous statement, especially following such mind-numbingly detailed analysis. What is greater than 100 years old is a hole in the ground and supporting columns which keep it from filling back in. Every aspect of the system which contacts/interacts with the streetcars is much newer. How many years ago the system was originally built is completely irrelevant, except perhaps to say that by now we should really have this down to a science.

It's nice to come back and get a reminder of why I stopped checking in here regularly.

FCMB presentation to the board about the derailment issues: http://www.mbta.com/uploadedfiles/About ... 20POST.pdf" onclick="window.open(this.href);return false;

7 of 10 derailments in the reporting period involved Bredas. Of the 3 Kinki derailments, 2 did not involve any physical plant faults: 1 was result of a large snow/ice mound blocking a grade crossing after one of the 2015 blizzards, and 1 was a result of human error from a mis-thrown switch. 5 of the Breda derailments were track condition-induced (i.e. the tolerances issue described above), while only 1 was a purely lineside fault (faulty switch).


In the meantime, expect speeds to get slower and slower because one of the few moves they have as the cars age is to enforce speed limits ever more strictly.

deathtopumpkins wrote:The age of the system is ABSOLUTELY relevant, when it means that you have curves tighter than anywhere else in the country, and when it means that you have clearances that require custom-designed equipment.

Much of the central subway was designed to be eventually converted to RT, and the oldest part of the central subway actually carried Elevated trains (note Kenmore having a similar track plan to the old Ashmont). Gerry used to indicate what the actual tightest curves and such were but I don't remember them, if only he'd chime in . Outside of loops I believe the tightest curve is Boylston, where I've not heard of a derailment occurring. Streetcar equipment, new and old, is designed to navigate tight loops and 90 degree turns at intersections. The latest two types of standard American streetcars, the Presidents' Conference Committee car and the United States Standard Light Rail Vehicle, operated in Boston as did PCC cars built for service in Dallas and unique cars built specifically for Toronto, so the "custom-designed equipment" argument is baseless. Age of the system is irrelevant, maintenance practices are not.

deathtopumpkins wrote:Also, no, there probably isn't any track or other hardware that dates back to the 1890s, but it's not like every so often the T rips the entire thing out and rebuilds from scratch. You're always going to be a lot more prone to things like minor derailments on a system that's been continually in use for 100+ years, even if you maintain it and replace things periodically, than on a system that was built new, from scratch, with generous clearances and curvature, intended for modern light rail vehicles instead of trolleys from 100 years ago.

Nonsense. You don't build a system based on what equipment you plan to run on it. The tail doesn't wag the dog.

deathtopumpkins wrote:What's ridiculous is you ignoring that.

Boylston curve, tighter inbound side, is actually a relatively generous 81 ft. curve radius. That's basically par with the turning radius of an off-shelf 100% low-floor car. Park St. loop @ 42/50 ft. (it's irregular-shaped east vs. west ends with different curvature) and Lechmere inner loop @ 45 ft. set the design standards for the Green Line. Boylston isn't a design constraint for the cars, which is why it's had relatively few derailments over the years. It's strictly a speed constraint. But speed constraints are why the Bredas are not derailing at increasing rates there while they are on more tangent, less-restricted segments.

Didn't they screw up ashmont or mattapan turnaround radius?

F-line to Dudley via Park wrote:FCMB presentation to the board about the derailment issues: http://www.mbta.com/uploadedfiles/About ... 20POST.pdf" onclick="window.open(this.href);return false;

7 of 10 derailments in the reporting period involved Bredas. Of the 3 Kinki derailments, 2 did not involve any physical plant faults: 1 was result of a large snow/ice mound blocking a grade crossing after one of the 2015 blizzards, and 1 was a result of human error from a mis-thrown switch. 5 of the Breda derailments were track condition-induced (i.e. the tolerances issue described above), while only 1 was a purely lineside fault (faulty switch).


In the meantime, expect speeds to get slower and slower because one of the few moves they have as the cars age is to enforce speed limits ever more strictly.

Wow. What design changes were made to the center trucks besides grinding the wheels? Maybe the center trucks should be "computer steered"? I imagine this could be done with four hydraulic pistons...of course I am not a mechanical engineer, lol!!

How do the Type 9's avoid this problem?