Acela & Amtrak NEC (HSR BOS-NYP-WAS and branches only)

Horrible to see a bogie separated from one of the cars!!!

That's a real safety 'feature'. If they get launched like a missile off the tracks you don't want it on wheels to go even further afield once it re-touches terra firma with an uncontrolled landing. Or have those trucks become projectiles in their own right that could penetrate the floor if they smack hard into something. So the trucks are designed to be the first thing that detaches when shit goes horribly wrong. The underside of the car gets awfully sparky, but the frame holds while it skids to a halt. As we saw with the Philly wreck, an Amfleet frame can hold intact and preserve the integrity of the passenger cabin even when hitting a catenary pole sideways at 100 MPH and having its entire stainless steel skin torn off from the force of impact. It probably wouldn't have been survivable for anyone in that car if 2 tons worth of bogies went crashing into the car after the shell was torn off...but despite the loss of life, most people in that obliterated car walked away from it because of the by-design sequence of what broke off first to keep the frame from ever snapping.



I would love to be there in a truck munching on some donuts when they try to fish that P42 out of the ditch. Looks like the trees broke its fall gently enough that it sustained little in the way of major damage, but...whoo, that's an awkward posture it landed in a long way down. They're gonna have a fun time figuring out how to do that recovery way out in the middle of nowhere.


EDIT: apparently there is a private driveway closeby at the bottom of the ditch that empties onto the main state highway, so they should be able to get the locomotive stabilized midair with a crane hook from track level then gingerly swing it around with ground equipment to load it upright onto a flatbed truck. Probably have to cut some trees down to clear a wide enough path to the main road, but it's a LOT easier to maneuver from below than from above. Things only get nasty if it's leaking fuel while tipped in that awkward angle, but so far no reports of that being an issue.
 
Last edited:
Looks like Vermonter is only operating between Springfield and Washington, DC while they get this cleaned up. Too bad for anyone who was planning on traveling north of Springfield. I wonder how long the disruption will last. Any guesses?
 
Looks like Vermonter is only operating between Springfield and Washington, DC while they get this cleaned up. Too bad for anyone who was planning on traveling north of Springfield. I wonder how long the disruption will last. Any guesses?

Not long. That's NECR's mainline so they're going to want that scene cleaned up pronto, or else all their carloads going to/from Canada are going to have to be re-routed via Vermont Rail System out of Burlington, down to Rutland, and back to NECR at Bellows Falls...costing them an extra day on the deliveries. They've probably already re-railed and towed the rear 3 Amfleets that remained on the track off-scene and will be getting the one straddling the tracks off there tonight. NTSB is not going to be in "preserve the crime scene" mode after getting cursory look, because they need to be able to get heavy track equipment in there to secure the Amfleet that's hanging on the ledge and help stabilize the locomotive from above and below. Can't wait till nature takes its course and sends them the rest of the way down. Not when it was oversaturated ground on the adjacent embankment that caused the rock slide that caused the accident.

NECR will then come in the second they have the track clear and re-lay the broken section in a few hours' time. Then will start running its backed-up freight loads through at *very* restricted speed to clear the freight backlog while crews below take another couple days to figure out how to get that P42 down to the ground, upright, onto the flatbed truck, and down the driveway...without spilling any fuel. The trackbed itself was not compromised, so they'll be safe to run and will work on stabilizing the embankment later (it's supposed to be dry all week). They just may have to have inspectors do a drive-by on a hi-rail truck before every train to make sure the embankment is holding OK until they have the time to do a thorough job on it.

NECR most definitely gets first crack at running because they're the line owner and they're not so large that having to wait and re-route all their Canada-bound and Canada-originating carloads isn't taking its pound of flesh from them. I would guess we're not gonna see the Vermonter running until Thursday or probably Friday, since NECR would have to be given any uptime Wednesday to run a bunch of freight extras. It's a long weekend and foliage season is fast approaching peak; bigger-than-usual tix sales this weekend that they'll want to be able to honor. So no question this outage isn't going to last more than 2-3 days tops.


For the NTSB...pretty open-shut case. Act-of-nature freak accident. NECR would've already run through there once after sunrise, and the track was clear. The rockslide had to have happened as little as minutes to an hour prior, and no one would've seen/heard it since there are no abutters. The train never had a chance to slam brakes coming off that blind curve at high-50's MPH. There's no signal system here because total traffic is comfortably below the threshold for triggering the PTC mandate, but it wouldn't have made any difference because track circuits can only detect a break in the rail...not something piled on top of it. And there's only so many rockslide detectors one can deploy; one unstable outcrop, sure...miles and miles of hillside that only acts up in 10-year rain events, impossible. The equipment performed superbly; every structural safety feature in the cars and loco exceeded all reasonable expectations for a crash of that severity, and if the front two cars and loco aren't too fucked up some of them might even be able to run again after the NTSB releases them and Amtrak can see if it's worth the effort to do an in-house rebuild.

The track's in rockslide-prone country that absorbs rainfall oversaturation poorly, and as EGE notes it was only a couple miles from the scene of a gruesome Montrealer accident far worse than this one that was likewise the sole work of Mother Nature. Mountainous terrain's gonna behave like mountainous terrain is wont to behave. The Central Vermont Railway has been running through this exact spot continuously since 1848 as New England's single longest north-south line...from the docks of Long Island Sound in New London to Cantic, Quebec. It obviously has taken a lot of punches through the centuries and absorbed it well to have lasted this long as a load-bearing trans-New England mainline.


Per reports there was a charter group riding this morning's train that had the rear business car reserved. That one was packed, and the other cars relatively empty as you'd expect for early on a Monday morning. Since they board rear-to-front at St. Albans to reserve as much space as possible for the Regional/Shuttle crowds that start getting on in Springfield--and keep the front car totally closed off until Springfield--nearly every passenger was in the cars that stayed on the tracks. There's only small chance that more than single digits were in the car that went sideways...possibly none at all. And definitely nobody in the front car that went over the ledge. Probably would've been a loss-of-life accident if the boarding procedure in that part of VT was anything other than back-to-front.

The crew?...well they were the real lucky duckies today. Especially the engineer in that loco who owes his or her life--and well-deserved hazard pay--to all the tree branches that broke that thing's fall on the way down. Strong branches, and lots of them; a GE Genesis weighs 268,000 lbs. and got stopped cold in a vertical fall without ever touching the ground.
 
Not long.


For the NTSB...pretty open-shut case. Act-of-nature freak accident. NECR would've already run through there once after sunrise, and the track was clear. The rockslide had to have happened as little as minutes to an hour prior, and no one would've seen/heard it since there are no abutters. The train never had a chance to slam brakes coming off that blind curve at high-50's MPH. There's no signal system here because total traffic is comfortably below the threshold for triggering the PTC mandate, but it wouldn't have made any difference because track circuits can only detect a break in the rail...not something piled on top of it. And there's only so many rockslide detectors one can deploy; one unstable outcrop, sure...miles and miles of hillside that only acts up in 10-year rain events, impossible. The equipment performed superbly; every structural safety feature in the cars and loco exceeded all reasonable expectations for a crash of that severity, and if the front two cars and loco aren't too fucked up some of them might even be able to run again after the NTSB releases them and Amtrak can see if it's worth the effort to do an in-house rebuild.

The track's in rockslide-prone country that absorbs rainfall oversaturation poorly, and as EGE notes it was only a couple miles from the scene of a gruesome Montrealer accident far worse than this one that was likewise the sole work of Mother Nature. Mountainous terrain's gonna behave like mountainous terrain is wont to behave. The Central Vermont Railway has been running through this exact spot continuously since 1848 as New England's single longest north-south line...from the docks of Long Island Sound in New London to Cantic, Quebec. It obviously has taken a lot of punches through the centuries and absorbed it well to have lasted this long as a load-bearing trans-New England mainline.

FLine -- there is a technology which can detect these kinds of events-- back after the fall of the Soviet Union there was a program to promote dual use of Defense Technology

The FRA in connection with High Speed Rail projects funded some small-scale work to utilize radar to detect objects on the track in time to slam on the brakes -- it works in principle

And before anyone says radar doesn't go around curves it does when the waves are guided by conductors -- aka rails
 
How much does that cost though. That rail line only has a single southbound and a single northbound Amtrak trip each day and a few freight trips. I'm just not sure the cost of that system would be justifiable on a line that doesn't see very heavy use.
 
How much does that cost though. That rail line only has a single southbound and a single northbound Amtrak trip each day and a few freight trips. I'm just not sure the cost of that system would be justifiable on a line that doesn't see very heavy use.

City -- the proposed rail radar -- is vehicle based -- it doesn't matter whether a line is busy or very very quiet

As the vehicle proceeds along it sends out pulses of radio waves that travel along the rails

The receiver then listens and either gets nothing back if the line is clear
or
a reflection comes back with the distance to the reflector and closing rate is indicated -- that's what radar does best

Signal processing could extract the size of the reflecting object thus indicating whether the entire track was blocked by a landslide or just some kids throwing empty paint cans on the rails to be mischievous

Beyond that its either up to the engineer on typical rail or an automated system on maglev or other very high speed rail
 
How much does that cost though. That rail line only has a single southbound and a single northbound Amtrak trip each day and a few freight trips. I'm just not sure the cost of that system would be justifiable on a line that doesn't see very heavy use.

It's not even that. It's the Green Mountains, period. The entire range is nothing but brittle old chunks that have been flaking off ever since the glaciers receded. All that mostly new-growth forest does is put putty on the outcrops. Putty that stretches, contracts, stretches, contracts with each winter freeze-thaw and each summer heat. Saturate it with rain, and the putty gives. The Montpelier area has rockslides all over the place. It just does. Vermont was the state hit hardest by the rains from Hurricane Irene in 2011 because of all the rockslides that knocked out roads. The problems they had in getting rescue crews out there for Irene ended up not being the severity of the breaches, but how widespread they were. It was little slides everywhere over hundreds of square miles, too many for crews to know where to begin.

NECR is in a near-constant state of patching and shoring up alongside its ROW; VTrans is in a near-constant state of patching and shoring up alongside Route 12 and other highways; the utility companies are in a near-constant state of patching and shoring up the bases of their high-tension lines. FWIW, MassDOT's maint headaches on the Pike up in the Berkshires aren't a whole lot different from what's VTrans' daily struggle up in the Greens.

The area's been settled since shortly after the French & Indian War; they know which individual hills are the most-prone. Was reported today that this particular hillside that gave way yesterday on NECR has had zero known rockslides in the entire 167-year history of the Central Vermont Railway, per geological records maintained (and you can see why they kept such thorough records) by the RR's various owners from chartering to present.

I'm not sure how anyone reasonably expects Jetsons Shit tech to detect every rockslide, mudslide, and dirtslide across the entirety of the Green Mountains. They aren't easily predictable. This one not only happened on a hill that was never known to be unstable, but it wasn't a large slide and didn't happen after any unusually decade-severe rain event. You're talking like 1 dumptruck's worth of rain-soaked dirt with embedded small/mid-sized stones and tree limbs that actually reached the trackbed, with most of it sliding to a rest further down the hill from where it began but not at the bottom where the tracks are. The train blasted the pile clear off the tracks. Look at the pics; there's a few stones laying about that were ejected from the force of impact, but otherwise train > rockslide after the energy was distributed. Unfortunately, rockslide > P42 plowblade + trucks when it came to staying on the rails...so it was just enough of a pile in just the wrong-enough place at just sudden-enough time to do what it did. All you can do is be thankful that the equipment once again performed like champs at distributing the crash energy, and that luck was on everyone's side re: that train having rear-boarding first and the locomotive making the fluky-softest possible landing one can make after a nose-first vertical drop.


EDIT: This is what the un-cleared portion of the rockslide looked like...barely taller than the doorstep on that Amfleet. Now try to figure out how you expect to detect something so small that they're now saying slid almost certainly no sooner than several minutes before the train.

dt.common.streams.StreamServer.cls
 
Last edited:
My main though about usage of the line was that without knowing how the system worked what the issues just around cost compared to line usage looked like. Thanks for the information though and I fully understand the reality of rockslides in VT. I'm actually living in VT currently and was living here when Irene came through. The other thing to consider is that the track speed in the area where this happened is 55 mph which means that the train would need a very early warning of any obstruction in the tracks to stop before hitting the obstacle.
VT_Rail_Speeds_v1.2_web_sm.png
 
55 MPH coming off a curve. The engineer would've barely had time to say "Oh, shit" and press the emergency brake button before it was all over. If the lower bound on time of the rockslide was seconds, not minutes, seeing the rockslide actively happening in the distance may not have been enough time to stop the train.

NECR is derailment-prone here even though they maintain some of the most pristine track in the region. They prepare for it; they insure themselves for it. At freight's 40 MPH max authorized speed on Class 3 track and length/heft of trains they probably wouldn't have fallen overboard like the 5-car nearly empty Vermonter did...but that's life on the Central Vermont Railway. And life in general on railroads that have to go through tough terrain, of which there are thousands of miles of in the U.S. and Canada. We don't experience that in Eastern MA, but this is a way of life in both Berkshire County and large swaths of the 3 Northern New England states. It's way of life for MassDOT on the far western Pike, way of life for CSX on a couple particularly tricky portions of the B&A out in the Berkshires, way of life for Pan Am & Norfolk Southern out by the Hoosac Tunnel approaches of the Patriot Corridor. People and goods gotta move nonetheless, and I'd expect this long weekend's Vermonter is going to be packed like it usually is for a peak foliage weekend.
 
I'm wondering what's possible if you budget, say, $1 million for 1000 cameras with 802.11 at perhaps $1000 each (including solar panels to keep the batteries charged, installation costs, physical mounting, weatherproofing, etc). A typical GoPro camera costs less than $500, and so does a smartphone.

I'm thinking you wouldn't worry about the cost and complexity of having a network so that the dispatching center would be able to view all of the cameras at all times; you'd just set things up so that a computer in the locomotive could display what the next few cameras along the ROW are seeing.

Of course, that still leaves the questions of whether the cheap, readily available cameras would be robust enough (this might be where you want a GoPro and not a smartphone, but then you still need to find a sufficiently robust solar panel and make sure water coming in through the USB port doesn't destroy the GoPro), whether one camera per 500' to 1000' of track would show enough detail to actually see this landslide, and whether ``blind curve'' would overrule ``we've never seen a landslide here'' in deciding whether to install the camera in the first place.
 
I'm wondering what's possible if you budget, say, $1 million for 1000 cameras with 802.11 at perhaps $1000 each (including solar panels to keep the batteries charged, installation costs, physical mounting, weatherproofing, etc). A typical GoPro camera costs less than $500, and so does a smartphone.

I'm thinking you wouldn't worry about the cost and complexity of having a network so that the dispatching center would be able to view all of the cameras at all times; you'd just set things up so that a computer in the locomotive could display what the next few cameras along the ROW are seeing.

Of course, that still leaves the questions of whether the cheap, readily available cameras would be robust enough (this might be where you want a GoPro and not a smartphone, but then you still need to find a sufficiently robust solar panel and make sure water coming in through the USB port doesn't destroy the GoPro), whether one camera per 500' to 1000' of track would show enough detail to actually see this landslide, and whether ``blind curve'' would overrule ``we've never seen a landslide here'' in deciding whether to install the camera in the first place.

You're seriously going to do that in extremely rural areas where ALL such infrastructure for hundreds of square miles is at equally random risk? That is beyond impractical.

I said it in the last post:

If the lower bound on time of the rockslide was seconds, not minutes, seeing the rockslide actively happening in the distance may not have been enough time to stop the train.
How is a camera and how is any sort of detector going to stop this accident from happening exactly the same way in that situation? This may have exactly been that situation, except the track just wasn't tangent enough for the engineer to see it actively happening and not-stop the train in time. He/she all the same still winds up staring out the window of a P42 breaking its fall on a vertical drop with tree limbs smashing through the window. Everyone survived; there were no serious injuries. This may have been illustrative of the risks, but there was no tragedy here and no potential Vermonter patron is going to be swayed by this. If they know the region well enough to have to a need to travel to/from there, they know how to roll with the punches the Green Mts. occasionally throw their way.


We're overthinking this to an absurd degree. Either you de-populate every chippy old mountain in New England and close every transportation corridor through them, or you live with it the best you can. People in VT, NH, ME, and Berkshire County wouldn't want it any other way. I don't think us in the gently rolling hills of Greater Boston have any leg to stand on nanny-stating them so much as they have the right to nanny-state us about sea level rise. Every terrain has its challenges for people to live in: the coast, the mountains, the fault lines, the snowbelt, tornado alley, the hurricane-prone, and on and on and on. There isn't enough silicon on planet Earth to monitor it all.
 
Per Kevin Burkholder on Facebook:
Update on Amtrak Vermonter cleanup... Two of the derailed cars are in process of being rerailed; track has been rebuilt at site, awaiting ballast; rock/cliff face that caused derailment is being cut back significantly; and the fate of the Amtrak P42 #102 apparently is a write-off...to be cut up in the woods where it rests over the steep embankment. Reportedly too dangerous to raise from there and multiple fiber optic cables at risk. More close in air photos in a bit.
 
Scratch that.
Amtrak #102 Overnight Update 0234 hrs: The scrapping is under review...AND after further review, it does now sound like crews are going to TRY to drag it down the rest of the embankment, across the Bull Run stream and truck it out from Bull Run Road, making a viable attempt at recovering instead of scrapping. I'll be watching
In comments, he notes the "two lane dirt road leading to Route 12A, which has
restricted tonnage bridges....they must have some magicians with them"

https://www.facebook.com/photo.php?fbid=10207436729093468&set=gm.1095127790505143&type=3
 
You're seriously going to do that in extremely rural areas where ALL such infrastructure for hundreds of square miles is at equally random risk? That is beyond impractical.

I said it in the last post:

How is a camera and how is any sort of detector going to stop this accident from happening exactly the same way in that situation? This may have exactly been that situation, except the track just wasn't tangent enough for the engineer to see it actively happening and not-stop the train in time. He/she all the same still winds up staring out the window of a P42 breaking its fall on a vertical drop with tree limbs smashing through the window. Everyone survived; there were no serious injuries. This may have been illustrative of the risks, but there was no tragedy here and no potential Vermonter patron is going to be swayed by this. If they know the region well enough to have to a need to travel to/from there, they know how to roll with the punches the Green Mts. occasionally throw their way.


We're overthinking this to an absurd degree. Either you de-populate every chippy old mountain in New England and close every transportation corridor through them, or you live with it the best you can. People in VT, NH, ME, and Berkshire County wouldn't want it any other way. I don't think us in the gently rolling hills of Greater Boston have any leg to stand on nanny-stating them so much as they have the right to nanny-state us about sea level rise. Every terrain has its challenges for people to live in: the coast, the mountains, the fault lines, the snowbelt, tornado alley, the hurricane-prone, and on and on and on. There isn't enough silicon on planet Earth to monitor it all.

Fline -- the cab-based rail radar system would have provided all the warning that was needed -- since it can "see around curves and up and down inclines and declines" -- unless the rocks were actively falling while the train was within the braking distance in which case the crash would have been unavoidable
 
It's not the space shuttle Endeavor parading through the streets of Los Angeles, but it's what passes for excitement in central Vermont. . .

3704.1444525253.jpg
 
It is laying on its side on the flat bed like a drunken dinosaur! Wonder if they will try to fix it and put it back into service or junk / scrap it for parts.

F-Line, would you know?
 
It is laying on its side on the flat bed like a drunken dinosaur! Wonder if they will try to fix it and put it back into service or junk / scrap it for parts.

F-Line, would you know?


http://www.wptz.com/news/wrecked-amtrak-cars-set-to-be-repaired/35800290

A representative for Amtrak said all but one of the coaches will travel by rail to Bear, Delaware. The other coach will be transported by flat car.

The engine, which suffered the most damage, was the last car to be removed from the ravine. On Saturday, dozens of people watched it get hauled from the crash site to the center of Northfield. Amtrak said it will be moved by flat car to its largest maintenance facility in Beech Grove, Indiana.

Amtrak does not know how much it'll cost to repair the equipment. The rail service hopes all of the cars will eventually be returned to revenue service.
 
It is laying on its side on the flat bed like a drunken dinosaur! Wonder if they will try to fix it and put it back into service or junk / scrap it for parts.

F-Line, would you know?

Here's what it looked like after they got it upright. It passed through MA yesterday yesterday via NECR to Palmer and CSX to Albany. . .

3459.1444618782.jpg

8752.1444618920.jpg



Now, from the RR.net thread on the accident the longtime RR employees said not to get caught up too much in looks. Stuff way more 'totaled' than this has made it all the way back like nothing ever happened. The frame wasn't compromised at all except for the engineer's cabin which obviously took a large tree from all the branches sticking out of that dented corner. Most of what you see on that seemingly destroyed side is superficial damage to the skin. Because 102 leaked zero fuel or oil the engine compartment wasn't compromised at all despite the nasty fall and couple days suspended at a 100-degree angle; all the mechanical components should be fine. They gave it par odds of being repairable.

GE Genesis frames can be hard to repair because they're monocoque (single-piece) design, so that would be the only reason why #102 wouldn't run again. If by "#102" you mean the frame with #102 painted on the side. Amtrak has many stored P42's at its main shop in Beech Grove, IN...some of them sound frames that are missing engines due to fire, others like this, others that are completely totaled and just there to strip for scrap. So if 102's frame is too far gone, it'll almost certainly instead become an engine donor that gets a different stored locomotive put back into service as a like-new unit. Either way they're not going to lose one off the active roster (though I'd expect it to take a year or two before they've had enough shop time to do a special heavy-repair project like that).

That's why they made every painstaking effort to rescue it instead of blowtorching it onsite. Cut it up onsite and they end up cutting up the pristine-condition engine and throw away any shot at returning "a" locomotive to service by matching up parts, whether as #102 or under a different number. Here they're almost certainly not going to need to write off a couple million dollars worth of lost equipment.
 

Back
Top