Orange Line questions

[ThinkBoston]

Thank you, 'F-line to Dudley via Park'. I appreciate your effort.

As you deduced from the OL fleet replacement, they have intentions to operate additional trains, though only 3, not 4, as revealed in this document -->http://www.mbta.com/uploadedfiles/About ... Report.pdf. That is the source from which I made the comment about the now 'official' 4 minute headway capability for the OL.

Though I am now confused as to exactly what technological changes were made. The 2009 report states, "The initial commitment as described in the September 2000 ACO requires signal improvements be made to the Orange Line, such that peak period headways can be improved from 5 minutes to 4 minutes," and, "The signal improvements to allow for the improved headways have been completed."

Of course you cannot likely provide the policy rationale for such antiquated headways, which was a part of my inquiry. It seems the problem is that we are using an ATO system at all. Signalling is a good measure to assist in the operation of the system, but making use of automated train control is unnecessary on a system which combines such low speed operations and so little inter-connectivity among lines. The reason ATO came about was due to the risk inherent in the operation of much faster and much heavier trains on, increasingly, more intertwined lines. The Orange Line is not such a line, using any stretch of the imagination.

Perhaps the MBTA just likes acronyms. Can we can sell them on a CTO system, with a computer aboard each train to detect spacing and apply appropriate adjustments to stay within a safe distance of the train in front? It's the most advanced system in the world, though it might cost $12 million per train to install, and it qualifies as a 'fail-safe' system, which means that the on-board CTO has to take a positive action for the train to proceed; a CTO shutdown would halt the train. It's fairly easily to implement a Cerebral Train Operation (CTO) system.

I appreciate the dollar figures, and the link, regarding the supposed expense to upgrade to an ATO that allows 2 minute headways. Of course the Green Line already operates with a headway less than 2 minutes, included in that scenario is that the outbound 'E' trains literally crossover the mainline inbound track when branching off (unless that's changed since the time that my source was dated). Boy, the Green Line must have some super duper futuristic beyond all universal intelligence thingamagig being used for its operation.

I've browsed for the cost of laying fiber optic cables, and it doesn't even approach the wild numbers given for the upgrades. Therefore the money must be in the on-board computers, but you get some MIT nerd types together and they'll create all of the software and hardware necessary to operate the system so that it'll even detect which passengers want coffee and danish and serve it right up.

In the meantime, let's go with the CTO system, cause we might can get it a lot cheaper than the $12mil/train estimate.

And, yes, several of the commuter rail projects, as does the GLX, stink. Gargantuan waste of funds; which is reason to get upset every time I read a comment about the MBTA not having any money.

[F-line to Dudley via Park]

The signal improvements the OL just made were installation of ATO from Haymarket to Oak Grove so the whole line ran on the same system. Prior to that the subway and 1987 Southwest Corridor extension ran on ATO while the 1975 northern extension still used the Blue Line-style trip arms + wayside signals. The operator was required to manually turn on/off the ATO box at Haymarket. The trip arms and heaters were at end of useful life, and were inadequate for traffic volumes. Prior to 1987 when the southern end still ran over the Washington St. El the whole line was trip arms and there were fewer stations, so the load and travel time on the north end changed dramatically when the south end of the line was remade. The signals needed replacement. They were years late initiating the project, though, and replacing all of the cabling slowed it down so much further that it finished years behind schedule. Really, it all should've been running ATO by the late-90's but ended up taking an extra 10 years to fund. But the ATO is thankfully controlled by new fiber optic cable there so if they ever did upgrade to CBTC they won't have to replace that cable to install it Haymarket-North, only the cab signal track mounts. That's a ton cheaper than the wholesale replacement they have to do everywhere else.

-----------

Now, the car fleet expansion matters because the 120 Hawker-Siddeley cars also date to pre-1987 when the Washington St. El was operating, and the whole line was operating with 4-car trains. The supply has been completely inadequate for over 20 years, so they are not able to max the headways today under the current ATO with what they've got. It was thought for awhile that they'd be able to rehab 24 Blue Line 0600 cars when those were replaced to go on the Orange Line (same Hawker-Siddeley car make, just slightly different car lengths) and provide the needed extras to trainline with the 01200's. But the rust was too bad on them to retrofit and rebuild, so the fleet needs a full replacement. 146 cars would make a substantial difference, as +3 operating sets on the line is 1,044 more passengers it can carry at any given moment at peak hours. Plus the spares that they can run more off-peak when they're doing maintenance on the ones out-of-service. Which in turn is going to create more overall transit ridership than just 1,044 bodies because fewer packed, long-wait cars will encourage some folks who don't ride at all to start riding.

-----------

Going whole-hog with a CBTC install is probably too big a leap for now. To get utilization of the 2-minute headways you'd have to increase the size of the car fleet by well over 50 cars and have the power capacity and maintenance space to handle all that. That's why the numbers are so expensive. Say you pay for CBTC at the going rate for just signals. So what? You still have the same number of cars on your overstretched fleet, the same crumbling old power draw that can't support running more consists on the same circuit breakers, and the same yard space and carhouses to run everything out of. To even tap the upgrades you have to spend several times over what the new signals cost to buy new cars, expand the maintenance facilities, and upgrade the power draw to get any payoff out of it. If those secondary factors don't support more than what ATO runs, then there's no point installing more than ATO. The current Orange Line and Red Line downtown power cables and main power feed are in such terrible shape that they are a huge fire hazard the D'Alessandro Report slammed last year as one of the biggest safety risks on the entire T. It's going to cost a fraction of a billion to replace, handle current loads without risk of power outages or fire, and upgrade the trunk line for future-proofed capacity that would support future upgrades like 2-minute headways. All of these extra contingencies make it a massive infrastructure investment on a 100-year-old system. When they're this late on a life-and-limb fix like the power feed deathtrap they don't have choices to dream about future signal tech. It's far too big a leap for their means. But, as mentioned, the new northside ATO is fed by high-bandwidth fiber so they've already fought half the signal installation battle when it's time to consider CBTC. So going ATO was the correct decision for now vs. band-aiding the old trip arms for another 20 years.

-----------

The Green Line is totally human-controlled and operates line-of-sight with manual block signals. That's the only way it can retain short headways, but it is a literal 19th century system with some critical limitations:

1) Capacity is constrained because you have to stop and protect every few feet when line-of-sight is compromised by curves, junctions, station approaches, etc. Even when there is verifiably no train on the next block. It's the opposite extreme of operator overcompensation from the Red Line's too-long ATO blocks.
2) Operator error is a constant risk. There are no trip arms, just wayside signals. If they miss a red because they're not paying attention...*KABOOM*. The line's had 1 fatal accident and 1 very high-profile non-fatal accident in the last 3 years from operator error/distraction and speeding. The fatal accident has required a speed reduction on the D line as further overcompensation.
3) Little system redundancy. Operators can be fully at-attention and not distracted at all, but still be at risk of collision due to signal failure. A burned-out bulb, or even low-visibility weather on the D making it extremely hard to see the signals. When a signal fails too soon for anyone but the next train to notice, that's a particular problem. ATO/CBTC/trip arms at least throw up an automatic stop when the signal system itself gets tripped by a suspected problem on the next block. The operator has warning from the auto stop to proceed manually at 5-10 MPH line-of-sight through the next block.
4) No central traffic control. Green Line schedules are managed pretty much the same way they were a century ago: pen, paper, walkie-talkies, and human inspectors stationed along the line. The inspector's got to make the scheduling decisions on judgment almost like a bus dispatcher would, which is why you have late trains randomly short-turning and expressing. That's what happens when the trains can't detect a signal and the signals can't detect a train...the humans make all the decisions. The stop signals are spaced to paranoia to overcompensate for potential of human error happening on dispatching or operator end at any moment on a scheduled trip.

That's not a rapid-transit signal system, and the Green Line is failing to do its duty because it has to be a rapid-transit line on the D, subway, and new northside extensions. It has the most to gain from a CBTC system catered to light rail. Automatic speed enforcement alone ups the line's safety a lot by removing the most serious operator error risk. Bi-directional communication eliminates the pen-and-pad dispatching, which will make things a lot more efficient. A line with that many vehicles and branches needs computer-to-computer decisions to manage the flow and bunching. And it'll eliminate a lot of cascading time eaten up on the schedule by mandatory line-of-sight stop and protects because the signal system can see the next block and the operator is prevented from blowing a red. No more of that random pause between Boylston and Arlington on the outbound side or Kenmore-Hynes inbound, and a lot better-managed congestion at Copley Jct. And it cushions the subway schedules from the delays that the signal system can't control, like road traffic fouling the B/C/E. To have the D, subway, Huntington Ave. subway, and northside extensions all automated keeps the B from being the proverbial monkey wrench constantly screwing up the whole works.

-----------

CBTC is a Positive Train Control system, so exactly like the CTO you describe. Is a little acronymy, as what the rest of the world calls "ATO" really means "driverless trains", while what the T calls "ATO" is just basic "Automatic Train Control" continuous speed enforcement. This is more like the NEC's ACSES overlay coupled with an ATO system having greater bandwidth to send different commands either direction. The underlying tech isn't revolutionary at all...more like the 'Version 3.2' evolution of systems that've been in wide use for decades, just beefed up with more nimble computer back-ends replacing the "dumb" analog controls and with the various systems integrated together a lot more nicely. But no legacy metro system is ever going to have driverless trains or something bleeding-edge like that running through centrury-old subways. All-new transit system builds can consider that; the systems with a mishmash of legacy infrastructure have to keep it simple and make evolutionary upgrades with what Just Works. Like Amtrak did putting retrofitting Pennsylvania RR's venerable old cab signals with ACSES. This kind of "Super ATO" is the only thing that would work with any kind of reliability on all the big classic systems like Boston, New York, Paris, etc.

[jonnhrr]

CBTC has not fared well in the Philadelphia subway-surface system (similar in concept to our Green Line) although I don't know if that is due to the technology itself or how the system was implemented. That makes me a little leery of CBTC.

Jon

[F-line to Dudley via Park]

CBTC has not fared well in the Philadelphia subway-surface system (similar in concept to our Green Line) although I don't know if that is due to the technology itself or how the system was implemented. That makes me a little leery of CBTC.

Jon

SEPTA's is RF based. The only track circuits they have are the unmodified old system that switches the wayside signals. That's why it's been such a debacle. PTC test installations that have relied on RF or GPS in lieu of track circuit hardware have had similarly rocky start despite the technology in theory being straightforward. Whereas ACSES has been rock-solid from Day 1 because it's nothing but an overlay on the old PRR pulse-code cab signals perfected in the 1920's. Fixed track hardware is always, always more reliable than wireless.

SEPTA got distracted by doing something futuristic for futuristic's sake when the same exact thing is accomplishable with tried-and-true track mounts driven by computers instead of analog. They richly deserve the criticism they're getting for that piece of crap system. Which is only CBTC in the barest sense of having computers. It really does absolutely jack more sophisticated than our heavy-rail analog ATO, and they have to maintain all of the old wayside signals in their entirety because of the reliability problems. Another reason it can be ridiculed as a shiny hi-tech thing that serves little purpose beyond being a shiny hi-tech thing that is expensive and therefore "better".

What the T would utilize, and the MTA L train CBTC trial does utilize, is the heavy-rail equivalent of ACSES. Computer-to-computer communication, digital instead of analog signals for wider bandwidth...but same old pulse code cab signal method and hardware that we've been using for 25 years on Red and Orange (albeit with replacement track mounts and cabling that can pump out digital instead of analog). This stuff can even retain dual operation of the analog ATO cab signals on the same feed so the old system can stay in full concurrent operation until all cars have been switched over to computers and absolutely 100% of the new system's kinks get ironed out. Without the senseless waste of SEPTA being backed into retaining a completely incompatible backup wayside system with completely different hardware. That's how much track circuit CBTC is just an evolutionary upgrade from the past system. Much like there's still segments of the NEC with newer ACSES installations that have sunset exceptions allowing use of ACSES-less commuter locomotives for a couple years until all fleets are upgraded. What SEPTA did junking the proven and going with a complete leap of faith into untested wireless was completely unnecessary in the face of more obvious battle-tested options.

[ThinkBoston]

F-Line to Dudley, I appreciate you taking the time to share your knowledge. I've been trying to further an argument for scraping the 'as-is' GLX in favor of using the OL up through to West Medford, while reserving the Lechmere stub of the GL for a proper use of light rail along the main thoroughfares in Somerville and Cambridge. While on balance the OLB (OL Branch) is far superior to the GLX, comparing costs and benefits. But, it is excruciatingly difficult to get people to understand balance (a result of them not having to incur the costs). So, I'm attempting to further improve the OLB plan.

I think (but am not certain) you understood my reference to CTO (Cerebral Train Operation), i.e. human control. And, as you described the Green Line operation, that is what we use there, which is fine for about 75% of the Green Line operation. The Green Line suffers from human politics, and it's only getting worse.

The one theme which kept recurring in your post is that there are both crucial and critical upgrades and/or maintenance duties which the MBTA has chosen not to perform. I say chosen, on account of their decisions to instead support the financial backing of projects which do not address those needs, and which are often tragically poor and absurdly expensive solutions to other transit needs: GLX, South Station Expansion, more Commuter Lines, etc.

I suggested in my previous post that we should allow the Orange Line to use CTO (human judgment) instead of ATO, meaning that, for such slow speeds and a single file route, the line should be operated under the control of a human operator, with the aid of track signalling. For some reason, we allow this operation on the Green Line and yet not on the Orange. Why? Is it only that the MBTA has not had the foresight to budget for an increase in train service, and thus have more trains available to run smaller headways? If so, skip the next two paragraphs.

If it simply policy, it seems that there is extreme over caution on one line and then extreme hazardous operation on the other. The minor differences in speeds do not justify the major discrepancy in headways and control. Even going with the newly acknowledged minimum headway for the Orange Line of 4 minutes, that 240 second headway is way out of proportion to the average of 94 seconds on the Green Line, which has in addition an at-grade junction aside from the other branches.

How can a sane person justify operating one system with four service lines and no automated stopping at AVERAGE intervals of 94 seconds while insisting on using automatic breaking on other systems with 240 second intervals? If the MBTA insists upon using the ATO on the Orange Line, then it should create smaller blocks, it's that simple. It should never have implemented an operating system which was going to so drastically limit its capacity, and at the very least should acknowledge that error now in order to provide better service at a reasonable cost to the public.

If the condition on the Green Line is so insane, why are they spending $1 billion to add more demand to it? They plan on running 6 min. frequency 3-car trains to College Ave. or Mystic Valley Pkwy, which will create a worse supply to demand ratio than the OL, RL or BL are able to currently provide on any of their branches. Without counting Lechmere, boardings on the West Medford extension are predicted to be 13,149/day for which a peak hourly supply of 3030 will be provided with the 'D' line, that's more crowded than the Alewife and Braintree branches of the Red Line.

Of course today, the 'E' and 'C' run to at least North Station, with 'D' and 'B turning back at GovtCenter, but with the new use of the 'D' branch on the extension, it's likely 'they' planned to cut the 'C' service back to GovtCenter. That plan will be short lived, as they will most likely have to send either 'B' or 'C' trains to run on the extension (with the 'E' already serving the Union Square branch). That means that three lines will be running all the way to Lechmere. That will mean the Green LIne's trunk line with 3 or more service lines operating increases by 1.85 miles, from 2.3 to 4.15 miles (to where the Union Square branch forks off).

With safety measures greater than those of the Green Line, the Orange Line could currently operate easily with headways of 3.5 minutes and less. These aren't massively heavy trains barreling along at 60+mph. Scrap the ATO's premature automatic braking and use rationale measures to 'drive' each train safely. A 3.25 minute headway would have 26 trains running during peak hours. The best use of the Somerville/Medford rail corridor is with one of the heavy rail lines. Instead of wasting the money to elevate a new Lechmere Station ($31mil) and extend the GL viaduct ($64mil), a perfectly sensible connection to the Orange Line sits waiting to be made. $95mil would go a long way toward upgrades to the Orange Line.

With 9 of 26 trains running to West Medford (and actually ALL the way to West Medford), it would provide much greater hourly capacity than the Green 'D' Line could, 5,023 v. 3,030. The only thing favoring the GL would be the frequency, but if cars are filled up it nullifies the difference. The West Medford branch frequency would be little more (25 sec.) than that of the Ashmont and Braintree branches which have 42,000 daily boardings. Malden would remain as it is today with 17 trains and a 5 min. headway. Frequency could increase when a proper implementation of ATO or CBTC on the Orange Line provides the option to add more trains.

The money being wasted to insert the 'square' Green Line into a 'round' hole best suited for heavy rail could be used instead to upgrade the power supply or implement sensible ATO technology on the OL. Other savings would be found in that Green Line (light rail) coaches are much more expensive that heavy rail coaches.

From a MBTA document, dated September 20, 2010:
At its meeting in April 2008, the MBTA Board of Directors approved the most recent Capital Investment Plan (CIP) for Fiscal Years 2009 through 2013. In this CIP, the MBTA has programmed over $133 million for overall Orange Line vehicle procurement.(2) The prior CIP (FY 2008 through 2012) had $73 million for the Orange Line project.
Footnote 2: These costs include $80 million for the vehicle procurement [146 coaches], $10 million to make the signal systems compatible with the new vehicles and over $43 million for ancillary improvements associated with the new vehicles such as power upgrades, bridge modifications, modifications to the maintenance facilities, etc.

[jonnhrr]

Thanks for the explanation F-line. SEPTA seems to have a reverse Midas touch, I have read many complaints also about how they implemented cab signaling on the regional rail, seems everything they do just makes the system slower.

Hopefully MBTA will learn from SEPTA's mistakes.

Jon

[MBTA3247]

Instead of wasting the money to elevate a new Lechmere Station ($31mil) and extend the GL viaduct ($64mil), a perfectly sensible connection to the Orange Line sits waiting to be made. $95mil would go a long way toward upgrades to the Orange Line.

Most of that money (if not all of it and then some) would probably be used to build the bridge or (more likely) tunnel needed to bring the OL from Community College across the Eastern and Western routes and then along New Washington St to where it can get onto the Lowell Line ROW, with all the intendant mitigation needed to keep commuter rail service running during construction.

[BostonUrbEx]

What if the Lowell Line continued straight south instead of curving around BET, crossing the Fitchburg, and curving back in along the Fitchburg approach tracks to North Station. Then the OL could have a relatively brief hop over the Western Route/Eastern Route approaches to the Lowell's current embankment. Only thing is, OL headways from Sullivan to Malden should definitely be maintained if not improved. How many new trains could possible introduced from North Station to Forest Hills to accommodate the new branch?

[MBTA3247]

What if the Lowell Line continued straight south instead of curving around BET, crossing the Fitchburg, and curving back in along the Fitchburg approach tracks to North Station. Then the OL could have a relatively brief hop over the Western Route/Eastern Route approaches to the Lowell's current embankment.

Doable in theory, but I think the old yard space has been sold off, even though it hasn't been redeveloped yet.

[ThinkBoston]

Instead of wasting the money to elevate a new Lechmere Station ($31mil) and extend the GL viaduct ($64mil), a perfectly sensible connection to the Orange Line sits waiting to be made. $95mil would go a long way toward upgrades to the Orange Line.

Most of that money (if not all of it and then some) would probably be used to build the bridge or (more likely) tunnel needed to bring the OL from Community College across the Eastern and Western routes and then along New Washington St to where it can get onto the Lowell Line ROW, with all the intendant mitigation needed to keep commuter rail service running during construction.

This is what disappoints me about this forum, a lot of assertions as to extravagant costs without any evidence to support it when wishing to put down one argument, while absolving favored arguments of any costs-benefit assessment no matter how horrific their costs.

[Finch]

Regarding signals, headways, and safety on the Orange Line.

From my understanding, ThinkBoston, your concept for Orange Line signalling would rely on the operators to slow or stop their train based on a prompt from the signal system. In other words, there would be no automated interference with a train's operation if it violated a signal. Am I understanding you correctly? If so, I just don't see such a system being palatable to many people. In previous posts, Green Line crashes caused by operator error were mentioned. Nobody wants that possibility to exist, regardless of the responsiveness of the vehicles, speed of the line, or any other factor. The consequences of a big operator mistake are dramatic. For many people inside and outside the industry, that's all they need to know.

It's worth noting that a heavy rail train is significantly less responsive than a light rail car (or mated cars). As you suggest, slower speeds make this less of an issue. But a rapid transit line is supposed to be just that: rapid. Might there be a conflict between the urge to make the trains go fast and the riders happy, versus maximizing capacity by operating trains close together and at slow speeds?

[3rdrail]

I agree with ya, Finchy, and if I may, I'd like to add that it seems as if Think Boston is not really thinking Boston at all. The Boston system is a catch-up, patch-up, make-up system. It's over a hundred years old. Were the lines all newly constructed with a world of possibility ahead with all of them, I'd would say that what is done for one, should incrementally be done for all-particularly where safety is concerned. The problem is that that's not the case. Everything should be ATO'd. Historic tragedy has proven that. We're dealing with infrastructure which was conceived and developed in many cases not long after the Civil War. And to add insult to injury, all of the systems within the system are different. What works in one does not necessarily work in the other, or at the very least, tedious stop and go testing must be done from start with each and every one. I won't even get into budgets and politics, which is another whole quagmire which the MBTA has to field. I wish all was as cut and dry as you protest, but alas, it ain't.

[BandM4266]

Forgive me if I have the wrong car number! But on what I beleieve was 01214 there are 8 small rectangular boxs on the ceiling. 2 near either end and 4 clustered in the center door way. Are these Security cameras or something else?
I did not see anything on the other cars in the set I was on. I would also send a picture but have no idea how to send from my phone.

[typesix]

It's worth noting that a heavy rail train is significantly less responsive than a light rail car (or mated cars).

What do you mean by less responsive? Other than emergency braking rates, modern streetcars and heavy rail cars have similar acceleration and braking rates.

[BostonUrbEx]

Forgive me if I have the wrong car number! But on what I beleieve was 01214 there are 8 small rectangular boxs on the ceiling. 2 near either end and 4 clustered in the center door way. Are these Security cameras or something else?
I did not see anything on the other cars in the set I was on. I would also send a picture but have no idea how to send from my phone.

They are indeed security cameras. There are 2 or 3 cars that have them (I'm pretty sure it's just 2, and I'm not sure if it is a married pair or not).

It's a pilot program which I'm sure we'll see expanding soon.