CBTC on Mattapan Line

[Gerry6309]

There is a way to straighten out the northern section of the Green Line by making the present NB Brattle track the SB main. This would require some subway construction at the north end of City Hall Plaza to avoid a tight curve there. You would give up Brattle and Government Loops and most of the useful platform space at Govt. Ctr. Park Street could become an island platform with the southbound altered to meet the existing fence rail across the present loop. No more loop, no more ability to pass/hold trains. One breakdown is all that it would take to cripple the system, since everything would run the entire length of the subway. The existing system may be a dinosaur, but it works.

FACT:

It takes 30 seconds to run around a loop.

It takes four minutes to change ends, assuming that the new cab isn't f…d up and that the dead end isn't full.

THE GREEN LINE IS 125 YEARS OLD and GLX is essentially putting something back that was eliminated to speed up service in 1932!

Those who don't learn from the mistakes of the past are doomed to repeat them.

[BandA]

I don't understand why the GLX is attached to the green line; Both green and orange meet at North Station, they could have converted Science Park to high platform & ordered new Orange cars with 3rd rail + catanery like the Blue line (used to have?).

And attach "D" to Orange line (would require new tunnel from Hynes to Back Bay, diagonally under the Pike, about 1 mile?, and moving Orange under Commuter Rail at Back Bay, increasing capacity at that station)

That would allow breathing room on the Central Subway for more B, C, E service, and converting Roxbury Silver Line to Tremont line streetcar.

[Bramdeisroberts]

Three P's

Perception

Pretension

Politics

When you're trying to sell a new T-line to the transit-ambivalent (if not transit-phobic) folks of West Medford, the last thing you need to do is go filling their heads with the idea of being just a few subway rides away from "those people" in Jackson Square, Roxbury Crossing, etc. Folks from Brookline, Brighton, BC, and Newton are a lot less scary.

It's the same reasons that killed the red line extension to Arlington and Lexington.

[The EGE]

Heavy rail was never a plausible option, for engineering reasons. A Blue Line extension was considered and rejected early in the alternatives; it would have been costlier to add a new tunnel under the Charles for no additional benefit.

An Orange Line branch would have been even worse. It would have halved frequencies from Community College north (a very big deal), cost more to build (heavier bridges, longer stations, etc), and also required a new Charles tunnel. The Green Line takes a tightly curved and very steep path to get from North Station to the viaduct; I don't believe you could run Orange Line stock in regular service on it. Orange Line trains may also be too heavy for the Lechmere viaduct. The Orange Line could not have possibly supported branches to both Medford and Union Square; in order to serve both, it would have required a tunnel under Winter Hill (an alternative rejected for the Green Line as being much more expensive). Additionally, using the Orange Line would have eliminated any chance of potential Green Line flanks to Porter, Sullivan/Everett, or the Grand Junction.

[MBTA3247]

The Lechmere viaduct was originally intended to be another branch of what became the Orange Line, so it should have no problems handling OL cars.

[Bramdeisroberts]

Heavy rail was never a plausible option, for engineering reasons. A Blue Line extension was considered and rejected early in the alternatives; it would have been costlier to add a new tunnel under the Charles for no additional benefit.

An Orange Line branch would have been even worse. It would have halved frequencies from Community College north (a very big deal), cost more to build (heavier bridges, longer stations, etc), and also required a new Charles tunnel. The Green Line takes a tightly curved and very steep path to get from North Station to the viaduct; I don't believe you could run Orange Line stock in regular service on it. Orange Line trains may also be too heavy for the Lechmere viaduct. The Orange Line could not have possibly supported branches to both Medford and Union Square; in order to serve both, it would have required a tunnel under Winter Hill (an alternative rejected for the Green Line as being much more expensive). Additionally, using the Orange Line would have eliminated any chance of potential Green Line flanks to Porter, Sullivan/Everett, or the Grand Junction.

See, if it were me doing the designing, I'd have kept Lechmere ---> Union Station as part of the Green Line, because it keeps open the possibility of extending to Porter, Alewife, and out along the Minuteman to Bedford, where it could be a northern mirror of sorts to the D line. It even leaves the door open to some sort of service from Cambridge to Kenmore via the GJ, if they wanted to get really creative/spendy.

The Medford/Tufts extension on the other hand would be perfect for the Orange Line, splitting off of it with a wye between Sullivan and Community College, and leaving open the possibility in the future of a further extension to Winchester and Woburn Center or Anderson. Seeing as to how under-capacity the Washington Street Subway is because of the OL's anemic fleet size, there's no reason why the T with a proper equipment roster wouldn't be able to push Red Line-level service to the Malden/Reading and Medford/Woburn branches.

On the bright side though, those Taj Mahal GLX stations in Somerville would be a cinch to convert to high-platform OL stations if the T ever felt like spending the extra couple hundred million or two to build the bridge over the BET yard to connect the two.

[sery2831]

This is not a GLX thread... lets get back to CBTC discussion.

[Gerry6309]

This is not a GLX thread... lets get back to CBTC discussion.

Mea culpa.

I got drawn into discussions of the value of the loops, and started this down the wrong path.

From a practical standpoint, there have been enough rear-enders on the Mattapan Line that some sort of collision avoidance device would be useful. What I cannot understand is the MBTA's reluctance to install crossing protection at Central Av. A trolley museum in Connecticut has two active protected crossings of public roads (gates and flayers) while the MBTA has a much busier unprotected crossing. This crossing's lack of protection means that the cars must wait for an opening or a courteous driver (rare in this area) and must stop and proceed even if the crossing is clear. With the nearside stops and dead-end traffic, Capen St. is not an issue.

[jonnhrr]

Given that the touted advantage of CBTC is to squeeze more TPH out of a line (whether it actually does this is another issue) wouldn't the T be better off just installing a conventional low-tech signaling system, something proven? After all TPH is not an issue on this line, it has more than enough capacity for the traffic.

Unless they are looking at it as a test bed for CBTC in the Central Subway which makes sense.

Jon

[MBTA3247]

Unless they are looking at it as a test bed for CBTC in the Central Subway which makes sense.

That's exactly what they're doing.

[F-line to Dudley via Park]

Given that the touted advantage of CBTC is to squeeze more TPH out of a line (whether it actually does this is another issue) wouldn't the T be better off just installing a conventional low-tech signaling system, something proven? After all TPH is not an issue on this line, it has more than enough capacity for the traffic.

Unless they are looking at it as a test bed for CBTC in the Central Subway which makes sense.

Jon

It's not anything inherent with "CBTC increases TPH". It's the Green Line: what kind of traffic intermixes on the Green Line, and what the Green Line's theoretical capacity ceiling is. Vs. whether there is a CBTC design/layout that can be fitted to the Green Line that's equal-or-better.

The Green Line already has one of the highest TPH's of any rapid transit system in the world, but the way it goes about that is pretty much managed chaos. Grade-separated branches with a dispatch-controlled signal system (D) of more-or-less precise scheduling merging at the same place as two street-signaled lines with crapshoot scheduling. Then a street-running line (E) that begats a reservation-running line all under street signaling transitioning into a flow-corrected subway line before merging at the midpoint. Then 3 downtown turnbacks. All in a 100-year-old subway with bizarre dimensions and its own set of flow kinks related to its archaic age and cobbled-together design that was adapted from downtown loop circulator to ill-fitting end-to-end circulator. Operating line-of-sight on incredibly short signal blocks still assisted to large degree by human dispatchers squawking over the radio while keeping hand count of train positions to keep some semblance of order. How does one unify all this under one computer brain?


On lines where CBTC is being implemented or strongly considered you've usually got a conventional HRT line (e.g. NYC Subway , or Red/Blue/Orange) with conventional end-to-end grade separation, conventional signaling on ALL branches and routings that's under the complete control of central dispatch (i.e. no "unknown" control like surface traffic), conventional controlled stops (NYC- or Blue-style mechanical trips, or Red/Orange-style analog ATO), conventional-length signal blocks for a 6+ car HRT train where even the shortest blocks are 'long' compared to the Green Line, and 110 years of more-or-less 'standard' heavy-rail subway design. It's orders-of-magnitude more straightforward to fit to a computer because all properties of the line are homogenous 'known-knowns' under central control. Even factoring in variable station dwell times and the crazy quilt of merging/diverging lines like in NYC doesn't seriously test the boundaries of automated dispatching. Therefore, a well-designed CBTC system that takes into precise account every 'known-known' of a subway line can quite easily equal the TPH of the previous system and potentially exceed it if the design shoots for that goal. Without question it would increase TPH potential on Red, Orange, and Blue. Especially Red.


Or...

On other mixed-traffic LRT systems that feed into a signalized subway, you can also do CBTC effectively because the 'managed chaos' variables are more self-contained than the Green Line. The subway's layout is likely a newer design still carrying the traffic it was originally designed for (i.e. trunkline not loop, with sight distances between blocks designed for LRV's and/or multi-car consists not open-car 1890's trolleys). There may be fewer branches, or less-irregularly spaced branch meges making the subway co-mingling easier to anticipate. And EVERY other trolley subway--MUNI, SEPTA, Toronto--has far fewer TPH as its starting point than the Green Line does. So light rail CBTC starts out with somewhat higher degree of design difficulty than a 'closed' centrally-controlled HRT system...but way less design difficulty than the Green Line.


The Green Line is just off-the-charts in comparison. And while other systems worldwide share SOME of its quirks and may even hit the severe end of the scale on SOME of those quirks...the Green Line is literally the only one in the entire world that presents ALL of those quirks in that messy a combination. And thus has no precedent for how to design a signal system to deal with ALL of those quirks. And therefore there's not even a reasonable expectation that there's a perfect-enough CBTC design to be had today that could maintain current TPH. Let alone enhance it. Because TPH probably already is at the bleeding-edge limit of what you could run heavy LRV's more-or-less safely through those tunnels.



Now...it could be possible to do CBTC on the D. CBTC in the Huntington tunnel. CBTC from North Station outbound to the GLX branches and anything furher. And then just stop there and transition onto the current signals for Kenmore-NS. Those outer signalized parts of the Green Line by their very nature will never have more than a theoretical minority of the TPH and the train spacing that the Kenmore-North Station gut of the Central Subway has. Those are well within the 'known-knowns' design tolerances of CBTC deployed elsewhere, and just a partial install of CBTC to those spots could bring real benefits.
-- Safety, collision avoidance.
-- Lower maintenance costs. LOTS less trackside hardware to maintain with CBTC. Which is why post-Sandy NYC is stepping up its investment. And why the T may want to strongly consider it on Blue even with its light traffic volumes; that's a few hundred signal heads, mechanical trip arms, trip arm heaters, and a whole lot of electrical cable taking a pounding from salt air which it could entirely retire in favor of sporadically-placed low-voltage radio transponders.
-- A return to 50 MPH speeds on the D, and an expansion of 50 MPH territory to more places on the D and certain places on GLX.
-- Additional precision on the branchline dispatching, making threading the needle into the Central Subway easier to do without bunching.

You still would be under human control with all the safety implications therein on the most crowded parts of the subway, so it's not a complete solution. But it's a beneficial start, and #3 and #4 above do very tangibly improve Central Subway service by-proxy. I suspect even if they did start rolling it out with a goal of converting the central subway we'd still only see CBTC on the D first for several years until it's perfected. So it probably would've been a Phase I: D, GLX, Huntington anyway. They just may want to re-approach this idea that "Phase II: Central Subway" has no firm commitment to ever proceed until they can guarantee a Do-No-Harm design for TPH. If they do find the way, it's a software solution; the trackside, onboard, and dispatcher hardware is the same as what they used for the branchlines, and thus compatible. So they lose nothing doing the branches then abandoning plans to do the Central Subway. Or doing absolutely nothing to the Central Subway other than programming some barest-minimal acceptable collision avoidance and leaving everything else alone.

[MBTA3247]

-- A return to 50 MPH speeds on the D, and an expansion of 50 MPH territory to more places on the D and certain places on GLX.

Wasn't the speed reduction from 50 to 40 due to the cars hunting at higher speeds? Changing out the signal system won't do anything for that.

[typesix]

Yes, due to hunting because of track conditions, which were supposedly fixed a couple of years ago. Could be the 40 mph limit stands because of reduced power draw and reduced maintenance.

[jbvb]

In order to maintain previous TPH levels, any CBTC will have to be designed with TPH as a priority, and budgeted as such. I was commuting on the Red and Orange lines when the current ATO was put into place, and TPH went down due to the new block lengths. Passenger convenience also went down, with the introduction of many "so near and yet so far" stopping points - halfway into or out of the station, 100 feet short of the station etc.

CBTC is considerably more scary to design than any of the older relay-based signal systems because of the number of variables and edge cases that have to be recognized, analyzed and handled. In other forums here covering the Northeast Corridor in New Jersey, there are frequent gripes from engineers about getting delayed by cab signal aspects which don't actually contribute to safety. After 30 years as a software engineer, I'm pretty sure those aspects exist because there wasn't the budget or time to make the design more complex in order to eliminate them.