Green Line Reconfiguration

[davem]

I think I see where the disconnect is: I've always assumed that a Huntington Ave-Tremont extension would be built at the same time as an Essex St subway. So what you are proposing combines them into one project and one tunnel.

No question construction of one larger tunnel along the Pike/Bay Village would be cheaper and easier to build. With Essex, though, I wouldn't use the POS provisions but have a new slip using the old Boylston St portal which is back near Arlington (avoiding the burial ground). This would give you the space to dive as deep as you need. I still think this would have other benefits like an underground mezzanine that can be used by nonpaying pedestrians during winter months.

I see your reasoning now (though to be fair in my version the Copley Junction was the first to go by extending the Huntington Ave subway to Tremont St and this doesn't prevent a tunnel to Dudley mind you) and I do see the downsides to the Essex St alignment. That said I still stand by it.

Okay, I understand our argument a lot more now. I still think the engineering difficulties of going via Essex are a huge issue, as well as the problem of rerouting one of the lines that has always gone to Government Center to the Seaport screwing with peoples commutes. I suppose the latter point could be mitigated by having the Seaport line run as a new service from Reservoir, or preferably Harvard (or both).

If your Essex plan also includes reactivating the Tremont tunnel, extending the Huntington tunnel, and building the Brookline Village connector, then really what needs to be done is a true cost/benefit study. Essentially it's equal lengths of new ROW: either deep bore under Essex, or cut-and-cover plus surface through the South Bay interchange. You would get a minor cost savings going via Essex as well since you wouldn't need the expensive Bay Village superstation I've designed, just a simple split with a V platform.

 

[Alon]

I know I'm not stating anything new but one of the issues with the Boylston tunnel (besides the infamous capacity killing at grade junction before Copley) is that there are simply too many trains that are forced to travel down it. Now I know we could all debate at length what could be done to increase capacity short of adding a new tunnel. We could talk about dispatching on the surface segment, we could talk about CBTC and we could talk about enhancing the Worcester line with indigo service, etc. and all of those thing are relevant and important and will help. But at the end of the day I think the essence of the problem is that we cram too many trains down that tunnel for it to operate well. In short,[/I]We have a capacity problem[/I]

It's not "we could talk about." You don't, in fact, talk about it. Or Van brushes it off as thinking small, even though it provides equivalent capacity to a Green Line branch without any of the flat junctions on the way.

Out here in the first world, the Boston Green Line would not be particularly special. And Park Street and Government Center wouldn't even be considered terribly busy stations. T-Centralen, Stockholm's central subway station, has 168,000 boardings per weekday. There exist stations lying on one branch of one line, like my university's station, with more ridership than any single T station.

There are fewer trains in that tunnel now than there used to be. The capacity's a kludge because of the surface running and the double-berthing, but it's there. The problem Boston has isn't capacity, but coverage, and that's what things like Dudley/Blue Hill, Lynn, Arlington, GLXes, the Urban Ring, and the NSRL are about. I don't actually mind the D-E connection too much, as long as it doesn't hog funds that those other projects need more, and doesn't make it harder to build the NSRL tunnel.

What I do mind is the Seaport. A city with a competent program for transit construction wouldn't build rail there, TOD or not. A city with competent planners wouldn't even think about reverse-branching the Green Line to connect things directly with Back Bay at the expense of actual Downtown Boston, and would certainly not think about wasting $2 billion on the Silver Lie connection.

Alon, I know you are going to disagree with me on that point but I can say that it's not just me who says that. A lot of people have looked at the same issue carefully and considerately and have come to the same conclusion. That's why there are so many proposed solutions floating around (some of them crazy) that all attempt to solve this core issue of undercapacity. For example, we have that Wentworth professor who has recently given a presentation about sending the Blue line down Newbury Street to eat the D line and on the Transitmatters podcast one of the show's presenters has advocated that we turn the green line into heavy rail. While those are potential solutions (with many attendant issues) they pale in comparison to this proposed solution that does more for less with fewer downsides.

Yes, people say that. That doesn't mean they're right. There's no organization before electronics before concrete thinking in the US, because the transit managers are dead set on pretending it's the 1960s, and the railroad managers are dead set on pretending it's the 1930s. As a result, a very large number of transit activists are certain that things that happen in your average European transit city hundreds of times per day are impossible. Hence the "it's just a commuter line" derision. No, there's no such thing as "just a commuter line." In London, Paris, Berlin, and Munich, commuter rail has about as much ridership as the subway. Provide frequency, mode-neutral fares, frequent urban stops, and reasonable speed (which no diesel equipment can provide with frequent stops), and people will ride it. Organization first. Then electronics. Concrete last.

 

[vanshnookenraggen]

If your Essex plan also includes reactivating the Tremont tunnel, extending the Huntington tunnel, and building the Brookline Village connector, then really what needs to be done is a true cost/benefit study. Essentially it's equal lengths of new ROW: either deep bore under Essex, or cut-and-cover plus surface through the South Bay interchange. You would get a minor cost savings going via Essex as well since you wouldn't need the expensive Bay Village superstation I've designed, just a simple split with a V platform.

I think we've finally met in the middle. I fully realize just how expensive and difficult the Essex St alignment would be. I'd just like to know from an actual engineering study which one would be best.

And like I mentioned, if there was a serious plan for developing air rights along the Pike through Chinatown and the South Bay interchange then a Bay Village alignment would be a clear winner.

 

[vanshnookenraggen]

Alon I just want to point out that I am FULLY behind you in your NSRL, improved CR service view point. I actually think that building the NSRL with infill stations and upgraded headways (via xMU or push-pull systems) is the best investment Massachusetts can make REGIONALLY. I truly believe that if built the NSRL can make an actual impact on the rapid transit congestion in central Boston.

If Boston had a version of the London Overground... game changer.

But that's another topic for another thread.

This thread is literally just taking all the Green Line ideas and putting them in their own specific thread. So when you say that upgrading the Worcester Line would effect the Green Line congestion there is a disconnect between what we are talking about and what you are talking about.

There already is a NSRL thread and perhaps we should continue the conversation there.

 

[Alon]

If Boston had a version of the London Overground... game changer.

But that's another topic for another thread.

This thread is literally just taking all the Green Line ideas and putting them in their own specific thread. So when you say that upgrading the Worcester Line would effect the Green Line congestion there is a disconnect between what we are talking about and what you are talking about.

Sure, but.

The issue here is that transit should work as a system. As a simple example of this, I'm very down on using a Green Line branch to connect Allston to Harvard, even though I think such a connection is useful, because it should be part of another line.

The point I'm making with the Worcester Line is that, since it parallels the busiest Green Line branch and serves the areas seeing the most ongoing and expected future development, it could take meaningful traffic off the line, decongesting it to the point that the flat junction at Copley is not a problem. Obviously, current traffic fits within current capacity. This isn't the Lexington Line in New York. The issue is extensions, then.

F-Line-style extensions aren't a capacity problem, because the Tremont/Boylston junction is grade-separated. So really the question is whether the useful extensions to the existing lines create a capacity crunch at Copley. I think everyone in this thread agrees that two such extensions are at least plausible (I think both should be built but I don't know if everyone agrees): E to Arborway, D.5 diverging from D at Newton Highlands to serve Needham. Both would add substantial traffic to Copley, so it's worth asking whether E to Tremont would be required.

I believe it would not be. I am not certain of that. My reasoning is as follows:

- In the presence of commuter rail modernization, Riverside loses value, since it no longer provides better service than Auburndale. Eliot, Waban, and Woodland are all low-ridership. This means that D.5 can be achieved by halving the frequency of the existing D branch and diverting half the trains to Needham, with each half running every 10-15 minutes.

- Since commuter rail modernization hits the B branch the most, it equalizes ridership across different branches, which makes it possible to run each branch at the same frequency, which simplifies scheduling greatly.

- An E branch extension should be entertained only if it has dedicated lanes all the way. This would improve its reliability somewhat, even as traffic went up substantially.

- In the presence of an urban ring, parts of the E branch would be put underground. I think the urban ring is crazy using the original definition of "should be funded only after the NSRL is in place," but solid using the definition of "could get decent ridership for the cost." At least in my conception, the urban ring zigzags away from the 66 route to hit the Orange Line and commuter rail at Ruggles, running under Huntington between Mission Park and the Museum of Fine Arts. This would naturally include an extension of the E tunnel, which would improve E reliability even more.

I've made an unspoken assumption so far, so let me be explicit: the problem is not physical capacity, since the Green Line used to have way more traffic than it does today measured in tph, but reliability. The line's current operations require relatively constant headways. Past operations did not. Current operations also involve far more car traffic than there used to be, which makes headway maintenance harder.

The solution, which I haven't talked about so far, is signal priority on all street-running segments. It's practically free. It just requires picking a fight with NIMBYs who think a train with 200 people should wait at a red light for a handful of cars with 1 person each.

It sounds like a small deal, but good signal priority is an immense speed booster. Again using Vancouver as an example, the east-west arterials on the Westside have signal priority over all north-south traffic, except at the arterials (of which there are 3), where the buses stop anyway. This means that, unintentionally, the buses have signal priority. The 4th Avenue limited-stop buses maintain average speeds of almost 30 km/h between Burrard and UBC; the 99, which runs on more densely developed Broadway, still averages 21 km/h.

Now, those Vancouver limited-stop buses may be fast, but they bunch. They're still buses, running at very short headways. On 4th Avenue, some (maybe half?) aren't even articulated. The bunching is caused by congestion in the more central parts of Broadway, and long boarding times at the UBC Loop. That's why I'm so adamant about all-door boarding everywhere: it both speeds up boarding and makes schedules less variable, since small perturbations no longer compound. (The 99 allows all-door boarding, but the 4th Avenue buses do not.)

Finally, again thinking systemwide, Tremont Street Subway reactivation conflicts with the D-E connection idea (to say nothing of the fact that the Urban Ring conflicts with it, too). Traffic projection for light rail to Dudley is already higher than the current traffic of any single Green Line branch. An extension on Blue Hill to Mattapan would make it even busier, while an E reroute to Tremont would make the traffic volume into Tremont higher than the traffic of the B and C branches, especially since commuter rail could take traffic off the B but not the F and only marginally the E-to-Arborway.

Put together, I think the correct plan for the southern/western side of the Green Line is heavier on organization than on concrete:

- Full signal priority on all branches.
- Stop consolidation on the B and C branches, to one stop per 400-500 meters.
- E restoration to Arborway, on dedicated lanes.
- F branch to Mattapan via Tremont and Dudley.
- Possibly, a one-stop extension of the C branch to Chestnut Hill Avenue.
- Elimination of the Needham Line and its replacement with a branch of the D branch.
- An extension of the E tunnel to Mission Park, running partly alongside the Urban Ring subway.
- F branch to Forest Hills via Dudley.

The first five are reasonable, and the last three are crazy. I've been convinced that the F branch to Forest Hills is not useful today, but think it could be useful in a future with commuter rail modernization, since Forest Hills would gain importance.

At the northwestern end, the GLX should be pushing farther out, to West Medford and Brandeis via the Watertown Branch. These are both crazy: West Medford isn't useful as an outer anchor except in the presence of good commuter rail, and the Watertown Branch, while useful, isn't a critical priority.

 

[fattony]

Holy wall of text Batman!

Ladies and Gentlemen, we officially have "F-Line: Euro Edition"

 

[Equilibria]

In the presence of commuter rail modernization, Riverside loses value, since it no longer provides better service than Auburndale. Eliot, Waban, and Woodland are all low-ridership. This means that D.5 can be achieved by halving the frequency of the existing D branch and diverting half the trains to Needham, with each half running every 10-15 minutes.

Umm... no. Riverside's value has nothing to do with serving Newton Lower Falls and the parts of Auburndale nearest it, though I certainly get a lot out of that personally. Riverside is an immense park-and-ride operation. The lot is full most days and for Sox games it spills into the neighborhoods. The MBTA is about to build a multimodal garage and bus station facility (a la Alewife) and a developer has a 250,000sf TOD going up. Aurburndale is a nice neighborhood center, but it has no parking, no bus connections, and is located more than a mile from highway access. Increasing service there would be nice for the folks who live and work next to it, but Riverside is and always will be the more important station.

That's not to say that some schedule balancing wouldn't accompany Green-to-Needham, but Riverside's ridership alone justifies frequent service on that branch. It was the 12th busiest Green Line surface station by weekday boardings in 2013, in a pack with Northeastern University, Brigham Circle, and BU Central.

 

[Alon]

Umm... no. Riverside's value has nothing to do with serving Newton Lower Falls and the parts of Auburndale nearest it, though I certainly get a lot out of that personally. Riverside is an immense park-and-ride operation. The lot is full most days and for Sox games it spills into the neighborhoods. The MBTA is about to build a multimodal garage and bus station facility (a la Alewife) and a developer has a 250,000sf TOD going up. Aurburndale is a nice neighborhood center, but it has no parking, no bus connections, and is located more than a mile from highway access. Increasing service there would be nice for the folks who live and work next to it, but Riverside is and always will be the more important station.

What I'm alluding to is that the park-and-riders could park elsewhere, closer to where they live. There aren't 2,000 people commuting to Boston per day living closer to Riverside than to any other station; it has an extended draw, which could just go elsewhere if the Worcester Line stopped sucking. Hell, some of this draw is from Needham, which would get its own Green Line branch.

That's not to say that some schedule balancing wouldn't accompany Green-to-Needham, but Riverside's ridership alone justifies frequent service on that branch. It was the 12th busiest Green Line surface station by weekday boardings in 2013, in a pack with Northeastern University, Brigham Circle, and BU Central.

Okay, but another way to think about it is that, per the 2014 Blue Book, the D branch has 20,075 boardings at Newton Highlands and points farther east, and 4,557 at stations farther west.

 

[Matthew]

The T is [slowly] working on signal priority, all-door boarding, and stop consolidation. They need constant cajoling. Help is always appreciated.

Regarding Riverside, many of the people who use that station are actually commuting to Longwood. Which means that the Yawkey station is not too bad of an alternative.

Historically, Riverside was connected to the railroad lines that ran along the Framingham and Highland lines. Riverside has been envisioned again as a potential terminus for frequent multiple-unit service along the Framingham line, so that could be another way to decongest a bit.

 

[Transitmass]

Where is F-Line? This thread is ready built for him. I'd love to hear his thoughts.

 

[George_Apley]

Where is F-Line? This thread is ready built for him. I'd love to hear his thoughts.

According to him he's been busy with real world stuff. He's had to cut himself off from his aB addiction to maintain productivity.

 

[vanshnookenraggen]

Historically, Riverside was connected to the railroad lines that ran along the Framingham and Highland lines. Riverside has been envisioned again as a potential terminus for frequent multiple-unit service along the Framingham line, so that could be another way to decongest a bit.

Using Riverside as a terminal for xMU service with some infill stations along the Worcester Line through Newton and Brighton would have an impact on the Green Line. It would also require the single platform stations through Newtown to be rebuilt and that would be expensive since there isn't any room along the Pike to do so. You'd have to shore up Washington St by digging out the embankment to make room for second platforms.

 

[vanshnookenraggen]

Another (crazy) alternative is to actually turn the Green Line into a modern light rail system and totally separate street running trains from subway trains. This was the idea back in the 1920s and later extensions even planned for it (Kenmore, Prudential and Symphony were all built to be converted to high platform operations).

That would then require two tunnel extensions, Huntington Ave to Brookline Village and Comm Ave to somewhere in Allston. Whatever is left of the B line would become a shuttle between BC and Packards Corner (or wherever) and the C would terminate at Kenmore like it was designed to do when that station was built.

With just two or three branches operating along Boylston St the impact of the Copley Junction would be reduced. Since any trains coming up from Dudley through Tremont St would use the outer tracks at Boylston/Park you wouldn't have that service interrupting Boylston St trains as much.

True it would force B and C line riders to transfer but operations along the subway portion of the Green Line would be much more efficient.

 

[davem]

I've always thought that sending the B over the Grand Junction, and a new grade separated line from Harvard through a Comm Ave tunnel woul work well. Transfers would happen at a superstation at the BU Bridge, combined with a CR station.

 

[Equilibria]

What I'm alluding to is that the park-and-riders could park elsewhere, closer to where they live. There aren't 2,000 people commuting to Boston per day living closer to Riverside than to any other station; it has an extended draw, which could just go elsewhere if the Worcester Line stopped sucking. Hell, some of this draw is from Needham, which would get its own Green Line branch.

No, but Riverside is incredibly well-located to take traffic off of multiple interstates. It's possible that many of those who currently park drive past a Worcester Line stop on the way in, but I don't think it's likely that they would choose a longer drive and an equally long train ride for the sake of schedule flexibility. It's more likely that they're commuting to Brookline or Longwood and can't get there on the Commuter Rail (Yawkey really only serves one corner of the LMA, and that's with a long walk).

Okay, but another way to think about it is that, per the 2014 Blue Book, the D branch has 20,075 boardings at Newton Highlands and points farther east, and 4,557 at stations farther west.

Well, you're comparing 4 stations to 9 stations, but it's true that Waban and Eliot are low-ridership. That doesn't negate the need to serve Riverside with frequency. If DMU/EMU service were to provide a faster ride to Downtown, then sure, I could see lowering the Green Line frequency a little. That's very different than cutting service and hoping people end up at their local commuter rail stop, though.

 

[Alon]

Using Riverside as a terminal for xMU service with some infill stations along the Worcester Line through Newton and Brighton would have an impact on the Green Line. It would also require the single platform stations through Newtown to be rebuilt and that would be expensive since there isn't any room along the Pike to do so. You'd have to shore up Washington St by digging out the embankment to make room for second platforms.

Don't build a second platform - just switch the southern track and the platform, so that the platform becomes an island platform. Access shouldn't be a problem, since all the stations are already located at overpasses, and the platforms are actually on the Turnpike side of the ROW, so there's no direct access to them except through the overpasses anyway.

It'd be narrower than ideal, but for a raised island platform, it's enough - the Newtonville, West Newton, and Auburndale platforms are about as wide as the Penn Station platforms. The platforms all have to be rebuilt high anyway, since low platforms are major time wasters in extra dwell time and schedule risk. All of this is a few million dollars per station, as on the Fairmount Line.

There's also space at the potential infill stations. Newton Corner, at Centre Street, has space for the southern track to swing south to make room for an island platform, as at the existing Newton stations. Faneuil, at Parsons Street, is on a viaduct with a few meters on both sides of the track for platforms. Brighton, at Market Street, is more constrained, but has wide track centers and a total ROW of 12 meters, which is wider than what Newton has today. Allston, at Franklin and Cambridge Streets, has four tracks and about 28 meters, wide enough for anything. There's even about 12 meters at Commonwealth and the BU Bridge for a BU station with a transfer to the B branch (although constructing the necessary junction toward the Grand Junction, to permit a circle line through the NSRL and the Grand Junction, would be nontrivial). At Mass Ave and Hynes, there's maybe 11 meters, which is as constrained as Newton.

 

[George_Apley]

Brighton, at Market Street, is more constrained, but has wide track centers and a total ROW of 12 meters, which is wider than what Newton has today.

The Brighton infill is already scheduled a bit further east at Everett street. Part of the New Balance project. I doubt there will be access from Market. The Allston infill doesn't have a location pinned down just yet, but probably somewhere in the vicinity of Agganis Arena.

 

[Alon]

The Brighton infill is already scheduled a bit further east at Everett street. Part of the New Balance project. I doubt there will be access from Market. The Allston infill doesn't have a location pinned down just yet, but probably somewhere in the vicinity of Agganis Arena.

I know, I know. The New Balance station idea is stupid - there's no street access from the south.

The West Station idea is also stupid, but it most certainly has enough space. Also, it could work as infill between BU and Allston, in the event the MBTA brings commuter rail to the 21st century, because then it could be used for TOD.

One thing to be wary of is the introduction of intercity trains to the corridor. There's no room for a new ROW for HSR from Boston toward Worcester and Albany; trains would have to use the Worcester Line until about Auburndale, where they could diverge to a new ROW near the Turnpike. If trains are expected to run at 15-minute frequency, then the maximum difference in schedule between intercity and commuter trains between Back Bay and Auburndale is 11 minutes, which corresponds to 8 local-only stations at 160 km/h and 10 or maybe 11 at 130 km/h, including the 4 preexisting ones. 10-minute frequencies cut the number of allowed intermediate stations to 4 at 160 and 5 at 130, unless there's an overtake. The only feasible location for an overtake is Allston, so this means 4 or 5 stations on each side of it. Anything more than that requires 4 tracks all the way, on a line that has no room for more than 2. So intercity trains put a tight limit on how many commuter trains the corridor can run.

 

[JeffDowntown]

I think we've finally met in the middle. I fully realize just how expensive and difficult the Essex St alignment would be. I'd just like to know from an actual engineering study which one would be best.

And like I mentioned, if there was a serious plan for developing air rights along the Pike through Chinatown and the South Bay interchange then a Bay Village alignment would be a clear winner.

You are going to get huge resistance from the row house owners in Bay Village for an along the Pike routing.

The T has a horrible track record at waterproof tunneling with cut and cover construction. The ground water table in that area is critical to the foundations of all those historic row houses. Trust me, the T would be in court for years trying to get approval for the tunneling --that neighborhood is organized.

(Deep bore Essex alignment is actually easier to waterproof, because it is built like it has to be waterproof).

 

[davem]

You are going to get huge resistance from the row house owners in Bay Village for an along the Pike routing.

The T has a horrible track record at waterproof tunneling with cut and cover construction. The ground water table in that area is critical to the foundations of all those historic row houses. Trust me, the T would be in court for years trying to get approval for the tunneling --that neighborhood is organized.

(Deep bore Essex alignment is actually easier to waterproof, because it is built like it has to be waterproof).

The only tunneling that comes near the rowhouses is the short segment from the old portal to the Pike. Which would have to be excavated for anything that reuses the portal (IE: retiring Copley Junction). Along the Pike, it's only a matter of moving the retaining wall ~30'.

The groundwater issues can be solved by having the sump pumps in the tunnels recharge the water table instead of flow into storm drains. The only reason this was an issue in the first place was because the hydrology of the filled land wasn't really understood until recently. Over the years asphalt came to replace more porous paving surfaces, while all of the storm/roof/whatever water was sent further and further away. This caused the water table to lower, exposing the fragile wood piles to the air, causing them to rot. Now that they know not to do that it's really a non-issue. IIRC, all new construction has to address this, the tunnel would be no exception.