Underrated Transit Projects

In between the airport connectors and mixed-traffic streetcars are some public transit proposals that would be potentially high-performing. This is a list of potential lines in the US that don’t get nearly the exposure that they deserve.

The basic rule of this post is that if it’s being built, or is on an official urban wishlist pending finding the budget for it, then it’s not underrated. Some of the most important transit projects in North America are in this category: Second Avenue Subway’s current and future phases, the Regional Connector, the UBC SkyTrain extension. What I’m interested in is lines that are only vaguely on any official wishlist, if at all, but could still get very high ridership compared to their length. It is possible that these underrated lines would turn out to be worse-performing if a study were undertaken and the costs turned out to be very high, but in no case was there an honest study. Sometimes there has been no recent study; other times there is one but it sandbags the project.

Finally, I am not including commuter rail projects on this list. Under current regulations and operating practices, nearly all North American commuter rail projects are wastes of money. Conversely, nearly all projects that assume modernization of practices are underrated. This swing, based almost entirely on organizational question, is why I’m excluding these projects from this list. The subway and light rail projects below are less sensitive to organizational questions.

Utica Avenue Subway

Location: New York

Concept: an extension of the 4 from Crown Heights along Utica Avenue to Kings Plaza, about 7 km. If Second Avenue Subway’s Phases 3 and 4 are built, then a branch can be built from Second Avenue to Williamsburg and thence under Bushwick, Malcolm X, and Utica, taking over the entirety of the line, with the 4 cut back to its current terminus; this is an additional 9 km to Second and Houston.

Why it’s underrated: the second busiest bus route in New York, the B46, follows Utica: see here for New York bus route rankings. The busiest follows First and Second, which are getting a subway. Two additional routes in the top ten, the B44 and the B41, follow Nostrand and Flatbush respectively, fairly close to Utica. The B46 has 48,000 weekday riders and the B41 and B44 have another 70,000 between them. Since subways are much faster than city buses, the expected ridership is much higher than 120,000, measured in multiples rather than in a percentage increase. In addition, the 2, 3, 4, and 5 are all busier coming to the Manhattan core from Uptown than from Brooklyn, so adding to their ridership from the Brooklyn end balances the loads better, and avoids the required increase in operating costs for the new riders.

What is being done right now: nothing.

Geary Subway

Location: San Francisco

Concept: a full subway from Market Street to the Outer Richmond District, about 9 km. This can connect to the BART subway, the Muni Metro tunnel, or a second Transbay Tube if one is built.

Why it’s underrated: the 38-Geary is the busiest bus route in San Francisco, with 57,000 weekday riders between the local, the limited, and the express buses: see here for San Francisco bus ridership. Parallel corridors are also busy: the 1-California has 29,000, the 31-Balboa has 10,000, and the 5-Fulton has 17,000. Some of the census tracts along the middle of the route, in Little Osaka Japantown, rank together with Los Angeles’s Koreatown as the densest in the US outside New York. BART’s current limiting factor is not the Transbay Tube, but the grades farther south in San Francisco, which lengthen the braking distance and make it impossible to run a full 30 trains per hour through the core segments; a Geary branch leaving south of Montgomery Street would reduce service to points farther south, but improve capacity for riders heading from Oakland to the San Francisco CBD.

What is being done right now: there were never subway plans, but there were light rail plans, which due to local merchants’ opposition to loss of space for cars were downgraded to a rapid bus. The city’s FAQ on the subject even has the cheek to portray the Boston Silver Line and the Los Angeles Orange Line as successes.

Downtown Relief Line

Location: Toronto

Concept: there are several different alignments, but all feature an east-west line somewhere between Queen Street and Union Station, with one or two bends to the north to intersect the Bloor-Danforth Line. The latter two alignments (using option 4B for the second one) feature about 12 km of tunnel; I do not know how much the first one has.

Why it’s underrated: only one subway line serves Downtown Toronto, the Yonge-University-Spadina Line. Bloor-Danforth is too far from the CBD, and requires a transfer. The transfer points are very crowded: as far as I can tell from this list, the central one, Bloor-Yonge, has 200,000 weekday boardings, apparently including transfers. Without figures that include transfers in other cities I can’t make comparisons, but I doubt any two-line, four-track station in New York has this many riders. Union Station is quite crowded as well, and DRL proposals include transfers to outlying commuter rail stations. Ridership on parallel streetcars is very high: there are 53,000 on King Street, 44,000 on Queen, and, if a more northern alignment for the DRL is chosen, 32,000 on Dundas.

What is being done right now: more studies; construction will almost certainly begin any decade now. Neither David Miller’s Transit City light rail proposal nor Rob Ford’s replacement of Transit City with subways included the DRL.

125th Street Subway

Location: New York

Concept: either Phase 5 or Phase 2.5 of Second Avenue Subway, going west along 125th to Broadway, with a station at each intersection with an existing north-south subway.

Why it’s underrated: east-west transportation in Manhattan is slow, even by the standards of Manhattan buses. The 125th Street buses in my experience are slower than walking; despite this, the various routes have about 90,000 weekday boardings between them, of which about 30,000 come from 125th Street itself. Second Avenue Subway Phase 1 is going to substantially improve east-west transportation, by serving Times Square and offering a two-seat ride from the Upper East Side to the Upper West Side and Central and West Harlem; however, passengers from East Harlem will still have to take a major detour to avoid the crosstown buses. While SAS offers a relief to the 4/5 and 6 lines, the 2/3 and A/D express lines are overcrowded as well, and a connection at 125th Street would divert some East Side-bound commuters.

What is being done right now: nothing, although (some) railfans who work at the MTA privately want to see such a line built.

Silver Line Light Rail

Location: Boston

Concept: replacement of the Silver Line buses along Washington Street with light rail, feeding into an existing Green Line portal, about 4 km of light rail.

Why it’s underrated: the Silver Line buses are the busiest in Boston, with 15,000 weekday riders on the buses to Dudley Square: see PDF-pp. 47-48 of the MBTA Blue Book. The ridership doesn’t justify a subway, but does justify dedicated lanes and rail. The Green Line tunnel has some spare capacity, has a portal pointing in the correct direction, and could take an additional train every 6 or 7 minutes, which would give riders in Roxbury faster trips through Downtown Boston.

What is being done right now: nothing – a study sandbagged the rail bias factor and assumed only 130 new transit riders on a Silver Line light rail service, making the project appear cost-ineffective.

Triboro RX

Location: New York

Concept: a circumferential subway line, with about 1 km of new tunnel and 35 km of route on preexisting rights-of-way, abandoned or lightly used by freight trains today.

Why it’s underrated: the biggest cost driver, right-of-way formation, is already present. The right-of-way in question has a few daily freight trains, but the most critical link, the Hell Gate Bridge, is four-tracked, and freight trains can be kicked out from their segment of the bridge and moved to the Amtrak tracks. The work done by Michael Frumin and Jeff Zupan in the late 1990s estimated about 150,000 commute trips per weekday (76,000 commuters each making a roundtrip per day), which is low for a greenfield line of this length but reasonable for a line on existing rights-of-way.

What is being done right now: nothing, although ever since Lee Sander mentioned the line in 2008, politicians have paid lip service to the concept, without committing funding.

Boston Circumferential Line

Location: Boston

Concept: a circumferential subway, from Harvard Square to Dudley Square or the JFK-UMass subway stop, roughly following the 66 bus route where it runs and intersecting the busiest stops of the Green Line branches and some commuter rail stops. This is about 12 km.

Why it’s underrated: although the busiest Boston bus is the Silver Line to Dudley Square, the next few are circumferential, particularly the 1 and 66, and secondarily the 23 and 28; together this is about 50,000 riders. Boston’s street network is hostile to surface transit except on a few major streets such as Washington, which is why there is no hope of making such a line light rail, which would fit the projected ridership better. A route that parallels the 66, at least until it hits the E branch of the Green Line, would intersect the B, C, and D branches at their busiest respective surface stops, and improve connectivity to Cambridge, which is increasingly a major business district of the Boston region in its own right.

What is being done right now: BRT, on convoluted alignments that don’t exactly follow either the 66 or the 1 where they are parallel but instead make detours.

Nostrand Avenue Subway

Location: New York

Concept: an extension of the 2/5 from Flatbush to the southern end of Nostrand Avenue, about 5 km.

Why it’s underrated: all the reasons that make Utica so strong apply to Nostrand secondarily; the present bus ridership may be high enough to support two subway lines rather than one. The present terminus was built as a temporary one, which is why it has side platforms rather than an island platform.

What is being done right now: nothing.


  1. Zmapper

    “Green Line tunnel has some spare capacity… and could take an additional train every 6 or 7 minutes”

    According to the schedule, during peak hours the B, C, and D lines each operate about 8 trains per hour, and the E line 10, with about 34 trains per hour scheduled through the tunnel. Could the Green Line tunnel handle about 40 trains per hour?

    • threestationsquare / Anon256

      The Green Line is four-tracked between Boylston and Park St, with a turnaround loop just north of Park. The track layout is such that the trains terminating at Park would have to be from the currently existing lines rather than the new (restored) Washington St direction, but as long as riders accept that the infrastructure is there for significantly more trains.

      • threestationsquare / Anon256

        Here is the track map. The disused Pleasant St Incline (which the Washington St line would use) connects to the “outer” tracks at Boylston, only one of which is shown on that map (but both trackways are still there).

    • Ryan

      Absolutely not.

      The Green Line can’t even handle the amount of traffic it sees today for a variety of reasons that mostly boil down to institutional heel-dragging: its signaling system is a literal relic from the 19th century and nobody in any sort of decision-making position has displayed anything close to the level of hair-of-fire urgency that problem merits, never mind any of the other problems.

      It’s really quite foolish to talk about concrete expansions to the Green Line without bothering to acknowledge the fact that the Green Line’s electronics are more than 100 years out of date.

      • Adirondacker12800

        Humans can sometimes perform better than signal systems when it’s low speed subway.

        • letsgola

          Yep. Part of the reason the MBTA has never installed modern signalling on the Green Line is that the short headways and double berthing make it very difficult to install a system that wouldn’t reduce capacity.

          • Ryan

            The beautiful thing about this discussion is that I don’t even have to make the argument as to why this is a stupid attitude to have, because Park Street right now is making my counter-argument for me better than any number of words on a blog could, just as it has pretty much every weekday rush hour, every morning and every evening, for a long time now.

            (Hint: if modern signaling systems will “reduce” line capacity, your line is probably over capacity.)

          • letsgola

            Park St station may be over capacity, especially from a platform loading perspective, though the MBTA doesn’t seem interested in trying to change that (adding another platform on the west side of the C/D track). The signalling capacity issues come in at the stations too, since when they double berth they close up very near to the train in front.

            But from a train volume perspective, the 4-track section between Park & Boylston is not at capacity. If it were, how could the 2-track section between Boylston & Copley even function?

          • Ryan

            “But from a train volume perspective, the 4-track section between Park & Boylston is not at capacity. If it were, how could the 2-track section between Boylston & Copley even function?”

            Simple. It couldn’t, it can’t, and it doesn’t. You can’t call what happens there every single rush hour “functioning.” The entire line is in a state of routine failure. You see it at Park Street where it’s most obvious, you also see it at Boylston, you don’t see it in Government Center only because someone decided that shutting that station for two years was the best possible option, but you see it everywhere in the Central Subway.

            The entire line is in desperate, dire need of help long before we can realistically talk up new branches of the Green Line. The signaling system alone isn’t a cure for what ails us, but it’s the most obvious place to start and that makes the failure of anyone to treat this with any sense of urgency all the more insulting.

          • letsgola

            If you have a signalling system that would make the Green Line run better, the MBTA would probably love to hear about it! 😉

          • Ryan

            CBTC is not exactly “new” technology at this point and there are plenty of options out there which would be compatible with the Green Line.

            I’m sure I’m not the first, fifth, tenth, or one hundredth person to mention CBTC to the knuckleheads in management. Does it really matter?

          • letsgola

            The MBTA has done studies on CBTC. My understanding is that the CBTC vendors don’t think they can make the current headways and double berthing work safely and reliably. Remember that automated signaling, even CBTC, is going to fundamentally increase train spacing on the Green Line b/c the braking algorithm is going to err on the side of caution.

          • threestationsquare / Anon256

            @Ryan: It’s bad enough when people deny the feasibility of transit operations that exist in other countries or existed at other points in history, but this is the first time I’ve seen someone flatly deny the feasibility of a transit operation that exists in their very city successfully transporting hundreds of thousands of people every day. It’s not fast or reliable but it’s there, it functions, people get where they’re going in one piece.

            In 1911 Park Street handled 211 terminating streetcars and 105 southbound through streetcars (and likely a comparable number of northbound streetcars) between 5pm and 6pm each weekday, that’s a streetcar every 17 to 34 seconds on each track. Today’s 34tph service pales in comparison.

          • Max Wyss

            Considering that the speeds are rather low in the tunnel, it might be worthwhile considering to bring the light levels up to a street at night, and then operation on view would be reliable and safe. That would mean that a train every minute would be possible (if the station stops won’t take too much time).

            I am sure that back in 1911 the lines were not signal controlled. Also, (I may be wrong here) there were no trains, but single vehicles only.

      • betamagellan

        Agreed about the signaling system, but it’s also worth noting that a lot of the congestion comes from the flat junction where the Huntington meets the Boylston subway. A Washington Street subway would enter in a four-tracked Tremont subway, as threestationsquare/Anon256 notes.

        In terms of relieving some pressure off the Boylston subway and eliminating that flat junction a new Stuart Street subway, continuing from Huntington and leading into the four-tracked Tremont Subway, was first suggested decades ago, and underground Stuart Street alignments were mooted for future Silver Line Phases. I’m not sure if routing the E line through a new Stuart Street subway quite qualifies as an underrated project, but it could dovetail nicely with a Washington alignment.

  2. Patrick O'Hara

    “The right-of-way in question has a few daily freight trains, but the most critical link, the Hell Gate Bridge, is four-tracked, and freight trains can be kicked out from their segment of the bridge and moved to the Amtrak tracks.”

    The Hell Gate Bridge currently only has four tracks, but it is wide enough to support a fourth track, but you would have to install an additional track along the entire right of way, increasing the costs.

    You would have to figure out a way to get freight off of Long Island, however, as giving it over to the Triboro RX would block off both ways freight can currently get off Long Island (and the third if the the cross-harbor rail tunnel goes in). That’s obviously an obstacle that will have to be planned around, and I can’t think of too many cheap ways to remedy that, unfortunately.

    • Alon Levy

      The track isn’t what’s expensive – it’s just steel and sleepers. The ROW is the expensive part.

      Freight could still run on the Amtrak tracks, it’d just take more time to get there from Lower Montauk to the Main Line to Sunnyside Yards.

      • Adirondacker12800

        There’s all the freight traffic that would be coming through the tunnel from Jersey City too. All the shippers who are drooling to take stuff off containers ships in Port Newark and send them to New England via that route. It has it’s charms, It gets trucks off the crossings. When I’m in the mood to watch New York City tax maps slowly load I’ll go see how wide it is.

        The tunnel from North White Plains to Linden looks better and better. Abandon the Hells Gate bridge for intercity traffic and let Metro North, freight and the subway fight over how to divide it up.

        • threestationsquare / Anon256

          Cap’n Transit looked into the available width a few years ago. There’s a section of a bit under a mile between the Brighton Line and the Culver Line where the Bay Ridge Branch ROW is only two tracks wide and would need to be widened to host both freight and the TriboroRX / L train extension, or the first phase would have to end at the Brighton Line. Other than that the ROW is at least four tracks wide all the way from the waterfront to Fresh Pond Yard, where freight could switch to the Lower Montauk to Sunnyside and Hell Gate.

          • Henry

            The thing is, we have no idea what kind of impact the Cross Harbor Tunnel would have on the rail network in the city. All of a sudden, freight into the city, along the NE Corridor and into Long Island becomes a much more feasible and attractive option, especially considering the chronic truck congestion in the metro area. Will Amtrak, MNR, and freight be able to share two tracks if such a tunnel is built, especially considering the fact that the first two are considering significantly upping their services in the coming years?

            Is there a benefit to using subway cars as opposed to carving out the interior of an M9 and putting in metro-style seating?

          • Adirondacker12800

            Jersey City to Brooklyn would be freight only. The freight wouldn’t be getting to Jersey City on the Northeast Corridor, It would be getting to Jersey City the same way it gets to Jersey City now. And the way it leaves Jersey City now. Sending Long Island garbage to Pennsylvania that way is a lot shorter. And New England beer.

            Most of the fantasy proposals have the Triboro line crossing the Bronx on 161-ish. There’s abandoned railroad ROW that heads that way from the Hell’s Gate Line. Connect with the 6, the 2 & the 5, the D and the 4 and maybe continue onto Manhattan to connect to the A and the 1.

            M10s. The third rail is only installed on a short section near the LIRR main line. The rest of it will have catenary. Well M10As, the kind without third rail shoes and more doors.

          • Henry

            I say two tracks to specifically talk about the section between the Main Line and the Bronx; if Penn Access, Cross Harbor, and the RX all happen, we’d probably end up with two pairs of tracks across the Hell Gate; one for RX and one for Amtrak, MNR, and freight. This isn’t as much of a problem without a Cross Harbor, but my question is whether or not closing this gap in the freight network will oversaturate the two non-RX tracks across the Hell Gate.

          • threestationsquare / Anon256

            Freight has to share track with Amtrak or Metro North to get out of the Bronx northbound anyway, so if this is a bottleneck there will likely be others. Also last I heard the Cross Harbor Tunnel was supposed to be single-tracked, which seems more likely to be the bottleneck than sharing two tracks with Amtrak/MNR.

          • Alon Levy

            The EIS I saw for the Cross-Harbor Tunnel included both a single- and a double-track option, although as I recall the single-track option was the preferred alternative.

          • Adirondacker12800

            A lot of what will be going east through the Cross Harbor will be food. Except for the through traffic from New England and tiny little bit from Upstate New York and Quebec the stuff going west is going to be garbage. Two tunnels have much better reliability and flexibility. They cost less than twice one tunnel. Food in and garbage out are pretty high on the list of things that need reliability.

      • Patrick O'Hara

        At that point, you’re bringing Amtrak into the freight business, and you’re opening the door to a plethora of freight trains (a href=”http://www.thelirrtoday.com/2014/01/ny-110-deriailment-gums-up-mainline.html”>which have the tendency to derail more often than passenger trains) and all interested parties have indicated that they would like it very much if freight stayed as far away from HAROLD as possible.

        Plus, you then have dozens of customers along the Bushwick Branch itself that would be cut off, plus you close off the route to NYNJ, making the Selkirk Hurdle the one and only way to get freight trains east of the Hudson River.

      • lop

        Is there room on Hell’s Gate for Amtrak intercity, MetroNorth PSA, Triboro RX, and existing freight?

        • Adirondacker12800

          It depends on how frequent service is on each. If there’s freight coming across the harbor things are going to be different than if there isn’t. No freight during rush hours and lower frequency on Metro North along with scheduling legerdemain might make it possible. Depends on whether or not you want to mix freight with Amtrak and Metro North or mix freight with Triboro or mix Metro North, Amtrak and Triboro and leave a track for freight or… Depends on demand growth too. And what else is done in the region. Connecting Jamaica to Newark via Brooklyn diverts traffic from Midtown which leaves more space for Amtrak and Metro North in Penn Station, which then uses up capacity on the Hell’s Gate line that freight could be using….

        • threestationsquare / Anon256

          There are four trackways, so basically yes – two for Triboro rapid transit and two for Amtrak+Metro-North+off-hours freight. The freight is going to have to share with Amtrak or Metro-North somewhere north of the bridge anyway, there’s no other way out of the Bronx.

    • johndmuller

      The Triboro Rx has never really moved me that much. I have several separate problems with it, but overall, I see it as equivalent to a sort of rails to trails thing where we’re taking a potentially irreplaceable heavy rail link and junking it up with a hard to evict tenant that doesn’t really have any more pressing a need than the current freight users.

      The sections in Brooklyn and Queens provide useful interline connections and a new Brooklyn-Queens link. Insofar as the Rx could be built there without impinging on ROW that would be needed in conjunction with the harbor freight tunnel, I am all for it, but where there is insufficient space, the Rx should deal with expanding the ROW up front, not by taking the cop-out of kicking the can down the road.

      As for the section in the Bronx, I wonder if the idea of using that big impressive underused viaduct and monumental bridge, as well as the catchiness of the TRI-boro title have gotten out ahead of the actual utility of the project. It’s not so much that the Pelham Bay line already serves the same area of the Bronx that the Hell Gate line does (Co-op City residents might prefer the improved connection though), but that this routing doesn’t do for the Bronx what it does in Brooklyn and Queens – namely make a lot of cross Bronx connections. A route that crossed the East River on its own between Shea, College Point and Clason Point could cover some unserved territory on its way to Co-op City, where it could then hook up with a cross street like Gun HIll Road to link all the various Bronx lines (or build out the Burke Ave. branch of the IND concourse line, join them all together and make a ridiculously unwieldy Quad-boro loop line).

      Of course, one could form some kind of new discount railway company to operate on the RR ROW’s as a railroad (honoring farecards) rather than a subway and a lot of the ROW issues would go away, you’d just need to figure out how to power the trains.

  3. threestationsquare / Anon256

    Most of these seem very overdetermined and I think have pretty broad consensus support amongst transit advocates outside the relevant agencies. The Boston circumferential line feels like the odd one out though; your and my social circles are probably disproportionately likely to see justifications for such a line, but trying to take a broader view it seems strange to spend billions on a circumferential subway while the radial Green Line routes remain very slow, overcrowded and unimproved. (If anything subway conversion of the B branch might be more justified.) Also while it’s true that much of Boston’s street network is hostile to surface transit, including along the 66 corridor, I think Mass Ave is wide and straight enough for significant bus or LRT improvements of the 1 corridor, which would weaken the case for a circumferential line further out. (Not sure why you mention the 23 and 28 which seem more radial, but their streets are relatively wide and straight as well, and furthermore they potentially make logical eventual extensions of your Washington St light rail route.)

    Given the ridiculous costs of anything in New York City, should consideration be given to a single subway extension of the 2/5 under Flatbush instead of both the 4 under Utica and 2/5 under Nostrand? The case for building both seems pretty strong though (and the construction cost problem needs to be dealt with regardless). The other thing I’m confused about is why your wishlist SAS-Utica connection involves a new subway paralleling the J/M/Z from the Lower East Side to Myrtle Ave, rather than taking advantage (with upgrades if necessary) of the existing underused elevated line.

    • Alon Levy

      Yes, the Boston circumferential is almost certainly the weakest on the list, and I hesitated to include it. As a pure conversion of the 66 and 1, it’s probably not worth it. The ability to replace the 23 and 28 with Red Line to circumferential makes it a bit more interesting, but it’s possible that those lines have too much intermediate-point ridership, far from the Red Line.

      The 2/5 under Flatbush is always one of the proposals, usually competing with Nostrand. I’m in favor of Nostrand mainly because the bus ridership statistics go Utica > Nostrand > Flatbush. I’ve read somewhere, I don’t remember where, that Flatbush also has a water table issue, making it harder to construct than Nostrand.

      Finally, the reason for the SAS-Utica is that it parallels the L from Manhattan to Brooklyn, and provides an easier way into Midtown. But this is very much the less important part of Utica.

        • Henry

          The Williamsburg currently has speed and weight restrictions on it after its rehabilitation way back when. In any case, SAS-Nassau was ruled out due to excessive ground disruption, and it’s not as if the Jamaica Line is suffering from a lack of capacity.

          I’m personally of the opinion that a Hanover-Transit Museum connection would be more useful, with a Utica spur from the Fulton local, but I have no idea how practical that would be.

        • Alon Levy

          It would, yes, but then there’s the whole too many lines on one track issue. I suppose it’s less problematic if it’s the underused Williamsburg Bridge, but three services (J/Z, M, and the new one) sharing the same track can lead to problems. The N/Q/R works by having not a lot of frequency on the N and R, and by interlining them with other services everywhere except southern Brooklyn, which has too many branches to begin with.

          • Adirondacker12800

            If southern Brooklyn has too many branches then there’s no need for extending the 2/5 and the 4 is there?

          • Adirondacker12800

            too many is different than tons of. I may not have arranged them the way they are but they have good service. Less wouldn’t be a good thing, it would be too far to walk to a train. If they have too many southeastern Brooklyn is about right. Let ’em take the bus to the end of the line. Like it would be in southwestern Brooklyn if lines disappeared.

          • Alon Levy

            “Too many” in context means that it is impossible to serve them frequently without causing capacity crunches on the trunk. As a result, some of the express lines (i.e. Sea Beach) are inactive, and the branches that are active get about a train every 8-10 minutes.

            Southeastern Brooklynites indeed take a bus to the end of the subway… or, if they’re going to a non-Manhattan destination, they just drive. The benefit of Utica plus Nostrand plus Triboro is that suddenly they can realistically take trains to Queens without cursing their multi-transfer ride or dreaming of moving to Long Island and buying a car.

          • Adirondacker12800

            That’s what they get for tearing down the 5th Avenue El. Making all the trains to Coney Island run on one line isn’t the solution. The solution is to double deck the tunnel that the LIRR/Metro North/NJ Transit/SIR should dig under the East River and move the trains on the West End/D line to the new tunnels. Change to the D at 39th from the Staten Island train, change to the N and Triboro at 65th.? Up Second Avenue in Manhattan? Or a super express to Wall Street and Grand Central?

          • Eric

            There’s not room to add a third and/or fourth track on the Williamsburg Bridge?

          • Adirondacker12800

            Add two tracks to the Williamburg, where do the trains go when they get to Manhattan? Up the Third Avenue El?

    • betamagellan

      For what it’s worth, the regional plan which suggested an rail urban ring (the same one which sandbagged the Washington St. line, interestingly enough) estimated something like $2.8 billion (2003, IIRC) capital cost and 134,700 riders and a net increase in system ridership of 54,600. It’s high-cost, but also high-benefit (though I agree that a lot in terms of transit funding and operating culture in Massachusetts would have to change first).

      Agreed on 23/28 being a logical extension of a Washinton line—wasn’t a BRT extension down to Mattapan suggested (and shot down due to traffic worries) recently?

  4. Eric Fischer

    Nice list! But “never subway plans” on Geary? There aren’t any current plans, but there were in the 30s, 40s, 50s, 60s, and 70s, and as recently as 2002.

  5. Adirondacker12800

    Just because a bus route is busy doesn’t mean the solution is a subway. Lots of people are getting off the bus to use the subway and vice versa. But not as many as use the bus line. Lots of those trips are local that wouldn’t be served by the subway. Down Utica Avenue probably makes sense, the Crown Heights subway station is much busier than the stations around it. While new tunnel could be built to IND/BMT standards and have IRT trains running in it IND/BMT cars on the IRT doesn’t work as well. There’s that pesky problem of there already being a busy subway line under Bushwick Avenue. Extending down Nostrand or out along Geary probably make sense… I thought the capacity problem on BART was getting people on and off the platforms on Market… Ridership on the 42nd Street and 14th Street buses isn’t all that much different from crosstown streets without subways…. Pity that they didn’t sit down and come up with a more rational integrated system back in the 1890s…

    • Henry

      All train extensions in New York City, regardless of division, are designed to BMT/IND standard, and have been since the Dual Contracts. The Astoria Line used to be IRT gauge, but when it was handed over to the BMT they just shaved the platforms and called it a day.

      • Alex

        So, theoretically, you could just shave back platforms on the 7 line extension and an R-160 would fit there? Interesting.

          • Adirondacker12800

            I think you could leave Main Street and Willets Point/Citifield alone. The cars would be coming in from the Corona Yard wouldn’t they? I think that’s the only place on the line to yard them. I seem to remember it took a day to get rail from the Livonia yard in Brooklyn, where the rail was delivered, to the extension.

          • Patrick O'Hara

            You can’t leave Flushing-Main Street and Mets-Willets Point alone, as they would be out of line with the rest of the line… the widths have to be the same along the entire line…you can’t run R188’s to Flushing and Mets-Willets Point and expect the passengers to jump across the chasm at all of the other stations.

          • Alon Levy

            Flushing and Willets Points can be shaved back, too – the only portion of the line that can’t accommodate R160s is the tunnel from Manhattan to Queens, i.e. the most important infrastructure on the line.

          • Henry

            LIRR’s too expensive and doesn’t have enough capacity. Plus, there’s nothing really wrong with the 7 as it is, and I don’t think you could maintain the structural integrity of the Steinway tunnels by shaving them back any further (since they were originally tramway tunnels, and are thus already shaved back).

            There’s also the question of how practical such a conversion would be, since you’d then get stuck with both having to order a massive CBTC compatible B Divison fleet for the 7 and need to stick the existing Corona Yard fleet somewhere else, at a time when the oldest cars in the A Division still have 10 or 20 years in their working lives left.

      • Adirondacker12800

        but you can’t have the 4 train serve the new station and the train from Second Ave, whatever letter that gets, at the same time. Unless you want to build four platform stations out in the suburban end of the line. And there’s still that pesky problem of there already being a subway under Bushwick Ave. Cutting frequency on busy line to serve a new terminal wouldn’t go over well so there would have to be new tunnels under it. That might get pricey.

        • Henry

          Alon’s plan specifically states that upon the connection to SAS, 4 service to KP will be cut back to Crown Heights again.

          What subway line under Bushwick Av are you talking about? The JMZ is aboveground and a block away, and the L is nowhere near the area being discussed.

          • threestationsquare / Anon256

            Crown Heights/Utica Ave station lies entirely west of Utica Ave and would not under any plan be altered for Division B trains. Alon’s Utica Phase 1 would involve a curve east of this station, and his Utica Phase 2 would involve the elimination of this curve and the construction of a new two-track Division B station under Utica Ave at Eastern Parkway (with a transfer passage to the existing 3/4 station) and tracks continuing up Utica to Williamsburg and SAS.

            I still don’t really see the point of this Phase 2 though, the 4 train seems fine for Utica.

          • Adirondacker12800

            Thats sorta what I come up with. Sounds pricey to me. Building the connection won’t be cheap and then abandoning it before it’s end of life doesn’t sound like very good use of scarce transit dollars.

          • Henry

            The connection wouldn’t necessarily be severed, it just wouldn’t be in service. We have plenty of unused connections lying around the system, including an inter-division one between the Broadway and Flushing Lines.

        • johndmuller

          Adirondacker12800: “but you can’t have the 4 train serve the new station and the train from Second Ave, whatever letter that gets, at the same time. Unless you want to build four platform stations out in the suburban end of the line.…”

          You can accommodate A and B line trains on the same stretch of 4 track line with just two island platforms for any common stops.

          If you are going to run both lines together on Utica (and having an express option for an outer line is a plus), you might want to take the 3instead of the 4 to run with the SAS to be less duplicative in Manhattan.

          • Adirondacker12800

            Four track stations are so cheap to build and maintain. Just so people don’t have to change trains like they do all over the system.

    • Alon Levy

      The 42nd Street buses aren’t actually that busy – the 23rd and 34th Street buses each have more ridership. The 14th Street buses are very busy, but most of that ridership is on the M14D, which goes down Avenue D, which isn’t well-served by subway.

      • Adirondacker12800

        Running a subway under Second Avenue doesn’t help people get up and down Avenue D just like running a subway under 14th doesn’t help them. Running a subway across 125th helps get people off the shuttle and the Flushing line. It will keep people off the buses that cross Central Park. I’m not sure it’s going to be particularly useful for people who use the 125th Street buses for short trips in the wider neighborhood… getting on the bus on 135th on the west side and getting off on 125th on the east side is going to be faster than two subway trains.

        • Henry

          I don’t think that it would take that many people off the Central Park buses; it’d only save you time if you were heading from a point south of Central Park to a point north. Otherwise, you’d be wasting too much time going north, transferring to a line further east or west, and then going back down.

  6. Rob

    City Branch Subway/LRT/BRT
    Location: Philadelphia

    Concept: A grade separated, submerged, underutilized freight rail branch, running 3 miles or ~50 blocks from the former Reading Terminal approach in Center City (converging with the passenger Reading Viaduct) westward above Callowhill St. and along the Ben Franklin Parkway towards the CSX Philadelphia Subdivision. Proposals include an LRT through the cut connecting with the 15 trolley route at Girard Ave, a BRT corridor rerouting crosstown lines through it, or a Broad-Ridge Spur subway extension, the first rapid transit expansion in 40 years.

    Why it’s underrated: Like Triboro RX, the RoW (up to 4 tracks) is virtually intact, allowing any transit to coexist with the single CSX track. Even a tunnel to the underutilized Ridge Spur would probably prove inexpensive as well — subway’s total capital cost could very well be under $500 million, less for LRT/BRT. It would directly serve a transit-starved, rapidly developing community and provide rapid access for millions to the city’s greatest museums, a potential casino, and the Community College of Phila.
    SEPTA had actually studied using the line in the past, but only as a component within the large, unrealized Schuylkill Valley Metro plan and pipe dream (frequent intercity LRT Phila-Reading). The inevitable failure of that project spelled the end of service proposals on many potential transit alignments including the City Branch. Since that time, around 2004, no updated alternatives analysis or feasibility study has been completed, and SEPTA has barely moved on any rail expansion anywhere.

    What is being done right now: Nothing. In recent years, various community organizations have tried to turn the Reading Viaduct into a High Line-style park, but one group, the Friends of the Rail Park, has advocated to turn both the Viaduct and the entire City Branch over to parks and bike trails. Now the singular group for the park plans, FoRP’s ideas have gained support with at least one important group, the Center City District. Not only is a City Branch park an egregious misuse of valuable infrastructure and a redundant duplication of parks and bike lanes on the surface, FoRP’s assertions that a dark, isolated, underground trench would somehow be inviting and safe for anyone to use is unfounded. In response, SEPTA and the Planning Commission have stated they would prefer to save it for BRT, but without tangible studies it’s up for grabs. Given that the city can only work feasibly on one subway project at a time, the City Branch would have to be evaluated against a proposed subway extension to the Navy Yard, which at $700-900M is no small investment.
    Reference: http://hiddencityphila.org/2012/06/what-to-do-with-the-city-branch-return-it-to-transit-as-light-rail/

    • Alon Levy

      Yes. I was thinking about including it, I just wasn’t sure what it would connect to. I guess it could be a BSL branch, but then it would enter the line just south of where the Broad Ridge Spur diverges. It could bend north a bit to actually serve the spur, which would be optimal, but that might require more tunneling (which isn’t necessarily a bad thing, I’m just unsure how much tunneling it would require).

      But connecting to a trolley could also work.

      • Rob

        I think it would be alright to cross over the Broad St trunk and join the Spur at Ridge and Noble, for a possibly new station, because the Spur is terribly underutilized as it is, and integrating it into the City Branch gives it a better purpose than just an infrequent transfer to PATCO. The City Branch could then also interchange with the BSL’s Spring Garden Station, but the transfer passage would be pretty long. To directly connect with the Spur at the flying junction would be a much harder affair.

  7. sajohnston

    Having gone to school in Morningside Heights, I’m a big fan of the 125th Street subway idea. If and when Metro-North finally gets around to bringing the Hudson Line into Penn Station, you’d also be able to connect to those trains on the far western end of 125th. My personal fantasy (and it really is just that, I have no idea how good ridership would be) would be to see the 125th St. line extended under the Hudson to Edgewater, as a kind of substitute for the C that was originally supposed to cross the GWB. Potentially you could even use the old NYS&W tunnel to get under the bluffs (though it would probably have to be expanded) and connect to HBLR.

    • johndmuller

      The 125th St. crosstown line has a lot going for it in terms of interconnections. I imagine that it would be awfully expensive as well. While pedestrian street life on 125th would be greatly disrupted and most of the merchants would fail to last through to the promised land at the end of the rainbow, the everyday street traffic would probably not be as bad as you would think during construction – people do adjust (at least that seemed to be the case when they built the Metro in downtown DC). Afterwards, I would expect a major makeover of the environment and at least a minor real estate boom.

      Despite calling this the 125th St. crosstown line, there may still be some alternatives as far as the exact route is concerned. Probably because of the terrain, 125th deviates from the grid on its western end, following a little valley west northwest to end up at about 130th St by the time it gets to the river. As the 125th St. station on the 1 line is on that course, the crosstown line would probably want to do that too in order to have a suitable transfer point. The 1 line is briefly elevated at that location, so the transfer would not be so straightforward. Perhaps a different station would work better.

      Continuing the line under the Hudson here would I suppose, be possible, although I’m not sure where the final destination on the NJ side would be. Going a bit north after resurfacing would get close to Ft. Lee and the I-80 corridor, where some park and rides could be constructed, or one could go somewhat south and get to Secaucus or a little further and turn the Meadowlands parking lots into a ‘giant’ park and ride.

      Another alternative is provided by the IND trunk line at St Nicholas Ave., which was designed to accommodate a future interchange with a line coming up through Central Park and is consequently equipped with at least 6 tracks up to about 145th St., where the A-C and B-D routes diverge. The crosstown line could turn north onto these tracks without creating any capacity overloads. Further north along this path, at 174-175th St., there is a spur which could be extended to feed onto the GW Bridge, which was built with the capability to handle a heavy rail connection like this. Reconfiguring the lanes on the GWB would be a real political hot potato, but it might be possible to use reversible lanes to add two tracks and still maintain the current auto capacity in the peak direction.

      • Alon Levy

        There already exist underground-to-elevated transfer stations: 74th Street-Broadway in Queens, and Broadway Junction in Brooklyn. Large numbers of people transfer from the 7 to the E and F at 74th every morning; as I recall, there’s an at grade station headhouse where the transfer is done. Doing the same at 125th Street and Broadway would require some work on the station, but nothing extraordinary.

        The GWB line I think is a good idea, but a) it can connect to the C already, and b) it’s a much lower priority than the urban lines. The basic issue is that there isn’t enough along Route 4 to justify a new line, today. If the cities along the route agree to add sidewalks, make it easier to walk from the Paramus shopping centers to the street, and so on, then it’s justified, but otherwise, it’s a problem. There’s also a capacity issue on the A/D, which almost all riders would transfer to from the C, so it requires 125th Street to be built first to help distribute the loads…

        • Adirondacker12800

          Bergen County, with lightly populated rich places and lots of parkland, compared to the Bronx,… is the 25th most densely populated county in the US. When they invent the magic tunnel boring machine, that makes it really cheap to build tunnels, it might make sense to burrow out to Paterson. There aren’t a whole lot of people in Bergen County who want to get to 145th and Broadway or get to Grand Central on a local when they can get there faster by going through Penn Station. Even fewer who want to get to Wall Street on a local via Times Square or Grand Central when they can do that via Hoboken or someday Newport. Getting to downtown Brooklyn is more useful than being able to get to 125th and 3rd.


          Magic tunnel boring machines because the Port Authority went and used the capacity that the trains could use to build the lower level of the George Washington Bridge. On the other hand the Port Authority’s bus terminal hovering over the Trans Manhattan has more passengers than there are Rockland County residents that work in Westchester. If there’s gonna be a cross Hudson rail line, one from Bergen to Upper Manhattan probably makes more sense than one from Rockland to Westchester.

          The shopping centers are unfixable. They are plopped in the middle of a sea of parking. The schlep from the Bronx to go to a chain store that has a branch in the Bronx isn’t worth it. The minimum wage jobs available in the food court are the same kind of jobs available in the Bronx. People in Bergen County aren’t going to get on the train so they can go to a chain store that has a branch in Bergen County.

          • Alon Levy

            It doesn’t have to be a magic TBM: a Route 4 line should be elevated from the bridge to the edge of Paterson.

          • Adirondacker12800

            It’s not an Interstate grade highway. It’s not even a Robert Moses kinda thing. It’s a high speed divided highway with lots of retail along it. The median is a Jersey barrier and there’s a driveway every few hundred feet. What compelling destinations are within walking distance of Route 4 to entice Manhattanites to Bergen County? What makes going to Route 4 a good choice when someone in Bergen County wants to go to Manhattan? 130 blocks away from where they want to go? Versus going the station a few blocks away that has direct service to Penn Station and Grand Central?

          • threestationsquare / Anon256

            @Adirondacker12800: Well, Route 4 reportedly has profitable jitneys running every 2 minutes all day from Paterson to the GWB (including getting off the highway to stop at Garden State Plaza), so evidently somebody’s going that way. There is indeed not much room in the median, which is why Alon said “elevated”. I’m not sure such a line would represent enough improvement over the jitneys to justify the cost, though.

          • johndmuller

            First of all, the Bronx does have quite a lot of parkland – for example, Pelham Bay Park, Van Cortlandt Park, Bronx Park , with the Zoo and NY Botanic Gardens. As for the Landed Gentry, you’ll have to settle for Riverdale, which I suppose is a little light on the land part of the ratio. The Bronx does have some IND lines as well; and if you think the roads are chaotic in Bergen County, …

            As for the topic, if the SAS were built with the 125th St. Crosstown connecting to the IND north up to and over the GWB, I think that a new Park and Ride Palace ™ where the NJ Turnpike and I-80 join would be an absolute gold mine (and/or sites like the Meadowlands existing parking and/or someplace around Hackensack). Assuming you route an express over there, it would be only 6 stops to Lex and 125th (throwing in a stop at Ft. Lee), then onward down the SAS proper to 42nd, 34th, Wall St. etc. I’d be surprised if you could beat that in your car.

            Granted you would have to drive to those park and ride sites, so it might develop into a sometimes thing, whether you kept on driving or got off to park, depending the weather, your mood, exactly where you were going and most important probably your reading of the traffic that day. It might start out as a kind of self regulating traffic modulator, but it could end up capturing many of the erstwhile drivers who work near one of the IND mainlines, once they’d given it a try.

            If the MTA got a piece of the parking action, it might even be a (gasp!) profitable enterprise.

          • Adirondacker12800

            There aren’t going to be trains over the George Washington Bridge. They used the capacity to build Martha, the part under George. If you don’t want to tunnel all the way to Paterson you need Reardon Metal to be able to put the stanchions for the elevated, in the few inches that is the top of a Jersey Barrier. That’s the unused space along Route 4. They’ve been talking about increased capacity across the Hudson for 30 years. They’ve been talking about the Second Avenue subway for 85. I suspect NJTransit and Amtrak to Grand Central is going to happen before Second Avenue Subway to 42nd Street.

      • threestationsquare / Anon256

        In fairness it’s not quite that simple, the six-track section ends a bit north of 135th so that junction would need significant rebuilding to accommodate trains from SAS without decreasing capacity on the existing lines. The flying junction needed to join the line south of 135th would also be difficult and disruptive to build. Of course this has to be weighed against the disruption of tunnelling further west under 125th, and both alternatives deserve consideration.

        • johndmuller

          That’s the same map I’ve been looking at; so the 145th to 135th is somewhat speculative, but I imagine the hard part is not so much figuring out the new configuration, but accomplishing it with minimal disruption to operations. Further south, if they were already planning to do something like this, there may even be some groundwork already laid (or at least sketches of possibilities), but even if not, it looks as straightforward as one could hope for.

          I also like to think that the bridge could be reconfigured with a minimal cost to auto throughput, although the construction would be nightmarish in all likelihood. The simplest plan I have is to replace the gap between the east and west roadways on the lower level and one lane from each side with the railroad ROW and if necessary, make the 2 center lanes on the upper level switchable so that during rush hours, there would be an extra lane in the peak direction. That would be no net loss in the peak direction and a minus 2 lanes in the off peak direction. I think there are 7 lanes each way now, this plan would have 6 each way normally and 7 and 5 lanes during rush hours (or just 7 and 5 all the time switching at midday and midnight) using something like the movable Jersey barrier on the Tappan Zee.

          • Adirondacker12800

            If they could easily implement reversible lanes on the GW they would have. Decades ago.

          • Henry

            Considering that accident rates went up significantly after the Tappan Zee had its zipper lane installed, that may not be the kind of precedent we want to follow.

      • Henry

        125th might very well be difficult, but for a completely different reason; the street is in a valley (hence why the street grid in the west is irregular) and is the location of a fault line.

  8. tompw

    “The basic rule of this post is that if it’s being built, or is on an official urban wishlist pending finding the budget for it, then it’s not underrated”
    Toronto’s downtown relief line (DRL) is on two official wishlists: the City’s, and that of regional planning body Metrolinx. For the latter, it’s on the recently-amended version of their Regional Plan “The Big Move”. So it’s not really under-rated.

    Union station is crowded with GO Train riders – if they take the subway, they are going against the peak subway flow (which is southbound).

    Metrolinx is currently studying all the things that could be done that would serve a similar function to the DRL. That includes things that relieve pressure on Bloor-Yonge but are quicker/cheaper than a new subway line, and things that allow for shorter DRL when it eventually gets built. (And also, how to maximise the benefits of the DRL).

    The City of Toronto is currently studying exact alignments for the ‘core’ section (downtown east and north to Bloor subway). This will include potential for extensions west from downtown and north to the Bloor subway; and further north (on the eastern side) to the Eglinton LRT (under construction)

  9. Henry

    I personally would’ve picked a Queens Blvd Bypass over the 125th St line, but that’s just a bias of mine.

    Would Nostrand and Utica really take precedence over, say, Third/Webster in the Bronx? After all, the Bronx-bound subway lines are much more crowded than their Brooklyn counterparts.

    • threestationsquare / Anon256

      The crowding numbers for the Bronx-bound lines in the Hub Bound Study / Cap’n Transit’s post are for the lines crossing 60th St when they’re full of riders from Upper Manhattan as well as from the Bronx. There’s not really room on the existing lines through Upper Manhattan for more Bronx-bound riders, so any expansions in the Bronx require expansions in Manhattan as well, greatly increasing cost. Maybe SAS will have spare capacity after it opens but maybe it won’t. Utica/Nostrand on the other hand can be built without overburdening the existing Manhattan infrastructure.

    • Alon Levy

      Well, Nostrand and Utica have more riders. The Uptown/Bronx lines have bigger capacity problems, which means they need to get more trunk lines in Manhattan, but they have a lot of branches serving most of the Bronx. The main way to relieve the capacity problem is SAS, which removes a lot of the traffic originating on the Upper East Side, and I think the 125th Street crosstown would also help redistribute some traffic from the 2/3 and to a lesser extent the A/D. (The A/D is very fast from 125th to Midtown; the 2/3 is circuitous and has gnarly curves.)

      In Brooklyn the only real capacity crunches are on the L, which can probably be solved with electronics (and if not, then with digging 50 meters of tunnel at the end of 8th Avenue for tail tracks), and the A/C. But southeastern Brooklyn has a huge service gap instead.

  10. Moaz Ahmad

    Toronto’s experience with the Sheppard subway line suggests that building a subway line to replace a busy suburban bus route is not always the best idea.

    Both the Sheppard subway line and the Downtown Relief Line were part of Toronto’s Network 2011 plan (introduced in 1985) along with a short extension of the University-Spadina line and the Eglinton West line.

    Toronto has unfortunately let politics get in the way of planning. The Network 2011 plan is an example. It was going to connect the “centres” of 4 of the 6 cities that made up Metropolitan Toronto… North York & Scarborough (via the Sheppard line), York (via Eglinton West) and East York (via the 2nd phase of the DRL) to the subway network. The Eglinton West line was originally going to be a bus rapid transit line out to Pearson Airport and Mississauga but was upgraded to subway to please the City of York and show that the government was ready to invest in a lower – income area.

    When a new government was elected in the 1994 they decided to cancel Network 2011…but Metropolitan Toronto councillor Jack Layton made a deal with North York mayor Mel Lastman to support the Sheppard subway line. Unfortunately the need to connect the line to the existing network (and the Davisville Yard) necessitated a costly wye at Sheppard and Yonge. Combine this with budget cutbacks by the new government and the line was cut back to Don Mills Road , which is why Toronto calls it the Sheppard ‘stub’way.

    I personally think Toronto would have been much better off building a crosstown LRT on Sheppard Avenue extending westwards to the old Downsview Airport and eastwards to Scarborough. The De Haviland plant at Downsview could have been repurposed to build LRVs and there would have been lots of room for a yard and maintenance facility.

    As for the DRL…my personal opinion is that it is long overdue and should be built as a crescent shaped line running from Don Mills Rd and Eglinton to the Canadian National Exhibition grounds, via Toronto’s downtown.

    Cheers, Moaz

  11. Matthew

    Brookline killed the Longwood avenue subway idea pretty quickly, they would freak at the notion of digging up Harvard street. I don’t think there’s any chance of replacing the 66 with a subway.

    I think the easiest and most compelling transit improvement is installation of proper signal priority and PoP to speed up the radial green lines, and upgrades to buses as well. Boston may never dig again, way things are looking.

    • threestationsquare / Anon256

      I’d say the most important subway project in Boston is the Blue Line to Charles, which is still on some official wishlists. While currently sandbagged, it could still plausibly happen within the next decade or two if it gained a political sponsor.

      • Matthew

        Yeah, that and Blue line to Lynn. Also Orange line to Roslindale. All these have appeared on official wish lists at some point though.

        One idea that hasn’t appeared officially is to build a light rail subway alongside the Mass Pike trench from Chinatown to Prudential. This allows reuse of Tremont street tunnel, retirement of flat junction at Copley, relief of central subway, and new service patterns between points west and South Boston, as well as Roxbury. There would be a three way junction/station under Eliot Norton park. Presumably this is one of the few places in Boston where utilities are well known thanks to urban renewal clearing.

  12. Michael Noda

    I regret to inform you that, having fallen off of all of the wish lists at SEPTA, DVRPC, and the Mayor’s Office of Transportation and Utilities, the Roosevelt Boulevard Subway to Northeast Philadelphia is now eligible to appear on this list.

    When it was last studied in the early 2000s, it had a higher projected ridership than any transit project proposal in the nation, with the sole exception of Second Avenue.

    • Rob

      It’s DVRPC that’s trying to shift the dialogue away from a subway extension; they want a “fresh start” with alternatives analyses on SBS-style BRT-lite, BRT, and light rail exclusively — which is very inadequate for even the 2000s ridership projections. Apparently DVRPC thinks it’s impossible for the 5th largest US city to spend a few billion on rapid transit expansion or any rail-related capital project, but not impossible to spend over $5 billion on the PHL airport expansion.

      • Michael Noda

        DVRPC has it half-right; transit on the Boulevard needs help a lot faster than even the most optimistic subway construction schedule can provide it. Bare minimum would be BRT or BRT-lite from Holme to Bustleton (and then down Bustleton to FTC). But dropping all consideration of a subway extension in favor of the immediate need for low-order transit is pants-on-head stupid.

  13. Eric

    What about an elevated line along the whole length of Utica and Malcolm X, connecting to the J/M/Z?

    This would be much cheaper than a subway extension of the 4 or SAS, and would also serve a greater number of north/south riders. Utica is rather wide so an elevated line would be tolerable in visual terms.

    Of course NIMBY would oppose it, but if it’s elevated or nothing…

    • Michael H (@ComplxBlackness)

      Cheaper it would be, but also very awkward with the hillyness of Crown Heights and the switches required to turn it and the M between their respective lines and Broadway along the main Broadway elevated.

  14. betamagellan

    Looking to smaller cities two routes come to mind:

    Although I have no idea about the status of the project (it might be too active to count as underrated) or which alternative is preferred (I suspect the more at-grade ones are, though), a Belltown-Ballard line in Seattle would connect some dense, constricted by traffic bottleneck, and currently rail-free neighborhoods. Should one of the more ambitious alternatives eventually get the go-ahead, Link would finally have a corridor that doesn’t feel like a pale shadow of the rejected circa-1970 metro plan.

    Going even smaller there there’s the re-use of the old Chicago & Northwestern railway corridor in Milwaukee as light rail, which is still around as a vague “recommended electric bus guideway or light rail corridor” in the Milwaukee’s 2035 regional plan but unlikely to ever see the light of day due to the planned installation of a streetcar on a parallel, street-running route and the popularity of the corridor as a bike highway (light rail and the bike trail could coexist, but there would definitely be resistance from active transportation types). With the exception of street-running to reach UWM and downtown (not pictured), the line is entirely grade-separated, allowing for quick travel times. Although it borders the lake for part of its route, density and commercial activity is also concentrated along Prospect and Farwell, which run one and two blocks west of the C&NW route (respectively); riders from the segment bordering the Milwaukee River would likely be captured by a station at North or one of the at Kenwood & Oakland, making it a non-issue. In addition to high adjacent population densities, this is also the only part of Milwaukee that’s fully walkable and where parking is difficult. There’s also a large student population, major employment centers at each edge of the line, and even without any fixed-guideway transit there’s been a lot of densification since the line was first seriously studied circa 1990. I can’t find the regional light rail study conducted then online, but IIRC ridership density along this segment would have been comparable to a heritage system like Muni Metro or the Green Line, with the caveat that this was when there was still an optimistic bias in projections.

    The corridor, though, would not have any effect on regional highway congestion (it serves neighborhoods with no direct highway access) and even though its ridership density was highest it didn’t have the highest number of riders per corridor analyzed. This led to a futile attempt to link Milwaukee and Waukesha by light rail, which in the long run ended up paralyzing any decent attempt of building passenger rail in the Milwaukee metro area. This, plus a change in transit fashion, has led to the current streetcar plan. I can understand using slower, lower-capacity vehicles on a line that doesn’t have to compete with expressways, but could you at least put them in the ditch where they could at least run quickly and reliably enough to have a hope of competing with the automobile?

    • Eric

      In addition to Downtown-UWM, Downtown-Zoo in a segregated lane of Wisconsin Ave is a viable route due to the hospitals and universities along it. In fact these two segments could be joined into a single quite cheap light rail line joining most of the Milwaukee area’s main destinations.

  15. betamagellan

    Ooh, and three from LA:

    Tejon: HSR, not transit, but the undisputed King of the Sandbagged

    Vermont Subway: Has this been actively discussed since the early nineties?

    Trenched Red Line Extension: Although most of the discussion about dealing with the Orange Line capacity issues centers around light rail conversion, I believe that for a while there was serious consideration of the Red Line in a closed trench. If there’s going to be an upgrade, a partial conversion of the Orange Line to Red Line Metro might be a good idea. Although I couldn’t find per-station ridership data after a couple of minutes’ Googling, I did find a suggestion that frequency from Reseda west be increased and that most Orange Line riders arrive by bus. Cutting out an Orange Line/Red Line traveler would likely mean not only higher speeds one less transfer penalty for many riders. Furthermore, not that far east of Reseda the Orange Line splits between Time Warner and Chatsworth anyway, which is a BRT-appropriate branching service pattern.

    • Alon Levy

      I was thinking of the Vermont subway, too, but I didn’t find bus ridership numbers for Vermont and Western. This tool says 26,000 Vermont (204), 14,000 Normandie (206), and 21,000 Western (207); it is also a difficult to use, and I can’t even figure out what all the Wilshire buses are.

      • EJ

        If the bus ridership numbers are that high, then it definitely justifies a subway. The average speed of a “Metro Rapid” bus is 8 miles per hour, which means it’s even less during rush hour. Skipping stops and signal priority doesn’t help much if you’re stuck in the same traffic as everyone else, so actual rapid transit would attract a lot more riders.

        • betamagellan

          It also helps that the old Vermont subway plans weren’t totally subway—I think the southernmost five or six kilometers were envisioned as elevated rail.

          • Alon Levy

            How far do the Vermont subway ideas go? Do they go south to the Green Line at Athens, or all the way to San Pedro?

          • betamagellan

            @letsgola Originally (and this was as of the initial subway-construction push in LA, which still remains in the background in long-range plans) it was to through-run along with what’s now the red and purple lines, so third rail was assumed (although the 51 line in Amsterdam switches from wide-median-running-with-overhead-wires to grade-separated third rail at Station Zuid, so still through-running isn’t impossible even if at-grade).

            There was someone who posted a ton of renderings and info from the old plans at skyscraperpage, though I can’t find them at the moment and the image links might not still work. I’m fairly certain it would have only gone to the Green Line, though.

          • letsgola

            There’s also the Blue Line in Boston, which switches from third rail to overhead at Airport. You’d have to buy new rolling stock, but you’d have to do that anyway for the extension. You’d still have issues w/ captive fleet unless you retrofitted existing cars, but the cost savings from going at grade might be worth it.

          • betamagellan

            @letsgola Although elevated rail does have advantages in terms of speed, reliability, and capacity, plus the rolling stock issue you mentioned. I’ve also been stuck at Station-Zuid when there was an issue with the pantograph failing to retract (the tunnels it runs through don’t have clearance), so you add mechanical complexity, too; the Yellow Line in Chicago used to switch between overhead and third-rail on the fly and I recall they switched to all third-rail partly for speed and reliability reasons, too. The fact that a third rail/overhead is done so rarely makes me think that these little issues add up in the long run

          • Adirondacker12800

            Metro North switches from third rail to catenary of vice versa all day every day. They or their predecessors have been doing it since the New Haven electrified.

      • anonymouse

        There’s also the 754 and 757 which are the Vermont and Western rapid buses and add 20k and 16k respectively. As for the southern end of the line, the Green Line makes a natural terminus, but if you want to go further south, the Harbor Gateway Transit Center would also be a logical endpoint and transfer to buses heading further south.

    • Adirondacker12800

      One of the reasons the Triboro will be cheap because it could use an existing bridge to get across the East River. The bridge goes to the Bronx

    • threestationsquare / Anon256

      No. A major point of the 125th St Subway is to get riders from the various west side lines to points on the east side of Manhattan between 125th and 59th (without requiring travel via Midtown). Connecting to Triboro would force these riders to transfer twice for the benefit of the far smaller numbers heading from the west side to the (mostly residential) parts of Queens served by the TriboroRX route. Furthermore, the TriboroRX route is on a very high viaduct on Randall’s Island after crossing the Hell Gate Bridge, and the construction/grade issues to get from there to a subway under 125th are nearly insurmountable. SAS by contrast will already have a station under 125th between Park and Lex (oriented east-west) after phase 2, so extension west under 125th is a straight shot.

      • Adirondacker12800

        To get from the west side of Manhattan to the east side of Manhattan doesn’t require a subway line that goes to Queens. One that goes to Wall Street might be nice. Or Brooklyn. Using the Hell’s Gate Bridge and the viaducts leading up to it makes service between Queens and the Bronx really cheap. People in Queens who want to go to 125th Street, both of them, can take the Triboro to a station in the Bronx and use one of the existing lines to go a few stops. There is a transportation method that is great for carrying a few people to places not many people want to go, it’s called a car.

      • Henry

        I mean, presumably a line from Randalls to 125th will also allow for crosstown subway travel. I’m not entirely convinced of the utility of the Bronx leg of the RX, especially considering that the most recent iterations have it going to Hunts Point instead of actually connecting things in the Bronx.

        If it is feasible, it wouldn’t be so bad to have both Second Avenue and RX trains going to 125th and Broadway on the same pair of tracks.

        • Alon Levy

          The utility is that it provides a direct connection from the Bronx to Queens. At least in the original RPA version it hits Yankee Stadium, Melrose, and the 2 and 6, though the connection to the 6 requires a lot of walking.

          • Adirondacker12800

            One of the iterations is to use the abandoned ROW that runs under/along St. Mary’s Park. It’s very close to 143rd on the 6. It misses the 2/5. But that doesn’t mean a new station on 6 or 2/5 couldn’t be built.

          • Alon Levy

            …why did I think that it’s the other way around – that it’s very close to the 2/5 but misses the 6?

          • Adirondacker12800

            I dunno. There’s something going on where it passes under the 2 and 5. Old maps show the Third Ave El having service on Westchester Ave in addition to Third. I hazard a guess that is a remnants of the connection.

  16. letsgola

    I don’t think we have anything in LA that would qualify, since most valuable projects are being actively discussed. One possibility would be rapid transit type service on the core of some Metrolink lines.

    Boston Silver Line would absolutely make sense as a Green Line branch. From a track perspective, there’s no way the four-track section between Boylston and Park St is at capacity. I suppose someone could raise questions about platform berthing capacity at Park St itself, but eastbound seems to work today with almost no trains using the inside track. Problem westbound is people have to pick the right branch though.

    Personally, I don’t see Boston’s Urban Ring being justified at subway level costs of construction unless the cities agree to major upzoning along the route. A southern terminal at Dudley or Andrew would be weak; you might be able to get more by going the extra mile for an actual stop at UMass. The logical connections (Northeastern, Longwood Medical, and Kenmore) in Boston don’t lend themselves to a great route, while following the 66 wouldn’t hit many big destinations. Mass Av could have bus lanes for the entire route of the 1, at least during peak hour, and that would make it a good option. Unfortunately, routes like the 47 and the 66 are always going to be tough surface routes, since Boston just isn’t designed for cross-town travel.

    Sort of off-topic, but how is the LA Orange Line not a success?

    • Henry

      LA’s Orange Line is the best example of how not to go about building BRT. It was really meant for rail or light rail, but the only reason they couldn’t do either was because it was explicitly banned. At this point, Orange Line capacity is completely maxed out, and it should probably be replaced by rail in the future, as some of the third-world BRT cities have already started doing.

      • letsgola

        Orange Line capacity is not maxed out. They’re running 60′ articulated buses on 4 minutes headways. They could decrease headways or they could start running two buses each trip (platforms are long enough to double berth). 4 minute headways are an artificial constraint imposed by LADOT’s unwillingness to incur any more motorist delay on north-south arterials.

        The Orange Line is pretty fast and reliable, it’s attracted more ridership than people expected, and it’s led to people demanding more investment in transit in the SF Valley. That seems pretty successful to me.

      • letsgola

        Edit: I would say that Boston’s Silver Line is a good example of how not to build BRT. Curb running, so it competes with right turns and parking, plus it doesn’t even have bus lanes in Downtown Boston or Dudley Square where they’d actually be useful.

        • Henry

          I largely meant “not to go about building it” as “crappy reasons to bulid it”. Wording mishap on my part.

    • Alon Levy

      The Orange Line’s having capacity issues because of the lack of signal priority, and that’s with ridership that isn’t even that impressive by Geary, First/Second, Utica, or (Vancouver) Broadway standards.

      And yes, LA doesn’t have anything like Utica that’s not already under active consideration. The Vermont subway is the main example, but it’s in the second tier category together with the Boston circumferential. This mainly speaks to LA’s success in actually finding money for its most important routes, on a timescale of a few decades, together with low-performing stuff like the Foothills extension.

    • LAifer

      Let’s please note forget a Sepulveda Pass subway/LRV/BRT. Lots of projects being built in LA right now, but this is the most glaring missing piece. We need something to connect the SFV with the Westside that isn’t more lanes on the 405 or more freeway lanes elsewhere (will just get jammed up as quickly as they’re built).

      • letsgola

        I think we can leave Sepulveda Pass out though, since it’s *definitely* on everyone’s wish list and would likely receive funding from Measure R2 (god willing). It’s not imminent, but it’s got a better shot than most of the projects on this list.

  17. mdahmus

    Austin’s 2000 light rail plan (or what can be built in-street as the off-street portion is now the Red Line). Hurts my feelings it wasn’t included!

  18. Christof Spieler

    Chicago: conversion of Metra Electric to rapid transit (no need to ditch the overhead wires, just run frequently and upgrade stations) with a northwards extension in subway under North Michigan Ave.

    • Alon Levy

      Pretty much any good commuter rail project would be in this category: Metra Electric, Clinton Street terminal with more Metra electrification, North-South Rail Link, Penn Station through-running with connection to Grand Central, Erie-Fulton-Flatbush, Caltrain electrification with timed overtakes and a Transbay connection done right, Metrolink electrification (and the Harbor Sub line) once the run-through tracks open. There’s a reason I deliberately excluded commuter rail from this exercise – the projects either stink (under current operating practices and frequencies) or are amazing (under modernized practices).

      • betamagellan

        Thinking of the semi-official and perpetually under study ideas for improving Metra, the big problems are:

        •For the Gray Line/Gold Line proposals on the south/southeast sides, rolling stock—I haven’t seen anyone ever mention switching to single-level (FRA compliant or not) vehicles optimized for frequent urban transit. They all tend to assume the continued use of the current, heavy bilevels.

        •On the non-electric lines I’ve never seen any official document mention anything about level boarding, ever. Nor have I ever seen anyone but me bring it up in railfan discussions.

        •Freight loads on the BNSF and UP-W lines, at least, are heavy enough that full passenger/freight temporal separation is infeasible, making non-FRA rolling stock impossible without lots of new concrete. EMU’s aren’t generally discussed either*—every official or semi-official discussion about Metra electrification has assumed a switch to electric locomotives.

        •No one’s looking at staffing, operating practices, fare collection, etc., and the fact that the somewhat antiquated practices help keep Metra a suburbs-to-downtown-only service means that their political constituency is broadly happy with it.

        • Adirondacker12800

          The destination in the suburbs is “my bed” If the other attraction in town is a Rite Aid and a Safeway along with a bank, the post office and other small retail and the attraction in the next town is a Walgreen’s an Albertson’s along with a bank, the post office and other small retail there’s really not much of a reason to go to the next town up or down the line. And even less to go the town five stops away where they have a Rite Aid, an Albertson’s along with a bank, the post office and other small retail. People in the suburbs take the train because it’s faster than driving. For the occasional destination five towns away – the restaurant with Albanian Mongolian fusion – the walk to the station, the wait for the train, even if it’s on a fairly frequent schedule, and the walk to the restaurant takes longer than driving there. Everyone in the suburbs has their very own car or access to one. Congestion isn’t very bad and the parking is easy. There’s never going to be much demand in the suburbs for suburb-suburb travel, except maybe for 9-5 work destinations. It’s too easy to drive and too easy to park. For the small amount of destinations.

          • betamagellan

            Metra runs through the City of Chicago, has stations in dense areas both in the city and in the suburbs, and reverse commuter ridership has been rising (it’s often at the cost of a bus transfer, though the dedicated “shuttle bug” services to office parks make it fairly easy). Although some of the big reverse commute markets are horribly located for Metra-based reverse commuters, there’s enough employment and job density to make Metra improvements worthwhile on at least some of the lines.

      • nei

        Caltrain electrification and an extension to downtown is planned, with reasonable schedule patterns, it should be one of the better transit projects underway.

      • Michael Schaeffer

        Then write a post about the amazing commuter rail projects under modernized practices.

  19. threestationsquare / Anon256

    Alon, do you know of many cases of such clearly worthwhile projects being neglected by official planners outside North America? To what extent is planning really worse in North America, and to what extent does it merely seem that way from the point of view of the transit blogosphere?

    • Alon Levy

      In Tel Aviv, the proposed subway route isn’t the best, and the subway-surface technology is wrong for what the city needs. There’s a plan to electrify commuter rail in the Haifa area, but oddly not in the Tel Aviv area, or nationwide (the national network’s small and dense enough to justify it). The Jerusalem light rail plan is based more on the national need to annex the East Jerusalem settlements than on the urban needs of the city; in both Jerusalem and Tel Aviv there are some neglected routes, purposely in Jerusalem’s case and accidentally in Tel Aviv’s.

      In Singapore there are some questionable investment plans (i.e. the self-intersecting Downtown Line), but I don’t know if there’s any equivalent of Utica or Geary there.

      • Shlomo

        There is a plan to electrify all rail in Israel (commuter rail is not really distinguishable from intercity passenger rail in Israel), but not all at once.

        In both Tel Aviv and Jerusalem, you are thinking of the first route in a system that supposedly one day will contain multiple routes. (In Tel Aviv, the planning of the first route is so incompetent that it looks like the second route might even end up being finished first.) When those routes are taken into account, the coverage is much better.

        In Jerusalem, the topography imposes severe constraints on both the routing and viability of light rail (many neighborhoods which exist or could be build close to a light rail station do not actually have easy pedestrian access to that station, due to the hills and windy streets).

      • threestationsquare / Anon256

        Singapore’s decision to abandon the form KTM railway line and cut back service from Malaysia to the isolated “Woodlands Train Checkpoint” seems rather strange by the standards of most countries, especially given the lack of apparent plans for other uses of the ROW or for replacement links with Malaysia.

        The UK has construction costs that sometimes approach those of the US, which might suggest similar issues with poor planning or mismanagement, but projects like Crossrail seem on the whole well-planned given their circumstances, and I struggle to think of obvious omissions from the official wishlists — certainly nothing as glaring as you list here.

    • betamagellan

      Having recently moved to Amsterdam, I’ve become aware that there’s a slate of metro projects that are in the same development hell as most American transit projects. However, the cost-benefit of a lot of these is dampened up by Amsterdam’s difficult hydrology, which makes tunneling very expensive (that said, I’m not all that impressed by the reliability and upkeep of the metro as is).

      This seems to be Amsterdam-specific, as transit in Rotterdam is the next best thing to a transporter, though there are a lot of complaints about the regional rail operator and its unwillingness (likely in part politically-constrained inability) to add capacity.

      • Eric

        Does Amsterdam really need any tunneling after the north-south line is completed? When I visited, the existing subway seemed pretty empty. The city is small and the trams seem to do a good job for many trips.

        • betamagellan

          The subway’s pretty well-used, but it’s emphatically not-for-tourists, mainly connecting the main regional/intercity train stations with office clusters (also not in the central city, partly due to aesthetics and partly due to hydrology) and peripheral, often lower-income, districts (which might be why the Metro’s maintenance and cleaning seems a little deficient compared to other European cities). I think there’s a decent case for extending the North-South line into Amsterdam-Noord, a west lines to Osdorp/De Akers and Geuzenveld-Slotermeer, and a Sloterdijk-Amsterdam-Noord connection (all of which have been floated in the past), but those would all likewise be mostly useless for tourists and potentially sunk (ha!) by horrible tunneling costs.

  20. Robin

    Calgary: 8th Avenue Subway. A ~1.6km LRT tunnel for the NW-S line through downtown. It would end interlining of the two lines along the surface ROW on 7 Avenue which has been at capacity for over five years.

    It might not apply since it is on long term city wish lists but it seems to be in a similar place as Toronto’s DRL. There hasn’t been any serious movement on the project despite the need and it remains behind almost every other project in the city as a priority.

  21. Jonah

    Great post!

    Random note – For the sake of clarity, San Francisco’s Japanese neighborhood is usually referred to as “Japantown.” The name “Little Osaka” is often used as a lesser nickname for Sawtelle in Los Angeles, one of the region’s 3 or so Japanese enclaves (along with Little Tokyo and Torrance.)

  22. Nathanael

    This is a list of projects which should obviously be built, yes. Several have been aggressively sandbagged for reasons I do not understand, including the Geary Subway.

  23. Pingback: What Elites Do Instead of Providing Services | Pedestrian Observations
  24. Pingback: The Utica Subway | Pedestrian Observations
  25. Pingback: The RPA’s Fourth Regional Plan | Pedestrian Observations

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.