Are Express Trains Worth It?

So, you have your urban rail line. It’s mostly above ground, so constructing new express overtakes is feasible. It has decent frequency, and carries trains to destinations at a variety of distances from city center. But it’s not an overcrowded subway line that brushes up against line capacity, requiring all trains to run at the same speed. Do you run express trains?

I’m going to focus on regional rail in this post, since with two Tokyo-area exceptions, proper subways are incapable of running express trains without dedicated express tracks due to their high frequency. On a line with a train every 10 minutes it’s feasible to mix trains of different speeds with timed overtakes; on a line with a train every 2 minutes, it’s not. I’m going to use the LIRR and Caltrain as examples, and then apply the derived general principles to other cases in the US, including future regional rail schemes.

The basic tradeoff of express service is that it provides faster service to the express stations at the cost of frequency at the local ones. This can be done in two ways: expresses that stop once every few stations, and local-then-express patterns. Jarrett Walker calls this limited versus express, based on bus service patterns; with trains, both types are called express. The subway in New York, the Chuo Rapid Line, Seoul Subway Line 1, and Caltrain baby bullets are examples of the first kind; the Caltrain limited-stop trains and the peak-hour trains on some LIRR lines are examples of the second kind.

Express trains of either kind but especially the first reduce line capacity, even with very long overtake segments. If train X overtakes train L, then there needs to be an available slot ahead of train L, and after the overtake there’s a slot opening up behind L. The Chuo Rapid Line runs a mixture of local (“rapid”) and express (“special rapid”) trains for most of the day, but at rush hour, there are only local trains, peaking at 28 trains per hour; on the shoulders of rush hour, there are some express trains, with total traffic of about 20 tph. The LIRR runs 23 tph on the Main Line at the peak, so this is an issue, which the LIRR unsatisfyingly resolves by running trains one-way at rush hour. It’s less an issue on Caltrain given constructable overtake locations, but right now the overtake locations are inconvenient and the trains are pulled by diesel locomotives, increasing the stop penalty and reducing the capacity of a mixed local-express line.

The second kind of express service is bad industry practice and should not be used. It avoids the capacity problems of the first kind at low traffic levels, but at high traffic levels the speed difference is still too large. It is used when the trains are a special CBD shuttle and makes it impossible to serve passengers who are cheap to serve, i.e. those getting off short of city center. Caltrain’s limited-stop trains do this because of capacity problems during rush hour, when they need to get out of the baby bullets’ way. The LIRR does this because of a cultural belief that trains exist only to shuttle people from Long Island to Manhattan and back; due to the same belief, it runs trains one-way at rush hour rather than giving up on rush hour express runs as JR East does.

The first kind of express service may or may not be warranted. It depends on the following questions:

1. What is the line’s expected traffic level? Low traffic, up to about 4 tph for a regional line, favors an all-local configuration to prevent cutting local stations’ frequency unacceptably. Very high traffic favors all-local configuration for capacity reasons, or else investment into long overtakes or even full four-tracking. Intermediate traffic, in the 6-12 tph range, is the best zone for express trains.

2. Have local trains already been sped up by use of good industry practices? Level boarding, high-acceleration EMUs, better track maintenance allowing higher speeds between stations, good timetable adherence allowing less schedule padding, and infrastructure preventing delays on one train from cascading to others allowing even less padding can all significantly reduce the speed difference between local and express trains. In some extreme cases, a local train can end up not much slower than an express train hauled by a diesel locomotive.

3. How long is the line, and how many stations does it have? Longer lines and shorter interstations both favor express trains, all else being equal. Intercity rail, which also has higher stop penalties because of the higher line speed, deserves more than one stopping pattern even at low frequencies.

4. How big is the difference between minor and major stations? It is crucial not to confuse current ridership with ridership potential, since lines with express service often pick winners and losers, after which the better-served express stations steal riders who live closer to bypassed minor stops. This is common on Caltrain, where some but not all express stops are major job centers.

5. Can intercity trains plausible substitute for express service?

It is question 4 that makes the difference in many cases. On the LIRR, the Main Line has a clear distinction between major stops (Mineola, Hicksville) and minor ones (all the rest). The Montauk Line does not. Note the ridership levels of the stations, going eastward from Jamaica to the end of electrification:

Main Line:

Hollis: 114
Queens Village: 791
Floral Park: 1495.5
New Hyde Park: 1725.5
Merillon Avenue: 766.5
Mineola: 5174
Carle Place: 386
Westbury: 1951.5
Hicksville: 8107.5
Syosset: 2748.5
Cold Spring Harbor: 2083
Huntington: 5556.5
Bethpage: 2481.5
Farmingdale: 2312.5
Pinelawn: 25
Wyandach: 1758.5
Deer Park: 2708.5
Brentwood: 1375
Central Islip: 1787
Ronkonkoma: 8639

Montauk Line:

St. Albans: 93.5
Lynbrook: 2738
Rockville Centre: 3425
Baldwin: 3371.5
Freeport: 2514.5
Merrick: 3383.5
Bellmore: 3267.5
Wantagh: 2890.5
Seaford: 1804
Massapequa: 2959.5
Massapequa Park: 1672.5
Amityville: 1542.5
Copiague: 1430.5
Lindenhurst: 1791.5
Babylon: 3293

There are three ends of electrification: Babylon, Huntington, and Ronkonkoma. All have markedly more ridership than nearby stations, especially Ronkonkoma, though in all cases it’s an artifact of their being the ends of electrification, with many people driving in from farther east. Ronkonkoma has nothing nearby that justifies its ridership level, the highest of any suburban LIRR station; it’s a park-and-ride that has a lot of ridership because it’s the end of electrification and has express service.

In contrast, in Mineola and Hicksville, there really is a concentration of activity justifying their status. Both have trivial transit usage as job centers, but there’s enough of a core, especially around Mineola, to justify higher service, and Hicksville is also the junction of the Main Line with the Port Jefferson Branch: see the census bureau’s OnTheMap tool.

But there are no special stations on the Montauk Line. Excluding St. Albans, which is in New York itself and has to compete with cheaper and more frequent if slower bus-to-subway options, the ratio between the busiest and least busy stations is 2.4:1. A similarly flat situation occurs east of Hicksville, excluding the two end-of-electrification stations.

What this means is that the LIRR should only run local trains on the Babylon Branch and east of Hicksville, while maintaining express service on the Main Line west of Hicksville when there’s enough capacity for it. A similar analysis of other lines in the New York area should give the following answers:

Hempstead, West Hempstead, Long Beach, and Far Rockaway Branches: all local due to short length.

Port Washington Branch: probably all local due to short length, but if additional local stations are added in Queens, then some express trains to Great Neck may be warranted.

New Haven Line: very long, sharp distinction between major and minor stops all the way but especially west of Stamford, high frequency, four tracks give enough capacity for everything. The current configuration of nonstop trains to Stamford continuing as local to New Haven and local trains turning at Stamford is fine, except that the express trains should also stop at New Rochelle (a junction with the Hell Gate Line, which deserves service, but also a major stop in and of itself, with the third highest weekday ridership of Metro-North’s suburban stations) and maybe also Greenwich; HSR overtake considerations may require stopping also at Rye and Port Chester.

Harlem Line: generally favors local trains, except that White Plains is a major job center and thus a far more important stop than all others, independently of its better service. There are four tracks south of Wakefield, favoring express trains, but conversely charging subway fares and allowing free transfers to the subway would lead to a ridership spike as people switch from the overcrowded 4 and 5 trains. There’s a big dropoff in ridership north of North White Plains, so the current configuration of locals that turn at North White Plains and expresses that go nonstop south of White Plains is fine, as long as off-peak frequency is raised.

Hudson Line: favors express trains because of length and four-tracking. Although on paper there are more and less important stations, this is an artifact of service patterns. The secondary stations in Yonkers serve higher density than the busier stations in the proper suburbs, and the dense parts near Tarrytown are actually in Sleepy Hollow, about equidistant from the Tarrytown and Philipse Manor stations: see the New York Times’ population density map.

Erie Lines and West Shore Line: probably all local since the population density thins too uniformly going north, with Paterson as the major exception. There are somewhat denser anchors at the outer ends of some lines – Spring Valley and Nyack – but Harlem Line-style nonstops run against a capacity problem, coming from the fact that this part of the network is necessarily highly branched.

Rest of New Jersey Transit: the main lines (Northeast Corridor, Morristown) are very long and have some distinguished suburban job and population centers (Metropark, New Brunswick Morristown) deserving express service, but the branches (North Jersey Coast, Montclair, Gladstone) do not. However, the fare structure and off-peak frequency lead to much less ridership on the inner-urban segments in Newark, Orange, etc., than would be expected based on population density. In addition, the difference between major and minor stops is fairly small on all lines when taking electrification into account, sometimes as small as on the Babylon Branch: see ridership data per line and per station.

Although my initial decision in my regional rail plan to pair the Erie lines with the Atlantic and Babylon Branches of the LIRR was aesthetic, creating a northwest-to-southeast line, in reality the systems are fairly similar in their characteristics. More or less the same can be said about the Staten Island-Harlem system. There are no direct connections to intercity rail except at Jamaica and in the Metro-North tunnel to Grand Central, the lines pass through urban or dense-suburban areas, the interstations are fairly short, and there’s relatively little distinction between major and minor stops. (White Plains is the major exception, and Paterson is a secondary one.) This makes the Lower Manhattan-based system much more RER-like than the Penn Station-based one, which is longer-distance and practically intercity at places.

Finally, the same set of questions in the other three major Northeastern cities generally lead to the conclusion that no express trains are needed.

In Boston, there’s too little difference between major and minor stops on each line (see PDF-page 70) – somewhat more than on the Babylon Branch, but much less than on the LIRR Main Line. The most prominent major station is Salem, but the low-ridership stations farther in on the Rockport/Newburyport Line are in working-class suburbs; the ridership there is depressed because of fare and schedule issues coming from competition with buses, and good regional rail would get much more additional ridership from Lynn and Chelsea than from Salem and the suburbs farther out.

In Washington, current traffic demand is so low that express service would seriously eat away at the frequency offered to local stations. MARC and VRE ridership is so low that any analysis of travel demand has to start from geographic and demographic information rather than from preexisting ridership; the only major outlying destination on any of the lines is Baltimore, which can be connected to Washington by intercity rail, and which conversely has much less Washington-bound commuter traffic than the Washington suburbs. The closest thing to justifiable express service is that when the commuter lines closely parallel Metro, they should have wider stop spacing.

In Philadelphia, on most lines, express service eats away at frequency too much. The one exception is the PRR Main Line, with the SEPTA Main Line a possibility. Many lines have sharp differences between local and express stations: for example, Cornwells Heights on the Trenton Line is much busier than the rest. But a combination of low frequency and lack of easy overtakes (on the Trenton Line, the inner tracks should be mainly used by intercity trains, with only the occasional regional rail overtake if required) makes this not useful. The PRR Main Line actually has less difference between major and minor stops than many others, but it is longer and has short interstations and higher frequency. The SEPTA Main Line has the frequency to support multiple stopping patterns, though the population density near the minor stations is high and the problem, as in other Northeastern cities, is high fares and lack of integration with urban transit.


  1. Tom West

    “The second kind of express service [local-then-express] is bad industry practice and should not be used”.
    I disagree (but of course :-)… it wouldn’t be a blog comment without some disagreement!). This pattern is used in commuter lines into London (UK) and Toronto (Canada). The purpose of an express is to save time. The local-then-express provides that time saving for everyone on the outer portion of line, while preserving the one-seat rides that commuters value so highly (more so than business or leisure riders). The limited-stop only does that for people at certain stops – everyone else must transfer to (possibly) achieve a time benefit. Unless your all-stops trains are very frequent, it is highly unlikely that everyone gets a short transfer time from the all-stops train to the limited-stop train.

    “It is used when the trains are a special CBD shuttle and makes it impossible to serve passengers who are cheap to serve, i.e. those getting off short of city center”
    Firstly, it’s not impossible. You provide an all-stops on the inner portion of the line, with a well-time transfer at the change-over between local-then-express and the inner-all-stops. (We have to get some better words here…). Because there’s only one transfer to schedule, you can make it very easy. (Further, because it travels down the line a few minutes after the express, it minimises the chance that it holds up the *next* local-then-express).
    Secondly, these people aren’t always cheap to serve… you are providing capacity all the way into the city centre, and therefore people who get off before the centre may leave an empty seat all the way to the city centre. Thus, the cost to the service provider is the same as if they rode to the city centre, but the revenue is less. (Of course, their seat may be filled by someone else getting on. This all depends on travel patterns. Dislaimer: Your City’s Travel Patterns May Not Match The Ones Described Here. Consult Your Transport Planner Before Taking This Advice. Etc…)

    The limited-stops express make most sense where you have many major trip attractors (generally job centres) along the length of the route. What you then end up with is a hub-and-spoke arrangement – it’s just the spokes are parallel to the lines connecting the hubs! Users generalyl have to travel a short distance along the line on a local (feeder/spoke) service to the nearest hub (express stop), then take an express (inter-hub) route to the destination (another hub).
    This pattern generally occurs either at a city scale (which is why you see it so often with buses – the hubs are major intersections), or at a national scale (the hubs are large cities, and you have inter-city and local train service). It’s not something you see very often at regional/metropolis level. (Caltrain seems to be the only exception I can think of).

    • Beta Magellan

      I suspect that one of the reasons GO sticks with local-then-express is because of its dinosaurian diesels. Metra is much the same way—in some areas it has very short station spacing, but, with the obvious exception of Metra Electric, is completely unelectrified. That, combined with the fact that all Metra lines (again, sans Metra Electric) lack in level boarding (i. e. cross-platform platforms would be a hassle in places with multiple tracks). Still, on many of the lines station spacing is usually long enough that zones can probably be eliminated while keeping travel times competitive.

      There might be an exception with Metra’s BNSF line—almost all inbound commuters are downtown-bound (so few intersuburban commuters), station spacing is very short for almost the entire stretch from I-355 to the City of Chicago, and many of the existing zoned expresses tend to fill to standing capacity. It’s a good candidate for modernization (i. e. EMUs and high platforms), but likely an expensive one due to heavy freight volumes (there’s also the issue of the BNSF line poorly serving suburban employment centers and the likely resistance of most station area communities to densification, meaning that, as of now, there’s little reason to enough expect all that much ridership growth east of I-355).

      • Tom West

        If diesel-hauled locos are the reason, then what about SE England? (All EMUs with acceleration that makes it hard to stand!).

        GO uses its local-then-express for the reasons I described – it works better for their customers (and hence for GO). If GO switched to EMUs, then it would still keep the same stopping patterns.

        • Beta Magellan

          AFAIK, English diesels aren’t heavy, slowly-accelerating North American models.

          GO would keep the same stopping patterns because it plans on switching from trains pulled by overweight diesels to trains pulled by overweight electrics. Although, based on what I’ve read, there’ll be some improvement from current schedules, it won’t be much (Caltrain judged electric loco-pulled trains not quick enough to solve their capacity issues).

          • Tom West

            I think you’ve missed my point 🙂 SE England uses speedy EMUs with rapid acceleartion, and they use the local-then-express model. So, I don’t think vehicle technology/acceleration is the most important thing here.

            No-one’s refuted my main point – which is that limited-stop generally means poor transfer times for connecting services, unlike local-then-express.

          • Alon Levy

            Limited-stop expresses require special infrastructure to time transfers, namely overtake locations. The same is true of local-then-express trains. The two-pattern limited-stop express pattern bandied about on Caltrain-HSR Compatibility gives passengers from beyond Hillsdale a zero-penalty transfer, while passengers north of Hillsdale can just stay on the local.

          • Tom West

            Alon – you say “Limited-stop expresses require special infrastructure to time transfers, namely overtake locations. The same is true of local-then-express trains.” Almost… the latter requires just *one* overtake location, and just *one* point where you need good transfers.

          • Alon Levy

            One point per change from local to express. For limited-stop express, it’s one point per overtake. But note that local-then-express may also require overtakes if there’s too much speed difference.

          • dejv

            Tom West:
            I think you’ve missed my point 🙂 SE England uses speedy EMUs with rapid acceleartion, and they use the local-then-express model. So, I don’t think vehicle technology/acceleration is the most important thing here.

            IMO you nailed it in the last paragraph of the thread starter. If you have rather uniform density and patronage of each station throughout the line, local-then-express seems better. Conversely, if there are more stations that stand out in terms of their ridership, the line is better candidate for local and limited pattern.

    • Alon Levy

      I don’t know what the schedules in London are, but in Toronto they’re pretty bad – it’s not just the diesels, it’s general North American commuter rail badness. GO arguably doesn’t have local trains at all, given the very wide stop spacing, especially within Toronto itself.

      The local-then-express deprives people riding outside the CBD the one-seat ride, and also tends to have a nontrivial transfer time. At Caltrain, it’s about 5 minutes. It requires additional station facilities to ensure zero-penalty transfers: the station has to have not only a track in each direction but also one or two center tracks for terminating trains, with island platforms between the center track(s) and the two running side tracks. Transferring from a local to an express is common enough that passengers just expect to do it on the 6, the R, and the 1; it’s not like those CBD-end transfers in Austin or (for people not working near Union Station) Toronto that make commuters just drive all the way in instead. In contrast, those 5-minute transfer times are a serious competitiveness problem for people who are not working in the CBD.

      The empty seat can easily be filled by an inner-urban commuter to the CBD. Again, this is routine on every subway system. If it doesn’t happen, it means the outer stations are so much busier than the inner ones that more people work at the inner stations than live there, and this means the inner station in question is a CBD, which has to be served by all trains. Definitionally the reason to run locals rather than local-then-express trains is to make it easier to use rail to get to job centers that aren’t true CBDs, such as Forest Hills or Rockville Centre or malls near train stations (this is how SkyTrain operates). If some outer stations are much busier than the inner ones then it also justifies limited-stop express runs as on the LIRR Main Line.

      • Nathanael

        “(for people not working near Union Station) ”
        This is the historical key to this. The “local in the suburbs, then express to downtown” pattern was devised in an era when *practically everyone* was working near Union Station. (Or the equivalent in another city.) The decentralization of the job market has substantially weakened the case for express trains.

        • Tom West

          “The decentralization of the job market has substantially weakened the case for express trains.”
          That would be true if the new job centres were near the stations skipped by express trains, or if people living near those stations worked at the new job centres. However, neither is the case.

      • robertwightman

        The problem is that you have to haul those empty seats a long way if every train starts at the outer end and you still want to carry the same number of passengers. GO was never designed to be a major carrier of people within the city proper. As long as it follows those stupid FRA/TC rules it never will. (I love your opinion of the FRA. The same can be said about TC Transport Canada.)

        Since 95% of GO’s rail users get of or on at Union Station their “local then Express” seems to work well. If we could only get decent EMUs then we could run a much more useful service that also served the city better. At least GO only uses 3 person crews, motorman, conductor and door operator (Customer Service Agent.)

        • Tom West

          “we could only get decent EMUs then we could run a much more useful service that also served the city better”
          EMUs would allow slightly faster service (about 5 minutes saved for an all-stops service from the outermost stations). A much more useful service would run every 15 minutes on all lines, and that can be done with diesel trains and the current signalling system.
          (All-day service on the non-Lakeshore lines requires the samre track upgrades whether you use EMUs or diesel locos).

    • BruceMcF

      As far as language, I am familiar with the Australian styling of Limited, Express and All-Stops, and into that mix I would drop the “local-then-express” and “express-then-local” as semi-express services.

      There are, of course, multiple patterns of semi-express. There are (all inbound, invert for the outbound):
      (1) All-Stops then Express. These are where there is are clear priority destination stations and a drop-off in ridership outside Marchetti’s constant, with the express portion both pushing the threshold Marchetti’s constant station out and serving the destination district as a true express, since the patronage from outside the destination district will detrain in increments at each in-destination-district station.

      (2) Express then All-Stops. This is particularly suited to the situation where there is a zone with substantially higher walk-up patronage closer to the destination district and substantially higher transfer patronage (recruiter buses, park and ride, etc.) further away, especially if the outer district is unlikely to effectively sustain a 10min all-stop frequency for trains with the capacity to function as locals in the destination district. This is clearly an issue for a geographically smaller urbanized area than NYC, but it IS an issue that is encountered. The express stations may then recruit from locals that themselves do not actually run into the primary destination district.

      (3) All-Stops then Express then All-Stops. Where the destination district line collects traffic from multiple inbound lines, typically in a Y alignment, this has the Express component for the same reason as (1) taking the express portion from partway down one arm of the Y to the junction and then running all-stops along the common portion. For the Y configuration, the capacity limit on the individual arms for mixed all-stops and semi-express is not a binding capacity constraint, which is on the common trunk corridor. The complementary All-Stops service may have its outer terminus at the beginning of the Express portion of the semi-express, so that the Semi-Express Y stretches out further than the All-Stops Y, and have a locked cross platform transfer at an Express station in the middle of the Express part of the Semi-Express.

      (4) Station skipper. The All-stops and Express route portions are more finely grained, with two versions of the route ensuring that all stations are served each two frequencies, the Express stops every frequency, three versions ensuring that all stations are served each three frequencies.

      None of these ~ not even station skippers, and I am not a fan of station skippers in general ~ can be ruled out a priori … each needs to be evaluated in terms of the population and trips distribution of the area being served.

  2. Christopher Parker

    You left out the biggest reason for express service: improved equipment utilization.

    Suppose you leave the center city with a fully loaded 10 car local. At each stop more and more people get off until half way to the end of the line the train is half empty. Keep doing this all day or all rush hour. Even though the train leaves full, your load factor will be 50%, assuming an even distribution of passengers at stops. Toward the end of the line you are mostly hauling air.

    Now suppose you make each of these trains a five car train that runs express to the middle of the line and then makes the stops. Your load factor will be 75% — much more respectable. In the mean time you can take the five cars you didn’t bring and run another train for the stops closer to the city — but this time the train reaches it’s end point in half the time, since it only goes halfway. It can then turn back and make another run. So if you had 6 trains before, each dedicating 5 cars to the closer commuters (for a total of 30 cars), now you can run that service with only 15 cars, since the trains complete their turns in half the time.

    What’s the most expensive cost in providing train service? That’s right — the equipment! Between buying it new and maintaining it, in rough numbers it can be 40% of the cost (your mileage may vary, so to speak . . . ). By running express trains you can save money. Marginally more crew expense can save the biggest expense.

    Maybe in a perfect world the service wouldn’t be resource constrained, but it always is, so running it more efficiently allows for more service.

    • Alon Levy

      If the problem is that the trains empty out as they leave the center, then short-turn some of them at appropriate places, a common practice on subway systems that were built in stages and have turnback facilities at through-stations that used to be termini. Branching can be especially effective at splitting demand to lesser-used outlying destinations, and avoids the line capacity problem coming from having a train turn on running tracks.

      The advantage of this is that for the same cost, it increases frequency to inner-suburban stations, at the expense of reducing frequency to outer-suburban stations. But it’s the inner-suburban stations where higher off-peak frequency can be the most effective at promoting off-peak transit usage. New Haven-New York is an intercity run using commuter rail equipment and branding; it’s New Rochelle-New York that’s more interesting as an S-Bahn city pair.

      • BruceMcF

        And branching lines are where running semi-express services along the outer branches can be effective in recruiting patronage that would not be interested in riding at the time of travel of an all-stops from their origin to their destination. That is, however, a Local-Express-Local service pattern, with the express on an inner segment of each branch, and all-stops on the trunk line for the capacity ~ and because within the destination district that allows it to serve as an additional all-stops frequency, replenishing some of the patronage that gets off at earlier stations in the main destination district.

      • robertwightman

        The trouble with short turning commuter trains is that those who travel to the end of the line might not get a seat because all of the local passengers will get on the first train that comes regardless of its destination. In the a.m. rush with its shorter, heavier peek many inbound passengers would not get a seat. This is a good way to deter ridership on a commuter rail system that supposedly tries to deliver a seat to everyone.

  3. Christopher Parker

    One reason the CalTrain limited service might make sense when it wouldn’t elsewhere is that the commuting pattern in Silicon Valley is not as centered on the San Francisco and San Jose centers.

    • Caelestor

      Caltrain limited service makes some sense because SF overshadows every other stop north of RWC job-wise, and the mid-peninsula overtake that would enable limited stop service and timed-transfer overtakes hasn’t been constructed yet.

      Currently, traditional peak service is fairly reasonable, though Caltrain overemphasizes fast service to SF at the cost of reasonable headways at many stations. It wouldn’t be to hard to rectify the current shortcomings: having both Baby Bullet trains make the 6 same stops (Sunnyvale, Mountain View, Palo Alto, Redwood City, Hillsdale, Millbrae) and turning the existing :50 limited train into another local-express train would be the two biggest changes.

      On the other hand, the reverse peak timetable needs a serious overhaul, since it’s based around the flawed premise of express service to SJ, when most riders disembark between RWC and MV. I’d keep one baby bullet service to SJ to placate political interests, but the other 4 trains should run almost, if not entirely local in Santa Clara Country.

  4. Mike

    SEPTA’s use of expresses mostly stems from the intercity history of their longer lines. It’s also a reflection of their failure to take advantage of the center city tunnel and run a frequent, S-bahn type service.

    • mpdl3280a

      No, the reason SEPTA must still continue to use expresses is the continued use of redundant legacy stations leaving extremely short spacing more comparable to rapid transit. Running times to Center City are actually slightly worse then they were almost 100 years ago. Also the heavily branched nature of the system leaves adequate frequency on the Main Lines, as the outer branches aren’t expressed.

  5. Henry

    For the most part, the LIRR service pattern works exactly as you recommend – most Long Beach, Far Rock, Hempstead, and West Hempstead trains are all local (at least within Long Island). It’s when you get west of the county line that you start getting the really funky service patterns – there are almost no trains during the peak that make each and every local stop on the Main Line, and some trains run express to Atlantic (which doesn’t make sense, given the fact that the line only has two tracks and that the stopping penalty for the two stops isn’t outrageous).

    The Port Washington Line also usually has express trains skipping Flushing, Murray Hill, and sometimes Auburndale, but that’s because the bulk of its ridership is actually in Bayside and points east, and because Murray Hill has a short platform and is situated on a large curve that requires more dwell time (people almost never get in the right cars to disembark)

  6. Ben Ross

    Two minor facts – The LIRR has another electrification terminus you don’t mention, at
    East Williston.

    MARC runs two expresses each rush hour. In the pm, one train leaves Washington and goes non-stop to BWI (a major stop) and makes all stops beyond BWI. A second train just behind it serves just the stations between Washington and BWI.

  7. Ant6n

    “This can be done in two ways: expresses that stop once every few stations, and local-then-express patterns.” I find this a bit unclear (taken from NYC transit dictionary?), what exactly do you mean?

    • Alon Levy

      See graphic on Human Transit. The first kind of express is what Jarrett calls rapid or limited, the second kind is what he calls express; this is the standard terminology for New York buses and I think also West Coast buses (certainly true of Vancouver).

      • Ted K.

        Here’s another example – San Francisco’s Muni .

        Two corridors are served with local, limited, and express service (e.g. #14, #14L, #14X and #38, #38L, #38AX, #38BX). Other corridors have local + limited or local + express (see link above). The “Description” tab on the page one goes to after clicking on the name link (left column) makes the differences clear.

        I personally use the #14L to get from the outer terminal to Eighth + Mission (near the Main Library) in about forty (40) minutes. The #14 would add fifteen to twenty (15 – 20) minutes on to that trip time.

  8. Ryan

    The Metro-North practice of running local-then-express out of New Haven is one of the few things I don’t have a problem with – while I agree there’s probably value in having trains out of New Haven that only stop in Bridgeport, South Norwalk, Stamford, Rye and New Rochelle, those trains should be an entirely new service rather than a half-assed reconfiguration of the existing local-then-express runs. The local-than-express runs from New Haven should stay as they are, and bogging down those runs with an extra stop (or two, or three) is going to cause more problems than it solves.

    The New Haven Line as a whole really ought to be treated as two separate lines – the Stamford Line, and the line between Stamford and New Haven. As one complete unit, it’s really frankly too long to be treated as anything other than an intercity line, and I think we can agree that trying to run an intercity line as a regional line is a bad idea.

    The same is true in Rhode Island, with the South County Commuter Rail – which, as a Rhode Island line, should be the vehicle for moving people between South County and Providence as its own service distinct from local trains out of Providence – not an abortive extension of those locals. The entire run to Boston out of Kingston is too long to be attractive to anyone other than broke URI students and the entire run from Westerly is just too long, period. Increasing speeds would help some, but the real solution is to adopt a local-than-express pattern and have South County trains running express north of Providence. Similarly, having even a nominal regional rail schedule into Kingston, Westerly, and Mystic (via an extended Shore Line East) would reduce the need to bog down any future Northeast Regional trains with stops in Westerly and/or Mystic that they probably don’t need to be making.

    • Alon Levy

      I don’t actually mind the Stamford expresses that much – ignoring HSR compatibility considerations, there are three extra stops south of Stamford that I’d insert: Greenwich, New Rochelle, and Fordham. This is specifically because the distinction between major and minor stations north of Stamford is not as big as that between New York and Stamford. True limited-stop expresses stopping at Stamford, Bridgeport, and New Haven – let’s also add South Norwalk for good measure – are intercity rail anyway, and should be run as such, perhaps continuing onward to Shore Line East.

      However, the locals only start stopping at Mount Vernon East; passengers traveling diagonally between the Harlem and New Haven Line need not apply. If it were possible for trains to stop at Wakefield then that would be the natural connection point – maybe new platforms at the interlocking that let New Haven Line trains stop near the preexisting Wakefield platforms. Either way, Fordham’s important enough as a reverse commute origin, and could be very important as a local station if in-city fares were the same as on the subway, and all local trains at least should serve it.

      It’s the Babylon Branch local-then-express-then-local-then-express trains that are truly mad.

      • crazytrainmatt

        I understand that the New Haven trains used to stop at the branch point (Woodlawn), until Fordham lobbied to have it changed there instead, and because of the operating agreement, only one stop in the Bronx is still allowed. In addition to the craziness about not being able to board inbound New Haven line trains at Fordham, Metro-North won’t ticket you a diagonal trip from the New Haven to Harlem lines, transferring at Fordham, even though it’s legal — you need to buy two tickets. I think you can do it with monthly passes since they are cross-valid, although connections can be ~20min.

        Not a good start, given that East Side access and Penn Station access are going to create two more transfer points for diagonal trips involving the LIRR and NJT.

  9. Richard Mlynarik

    I’m not going to presume to comment on most of your examples in light of that “Wovon man nicht sprechen kann, darüber muß man schweigen” jive.

    But I’ll generally note that it can be productive to think about some of the routes involved as several different sectors/markets glued end-to-end and also as different ones overlaid on top of each other. In addition, there is a qualitative difference between “900 pound gorilla” markets and lighter (it goes come back to gravitational analogy, doesn’t it?) regional ones.

    “Glued together”, one might have several intermediate regional sub-centres, perhaps meriting their own regional “S-bahn”.

    If the centres are sufficiently separated and their economic spheres of influence compact, one ends up with the model combination of “inter-city”-like trains (“limiteds” in the odd jargon we’re using here, serving the “overlaid on top” market) running between important stops, and “local” trains shuttling back and forth in captive orbits around the massive stops.

    If the zones surrounding regional centres start to bleed together (aka “continuous sprawl hell”), then the the regional “S-Bahns” start to overlap and, for basic logistical reasons, one ends up with major city pairs served by both “limited” and “local” trains.
    In the Caltrain example, a hypothetical Palo Alto S-Bahn and Redwood City S-Bahn and Mountain View S-Bahn might end up signing some sort of co-operative joint service agreement.
    Likewise the “SF-Redwood-Palo Alto Regional Express” and the “Palo Alto-Mountain View-SJ Regional Express” services have some overlap, so some (but not necessarily all) of those service runs might be physically implemented using single through-running train.

    If the regional centres are sufficiently close or the number of minor local-only stations is sparse, the distinction between “limited” and “local” becomes unhelpful and it makes sense to run a single combined (“all limited” or “all local”, name it as you will) service between them.
    As “Caelestor” correctly notes above, the part of the Caltrain line between Palo Alto and San Jose matches this last scenario, with relatively few and widely-spaced stops and relatively even potential (if not currently, for reasons of poor existing service) demand.

    Turning to “900 pound gorillas”, where one centre dominates a region and extends its commute-shed and economic zone of influence past those of its satellites, it’s practical and useful to design some service in a “collector-distributor” manner, what you’re calling “local-then-express” here. (I was confused by this when just skimming through at first, since I though it meant a service pattern of a local train followed by an express train.) The mega-centre draws in passengers from all over, but the more distant the stops the sparser the pickings, so one sweeps up riders from the outskirts until the train is sufficiently loaded, then zooms offs and dumps them all in the black hole lurking at the heart of the system. But note that this service combination of “collection” and central “distribution” needn’t be accomplished by a single train: it might be a sub-centre’s existing S-Bahn “local” that sweeps up its suib-region, then meets up (across the platform, how else would one even dream of doing it?) with the mega-regional “limited” inter-city-style train which performs the distribution. One service pattern, one passenger flow, but overlaid on two physical trains and two independently valuable services. Glue enough of such bled-togther sub-centres and you end up with locals running end-to-end meeting up with and being overtaken by limiteds running end-to-end.

    When there aren’t insufficiently powerful sub-centres, or where the mega-region suffers from total mega-centre dominance, the S-Bahns wither away to nothing and one ends up with “local-then-express” service provided with one physical train because there are no viable independent utile pieces from which to fashion it.

    (PS The correct corridor service for Caltrain is the paired combination of “local-then-express” SJ-local-Redwood-limited-SF overlaid with Redwood-local-SF and timed transfers in Redwood. Off-off-peak this can be cut back to SJ-local-Redwood plus Redwood-local-SF and “transfer” to the same train on the same platform, which of course just means SJ-local-SF. “S-Bahn SF” and half of “S-Bahn Redwood” provided by Redwood-local-SF. Half of “S-Bahn Redwood” and all of “S-Bahn PA” and “S-Bahn MV” and “S-Bahn SJ” provided by SJ-local-Redwood. “Limited-then-express” to SF mega-centre provided by SJ-local-Redwood-limited-SF. Route-wide “limited” provided by the same train, an acceptable compromise given the wider station stops SJ-Redwood which blur “local” and “limited”.)

  10. Miles Bader

    The Tokyu Toyoko line runs express trains during peak hours (24tph); about every 3rd train is an express. The expresses themselves alternate between “ordinary” expresses (skip somewhat more than half of stations), and “commuter limited” expresses (which skip most stations).

    This works really really well generally.

    [See the 08: slot in the first table (Shibuya direction) here:

    Oh, and the Toyoko line uses timed overtakes, there are no dedicated express tracks.

    • Andrew in Ezo

      To add to Miles’ comment, another line in the Tokyo area, the Tobu Railway Tojo Line, has a configuration more along the “900 pound gorilla” example given by Richard, that gorilla being the Ikebukuro terminus, and the line extending out to the Saitama suburbs (the line itself is 75km long if you count the rather rural extreme outer end which has sparser service), but with some “sub-center local train in their captive orbits” per Richard. The interesting thing is the last 10km of the line into Ikebukuro is served only by local trains both peak and off-peak. The expresses and rapids make numerous stops in the outer suburbs and then run as limited stop services the remainder of the way, making their last stop before Ikebukuro at that 10km point at Narimasu- passengers needing to get to those inner-city destinations make a cross-platform transfer to a local here. Most locals turnback at Narimasu, but a couple per hour go on to Kawagoe City. As for the middle portion of the line, the local services are provided by run-through trains off the Tokyo Metro Fukutoshin Line* and Yurakucho Line (via the junction at Kotake Mukaihara), so these trains serve a dual purpose of providing local service and as a connection to destinations in southern Kanto and the Tokyo CBD.

      *some of these trains continue on to the Tokyu Toyoko Line in Miles’ territory.

    • Andrew in Ezo

      To supplement the schedule linked by Miles, here is a string diagram of the Toyoko Line for the AM peak:
      The bottom in the terminal at Motomachi Chikagai, the top is Ikebukuro (Fukutoshin Line). The colored lines represent the following services:
      -Top legend (Fukutoshin Line)
      Red: express
      Green: commuter express
      Black: local

      -Bottom legend (Toyoko Line)
      Pink: commuter limited express
      Brown: express
      Blue: local
      Grey: local (terminates at Shibuya)

      As you can see, the greatest number of timed overtakes are performed at Jiyugaoka (自由が丘). This is a good place to do this as this is also a transfer point to the Tokyu Oimachi Line (an important cross-city route in southern Tokyo).

      The other two string diagrams are for off-peak and (the last one) evening rush services.

      • Richard Mlynarik

        Miles and Andrew: thanks so much for the Japanese examples. (Especially the in the One True Global Universal Language of time-distance string diagrams.)
        I’m too old and stupid to ever make any linguistic headway on being able to web search anything in Asia for myself, and while I’ve done a little Japanese tourism (and of course have peculiar proclivities that make me pay more attention to train schedules than most), it’s the unglamorous everyday workhorse lines which are most valuable and interesting.
        Forever marooned in iso-8859-1! Thanks again and please do continue.

        • Andrew in Ezo

          Richard, I share your interest in the language of string diagrams. Thankfully there is even a monthly magazine here that covers such issues (and others that occasionally publish articles about them), and those lucky enough having an acquaintance working for a railway often leads to acquiring expired diagrams from the previous timetable regime:) My next study topic is how the diagram is modified in cases of disruption due to accidents, weather, etc.- a whole field in itself and quite critical especially with the expansion of run-through services.

  11. JJJJ

    Local then express is very beneficial in locations with lots of older dense suburbs, like the NJT northeast corridor line. Your express from new york running to new brunswick, princeton and trenton is still going to be packed, and it doesnt encourage sprawl because those are dense cities.

    That being said, NJT does NOT have a good transfer policy. Try to go from elizabeth to new brunswick during rush hour – you cant unless you go back into newark. Theres no shared stop where you can transfer from an “inner” train to an “outer” train (ie, rahway or metropark).

    For the MBTA, they SHOULD offer more express service to worcester and providence for the same reason, those are “real” cities.

    If anyone is going to suffer a penalty, it should be those who live in sprawland.

    • Alon Levy

      The NJT Northeast Corridor Line is a lot like the Metro-North New Haven Line and the LIRR Main Line: it has a lot of big secondary nodes. The platonic ideal of an express train matched to service needs should stop at Metropark rather than expressing from New Brunswick to Newark; arguably it should also stop at Elizabeth. It requires a bit more schedule discipline on NJT’s behalf because it assumes express commuter trains share tracks with slower trains at least between New Brunswick and Metropark, but current traffic levels permit this with a lot of slack. However, it’s more compatible with future HSR, since the local-express speed difference is still much smaller than the express-HSR speed difference, so HSR should hog the inner tracks and NJT should stay on the outer 2-4 tracks.

      All of those lines – NJT NEC, Metro-North New Haven, LIRR Main – are long, and in the case of the NJT NEC, the stop spacing south of New Brunswick is also very wide. About the only thing other than HSR that could reasonably skip Hamilton is a renewed Clocker, and even that I’m doubtful about. (Adirondacker will say such service should run via West Trenton anyway.)

      • Adirondacker12800

        Pity NJTransit didn’t consult with you. They are busy dividing the line into Trenton–South Brunswick, South Brunswick-Rahway and Rahway-New York. They are going to leave the North Jersey Coast divided into New York-Rahway, Rahway-Matawan. Matawan-Long Branch and Long Branch-Bay Head.

        The West Trenton line has enough people clustered along it that it would have decent frequencies. More than half of the people who use Trenton and Hamilton originate in Pennsylvania and are destined for Manhattan. It makes sense to run a local in Bucks and Montgomery along the Trenton line and another along the West Trenton line that then expresses to Newark and New York. ( Newark so the Wall Streeters can change for PATH )

        • Joey

          I’m still not seeing why NJT service along the West Trenton line is so important. There are a few towns along it but for the most part density falls off sharply once it splits from the Raritan Valley Line. That, and the fact that it’s always going to have a lot of freight.

          • Adirondacker12800

            To get the 5,000 Pennsylvanians who commute to Manhattan on a train in Pennsylvania instead of clogging the bridges to New Jersey.

          • Joey

            Then why not just extent the NEC trains into Pennsylvania rather than developing a new route which misses most of the population?

          • Adirondacker12800

            Both of them, as in “along the Trenton line and another along the West Trenton line”

            SEPTA ridership on either line is about the same which implies that population is about the same. 2,500 current riders along each implies there is enough demand to do something cheap and simple.

            the NEC trains into Pennsylvania

            NJTransit offered to extend service to Morrisville, just across the river from Trenton when they were expanding the yards in Morrisville. SEPTA refused. So the empty trains go across the river to turn around or get stored and Pennsylvanians have to go to Trenton or Hamilton to get to New York.

          • Joey

            So once again it’s a question of agency turf wars? How would the West Trenton line help that any? And if your goal is to serve destinations along SEPTA’s West Trenton Line, why not just run trains along the NS Morrisville Subdivision from Morrisville to Woodbourne (along which there’s plenty of room for additional tracks if needed) rather than trying to restore service all the way from Bound Brook to West Trenton?

          • Alon Levy

            For what it’s worth, of the 2,000 Manhattan-bound commuters who live in Bucks County, the greater concentration seems to be closer to the West Trenton Line than to the Trenton Line. The biggest source township is Lower Makefield, with about 600. That said, in the parts of Bucks County near Trenton the two lines are closely parallel.

          • Adirondacker12800

            the NEC trains into Pennsylvania

            Why not send the nearly empty SEPTA trains going to Trenton in the morning all the way to New York? And orginate the nearly empty SEPTA trains leaving Trenton in the afternoon in New York? Those trains match the platform length in SEPTA territory better than NJTransit 12 car multilevels. Match the demand better too.

          • Joey

            Sure. If you get rid of the boundaries and through route trains it doesn’t matter what they’re called.

  12. Nathanael

    It’s worth noting that on some lines there are four tracks on the inner trunk, and two (or fewer) tracks on the outer branches. SEPTA in particular tends to have this structure.

    When you have this structure, it simply lends itself to running one service “local then express” — local on the outer branch, “express” on the trunk; versus another service “local on the trunk”.

    By contrast, a two-track line which splits into two branches lends itself to all-local service on the trunk, with frequency dropping in half as you go onto the branches.

    I guess the main complaint I can make about your analysis is that it’s an analysis of a *line*. If you have a network, other considerations start to predominate.

    As someone else pointed out above, if you have a line of uniform density residential, with a concentration of jobs at one end, then “local in the outer area, then express to downtown” looks extremely logical, to operate the outer section as a “segregated service” from the inner section. I think this accounts for the development pattern of the lines in London and some other older cities, which did have exactly this pattern of residential and job development when the railroads were being developed (partly due to limits on the practical heights of residential buildings at the time, and partly due to job concentration).

    Nowadays the residential and job patterns are far more complicated than that. And mostly, they don’t seem to encourage the “express to downtown” style of trip; limiteds are more appropriate.

    As a historical note, “limited” and “express” were originally *intercity* railroad terms. The “local” stopped at every stop. The “limited” stopped at, well a “limited” collection of stops, which often meant only the “big” stops, every fifth stop or something similar. Though sometimes a “limited” was effectively an express. An “express” was specifically advertising that it was blowing past major stops in order to run as fast as possible between two points….

  13. JohnDMuller

    Metro North appears to operate its schedules in a manner seemingly designed to minimize trip times for as many people as possible. In the rush hour, this takes the form of picking up passengers from only a handful of adjacent stations and then running them into Manhattan nonstop. I assume this is what you call local then express. In the peak of rush hour most trains are essentially “specials” to/from one’s own neighborhood enclave with little cross pollination. They run some all-stops locals also, but these would mostly for those who need the innermost stations. This service pattern makes for a decent time-savings for those in the outermost suburbs, noticable savings for those closer in and vanity strokes for most everyone for having what feels like their own private express.

    Even with all these special trains, the schedule is surprisingly uncomplicated, with approximately zero overtaking. On the Hudson line for example, they could probably run the four track section with only two tracks. In practice, I suppose that the diesels from the uppermost section use the inner tracks in the lower section along with the Amtrak trains and the electric MU’s run on the outer rails

    The Harlem and New Haven divisions have likewise overtake free schedules, although they are more prone to make their outermost commuters change-off from diesel to electric in mid-ride, except for a few rush hour specials. There is more Amtrak traffic on this road, however.

    Even with these fairly straight-forward schedules, it still must be “interesting” to manage the flow of all three lines through the Park Ave. tunnel and the GCT maze during the peak hours.

    • Alon Levy

      On the Babylon Branch, they apparently can’t even do that – maybe because it’s two-tracked, maybe because it has too much traffic density (12 tph peak, with fairly closely spaced stops). So the trains don’t actually make three local stops and then go express. They make three local stops, skip the next few local stops, then make local stops again, and then go express.

      Metro-North is doing some pretty spectacular thing with the weave going into the tunnel at rush hour. Really, that junction has to be grade-separated – there are too many scheduling constraints and too many conflicts otherwise.

      • Zmapper

        Sorta off topic: Which tracks leading into GCT does each MN line use in each direction, by time of day (peak, off peak, weekend)? I saw a Youtube video once of what looked like a northbound Hudson line train using the westmost track while a southbound Hudson line train used the west-center track, and I wonder if that switch of directions happens regularly.

        • JohnDMuller

          The Hudson division of Metro North uses that westernmost track a lot, presumably in the peak direction of rush hour, when I believe that MN uses 3 of their 4 main tracks in the peak direction through upper Manhattan. As the Hudson division is the first to split off going out of the city, the westernmost of the center tracks would be a logical place for the off-peak direction trains of all 3 lines to use so that the Hudson division would not tangle up the others. The Harlem and New Haven divisions have the rest of the Bronx to get out of each other’s way, or at least until Fordham station, where about half ot the trains stop.

          Likewise, the peak NH trains are presumably using the easternmost track at the Manhattan end and there is a partial flyover where the Harlem and NH split to ease that crossing. Still, another flyover in the southern Bronx would make it much simpler.

          Inside GCT, the trains load/unload from platforms that are mostly grouped by division (the Hudson division generally on the west side here too), but I think that the switching network could handle anything that the traffic controllers would want to deal with.

      • Adirondacker12800

        Metro-North is doing some pretty spectacular thing with the weave going into the tunnel at rush hour.

        I thought it was gonna be a piece of cake to add 20, 25 trains an hour from Long Island to the mix and it was all dastardly turf war that was preventing it.

          • Adirondacker12800

            They wouldn’t. But once they go south through it, all of them go to Grand Central Terminal – none of them turn back at 125th.

          • Joey

            So what project are you talking about then? The capacity constraint is between 149th St and 59th St. There’s plenty of capacity both north and south of there.

          • Adirondacker12800

            There are extra tracks under 53rd Street too.

   on the IND station at 53rd and 5th:
            The upper platform is approximately 60 feet below the street, the lower level approximately 80 feet below the street.

            or 53rd and Lexington:
            The station is very deep (approximately 80 feet below street level)

            And the ones under 60th might complicate things a bit. Go ahead, get trains into the lower level from Queens.

          • Joey

            You really think they didn’t take that into consideration in developing that alternative? Evidently you haven’t looked at the EIS much. Please see figure 2-15, pdf page 31, of Chapter 2.

  14. Ryan

    Since we’re talking about ESA anyway – are the new LIRR platforms positioned in such a way as to make a direct rail link between GCT and NYP impossible now? I can’t find a clear answer on this one way or the other, but maybe I haven’t been looking hard enough.

    • Joey

      It should be possible to connect the ESA tail tracks to Penn, but it doesn’t allow any useful through-routing that isn’t already possible. If a NYP-GCT connection is built, it should connect to the lower MNR level, since this would allow through routing between NJT and MNR, even though it would be harder to construct.

      • Ryan

        I’d argue that a connection to the lower MNR level from Penn Station is actually impossible to construct just based on the fact that it would need to rip through the existing lower level of the station to do it. Historical would freak out, lawsuits would start dropping like candy out of a pinata and the prospect of actually going through the lower level of the building leaves an extremely sour taste in my mouth to begin with.

        On that basis, I’m alright with the extra level being built beneath GCT. Failure to actually connect it to the existing tracks (never mind provisioning for MNR trains to go to the “LIRR” level or vice versa) is still unforgivable, but I’m okay with having an extra new level built beneath the existing station, especially if it’s being built with the capacity to one day perhaps run trains directly between NYP and GCT without having to reverse direction.

        Assuming that Penn Station Access actually gets completed, there’s really NO through-running set up, useful or otherwise, that wouldn’t already be possible at Penn.

    • Alon Levy

      It is very hard, though not impossible. Difficulties include,

      1. The multilevel configuration of the ESA cavern, requiring more infrastructure if all tracks are to connect to Penn,
      2. The TBM that was parked to the south of the tunnel to avoid the cost of extracting it, and
      3. The large difference in depth between Penn Station and the ESA cavern, requiring a constant grade of about 3% to connect the two.

      • Ryan

        I suppose the silver lining is that you could just use that TBM to do all the digging and not have to worry about dropping another one in.

        That having been said, is a constant 3+% grade really that much of a problem if the only trains using this connection are all commuter EMUs? High-speed rail doesn’t want to have to stop twice and wouldn’t be using the connection at all, the diesel services that MNR/LIRR/NJT run today could be explicitly disallowed from using the connection (and allowing diesel trains into the tunnels is already a stupid idea anyway), and I’m not sure the potential problems associated with stopping on a 3% grade couldn’t be overcome by competent dispatching just based on the distance in question.

        The psychotic alternate option is to suggest another set of lower-level platforms beneath the existing NYP platforms, which would reduce the grade needed and buy us more running room, but the infrastructure cost there is likely too expensive unless NYP is really, truly and definitely maxed out.

        • Alon Levy

          The TBM was deactivated, precluding future use.

          3% grade isn’t unusable, but it’s pretty high and because it’d be the average, it’d reduce the flexibility of the tunnel to change its grade relative to the average to avoid other infrastructure.

          The lower level is easier, and because it’s not even level with the food court level of the station, it’s unlikely the main station would have to be touched if the TBM launch box is at the Penn Station end (as it should be – there’s free land at Hudson Yards and in Jersey). The food court might have to be shut down for renovations, but it could be underpinned and rebuilt to look like the original. And honestly, although the NIMBYs might care about temporary food court shutdowns, the part of Grand Central that people actually like is the main concourse. The food court is your standard mall food court, and to be honest kind of sucks compared to the food options at Penn Station.

          • Joey

            The lower level tracks already extend under the food court somewhat. Would there even need to be underpinning? Though in truth it would probably be a good idea to rip out the floor and build a new set of platform stairs (or a few).

          • Ryan

            I could care less about the actual food court “offerings” (I agree with you, they aren’t anything to write home about) – I care about the building itself and the integrity of the actual lower concourse. If we can run tracks under it without royally fouling up the floor integrity, that’d be the best option.

            Based on gut feeling (I haven’t seen the actual measurements, nor have I been underneath the floor), the actual clearances between the expected height of the train and the bottom of the lower level seem absurdly tight to me. I’d be far more comfortable with the proposition of lower level if there was an absolute guarantee that there was at least 60 cm (and preferably 100 cm) of clearance between the expected top of any given EMU car and the bottom of the station floor. I’m not sure if it’s just a matter of perspectives and not having the actual numbers, but I can’t imagine there being more than 30 cm of clearance between the top of an M8 and the bottom of the concourse. That’s enough to make me worry.

          • Joey

            There is good reason to believe it’s tight, yes, but probably not any tighter than the 47th St cross-passage or various other points in the Park Ave Tunnels. The track in there is ballasted, so if needed it could be ripped out and replaced with thinner slab track to add vertical clearance.

          • Adirondacker12800

            and you have to get around that pesky crosstown train that runs between Times Square and Flushing.

          • Joey

            Challenging without a doubt, but possible. The ARC AA found it feasible but rejected Alternative G on other grounds.

          • Adirondacker12800

            According to, the 7 is 20′ deeper than the mainline lower level.

            60 feet below street level to the top of the platform to 80 feet below street level to top of the platform. There’s people and trains above the top of the platform at 80 feet which would be under whatever is supporting the rails at 64 feet below street level.

            I’m not in the mood to do the arithmetic to figure out how far away top of a 7 is from the bottom of Metro North train’s wheels. It’s roughly 8 feet. from the platform to the top of the car. So you have 8 feet from the top of the car to the underside of the wheels. I suppose some Reardon Metal would solve the problems.

            People with access to accurate three dimensional drawings of what goes on down there have been saying “deep under Grand Central” for 45 years. And for the past 45 years people have been airly saying “just use the lower level”

          • Alon Levy

            The large-diameter TBMs in Barcelona have less than 12 feet between the top of the lower level car and the bottom of the upper level car. And that’s for trains powered by catenary.

          • Adirondacker12800

            Designed and built that way. How disruptive would it be to to the Flushing line out of service for a year or two while they slice the top of the tube off?

          • Alon Levy

            I don’t know. How disruptive was the construction of the Dual Contracts? How about the IND? The subway has 20 feet between levels at Times Square…

          • Joey

            Looking at Figure 2-11 of the ESA EIR, it appears that there’s plenty of room to sent a TBM above the Flushing line station box. Shame the ARC documents aren’t available anymore.

          • Adirondacker12800

            So after four or five decades, hundreds if not thousands of people who build things for a living, looking at accurate drawings, haven’t come up with this idea? Odd isn’t it? Or they look at it and decide there are different alternatives that can be used.

          • Joey

            The ARC people came up with it didn’t they? These types of ideas rarely get much traction if there’s not political will to push them forward.

          • Ryan

            Well, that’s aggravating. And also rather stupid, in my opinion.
            There’s no possible way to re-commission it? It’s just going to sit there, collecting rust forever?

  15. Zmapper

    The bus parallel would be limited-stop and express services.

    Limited-stop routes which stop roughly 1-3 times a mile, and usually overlay a local route stopping 4-6 times a mile, are generally not worth it. As the limited-stop route is scheduled independently from the local route, this effectively results in guaranteed bus bunching and reduced frequency for customers. Branching divides frequency, which means that instead of a combined route at a 7-10 minute headway at all stops, the major stops can be scheduled for 2 buses every 15 minutes and the minor stops have 1 bus every 15 minutes.

    Limited-stop routes are perceived by customers to be more frequent than local routes, even though the in-vehicle time savings in practice is easily outweighed by the additional waiting time. As an example, consider the difference between RTD (Denver) Routes 15 and 15L, from Colfax/Broadway (large transfer hub, and last timepoint in Downtown) to Colfax/Billings (last stop many buses terminate or diverge to different destinations, a large RTD maintenance facility is also located here). The distance between the two stops is 8.7 miles. It is likely the largest difference in time and slowest speeds are during the PM peak leaving Downtown, so I will compare the next bus after 4:00 pm for each route. The Route 15 trip departing at 4:11 takes 51 minutes, while the Route 15L trip at 4:07 takes 43 minutes, for a savings of 8 minutes, or roughly 1 minute per mile.

    Using the commonly accepted conversion rate of 5 minutes walking time for every 1/4 mile, a customer would have to be travelling at least 5 miles before the limited is quicker than a local, before factoring the reduction in frequency of both routes, and that walking time plus the additional waiting time is perceived as more time consuming than in-vehicle time (Wind, temperature, precipitation, etc don’t affect customers inside a vehicle).

    The other option is an express route, which varies from merely operating along a less congested parallel street closed-door at a city traffic average speed of 20-30 mph to using a free-flowing freeway lane at roughly 55 to 65 mph. Because of the drastic difference in top speed, and that freeway routes can’t be scheduled to stop for customers on the freeway (at least while spending the same amount of money as a typical sign+bench/shelter stop), express routes along freeways save a substantial amount of time.

    As an additional comparison, Route 20, which operates along the parallel and less congested 17th Avenue two blocks north of Colfax, takes 20 minutes to travel between Champa and Colorado, while Route 15 takes 24 minutes. Merely using a less congested parallel street for a little over a quarter of the distance saves half the time a limited-stop route along the main street would take, before even considering options such as closed-door service along the parallel route.

    A customer only cares about two stops, the one they boarded at and the one they will alight at. From a customers perspective, every other stop takes up time with no direct benefit. Every other customer also thinks the same. As mass transit involves… the masses, enough customers to productively operate a vehicle are required in order to justify service above and beyond local routes along a frequent grid. A traveler from Downtown Denver to Aurora doesn’t care about the stops at Josephine, Colorado, Quebec, etc. The hypothetical ideal would be for closed-door routes from the customers origin to their destination, at the time they wish to travel (basically, an individual car). If enough people travel from one destination to another, like from a specific suburb to the CBD at peak hours, then transit can closely replicate an individual car in time with the additional cost saving benefit of the economies of scale found in moving more people per vehicle. In order to have those same economies of scale at other times, the route must behave slightly less like a private car by stopping at those “extra” stops for the other customers.

    • Alon Levy

      Limited-stop buses are useful on very thick routes. Broadway in Vancouver supports an express bus (99) every 10 minutes and a local bus (14 to downtown, 9 straight east) every 10 minutes well into the night; at rush hour, each comes every 3-5 minutes. The key features of Broadway include,

      a) Relative length: the route is 13 km long;
      b) Long trips: a majority of passengers on the 99 ride the entire length of the route, from the subway connection to UBC;
      c) Spiky development: Broadway has a clear set of major cross streets, with much more development at those intersections than elsewhere; and
      d) Thickness: ridership on Broadway is so high that a single bus line couldn’t possibly serve everyone without horrifically bunching, so a split into multiple routes actually improves service.

      I don’t know if there’s any street in Denver that meets those criteria. In the bigger US cities, some corridors do: Wilshire in LA, Geary in SF, 1st/2nd and the other top routes in New York.

      Express buses have a problem with low turnover. The 99 in Vancouver also has low turnover, but it makes up for it in crowding, so overall cost per passenger is low. (In contrast, Vancouver’s other two low-turnover grid routes, the 25 and 49, have higher operating costs than other buses of comparable ridership.) Urban transit operators usually get paid the same no matter if you ride 1 km or 10 km, so buses that shuttle people over short distances have an advantage over longer-range buses, even if the longer-range express buses charge more. The Manhattan crosstown buses have extremely low cost per rider by local standards simply because they’re short rides. Of course this isn’t an issue on commuter rail, where fares are closer to proportional to distance.

      • Zmapper

        The 15/15L along Colfax Ave is the second highest ridership bus route in Denver behind the free 16th St Mall Shuttle (which basically operates as a 1.6 mile long sidewalk with a motor). As the arrow-straight main street for the area with four lanes of traffic and on-street parking, Vancouver’s Broadway and Denver’s Colfax are roughly similar, with the caveat that Denver’s Colfax is more 50’s era single story auto-oriented low rent commercial than Vancouver’s Broadway, which appears to be fronted with more multi-family housing and has a more intact pedestrian realm.

        As to your main points:
        a) – Roughly the same distance between Downtown Denver and the Anschutz Medical Campus, which is on the other side of I-225 from the timepoint I used in my analysis.
        b) – While I don’t have on/off data on hand, the 15/15L likely has more turnover than the Vancouver 99, but I suspect a significant number of customers do primarily ride from locations in Aurora to Downtown Denver, or from Denver to Anschutz Medical Campus.
        c) – Less prominent in Denver, partially due to the relative lack of density.
        d) – A check of Translink’s website shows that Broadway has almost one scheduled bus per minute, with the local route 9 every 4-6 minutes. Below every 8-10 minutes, multiple routes with different stopping patterns can be operated without making the customer wait for longer than a few minutes. At one bus per minute, bunching is inevitable anyway because of traffic signal cycles, so there is almost no advantage to using one stop pattern.

        Both the RTD 15 and the 15L operate roughly 16 buses per hour during peak hours total, or one every 3.25 minutes combined. This drops to 8 buses total on weekends, with each route every 15 minutes. While a reduction in headway from 4 to 2 minute headways is negligible, a reduction 15 to 7-10 is large enough to be noticeable.

Leave a Reply

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

You are commenting using your 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.