Boston Regional Rail

At TransitMatters, we have finally released our regional rail paper, recommending improvements to the MBTA that regular readers of this blog are probably familiar with. Alert readers might even want to probe which parts were written by me and which by others; the main document underwent several edits but some stylistic differences might persist, and the appendices were mostly written individually. We are suggesting the following two-step process:

1. Modernize the system based on best industry practices. This includes full electrification and fleet replacement with electric multiple units (and not electric locomotives), high platforms at all stations, and high frequency all day, every half hour on every branch interlining to support a train every 10-15 minutes on urban trunk lines. In some areas, such as Revere, there should also be infill stops. The capital cost, excluding fleet replacement, should be on the order of $2-3 billion, but the first priority, the Providence Line, is maybe $100 million excluding rolling stock, mostly going to high platforms.

2. Build the North-South Rail Link, with four tracks connecting the South Station and North Station systems. This takes longer than electrification, so planning should start immediately, with the intention of opening somewhat after the entire system is wired. The capital cost should be $4-6 billion, per a study that we’re referencing in our report.

In my mind, regional rail serves three main markets:

1. Local trips on trunk lines, connecting to urban neighborhoods and subway transfer points. The main benefit of regional rail is that it provides an express subway at very high frequency, just as I use the RER to get to Western Paris faster than I would on the Metro. In Boston, areas that would benefit include Forest Hills, Allston and Brighton, Hyde Park, Dorchester and Mattapan along the Fairmount Line, Chelsea, Revere, and Porter Square. Residents of these neighborhoods are likely to travel to other neighborhoods and not just to Downtown Boston.

2. Suburban trips, which are dominated by peak commutes; I complained here that US commuter rail demand is peaky, with 67-69% of suburban trips on the LIRR and Metro-North and 80% on the MBTA occurring in the morning peak compared with around 47% on Transilien, but this is in large part about land use and not just frequency. We’re calling for replacing park-and-rides with town center stations in the report, but absent extensive transit-oriented development, suburban trips are likely to remain peaky and CBD-bound. This is the only market North American commuter rail serves, and its users are territorial about what they view as their trains. However, electrification would speed up these trips materially (the Sharon-South Station trip time would go from 35 to 23 minutes), and the North-South Rail Link would offer North Side suburbs access to the CBD, which is too far from North Station.

3. Intercity trips, which are not peaky except insofar as some people commute. Those tend to dominate off-peak ridership today: per a CTPS study from 2012, about half of the Providence Line’s off-peak ridership originates in Providence itself, which also accords with my observations taking the line on weekends. These trips gain less from high frequency, but need a consistent frequency all day, every day, at worst every 30 minutes, ideally every 15 or 20. Regional rail modernization also speeds these trips the most.

Bear in mind that even though the report just came out, the actual writing was for the most part done in November. This means that the technical aspects of scheduling reflect my thinking in November and not now. At the time, I hadn’t thought about peak-to-base ratios systematically, so my sample schedule for the Providence Line has a train every 15 minutes on each branch (Providence and Stoughton) at the peak and a train every 30 minutes off-peak. I had been assuming a peak-to-base ratio of 2 would be appropriate, by comparison with schedules in Tokyo and on the RER here in Paris. I knew that the ratio was lower in some other places I think highly of, including London and the German-speaking world, but my assumption had been that demand would be so peaky that the maximum acceptable peak-to-base ratio was the correct one.

I’ve argued before that the peak-to-reverse-peak ratio must be 1 or as close to it as practical, in order to avoid parking trains in city center midday. The capacity problems at South Station, which averages a train arrival per platform track per 35 minutes at the peak even though the system is capable of 10-minute turnaround times, come from trains going from the platform tracks to the layover yard during the peak, crossing the station throat at-grade and delaying peak arrivals.

But recently, I started thinking more carefully about operating costs, and wrote this post about peak-to-base ratios. I no longer think peak-to-base frequency ratios higher than 1 are supportable. The marginal labor cost of midday service when there’s a prominent peak is very low, since the railroad would be replacing split shifts with regular shifts, and this encourages running the same frequency during rush hour and midday, if not during the evening and on weekends. And as I explain in the linked post, the cost of rolling stock purchase and maintenance encourages running trains as often as possible. Only energy costs scale linearly with service-km, and those are low: at New England’s current electricity rates, it costs $180 to run a 320-ton 8-car EMU between Providence and Boston each way, and at current fares, inducing 16 extra passengers from the extra frequency is enough to make this pay.

In the report, we talk about American commuter rail operating costs, mostly because that’s what’s available. SEPTA’s are $311/car-hour, whereas those of the LIRR, Metro-North, New Jersey Transit, Metra, and the MBTA are $500-600/car-hour. Per car-km, SEPTA costs about $9 to operate. But a system built around cost minimization, with a peak-to-base ratio of 1 (thus, relatively empty off-peak trains), can get this down to about $2/car-km, or about $180/car-hour.

The reason I think the MBTA could run modern regional rail for $2/car-km, where the RER costs $6/car-km and the Singapore MRT $4-5/car-km, is that the schedule is faster. The costs of rolling stock and labor are based on time rather than distance, and the regional rail system we’re proposing has aggressive schedules, averaging 90 km/h between Boston and Providence. Even energy costs can be contained, since a fast schedule implies relatively few stops. For the same reason it’s easier to make a profit on high-speed rail averaging 200 km/h than on low-speed rail, it’s easier to make a profit on a 90 km/h train at the boundary between regional and intercity scale than on a 40 km/h local train.

In general, I believe that transit planning has to be opportunistic: no city is perfect, so it’s always necessary to find workarounds for some local misfeatures, or ways to turn them into positives. In Boston, the misfeature is very low suburban density, making intense regional service modeled after the RER less useful. The opportunity lies in retooling lines that serve low-density suburbs as intercity lines, connecting Boston with Worcester, Providence, Lowell, Nashua, and Hyannis. With the exception of Worcester, which is on a curvy line, these cities can be connected to Boston at an average speed of 90 km/h or so: the stop spacing is so sparse, and the lines are so straight, that long stretches of 160 km/h are feasible.

But none of this can happen under the present-day operating paradigm. The opportunity I’m describing relies on postwar travel patterns and to some extent even on 21st-century ones (namely, university travel between Providence and Cambridge), which requires reforming frequencies, rolling stock, and infrastructure decisions to incorporate best industry practices that emerged from the 1970s onward. The MBTA can offer a fast, affordable, frequent regional transportation system from as far north as Manchester to as far south as Providence, but for this it needs to implement the regional rail improvements we’re proposing.


    • Alon Levy

      Infill stations or infill development? The former: because on the Providence Line there’s no good place for it – the density drops too fast south of Readville. The latter: because New England suburbs are NIMBYvilles.

  1. Joel N. Weber II

    In the case of Stoughton, running a Princeton Dinky equivalent shuttle with Canton Junction transfers in the evenings and on weekends to improve frequencies to Providence might make sense.

    There’s plenty of demand for more homes in the greater Boston area, so the misfeature of insufficient density around stations is only a problem if NIMBYs turn out to want to preserve that misfeature.

    How many megawatt hours are you assuming in your $180 spent on electricity? Is that roughly one megawatt hour at the residential rate I pay of somewhere around $.20 or $.22 per kilowatt hour including generation, transmission, distribution, etc (in which case somewhere around 5 tons of Tesla batteries would probably be able to cover it, given that the Model S battery pack probably weighs around 1000 pounds and is available in a 100 kwh version)?

    • Alon Levy

      I was assuming around $0.15/kWh, the transportation rate. The rule for estimating electricity consumption is \frac{1}{2}mv^{2}; even at 160 km/h, air resistance is small enough that it doesn’t add that much. Between Boston and Providence there are 13 stops; one interstation (to Back Bay) is 100 km/h, two (Back Bay-Ruggles, Hyde Park-Readville) are 130 km/h, ten are 160 km/h. A 320-ton train doing these stops needs 3.8 GJ to accelerate, which is about 1 MWh. Extra electricity consumption comes from the fact that actual electricity consumption in acceleration is about 20% higher than the kinetic energy formula due to air resistance at high speed.

      • Joel N. Weber II

        Are you assuming that the regenerative braking is used to heat an external resistor array instead of to put power back into the catenary or recharge batteries?

      • Joel N. Weber II

        The Amtrak System Timetable at lists Boston to Providence as 43 miles and Boston to Brunswick as 145 miles; if Boston to Providence has .2 MWh of wind resistance and it scales roughly linearly with distance, Boston to Brunswick would be around .7 MWh of wind resistance.

        The Tesla Semi has a 20 second 0-60 time when loaded to 40 tons (the maximum in many US states with no overweight permit); 20 second 0-60 would be 3 MPH/s which seems to be about what typical EMU passenger trains target. A 320 ton train would therefore need 8 Tesla Semi drivetrains to achieve 3 MPH/s. There’s a rumor that the Tesla Semi may have something like an 800 MWh battery pack, in which case 8 of them would provide 6.4 MWh. These numbers suggest that if one were to install enough Tesla batteries in a train to get 3 MPH/s that may have the side effect of providing plenty of range for commuter rail and shorter Amtrak routes.

      • Joel N. Weber II

        At $.15/kwh it might be worthwhile to look at whether there’s cheap land close enough to the railroad that could be filled with MBTA owned solar panels. For battery trains that might mean rotating different sets through the outer end of the line layover facility for 30-60 minutes per set during the mid day, and perhaps even rotating sets between different lines since some outer layover facilities will work better for solar installations than others. If overhead wires end up being preferred there’d be even more flexibility in locating the solar panels.

  2. adirondacker12800

    We’re calling for replacing park-and-rides with town center stations in the report

    Speed up the trip from the town center station suburbanites who had been fighting traffic and finding parking in the CBD or taking the bus to the CBD will discover the charms of a faster train. And want to park, for free, all day, in the lovely town center. Or three blocks away where it begins to turn into single family residential. You want both. Someplace where the people who want to drive to the station have someplace to park and town center stations. Which to build first, the locals probably have a better idea of what would make the most sense. Probably park-n-rides first so the screaming about all those outta towners parking all over the place are avoided.

    To beat a dead horse, go look at what happened when MidTown Direct opened.

  3. Jeffrey

    I am pushing for the State of Rhode Island to get their own electrified intrastate commuter rail-rapid transit line similar to the Denver Regional Transportation District using their type of Silverliner V railcars. The rail line which will be jointly managed by both the Rhode Island Department of Transportation-RIDOT & Public Transit Authority-RIPTA & operated by the Providence & Worcester Railroad-P&W & will be called the “RHODE ISKAND RAILWAY TRANS-STATE (Woonsocket-Westerly) & TRANS-METRO (Woonsocket-Wickford Junction & Woonsocket-Quonset-Point) Commuter Rail-Rapid Transit.
    I am also pushing for the States of Rhode Island & Massachusetts thru the Rhode Island & Massachusetts Departments of Transportation, Massachusetts Bay Transportation Authority & Rhode Island Public Transit Authority to electrify the Boston, Attleboro, Providence, TF Green Airport & Wickford Junction Commuter Rail Line using AEM7 Electric Locomotives & Double Decker Commuter Railcars.

    • Alon Levy

      AEM7 locomotives are garbage. That 46-minute Boston-Providence schedule isn’t possible even with newer, more powerful locomotives – about an hour is the best that can be done. You need EMUs to run this fast.

      • plaws0

        Not sure they are garbage. *Amtrak* was able to keep them running for 30+ years after all (as opposed to the HHP-8s). That said, they are *still* 30-year old locomotives. Lots of newer models in the dealer showroom. The upside of locomotives is that they ease the cost of transition from Diesel-electric to straight electric since you can keep using your not-so-old coaches. But yeah, EMUs in the long run. Certainly not “Diesels under wire” as the T does now on the Providence line.

        • Alon Levy

          The Sprinters have 6.4 MW and can’t do Boston-Providence in under an hour. The AEM-7s have 7,000 hp, which is 5.2 MW, and are unreliable because of their age, which means slower technical travel times as well as more schedule padding; they may not provide any time benefit over diesels.

          The coaches should be replaced, too. They have narrow car-end doors and obstructed passageways, especially the bilevels, which makes 30-second dwell times hard to support in practice.

      • Jeffrey

        Maybe the Rhode Island & Massachusetts Department of Transportation, Massachusetts Bay Transportation Authority & Rhode Island Public Transit Authority if interested in getting Electric Multi Unit-EMU railcars, maybe they should look into what I suggested with my RHODE ISLAND RAILWAY TRANS-STATE & TRANS-METRO COMMUTER RAIL-RAIPID TRANSIT. Use Denver RTD Hyndie-Rodem Silverliner V railcars.

        • F-Line to Dudley

          Silverliner V’s are an absolute dumpster fire of a make, thanks to Rotem’s utter incompetence as a builder and SEPTA’s all-world project mismanagement of that procurement. The cost of ownership keeps soaring for SEPTA as the fixes-for-the-fixes keep mushrooming in the wake of last year’s fleet outage fiasco. They’re going to have to have major components replaced at a much more rapid pace than any comparable EMU just to live out their 25-year rated lifespan. They won’t get a shot at any midlife overhaul because of that. The Rotem SLV is already for all intensive purposes a dead-end lineage.

          Denver got a deal on its starter fleet by taking the parasitic option of SLV’s as-is. It doesn’t mean they aren’t holding their nose tight too at Rotem’s “handiwork”, but as a means to an end it got them their starter fleet within cost. And being a much better-overall run agency that also had the luxury of advance timing before its service start to anticipate the maintenance needs of these cars…they’re doing a better job babying their Silverliners to higher MTBF than SEPTA is institutionally capable of. Denver was in a good position to make that kind of pivot, and make it work for them. But that doesn’t mean they have any sort of intrinsically rosy outlook on those cars’ future…just that they had better flexibility to adjust.

          They’re not likely to ever do a follow-on order of the same thing when it’s time for the next big fleet expansion. For one, Rotem has so destroyed its domestic reputation with multiple botched commuter rail contracts that it’s extremely unlikely they’re even going to bid the same make again. SEPTA’s going back to the drawing board all over again when it’s time to replace the Silverliner IV’s. And nobody else is letting Rotem get a sniff as a low bidder for their rolling stock because of how thoroughly they’re now being punished on technical scoring in the RFP’s they’re bidding for.

          The next mass-market NEC-compatible single-level EMU that comes out may superficially look like a Silverliner V, if only because SEPTA may stick with the same general motif on the Silverliner VI design. But it will be yet another entirely new domestic make under the hood imported and FRA-adapted from something somewhere else. *Maybe* a moderately adapted AC-only M8 re-skinned in Silverliner clothing if Kawasaki makes a big push…but I’m not sure that’s a good thing to hope for with how morbidly obese the M8 turned out. Unfortunately that means that somebody’s still going to have to be the guinea pig for a first-time U.S. adaptation of some EMU make…and keep their own hands away from their necks while doing so. So as of March 2018 we’re still not exactly staring at anything resembling full-on “off-shelf” with domestic EMU purchase options.

          Maybe we’re moving in that direction if Bombardier comes up with a winner for NJ Transit’s MultiLevel EMU’s currently out for RFP, and NJT getting it right playing the guinea pig on a >100-car order nets something reliable that can be had by others at a good price point. Still way, way too early to tell. And too early to project whether similar orders would be a fit at other agencies. Though with that NJT procurement being active ongoing current events with reams of bidder paperwork coming available in the next couple months to pour through and analyze it will be useful grist for sizing up the domestic EMU market’s immediate prospects.

          • ckrueger99

            What about the aluminum vs steel body question? It’s hard having a mixed fleet between the two, as they can’t be washed/maintained out of the same shop (the steel particles get on the aluminum, causing rust to appear on the aluminum, if I understand correctly). FLIRT is aluminum and Stadler won’t make anything steel.

          • R. W. Rynerson

            The equipment decision in Denver was unorthodox in that the agency had to choose between two huge DBOM consortiums. Hyundai was a part of the winning group, which had other strong features. The losing group included Siemens and obviously RTD has had an overall good relationship with that firm. However, the losing group also included Veolia and the RTD experience up to that point with them — including the experience of many engineers and planners as bus commuters — was disappointing. A broad-based polling of staff opinions on the two packages was undertaken and these facts — plus the knowledge that SEPTA would already be de-bugging the new design — influenced the outcome.

        • Peter Brassard

          With Regional Rail (RR) that could be implemented almost immediately, why should Rhode Island develop its own system (whatever name its given)? A stand alone system, assuming it’s successful, would insure a massive number of transfers at Providence Station for Boston bound passengers. Wouldn’t it make more sense to just extend the Providence line RR (MBTA) all the way to Westerly and insert a lot of infill stations along the line. The RR proposal suggests 30-minute headways for Rhode Island, which is good for suburban, distant, or rural areas, but is not great for the urban core. A separate Rhode Island system would likely offer the same 30-minute headways. Where Rhode Island could have its own train is on the P&W or Blackstone Valley line (BL). The state wouldn’t have to electrify the BL right off the bat. If the BL terminated at the airport and the schedule alternated with the Providence line, a BL could make it possible to have 15-minute headways within the urban core from Pawtucket/CF to TF Green. This would be a boon to the airport and downtown, and the urban core. There’s nothing to say that the operation of the BL couldn’t be part of the MBTA system, especially if service between Providence and Worcester were to happen. Such a BL should be public with many infill stations along the route, unlike the apartheid transit proposed by Boston Surface Railroad.

          • Alon Levy

            I’m of two minds on this. I think that through-service is probably not a good idea, because of train lengths. There is a lot of demand between Boston and Providence, filling 4-5 eight-car tph at the peak, and maintaining okay ridership even off-peak (there are around 200 passengers per midday or reverse-peak train serving Providence). In-state, long trains are unnecessary. A two-car EMU running every 15 minutes between Woonsocket and Westerly – possibly even every 15 minutes between Central Falls and Warwick and every 30 minutes between Woonsocket and Westerly – should be more than enough. It wouldn’t even be worth it electrifying Providence-Woonsocket if Pawtucket-Westerly weren’t already electrified, using the 0.3%-ROI-per-car-per-hour formula.

          • Peter Brassard

            Fair point. The two economic drivers for New England are Boston and New York. This might be pie-in-the-sky-ish and decades away, but what if MBTA RR service ran through to New Haven replacing Shore Line East operations? Or, alternately Shore Line East was extended to Woonsocket or ultimately to Worcester? The suburban overlap between New Haven and Providence is roughly New London, though granted southern Rhode Island and eastern Connecticut is the least densely populated segment of the NEC. The current fragmented nature of commuter rail along the NEC, would be compounded, if Rhode Island had a separate system with Connecticut two systems, the MBTA in Massachusetts to Providence, not to mention everything south of New York. How would New Hampshire or Maine fit, would they have their own commuter rail systems?

          • Alon Levy

            You’re basically calling for a Northeast Regional, except fare-integrated with commuter rail. I guess between New Haven and New London you want this train to be long, because capacity is limited by the movable bridges, and then you can just declare that one Providence Line train per hour runs to New Haven (or that one New Haven Line train per hour runs to Providence). Or you could just say fuck it, Westerly isn’t a big draw, and run this train as a New London-Kingston bridge and tell longer-range intercity rail riders to just use HSR.

            North of Boston the situation is different, because Boston-Portland is a reasonable corridor for hourly trains. Hence my “one Haverhill train every hour goes to Maine” pitch for shoehorning Portland into the system.

          • adirondacker12800

            Having the intercity train stop at every platform that isn’t a flag stop between Haverill and Boston would be annoying. And you are either running empty seats between Haverill and Portland or the people going to Portland will be standing in aisle watching the locals sit in seats. The locals will figure out how to get seats. They could do something clever like only stop it in Woburn and Haverill, give a catchy name.. something like “The Downeaster”…. As if the Haverhill and Woburn lines are ever going to be busy enough to have to shoehorn an hourly train into anything.

          • Alon Levy

            The problem is that the Woburn-North Station trunk line is busy and needs trains to all run at the same speed. Winchester isn’t Westborough, and Lowell generates a decent amount of off-peak demand as well (and will generate even more if the line is sped up).

            You can skip stops north of Woburn, sure. You can even do a reverse Wildcat Branch and have the Downeaster take the route the Haverhill Line takes today, but that requires double-tracking an annoying pinch point alongside the Orange Line (you can schedule regional rail to Reading around it, but forget about mixing intercity trains in there).

          • adirondacker12800

            Fast trains can go slow. There’s four tracks of railroad across the Charles and even more in Cambridge. According to Wikipedia they want to replace the bridges with six tracks of railroad across the river. According to WIkipedia they are putting the second track back in, someday all the way out to Haverill. Fitting an express in once or twice an hour isn’t a problem.

          • Alon Levy

            The approach is four-tracked, the rest of the line isn’t. And they’re double-tracking the upper Haverhill Line, but next to the Orange Line I don’t think they are.

          • adirondacker12800

            So they are gonna spend a gazillion dollars electrifying everything but not spend a few million laying a mile or two of track. Okay.

          • Peter Brassard

            Adirondacker12800 Yeah you’re right I guess I’m a jerk, thanks, but few Rhode Islanders are using commuter rail other than to Boston because basically they have a non system with just two stations to Providence, one of which is in the middle of nowhere with great highway access and both have a crappy schedule just on weekdays. The original funded cost for Pawtucket was $40m. Amtrak slapped another $25m on top of the $40m just for fun. That’s why they cut the two other tracks. The urban core (where the population density is) is where rail service would be effective. But maybe I’ve got that wrong too.

          • adirondacker12800

            Last time I checked Providence is south of Boston. Why would railroad tracks north of Boston significantly affect service in Rhode Island?

  4. Benjamin Turon

    Seems like a lot of good ideas, especially electrification. From where I live in the Capital District, the big talk from Mass is about “East-West Rail”, a new intercity rail service Boston-Springfield, with some trains going on to Hartford and New Haven. I’m working on the idea of having some trains continue onto Albany and New York City to provide a Albany-Boston service and give the Berkshires their desired rail connection to New York City and Boston.

    Out of curiosity, whats the standard turnaround for a Acela on the NEC at Boston, NYC, or Washington?

    • adirondacker12800

      Spend a lot of money to get something that would still be slower than a bus?

  5. plaws0

    A lovely vision and do-able if the Powers That Be are able to ignore the “yeah, but the Big Dig!” whining. My fear though, would be that the B&M guys and the New Haven guys would work to destroy a cross-city system just like the Reading and Pennsy guys did in Philadelphia. That their respective railroads are 50 years gone made no difference.

    • Alon Levy

      I don’t think that the MBTA is territorial about North vs. South Side. The private railroads collapsed in Boston at the same time as in Philly. At the time SEPTA built the connection the private railroads were ten years gone, but now they are more than forty years gone. There are people at TransitMatters who work at the MBTA, and they tend to move between the North and South Sides of the network easily.

      • plaws0

        So how is the split in Philadelphia explained? Seems like, for ~30 years, SEPTA had almost exactly what you propose for Boston … and then they broke it all apart – no more running from end to end, all trains stop and reverse in the middle.

        • Alon Levy

          No, there’s still through-running end-to-end, it’s just not advertised. We go over some reasons in the appendix describing Philadelphia:

          1. The lines aren’t very coherent – there are self-intersections. This is because the PRR and Reading were competitors with overlapping territories and little service due east, in Camden. The New Haven/NYC and B&A neatly separated into South Side and North Side suburban services, so most through-lines would be relatively straight, encouraging through-ridership.

          2. The off-peak frequency and fares on SEPTA RR discourage riders who are not wealthy commuters, and those can just drive to non-CBD destinations.

          3. Not mentioned in our report: there are few jobs near the SEPTA railroad stations except Temple, Market East, Suburban, 30th Street, and University City. This is also a problem in Boston, but less so, since there are a fair number of jobs in Providence, and there’s also a lot of recreational ridership to Salem.

          • plaws0

            Lol! Are you telling me to RTFP??? 😀

            *downloads PDF*

            Seriously, thanks.

          • ethanfinlan1028

            Not just Salem. April-November the whole Newburyport/Rockport Line gets decent weekend day-tripper ridership, and that’s with current crappy schedules.

      • F-Line to Dudley

        Technically there hasn’t been any sort of ops “split” on the T since 1976. That was the year Boston & Maine negotiated via the bankruptcy courts its corporate reorganization that outright sold out the ownership of all northside lines, properties, and equipment to the T. While the T simultaneously gifted them a 10-year contract to be the contracted operator for both the north and south sides (displacing Conrail down south, who were getting out of the commuter contract business). The basic structure of the T’s contracted ops have stayed the same ever since (through Amtrak, MBCR, and Keolis), and the B&M employees who fanned out across the system 40 years ago largely passed intact from contractor to contractor for continuity. Some of them who came over in ’76 are still working there today scattered north and south.

        The only real cultural difference between sides is in dispatching, with Amtrak controlling South Station and the NEC and exerting a powerful influence over everything down south. North there’s no AMTK sticking its nose in everyone’s business, just a minor bit of Pan Am dispatch overlap where they co-mingle with the freight main (Andover-Haverhill + Ayer-Wachusett + immediate Lowell station approach). Crews more often than not stick to one side, but only because it’s too enormous a time commitment to maintain annual qualifications on all 13 lines on both sides. There’s ample fluidity to transfer between sides if a staffer wants to request an assignment that better matches their own work commute, however.

        Functionally, the only ops difference is that the northside still has a blanket ban on cab signals/ATC…a concession to B&M in the ’76 fire sale that’s passed in perpetuity to Pan Am today. Makes their northside PTC installation more radio spectrum -hungry, debugging-intensive, and expensive than what’s going in down south…but once the PTC is fully operational it’ll be a largely transparent distinction in-practice.

  6. Connor Harris

    Assuming the track owners could be persuaded, would it be worth extending Haverhill Line electrification the extra 130 km to Portland for electric Downeaster service? Granted, there’s not much between Haverhill and Portland other than UNH, but an electric train could reliably beat I-95, especially during summer weekend travel.

    • plaws0

      I don’t recall where the Commonwealth’s ownership ends and Springfield Terminal … sorry, Guiford … sorry, PanAm’s starts but recall that the non-T owners fought installation of heavier rail when the Downeaster was first proposed. Unless they are more willing to play ball 20 years later, I can’t imagine them being up for stringing wires. Wires need volume, too, which exist on Providence line – would they exist on a hypothetical electric line to Portland?

      *finds PDF, checks for details*

      • Alon Levy

        The volumes exist on all the regional lines. Boston-Portland and Boston-Springfield are dicier. If you ask me, they should electrify, run EMUs as if they were regional trains, and just declare that 1 tph on the Worcester Line extends to Springfield and 1 tph on the Haverhill Line extends to Portland; but that’s purely to keep up with electric trains on busy track, not because 1 tph is enough volume to electrify an intercity line.

        • po8crg

          Hybrid trains? Run them electric to the end of the wires, fire up a diesel generator to power the electrical traction out to Portland/Springfield and back and then return them to running off the wires. Would be more expensive trains but save on the wiring.

          To keep up with an EMU, they’d need a bit of extra power in wired mode (because they have the extra weight of the diesel generator to haul), but you don’t have to overpower the diesel to match that, just run a bit slower off the wires.

          If Hitachi can fit diesel generator rafts under passenger carriages on the narrow UK loading gauge, then they (and others) can definitely do it for Boston, and a diesel hybrid seems to be the current technical solution for low-frequency extensions or branches running off an electrified core.

          • Alon Levy

            I know there exist dual-mode DEMUs, but I don’t think they’re as high-performance as the single-mode EMUs that circulate in Central Europe.

          • po8crg

            Stadler are building dual-mode FLIRTs for Greater Anglia in the UK (these have the generators in boxed-off areas taking up passenger space, rather than on rafts under the solebar). They’re supposed to have the same performance as EMU FLIRTs when in electric mode.

            They’re also on order for Aosta in Italy.

          • Jon

            The thing to note as a proxy for the tractive effort curve is the starting tractive effort plus the HP rating.

            EMU FLIRT:
            200 kN starting
            3000 kW

            AOSTA Bi-Mode FLIRT
            200 kN starting
            700 kW peak diesel
            2600 kW electric

            TEX Rail DMU FLIRT
            160 kN
            700 kW peak diesel

            So the 0-10 MPH is probably very similar for all the models, but for 0-60 the EMU is going to win the race. If you’re just thinking about matching schedules in the electric territory, though, then the bi-mode is perfectly suitable.

            Note that the Hitachi Class 800 includes diesel spec was noticed as a problem for certain applications, so other regions ordered a new Class 802 with more and up-rated diesels. I am assuming an 802 would beat the 800 and the HSTs…

        • adirondacker12800

          The once an hour to Vermont, the once an hour to Montreal, the once an hour to Toronto, the once an hour to Ohio and the local to Buffalo can stop in Springfield. Extending one of the CDOT trains running between New Haven and Springfield might be a better viewpoint. But then there are enough people in Worcester, Springfield, Hartford, Bridgeport and Stamford that there can be a train that continues on only stopping in New York, Philadelphia, Baltimore and DC. The one that comes through a half hour later can be a “local” to Philadelphia and Wilmington.

          • Benjamin Turon

            From what I read in ‘Modern Railways’ the Class 800 dual-modes DEMUs don’t perform as well as the straight EMUs or even the old Intercity 125 HSTs. My understanding is the issue is acceleration, at first the Class 800 accelerates faster than the HST but then the HST catches up and overtakes the the Class 800. Dual-modes have their uses, but there is a trade-off, the new PRIIA dual-modes planned for Amtrak/MetroNorth/LIRR will be limited to 110-mph because of weight issues.

        • threestationsquare

          This is particularly desirable for Portland (and other potential north side intercity lines) as it would permit closing and redeveloping the surface Boston North Station (without forcing Portland riders to transfer at Malden or Haverhill or something).

    • F-Line to Dudley

      Haverhill is going to be the single toughest line to electrify on the whole system because of freight clearances that are due to be upgraded to double-stack between Andover, MA and South Portland, ME on Pan Am’s Ayer-to-Portland mainline in the next decade. That’s going to increase the vertical clearances from the current 17′ to 20’6″, which has been scoped out as feasible. But to do 25 kV electrification on top of that you need an extra 2’6″, for 22’…a full 5 ft. higher than today. That’s going to be incredibly difficult to swing under the Lawrence, MA street grid and the double-decker I-495 Merrimack River bridge in North Andover, especially when proximity to the Merrimack riverbank flood zone doesn’t allow for much if any trackbed undercutting. There’s other tricky spots in NH and ME too that’ll be pricey fixes, but none sitting on the razor’s edge of “possibly physically impossible” like the Lawrence-N. Andover stretch. Hasn’t been looked at closely enough to peg definitive cost & feasibility…but projects to be thoroughly ‘brutal’ with the pecking order for line electrifications no doubt needing to reflect the scope of that challenge.

      This is quite unlike the other freight clearance route co-mingles on the MBTA system, which don’t throw up any major complications for wire clearances. . .
      — Worcester Line. Double-stack clearances (20’6 + 2’6 = 22′ under wires) from Worcester-Westborough. Plate F (17′ + 2’6″ = 19’6″ under wires) clearances Westborough-Framingham. No height mods whatsoever required east of Westborough. Total of 6 overpasses Westborough-Worcester…only 2-3 perhaps needing any touches at all, and all solvable by minor trackbed undercuts. Inexpensive. (NOTE: Worcester-Springfield is a whole other poopshow with 35 overpasses west of Worcester that would have to be checked/modded to accommodate 22′ underclearance; definitely a little early to be pining for an all-electric Inland Route given scope of that challenge.)

      — Fitchburg Line. Pending double-stack clearances from Willows Jct. in Ayer to Wachusett. 8 overpasses, more touches required than Worcester but no particularly tricky structures. Moderate expense. No backbreaker, but because Fitchburg is the system-longest line and likely requires a second 25 kV substation to span the whole 55-mile distance this probably is an electrification that gets broken into Phase I & II builds–Boston-Littleton then Littleton-Wachusett–one substation at a time. The $$$ for the clearance work chunks together pretty nicely with the second sub for budgeting and construction.

      — Franklin Line. Plate F clearance from Readville to Walpole Jct. Only 6 overpasses…maybe 3 at most that need any touches, all easily fixable with trackbed undercuts. No sweat (at least until the Dedham NIMBY’s start screaming bloody murder about “unsightly” wires).

      — Lowell/New Hampshire Main Line. 2.5 mi. overlap with Pan Am freight main (double-stack to-be) from Lowell station to North Chelmsford along the Merrimack River…5 street overpasses of varying minor difficulty on that stretch. Plate F clearance everywhere else from Somerville to Concord, NH. Many, many small overpasses…none intrinsically tricky, but on sheer quantity of touches this is the priciest of all mainlines to wire up. But also one of the most mission-critical wire-ups of ’em all, so not a price to quibble with.

      At the very least, Haverhill’s likely to be the last northside line to get an attempt at a wire-up because of the expense and difficulty involved. You will need the extra time afforded by punting it to the back of the queue to come up with a solve, and the Downeaster can’t sort out NH & ME until those killer pinches along the Merrimack are secured first. This doesn’t affect the inner half of the line to Reading, which can run under wires on its own high-frequency schedule. Probably just means the outer-half schedule gets sheared off from Reading and re-routed on the Downeaster’s path via the Lowell Line to Wilmington (like it does on select Haverhill expresses today and used to pre-1979 on all schedules).

      And no, you can’t relocate the freights. Pan Am is the only direct connection to Maine & the Canadian Maritimes from the Class I freight mega-hubs in Albany + Worcester County, MA. Maine’s economy is skewed to unusual degree of direct dependence on that one shipping lane, and growing the intermodal carloads to Portland with double-stack capacity is critical for Greater Portland’s economic growth being able to carry the rest of the state’s economically depressed rural counties on its back going forward. In absolute terms this isn’t a particularly challenging corridor for long-term schedule coexistence with freights (especially if Pan Am got a management makeover to reform its infamously sloppy ops practices). This is still adorable “New England-busy” freight volumes we’re talking, not the Alameda Corridor. But the loading capacity is economically critical, or you wouldn’t be able to float a Maine economy that’s worth visiting on a Downeaster train. If Haverhill ends up the only really brutally tough case clearance case to tackle on a 13-line commuter rail system, they lucked out big in the end.

      Deep-future, if you want a substantially faster and more robust Downeaster you’re probably looking at wholesale re-route of it in MA onto the far more tangent Eastern Route through Newburyport and re-extended to Portsmouth. Then figuring out how to reinstate the abandoned track connections from Portsmouth back to the Western Route via either A) Newington-Dover (6 mi. of new track along NH 16 expressway median + chunk of preserved ROW in downtown Dover) or B) Kittery-North Berwick (16 mi. of new track along ME 236) rather than trying to squeeze blood from stone trying to push the Western to >65-80 MPH through Haverhill, Exeter, and Durham. You can easily sustain triple-digit speeds the Eastern between Beverly and Portsmouth, so that’s a higher-leverage prospect to keep in mind if one has visions of the Downeaster buffing out into more of a Northeast Regional-class service come the 2030’s.

      • Connor Harris

        If the freight clearances through Lawrence are an issue, would it be possible to build a 3 km freight bypass hugging I-495 to the southeast from roughly Exit 41 to the Merrimack? It may even be possible within the freeway right-of-way if the exit ramps are realigned.

        • Alon Levy

          The freight clearances aren’t actually an issue – the route isn’t clear for double-stacked freight and Pan Am doesn’t have money to raise clearances itself, it’s just on the state’s wishlist. It’s not like Class I freight, where the companies actually do invest in fixed plant themselves.

  7. ckrueger99

    Thanks for the (mostly) favorable comparisons with SEPTA. Would you consider doing something along these lines with someone from DVRPC for Philly? Maybe with Jim Saksa?

    • bensh3

      The unique circumstances of SEPTA Regional Rail suggest slightly different recommendations. Compared with MBTA, SEPTA has shorter average stop spacing and route length, 50% of track mileage lying within the city, a less fleshed-out subway system, weaker intercity service, less share of CBD commuters, and almost no service to major postwar employment clusters. This heightens the emphasis for SEPTA to reduce track-km costs by cutting labor/maintenance costs rather than relying on fast schedules, in order to run with complete fare parity with transit, 1:1 base/peak ratio, and a greatly expanded number of infill stations. Because of the extreme dichotomy within its network, this also suggests that SEPTA can focus improvements to its inner 5 lines and short-turn segments of Thorndale, Trenton, and Wilmington, repairing lines based on length instead of ridership, creating a frequent service overlay similar to what others have proposed in the past, like PCPC’s Cityrail concept.

      But SEPTA was consistently evaluating frequent service plans from Vuchic and co after the tunnel opened too, most notably with the Metrorail proposal of 1993. Even before questions of institutional inertia, SEPTA was never at a level where they could commit the capital improvements needed for frequent rail.

  8. James Sinclair

    I think you miss step 1, which is improve off peak service.
    Trains are very rare on weekends, fares are high, and on Sundays, parking in Boston is free.

  9. A.G.

    The report mentioned having a Central station within the NSRL, with a same direction cross-platform transfer at Central, and opposite direction cross-platform transfers at both North and South Stations. Do you have a track diagram; would this involve two flying junctions, or transfers at grade?

    • plaws0

      Central Station (between North Station and South Station) goes back to the earliest proposals I remember seeing in the 1980s. Never thought about the platform configuration – always assumed the tunnel would just be four across side-by-side … which doesn’t actually make much sense. Am I remembering right that the proposals in the 80s called for a 4-track tunnel?

  10. johndmuller

    Alon says:
    “, , , the cost of rolling stock purchase and maintenance encourages running trains as often as possible.”

    This notion is part of a chain of logic which turns on the usual pattern of behavior of transit agencies to replace entire batches of trains together. In other words, when the Model A’s get too old, we’ll replace them all with the new Model D’s we are now ordering.

    If the Model A’s consisted of x full time cars and another y peak time only cars, this logic goes, the part time cars were obviously much more expensive per mile run than the full time cars. Nevertheless, they all wore out at the same time. Train car life is independent of miles used – QED.

    I think you bundled maintenance costs in there somewhere/somehow too, with a similar conclusion.

    This could all be literally true, and if so, more power to you; run all the cars as often as possible to get the maximum money’s worth out of them. Certainly some stuff in the cars will deteriorate just with time passing sitting around waiting to be used.

    I do find it to be a little counter-intuitive, however. I think more likely that there is some natural selection going on and that some lesser performing cars with chronic, but not overly critical problems, like rattles, flaky heat or AC, bad intercoms, etc, will find themselves in the Peak Only group, perhaps (making no bones about it) “instead of getting their problems fixed” or maybe just (rationalizing) “until more important jobs are done”. Over time, the number and deficiencies of these sub-par cars in the Peak group will grow until maybe even the off peak cars must include some lemons.

    If my cynical hypothesis is correct, then if we start using them full time the rattles will become worn out linkages, connectors or whatnot, the flaky stuff will become bad and the bad stuff will just stop working at all. The things that were otherwise never going to get fixed will have to get fixed, and some of the cars will just become unusable.

    Somewhere in between is probably where the truth lies. When all is said and done, some cars are probably obsoleted while they still have miles left in them, but others were just hanging in there only because they were hardly getting used at all. So some of the peak cars could be used more at little to no additional cost, but others not so cheaply.

    • Alon Levy

      That’s an interesting hypothesis. I suspect it’s false, but my reasoning is not airtight. If the reason trains get replaced in a batch is that agencies assign the less reliable trains to less frequent runs, then we should see the following two things:

      1. Less reliable peak than off-peak cars even within the same rolling stock class.

      2. Lower mean distance between failures on services with lower peak-to-base ratios, since those lower peak-to-base ratios make it harder to use worse equipment less intensively.

      I can’t confirm or falsify prediction #1. I don’t have data on this, and my anecdotal observations are ambiguous. But #2 seems false.

      • adirondacker12800

        The people who do 30 and 90 day inspections don’t work one day a month. Things rotate in and out.

    • Alon Levy

      1. Does Denver use OPTO?

      2. Denver’s rolling stock is not exactly low-maintenance.

      3. What are the turnaround times like?

      4. The peak-to-base ratio on some lines is 2 and on others is 1. It’s not quite 1:1.

      5. Small systems are likely to have more managerial overhead and fewer trains to split track maintenance across (this is Lisa Schweitzer’s explanation for LA’s highest-in-the-country subway operating costs).

      • Bjorn

        1. It’s 3PTO. One driver, a conductor, and a security guard. The stated reason is because it runs between two high-profile stations; (it’s overkill, I know).
        2. What’s wrong with the rolling stock? Does SEPTA have problems with it?
        3. 11 minutes at DUS, 20 at DEN. In defense of the 20 minute turn at DEN, the additional train provides a “live” extra an hour away from the yard which could function as a backup (to cover an incident, turn times could presumably be shortened to five minutes and the “extra” train pushed into service).
        4. The main A line is 1:1. I highly doubt that the marginal 1/9 trains (7 on A, 2 on B) is why costs are astronomical.
        5. Fair, and agreed. RTD’s DBOM contractor, DTP, is having issues with the grade crossing system and has to staff flaggers there (or at least the PUC is requiring them to do so) until they can get the regulatory boards to sign it off.

        • R. W. Rynerson

          An added reason for the long layovers in Denver is the amount of single track. You are right regarding the use of it for recovery from delays. With so much single track, a sticking door on one train may cause cascading delays on the whole line, even during wide headway periods of the day. (After the track planning was done a station was added to please real estate developers, repeating the 1994 mistake on the single-track Light Rail Welton Street segment by adding runtime. The 2013 retrofit of quiet crossings and a speed restriction at Independence Street on the RTD Light Rail West Line changed the planned runtimes at the last minute, but as there was double-track in the affected segment the slowdown was feasible. (Albeit the changes added another train to the cycle and resulted in ridership less than forecasted.)

        • Alon Levy

          The rolling stock has had a lot of problems on SEPTA. (It’s also FRA-compliant, which might raise maintenance costs; it certainly raises energy costs, since the trains are very heavy.)

          • R. W. Rynerson

            I have previous experience in Edmonton with set-up and corrections on new equipment being charged as an operating cost rather than as a capital expense. That may be what is going on here. It’s probably wise to wait a couple of years before making much use of Denver commuter rail stats.

  11. lake9856

    What do you think is the frequency standard for electrifying a regional rail line? Is 1 tph not enough? By the way, the single-track, hourly (sometimes half-hourly), two-car, single driver, mostly electrified rural/second-tier city regional train lines in Japan fascinate me. You pay the driver upon exiting just as if the train were a bus (at the less busy stations, the busier stations have turnstiles). The trains only run every half-hour or hour. Yet they are well-used, almost always on time, and have timed transfers to connecting busses. The stations all have raised platforms, obviously. Fare integration does not really exist, though, as there are distance-based fares for both busses and trains. This phenomenon seems to be more common in Asia. Can you explain how this system works? As far as I know, there are not many examples of these types of trains running these days outside of Japan. Am I right?

    • Alon Levy

      Switzerland is 100% electrified, legacy of coal shortages in WW2. So there definitely exist mountain railways with 0.5 tph under wire, it’s just not all that common.

      On a purely financial basis, EMUs have half the lifecycle cost of DMUs; if the peak-to-base ratio is 1, this means $1/car-km vs. $2/car-km. 1 tph corresponds to about 11,000 trains per year in both directions, so electrification saves $11,000 per year per car-route-km while costing $3 million per route-km to install. So the financial rate of return is, in theory, about 0.4% times the number of cars per peak tph.

      However, there is extra benefit to electrification coming from higher reliability and better performance. This is especially true on lines with frequent stops or grades, including urban lines and mountain railroads. Urban lines also tend toward longer trains: an 8-car train gives you 3% ROI per peak tph even without taking the speed difference into account – if you do take it into account, you get up to 4-5% depending on how fast you want your schedule to be, and then at 4 tph the ROI is 15-20%. But rural lines with one 4-car tph have 1.5% ROI, and if stops are widely spaced and the route is flat then the performance difference is small.

      These ROI calculations are for fleet-wide conversions. You lose out by operating a mixed fleet. So the first line to be electrified needs higher ROI to justify the mixed fleet, which ROI exists in Boston because the Providence Line is free to electrify and a number of other lines have that 15% ROI. The last line to be electrified can be justified with much lower ROI, for the same reason, and in Boston this is probably Greenbush or Haverhill; both lines also have substantial through-running onto stronger trunks (the Old Colony main north of Quincy and the Lowell Line via the Wildcat Branch), which means that in practice you get more than one km of electrified route per km of installed wire. I imagine that this is what leads Japan to have electrified rural lines – if the trains run through to an urban segment then they should be electrified anyway to avoid running diesel under catenary and also to avoid mixing trains of different speed profiles on the same busy trunk line.

      • lake9856

        That makes more sense. The train line that I am thinking of (the Oitto line) has frequent stops (every 2 km or so), and runs up a mountain. It also connects to an urban electrified train line that runs more frequently, but there is not a lot of through-running. Another factor involving electrification might be the price of fuel, which I know is quite expensive in Japan.

  12. D.Miller

    Is no one going ever going to bring up the fact that the MBTA spent much of the late 20th century moving commuter-rail stations away from historic town centers, and plopping them down in useless expanses far from anywhere? There are entire lines that no longer stop at a single genuine destination.

    As much as I support the goal of the North-South rail link and useful frequencies/fare and transfer integration for better travel within the 128 belt, this notion that (expensively achieved) high-frequency rail in the outer suburbs will magically achieve anything is just as fanciful as the notions that gave the world empty (and cut-back) trains wandering the Dallas and Denver suburbs.

    • D.Miller

      I guess you did mention this problem, which is good, because prior TransitMatters grand-vision statements have ignored it entirely. Still, I don’t know what is gained by grossly overestimating the ridership that will magically appear overnight from unwalkable non-places, once you’ve spent zillions on electrification and overservice.

      • Alon Levy

        Not only did we mention this problem, but also the fantasy map on page 35 (PDF-p. 18) cuts off Kingston to serve Plymouth (in the historic center, not the current station), places New Bedford stations in relatively walkable locations, and adds a lot of infill in walkable places in the Boston built-up area and in Providence. This is on top of stations that already serve walkable places, like Salem, Norwood, Brockton, and Waltham. Then there are Worcester and Lowell, which aren’t really walkable but generate a small but nontrivial amount of intercity ridership, and would generate more if trains came more often than every 1-2 hours.

        • adirondacker12800

          Worcester and Lowell may not be Times Square but they developed before people had cars.

          • Alon Levy

            Sure, but Lowell Station is in a pretty desolate location (I’d know, I used it to travel to LARP conventions), and Worcester Station isn’t in a really good location within the city either.

          • D.Miller

            Paradoxically given my criticisms, Worcester and Lowell may be the two stations on which I’m slightly more bullish than you.

            Worcester’s location is the least of that line’s troubles. All of downtown Worcester suffers from continuity/parking-crater problems, and this isn’t particularly worse in the direction of the station than in any other. The station isn’t objectively far from the downtown epicenter.

            Lowell’s problem is the god-awful pedestrian environment for the first 1500 feet after leaving the station, in any direction but especially heading into the historic district or toward the university. Fix that and the location itself suddenly feels a lot less repellant.

          • adirondacker12800

            They were all gonna make great big gobs of money out competing each other.

            It’s the one that made the most sense when everything contracted. It’s where the train station is going to be.

            Worcester has a great big pile of Union Station. Someone thought it was the achievable compromise some time. Where should they put the condos with retail on the ground floor, places that already have a train station or places that don’t?

        • D.Miller

          You are correct that I shouldn’t have replied based on a cursory skim and prior statements, and that I shouldn’t have done said cursory skim without coffee.

          While I am heartily in favor of TransitMatters’ broader aims here, this proposal is nevertheless marked by the same frustrating touchstones that characterize much of your writing: stridently specific engineering recommendations and demands for improbably high service levels, presented as the only viable options, and backed up with ludicrously optimistic cost and usage presumptions.

          I agree that semi-major urban endpoints (Providence, Worcester, Lowell, New Bedford, or hypothetically Plymouth, the last being probably near-impossible to restore) represent your best hope for achieving all-day “regional” ridership outside of the 128 density drop-off. I would certainly welcome and enjoy enhanced span that allowed car-free trips from the Boston area to other cities later at night than is currently possible. But there is a vast chasm between advocating schedules less erratic than today’s, and deeming frequencies “at worst every 30 minutes, ideally every 15 or 20” remotely warranted by latent demand (much less declaring them non-negotiable).

          Claims that electrification will cut 35% of travel time off of relatively long-distance (but still locally-stopping) trips similarly fail to pass the sniff test.

          • Alon Levy

            First, the claim about travel time reduction is about electrification plus level boarding. Level boarding saves you 15-20 seconds per station. As for electrification, the MBTA has some weak locomotives, whose acceleration penalty to 60 mph is 70 seconds (FLIRT: 13 seconds). With top speed going up to 160 km/h on the straighter lines, the difference is around 2 minutes per station from electrification alone. On top of that, EMUs are less maintenance-intensive than diesel locomotives and break down less often, so it’s possible to run them with less schedule padding; because EMUs accelerate faster, there’s also no need to run express trains, and the simpler stopping pattern improves reliability and allows cutting schedule padding even further. You add all this together and you get a difference of around 2.5 minutes per station, which is where the 30-45% reduction in travel time comes from.

          • adirondacker12800

            Purely anecdotal, watch the videos out the railfan window, of the trip from the end of the line to downtown, level boarding saves more than 20 seconds. More like a minute.

          • D.Miller

            Alon, this is objectively and provably incorrect, requires no extraordinary wonkery or access to “special knowledge” to demonstrate as incorrect, and is highly telling about the false and compounding assumptions that disturbingly thrive in certain corners of the transit-advocacy internet.

            Sharon, Canton, 128, and Hyde Park are each between 3 and 3.5 miles apart, and current timetables have them listed at intervals of 4 to 7 minutes each, depending on time and direction. You simply aren’t cutting 2.5 minutes (i.e. 35%-60%) off of each of these stops just with level boarding and improved acceleration to a hypothetical cruising speed you’ll barely if ever reach. Level boarding and EMUs are objectively superior; they aren’t that much better.

            This is and should be obvious to any well-travelled and observant adult whose life has afforded him/her opportunities for hum-drum, non-railfanning experiences using RERs, S-Bahns, Metro-North trains, and similar. Said adult will ultimately observe that dwell times, acceleration, and aggregate travel speeds are all affected by a multitude of factors and inputs, some structural and others seemingly arbitrary, but none so assuredly fixed as to be able to demonstrate the radical time savings you propose as inherent in and foundational to your desired scheme.

            Contained in your brief reply above is much that I find intellectually dangerous about the state of self-styled-“expert” advocacy, and which, the more I read, I seem to find in especially concentrated form in your specific-project-advocacy writings. You start from your deep and commendable reserves of “general knowledge” (i.e. objective on-the-ground specifics of transit infrastructure and operations, able to be processed as globally-applicable trends and to be compared-and-contrasted from one situation to the next). But you very quickly proceed to invoking claims of “special knowledge”, approaching “special abilities”: Each and every input — capital cost, run-timings, operational minutiae, fleet management and maintenance properties, and of course, “if you build it they will come” demand — are all declared advantageous edge cases far outside the standard deviation of real-world expectations, because reasons. Financial self-sustainment (capital and operational) invariably makes an appearance in the argument, again somehow inextricable from the extremely high service levels that miraculously require lower labor and maintenance costs than in any identifiable real-world example (the most pungent “special knowledge”). Each of these uniquely advantageous inputs and superlatively effective outcomes mutually reinforces every other, thus leading to a constructed narrative of unimpeachability and a resistance to any intrusions of logic or realpolitik that might yield “lesser” results than under your proposition.*

            All of this is, of course, catnip to those who would like to see Metro Boston as the Île-de-France of tomorrow or the Pacific Northwest as a budding Honshu, and will champion any logic that sells them the transportation component of their dream, no matter how in-credible. But it’s hard to even have a reasonable conversation inside the advocacy sphere, when one of the basic precepts (“our platforms and EMUs easily cut 2.5 minutes off of 5”) can be contradicted by a single ride on the Long Island Railroad.

            *(I know that all sounds rude and accusatory, but really, that’s not why I came here. I just find it frustrating and counterproductive that even good ideas — yes, better use of existing commuter rail assets; yes, more even scheduling and better span and improved intercity access; yes, a North-South link, especially versus an extremely costly South Station expansion — seemingly can’t find their way to useful advocacy without being run through the ideology/pie-in-the-sky mill and being divorced from all realistic outcomes.)

          • D.Miller

            Oh, and I forgot to note that Acela currently does South Station to Providence in 35 minutes, including some of the fastest running allowed anywhere on that oft-derided service.

            But we are expected to buy that you can add 6 or 7 mostly-evenly-spaced station stops, and still run that same track in 46 minutes.


          • Alon Levy

            The Acela doesn’t accelerate very fast (it’s not optimized for fast stops), and its higher top speed only really matters for a short stretch between Route 128 and Boston Switch. It also has longer dwells because it’s an intercity train. Add 7 stops at a 75-second stop penalty and 46 minutes doesn’t seem so outlandish.

            As for the interstations you mention, cutting the travel time by 2.5 minutes per station isn’t outlandish. To give you a taste of what the stop penalty is today, most outbound trains do Route 128-Canton Junction and Canton Junction-Sharon in 6 minutes each, or 12 minutes total. Inbound, Canton Junction-Route 128 is 5 minutes and Sharon-Canton Junction is 7-8 minutes, or 12-13 minutes total. But some trains skip Canton Junction, and those do Route 128-Sharon in 8 minutes in either direction. Put another way: the stop penalty of Canton Junction in the timetable is 4 minutes. This is on a train that in practice only goes about 130 km/h, a speed at which a high-performance EMU with level boarding has a stop penalty of about 70 seconds, including pad.

            This observant adult who rides the RER has observed 30-second dwells, at high crowding levels, many times. At rush hour the busiest stations – Gare du Nord, Chatelet-Les Halles, and Saint-Michel-Notre-Dame – have about 60-second dwells. The rest stay at 30. I’ve argued about this with Long Islanders – on Long Island dwell times are about 40-45 seconds, with level boarding and single-level EMUs. Part of it is that the LIRR has extreme timetable padding (relative to current M7 acceleration rates, not FLIRT acceleration rates), and this is usually taken as extra dwell time. Part of it is that even single-level American commuter trains optimize for seating capacity and not fast egress. Part of it might be manual flagging by conductors (on New Jersey Transit the need for the conductor to signal the okay to the engineer wastes a few seconds per stop).

            What the RER doesn’t have is high acceleration. That’s true. French trains aren’t especially high-performance, and there is a lot of schedule padding, especially on lines that mix local and express service on the same tracks, like the RER B. The 7% schedule padding we use is cribbed from Switzerland, where the infrastructure is designed to prevent a delay on one train from cascading to other trains, e.g. in the configuration of flat railroad junctions.

            I also don’t get why you think I’m using especially low labor costs (I and not we, because we didn’t say anything about operating costs in the report). On the contrary, I’m specifically assuming high train operator compensation as in today’s union agreements, and revenue hours per driver numbers that aren’t outlandish by the standard of low-peak-to-base-ratio operations. I’m assuming that there are no conductors, as is the case here and in the German-speaking world and on some Japanese and British trains. And the procurement and maintenance costs I use come from actual contracts that I’ve found in the trade press.

          • D.Miller

            Sigh. I don’t know why, but I really thought I might get a response that amounted to more than increasingly wild handwaving.

            Only one of Acela’s two stops between South Station and Providence could be remotely relevant to discussions of dwell times or comparative acceleration, and the dwell at 128 (high platforms!) is not remotely significant enough to indulge your fantasy of beating the express at its own game.

            I could look up Acela’s acceleration out of 128 or its habitual running speed in J.P. or through Attleboro, but I’m not going to, because your claim is insane. No, you are not going to run commuter EMUs as fast as Acela over a 44-mile distance. Not when 16 seconds out from station stops. Not when a minute out from station stops. Not at the mid-point between station stops. Not anywhere.

            That the second sentence of your reply seems to claim otherwise goes way beyond “special knowledge” or “special powers”, and arrives at “risk the credibility of an entire regional proposal and any of its supporters on a prima facie absurdity”.

            More handwaving in the second paragraph: The timetable does show some southbound trips from 128 to Sharon taking 12 minutes when stopping in Canton. But it also shows trips that take 11 minutes, and some that take only 10. All the trips that bypass Canton take 8 on the timetable. So the actual penalty ranges from 2-4 minutes, and likely has as much to do with schedule padding as with the train’s acceleration capabilities, or (at all but the peak of the peaks) with boarding delays.

            But you double down on your insistence that your idealized EMUs reduce this per-stop penalty to 70 seconds, total. Versus running straight through at (effective) top speed. And you include in this 30 seconds at a total standstill. (That last part seems perfectly reasonable to this regular user of MBTA subways, whose drivers don’t dawdle with the doors. If you’d claimed 10 seconds of total stoppage time, I’d have to further call you out.)

            *However*, your 70-second total penalty argument now rests on the EMU braking (from effective top speed of, you claimed, 80mph) in 20 seconds or less, then accelerating (back to 80mph) in another 20 or less.

            And even once you’ve scraped your passengers off the floor, your 2.5-minute time saving math *still* doesn’t work, because today’s trains already make this same extra stop, on some trips, in less than 3.

            So I continue to call bunk.

            >>>>I no longer think peak-to-base frequency ratios higher than 1 are supportable…. The cost of rolling stock purchase and maintenance encourages running trains as often as possible…. Inducing 16 extra passengers from the extra frequency is enough to make this pay…. All day, every day, at worst every 30 minutes, ideally every 15 or 20.

            Each of these assertions is a fiction. Cumulatively it becomes an insanity.

            I don’t really know what to say. If you honestly believe:
            – that wear-and-tear is not a thing;
            – that 4 t.p.h. all day with skilled labor on 14-hour shifts (or twice as many 8-hour shifts) with scheduled recovery time does not cost significantly more than 1 t.p.h. plus split-shifts to cover the peaks;
            – that high-frequency service will self-justify even at intercity distances (where no trip is ever perfectly spontaneous) and with most of those trips now running 4x emptier than a standard off-peak service; and
            – that this will all pay for itself (!) because your EMUs stop and start so much faster that they bend time and space….

            If you believe all that, then it scares me that you have followers or endorsers at all. Because there is no reasonable advocacy taking place, and precisely nothing will come of Transitmatters’ otherwise laudable aims.

          • Alon Levy

            A 40-second acceleration and deceleration penalty (plus 30 seconds of dwell) doesn’t mean acceleration takes 20 seconds. It means the penalty is 20 seconds. If the acceleration rate is constant, it means acceleration takes 40 seconds. In reality, acceleration is not constant, but is higher at lower speed, where it matters the most. It’s 47 seconds, during which the train travels 1 km; you can watch videos of FLIRT acceleration if you don’t believe me. So with a top speed of 130 km/h, the acceleration penalty is 19 seconds. The deceleration penalty is smaller, about 16-17 seconds. So, a total of 65-66 seconds, or 70 seconds with a 7% pad. At 160 km/h, the penalty is 75 seconds plus pad, so around 80-81 seconds.

            There’s a study for Fairmount Line DMUs from last decade, which lists the acceleration profile of the MBTA’s existing diesel locomotives as well as that of the DMUs they wanted to get, a mixture of powered and unpowered Colorado Railcars. To a top speed of 60 mph, the acceleration penalty alone was 70 seconds for a diesel locomotive-hauled train and 43 seconds for a DMU; at that speed, the acceleration penalty of EMUs is 14 seconds (and you can watch both FLIRT and Silverliner V videos to confirm). This is how we see most (not all) trains exhibit a 4-minute stop penalty at Canton Junction.

            As for off-peak operating costs, going in order:

            – The shelf life of a train doesn’t seem to depend much on how much it is used. London and New York replace trains on approximately the same schedules. This seems to be taken for granted in studies of high-speed rail operating costs in Europe: the cost of procuring and maintaining the train is taken as a constant, and the annual distance driven is a variable. Wear and tear depends on time and not just on distance, and evidently train maintenance contracts here do not specify anything about frequency of usage.

            – Split shifts cost a lot more than you think. The marginal cost of labor on a midday train isn’t zero, but it’s much lower than the average cost of labor. The MBTA has to supply pool tables for train crews to entertain themselves in between the morning and afternoon shift. And in New York, the express buses, which mostly run peak-only and use a separate pool of equipment and drivers from the local buses, have very high labor costs.

            – Providence-Boston trips can be pretty spontaneous. I’ve even been occasionally discouraged from taking New York-New Haven trips over hourly frequency. You don’t need to induce that many passengers to break even on marginal energy and labor costs going from a train every 2 hours to a train every half hour.

            – EMUs don’t bend time and space. They just accelerate at 1.2 m/s^2, and the higher-powered ones used for low-ish-density German and Swiss suburbs have power-to-weight ratios in the 20 kW/t area (FLIRT, DBAG Class 423, Coradia Continental, Talent 2).

            And this isn’t some cranky opinion that only I share. The idea that the peak-to-base ratio should be low is increasingly common wisdom at TfL. The timetable for Crossrail was recently revised to 24 tph peak, 20 off-peak through the central tunnel. And before you bring up land use, compare Cockfosters (1:1 peak-to-base ratio on the Piccadilly line) with Boissy-Saint-Leger and Saint-Germain-en-Laye here (Boissy gets 2:1 on the RER, Saint-Germain itself is 1:1 but the next station on the branch is already 2:1), or with every outer end of the New York City Subway (on average they’re about 2:1), or with the Chuo Line in Tokyo (also 2:1).

          • adirondacker12800

            e MBTA has to supply pool tables

            Pool tables are cheap and last forever. Fancy smancy ones are $2,000. I doubt the ones the MBTA bought are fancy ones. Could just make them stare blankly at the wall but that would make them grumpy.

            discouraged from taking New York-New Haven trips over hourly frequency.

            A Regional or an Acela or both wander through most hours of the day. Twice an hour sometimes thrice an hour frequency.

          • D.Miller

            My basket of apples is suddenly overloaded with oranges.

            It would not surprise me if the New York Subway, London’s heavy lifters, or the very-highest-demand routes across the Île-de-France boasted enough all-day demand and network usefulness to justify very high service levels that could be boasted of in your preferred ratios. That is not, however, remotely relevant to services like a hypothetical MBTA regional rail system, which you yourself have peak frequencies driven by suburban commuting, with an entirely different market (less frequency-dependent and longer-haul intercity trips) responsible for whatever medium-level all-day demand you’ll ever be able to unmask.

            (Your claim still isn’t even true, of course. New York is full of subway segments will much lower off-peak:peak ratios than 1:2. The famously irritating Far Rockaway branch of the A train is like 1:3.5, and frequency in the peak-est hour on that branch isn’t exactly impressive to begin with.)

            Now from the top:

            – More hand-waving, time-and-space bending: Now, apparently, only the first 19 seconds of the acceleration matters to trip time. A train still speeding up from half-speed is exactly the same as a non-stopping train barreling through at full. Anything to massage claims of dramatic time savings to maximum effect.

            – No one extolled the virtues of MBTA’s aging diesel fleet. Still doesn’t explain why you think you can save 2.5 minutes on every stop, when those very locomotives cause have penalties of only 2-3 minutes on many runs today. Hand-waving away that you’re now calculating negative time. The only explanation for “4 minute penalties” is schedule padding. In reality, the stop is not presently slow enough for you to save nearly the time you claim.

            – Irrelevant pretty much in total. Replacement schedules in New York have even less to do with the actual condition of the vehicles than with the total number years in service or the number of miles driven. Between London and New York, though, I bet I can guess who spends more on preventative track work and on consistent vehicle upkeep and system refreshment, and I bet that number is pretty darned non-negligible.
            The only type of vehicle for which it can be credibly claimed that mileage barely matters to maintenance cost is an airplane, because there are relatively few moving parts and normal operation involves relatively little friction. You are the first person I’ve ever seen claim mileage doesn’t matter to trains, that maintenance contracts are divorced from any expectation of how and when the vehicles will be run, or that “1:2” ratios in the absence of the demand to remotely warrant such ratios come with zero maintenance burden. “Special knowledge” in extremis.

            – An entirely separate peak fleet and labor pool is relevant here how, exactly? You still haven’t addressed the universal 14-hour shifts or double/tripled/quadrupled/quintupled pool of 8-hour-shift employees you would need to keep up absurdly and unwarrantedly high frequencies all day. Because that would defeat your self-sufficiency narrative pillar.

            – At no point have I advocated maintaining multi-hour gaps in off-peak service. The fact remains that if you’re headed from Providence to Boston, on even a semi-spontaneous basis, you’re going to be gone for much of the day or evening, so you are willing and able to plan around a schedule, even if that schedule is less than walk-up-and-go. What matters is that the schedule is predictable with a span such that you can reliably get back. It doesn’t even have to be the fastest trip imaginable, as long as it is faster and easier than driving most of the time.
            In fact, Lowell and Providence already come close to achieving this, thanks both to monstrous traffic much of the day and rails in much, much straighter lines than any highway you can drive. Any speed improvements at all are icing on the cake, but scheduling and span improvement probably matter more. (By contrast, Worcester is so painfully slow that there is little you can do to make it more appealing off-peak, unless you treat speed improvements as paramount.)
            Our friend Adirondacker helps to make this point, btw, by noting that you complained about Metro-North frequency deterring you, without seriously considering a much, much faster trip on a possibly more frequent Amtrak train. It turns out that you too prioritized factors other than trip time or immediacy of departure when deciding how to travel.

            – Power-supply jargon meant to impress, not impressing. If you were making it up I wouldn’t know or care. I never brought up electricity costs. This is general knowledge functioning as if it were special knowledge, because you are trying to use it to again “prove” your reality-distorting claims of impossible time savings, of trains that cost ~nothing to run in massive overservice (in part because they’re so darned sprightly).

          • Alon Levy

            Please try to understand acceleration numbers better instead of saying “nah that can’t be true.” A FLIRT takes about 46-47 seconds to accelerate from a standstill to 130 km/h, during which period it travels 1 km. Were it traveling the same kilometer at 130 km/h, it would take 27-28 seconds. Thus, the penalty is 19 seconds. The MBTA diesel fleet has a penalty (not just acceleration time) of 70 seconds to 60 mph.

            When you tell me I’m the first person who makes these claims about peak-to-base ratios, you tell me you don’t read London Reconnections, a blog that generally views itself as explaining TfL’s reasoning rather than criticizing it.

            I bring up the Underground vs. the subway in New York, the RER, and the Tokyo rail network to argue that high peak-to-base ratios are cultural. New York has certain internal standards for subway crowding, and based on these it runs most individual lettered routes every 10 minutes off-peak. If you propose more frequency, people push back by citing the standards. The same reasoning is used on commuter rail all over North America, regardless of development typology; the insiders at GO Transit dislike the Toronto RER project with its 4 tph off-peak frequency to dense suburbs and are also opposed to infill stations. Paris and Tokyo have cultures of high-for-urban-rail peak-to-base ratios, like New York; Switzerland has a culture that aims at a peak-to-base ratio of 1 (but peak trains may be longer), and Munich and Berlin have low peak-to-base ratios on the S-Bahn.

            London and New York spend similar amounts of money on rolling stock maintenance per car-km (link, PDF-p. 10 – London is clearly labeled, New York is the highest-cost system labeled Am). Both are on the low side by first-world standards; it’s consistent with London being more efficient through running trains more intensively and New York deferring maintenance, but it’s not the only plausible explanation.

            The separate fleet and labor pool is relevant because that’s to a good approximation commuter rail service today in Boston. It’s useful to look at a mostly peak-only service to see whether it saves money, and it doesn’t. The pool tables are of course cheap, but paying workers to play pool rather than drive trains is not so cheap (and if you think you can retain drivers in the long term with split shifts and no in-between pay, you’re delusional).

            I didn’t ride Amtrak because I missed a train at 125th Street, which Amtrak doesn’t serve. Amtrak is also more expensive, not much faster, and as far as I remember not really timed to produce even intervals with Metro-North between New Haven and New York (they don’t care). I have ridden Amtrak between Boston and Providence when the MBTA was at inconvenient times.

            Power-to-weight ratios aren’t jargon, and if you’re uncomfortable talking about them, then don’t argue with me about acceleration rates. If you wouldn’t know if I were making it up, then you’re literally telling me you don’t really understand the subject of this conversation, which is buttressed by the fact that you don’t get how 47 seconds of acceleration time could produce 19 seconds of acceleration penalty. You’re just seeing a conclusion that runs against a culture that you’re used to and assuming that it must be wrong.

          • D.Miller

            Before concluding that you’re hopelessly obstinate and that your advocacy is divorced from useful reality, I’ll try one alternate approach to the “2.5-minute savings” problem.

            As far as I can tell, every inbound train that stops at both Canton Junction and 128 lists a 5-minute interval between the first departure and the second. With no exceptions.

            So therefore, under your theory of “2.5-minute savings” per stop, you *must* imagine that once each train shuts its doors in Canton, this happens:
            – the train accelerates
            – it travels precisely 3.4 miles
            – it decelerates
            – it opens its doors at 128
            – it boards passengers
            – its shuts its doors again…

            Under your theory, all of that now happens in 2.5 minutes, down from the current 5.

            A full 50% reduction.

            All because your preferred brand of EMUs is so darned awesome.

            This is simply not a defensible theory.

          • Alon Levy

            The unpadded FLIRT travel time between Route 128 and Canton Junction is 3:15; the padded time is 3:29. The current schedules average 5-6, with pad and rounding artifacts. So, about 2 minutes of difference. Then Canton Junction-Sharon is 3:10 unpadded, 3:23 padded; the current schedules average 6-7.

          • D.Miller

            More trafficking in ambitious-to-the-point-of-nonsense hypotheticals as if they were actual facts. No “FLIRTS” have ever run here.

            You’re suggesting an *average* speed of >60mph, over your 3:15 time span, *including* all acceleration, deceleration, stopping, and boarding time. Without precedent or proof.

            And wording it as if it were an on-the-ground fact today.

            And you can’t even admit aloud that every single northbound trip (from either branch) currently reads as 5 minutes. None are 6.

            Your advocacy is clearly non-serious. Have a good evening.

          • Alon Levy

            FLIRTs have run elsewhere in the world, and there are videos of their acceleration capabilities. The laws of physics are the same in the US as in Hungary and Switzerland.

          • D.Miller

            Also, most of the above is what is gently referred to as tautological reasoning, with a generous helping of confirmation bias and an increasing employment of special knowledge, less for anyone’s edification than to deflect against what you see as an intrusion into the airtightness of your latest ineffable proposal masterpiece. This is not the first time I’ve seen you defend a proposal thusly.

            I do understand that a train accelerating and train already at running speed each cover the same distance in two differing and measurable amounts of time. I disbelieve your claims as to what those relative times are. I have also yet to see any proof that such station-interval patterns and sustained speeds are being achieved, anywhere, over station spacings of 3 miles and change. The burden of proof here is not on me.

          • Alon Levy

            FLIRT acceleration:

            Diesel loco acceleration: 0.22 m/s^2 initial (link, PDF-p. 27), which imposes a lower bound of 60 seconds of acceleration penalty to 60 mph.

            And it’s nice that you say you understand the difference, but in the last comment you did not and you were puzzled as to how a 47-second acceleration time would produce a 19-second acceleration penalty.

          • adirondacker12800

            You said you didn’t take trips not that you yet again didn’t yourself to the station on time and blame that on the railroad.

          • Alon Levy

            No, I blame the 125th Street buses for being slower than walking for my missing the train; I blame Metro-North for the fact that the next train was an hour later and not a more reasonable interval later, like, say, 20 minutes.

          • adirondacker12800

            You knew the buses were slow yet assumed they would be faster than walking. Or chose not to hail a cab. Or use your smart phone to summon an Uber. Or use a Fordham Road bus. Okay.

          • Alon Levy

            Why would a Fordham Road bus connect me to the New Haven Line? The trains to New Haven run nonstop from Harlem to Stamford.

            And yeah, I should have remembered that the buses are slower than walking on 125th. My bad for forgetting how bad the MTA was. But that still doesn’t justify hourly frequencies.

          • D.Miller

            I don’t misunderstand; I challenge the airtightness of your math. There is no odometer in the Youtube video to show when 1km has been traveled. And I guess you’re proposing to send extremely short trains around constantly, just to take advantage of the most edge-case math (because apparently that matters a great deal with the Stadler design). That might work when you intend to flood the suburbs with 4 underused t.p.h. in the off-peak, but it wouldn’t come close to cutting it at rush hour, or in any rapid-transit-like service within 128.

            A final coda: You would be correct to observe that I am annoyed, and sucked into replying on the basis of that annoyance. This is not personal, but it is exacerbated by the way this discourse has revealed your indulgence in one of my primary peeves: Choosing a technology and then reverse-engineering an entire, massive proposal to best suit that technology. (This is a form of emotion-driven bias masquerading as objectivity and principle, and bugs me for its lack of self-awareness and predilection toward intellectual dishonesty and adverse consequences.)

            It is increasingly clear you have done this to push the Stadler (or similar) as the be-all-and-end-all solution to Greater Boston’s problems. It isn’t enough to neutrally assess demand at the 128 quasi-rapid-transit level, at the decentralized 495-belt level, and at the intercity level, or to enumerate the (objective, valid) advantages of EMUs over other fleet choices. For whatever reason, you chose to tweak the comparative run-time math way beyond reason, propose improbably high levels of service criss-crossing the entire system, pluck irrelevant peak/off-peak ratio comparisons from some of the busiest urban places in the world, minimize demand-suppressing flaws in the existing network, twist into knots to declare maintenance and high full-time staffing levels effectively free under your plan. And insist that ~350 miles of electrification are a dealbreaker for any improvements at all.

            This last point is particularly annoying to me. When you dump on the Fairmount DMU plan, you are saying that you’d prefer to see an underserved community, in a place where urban-level service could literally change lives, suffer *for decades*, rather than see anything implemented in its midst that falls short of your tautologically-untouchable plan.

            That pisses me off to no end. And that false sense of certitude spreads across this plan.

            Why can’t you make a case for your preferred method and outcomes, without infusing it with layers of bullshit math and abjectly false financials?

          • Alon Levy

            Take the speed at regular intervals. Multiply the average speed by the length of each interval. You get that the FLIRT reaches 160 km/h in 2 km and 130 km/h in 1 km. In general, it would be useful if you tried engaging in the supporting references I give more and went on tirades about how it is clear it’s all bullshit math.

            I also don’t know why you think trains have to be short – coupling multiple high-performance trains together is routine in Germany and Switzerland.

            When I dump on Fairmount DMUs, I dump on bad transit. I dump on a train with twice the lifecycle cost of an EMU (same link also says that train maintenance costs include a large fixed element). I dump on the idea that spending $40 million on electrification is an extravagance but spending $120 million on procuring four 8-car DMUs is prudent. And I dump on the idea that diesel trains belong in a neighborhood with high asthma rates.

            Americans are allergic to electrification because they think it’s more expensive than it really is, which is part of that culture of “we don’t do things this way in this country.” So far about 100% of the criticism I’ve seen of what I’ve written re Boston regional rail, in this thread and on Streetsblog, is utter BS. It’s anons trying to pull rank and saying “no it isn’t” whenever I link to cost or performance data, because, like you, they think math is magic.

          • adirondacker12800

            If your goal is New Haven avoid the trains destined for New Cannaan, Danbury or Waterbury. The ones destined for Boston, Springfield, New London or Old Saybrook would work in addition to the ones terminating at New Haven.

          • Alon Levy

            From Washington Heights and Inwood it’s a lot faster to get to 125th to get to an express train than to get to Fordham to get to a local train and change at Stamford.

          • adirondacker12800

            But the bus on 125th is too slow.Slower than walking,. You were too stupid to get out and walk. Or leave earlier. Or take a cab. Or use Amtrak instead. Or go to Fordham. Other people manage this or they would stop running the train from lack of interest. It’s very likely you have something in your pocket that would have been a supercomputer in 1980 that has high bandwidth connections to Internet. People managed to do this when it was paper schedules and wind up watches. Unless they were rich and had one of those self-winders. . . The ADA isn’t required to accommodate stupid.

          • Alon Levy

            Yeah, what can I say, Vancouver spoiled me and I forgot that crosstown buses in Manhattan are slower than walking. That doesn’t mean hourly regional rail frequency doesn’t suppress demand.

          • adirondacker12800

            That you have to pass out of the NYC MSA, go through the Bridgeport MSA to finally arrive in the New Haven MSA probably has more to do with it. People live full and rewarding lives without leaving metro New Haven often. It’s kinda far away. Heading south of of New York you’d be in Philadelphia. Suburban parts of Philadelphia but Philadelphia.

          • D.Miller

            Better than sucking in diesel bus fumes in traffic on Blue Hill Ave for decades while we wait around for everyone to drink the Alon Levy kool-aid, methinks.

            I’m sorry you are unable to acknowledge the role of your endless edge-case cost and service extremes in the b.s. cycle. This infects every one of your big-reveal proposals, which is why they never go anywhere, and which then undermines any useful expertise or criticism you may have to offer on real-world matters.

            But of course it is *everyone else* who is wrong.

          • Alon Levy

            Why do you think electrification takes decades? Wiring New Haven-Boston for the Acela took 4 years as I recall.

            The fact that Bostonians all think electrification costs more than DMUs is not my problem. Generally, Americans have weird ideas about how much things should cost, and as a result the ideas they pitch end up saving money in the wrong places, which is how the cost problem got so bad. Forgive me for not paying too much attention to the “we’ve always done things this way” crowd in a place where subway tunnels cost a billion dollars per kilometer. And when I point out to you how wrong you are, your response is always “no it isn’t.” Never any link to countervailing evidence. If you’re familiar with evolution/creationism debates on, it’s called the argument from incredulity: “I can’t personally believe the eye could have evolved through natural selection, therefore it didn’t.”

          • D.Miller

            How about starting here: The burden is on you to prove that your insanely high proposed levels of service will magically accrue the ridership to pay for both the infrastructural investment and the operations.

            All of your proposals feature this. You claimed the Pacific Northwest was like Italy, demand-wise.

            Your response to any and all issues taken with your approach is “I am expert, and I have a fan-file of mostly-irrelevant links”.

            I guess operating in fantasia is better for the Patreon crowd. If you ever got anything done, and your legions of riders failed to show, people might stop Patreonizing you.

          • Alon Levy

            I claimed the Pacific Northwest’s HSR study made a unit conversion error in its operating costs. Most of what I say is backed up by numbers, which by your own admission you’re not really trying to evaluate (hence stupid questions like how a video of acceleration times says anything about acceleration distance). The things that I say that aren’t I try to label as theory, like my latest post. People do check my numbers, just not you. What do you think Anonymouse and Threestationsquare and Sandy Johnston and Sky Rose and Clem Tillier and Richard Mlynarik do?

          • D.Miller

            Anyway, I’m done.

            I saw your (characteristic) problem of hyper-optimistic inputs in all facets of your plan. I tried to engage with them. I got back increasingly offensive resistance in the form of “pulling rank”, and accusing all others of cognitive biases while entirely denying your own.

            I just hope TransitMatters hasn’t decided to wager all its hopes, dreams, and efforts on the Alon Levy snake oil.

          • D.Miller

            Ah, yes. If memory serves, you found a single potential conversion error, then flimsily seized upon to allege a massive and systematic overestimation of costs.

            And what was your “proof” again? Oh, yeah, that the study didn’t predict PNW HSR making a massive profit! (See, your tautological self-reinforcements invariably rely on a starting presumption of unprecedented, profit-yielding ridership. Somehow this is where you see “evidence” as trivial.)

            You also suggested that the PNW corridor was analogous to Milan-Rome, i.e. large and dense urban areas sitting on the travel spine of a 60-million-person nation and attached to the rest of continental Europe. Which suggests that, on the most basic of transportation fundamental — how population and land-use geometries immutably impact demand; how network effects do or don’t function — you simply do not know what the fuck you’re talking about.

            I am not a psychologist. So I don’t know off the top of my head the name for the phenomenon in which a clearly very smart person decides to selectively privilege any and all superficially fact-resembling assertions that prop up his pet obsessions, while treating every inconvenient reality as the mathematical failing of his intellectual inferiors.

            But the more I read, the more I realize you are a textbook case of this.

          • Alon Levy

            Didn’t you say you were done?

            My proof was that the study said the trains would be 30% full. However, the cost per seat-km was 12% as high as the cost per passenger-km. This came on the heels of the mistake in table 5-10, which WSDOT acknowledged (while maintaining the mistake was restricted to that table, and then promising to send me a followup explaining the table 5-1 vs. table 5-10 discrepancy and never following through).

          • D.Miller

            What you fail to mention is that daily demand was given a numerical value, and that value was in the mid-4-digit range. Trains were only “30% full” because the study, unlike you, knew there would be no reason to run dozens of daily trains in each direction.

            The lack of adequate demand to justify the massive expenditure of from-scratch HSR wasn’t a conversion error or a rounding error. You would need demand to be orders of magnitude higher (like, say, in denser Italy) for HSR to remotely qualify as a need to be met at any real-world cost.

            For someone whose header claims to be “against boondoggles”, you make an awful lot of excuses and exceptions for when you’re the rhetorical boondoggler.

          • Alon Levy

            I at no point criticized the study’s ridership model except on second-order issues, i.e. intra-metropolitan ridership and the impact of station placement, which the study itself acknowledges the model does not look at. I criticized the study’s claims about operating and capital costs. I did not say “30% occupancy is unrealistically low,” but rather “the study says 30% but appears to calculate operating costs per passenger-km as if the occupancy is 12%, which suggests a unit conversion error.” At 30% occupancy, HSR is operationally profitable based on the study’s assumptions on ticket fares and per-seat-km operating costs in the corrected table 5-10.

    • plaws0

      >> the fact that the MBTA spent much of the late 20th century moving commuter-rail stations away from historic town centers, and plopping them down in useless expanses far from anywhere

      I’ve not lived in the Boston area in many years (or decades) but I don’t remember them doing this. I rode the Franklin Line (even before it was extended!) and recall an in-fill station near Rt 128 to serve a newish office park (then) but I don’t remember the T moving stations out of town? In fact, except for the Kingston/Plymouth thing on the re-activated Greenbush lines (that happened years after I left but was being planned then), I think all those stations were where the New Haven (or Old Colony) put them initially.

      What are some examples where they moved stations out of town?

      • D.Miller

        The rebuilds and reactivations are definitely the worst offenders: Newburyport is bad, and all three Old Colony lines are awful. Aside from the Brockton stations and (arguably, barely) Bridgewater State and the twin flanks of Weymouth, I don’t think there’s a single genuine destination you can reach from any of the Old Colony branches without getting picked up in a car.

        Every other line, I think, has one or multiple examples of relocations in recent decades that demonstrate explicit exurban-commute bias. Ashland and Westborough are the particularly egregious examples on the Worcester line that first come to mind.

        But perhaps to your implied point, the existence or absence of tiny-village-crossing stations in Littleton or Acton may be neither here nor there. Waystations with that degree of remoteness created near-zero demand even in the pre-automobile era; the happenstance of 19th-century track location is entirely irrelevant to the behavior of anyone coming or going from those corners of the metropolitan area today.

        I tend to think that there are benefits to the way Greater Boston developed, unevenly, around 400-year-old towns and villages, although that ironically makes us look “sprawlier” on aggregate-CSA lists than the endless cul-de-sac nightmares around DC or the pervasive automotively-scaled rectilinear sprawl in certain west-coast cities. (This is the risk of “ranking” densities over CSAs comprising thousands of square miles, with no regard to actual form or to dissimilar concentric distributions of population.) But even if our sprawl outside the 128 ring is irregular, it is still a form of sprawl, and therefore often devoid of reasonably transit-serviceable destination clusters.

        The MBTA has made this worse with recent station placement and re-placement. But even reversing this trend won’t suddenly make high-frequency Regional Rail the instant demand-slam-dunk that some would like to believe it would.

        • Antonio

          Some of the placements definitely are not great; however, I aside from the Old Colony lines which really are terrible on station placement most of the other lines as I am looking at them either have stations in town centers or purpose built park and rides near major highways. The main wtf stations that I noticed were Newburyport which doesn’t appear to be too hard to adjust and then the Fitchburg line which has a bunch of weird stations surrounded by essentially nothing which seems to be related to the fact that between Waltham and Leominster there just wasn’t a lot of pre-WWII development. There are plenty of stations for example North Billerica which doesn’t have much nearby, but is really in the best location possible all things considered. The vast majority of stations though are clearly in or as close to walkable somewhat dense development as possible and I think going forward the T should be able to incrementally improve station placement.

          • D.Miller

            Except that every time they have touched a station they have done the precise opposite. Westborough is a town; the T’s relocated station (built just 15 years ago) is not only in an abject nowhere, but is harder to reach even by car, from any direction.

            I live in the city so am not intimately familiar with every commuter rail line or station, but I recall (from past investigation) every line having examples of stations that have closed in real places and reopened in non-places. That you were able to observe a reasonable proportion (and on some lines, preponderance) of more reasonable/historic station siting is a testimony to the sometimes-beneficent impact of inertia: These are the stations the T has never tried to mess with.

            Even the station in superlatively central-urban Chelsea is being moved, as we speak, to a lousier location, for the stated purpose of shortening the time a complex street crossing has its gates down. Relocating the platforms a bit further from 6th Street and building infrastructure to make (grade separated) Washington Ave the primary access point never seems to have been seriously considered.

            But apparently we are to believe that the T will entirely reverse this trend of gradual exurbification, then in the medium-term future embark on a massive program of station site improvements, including a non-negligible number of rail-bank reactivations and ROW takings where necessary, on top of tens of billions in electrification costs and fleet replacements and finally building the downtown through-route that has been talked about for a century, so as to *marginally* improve the bi-directional utility of a new network that still has to contend with the usability demerits of late-20th-century development patterns, but which Alon somehow believes can “make a profit” running trains all over the place at irrationally high frequencies.

          • Ethan Finlan

            IIRC, Newburyport was supposed to be in its original, close-in location, but the neighbors were Not Happy, and now they are Not Happy that the station is so hard to get to…

            Where do you get the “tens of billions” figure for electrification?

          • D.Miller

            I was including the tunnel, the fleet, the countless presumed facilities modifications, the service levels that won’t actually pay for themselves, etc, etc, etc.

          • orulz

            With better performance due to electrification, could you just re-add the town center stations (with no parking) while keeping the modern park and ride stations, without blowing up the schedule, and wind up with a win-win? Newburyport is at the end of the line so it makes the line longer, but doesn’t add to anybody’s commute. Westborough would be two miles east of the park and ride, which seems reasonable spacing for an electric commuter line.

            Of course TOD apartments on the parking lots would be the next step.

          • D.Miller

            Everything in your first paragraph is factually viable and perfectly laudable, aside from the presence of the word “just”. I will never quite understand internet commentators who think that hundreds of millions of dollars in additional infrastructure can just be willed into existence with a couple of sentences, once every other precept of the very expensive but symbiotically perfect napkin plan has also come to fruition.

            “Middle-of-nowhere parking-lot TOD”, on the other hand, is one of the most insidious fallacies of the virtue-signaling crowd. All the disadvantages of apartment life with zero of the benefits. Invariably populated by those with little need to commute in linear fashion, but who wound up there only because bare-bones units in terrible locations were cheaper than other options. Car dependent in extremis, because there is literally no there there. Every example in existence is terrible. Focus on density in actual places.

  13. Pingback: Nesi’s Notes: March 3 | WPRI 12 Eyewitness News
  14. Joel N. Weber II

    Is there any reason, in a world with sensible labor contracts, why human beings who act as train conductors or locomotive engineers from 5 or 6 AM through 9 or 10 AM couldn’t spend the late morning inspecting / cleaning / maintaining the rolling stock instead of playing pool?

    Since Forge Park / I-495 isn’t within walking distance of much of anything, maybe we could replace it with a Kenwood Circle station in Franklin and then extend the Franklin Line along the Air Line alignment toward northern Rhode Island?

    • Mike

      I might be wrong but I assume they’re playing pool because they’re not being paid. If you end up paying them they would be better utilised driving trains than anything else.

      • R. W. Rynerson

        Depending on the labor agreement, people working splits are often compensated at a low rate for the interval of non-work. There also may be limits on the number of hours measured from the start of the first piece to the end of the last piece. And, of course, there are requirements for sleep (a split day that is too long can encroach on required time off-duty). Contracts also may stipulate minimum percentages of shifts that will be straights. This is one of the most arcane parts of transit and railway labor negotiations! Systems that try to chisel these costs down too far, such as the 1990’s private contractors in Denver, can expect insane levels of operator turnover or shortages.

        King County Metro runs very good midday service which has contributed to high ridership and has helped to warrant continuing their trolley coach network. However, one reason is that their labor agreement has very high minimums for straights, so they might as well have them driving.

        I smiled at the pool reference because both transit agencies that I worked for experienced brief media kerfuffles about providing nice facilities for the operators. When an afternoon snow storm hits Denver and extra buses are needed before the peak hour, or when a driver gets into her car and it won’t start – requiring someone to cover her run – customers can be thankful for those pool players, paid or not.

    • Bjorn

      Labor specialization. People who would make good train drivers may or may not make good train cleaners or mechanics.

      Train cleaners also usually get paid less than operators, but more than comparable janitorial work such as in an office.

  15. R. W. Rynerson

    ” This is the only market North American commuter rail serves … ” is an over-statement. Colorado’s commuter rail lines are designed for multiple purposes, as are Utah’s and to some extent even New Mexico’s. They can be criticized for serving undeveloped areas, but that’s a political chicken-and-egg decision that people were aware of when decisions were made.

    • Alon Levy

      Which markets does RTD serve? I know in Utah and New Mexico the lines hit adjacent cities and have some intercity service, analogous to New York-New Haven or Boston-Providence, but how many people actually make use of it?

      • R. W. Rynerson

        The loads for Denver International Airport are often the peak direction on the A-Line. The B-Line is a short trip into the suburbs, but in addition to conventional commuting is carrying student traffic and sports crowd traffic. The G-Line to come has employment within walking distance of two stations and a new community college campus at a third. The R-Line under construction will carry students and sports crowds and some reverse commute work trips. Other than the A-Line, none of this is remarkable, but neither were most of the other states’ commuter rail lines when they started. When the B-Line is eventually extended into Boulder County it will serve intermediate stations with employment.

        In some respects the RTD commuter rail system is a miniature S-bahn, coordinated with bus and Light Rail networks in one tariff chart, but we don’t have a category for that in North America. The contract operator has had to experience a learning curve when railway-experienced personnel were hired and had to learn some new behaviors.

        An aside regarding sports traffic: this is important to meeting Denver regional air quality standards. Puritans are always concerned that transit customers might be enjoying themselves, but a number of hard-headed people concluded that exceeding EPA standards when major events coincide with commuter peaks was not worth the risk of depending entirely on autos and Diesel buses. Clock headways through middays and early evenings automatically serve numerous small events that would not warrant special bus service. A commuter-peak only rail service would be of little use for these trips.

Leave a Reply to Alon Levy Cancel 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.