Notes on Boston-Springfield Service

Readers of this blog know I have a particular interest in intercity rail in New England stemming from growing up in New Haven. So when Eitan Kensky sent me a February presentation I hadn’t previously seen from the Northern New England Intercity Rail Initiative (NNEIRI, not to be confused with the Northern New England Regional Rail Association, or NNEPRA, which runs the Downeaster), I was seriously intrigued. There have been numerous efforts over the years to revive the Inland Regional service that Amtrak and predecessors once ran between Boston, Worcester, Springfield, Hartford, and New Haven, and this document presents the general outline of the group’s current vision for the return of such service. Much of the research seems to have been done by contractors HDR, and the predominant vision is clearly that of MassDOT, with secondary input from Vermont and other stakeholders.

NNEIRI study area map

NNEIRI study area map

Massachusetts has, of late, been focused on two major goals for non-Northeast Corridor intercity rail: a link to Montreal and restoration of Inland Regional service. The current study (logically) links these two together. Tough previous service to Montreal has run along the Central Vermont line, turning off the Boston & Albany at Palmer to serve Amherst before heading north through Vermont, the current vision has Boston-Montreal service using the recently rehabbed Connecticut River Line from Springfield to Greenfield before continuing north. It’s a little bit longer, but serves Springfield, Holyoke, Northampton, and Greenfield instead of just Amherst, and takes advantage of the state-owned Conn River trackage.

The predominant challenge to intercity rail in New England is that the trackage is in most places exceptionally curvy. The build alternatives envisioned for the NNEIRI service thus focused on regular-speed trains, with no ambitious plans for even moderate-speed (110 mph) options. It’s important to remember that “maximum speed” here means Maximum Authorized Speed, or MAS, rather than average speed. On curvy legacy tracks the trains are unlikely to obtain the maximum speeds for very long stretches, given FRA restrictions on tilt for conventional equipment (although the 90 mph MAS alternative does make brief mention of the possibility of acquiring tilt equipment).

Table of service alternatives

Table of service alternatives

A Boston-to-Springfield time of right around 2 hours would be extremely competitive with driving, which is about an hour and a half without traffic (yeah, right) and realistically usually at least a half-hour longer. It’s also about the same time as Peter Pan’s bus offerings, but a train would presumably offer a much higher level of comfort and reliability.

Costs would fall in the billion to billion and a half range for the bottom two alternatives, which seems on the high end for relatively simple double-tracking work within an existing right-of-way; I assume most of the capital expenses would be on the Vermont sections, since the B&A right of way is built to accommodate at least two tracks.

nneiri costs

Overall, the conclusion seems to have been that bumping MAS from 79 to 90 mph would result in considerable extra expense with little time saved or gain in ridership. The study team’s Draft Build Alternative is a modified Alt 2, with 79 mph MAS and slightly fewer trains:

draft build alternative

Eight trains per day would run through from New Haven to Boston, a kind of mini-Inland Regional service. These trains would function as extensions of the current New Haven-Springfield shuttle service. There would be one round-trip per day from Boston to Montreal, and another from New Haven to Montreal, while the Vermonter would continue as it currently operates, with an extension to Montreal. Springfield would get 9 round trips per day to Boston, and presumably the New Haven-Montreal train would have a timed connection with a westbound Boston-New Haven train at Springfield, giving Boston in effect two daily round trips to Montreal.

All trains are envisioned to make all local stops, which is interesting to me; I would have run the Inland Regional/shuttles as expresses in Connecticut, stopping only at Hartford. As it is, the additional 9 corridor trips will provided important added frequency to the NHHS/Hartford Line service that should be beginning in 2016. A 2011 NHHS document envisions full cross-ticketing between NHHS and shuttle/Regional trains, and the boost from NHHS’ 25-32 trains per day at launch to 34-41 including the corridor services is nothing to sniff at. However, that many trains would clearly require Connecticut to finish double-tracking the Hartford Line between Hartford and Springfield. That task isn’t itself all that complex but has been deferred to Phase II of the NHHS project (though it is included in Governor Malloy’s 5-year transportation ramp-up plan) because of the  considerable expense of rehabbing the Union Station viaduct in Hartford–which is, somewhat amazingly, believed to no longer be able to hold two trains at once–and the bridge over the Connecticut River.

Interestingly, study staff clearly believe that Springfield-Boston service alone would be a poor use of resources, labeling it “Low Ridership” and “Ineffective and Costly.” As Alex Marshall pointed out on Twitter, much of the envisioned ridership to New Haven is surely people from Worcester or the Metro West region who want a two-seat ride into New York City without doubling back into Boston to catch an NEC train.

Likewise, the study labeled plain Boston-Montreal service “Low Ridership,” while noting the potential for higher ridership in the New Haven-Montreal corridor. Despite decades of pleading for Montreal service, planners still seem to believe that Boston doesn’t quite deserve it. That’s not particularly surprising to me given how slow such service would be and how sparse population is along the corridors between the two cities. So for now, there will likely be just the one round trip per day, plus the possibility of a two-seat ride via transfer in Springfield, and that situation seems likely to stay the same for quite a while.

Other notes

Finish the Cross

As currently planned, the NNEIRI system looks like a sideways T, with the long axis pointing to Boston. I’m on record as a (self-interested) proponent of Albany-Boston service, and I think some of the improvements proposed here strengthen the case for finishing off a cross-shaped network with trains from Boston to Pittsfield and Albany. Double-tracking the Boston Line from Worcester to Springfield would leave less than 100 miles of single track from Springfield to Albany (it’s 102 track-miles, but there are existing sidings and stretches of multiple track). If trains can do Boston-Springfield in 2 hours, a time of 4 hours to Albany should be eminently achievable even without much in the way of speed improvements. With significant speed improvements (most of the line west of Springfield is limited to 40-50 mph, even though the trackage west of Pittsfield isn’t all that curvy or steep) a time in the 3:30 range–which my previous post identified as the time necessary to be competitive–should be achievable. That would open up the possibility of Boston-Toronto service via the Erie Canal corridor cities–a potential market for an overnight train?

Boston Line Capacity

One of the major ongoing dramas in New England intercity rail has been CSX’ reluctance to share the ex-B&A right-of-way with passenger service. Given current constraints, it is somewhat understandable; it’s a steep, curvy line that has suffered from decades of deferred maintenance (yes, part of that is CSX’ fault, but the neglect predates CSX ownership). CSX runs 25-30 trains per day on the line, which approaches the capacity of a mixed-use single-track line, even one equipped with advanced (by freight rail standards) CTC signaling:

From NCHRP Report 773

From NCHRP Report 773, “Capacity Modeling Guidebook for Shared-Use Passenger and Freight Rail Operations”

Double-tracking the line, however, offers enormous potential, jumping the capacity from an estimated 30 trains per day to 75. In other words, CSX could double current traffic–a situation no one sees as being around the corner in New England–and there would still be 15 slots per day for passenger traffic. More realistically, a fully double-tracked B&A could easily accommodate 40 freights, the 8 proposed Inland Regional trips, 6-8 trips to Albany, and the Lake Shore Limited–a total of under 60 trains per day west of Worcester.  Of course, fully double-tracking the line requires the states of Massachusetts and New York to cooperate, and the Cuomo administration has shown little interest in efficient passenger rail.

Pessimistic SPG-NHV times

The table of travel times above envisions a trip time of 1:40 from Springfield to New Haven given all local stops. This seems somewhat pessimistic to me, as the current shuttles and Vermonter are scheduled for 1:20 to 1:30 over the same route; perhaps the longer time takes into account that a few stops will be added under the NHHS scheme, but those should be counterbalanced by improved track speeds; it’s not a big deal, but I’m somewhat confused.

Who’s going to operate it?

Most commentary I’ve seen has assumed that any extension of rail service from Boston to Springfield would be operated by the MBTA. Running the trains through to New Haven would seem to preclude that possibility. Amtrak would seem the most logical choice, but the northeast state haven’t been thrilled with it of late; Connecticut, for example has opened the NHHS service to a bid competition. The NNEIRI network is an extremely complex system, involving at least three states, plus the province of Quebec and federal authorities regulating border crossings, the private railroads owning the tracks, and various other stakeholders. So perhaps now is the time to revive my call for a unified Northeastern passenger rail authority.



14 thoughts on “Notes on Boston-Springfield Service

  1. Assuming a decent transfer at Back Bay, how far west of Boston would you have to go for this to be faster than going east and taking the Northeast Regional or Acela?

    • That’s an excellent question that I had in my mind while I was writing. If times from BOS to NHV are really in the 3:40 range, you’d have to right around Worcester before it evened out. Times from BOS to NHV on the NEC seem to vary considerably, anywhere between 2 and 3 hours, and Worcester-BOS is about 1:30 and getting faster, making it essentially a wash. Worcester may also get a regional service to Providence at some point in the near future, which would probably be faster than going via Springfield.

      The other consideration is that the Inland trains, especially if not operated by Amtrak, may end up being considerably cheaper than the Shoreline ones, and you can change in New Haven to Metro-North, which Amtrak’s fare structure expressly discourages for obvious reasons. If I were marketing this service, that’s how I’d sell it–a discount option for NYC-Boston travel that takes longer than the Shoreline trains but is considerably cheaper, and more pleasant than a bus.

      • The breakeven point is way east of Worcester. Think about it: Boston-New Haven is about 2 hours on the Acela (and 2.5 on the Regional), when Amtrak feels like staying on schedule. If inland Boston-New Haven service is 3:40, then the breakeven point is 50 minutes west of Boston, and not 1:40 west of Boston. That puts it right around Framingham, assuming there’s a timed transfer at South Station/Back Bay, which, to give credit when it’s due, there sometimes is.

        Also, the mind boggles at how things got to the point that high-cost low-speed diesel service offers cheaper tickets than medium-speed electrified service. Electricity is cheaper than diesel, mostly cruising at 200 km/h north of Kingston is cheaper than the repeated acceleration and deceleration cycles of the B&A’s craptastically curvy track, and the crew costs less per kilometer when the speed is higher. So what gives?

      • (reply to Alon, it won’t let me respond to his comment directly for some reason)

        You’re right about the break-even point; blame a late-night brain fart for my bad math. I don’t know for sure that the Inland trains will be cheaper than the Shoreline ones; Amtrak currently charges around $20 for NHV-SPG on a single-ride ticket, which is kind of steep for such a short ride, but also pretty cheap by Amtrak standards. I just assume they’ll be marketed as the cheap option since the trip is longer for most riders, and the most convenient connection in NHV will presumably be to Metro-North.

        The larger question, of course, is whether Amtrak’s pricing policies on the NEC in particular make any sense. They probably don’t, but I suspect part of the reason they exist is to manage demand so that Penn Station doesn’t get totally overwhelmed. That certainly seems to be the case on the Empire Service, which could probably capture 100% of non-car Albany-NYC traffic with a few minor improvements, yet costs twice as much as competing buses.

      • Amtrak’s ticket policies do make sense given the extreme peakiness of demand, and working within the constraints of Congressional micromanagement that they have to put up with, as well as the constraints of capital investment. They’re limited in the range of prices they can use for yield management, and they’re limited by the number of Amfleets (and formerly, the number of electric locomotives as well). So they set the upper end of the price range to sell out their Friday and Sunday trains, which they can’t run more of due to limited rolling stock, and then the lower end can’t be less than 50% of that, which still ends up being relatively expensive. They figure that it’s more profitable to not turn away Friday/Sunday riders due to sold out trains, and instead lose some of the low end of the market on other days.

      • »assuming there’s a timed transfer at South Station/Back Bay, which, to give credit when it’s due, there sometimes is.«
        Given that we want timed transfers in Springfield and New Haven (or interlining south of NH), this might become a nice demonstration of the problem of fitting together conflicting transfer needs.

        Might both new Montral trains be able to offer a timed transfer in Springfield for a ninth daily trip New Haven–Boston?

      • I think the Springfield transfer is less of a big deal, at least southbound, since there will also be NHHS trains running at least hourly, and more frequently at full build-out. That being said, a timed transfer would be awfully nice, as would a timed transfer to MNRR trains at New Haven.

  2. Upgrading Framingham-Boston is going to be an expensive mess, though once done it would seem to have many benefits for local and regional service.

    • Yeah, no doubts about that. The state and the MBTA do seem to be moving relatively quickly on it, and it’s a suburban constituency that has the pull to demand priority, for better or worse.

      • I’m not aware of any efforts to fix the numerous problems of inner-Framingham infrastructure, that were left over from the devil’s pact struck by the B&A with William Callahan during the Mass Pike extension period. The railroad still has lineside signals, a 60 mph limit, some horrible inaccessible single-platform stations, and a lack of crossovers. The double-tracking project through Beacon Park is yet to be finished, 2 years after they claimed it would be.

        It shouldn’t take an hour an a half to get to Worcester — that’s only about 45ish track miles.

        It’s actually rather puzzling considering the constituency.

  3. That task isn’t itself all that complex but has been deferred to Phase II of the NHHS project (though it is included in Governor Malloy’s 5-year transportation ramp-up plan) because of the considerable expense of rehabbing the Union Station viaduct in Hartford–which is, somewhat amazingly, believed to no longer be able to hold two trains at once–and the bridge over the Connecticut River.

    I’m not sure where concerns over the viaduct’s potential (in)ability to carry two trains at once were first stated but my understanding is that concerns over the expense of a rehab project are entirely secondary to the fact that the Aetna Viaduct is a non-optional repair or replace job within the next 15 years or so and the Union Station viaduct is pretty much square in the impact zone of anything you’re going to do to the freeway.

    Based on that, it makes no sense to overhaul the rail viaduct now when there are at least two legitimate proposals (I’d argue tearing down 84 in Hartford entirely and building a surface boulevard instead is a third legitimate proposal but people disagree with me on this) on the table that would end up requiring you to tear the thing down and rebuild it anyway, essentially meaning you ended up double-paying for track improvements into Union Station.

    Now, obviously, if the preferred alternative for 84 repair/replacement that emerges is one which doesn’t end up significantly impacting the railway, then double-tracking the viaduct should move back into Phase 1 – but, far from that being clear yet, it’s increasingly likely that the rail viaduct IS coming down as part of 84 repair/replacement.

  4. FWIW, even the appallingly car-obsessed, anti-train Andrew Cuomo administration has ended up doing more for passenger rail than the three administrations before it. I think the trend in New York is strong enough that it’s going to overcome the personal biases of individual governors, and we’ll probably start getting some serious support in the next administration.

    In practice, we will only see improvements on lines which are purchased by the state. As long as a “freight railroad” owns the right-of-way, we’ll get nothing nothing nothing. The freights can be very friendly tenants, but they’re awful landlords.

  5. two points

    1.) the inland route needs to be restored because of the limit on the number of trains on the NEC in CT because of bridge opening … it is at or near the permitted limit now !
    2.) if the MBTA replicated the CalTrain “Baby Bullet” skip stop service between San Jose and San francisco (same distance, same FRA class 4 track, same locomotives) it could run commuter rail service between Worcester and Boston in less than an hour. Amtrak train with one stop in Framingham would be even faster

    • Boston-Worcester is currently Class 3, which is one of the reasons for the 60 mph limit. Hopefully they’ll bring the maintenance standards up soon and get decent speeds. Also, you’re probably not going to replicate the Baby Bullet stopping pattern because the development pattern is totally different (i.e. not 40 miles of solid strip malls and office parks).

Leave a Reply

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

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

Google photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s