Re: North-South Rail Link
As promised, lengthier (much lengthier) notes and thoughts on last night's meeting. I'll be repeating/summarizing things I heard and saw, but I'm definitely not in a position to assess most of it-- I'm no expert, just an enthusiast. Those who were also in attendance, please correct anything you see that you believe I'm mistaken on.
I arrived a bit before 6pm and browsed the poster displays, many of which were familiar from the PowerPoint that was released on the MBTA's website early this week (and rather well-hidden up there, if you ask me), though several of which I noticed were slightly different than what I recalled from the presentation. Tables that consolidated certain line items, for example. This sort of lack of consistency seemed to be a theme throughout the evening.
One item of particular interest to me was a pair of maps, depicting different service patterns under the 2-track versus 4-track alternatives. A number of interesting nuggets from those maps aside (both depicted the Needham Line and Franklin Line as continuing to use the NEC, both showed full-builds for South Coast Rail via Stoughton, the 4-track showed through-running Fairmount service that forks in the north, running both to Fitchburg and [I believe] Newburyport), I was frustrated to see that the 4-track map did
not indicate a Central Station.
This brings up my first pet peeve about the Reassessment Study (as it has currently been released): in three of the alternatives they present, two stations (South Station Under, and North Station Under/Haymarket-State Under) are assumed. But, without explanation, the 4-track Central Artery alternative plops in a third station, basically at Aquarium Under, which shoots up the price.
When I asked about this before the proper presentation, I was told (I believe his name was Tony) simply that the 4-track alternative required Central Station as part of the scope they were given.
Either way, the inclusion of Central Station only in the 4-track alternative creates an apples-to-oranges comparison that muddies the waters (another theme of the evening). If two stations are acceptable in the 2-track CA alignment, they should be fine in the 4-track as well.
And after all that, Central Station didn't even show up in the service plan diagram they presented. Hardly confidence-inspiring.
The presentation began at approximately 6:10. While on the one hand I understood why they wanted to start early (given the anticipated length of the public comments and questions), it bothered me that the presentation was advertised as starting at 6:30 but started 20 minutes before that. To be honest, I think it's disrespectful. But I digress.
The presentation itself was interesting enough. There were a number of slides that were not included in the online version, and my understanding is that it was also a noticeably distinct presentation from what had been presented earlier in the week. I took notes on things that I didn't recognize from the online presentation:
Significant emphasis on Southside Accessibility, on the current lopsided access to jobs depending on which side of Boston you live on.
There was a really cool graphic of Orange Line congestion at each station during the AM peak, with two series of bars, one for northbound and one for southbound, creating a pair of overlapping bell curves. I hope that the folks at TransitMatters will be able to get access to such data, because that graphic did a better job of telling the NSRL-Relieves-Central-Subway story than anything I'd seen previously.
Their estimates for tunnel construction includes flying junctions (definitely at the north portal, I was a bit less clear on the southside). Their argument was that these were necessary in order to support the kind of high-throughput service envisioned for the Link. (On the one hand, this obviously escalates the cost; on the other hand, it's not like there'd be a second chance to build this thing, so probably is fair to assume flying junctions.)
They proposed 17 trains per hour (tph) through the 2-track alternatives (this rises to 21 in the 4-track). They clarified that this is 17 MBTA trains per hour, with the assumption that Amtrak would send another 2-3 tph through, bringing us to about 20 tph. They asserted that, even with upstream improvements elsewhere in the system (double-tracking, resignaling, etc.), 20 tph is the most aggressive that could reasonably be assumed.
Citing Amtrak's NEC Future study, they assumed that Amtrak service would involve:
- 4 Express (Acela) tph to South Station Surface
- 2 Metropolitan (Regional) tph to South Station Under, continuing without stops to Anderson/Woburn and then Lowell (interesting)
- 0.5 tph of Downeaster-Lake Shore Limited
- One other line item I didn't get down (sorry)
(This was an interesting list. As a Providence commuter, I certainly wouldn't mind 30 minute Amtrak headways on the Regional, but I'm skeptical. Seems to me that the necessary infrastructure improvements to the NEC, B&A, Lowell and Haverhill Lines would have a pretty significant ripple effect on the rest of the MBTA system, which probably would impact the variables examined in this study. Also, while we're alluding to more widespread electrification, they noted that electrification of the Fairmount Line would be necessary during Link construction, in order to support continued Amtrak service. But, post-construction, they seem to forget that there is now a second electrified corridor on the southside. Plus, if the Worcester, Lowell and Haverhill (minus Reading) lines are all assumed to be electrified anyway... but I'm getting ahead of myself.)
They briefly discussed fares. They assumed subway fares for travel between stations which are currently Zone 1A, same fare structure as today for travel from suburbs to downtown, and a new distance-based formula for suburb to suburb travel that avoids hitting you with the Zone 1A premium. They also acknowledged that all this may change in the near future with new fare practices on the Commuter Rail.
They showed a pie chart (which I believe is available in the online presentation) of post-Link destinations for passengers coming from the northside. The majority of riders were still projected to get off at North Station. My question (which I did not get a chance to ask) was which alternative this pie chart examined-- if it was using the non-Central Artery alignments, wherein "North Station" becomes "Haymarket-State Under", then that's an obvious explanation for the supposed "lack" of changed behavior. (There are probably lots of questions to raise about that pie chart and its analysis, but that was something that jumped out at me during the presentation.)
In an obvious case of scope-creep muddying the waters, they presented pretty dismal-looking numbers about air quality. First, they assumed use of dual-mode locomotives, which would, in some cases, improve air quality. However, because they also simultaneously proposed (in most of the scenarios) a
major increase in the frequency of service, and because much of the network (they assumed) would still use diesel (but again, I refer to my earlier points about Worcester, Lowell and Haverhill), the air quality improvements, they claimed, would largely be negligible.
They specifically called out development opportunities at North Station, the Post Office next to South Station (which touches on another theme of the evening-- constant comparisons to SSX), and Widett Circle.
They noted that the NSRL would increase capacity (in terms of number of Commuter Rail seats arriving in Downtown during the AM peak) by 25%.
My back-of-the-napkin-style notes for the costs were:
- ~$2.4B for vehicles (dual mode locomotives and I believe also the additional diesel locos and coaches needed to support higher-freq service)
- ~$1.3B for various upstream improvements, which included ~$376mil for double tracking and ~$400mil for resignaling
- ~$500mil for electrification of Back Bay through the tunnel to the first stop after North Station on each of the northside lines
The stations under the South Congress alignment were a little bit cheaper, I noticed.
For all-day peak-level service on the 2-track alignment, they posited operational expenses of ~$929 million per year. [EDIT: This formerly said “billion” instead of million, which is my mistake. My apologies.]
An item of particular note, given all the brouhaha over costs: they put up a very interesting table that compared the costs, in 2002, 2018, and 2028 dollars, estimated from the three studies (the current one, the Harvard study and the MassDOT study from 16 years ago). When actually compared apples-to-apples, using the same years' dollars for each, all three proposals came out to the same ballpark for all of the 2-track alternatives. (The four-track was costlier in the current study than previous studies had estimated, by ~30%.) My interpretation of that is that there really isn't news here-- accounting for inflation, the costs of the 2-track alternative, as previously studied, is in the ballpark of $8B in 2018 dollars, which is also what the current study concluded (again, using 2018 dollars for apples-to-apples). $8B
still seems high to me, but that's a different problem.
The presentation ended, and question/comments period began.
At this point, my overarching reaction was that it seemed extremely arbitrary to count the costs of
some upstream improvements, while
not assuming systemwide electrification and
then also assuming all of the drawbacks of continued reliance on diesel locomotives. They gave a vague handwave about wanting to keep the numbers down to justify not assuming systemwide electrification, but the choices seemed arbitrary and unjustified. Again, muddying the waters with apples-to-oranges.
Several people delivered prepared or semi-prepared remarks during the questions period.
First off was former Governor Dukakis, well-known by now for his transit advocacy. His remarks were relatively brief but to the point: boy, how did these construction costs get so high?
Former State Representative John Businger made two key points (according to my notes): first, that the NSRL is not an MBTA project, but an NEC project; and second, that, contrary to what had been asserted during the presentation, the construction of SSX
would preclude NSRL due to the consumption (or waste) of public goodwill. He also asserted that SSX is never going to happen anyway, because the Post Office isn't moving.
Clay Schofield, who managed the NSRL project for the T back in the '90s, spoke at greater length to review the benefits of the NSRL, benefits which he felt were not adequately conveyed in the current crop of presentations. Those included (according to my notes): 41K cars taken off the road; 55K fewer riders on the Central Subway; 55K man-hours saved per day. He criticized the current proposed service levels, arguing that they and the ensuing ridership projections are not realistic, and noting that the service levels themselves are a major driver of cost. (I think his point, which could have been clearer, was that there are benefits to the Link itself, independent of the raised service levels-- that the Link can stand on its own merits, and doesn't need the raised service levels, which bring their own drawbacks, to raise it to the level of viability.)
He noted the lack of discussion of freight traffic through the Link. As an aside, he criticized the discussion of South Coast Rail, noting that the SCR proposal only calls for 2 trains each from Fall River and New Bedford, which hardly seems worthwhile. Finally, he argued that the improved efficiency resulting from the centralization of maintenance activities at the BET was not properly accounted for in the current analysis.
As previewed in the papers earlier during the day, Professor Barry Bluestone rose to call the study "extremely unbalanced" and accused it of trying to "scare the public." I found his remarks very well-done, so I will line out the major points:
- The presented inflationary numbers are unfair, he argued, because revenues, income, etc. will rise in parallel with these costs (essentially negating, or at least minimizing, the ballooning effect-- that's my interpretation, anyway)
- Inflation in recent years has been ~2%, making the 3.5% used in the study unfair
- He called for more pressure and oversight from MassDOT to keep costs from contractors down
- Echoing the earlier discussion of lopsided job access, he described how The Gap creates essentially two separate Boston job markets, saying that for 85% of the population, their home and place of work are located on the same side of The Gap-- i.e. not many people from the northside work south of South Station, etc.
- He noted that 52% of population growth will occur in towns like Lawrence and Andover (i.e. north of Boston)
Finally, in one of the striking moments of the evening, Bluestone presented numbers from a forthcoming study to argue that the biggest beneficiaries of the NSRL will, in fact, be those who do not use it. Drawing our attention to the group of people who
still will need to drive, he pointed out that current average speeds on the arteries are around 18 mph. Because most cars are inefficient at those low speeds, this creates a real cost (environmentally and through fuel costs). If, he said, auto use on those arteries were to drop by a mere 13%, it would double the average speed of traffic up to 36 mph. The resulting savings in fuel costs due to better mileage at higher speeds, he calculated, would be approximately $2 billion.
Brad Bellows, who has a long history with the NSRL, spoke next. I agree with Java King's assessment-- Bellows spoke very very well. As others have noted, he said that the original premise was for equal resources to be spent on benefits and costs, but in fact benefits analysis got approximately 4% of the pot instead. He argued that the estimate should reflect industry best practices, but clearly doesn't, citing a number of similar projects, mainly in Europe, that cost much less, or much less per mile. (A Dutch tunnel was 3 miles for $2B, and Crossrail, which is extraordinarily complex, is 26 miles with a large number of stations for only $40B-- by comparison, at the rates this study used, Crossrail "should" instead cost something like $200B.)
He alleged resistance from MassDOT to having international (specifically international) participation from experts with actual experience with this kind of project. Things like minimal time on the agenda, lack of proper IT support for a Skype call remote conference.
Why did this proposal assume locomotive-powered trains?, he asked. No modern high frequency regional system does that. Do we really think, so many years from now, we'll still be using locomotives?
He then made a really interesting point. The current push-pull system uses low-frequency high-volume trains. This requires, among other things, really lengthy platforms. Service run on high-frequency multiple-unit trains could use shorter consists and, therefore, shorter platforms-- thus potentially reducing one of the biggest costs for this project, the stations themselves.
Ed Mueller, who was co-lead in the working group, was the last person I heard before I had to leave, and unfortunately I could not stay for his full remarks, which mainly focused, as others have noted, on how the PR for this new study really was incredibly botched. Confusion over whether this was the actual report, whether the numbers were in 2018 or 2028 dollars, inconsistent presentations that focused on risks without discussing benefits. This, I felt, was actually the most severe criticism of the evening (though Bluestone's and Bellows' are tied for a close second)-- the study itself can be revised with better numbers, better practices, better models, and so on, but the rollout to the public is a moment that has come and gone and appears to have gone pretty badly.
For my part, I came away from the meeting pretty comfortable in the conclusion that systemwide electrification-- independent of the NSRL-- is something I could imagine happening in the next 10 years, while the NSRL is not. However, once the process of electrification begins, the entire analysis of the NSRL will need to be redone, and, I suspect, it will have another round at bat. And, particularly if Amtrak makes any significant improvements in either the NEC or Worcester Line corridors, that will change the equation as well.
Unsurprisingly, I find myself in continued agreement with the Transit Matters Regional Rail report-- NSRL will be a fantastic asset to our burgeoning regional rail system, but it is by no means obligatory for such a re-visioning and should not be allowed to stand in the way of other improvements-- nor should it be the primary focus.