North-South Rail Link Discussion

Balerion wrote:Rail link advocates continue push for $2 M. study

Progress on obtaining a study — any study — into a proposed rail link connecting Boston’s North and South stations chugged along as the proposal's working group met for the second time at the State House on Monday.

Gov. Charlie Baker said his administration still is weighing the potential scale and scope of a study.

Last week, Rep. Seth Moulton told POLITICO Massachusetts that he would be interested in a public-private partnership to develop the rail link, similar to a high-speed rail project in Texas. Bellows said on Monday that a public-private partnership has been explored with the North-South link project.

But Arena threw cold water on the idea. “We just don’t do that in Massachusetts,” Arena said of a public-private venture. “There is a bias against it, in my opinion. It’s not second nature. Maybe it’s the Yankee skepticism and hesitation.”

A minor update, but no real news just yet.

No private enterprise is going to build and run such a money-losing tunnel. And it is too short to be a standalone venture, and the idea was it was supposed to be a run-through operation. So this company would build the tunnel & recover their investment by charging MBTA/Keolis/Amtrak trackage fees?

Maybe build autotrain terminals in Lowell, Woburn, Mansfield, Franklin and Framingham-I-90. Auto trains run express every five minutes during rush hour to the other points whisking cars off congested highways & lower emissions.

Zurich just completed its version of the NSRL (including a 3mile S-shaped tunnel and flyovers on the approaches) for CHF 2.1Billion (about the same as USD $2.1 Billion)
http://www.railwaygazette.com/news/infr ... leted.html" onclick="window.open(this.href);return false;

All the solid red stuff on the bottom of the map below is how they get mainline trains that previously stub-ended into Zurich HBF to instead flyover and get into the tunnel portal on the south side of the HBF. The analogy in Boston would be flyovers to allow all Southside lines to flyover into a portal without doing crossing movements across the NEC, and for Northside lines to all feed into that end without fouling the Lowell Line (which we assume is the new intercity line)

The Tunnel is shown dotted, below (click to enlarge)

I'm surprised to see support for a 2-track N-S connector with 2 tracks for the subway. It seems a bit half-baked to me, honestly. If the T is going to spend earth-shattering amounts of money to build this, they might as well build it to support high frequencies. Merging multiple lines together at rush-hour frequencies with stops in the middle (SS Under, NS Under, maybe a Central station) will make 2 tracks tight. A breakdown at rush hour would cripple the entire system--remember, both directions are peak. The end result of a 2-track tunnel with transit would be that the tunnel is clogged, and the operations difficulties cannot be resolved by adding without cannibalizing subway service. Especially considering the dwell times--you'd have large numbers of people both boarding and detraining at the downtown stations, and having large numbers of other trains waiting behind that would slow things down.

Yes, the East River tunnel does it, but it has no stations, and only one peak direction (so one tunnel is a more viable option). Also, if I understand correctly it's a huge operational bottleneck that limits capacity. The perhaps more accurate comparison for track-number needs is the SEPTA Center City connection. It's very different, in that it's used by all SEPTA lines and has no intercity rail, but I think the service frequencies are the right order of magnitude. It's 4-track, and standing at Market East (a 2-platform, 4-track station in Center City) at rush hour it's clear there's no way you could shove that level of service onto 2 tracks.

With proper equipment selection and signalling - and stations designed for speedy boarding and alighting - you can get enormous throughput in a two-track tunnel. Boston would actually have the advantage that all service is truly running through - Penn Station and the Center City Tunnel both have some mix of tunnel-terminating and through trains.

Crossrail, scheduled to open under London in two years, will run 24TPH with the possibility of 28TPH in the future. Right now, the southside gets a max of 15TPH (7-8am) inbound and 20TPH (5-6pm); the northside tops out at 13TPH inbound and 14TPH outbound. So a two-track tunnel could - even with a mix of EMUs and dual-modes - reasonably handle all of Boston's current service plus Amtrak.

It's when you start getting into real service increases that you want a four-track tunnel and (not instead of) expanded terminals. Indigo service to Peabody, Waltham/128, Reading, Riverside, Readville. Clock-facing Lowell, Rockport, and Newburyport service plus Concord, Portsmouth, and Maine expresses/intercity. Central Mass. Three different service patterns on the B&A plus branch service to Clinton. Two new branches off the Franklin main. Cape Cod. Even real service on South Coast Rail. All of that is possible by 2040, and you'd need current terminal capacity plus 2-4 tracks through plus expanded terminals to deal with it all.

ferroequinologist wrote:I'm surprised to see support for a 2-track N-S connector with 2 tracks for the subway. It seems a bit half-baked to me, honestly. If the T is going to spend earth-shattering amounts of money to build this, they might as well build it to support high frequencies. Merging multiple lines together at rush-hour frequencies with stops in the middle (SS Under, NS Under, maybe a Central station) will make 2 tracks tight. A breakdown at rush hour would cripple the entire system--remember, both directions are peak. The end result of a 2-track tunnel with transit would be that the tunnel is clogged, and the operations difficulties cannot be resolved by adding without cannibalizing subway service. Especially considering the dwell times--you'd have large numbers of people both boarding and detraining at the downtown stations, and having large numbers of other trains waiting behind that would slow things down.

Yes, the East River tunnel does it, but it has no stations, and only one peak direction (so one tunnel is a more viable option). Also, if I understand correctly it's a huge operational bottleneck that limits capacity. The perhaps more accurate comparison for track-number needs is the SEPTA Center City connection. It's very different, in that it's used by all SEPTA lines and has no intercity rail, but I think the service frequencies are the right order of magnitude. It's 4-track, and standing at Market East (a 2-platform, 4-track station in Center City) at rush hour it's clear there's no way you could shove that level of service onto 2 tracks.

It's a unified bore for the 2 tracks that fan into 6+ tracks at the terminals, so having multiple crossovers makes the resiliency issue a lot more benign than having 1 track per tube...which hasn't killed Penn Station yet after 100 years of just 2 tracks, 1 per North River tube. The traffic levels here won't even begin to approach that.

And the only way you get ultra-high frequencies on this system is by retaining the surface terminals and playing mix-and-match. The terminal districts are the limiters to service levels on most of the mainlines, not the mains themselves. The tunnel uncaps it by letting trains bail out into a second terminal district before the great mash-up at SS and NS. You have to leverage BOTH the tunnel and surface terminals to fullest extent to get a 100-year service level increase ceiling. Run-thru by its lonesome and replacing the surface terminals isn't enough justification for building this thing. It's a lateral trade on capacity and/or modest increase where you're swapping one terminal district limiter for another terminal district's limiter...not uncorking unlimited system capacity by having everything at your disposal. This is not SEPTA Center City redux. We are not Big Digging the surface terminals. That myth needs to die.

The value proposition here is that you get the frequencies of your dreams, and can run the highest-demand ones as run-thru. But you are most definitely not doing "everywhere to everywhere" when the highest-demand pairs align with the compass. Branchlines and whatnot are probably going to keep turning at the surface as their regular patterns when there isn't a particularly high-demand slot meriting a run-thru. Lines that intersect a "north-south" subway orientation before hitting the terminals are probably going to retain a stable number of surface slots. North vs. south mismatches in service levels, mains vs. branches, 128-turning Indigos vs. long-hauls, etc. are going to keep portions of most lines' schedules doing a healthy number of surface slots for cleanliness amid asymmetry. For instance, there's only 2 mainlines south-- Fairmount and Worcester--that are Indigo candidates. Old Colony has Red out to 128 and 3 branches to feed; NEC has monster traffic loads and may still need to trade Needham over to rapid transit to manage its traffic levels through 3 SW Corridor tunnel tracks. But all 4 northside mains are pretty evenly-matched with Indigo potential, so matching 4 vs. 2 isn't going to be easy without creative dispatching. This tunnel isn't a replacement, it's a huge lift above-and-beyond. The tunnel doesn't need to carry the world on its back like Penn does for everything running thru or turning, like Center City does. This is not like those. It makes managing the system under unbelievable loads easier by giving 2 terminal options for everything.


So with that in mind, 2 bores of 2 tracks is not as high-stakes if you study the alternatives where 4 of them are RR tracks or it's 2 x 2 RR vs. rapid transit. And they'd probably do well to ballpark all the options out for simple due diligence. For one, how exactly does one distribute all the northside traffic that's going to be slamming the hell out of South Station when the Red Line and Silver Line are already having their dwell times decay precipitously from overload. There's no way forward for distributing all that load without Red-Blue, some sort of built replacement for Silver Line Phase III that preferably runs on rails, and substantial portions of the Urban Ring. Because you truly do not want to be riding Red as presently constituted with unlimited commuter rail capacity throwing several times as many people at both the above-ground and below-ground terminals. Load distribution may indeed force that issue, because it's not like the existence of the Link means people are going to start shooting in one ear and out the other like downtown is flyover country. No...they'll be crushing SS, DTX, Park, Back Bay, State, GC like they've never been crushed before. Taking on the challenge of the Link is going to virtually require getting just as bold with tearing through the rapid transit backlog. Especially on the circulator/interconnection type projects. This is a big, huge, massive commitment. And anything should be on the table for how to best equip downtown for the loads. Including study alternatives for how best to utilize the two 2-track tunnel bores.

Nobody's saying half-and-half is what will be recommended. But they don't really have a choice but to do their homework, because the Link is not just about frequencies via run-thrus. It's frequencies by doing away with the entire notion that Boston fas a terminal-induced finite frequency limit, because any northside train and any southside train will have two full-capacity terminals to choose from on the last-mile approach to sidestep the mash-up of all lines into one place.

Good point about having more than 2 tracks through the stations, which would make the North River Tunnels the better equivalent. According to wikipedia (sorry) there are 24tph in the peak direction at a maximum. That's 2.5 minute headways. Or stated another way, that's enough for 6 lines to have 15 minute headways. So you could route the NEC, Worcester, Franklin, Fairmount Indigo, B&A Indigo, and Needham lines through at 15 minute headways, which does seem adequate considering not all service will run through.

However, I still think the system of people's dreams would require room to grow. Maybe transit will develop such that more lines can support high service frequencies, or northside Indigos could be developed (maybe Waltham, Woburn, Reading, and Salem?). Making things 4-track would allow for the T to grow into the tunnel, as more lines get electrified or get dual-modes. I definitely agree an RT option is worth considering (there'd be too much money to not consider options), but for a project this big I'd think supporting expansive transit goals long-term would be the way to go. But like you said, once the options are analyzed, it might be how it ends up.

ferroequinologist wrote:Good point about having more than 2 tracks through the stations, which would make the North River Tunnels the better equivalent. According to wikipedia (sorry) there are 24tph in the peak direction at a maximum. That's 2.5 minute headways. Or stated another way, that's enough for 6 lines to have 15 minute headways. So you could route the NEC, Worcester, Franklin, Fairmount Indigo, B&A Indigo, and Needham lines through at 15 minute headways, which does seem adequate considering not all service will run through.

However, I still think the system of people's dreams would require room to grow. Maybe transit will develop such that more lines can support high service frequencies, or northside Indigos could be developed (maybe Waltham, Woburn, Reading, and Salem?). Making things 4-track would allow for the T to grow into the tunnel, as more lines get electrified or get dual-modes. I definitely agree an RT option is worth considering (there'd be too much money to not consider options), but for a project this big I'd think supporting expansive transit goals long-term would be the way to go. But like you said, once the options are analyzed, it might be how it ends up.

I think the north vs. south mismatches are going to inform some of that. Since lifting the terminal district service frequency ceiling explodes the amount the mainlines can handle, it does create some inequities for the routes where the mainline itself is the limiter. Such as. . .

-- Needham, as previously mentioned. Only having 3 tracks in the SW Corridor tunnel for all manner of NEC traffic puts Needham in the vice grip that prevents any meaningful service increases. You can start shifting thru Franklin trains over to Fairmount (eventually all of them) to clear one branch away for Providence and Stoughton/South Coast. And you can outright close Hyde Park station since Fairmount 2 blocks over will have vastly higher frequencies and suck in all the Cleary Sq. buses for a loop. But that only uncaps it south of Forest Hills. Inside the SW Corridor tunnel Needham is still the problem because it's the one that's crowded out, and it stops at Ruggles every time while others mix-and-match their skips. Something has to give there, and it won't be widening the SW Corridor because they already priced that out in a study about 10 years ago; the cost pushed $1B and the surface impacts were unacceptably destructive. So that's where the rapid transit trade-in comes in. You can handle 3 mostly slow tracks with that SW Corridor traffic only being Amtrak, Providence, and Stoughton/SCR and a reduced reliance on Ruggles. They'll sort themselves out on the station-less gap to 128 and the 4-track territory beginning at FH. But such a rapid transit trade-in becomes necessary for both Needham to get its slots and for keeping things loose and fluid through Roxbury.


-- The Western Route to Reading is pinched by the constrained space it shares with the Eastern Route around Sullivan Sq., single-track along the Orange Line, and grade crossings north of there. To get a full-featured Indigo route on that line (where it would thrive!) requires re-shearing the Reading and Haverhill lines back in half and sending Haverhill back to the NH Main and Wildcat on its pre-1979 routing. Thankfully that option exists, and it's arguably the best of both worlds because it would give Reading more frequencies and less overcrowding while shortening Haverhill travel times to a more tolerable hour and easing their overcrowding. But that's all you can really do with it, so it's not going to be able to take much advantage of the Link. Ripping out the 3rd Orange track is expensive, duplicating the electrification to Oak Grove is expensive...and ripping out the Orange track just pushes the headway limiter a little further upstream to the Melrose and Wakefield grade crossing clusters that slow everything down. The inner Western will still be dragging up the rear on natural capacity amongst the 4 northside lines. And in addition, simply feeding the Eastern Route's exploding volumes is going to create an inequitable pinch around Sullivan Sq. Where you don't have the option to rip out Orange Track 3 because the Community College viaduct touching down just south of the station seals off the room for turning out an extra CR track. Unless you do some hideously destructive blow-up/rebuild of the whole works.

So if it's going to be that construction-invasive in co-mingled Orange territory, and $$$ has to get spent zapping a few grade crossings further north in order to get throughput comparable with the other 3 mains...how much pain becomes worth it for the CR mode vs. putting the Orange-Reading extension back on the table? At least that one doesn't require touching a single thing to Oak Grove except grabbing the abandoned CR track for a Track 3 extension to Oak Grove, and if you have to spend money eliminating grade crossings does it matter if it's Purple Line or Orange that does it? You could probably apply the cost savings from not having to rip up Somerville, Medford, and Malden to the rest of the grade crossings CR upgrades wouldn't totally eliminate south of Reading and end up nearly cost-neutral. And you'd no longer have the problem of trying to match 4 north Indigos with 2 south Indigos. 3 vs. 2 means Riverside and Fairmount would basically switch off as heavy-traffic mains and feed the north trio as alternating branches a la Ashmont and Braintree where the schedule stays nearly as robust and the south destination pairs follow a logical switch-off. 4 vs. 2 means somebody on the northside is either getting their service shorted much more diffusely, or being forced to bail out with a surface terminal short-turn.


^^With both of these examples, NSRL isn't a panacea. It does induce some minor inequities that pick winners and losers by the way traffic is shaped, and requires some extracurricular mitigation to keep the transit fairly distributed. So there definitely are unintended rapid transit consequences. There's the need for billions in needed downtown radial transit to go back on the front-burner before the downtown subways collapse under load. Urban Ring, Silver Phase III or light rail facsimile, and Red-Blue. There's the need for maybe not billions, but definitely a due-diligence buildout of Purple-to-rapid transit transfers ringing the outskirts of the city to spread the load around away from the downtown singularity. Stuff like Blue-Lynn so the North Shore doesn't overwhelm the downtown airport routes, the North Shore's bus depot can distribute downtown destinations without sending half the buses through the airport tunnels or over the Tobin Bridge. Stuff Green from Union to Porter for a Fitchburg/Red/Green/77 superstation so one of the fastest-growing parts of Cambridge and Somerville have a node diverging all directions that doesn't require cramming down the gut downtown to pick a line. Red from Ashmont to Mattapan for a timed-transfer connection up the street to Blue Hill Ave. station. All of these considerations--some also lingering ever since the BTC drew up that rapid transit expansion map in 1945--also get pushed to the fore as unintended consequences of the NSRL and how it changes traffic flow.

It really is that big and transformative. The amount of stuff they will have to study just to make educated guesses at where the pressure points on the existing system are going to shift is going to end up being a lot similar to what LIRR is trying to wrap brain around now with East Side Access and where that one build completely changes the game on their routes (and where their riders will or won't transfer going forward vs. where they've transferred historically), impacts on certain subway lines, impacts on bus routes, impacts on Metro North, impacts on Penn capacity for Metro North and NJ Transit. It's so big every mode gets its demand book rewritten. That's what we're looking at here. It'll rewrite the transit needs list in ways that are hard to picture today with where demand clusters on the legacy system. It is most definitely not going to self-contain the demand changes within the Purple Line when it's built. It's a systemic revolution affecting every mode, with every mode needing to adapt and possibly expand as a result.

It's also worth bearing in mind that a tight, short-headway operation on two tracks requires a very short dwell time, not easily achieved with present equipment in the U. S. A valid comparison would be the S-Bahn tunnel in Munich, which handles 5 or 6 through routes on 20-minute headways each. They manage it under 3 conditions:
1. Subway-type cars with either 3 or 4 double-stream sliding doors and level boarding;
2. Island platforms for boarding and side platforms for alighting, encouraged by opening the right-hand doors a few seconds before the left-hand, and having no up escalators from the center platform and no down escalators to the side platforms (passengers needing to change trains exit to the left onto the center platform when the left-hand doors open).
3. DB standards of punctuality and equipment reliability which, although they have apparently deteriorated in recent years, are still far ahead of what we find in most American cities.
Of course provision of more underground tracks and platforms at NS and SS, and no Central station, would partially alleviate this, but trains would have to slow down approaching the interlockings rather than approaching stations at track speed.

Well, there'll be a study - at a cost of $2M:

http://www.bostonglobe.com/business/201 ... ost_Viewed" onclick="window.open(this.href);return false;

The article credits former governors Bill Weld and Michael Dukakis for pushing the idea. Dukakis says that the NSRL means you don't have to expand South Station. If you DO expand South Station (like Governor Baker wants), you need more train storage at Widett Circle. Boston's Mayor, however, wants to turn Widett Circle into a new neighborhood.

You need to do both. But South Station needs to be expanded more than the N-S rail link is needed. South Station and the storage yards are perfectly sized for service in the early 1980s!

Dukakis thinks you can magically run trains back and forth all day, like a subway, and don't need to store them anywhere when people are at work and aren't using the trains!

All the missed opportunities to store trains in the basement under new developments. Just have dedicated electric switching locomotives so you don't need to vent diesel fumes. Or vent the fumes. This would have worked where the Boston Herald was, and would still work at Beacon Park.

Well, hopefully the study will help them confront reality... informing government decisions is one of the main goals of a study. As much as it is $2M down-the-ropes, it really needs to be studied properly. There have been far more trivial things have gotten expensive studies, without even needing to look beyond Boston for examples.

Talking about sub-surface storage, I wonder if they could do two-level storage (surface+below) in either the BET area and/or the Southampton Street Yard/Widett Circle area. It'd be a challenge if not outright impossible to move trains into the sublevels without coming through the tunnel itself, and of course exclude diesel ops, but it'd be a way to maximize property. Of course, there'd be significant interruptions to the normal functioning of both yards and facilities in the construction process of such a thing, even if you tried to utilize areas not already occupied by facilities.

BandA wrote:You need to do both. But South Station needs to be expanded more than the N-S rail link is needed. South Station and the storage yards are perfectly sized for service in the early 1980s!

Dukakis thinks you can magically run trains back and forth all day, like a subway, and don't need to store them anywhere when people are at work and aren't using the trains!

All the missed opportunities to store trains in the basement under new developments. Just have dedicated electric switching locomotives so you don't need to vent diesel fumes. Or vent the fumes. This would have worked where the Boston Herald was, and would still work at Beacon Park.

Why would you necessarily need to store them near the terminals? Why not have most of the train storage be at layovers 25 or 30 track miles from Boston? The extra train-miles will cost something more. Using $50/train-mile as a reasonable guess at marginal operating costs and assuming that the runs to said layovers get a few hundred in fares (that's fewer than 40 riders each), that's about $2500 times maybe 35 sets (not counting ones that would go in for maintenance and those that are providing midday service already) times 260 days => about $23 million a year. By doing this, though, you can sell off the storage tracks at Widett Circle and BET (perhaps not all of BET), and every $20 or so you get from that reduces overall expenses for the T by $1.

If (value_of_BET_storage_track + value_of_Widett_Circle_storage_track + cost_of_South_Station_expansion + cost_of_North_Station_expansion - cost_of_NSRL - cost_of_layovers)/20 is greater than $23 million, then the MBTA is lighting money on fire by not building the Link.

How could you put them in layovers in the middle of the day at distant terminals and then surge them into outbound rush hour service from 4 to 6 PM? And also include time for them to get serviced (even at the distant layovers)? Not enough hours in the day or track capacity (to get them back into the city before the rush hour).

Congressman Moulton is "bumping" this thread per article in today's Globe: http://www.bostonglobe.com/business/201 ... story.html

Interesting excerpt - maybe he has credibility on the topic:

Most everyone thinks of Moulton as the decorated military veteran, but before he won his seat in 2014, the Harvard MBA grad spent a year working on a high-speed rail project in Texas. In other words, he knows a lot about trains and transportation.

I doubt working in transportation for only a year would certify real expertise.