Bypassing Hartford strikes me as unacceptable, given the size of that metro area. Bypassing Providence would not sit well with Rhode Island’s Congressional delegation, which might be small, but still has two senators.

On the other hand, building a new RoW won’t be cheap, and might meet some fierce local opposition.

I’m making a video about the old Pennsyvinia Catenary how it needs to be replaced and up graded south to Richmond VA. The idea would have the catenary taking on the best parts of tensioned catenary systems but still having the new catenary masts going into Richmond retaining their classic H beam Pennsyvinia Railroad shape.

At the same time, you’re using curved rights of way in Connecticut, where trains can’t run very fast; the freeways only give a reliable right of way east of New Haven. If there’s money for a tunnel between Long Island and New England, put it further east, and run trains on the LIRR mainline; remember that the LIRR was first built as a New York-Boston intercity line with a ferry connection. You gain precisely nothing from the Jamaica-Stamford tunnel – for service to JFK, run some trains DC-Philly-NY-JFK, since there is going to be more demand south of New York than north of New York.

Start with the corridor from Philadelphia to Boston. Yonah would disagree with me on this, but I’ll start by pointing out that five of North America’s busiest airports (Philly, Newark, JFK, La Guardia, Logan)–and at least four of its busiest international gateways–are squarely within the footprint. Those airports would still be immense trip generators, but stop thinking about that for the immediate moment.

If this corridor were in Germany, France or Spain, chances are very good that there would already planning for a new, passenger-only trunk line, to be served by EMU trainsets. Lightweight, high-speed EMU sets can handle grades as steep as expressways–let’s say 4% as a design standard. In those three countries, chances are very good that the new alignment would be adjacent to an existing expressway. That’s what Germany did when it (finally) built the ICE line between Frankfurt and Cologne. Start at Philly 30th Street, in a deep tunnel, and cross under the Delaware to the Jersey Turnpike r/w. Follow it to the south end of Newark Airport, stop in front of the terminal complex, then join the NEC r/w to Newark Penn. East of there, the line crosses in a new Hudson tunnel to an expanded NYC Penn Station. These plans aren’t new, so this is no conceptual problem. Remain at deep level, cross Manhattan and the East River, and head for Jamaica. Airtrack already connects from Jamaica to JFK, and could easily be extended to La Guardia. The new line continues east, into western Suffolk County, and stops again. Turn north and cross Long Island Sound, either on bridge or in tunnel. Hit land at Stamford, and provide for joining main line there, or continuing in new r/w. New r/w would continue in tunnel, then surface when joining r/w of Merritt Parkway. Bypass Bridgeport and remain in Merritt r/w, then enter New Haven. This would allow for connections to Hartford, Springfield and points north. East of downtown New Haven, rejoin the 95 r/w way and follow it at least as far as the Rhode Island/ CT border. Follow 95 r/w into Mass, then possibly follow in or beneath Fairmount Line r/w to South Station. A connection is planned in downtown Boston, to North Station, which will allow continuation to New England, Quebec, and the Maritimes.

Yonah has previously suggested center-to-center focus. In this corridor, there is a great deal of international air traffic generated in almost every corner. The existing corridor’s service is pitiful, and even with sprawl and American driving habits, there are likely at least a few million travelers per year who would gladly park on Long Island and have through-ticketing to an international airport. This action allows the existing line to focus on slightly more local service, meaning a narrower speed band, which also means greater track utilization. From what I’ve described, let’s say that the Philly-Boston service lines up like this: Philly 30th St – New Brunswick/East Brunswick/ Edison – Newark Airport – Newark Penn – NYC Penn – NYC Jamaica – (somewhere in western Suffolk) – Stamford – New Haven – Providence – Boston South. I mentioned lightweight EMU sets easily capable of a 4% gradient. That means that, if starting just above sea level and required to cross a 150 foot vertical clearance, the line would need less than a mile at either end to go over, or less to go under (and I’m stating here that a crossing of Long Island Sound has its own issues, but I am prepared to discuss that too).

How much would all this cost? It’d damn sure be more than $10.5 B. There are other business questions that matter more, though. A true HSR line in the Northeast wouldn’t just knock out the Boston-NYC shuttles. It could knock out almost all air traffic originating within at least 50 miles of the corridor. That’s not disaster for the airlines–it would open up dozens of gates for more lucrative long-distance flights, while giving each airport a far broader market. It would also remove a lot of the pressure for multibillion-dollar airport improvements, which means that airports could stay busy and generate more money with modal interchange than with expanded runways and gates. Residents would be happy at not only the reduction of noise (which could allow more 24-hour operation), but also at greatly improved property values (which means more tax revenue). For cities, greatly reduced demand for flight paths would mean removal of some building height limits, again generating more tax revenues.

Even considering this segment–Philly to Boston–there are some stunning changes. There’s no reason that a train from, say, Portland or Manchester would need to stop anywhere en route to Baltimore, DC or Richmond. If the corridor had a design allowance for continuous 225 mph service, then an hourly or twice-hourly Philly-Boston train (ten-car bilevel EMU consist) could not only carry over 1000 passengers, but it could easily run the 320-ish mile distance in less than two hours. Longer trainsets are not impossible. A 3-car unit from Portland, another 3-car unit from Manchester, coupled with an 8-car express at Boston, could make lower-speed stops and then travel at top speed without overcrowding the line. Planes can’t be coupled like trains.

There has been mention of discount airlines and the inability to compete with them. Discount airlines operate on the assumption that they will continue to benefit from an inequitable system of subsidies, which in the US favors road and air over rail and water. Given the current Federal budget deficit, there’s no justification to continue giving road and air such disproportionate support. There is also an established practice in Europe of airlines (Swiss, Lufthansa, Air France) either contracting for rail service, or even (Virgin) operating it themselves. This satisfies the Republican desire for private sector involvement, which has the potential to draw more political support.

In the late 70′s, the Carter administration proposed a massive electrification program of mainline railroads, to be paid for through royalties per ton-mile. It’s a form of tolling, and there’s no reason it couldn’t work on the New NEC. Airlines purchase gate access, but every one of the airports in this corridor is publicly-owned. The taxpayer deserves the best return on gate value, not what’s most convenient for a cherry-picking airline.

In theory, if trains were able to run 5 minutes apart, that results in 12 slots per direction per hour. A 15-minute service would still leave 8 open slots per hour. Again, in theory, there’s no reason why the new line couldn’t run a 15-minutes service, connecting with a similar but slower 15-minute service on the existing line. This would rely on transfers more than interline operation, but still offers the potential to be more cost- and time-effective than anything in the status quo. No-one (with the possible exception of discount bus operators) has anything to lose with this concept.

Actually, the train lengths are limited by platforms and station setup. For “intercity” services, the limit is 400 m, for commuter services, it is 300 m (or less).

Another reason is that the network is running at its capacity (or even above) on its businest segments, and it is not possible to add more trains.

The maximum speed is indeed 200 km/h, and will most likely remain at that (although they are working at 250 km/h for the ETR 610 through the Lötschberg base tunnel (and in a few years through the Gotthard base tunnel).

The higher axle load is indeed a crucial issue, and most likely the reason why there are not that many double-deck high speed trains around besides the TGV 2N and the JR East Max (whereas the latter can profit from the advantage of less stringent crashworthyness requirements because it is operating in a closed protected system, and can therefore be built lighter than “normal”.

Interesting… then why is the Swiss Federal Railway using and ordering mainy double-deck rolling stock? Not only for commuter services, but also for (Swiss) intercity services (travel times corresponding about to the Northeast corridor).

Honestly, I don’t know. A few guesses:
- Swiss regulations restrict train length, often for stupid reasons, like a bit overflow in computerized train control.
- Swiss rail has a medium top speed, about 200-250 km/h, which is within the capabilities of double-deckers. The top speed advantage of single-deck trains is therefore less important.

Also: yes, double-deck trains weigh less per seat, but they weigh more per axle, and that creates maintenance problems. Tellingly, today’s HSR companies focus on single-deck equipment: the AGV, the Velaro, Fastech, Talgos, CRH-3.