Quote:
Originally Posted by Alon
No. Not at all. Let's start with the obvious question:
where in the world do they run mainline trains like that? Not in Japan: even lines that have dedicated tracks and no branching, like Yamanote, Chuo-Sobu, and Keihin-Tohoku, are not run like that, but instead have printed schedules. And before you talk about double-tracking, you should know that the Yokosuka Line has a single-track outer end with a double-track segment in the middle and is run to a schedule that's adhered to. Line branching dictates schedules - you can't say "a train every X minutes plus or minus 10 minutes" when people board on a branch with only a few tph. You need to make sure the merges at the junctions go smoothly, without causing bunching on one branch while the other branch has a long gap in service. But even without branching, you can't have timed intermodal or cross-platform transfers without a schedule, and you can't make passengers wait around like they have nothing better to do with their time. But even on lines that really do run every 2-2.5 minutes at rush hour, without branching or track-sharing with other trains that need to fit into precise slots, there's a schedule.
I question how much you know about Japan given your comment at the end about the Shinkansen. Yes, the Shinkansen is more punctual than the legacy system, but that's true of dedicated-track HSR in the rest of the world, too. It's not about the number of tracks (it's not "virtually 100%" double-track, but exactly 100% - HSR lines are built double-track from the start); it's about longer station spacing allowing more space to recover from delays imposed at stations and much lower tph counts than legacy lines. The Shinkansen actually executes timed overtakes at local stations, which have 2 stopping platform tracks and 2 express tracks, as is common on other HSR lines around the world. Only the busiest stations have more than 2 platform tracks per line (they have 4 or 6*), and at those the track numbers appear on the timetable. A local train pulls up to a stop, the express train behind it goes through on the express track, the local train departs the stop behind the express. It's not about investment, either - the non-grade-separated, single-track lines are also very punctual.
Or look at Switzerland. There's 7% padding to allow trains to recover from delays, which is a lot less than Amtrak and Metro-North employ judging by how early they can be sometimes and how fast they are when recovering from delays. Complex junctions sometimes have an additional track to let a late train sit without propagating delays to the other trains. None of the stupendously expensive investments that the US is proposing; there isn't a single project in Switzerland that's budgeted at $13 billion the way Gateway is, and the one that's closest, the $10 billion Gotthard Base Tunnel, is over 50 km of large-diameter double-bore tunnel with over 2 km of rock overburden at one point.
Major stations are built to allow passengers to wait on one platform if required. For example, Stuttgart21, which is by German standards extremely expensive but still better-designed than anything in the US, has each approach track leading to two platform tracks straddling the same island platform. This way, if there's an unexpected change, it can be announced at the station without requiring passengers to scramble to another platform. Secaucus, which has a similar two platform tracks per one approach track concept (as is wisest on a line with 25 peak tph), does it the wrong way: the two tracks serve different platforms, so passengers need to know in advance where to stand. This reduces its usefulness as a station where trains can be rearranged if necessary to avoid cascading schedule delays.
The Northeast Corridor is indeed a complex line, but its traffic and complexity are not unheard of in other developed countries, which cram many more passengers with better punctuality. Learn from them; don't try to run it like an on-street light rail line.
*Tokyo has 10, of which 6 are only for Tokaido and 4 only for Tohoku; all trains terminate, and there are no track connections. Shin-Osaka has 7, but it's both a terminal for Tokaido and a through-station for Tokaido trains that continue to Sanyo.
|
My point is that, in the US in 2013, we have our own unique combination of highly advanced computing skills and worse than 1950s style line routing. In addition, we will not spend the money on the right-of-way modifications needed, so putting more of the burden on flexible scheduling is a very good idea, IMO, based upon return on money invested. All one needs are arrival display boards at each station, and, announcements on the platform level to make it work. Adhering to a schedule that will not work (defined is being unable to adhere to a schedule where train X is forced to be later than train X + 1,2,3, etc) due to variations in branch traffic timing, work orders, etc) IMO, is a far worst PR problem, and, lowers net line traffic, than building flexibility say +5/-5 minutes in lines with frequencies of trains every 3 or so minutes.
In the US, where passenger rail traffic interacts with freight traffic, frequent work orders due to decades long under investment in up keep and upgrades, one, IMO, should use the tools at hand first. The high speed rail work done that has been done abroad, and, will likely never be built here on any large, interstate scale* in over the next several decades, is just a dream in 2013. For systems who have routes which cannot be precisely timed, due to work orders, sharing tracks with freights, grade crossings, street travel in the case of light rail (some heavy rail, too), single track mains, among other things, flexible, computer assisted, real time scheduling would generate the greatest increase in efficientcies. This is the US reality (even the California HSR is going to share travel over track with road crossings as it travels up the shared route of the Baby Bullets into San Francisco).
I know more than you suspect about the skinkansen.
The key to the skinkansen being able to work along it's busiest spine- between Tokyo and Fukuoka- is that trains can pass one another via 4 track stations. In the majority of such 4 track stations on two line mains, for safety and speed purposes, trains do not have to run through next to a platform. Such platforms should only be used when express train needs to stop so that passengers need to feed or by fed by, local service (the situation in Germany likely has to do with the station being in a rather large metro area and has far more than two extra tracks to serve a 2 track main line.I have been talking about stations along routes- and of course, I am not talking about Nagoya, Kobe, Osaka, etc). This enables various express levels to be safely created in addition to local train service. This platform design has been used in Japan for years, and, used in many instances, prior to the Shinkansen, even by private railroads such as the Odakyu Railroad (which I occassionally used to ride going to school.)
In the best designs, IMO, the 4 track station on a 2 track main should only use platforms for tracks 1 and 2 when expresses stop and passengers can transfer to local trains (or due to the station itself being a high passenger volume traffic generator.)**
YT via Skyscrapercity.com-which has the best Japan rail transportation forums filled with innumerable examples of high speed shinkansen run t
hroughs, most videos being taken of trains that run on tracks without abutting platforms (from JAPAN | Highspeed rail...)
http://www.youtube.com/watch?feature...&v=EeLpc5_2xOY
Of course the Spanish, Italians, Japanese, Koreans, Germans, French, Republic of China, People's Republic of China, and, others now building dedicated HSR (Turkey, Saudia Arabia, to name two), have dedicated ROWS, many (or most) with run through stations. But, their timing precision is nowhere near that which the Japanese run. We're talking +/- seconds in Japan, and +/-1 or two minutes everywhere else (after the Japanese, I would put the Swiss and the Germans for the best real time schedule adherence).
*The Washington - NY corridor was completed by 1943 or 1944, built by the Pennsylvania railroad. At that time, the corridor was the most advanced passenger corridor in the world, IMO. The Congressional trains in the early 1950s and train 153, made the trip from Penn Station to Washington in 3 hours and 35 minutes with 10 -12 car trains. Today, the fastest Acela makes the trip in about 2 hours and 45 minutes. Pretty close to speeds in 2013.
**The major argument I hear and read about run through stations is the need to build over or under track access across the tracks, and, the necessary elevators due to physically challenged passenger users.
My ties to Japan are deep. My father is buried there.