leviramsey wrote:Excellent analysis. Thanks for this insight.east point wrote:The slow sections can be upgraded to 60 - 80 MPH and you get more bang for the buck instead of spending first on higher speeds. Going from 60 - 80 MPH takes much more track miles than 30 - 60 . 80 MPH/ You get more time reduction with the 30 - 60 increase. That means much more money !s available to do more miles of eliminating the slow sections. That is what NC DOT is doing.Higher max speeds are sexy and do have a marketing benefit. But upgrading the lower speed segments has far more effect.
That is how you increase the average speed better than ... pushing first to increase top speed. Know It is not as sexy but -----------
30 miles of 100 mph and 10 miles of 30 mph => 38 minutes (63 mph average speed)
30 miles of 200 mph and 10 miles of 30 mph => 29 minutes (83 mph average speed)
30 miles of 100 mph and 10 miles of 60 mph => 28 minutes (86 mph average speed)
( #harmonicmeans )
The analysis here is the same as for cars, so people thinking in car terms can understand up to here.
But for a train (or a plane), headway matters. If the train arrives at 6am [but] you wanted to get there at or before 12 noon, that's not that different from 6 hours of 0 mph.
The airlines have figured this out over the last 30 or so years (as they've converted routes from 1 flight/day with as big a plane as possible to several flights/day with smaller planes).
First approximation: desired arrival time at a destination is uniformly distributed over 7am to 7pm local time. So a 6am arrival is an expected wait time of 7 hours, 7am is 6 hours, 8am is a 7 hour wait time, 9am is a 8 hour wait time -- and 7pm is an 18 hour wait time (due to the effect of an overnight wait). Add 9 hours of 0 mph to travel and even if it takes the train only 1 hour to travel 150 miles, the effective speed is 15 mph. Add a second frequency, and you cut the wait time by half, which nearly doubles the speed for the shorter trips (the 150 mile trip with 1 hour of travel time goes up to 27 mph, which is faster than a teleporting ansible that works once a day).
Focus on acquiring rolling stock and paying for smaller upgrades (signals, passing sidings, etc.) to enable n+1 (at least where n is less than 10 or so; at n>10, the extra frequency might start to get beaten by moving the train faster) frequencies where it makes sense to. Maybe try to get a few miles here or there upgraded to 79 so you can say "travel at up to 79 mph" for marketing, but the real gains come from the extra frequency. At the same time, don't be afraid to increase fares to capture the increased consumer surplus.
As I noted, airlines have figured this out. Transit supporters who don't actually use transit (e.g. it's a politically symbolic gesture), which includes most decision makers at transit agencies in this country, don't get this and instead fixate on (if we're lucky) making the trains run faster. Amtrak having an airline exec for a CEO who presumably gets this is a huge step forward.
Making this knowledge part of the remedy is fraught, however. Almost every Amtrak train off the NEC is a tenant of a busy landlord, and the freights hate the interference in their running caused by Amtrak's frequencies.
Meanwhile, the Stimulus invested Billions in a handful of corridors, seeking to raise speeds and add frequencies. Some of this investment was no doubt the cost of getting the host railroads to agree to carry more Amtrak frequencies.
The Cascades service was to showcase the dual goals. A TV commentator was aghast and repeated, "They spent almost a Billion to save 10 minutes!" Well, not exactly. For the Billion they got two more frequencies each way. With a promise to increase On Time Performance to 88%. As well as new stations and two new trains. And yes, to save 10 minutes. Those two new frequencies -- early morning and end of day -- make possible same-day go-and-return business trips between Seattle and Portland. Before the new frequencies, riders got barely half a day at their destination. So the new frequencies, doubling the usable hours for these riders, as you suggest, had much the same effect of doubling the speed of the trains.
About 10 years ago, the Lincoln service St Louis-Chicago added two frequencies (plus the Texas Eagle) to its runs, and soon more than doubled ridership. Since then another Billion of Stimulus funds has been invested to raise speeds, up to 110-mph for some fairly long stretches. Officials promise trips end to end will be shortened by "about an hour" over this year and next. They haven't said if frequencies will be increased at all, so it will be interesting to see the ridership figures.
The Wolverines service Detroit-Chicago will also be reducing trip times, again by roughly 40 or 50 minutes iirc, but with no added frequencies announced so far.
The Piedmont service Raleigh-Charlotte, using Stimulus funding, will reduce trip times by half an hour or more iirc. The route has two Piedmonts plus the Carolinian, with promises now to add another frequency later this year and yet another next year. NC DOT ambitiously forecasts Piedmont ridership will rise from about 150,000 a year now to 250,000 with the 4th frequency and to more than 350,000 after the 5th frequency starts up, before reaching almost 450,000 a year after the full package of improvements (new stations at Raleigh and Charlotte) are finally in place. In other words, they expect to go from 3 trains a day to 5 a day and see triple the ridership of the Piedmonts.
https://www.ncleg.net/documentsites/com ... vision.pdf" onclick="window.open(this.href);return false;
(The North Carolina data argues for more frequencies in another way. At one Piedmont run a day, the "support" required for each roundtrip rider was $127. With two Piedmonts, each passenger needed a subsidy of only $67! With three Piedmonts, the expected subsidy will drop to $49, and with four Piedmonts, it will be only $39.)
From New Haven to Hartford, after "sufficient double-tracking and infrastructure upgrades", line capacity will be increased to allow 9 Amtrak trains a day up from 6 a day, along with a number of Connecticut commuter trains on the shared route. (Time savings are not worth mentioning until after a "phase 2" takes 11 minutes out of the 90 minute schedule.)
In New York, a lot of money is being spent along the Hudson River and around Albany to eliminate congestion, improve On Time Performance, and presumably speed up some schedules. But nothing here indicates more frequencies any time soon.
Of course, more of these projects might boast increased frequencies but for Amtrak's perpetual equipment shortage, made worse by Nippon Sharyo's face plant. So the Stimulus corridors could look better after the Siemens cars get delivered.
But Amtrak needs a lot more corridors with heavy frequencies. Where and how can that happen?
