MISC | Yield Management in Railways

[Suburbanist]

Originally Posted by trainrover

The French which, eh? Eliminating the margin of error shall probably proove to bite no matter whomever really HARD in the ass.

You don't measure margin of error of fast trains with full in-cab signaling based on time, but rather based on braking distance, wither using block signals with PTC or the more modern dynamic signaling.

A high-speed train can come to a halt from 300km/h in less than 70 seconds under emergency stop protocols with safety (though a bit of discomfort for passengers). You don't need more than that as a safety margin, you don't want to have the train ddriver deciding whether to stop or not, you want a system that detects a faulty/broken/crashed train or railway with intrusion of foreign object and then suddenly deploys emergency stop activation in your trains in milliseconds.

[trainrover]

So many clawbacks being reported are enough to make wonder whether redundancy is no longer devised. Although I haven't done the arithmetic, a 6' headway at great speed strikes me as pushing the braking distance should operations go awry, let alone a 1' one.

[K_]

Originally Posted by trainrover

So many clawbacks being reported are enough to make wonder whether redundancy is no longer devised. Although I haven't done the arithmetic, a 6' headway at great speed strikes me as pushing the braking distance should operations go awry, let alone a 1' one.

Well, maybe it's time for some arithmetic then.

3' at 360kph means 18km between trains. The normal stopping distance at that speed is about 9 km. So there is still quite a bit of safety margin there.

If you want to reduce headways to 2', you reduce train separation to 12km. That means that your safety margin gets reduced. There are two things you now need to do:
- Shorten your stopping distance. That can be done using eddy current brakes.
- Improve the signalling. That is what ETCS is for. 2' headways at 200kph are already in daily use. The LGV from Paris to Lyon uses 4 minutes as minimum headway, using a system that was originally designed for 5', and will go to 3' soon. It might even go to 2' as this is probably going to be a lot cheaper than building an additional LGV...

So 6' is being very prudent...

[chornedsnorkack]

Note that off-peak the headways are easy to keep. The xx:30 express train leaves Taibei 6 minutes before the xx:36 milk train, and never approaches. Since it is 24 minutes faster than the xx:00 milk train, it reaches Zuoying 6 minutes after the xx:00 milk train.

[makita09]

Originally Posted by trainrover

So many clawbacks being reported are enough to make wonder whether redundancy is no longer devised. Although I haven't done the arithmetic, a 6' headway at great speed strikes me as pushing the braking distance should operations go awry, let alone a 1' one.

You have the whole concept back to front. Braking distance is a distance in space. Headway is a distance in time. They are not the same thing. Signalling systems separate trains by the necessary distance in space, not time. The headway is something that is an output after the safety of the system has been guaranteed.

Breaking characteristics > Signal block length > then at this point we can go back a calculate minimum headway.

Rather than sticking with a misapprehension, there are lots of sources on the web help understand how signalling systems function.

[Silver Swordsman]

Implementing a new system will obviously cost manpower, time and money. Right now, in Taiwan, there simply isn't a need to ramp up frequency that much. Not saying that it can't, but right now, I don't think it's a feasible thing to do business-wise. Just because you can doesn't mean you should. However, there are recent reports that THSR plans to ramp up the number of services to over 200 per day... so I guess that will definitely mean higher frequencies. (200 trains north and south over 18hrs)

[chornedsnorkack]

Originally Posted by Silver Swordsman

Implementing a new system will obviously cost manpower, time and money. Right now, in Taiwan, there simply isn't a need to ramp up frequency that much. Not saying that it can't, but right now, I don't think it's a feasible thing to do business-wise.

The THSR trainsets are now all (30 existing, 4 under construction) 12 cars each.

Would it be useful to add smaller trains, e. g. 8 car trains, to THSR so as to increase frequencies without excessive cost?

[trainrover]

Good, sounds better I imagined the distance between HSTs running at full speed much less than 12Km.

[K_]

Originally Posted by chornedsnorkack

Note that off-peak the headways are easy to keep. The xx:30 express train leaves Taibei 6 minutes before the xx:36 milk train, and never approaches. Since it is 24 minutes faster than the xx:00 milk train, it reaches Zuoying 6 minutes after the xx:00 milk train.

In this case it makes sense. However they could use the extra 6 minutes to maybe add in another stop on the milk train. I don't know if the times of other trains in Taibei and Zuoying are coordinated with those on the high speed line.

On the Swiss network the departure times of trains are coordinated at the major hubs, and for that reason trains are run as closely as possible on the high speed line from Bern to Olten (with 2' headways), even though with six trains per hour they could spread them out more.

[trainrover]

11 Dec 2012 :

Proposed railway service penalties unnnecessary, CN Rail, CP Rail chiefs argue

[Silver Swordsman]

Originally Posted by chornedsnorkack

The THSR trainsets are now all (30 existing, 4 under construction) 12 cars each.

Would it be useful to add smaller trains, e. g. 8 car trains, to THSR so as to increase frequencies without excessive cost?

I've thought about that as well, but it seems that because the 700T is an EMU, rearranging trainsets is difficult. If they want 8car consists, they're going to need a completely new different trainset.

I think, for now, Taiwanese peeps still see trains as "trains" with a timetable, instead of a fast metro service (with insane frequency.)

[k.k.jetcar]

It's also helpful to have consistent train lengths when the occasional cancelled service (due to breakdown, accident, etc.) needs to be replaced with a spare trainset- you can provide seating for all passengers with reserved tickets who were scheduled to be on the cancelled service. This is one reason why JR Central only runs 16 car trains on the Tokaido Shinkansen.