Why do LD trains not handle intra-NEC traffic?

[David Benton]

Sticking points end up being things like slots on the Long Bridge and cost of a base to turn it at (I vote Lynchburg, actually, as if it were a Crescent stub even though we are saying it is a Cardinal)

Charlottesville is as far as it can go , in order to turn back the same day . It has to be back in NYP for the following days Cardinal.

[electricron]

What if any time a passenger booked a ticket from south of DC to an NEC station south of Newark, a seat would become available from that station to NYP? That way, it never takes seats away from LD passengers, but lets them sell space that would otherwise go unsold.

And what would happen if the original long distance passenger getting off before Newark cancels after Amtrak sold the seat to a regional passenger? Now there's one less seat available for a long distance passenger to buy getting off at New York City because that seat was sold for a regional passenger. The idea that long distance passengers buying seats months in advance don't change their plans is foolish.
Whereas I agree that there will be seats becoming available as passengers alight the long distance train, and Amtrak should be able to refill the seats, the question to be asked is when should Amtrak be able to resell them; a day, a week, a month, or a year before? The last thing I want is forcing a long distance passenger to transfer trains In D.C. because there isn't a seat available to go all the way to New York City because some seats have been sold to NEC regional passengers.

[leviramsey]

Under the current fare structure, it's not a huge problem. The LD trains running on the NEC are mostly only charging riders to ride to WAS. For MIA-WAS (Star or Meteor), leaving March 9, the fares are $114/$143/$310 (Saver/Value/Flexible); MIA-PHL is $116/$145/$316; MIA-NYP is $118/$147/$320. On the Regionals bracketing the Star's scheduled arrival in WAS on the 10th, WAS-NYP is $49/$88/$173.

Selling the seat from WAS-NYP is a gamble by Amtrak that a MIA-WAS passenger doesn't cancel and that no replacement MIA-WAS passenger can be found (under simplifying assumptions that an available MIA-NYP rider can always be found and that there are no stops between MIA and WAS where a rider can get off the train; if those simplifying assumptions don't hold, then the analysis points even more to "sell the seat" by reducing the risk and/or increasing the reward). As with all gambles, whether the bet should be taken (by selling the WAS-NYP seat) is a question of the risk and reward. The risk is the MIA-NYP fare ($118/$147/$320). The reward is the sum of the MIA-WAS and WAS-NYP fares (if WAS-NYP is at the Regional fares, $163/$231/$483, ignoring such possibilities as Saver MIA-WAS -> Flexible WAS-NYP). As long as the MIA-WAS passenger on Saver cancels 27% or less of the time, Amtrak's expected value is positive*. For Value passengers, Amtrak is still ahead even if they lose the bet 36% of the time, and a 33% cancellation rate among Flexible fares is still break-even for Amtrak. For any loss rate deemed acceptable, a fare for WAS-NYP can be defined which has positive EV for Amtrak (the limiting case would be charging the WAS-NYP rider the fare for MIA-NYP: then even if every MIA-WAS rider cancelled, Amtrak would be no worse for wear!). Amtrak presumably has data on propensity for reservations to be cancelled (if they don't, paying a team of software engineers and statisticians a million bucks a year to get that data may have the single greatest return on investment of anything Amtrak could do right now), which could easily be used to adjust offered fares for LD trains on the NEC.

*: Let's say that of a 100 seats sold on the Star WAS-NYP to fill seats vacated by MIA-WAS passengers, we have 27 MIA-WAS passengers cancel (all 200 reservations were Saver fares). Then Amtrak collects $13,222 in fares. If instead we had 73 MIA-WAS passengers and 27 MIA-NYP passengers booking seats originally taken by MIA-WAS and not sold from WAS-NYP, Amtrak would collect $11,508 in fares.

[ThirdRail7]

Well, we know southbound trains 89, 79, 51, 91,19 and 97 carry local travel over the NEC. Northbound, train 90 carries local passengers every day, while 80 carries local travel on Thursdays, Fridays, and Sundays.

For the next few weeks, train 50 and train 92 will get into the act.

92 will only take corridor passengers on Sundays. This may work out since 92's Saturday departures from MIA seems to have performed reasonably well over the route.

The average delay is 35 minutes since January and most of the delays occurred between April and September when you have trackwork and heat restrictions.

You have to keep a sharp watch on 50 though.

[Tadman]

Are the LD trains on the NEC late more often than the Regionals? Amtrak might not want to sell seats on trains are more likely to be late and make passengers angry when other trains are available.

That's a statistical certainty. If you have 500 miles of pre-NEC operation, the chance of being late compared to a train originating at WAS is very large.

I got boned by this theory last week when catching a Regional WAS-BWI. I forgot it was a Richmond origination, ran late, and I had to jump on MARC to ensure I got into BWI and through TSA on time. This is why I advocate that Richmond trains should be shuttles, meaning no motor change, no run-through. If one train is late, the corresponding train is not held (and inconvenience even more passengers), the late passengers from the first train have to catch the next train.

[Arborwayfan]

If some or all of the LD trains are too unreliable for Amtrak to want to sell reservations (ie too many pax would be delayed and unhappy), why not sell space on all the northbound LD trains on the corridor, but only at the last minute (once it's pretty clear when the train will actually pull in)? Basically, if someone wants a ticket on the next train from Baltimore to NYP and the Crescent has left DC and will be the next train from Baltimore to New York and has a seat for that person, sell it to them. The passenger has no reason to care if the train is on time, as long it happens to be passing through when the passenger wants to travel. Amtrak could also make station loudspeaker announcements to offer seats on passing LD trains to pax with reservations on the next regional train (and pax with unreserved tix, if they even still exist), and then when the conductor scans their tickets, open up their seats for slightly later walkup sales. Airlines often offer people who show up a little early an earlier flight if there are empty seat on an earlier flight and the airline can sell (or has oversold) seats on the passenger's original flight.

Short version: use the LD trains as somewhat unpredictable extra service for walkup pax.

[mtuandrew]

Tad: good idea to have intra-Virginia Regionals (that continue to DC.) Encourages transfers to Acela too.

[ryanch]

>Short version: use the LD trains as somewhat unpredictable extra service for walkup pax.

This! They could even require same-day non-refundable purchases.

[jonnhrr]

This is why I advocate that Richmond trains should be shuttles, meaning no motor change, no run-through. If one train is late, the corresponding train is not held (and inconvenience even more passengers), the late passengers from the first train have to catch the next train.

I guess as long as the majority of passengers are going to Washington or points short of DC. I don't know what the numbers are for these runs. I imagine it would be a detriment if a large percentage of riders were continuing on past DC and now it becomes a 2 seat ride.

[Tadman]

Here's the thing: On the vaunted European HSR/HST systems, 2-seat rides are normal.

Also, the ridership south of DC is less than north of DC. Now you can have a ACS motor with 8 cars pull up to a platform, cross over to a P42 with 4 cars, and your seat-miles better match your passenger miles, which helps the finances.

Finally, we've heard about some train storage space issues at Richmond and Tidewater locations. If you cut back the empty cars heading south and right-size the trains, perhaps there are additional train storage slots that open up.

[mtuandrew]

Here's the thing: On the vaunted European HSR/HST systems, 2-seat rides are normal.

Also, the ridership south of DC is less than north of DC. Now you can have a ACS motor with 8 cars pull up to a platform, cross over to a P42 with 4 cars, and your seat-miles better match your passenger miles, which helps the finances.

Finally, we've heard about some train storage space issues at Richmond and Tidewater locations. If you cut back the empty cars heading south and right-size the trains, perhaps there are additional train storage slots that open up.

That also opens up possibilities for sections that share a Long Bridge slot: four cars to NPN and four to NFK on the same train leaving WAS, splitting and combining at RVM. Maybe four cars to RKE and four to DAN (extended to CLT if North Carolina finds the money.)

[DutchRailnut]

in short, northbound the trains are never on time .
Southbound they do not want extra passengers filling up the toilets before long part of trip.
though enough to keep them shitters representable.