Hacker News Reader

Comment by magila | Hacker News Reader

> since those who want to go direct between SF and LA will always have a direct flight as an option

Therein lies the problem. The cost of the line can't be justified unless it provides a viable alternative to flying between SF and LA, because that's the only conceivable way to get enough ridership. The problem is the distance between SF and LA means being competitive with flying in terms of travel time is incredibly difficult. To achieve parity the trains will need to average around 220 MPH, which is incredibly optimistic. Existing high speed rail lines rarely run at 200 MPH even momentarily due to high operating costs.

All this means that most trains will have to run nonstop between SF and LA to even have a hope of competing with flying. If the trains fail to draw passengers away from the airlines, ridership will be far too low to sustain the line without absurdly high subsidies.

View on HN

> The cost of the line can't be justified unless it provides a viable alternative to flying between SF and LA,

I have no trouble imagining that the train can compete with flying between SF and LA. Even if the train has to stop in several backwater towns along the way.

The shuttle flights between SFO/SJC and LAX are completely unreliable. When I was doing the SF-LA haul regularly, I'd be delayed (delayed!) by 6+ hours at least once per month. That's longer than it takes to drive. Do you really think the high speed train will take that long to travel between LA and the Bay Area?

So what if the minimum time of the train is a little longer than a flight? The expected time enroute is still much less.

That assumption is: plane with current on-time arrival percentage vs train with mythical 100% on-time arrival.

There are a whole host of additional problems that affect rail travel that you never have to think about with air-travel.

Currently, Amtrak OTA is 80% - so that means 1/5 of your trips will experience a delay of some sort.

Amtrak doesn't own the right of way for most of their routes.

Where right-of-way is shared with freight, freight has priority.

That seems rather backwards. Passengers are the more time-sensitive cargo than freight.

Presumption would be it's much harder to stop/start a freight train vs. a passenger train.

Not to the freight line companies owning the networks (Apparently there is a law saying that Amtrack has priority, but enforcing it has as of now just lead to years of lawsuits, without useful results)

Agreed. Except that it's the freight railways that generally own the trackage in most (all?) places where it's shared with Amtrak. And the freight railways generate significantly more revenue from that trackage (about $75B annually vs $3B).

There's also the small detail that it's the freight companies that actually own the tracks in most cases so they make the rules.

Sure, but the freight companies own the track, so they prioritize their own trains.

Not if you look at the money tied up in the cargo, vs the fares refunded.

This is the problem with the stupid American capitalist system. You have these sorts of problems.

If it were all controlled by the Government, like it is in other countries, nobody would need to have 'priority'. You just schedule the trains properly.

Sounds great Benito.

This is presumably a reference to the famous Mussolini claim about having made the trains run on time[0].

[0] http://www.snopes.com/history/govern/trains.asp

How silly you are, to expect government to be responsible for coordinating things. The only reasonable purpose for government is to hand out candy to friends, relatives, and people who can keep you in government.

This is patently false - Amtrak has the right of way by law.

https://www.law.cornell.edu/uscode/text/49/24308

"intercity and commuter rail passenger transportation provided by or for Amtrak has preference over freight transportation"

What qualifies as commuter rail? I think most of Amtrak's cross-country tracks would probably not be considered intercity or commuter lines.

Amtrak is the contracted operator for a number of commuter rail services, such as MARC. Not sure if that counts as "by or for Amtrak".

The real issue is that often operationally, a passenger train has to stop for a freight train. If you have a 1000' siding that you have to have a 6 car passenger train pass a 30 car freight train at, the passenger train is going to have to stop for the freight train. That doesn't even take into consideration other factors like how freight dispatchers don't like to stop hazmat/"hot" trains for safety reasons.

According to this, the law is not enforced: http://www.sfgate.com/opinion/article/Amtrak-has-priority-ov...

You clearly have never taken Amtrak over distances. I have taken Amtrak numerous times Emeryville - Denver.

The reality is:

1. if the siding is only 1km long - the Amtrak train is taking the siding as the freight train will not fit on it.

2. if the dispatcher decides the freight train goes ahead of Amtrak -- there is no lawyer around to argue the point.

3. if there is a long section of single tracking, with freight backed up the opposite direction, the dispatcher allows the all the same direction freights to move. ( 2+ hour delays when that happens )

4. If there is enough of a delay and a freight train crew in front goes outlaw - then that train stops until a replacement crew is brought in... which usually results in the Amtrak crew going outlaw as well.

It has gotten a lot, lot better and the scenery is beautiful - but please don't kid yourself about the realities on the ground. Amtrak has to stay in the good graces of the railroad companies especially with a hostile congress.

I assumed that the Northeast Corridor (which I believe is the only place where Amtrak actually makes any money) would be much different from other routes but that actually isn't the case. Acela is better than the long distance routes but not by a huge amount.

Of course OTA % doesn't account for the size of the miss and I've heard plenty of stories of long distance trains sitting for hours while multi-mile long freights go by.

Some trains in the northeast corridor also get bounced by local commuter trains that have priority on the tracks.

Several times I've been on the Downeaster headed to Boston and gotten shunted onto sidings, to the point that we got an hour behind schedule.

"The shuttle flights between SFO/SJC and LAX are completely unreliable. When I was doing the SF-LA haul regularly, I'd be delayed (delayed!) by 6+ hours at least once per month."

I'm curious how that happened to you - I used to fly SFO-LAX from time to time, and was surprised that they didn't even really worry about me missing my flight - they just told me to catch the next one in line. United runs 27 flights daily from SFO to LAX, and presumably have some extra planes they can inject into the system if needed... (SFO is a United Hub)

If it rains at SFO or there is fog they have to close 2 of the 4 runways and now you got massive delays. I've had a flight delayed from 8pm to 1:30am last year. Having lots of flights going the same way doesn't help, if the flights cannot take off.

And what do you do if there's a suicide (unfortunately) on the HSR? The Caltrain that runs down the Peninsula from SF to SJ has a pretty high suicide rate; high enough, that there are people posted to look for potential suiciders.

Part of the cost is improving the peninsula rail corridor for full grade separation, along with the suicide prevention improvements Caltrain is already performing on the corridor. When it's built, there should be a lot fewer places for suicide and a lot tighter control.

>United runs 27 flights daily from SFO to LAX

That's just United. Southwest (and I'm pretty sure Jetblue and Virgin) has numerous flights to all airports in the greater LA area: https://www.southwest.com/thmpg/flights-from-SFO-to-LAX.html

> I'm curious how that happened to you

Mostly a combination of needing to travel at particular times to particular airports.

The last two flights from SJC to LAX each evening seemed to be the most often delayed. You can't exactly hop on the next flight if it doesn't leave until the following morning.

Morning flights to SFO are often delayed due to weather, and fighting peninsula traffic to get to work sucks. BUR-SJC on Southwest would have been optimal from a flying perspective, but LAX was a shorter drive.

"I have no trouble imagining that the train can compete with flying between SF and LA. Even if the train has to stop in several backwater towns along the way."

Especially if the train stops in "several backwater towns" (Fresno and Bakersfield ain't particularly glamorous, but I'd hardly call them "backwaters"; I digress...), this will automatically be more useful than just an ordinary flight between SFO and LAX for that reason alone.

In addition, the time required to travel from each airport to where you actually want to be can be much less, since train stations can be built much closer to major metropolitan areas. Plus, weather is far less of a problem; when the planes at SFO are grounded due to fog, the trains will still run on time.

For me, the line is effectively useless without Phase 2 (since that connects to Sacramento). With Phase 2 added, you'd have a high-speed connection to all four of California's major cities, significantly improving the already-pretty-great utility of the line.

That ignores the fact that there are already 6 airports in Southern California and 3 (5 if you include Central Valley airports) in Northern California. Do you really think getting to Disneyland or Pasadena will be faster if you take a train to downtown LA than if you flew to SNA or BUR or ONT?

And that ignores the fact that not all of those airports are going to have the same airline coverage as the major ones (namely: SFO and LAX) - and even if they do, certainly not with a significant selection of flights, and certainly not at readily-affordable rates. To the average traveler, they might as well not exist; even Amtrak or Greyhound are more viable.

Also, again: weather is still a major factor for any airport.

Absolutely false - for NorCal-SoCal trips, all of the aforementioned airports have plentiful flights at similar cost to SFO and LAX. And weather primarily affects SFO - OAK and SJC rarely have the same problems as SFO.

NYC to Boston is 230 miles, about 3.5 hours by "fast" train. It's about an hour flight. A lot of people still take the train.

If you can get the trains to take you directly into the city center, it can often be a better option.

That theory doesn't really work for LA because it's so spread out that you are very likely going to need to take a car to your final destination regardless of which mode of transportation you use to get into the city.

Depends what part of LA your in, but in most cases your absolutely right. Most areas are designed with cars being the prime transport, with sidewalks halfheartedly tacked onto high speed roads. I'd never walk along that by choice!

If you can get the trains to take you directly into the city center, it can often be a better option.

That's because Boston and NYC are both navigable without a car. Only one side of the SF-LA link can claim to be close to that ideal.

But the time and hassle of airport parking and security and delays might be able to keep a 3.5 hour train ride viable even in CA. That would require a train system that allowed passengers to step through a quick scanner with no line and onto the train without queueing or carefully organized boarding. Lots of train systems used to do that in the USA but the bureaucrats don't like passengers freely walking around platforms by heavy machinery or directing themselves or quick predictable security procedures. Acela has been abandoning efficient and established traditional practices to copy instead the airports. How long are they going to stand for a train system that isn't as awful as airport security and boarding?

Since politicians can't risk being blamed for trouble, we have no ability to control the bureaucrats and every high speed transportation system will inevitably get to be as bad as airports.

If rail could be built at the same costs as in France or Japan (or even, optimistically, Spain or Korea), to wit, $20-30MMM, you could justify it just for saving the cost of airport expansion. But at $80-120MMM it's lunacy to proceed.

You don't think Measure M passed last year would make any difference? http://theplan.metro.net/wp-content/uploads/2016/09/factshee...

You can't make a city transit and walking friendly by grafting rail infrastructure over a car-based design. You need to allow development on narrower streets without minimum parking requirements on every lot and you need more population density.

You'd have to actually change the rules and overcome the NIMBYs. Some problems can't be solved with cash.

Pretty much this, if its sub-5 story slum (aka most US cities & nearly all suburbs) overlaying rail is a bandaid. The city was built wrong in the first place, and needs to gain density, rail can keep a dense city economically viable by providing reliable, rapid transit. There is no good transit solution for suburban sprawl.

I live in LA and travel to the Bay for work frequently. I can walk to a Metro station from my house in Highland Park in 10 minutes and be at Union Station in 13 minutes. If I could spend 3 hours on the train to SF from Union instead of driving to Burbank, arriving an hour before my flight, and then spending an hour on a plane, I would do it every time.

> That's because Boston and NYC are both navigable without a car. Only one side of the SF-LA link can claim to be close to that ideal.

Just put some car rental places near the station in LA, just like is done for airports.

I think the point is that one of the things making train travel to, say, Manhattan attractive vs. air is that I can just walk, subway, or take a shortish cab trip to my destination from Penn Station. That's a big advantage over getting into Manhattan from any of the NYC airports.

If the answer to arriving in LA is "you need to rent a car in any case" (which it often will be), then you lose one area of train potential advantage.

I'm one of those people that go NYC to Boston and NYC to DC. I work in the dining car with laptop and WiFi. Only take planes when taking early AM flights to get to Boston or DC early in the morning.

And if you're coming from outside Boston to Manhattan Westwood station can be a lot easier than Logan airport.

I avoid flying like the plague but I'll do it if morning meeting and can't go down night before.

I live about 30 minutes from BOS by bus/rail. Let's say everything is timed perfectly: I get TSA Precheck, 10 minute curb-to-gate, 5 minute wait, boarding commences at T-20, first in line for takeoff.

That means I leave for the airport 75 minutes before my flight, fly for 1 hour, arrive in JFK/LGA/EWR, and then take about 30-40 minutes of public transit to my destination in Manhattan. And that's the ideal.

There's some break-even point for travelers between the more relaxed pace of train travel as opposed to the efficiency of air travel, but BOS - NYC is still within the realm of the train, even at Amtrak speeds.

Plus with a train you don't have all the stupid security that you do on a plane.

France has tracks with 320km/h.

Germany uses 300km/h on some tracks. Here the view from a train at that speed overtaking cars on the Autobahn:

https://www.youtube.com/watch?v=Js3XDTE5Iyg

China learned from it, here using a version of a German Siemens ICE train, Shanghai/Beijing in under five hours. https://www.youtube.com/watch?v=xvytUaRs2dU

320 kph is the maximum cruising speed. Average speed for a trip is typically much less than that because the trains can only hit 320 kph on certain stretches of track. Even 320 kph is about 30 kph short of what the California trains are planing to average. From what I understand the operating costs go up considerably between 320 kph and 350 kph due to the need for much more stringent track maintenance.

> 320 kph is the maximum cruising speed. Average speed for a trip is typically much less than that because the trains can only hit 320 kph on certain stretches of track.

France and Spain have those tracks in Europe.

The newer German tracks with 300km/h are operated at 300km/h with the ICE 3. Even through towns and tunnels. It's not the average speed, but a typical cruising speed on those tracks. France has more and longer high-speed tracks, thus the high-speed trains operate at higher average speed than in Germany.

Operating at 300km/h and beyond is indeed expensive. As are high-speed train in general. Still Western Europe now has an extensive amount of high-speed tracks:

https://upload.wikimedia.org/wikipedia/commons/b/bf/High_Spe...

UK HS2 project is being planned for 400km/h operation.

Oh, most new track is designed for operation speeds far beyond what's commercially expected: it's very cheap to do (the primary requirement, after all, is about curve radius, which just affects your choice of route). The signalling, the overhead line equipment, the rolling-stock… all of that is all designed for the initial operational speed, because that does become expensive (or non-existent) if you want to push the speed up.

> France has tracks with 320km/h.

350 nowadays, that's the rating for the latest LGV Est although the current trainsets are only rated for 320.

This is completely false. The entire flight demand of SF to LA is so pathetically low in comparison to the demand from the Cartesian set of intermediary routes that catering to just SF to LA would doom the project to failure. There isn't a successful HSR system in the world that hasn't already learned this lesson. Not only is Fresno crucial to ridership projections, but so are all the other small towns like Merced.

I would love to read about this lesson that other HSR systems have learned in more detail. Do you have any articles about it?

According to the figures released before the project, it will cost the same, but take twice as long to go from LA to SF compared to flying. No one is going to take the train instead of flying, except maybe John Madden.

I have a friend who cannot fly due to temperament. He's tried everything from hypnotherapy, to beta blockers, to getting near black out drunk. His fear, which he knows is irrational, has only worsened over time.

He would take such a train.

Another thing is that a one hour flight from SF to LA means, 30 minutes to airport, 30 minutes to get through gate, 15 minutes to board, 10 minutes on tarmac, an hour flight, 10-20 minutes for landing/deplaning, 15 minutes to ground transport.

So, two hours seems ballpark.

Trains typically don't have as much security and boarding is simplified. I'm thinking plane and train travel times door-to-door would be comparable.

Finally, train seats are usually much more comfortable than seats on domestic flights.

I think a substantial number of travelers would use the train instead of flying.

But wouldn't train connect city centers? Airports aren't really near anything.

This is not 100% true and the situations that "not near anything" is true for are varied. DFW was located "in the middle of nowhare" ca. halfway between two cities when it opened. I visited there more in the past than now but "anything" was building up around it when I did go sporadically. LAX is not close to Downtown LA but there are lots of office buildings out there. San Diego's and Pheonix's airports are very close to Downtown. In the case of San Diego it would be difficult to do anything but reuse the existing railways. I suspect if you look around there are a range of airport distances and a range of abilities to get high-speed railways close to city centers, rather than everything setup like Heathrow vs London.

It depends a lot too on how airports are connected to cities. While it admittedly probably takes a good hour from gate to hotel, the fact that I can hop on BART from SFO (or several train options from Heathrow to London) negates some of the airport vs. center city difference. Especially because it's not like you can just walk from a given downtown station to your final destination in all cases anyway.

In Southern California this is utterly untrue. Burbank, Ontario, Santa Monica, Long Beach, and Orange County/Santa Ana are all airports with multiple daily flights to/from Northern California, and between them you're close to 75-80% of the Los Angeles area. A train that stop at union station downtown will be close to one place - downtown. LA doesnt really work that way though - it's ridiculously spread out.

I am frequently traveling between Portland and Seattle. I've tried airplane, bus and train. The train is the slowest of the three. However, The Lyft ride from SeaTac to downtown Seattle costs more than the train ride. The train is also so much more comfortable. Unless I am under time pressure I take the train. The bus frequently smells like porter potty which eliminated that cheaper and faster option. Of course the LA-SF distance is about twice as far, but it's also not supposed to be a shitty Amtrak train that has to constantly yield to cargo trains. I am super concerned about the stops in irrlevant cities though.

>I am super concerned about the stops in irrlevant cities though.

And people can't understand why large swaths of the population think well-off techies are out-of-touch.

Not only that, but they presumably have never ridden on a train before.

Every Amtrak train I've been on (which, of course, is not the model of efficiency) takes on the order of two or three minutes for most intermediate station stops, if not less. The train stops, people leave, new passengers enter, and it starts rolling again. Stowing/gathering belongings, getting seated, lining up at the exits, and ticketing is all done while the train is in motion. It's not like having a flight layover.

You also lose time by slowing down and speeding up but that notwithstanding, you may have fewer intermediate stops with HSR (as is the case with Acela vs. Regional service) but you can't cut out intermediate stops. All you have to do is look at Shinkansen routes.

The relevant duration for an honest comparison is door-to-door timing -- lime left origin to time arrived at ultimate destination. Travel to/from airport, security, checked bag retrieval, early arrival for margin-of-error to not miss a flight, etc. favors rail, making the overall time-in-transit difference less.

Plus, we could see added environmental impact taxation for short-hop flights (carbon tax, if nothing else) that would raise airfare costs.

Presumably also you can pull out your laptop and do useful work for a much higher fraction of that door-to-door time as well.

This. I don't even drive to Giants games anymore (South Bay to Mission Bay area of SF), although I have full-season parking passes -- I drive to the nearest CalTrain station with safe parking, then I can read or do online work. And the parking passes sell at a profit.

According to a quick look, a flight between LA and SF takes approximately 1.5 hours. So, a train would be 3 hours?

I'd wager a lot of people would take that. Once you factor in arriving at the airport an hour before your flight, along with the time taken to get to/from both airports to the actual centre of the city as opposed to where the airports actually are, 1.5 hours actually starts to sound like an improvement.

I would be utterly shocked if even the top speed reaches 220 MPH.

> Existing high speed rail lines rarely run at 200 MPH even momentarily

Really?

Zaragoza to Barcelona does 310kph and very smoothly too. Much more comfortable than flying: https://dilemma-x.net/2014/03/14/high-speed-rail-in-spain-is...

That's 193 miles per hour, doesn't refute what your parent said

310kph is probably the maximum cursing speed. I can't find precise distance information for AVE between Zaragoza and Barcelona, but it's somewhere around 320-330 km. The minimum travel time I see listed is about 1.5 hours. So conservatively the average speed is around 220 kph or 137 mph.

Reply on news.ycombinator.com