Hacker News new | ask | show | jobs
by 0xjunhao 486 days ago
Maybe we should build more railways. The ground is more stable than the air.
3 comments

Just because the ground is more stable than the air doesn't mean there aren't train collisions.

Air travel actually has a lower fatality rate per mile than trains do. Rail crossings can be dangerous places.

If we invested as much into rail safety as we do into air safety, it'd be fair to compare them. Regardless, in the long term rail transportation is simply much more sustainable and increased investment is very much warranted
It's also much more pleasant. And in the rare cases when things do go wrong your chances of survival are much higher.
> And in the rare cases when things do go wrong your chances of survival are much higher.

That's a completely irrelevant metric.

The only thing that matters is my risk of injury/death per passenger mile. Trains are much worse. End of story.

What you're saying is that trains get into far more accidents, but that don't injure/kill you. On top of still being much more likely to injure/kill you.

And you think that's a good thing...?

You should compare to a country which invests significant effort into a safe railway, such as the UK.

There have been 7 passenger deaths in the UK in the last 20 years, with around 60 billion passenger km each year.

That seems to be of the same magnitude as aircraft passenger deaths per km, though I can't find those statistics in precisely the same form.

> It's also much more pleasant.

And much slower. Most people prefer shorter discomfort over a longer one.

Also sometimes more expensive because train service is much more monopolized due to the need for infrastructure betweem stations.

Granted the lines are a natural monopoly. I don't see why the usual regulatory approaches shouldn't work. The government could even provide the infrastructure similar to highways.

> sometimes more expensive

Examples? Because that seems patently ridiculous on its face given the differences in energy requirements.

There are plenty of places with functional rail systems to compare to. This stuff isn't rocket science.

Prices have little to do with cost when monopolies are involed, that's the point.

DB quote for "high"speed train from Frakfurt to Paris: from 151 €

Flight from Frakfurt to Paris: options as low as 76 €

This isn't a 100% fair comparison but hopefully serves to demonstrate the absurdity.

The Shinkansen has had one fatality in its entire existence. Domestic air travel is insanity, and I say this as a pilot and a flight instructor.
That's cherry-picking to the extreme.

Not only are most trains not Shinkansen, not even most trains in Japan are Shinkansen.

Only the Shinkansen (and TGV on my sibling post) can be compared to planes though. Local or IC trains might be compared to cars in the US.
Doesn’t the Shinkansen and other HSR systems get rid of at grade road crossings? Yes, without people actually crossing in the tracks, you’ll get less deaths that way.
Sounds like getting rid of at-grade road crossings is a good thing, who've thought?

Sounds like a good investment to me regardless of rail speed.

> Only the Shinkansen (and TGV on my sibling post) can be compared to planes though.

Says who? Right now, in reality, the only thing that makes sense is comparing existing flights with existing train routes.

And the TGV has only had deaths once during a test run.
I'm interested to hear more on why domestic air travel is insanity.
Railway crossings are not especially deadly if you're a train passenger. Most deaths are from train-on-train collisions or overspeed derailments.
I feel more comfortable with the thought of applying the emergency brakes on a train versus applying the emergency brakes on an aeroplane.
It wouldn’t surprise me at all if aircraft can generally slow down faster than trains, assuming level flight and grade (or the same climb rate)

It’s not really fair though, because the aircraft has so much force acting against it from how fast it’s moving. If somehow a jet was moving at train speeds and could only use its aerodynamic surfaces to slow down, it would probably take a while.

What percentage of train fatalities are non-passenger suicides? The statistical comparison isn't very fair.
There's about a million times more ground accidents than air accidents
HN is full of clickbait messages like this. US and Canada can use a lot more trains, but air transport is more suitable for medium to long distance, such as this Minneapolis-Toronto route. They are complementary.

The accident occurred on the ground, any aircraft is susceptible, “more trains” doesn’t make it less so.

The conventional wisdom is that high speed rail is competitive with flying when travel time is at most 4.5 hours. But China has been challenging that with some ridiculously long routes, such as 2760 km from Beijing to Kunming (10 hours 43 minutes). I think the idea is that if flight time is long enough that you can't travel and work reasonably on the same day, you have to dedicate the entire day for travel. Then the difference between a 3.5-hour flight and a 10-hour train trip is no longer that significant.
"Night trains" (where the train effectively doubles as lodging for 8--12 hours) make even longer-haul routes quite viable. Even without HSR that affords nearly 1,000 mi / > 1,500 km of range (80 mph constant / 130 kph).

At typical HSR speeds of 185 mph / 300 kph, that extends to ~2,200 mi / 3,600 km. That's sufficient for travel from SF to Chicago, or NYC to Salt Lake (with range to spare).

(Both calculations presume 12 hours and operating largely at top speeds, both of which may be atypical in practice, but do satisfy the maximum possible range question.)

Travel within major population corridors, generally the east coast (Boston, Minneapolis, Miami, Houston) or west (San Diego, Phoenix, Seattle, Salt Lake, Denver) should be highly viable.

And as is often noted, rail operates city-centre to city-centre, and typically has fewer security checks and delays.

> air transport is more suitable for medium to long distance, such as this Minneapolis-Toronto route.

Eh maybe. That's not much longer than Beijing-Shanghai. Rail can definitely be competitive.

> The accident occurred on the ground

Plane crashes generally do. Given that all plane journeys involve at least one takeoff and landing, it's fair to consider collisions and incidents during those as part of a safety comparison.

> Eh maybe. That's not much longer than Beijing-Shanghai. Rail can definitely be competitive.

Connecting two growing mega cities by rail in a country that can still build things vs two mid-tier cities in two different countries, neither of which build anything without a decade of delays and 3x the budget (before the projects get cancelled), is not a good analogy to use against flight IMO.

> countries, neither of which build anything without a decade of delays and 3x the budget (before the projects get cancelled)

That ends up being circular though. The main reason the US and Canada can't build anything is that they don't build anything.

It's also bullshit. Clearly both the US and Canada do build lots of things.
Like what?
China eastern airlines has its own desk at the airport to deal with Beijing-Shanghai flights, even after HSR was built out there are still more planes flying between those two mega cities than Minneapolis and Toronto (makes sense if you consider connecting flights wouldn’t benefit from HSR very much). Maybe when they get a maglev going…
Beijing and Shanghai are the two largest cities in China. There is enough demand to justify trains. There are also at least 50 daily flights between the two cities.

Now what is the demand between Minneapolis and Toronto?

Improved transport infrastructure tends to induce demand. We're familiar with this in terms of highway widening, from the paradox that widening highways tends to not improve traffic speeds. Famously Los Angeles today sees comparable net travel speeds as existed in the age of horse-drawn transport, though of course far more net daily passenger miles.[1]

Another example I like to cite is of Denver, CO, which grew roughly seven-fold in population in the decade after it was linked to the then-new US Transcontinental Railroad, 1870--1880:

<https://www.uncovercolorado.com/colorado-train-stations/>

Rail build-outs competing with existing air links is another matter of course, though experience in Europe, Japan, and China should help provide useful data.

________________________________

Notes:

1. Discussion of LA freeway speeds generally, noting several stretches (including those recently widened) netting < 20 mph: "Five years after Sepulveda Pass widening, travel times on the 405 keep getting worse" (2019) <https://la.curbed.com/2019/5/6/18531505/405-widening-traffic...>. I'm not finding the specific horses-to-cars comparison though I'm sure I've encountered it before.

The twin cities MSA has a population of about 3.7 million, the Toronto CMA 6.2 million, so Chinese cities at a similar metropolitan area population and distance would be Kunming and Changsha. Which see 20 trains a day per direction, not counting the slower sleeper trains (and also 10-12 flights per day).
The population comparison is misleading. Kumming and Chiangsa are part of Shanghai–Kunming railway. The trains between these two cities don't stop at either, but connect all the stations on the larger railway. Compare that to flights between Minneapolis and Toronto. The majority of passengers are point to point, as transit passengers from Minneapolis probably take hubs such as Chicago or NYC, not Toronto. There are up to 6 daily flights between MSP and YYZ, so up to 600 passengers daily. That'd fill up less than a train. That doesn't justify a high speed passenger train connection between the two cities.

The original assertion is that train can probably replace airplane on this route. That doesn't make any sense.

> and also 10-12 flights per day

That are probably those who want to travel directly between the two cities. So even though train option exists people still choose to fly.

> Kumming and Chiangsa are part of Shanghai–Kunming railway. The trains between these two cities don't stop at either, but connect all the stations on the larger railway.

Which would be the same for a North American rail network. Kunming was Minneapolis in this analogy, trains to Toronto could carry on to Boston or Montreal or New York just as trains to Changsa carry on to Shanghai.

> The original assertion is that train can probably replace airplane on this route.

I don't think anyone claimed that it would replace planes completely. It could be competitive, it could be an option for people who want it.

Can you really not imagine any reason why world-class rail between Beijing and Shanghai is more feasible than between Minneapolis and Toronto?
If you want to say something then say it. No, I don't think there's any particular physical reason for that; I think the poor state of north american passenger rail is almost entirely for political reasons and could be changed if the public had the will to improve it.
Beijing and Shanghai are two of the most important cities in the world. Minneapolis and Toronto are not. Of course it’s physically possible to build a high-speed train from Minneapolis to Toronto, but it makes little sense to do so unless you’re building all the O(n^2) point-to-point links between all the medium-importance cities in North America.
High speed rail requires density.

Toronto and Minneapolis are villages comparatively.

There are significant air routes in China between higher population cities than the MSP, YYZ pairing