Hacker News new | ask | show | jobs
by tialaramex 1526 days ago
For comparison, the Elizabeth line (Crossrail) under London will launch with 12 trains per hour (one every five minutes) and will eventually go to 24 trains per hour during peaks. CSX says 8-10 through trains per day use this route, plus 1-3 "coal and grain" trains and "numerous" local services. The reported scarcity of trains on the stream matches this.

Now of course much of the route Amtrak wants to use is single track, which is especially a problem for freight trains since they're so long that only very long purpose made sidings can possibly allow them to pass. But on the other hand it's also notable that CSX insists daytime is "peak" for freight and that doesn't make much sense. Since passengers mostly want to travel in daytime, it makes sense to shift freight to the night, not schedule all the freight for daytime and then insist that passengers be re-scheduled instead.

3 comments

Maybe we should just nationalize the existing rail network. America has a decent rail network but unfortunately much of it is private leading to this bullshit. Here's a fun rule of thumb for ya: Road wear is proportional to the fourth power of weight per wheel. A truck moving things does something like 400-600x as much damage to roads as a car.

Also tire dust is bad for you and other living things, train wheels have much less rolling resistance.

The us has the best freight railroad in the world handling much more freight than countries that have nationalized rail.

Freight and passenger rail rail do not mix. Amtrak needs to build their own track instead of complaining

This is American rail transportation’s learned helplessness. Freigh and passenger rail mixes just fine in other countries.
As far as I know, the UK is about the only country that's made this work reasonably well, and that's with pretty much all routes double-tracked or quad-track for the major ones, and much less freight rail than the US and worse passenger rail than Europe.
The exurban areas of Australia’s three largest capital cities. For example, the Central Coast line north of Sydney has half-hourly electric intercity trains interoperating with freight on a line that extends 165km from the city. Melbourne and Brisbane have similar lines radiating out to regional areas.

Within the Sydney metropolitan area, commuter trains are operating among freight trains at 5-15 minute frequencies, though with dedicated freight bypass lines in some places.

One issue in the USA, alongside private ownership of the rail lines, is oversized freight trains and resulting overbuilding required of passenger trains for crash safety. Unfortunately that rules out high-performance EMU designs as used in other countries. I believe Caltrain had to get an exemption for their ongoing electrification upgrades.

> Within the Sydney metropolitan area, commuter trains are operating among freight trains at 5-15 minute frequencies, though with dedicated freight bypass lines in some places.

AFAIK there a blackout periods for freight on the Sydney Trains network during the morning and afternoon peaks, though.

Plus the "dedicated freight bypass lines in places" nowadays is basically the complete route between the southern limit of the suburban rail network at Macarthur and Port Botany via Enfield Yard, plus a stub from Enfield towards North Strathfield. So major track sharing (especially with super-long interstate freight trains) only really happens from Strathfield on the line towards Newcastle, and that again is at least three or even four-tracked for parts of the way within Sydney (although unlike the southern half it's not exclusive freight infrastructure).

Plus whatever local-ish freight traffic might still exist around Western Sydney, towards the Blue Mountains, and along the coast to Wollongong.

Those overbuild passenger trains rules are no longer in effect, though the rules didn't change too long ago

Longer trains are more efficient. That is why we run them

It’s much easier to double, triple or quadruple track the beat up American railroad infrastructure if it was in public hand, working to maximise capacity for the whole market, rather than just every piece of track being used to maximise profits for a single company, acting as a moat to ensure as little local competition as possible.
Well the original railroads were mainly funded by taxpayer subsidies so I could say no CSX, etc needs to lay their own track, and fund it themselves.
For the most part they were not funded by subsidies. East of the Mississippi (very roughly) they mostly had to buy their own land and everything. Out west they were given free land, but the value of that land without railroads was approximately zero.
The problem is rights of way. Getting land is a lot more difficult than when the railroads first started and were given free land.

Where lines are constrained they can invest in parallel tracks and increase current capacity. Everybody could win in this way.

The only free land railroads got was worthless land out west. Of course once the railroad was built the land became valuable. Out east they generally had to buy land
America ranks #3 in terms of rail freight in terms of weight * distance behind Russia and China which sounds fine except there are few large counties. https://en.wikipedia.org/wiki/List_of_countries_by_rail_usag...

It’s hard to rank such systems but the US is a long way from #1. Being for example 113th in terms of miles of track per population.

Not sure of your source, but by most measures of efficiency US freight rail is the best in the world, and also carries 10x weight * distance per capita than Europe.
Weight * distance/population we again rank #3 behind both Russia and Canada. It says less about or rail system than our large size and relatively low population density.

In terms of efficiency we rank 31st in terms of miles of electrified track at a paltry 2,000km which significantly increases costs. https://en.wikipedia.org/wiki/List_of_countries_by_rail_tran...

Don’t get me wrong I have heard people say the US has the #1 rail network, but objectively I have never seen anyone actually back it up.

I'm wondering if this would hold true, if you'd 'overbuild' the existing rights of way with overpass-like structures where it makes sense, instead of for instance doing it all over again from scratch for hypothetical hyperloops, maglevs, etc. and have long distance passenger rail running mostly grade separated over them. This would also give better views! :-)
The United States already nationalized passenger rail-- Amtrak is the sorry result. We should privatize passenger rail, following the successful models in Germany and Japan. Amtrak should lose control of the nationalized Northeast Corridor and those timeslots awarded to bidding private companies.
Germany has very little private competition in the longer distance sector (>2h) and almost no private infrastructure. Both are controlled by the state owned DB and there are ambitions to make the infrastructure part even less profit oriented and talks about moving it out completely into a separate, even less private, entity. In the regional rail sector there are more private actors, yes, but they work routes tendered by the state with pretty strict requirements as to frequency, vehicles and branding. It's a bit more like contractors in housing construction (and most of the infrastructure is still directly or indirectly state owned).

One of the bigger issues with Amtrak is the lacklustre infrastructure, that is mostly owned by freight companies (with prominent exceptions like the NE corridor). I wont say that Amtrak is great, but infrastructure is the bigger problem. No freight company is interested in upgrading, electrifying, speed increases or even building new lines. There are few private actors interested in the passenger rail sector (with exceptions such as Texas Central and I wish them the best).

I also can't imagine great results with new private actors, since non-high speed rail suffers even more from the competition with the car due to a lack of attractive or in many cases usable public transit option near the start or destination. And high speed rail is really expensive to build, so I have my doubts that private actors will be able to secure funding without any previous examples.

To fix passenger rail in America, in my opinion, you have to at least have a major rework of how rail infrastructure works.

Canada is also doing a lot better with rail than America, with a major priority being the independence from freight companies in regards to infrastructure.

Where even the regional operators are at least partially owned by the states in turn.

Furthermore this is causing friction when the contract ends, and another bidder wins.

Compared with how it has been before this, when it was all state run, much hilarity ensues on all levels of operation.

Beginning with the engineers, now dispersed over different sub-contractors, not being able to assist when there is a shortage in another sub-contractor.

Leading to delays, because available ones have to be brought in by taxi from over 100+ of km away.

'Streckenkunde' == knowledge of tracks, stations and switching/marshalling/maintenance yards is degraded, because the sub-contractors don't do it all, everywhere, anymore.

Regarding maintenance, more empty movements to farther away, because not every shop can or will service everyones locos, trainsets.

For some 'unexplainable' reason, during the slightest bad weather chaos ensues, every year, again and again. No matter if cold, heat, wind, wet.

While aeons ago, they advertised with an engine plowing through the snow, caption: "Everybody is talking about the weather. We don't."

[·] https://www.spiegel.de/geschichte/68er-plakate-a-946587.html...

[·] https://www.youtube.com/watch?v=eo8l2qp2N8M

[·] https://www.youtube.com/watch?v=mGhJW5TvIuQ

This was the truth, at the time.

What we have now are a few high-speed tracks connecting the larger centers, and regional mass transportation in and around these. Outside of that it's patchwork, or doesn't exist at all. (Though it did! Once.)

In addition to that, it costs much more, and is inflexible to book.

It's FUCKED!

edit: Also 'type-ratings' for the engineers. Before pseudo-privatization and splitting in sub-groups, there was only distinction between Diesel(hydraulic) and electric locomotives, and passenger vs. freight rail. Those were about a dozen each, and they usually could drive all of them from their branch of diesel or electric. Today? Not anymore.

> there was only distinction between Diesel(hydraulic) and electric locomotives, and passenger vs. freight rail

That's oversimplifying things a little too much – type-ratings were a little more fine-grained than that even during the days of the Federal Railways.

Sounds a lot like the argument against a public USPS: put it in a position where it can't succeed and has no leverage, then point at it and say that a public model can't possibly work and that it needs to be privatized.

Privatization does essentially nothing to fix the issue of rail ownership or use. Making the service privately owned gives it no leverage to operate efficiently: that comes from regulation, which is a separate problem. The biggest difference with a privatized service is that a private service can be bullied out of existence.

USPS is awesome. But there are other examples of extreme incompetence in Government from local to Federal. SSA, USCIS, DMV (state), etc.

So both sides cherry picking their favorite gov agencies does disservice to improving things and doubling down on what works, and criticizing what doesn't.

Well, Amtrak was the nationalization of the services. The main issues with Amtrak today are that they do not run the tracks as well, so despite laws with contrary intent they never have priority on the rails.

That, and the requirements for high speed rail are pretty much opposite of the host freight railroads, which are content to have slow, low-standard track because it’s cheaper.

The NEC is decently run. The rail conditions are so bad that if we were to privatize rail pretty much everything outside the NEC would dry up.

Incorrect. Amtrak owns and operates the most valuable and viable stretch of rail in the United States, the Northeast corridor stretching from DC - Baltimore - Philly - New York City - Boston.
Hm. I meant to say that "The Amtrak-run NEC is decently run."

Is it true high speed rail, no, but at this point the startup costs of actual high speed rail are so high that private entities pretty much never take it on. (Brightline in Florida is not actually high speed rail due to the speed, Texas Central is floundering due to lawsuits, Las Vegas to LA is on constant life support, etc.)

The only profitable private railroads are the ones that started off with cheap property to develop near stations and never sold it off, essentially becoming landlords/property developers in their own right. But the horse has left the barn for that in the US and nearly all city center property is expensive for a singular entity to just buy up and redevelop.

I think the answer is more complex. In Switzerland, it's mostly state-owned (with a few local concessions) and it runs very smooth and we have one of the most punctual railroads in the world.

So just saying "state owned is bad" may no be the full answer, just a part of it. There are many factors that play a role and may lead to a different answer for a different country.

As a German I wouldn't call it successful. Maybe seen through pink glasses while stoned or drunken. Otherwise? Meh...
How does it work in the Netherlands? They do such an awesome job I can’t believe it. I think it is semi privatized. But I don’t know the details.
The rail infrastructure is owned, managed, and maintained by ProRail: a private company but its sole shareholder is the Government. Everyone running trains pays them a usage fee.

The vast majority of passenger trains and all railway stations are owned and operated by NS: again a private company with the Government as sole shareholder.

Various other private companies are running local services, and all freight traffic is done by private companies too.

> America has a decent rail network

According to what metrics? For example my country has over five times greater length of railway network per unit of area. For electrified rail, which is increasingly important for sustainability reasons, that advantage goes up to a factor of 200 (!).

I'm not 100% sold on these [0] values as I have seen different elsewhere. But according to this, the US is third behind China and Russia on rail ton-miles. I think those are reasonable numbers given a) the compactness of the US compared to Russia and b) that unlike China, the US has major ports on both coasts and doesn't need to traverse the entire country.

The reason I'm skeptical of the numbers is that UP/US DoT reported 2.7T ton-KMs for 2018 [1]. Given that there are good reasons for 2020 in particular to be low, I wouldn't hold 2020 out as representative.

[0] https://www.russia-briefing.com/news/russian-rail-freight-vo... [1] https://www.up.com/customers/track-record/tr120120-freight-r...

> But according to this, the US is third behind China and Russia on rail ton-miles

That's awesome if you're not interested in carrying people, but whether a network incapable of carrying people is "a decent rail network" is something many would dispute.

Also, Russia has ~44% of population of the US, yet it still outperforms the US in absolute terms? That's quite impressive for Russia in my book. Having said that, they've always been heavily dependent on rail to lower their transportation costs. Trucks and airplanes won't work for them nearly as well. So I'm not surprised if they're placed so high in rail freight ranking.

> The reason I'm skeptical of the numbers is that UP/US DoT reported 2.7T ton-KMs for 2018 [1].

I'm reading 1.7 on that page. Am I looking in the wrong place? It says "In 2018, 1.7 trillion ton-miles of freight (calculated by multiplying shipment weight in tons by the number of miles that it is transported) was shipped by rail, according to the U.S. Department of Transportation."

kilometers vs miles.
America has huge swathes of undeveloped and uninhabited land. Maybe over five times as much as your country.

What would the multiple be if you compared your country’s metrics to just the northeast corridor?

Adding up the metropolitan areas around the NE corridor, there are ~40 Million people living there, around half of Germany. The NE corridor includes 611km of electrified track (or 1500km for all electrified rail in the US, adding up all the numbers mentioned in the Wikipedia page).

In Germany 20000 km are electrified, so around 25cm per inhabitant. In comparison to the NE corridor with a number of 1.5cm per inhabitant, this is a huge difference. To account for other railways inside the NE corridor, we can also just use all electrified rail as a reference and arrive at 3.7cm per inhabitant).

And Germany hasn't been great about electrifying it's rail.

I'm confused: why are you only counting electrified rail? For example, all of Boston's regional rail is diesel locomotives.

(US transit agencies are unreasonably ignorant of best practices, including electrification and EMUs, but it's still rail.)

Electrified Rail is important for great passenger rail, by nature of it's much higher acceleration and lower cost for increased service. This is part of great rail infrastructure for me and much of the electrification in the USA has been rolled back due to the different needs of freight providers and cost-cutting measures. Countries like India and China are also huge on electrification for dedicated freight routes, so it's not just a passenger service thing.

It's also much easier to compare as a baseline of decent rail infrastructure, since it implies a minimum condition of the line and a certain amount of investment in the last 100 years (and it was much easier to compare for the NE corridor, since that contains most electrified rail in the US). Most countries that are considered to have a great rail network have a lot of electrified lines, beginning with Switzerland but countries as Russia have also invested a lot in electrification. Electrification is a lot of effort, and it will take multiple decades to achieve a decent percentage in the US if it were started right now with a lot of political backing.

Many transit agencies in the US, including the one in Boston, are planning electrified rail (as they're aware of the benefits as well) but are unable to construct it right now (and likely the next 10 years) due to funding and ownership issues.

It's not impossible to run decent service over non-electrified rail, but the slower acceleration, near impossibility of high speed as well as the increasingly low availability of DMUs make it harder and, coupled with the higher fuel costs, unattractive.

Properly assessing the state of the routes without using electrification as an easy shortcut was way too much effort for me.

In short: Just because you have a lot of gravel roads everywhere doesn't mean you have a decent road network

i think youve misread this issue. i can assure you that the us is aware of best practices. the private swctors awareness of the economics involves drives a lot of their behabior while the politicians lack of awareness of anything but the political costs drives the regulatory side.
I'm pretty sure we also don't have equal rail density in all regions. So this is just the average. Nevertheless I'm not sure how this can swing a factor of two hundred. (If I use rail length per capita instead, which somewhat compensates for unpopulated regions, the advantage is still ~1.4x for non-electrified and ~50x for electrified rail.)

Also, is there any statistics for this northeast corridor, regarding region area, electrified, and non-electrified rail length? I only know where to get national statistics, so that doesn't help me a lot here.

The Northeast Corridor where the Acela runs is all electrified. In fact the electrification of the tracks north of New Haven (so an engine switch was no longer needed) was one of the big benefits Acela brought to even the non-Acela trains.

I don't know what it looks like north (basically Portland) and south of the northeast corridor--or the non-coastal routes in New England.

Thank you for putting that out there -- I've been saying the same thing.

Then invest in increasing the traffic capacity and enforce Positive Train Control on all rolling stock -- https://www.aar.org/campaigns/ptc/

Add in electrification where possible and allow for other "carriers" to utilize the lines to maximize utility.

America's freight transport is cheaper than anywhere else in the developed world. Nationalizing it would increase that cost, not lower it. It would also push a lot more freight on to the roads instead of rail because road transport would get closer to the price of rail transport.
What is the mechanism that causes nationalized rail to inherently be more expensive?

I can see the argument for nationalized being cheaper due to not being for-profit, due to having larger negotiation power (via the government), etc.

What's the argument for nationalizing it increasing cost? Is it just "look at all these anecdotes?" or is there some fundamental economic reason.

Freight rail is perfectly fine spending minimum amounts of money on clapped out tracks that twist and curve. Freight rail is never going to compete with air freight on speed, and is also much faster than boats, so speed is not really a priority.

Competitive passenger rail is high speed. High speed requires double tracks (to avoid slowing down to let trains pass), straighter rights of way to go faster (increasing land acquisition cost), and very high maintenance levels for safety and comfort reasons. Hosting is freight trains on high speed rail tracks also significantly increases their already high maintenance costs because they are so much heavier and cause more wear and tear.

Nationalization, I think, is less the driving fundamental here than the inherent conflict between timely, regular passenger services and the American freight rail system for bulk freight. The only time the freight railroads really prioritized passenger services was when they delivered mail on those trains, which is also lightweight and needs fast delivery, but that has long moved to air freight.

>but that has long moved to air freight

Well, not even air freight. How much extremely time-sensitive information is actually transported physically these days?

I have a few things like my tax return but most of what I deal with these days is sent electronically if it's really urgent.

Mail services also included small parcels, and if anything those continue to need ever faster logistics.

When we talk about air vs rail we are talking about hours and days vs. days and weeks.

> What is the mechanism that causes nationalized rail to inherently be more expensive?

Hold that question for the next time the MTA gets caught spending stupid amounts of money to get nothing done.

The exact people you are agreeing with now will be happy to provide you a laundry list of ways government dysfunctionalality wastes money and gets taxpayers and riders less for their dollar if you ask in that context.

> What is the mechanism that causes nationalized rail to inherently be more expensive?

A management that is not incentivized to reduce costs, or a the very least, following the operating policies set by politicians coming first before reducing costs. (It's hard to reduce costs when how you operate is decided by politicians.) The same problem that plagues government run systems all over the world.

A great post on just a few of the silly policies is here: https://news.ycombinator.com/item?id=30978263

> A management that is not incentivized to reduce costs

This argument doesn't seem to me as if it's fundamental.

Politicians are at least in theory incentivized to reduce costs since citizens would rather not pay more taxes, and in theory a politician who enacts wasteful policies would not be elected.

On the flip side, companies are only incentivized to reduce the cost charged to consumers (or in this case, companies shipping freight) in the face of competition, and long-haul freight has a massive up-front investment cost of building out rails, so there won't ever really be that many choices. This is akin to the highway robbery ISPs can still charge, even though they are private companies and the moat of laying fiber isn't nearly as extreme as that of laying rail

> following the operating policies set by politicians coming first before reducing costs

Won't it be equally true that private and public corporations will have to follow laws? That seems identical.

> A great post on just a few of the silly policies is here: https://news.ycombinator.com/item?id=30978263

I'm trying to ask if there's a fundamental reason here, not a bag of anecdotes, and we were also talking about freight, not commuter trains, so that comment isn't particularly relevant to the comment tree you started about freight specifically.

> Politicians are at least in theory incentivized to reduce costs since citizens would rather not pay more taxes, and in theory a politician who enacts wasteful policies would not be elected.

Politicians are incentivized to give lip service to reducing cost, not actually reduce costs. They only need to look like they're more likely to reduce costs than their opponent, assuming the voter actually cares enough in the first place. Further, it assumes that politicians even know how to reduce costs when they have no experience what-so-ever in running a train company.

The trains on the Tube under London aren't 3 miles long.
That's not a comparison. Amtrak trains travel at like double the speed of cargo trains.

It's makes it much harder to schedule, since you have to clear half the track before Amtrak can enter it, otherwise it would reach the back of the freight train.