High-speed trains vs. airline service

[lpetrich]

Any good jokes about Ryanair airports? Google gives about 10,000 hits for Ryanair "out-of-the-way airports".

Here are Ryanair's other distant airports:

Brussels .. Zaventem: 13 km .. Charleroi: 60 km (R)
Duesseldorf .. D: 6 km .. D-M: 19 km .. Weeze: 84 km (R)
Glasgow .. G: 15 km .. Prestwick: 51 km (R)
Hamburg .. H: 9 km .. Luebeck: 55 km (R)
Lyon .. St. Exupery: 20 km .. St. Etienne: 49 km (R)
Munich .. M: 28 km .. Memmingen: 115 km (R)
Oslo .. O: 37 km .. Rygge: 60 km (R) .. Torp: 119 km (R)
Stockholm .. Arlanda: 36 km .. Vasteras: 103 km (R) .. Skavsta: 106 km (R)

ASCII-art diagrams:

Code: Select all

Barcelona   .#.......R.R
Brussels    .#....R
Duesseldorf ##......R
Glasgow     .#...R
Hamburg     #....R
London      ..#.RRR
Lyon        ..#.R
Munich      ..#........R
Oslo        ...#..R....R
Paris       .##.....R......R
Stockholm   ...#......R

Checking on other low-cost airlines, EasyJet uses Ryanair's London airports, but does not use Ryanair's other distant ones. The original low-cost airline, Southwest, also goes to relatively-close airports.

[lpetrich]

Just for the Hades of it, I decided to consider what would be necessary for transoceanic railroad service. Tunnels, of course, but what conditions would the tunnelers encounter?

The longest existing undersea rail tunnels are the Chunnel and Japan's Seikan tunnel between its Honshu and Hokkaido islands. Their stats:
Tunnel ... Total Length ... Undersea .... Tunnel total depth ... Seafloor depth ... Tunnel depth beneath seafloor ... Cost in present-day dollars ... Cost per unit length
Chunnel ... 50.45 km ... 37.9 km ... 75 m ... 30 m ... 45 m ... $18b ... $360m/km
Seikan ... 53.85 km ... 30.3 km ... 240 m ... 140 m ... 100 m ... $6.7b ... $120m/km

There are several proposed rail or road tunnels which are longer than these two:
Bering Strait: 80 km ... 55 m -- Bering Strait crossing
Strait of Malacca: 128 km ... 55 m -- The Strait Of Malacca Crossing
Korea Strait: 200 km ... 100 m ... -- Japan–Korea Undersea Tunnel
Strait of Gibraltar: 14 km ... 900 m / 23 km ... 300 m -- Strait of Gibraltar crossing

All of the existing and proposed ones are on continental shelves, with the Strait of Gibraltar being a borderline case. Continental shelves usually extend around 80 km from the shoreline, and they end in continental slopes, which are followed by continental rises. Continental shelves and slopes often have submarine canyons in them with a depth of as much as 2 km. Most of the oceans are even deeper, with an average depth of 3.8 km and over half being over 3.0 km deep. Oceans have mid-oceanic ridges with heights of around 2 or 3 km, and trenches that extend as far down as 11 km. Ocean-volcano islands like the Hawaiian Islands are giant volcanoes that extend up from the ocean floor.

Grades:
Continental shelf: around 0.5%
Continental slope: 1% to 10%, average 3%
Continental rise: 0.5% to 1%
Mid-ocean ridges: around 1%, likely greater in their central valleys (about 20 km wide)
Ocean trenches: outer: 5%, inner: 10 - 16%
Ocean volcanoes: around 12%

So undersea topography can be at least as bad as above-ground topography. Even worse, most potential routes cross the more troublesome features. North America - Europe crosses the Mid-Atlantic Ridge, Japan has a trench just to the east of it, etc.

[lpetrich]

Color-coded ocean depths: GEBCO world map

Pressure increases about 1 atmosphere each 10 m of water depth. The average ocean-floor depth of 4km thus has a pressure of 400 atm.

Geological hazards? Mid-ocean ridges and ocean trenches are tectonic-plate boundaries, complete with LOTS of earthquakes. Mid-ocean ridges also have lots of volcanic activity, though it's relatively-peaceful Hawaii-style eruptions with lots of lava flowing out like syrup and forming pillow-ish rock formations.

There are several possible ways of building a transoceanic line; I'll examine some of them here.

Tethered buoyant tunnels. These would be at a depth of 50 to 100 m to avoid interfering with shipping, while having a relatively low pressure. One can estimate how much buoyancy they will need from the weights and lengths of various trainsets. The lightest likely ones are high-speed passenger ones. Some TGV's are 200 m long and weigh 400 metric tons, making 2 tons / m. However, a typical suspension-bridge vertical cable every meter ought to do OK. However, these cables are a potential navigation hazard for submarines, which can go deep enough to run into the cables. US military submarines can submerge to depths greater than 800 ft / 243 mi. Will the cables need big Venetian blinds or some such attached to them?

Sunken-tube tunnels. Like the BART tunnel or Istanbul's upcoming Marmaray tunnel. At continental slopes and rises, they will need protection against submarine landslides.

Bored solid-rock tunnels. To get an idea of what to expect, we consider the structure of oceanic crust.
Sediments - average of 0.4 km, hardly any near mid-ocean ridges
Pillow lavas - 0.5. km - fine-grained volcanic rock that froze on the surface
Sheeted dikes - 0.5 km - medium-grained volcanic rock that froze on the way up
Continuous solid - 5 km - coarse-grained volcanic rock that froze in place

So unless one is willing to bore through mud or loose rock, one will need to bore one's tunnels at least 1 km below the ocean floor.

[lpetrich]

Now for some possible routes.

New York City - London. Distance: 5600 km
Much of this route goes over land, so we can take advantage of that:

NYC - Connecticut - Rhode Island - Massachusetts - New Hampshire - Maine - New Brunswick - Nova Scotia - Cape Breton: 1500 km
Cabot Strait: 120 km, depth 500 m
Cape Ray - Newfoundland island - St. John's: 540 km
Atlantic Ocean: 3100 km, depth 3000 m
Southeast Ireland - Cork - Waterford - Rosslare: 280 km
St. George's Channel: 100 km, depth 150 m
Fishguard - Swansea - Cardiff - Bristol - Swindon - Reading - London: 360 km
Total: 6000 km

One can get across the Mid-Atlantic Ridge above water by crossing it in Iceland, but doing so will require 4000 additional km, with 1000 more if one avoids most areas deeper than 1 km or thereabouts.

Los Angeles - Hawaii. Distance: 4000 km
All of it ison the Pacific Ocean floor, depth 4000 m
Fortunately, it has to cross no ridges or trenches.

Tokyo - Los Angeles. Distance: 8800 km
It is close to a route that requires only 4 sea crossings, and relatively shallow ones at that: Honshu - Hokkaido, Hokkaido - Sakhalin, Sakhalin - mainland, and the Bering Strait. Its distance: 11,500 km

Los Angeles - Sydney. Distance: 12,000 km
Pacific Ocean floor. Crosses a trench along the way.

Los Angeles - Auckland. Distance: 11,000 km
Pacific Ocean floor. Can avoid trenches by going to the east shore of the North Island.

Auckland - Sydney. Distance: 2200 km
Pacific Ocean floor. No trenches.

Tokyo - Sydney. DIstance: 7800 km
Mostly Pacific Ocean floor. Crosses the Mariana Trench.
It's possible to minimize water crossings and avoid trenches by using a rather circuitous route through Korea, China, Laos, Thailand, and Indonesia. The deepest water is between Borneo and New Guinea; the rest is all continental shelf. Total length: 16,000 km

[justalurker66]

Consider how many people want to hit the panic button going through the Chunnel or other tunnels, worried about fires, flooding, collisions, collapse and how they would escape (or even IF they could escape). Would a regular human be able to handle a long journey with no escape? Those with a fear of flying certainly would not be customers.

Otherwise such an intercontinental trip might work better than the same distance transcontinental. On a transcontinental journey there is always one more city that wants the train to stop. That wouldn't happen on an ocean tube train. (Although NY to London great circle goes through Boston and Newfoundland in North America and Ireland and Wales with about 1900 miles (3000 km) of open ocean on the 3400 mile / 5500 km trip. I suspect a less direct Boston to Bristol path going around Canada and Ireland (3200 miles / 5100 km coast to coast) would be followed. And at 250 MPH / 400 km/h it would still be a 13 hour trip in the tunnel - plus more on land.

Perhaps passengers could be sedated?

[george matthews]

Consider how many people want to hit the panic button going through the Chunnel or other tunnels,

I never heard of any. The Channel Tunnel takes about half an hour. Most journeys on the Underground in London take longer, and there are no safety tunnels. If people don't like the idea of the tunnel they can take the ferry. That's reasonably safe in good weather. Really, the actual passage is uneventful. I was once going through in a bus when there was a pause underground for about ten minutes. That was slightly upsetting, but if there had been a real problem there was a safety door no more than 100 metres away. I would have been interested in seeing the third tunnel.

I think a Behring tunnel might be possible but not soon. Taiwan to China might be possible but I think it has low probability, as also Japan to Korea.


A tunnel from Holyhead to Dublin is sometimes proposed (it would need regauging at least some of the Irish railways). There was more discussion when the Irish economy seemed to be booming but it was a financial boom that has now collapsed miserably.

[David Benton]

i think anything that involves more than 100 - 200 km of undersea tunnels would be hopelessly expensive . Certainly in the pacific , none of the routes would make any sense .
Australia , indonesia , singapore , thailand , China might , but the political problems are daunting , and you end up at the top en of Australia with bugger all population .

[justalurker66]

Most journeys on the Underground in London take longer, and there are no safety tunnels.

The underground stations in London are not miles apart. Every couple of minutes there is a place where people could get off the train and come up for air.

If people don't like the idea of the tunnel they can take the ferry. That's reasonably safe in good weather.

There are other options for a long ocean crossing. A half hour journey thinking about the safety features and looking forward to seeing the sky again is a lot more manageable than spending 12 plus hours in a place where it could take much longer to rescue you.

Really, the actual passage is uneventful. I was once going through in a bus when there was a pause underground for about ten minutes. That was slightly upsetting, but if there had been a real problem there was a safety door no more than 100 metres away. I would have been interested in seeing the third tunnel.

It would take a while for a good reputation to be built. People accept the underground portions of the tube system because of the countless journeys over the years that others have made safely. The Chunnel has a track record of years. Perhaps after a couple of decades a Boston to Bristol tunnel on a New York to London journey would be accepted (even though the travel time is slower than the airline alternative).

Shorter long tunnels need to be built to build up people's confidence. South Hampton (England) to Le Harve (France) as a ~100 mile Channel tunnel would be a step. Plymouth to somewhere in Spain would be ~500 miles and off the continental shelf - a deep water tunnel. Although I expect a journey to Spain would be more easily routed through the Le Harve tunnel or current Chunnel than engineer a deep water tunnel. Perhaps in another hundred years or so engineers will work up to the challenge of a Boston to Bristol tunnel ... but they'll probably figure out something better first.

[george matthews]

Shorter long tunnels need to be built to build up people's confidence. Southampton (England) to Le Harvre (France) as a ~100 mile Channel tunnel would be a step. Plymouth to somewhere in Spain would be ~500 miles and off the continental shelf - a deep water tunnel. Although I expect a journey to Spain would be more easily routed through the Le Harvre tunnel or current Chunnel than engineer a deep water tunnel. Perhaps in another hundred years or so engineers will work up to the challenge of a Boston to Bristol tunnel ... but they'll probably figure out something better first.

I know all too well how difficult it is to make money. I lost a lot from the original share issue of Eurotunnel. There certainly isn't enough traffic for any more connections to the mainland from Britain.

A tunnel to Ireland is feasible but the traffic is unlikely to support it, even if built with sections lowered to the sea floor.

You didn't mention India to Sri Lanka, recreating a link that existed during the Ice Age, before the seas rose. In British times there used to be a coordinated train and ferry service. Here too I doubt if there is the traffic to support it.

[lpetrich]

A Bering Strait route? Most of it is very thinly populated. From west to east:

Western end: Sapporo, Japan: 1.91m, Vladivostok, Russia: 0.58m, Khabarovsk: 0.58m, Yuzhno-Sakhalinsk: 0.17m
Magadan: 0.099m
Uelen: 776, Bering Strait, Nome, AK: 3505
Fairbanks, AK: 0.035m, Whitehorse, YT: 0.023m, Prince George, BC: 0.070m
Eastern end: Vancouver, BC: 2.12m, Seattle, WA 3.41m, Edmonton, AB: 1.03m, Calgary, AB: 1.08m

Some additional routes:
Holyhead, UK - Dublin, Ireland: 110 km, depth 100 m
Seoul, Korea - Yantai, China: 500 km -- intended to bypass North Korea
Taipei, Taiwan - Ningde, China: 270 km -- many Taiwanese won't want their moat breached, however
India - Sri Lanka: 31 km

As to cost, I think that the most optimistic estimate for a long tunnel is from the Seikan tunnel at $120m/km. The Gotthard Base Tunnel is about $200m/km, and the Chunnel is at about $360m/km. So using the Seikan-tunnel estimate, a Dublin-Holyhead tunnel would cost about $13 billion.

Turning to a transatlantic tunnel, the best way to cut costs is to avoid tunneling as much as possible, which is why I'd posted on a route that goes through Canada and Ireland. Its underwater length, 3100 km, is a little over half the length of an underwater route between NYC and southern England.

With Seikan costs, it would be nearly $400 billion, and with Chunnel costs, over $1 trillion.

On the North American side, one can cross the St. Lawrence River at a much narrower point at Montreal or Quebec City.
NYC - Montreal: 560 km, M - Quebec City: 200 km, QC - Lodge Bay, Labrador: 1400 km, Total: 2200 km
With the cost of the LGV Sud-Est, about $7m/km, that's $15 billion.

On the European side, it may be better to go to Dublin than to Cork, but doing so means joining the West Coast Main Line between Birmingham and Manchester and continuing southeastward to London. In Ireland, one would go from northwestern County Mayo to Dublin, about 240 km, and in England, about 400 km. Optimistic cost estimate: $18 billion, though much of it will be for the Dublin-Holyhead tunnel.

[george matthews]

On the European side, it may be better to go to Dublin than to Cork, but doing so means joining the West Coast Main Line between Birmingham and Manchester and continuing southeastward to London. In Ireland, one would go from northwestern County Mayo to Dublin, about 240 km, and in England, about 400 km. Optimistic cost estimate: $18 billion, though much of it will be for the Dublin-Holyhead tunnel.

The Holyhead-Dublin tunnel would be pre-fabricated sections lowered into a trench. It would connect to a North Wales route upgraded by electrification. When Ireland was booming the government there wanted that line speeded up and electrified. I can't remember if they were prepared to contribute. It would probably meet the main West Coast mainline at Crewe. However, by then (the rather far future) there should be the other High Speed route now being planned. That would go from Birmingham to Manchester.

The Irish economy is at present in terrible condition with government bonds downgraded yesterday to "just above junk", that is they can't raise any money. It was hard to finance Eurotunnel (some of it at my expence as the shares I bought are worthless). But Eurotunnel is very busy. I cannot see the same kind of traffic demand from Dublin.

[lpetrich]

Dublin's population is 1.66m, and the entire island's population is 6.26m (Republic of Ireland: 4.47m, Northern Ireland: 1.79m). That's less than the population of either the London or the Paris metropolitan areas. So it's rather hard to justify a 100-km tunnel. Northern Ireland is closer; from the coast northeast of Belfast, the distance across the North Channel to Great Britain is as small as 34 km: BBC NEWS | UK | Scotland | South of Scotland | Bridge to Northern Ireland mooted Cost: GBP 3.5 billion, or $6 billion.

Back to the original subject, I recently discovered
EU could ground short-haul flights in favour of high-speed rail | World news | guardian.co.uk
I tracked down its source: Transport: White paper 2011 - European commission
From that document,

2.5. Ten Goals for a competitive and resource efficient transport system: benchmarks for achieving the 60% GHG emission reduction target

Developing and deploying new and sustainable fuels and propulsion systems

(1) Halve the use of ‘conventionally-fuelled’ cars in urban transport by 2030; phase them out in cities by 2050; achieve essentially CO2-free city logistics in major urban centres by 2030 [10].

(2) Low-carbon sustainable fuels in aviation to reach 40% by 2050; also by 2050 reduce EU CO2 emissions from maritime bunker fuels by 40% (if feasible 50% [11]).

Optimising the performance of multimodal logistic chains, including by making greater use of more energy-efficient modes

(3) 30% of road freight over 300 km should shift to other modes such as rail or waterborne transport by 2030, and more than 50% by 2050, facilitated by efficient and green freight corridors. To meet this goal will also require appropriate infrastructure to be developed.

(4) By 2050, complete a European high-speed rail network. Triple the length of the existing high-speed rail network by 2030 and maintain a dense railway network in all Member States. By 2050 the majority of medium-distance passenger transport should go by rail.

(5) A fully functional and EU-wide multimodal TEN-T ‘core network’ by 2030, with a high quality and capacity network by 2050 and a corresponding set of information services.

(6) By 2050, connect all core network airports to the rail network, preferably high-speed; ensure that all core seaports are sufficiently connected to the rail freight and, where possible, inland waterway system.

[10] This would also substantially reduce other harmful emissions.
[11] Cf. Commission Communication “A Roadmap for moving to a competitive low carbon economy in 2050”, COM (2011)112.

From the Guardian article,

The EU transport commissioner, Siim Kallas, has announced a series of green transport goals including shifting the majority of flights longer than 300km to rail and phasing out the use of petrol cars in city centres by 2050.

"At Heathrow there are no new runways, but we desperately need to increase capacity and you can do this if you reduce short-haul flight connections," said Kallas. The commissioner added in an interview with the Guardian that the UK should look at the example of Spain, where high-speed rail has hit demand on a previously popular flight corridor.

"This has happened in Madrid and Barcelona, where 50% of the market has moved to high-speed rail. It is comfortable for everybody. Airlines can put emphasis on long-haul flights, which is better for their business.

[george matthews]

Northern Ireland is closer; from the coast northeast of Belfast, the distance across the North Channel to Great Britain is as small as 34 km: BBC NEWS | UK | Scotland | South of Scotland | Bridge to Northern Ireland mooted Cost: GBP 3.5 billion, or $6 billion.

But the North Channel is unsuitable for building a tunnel. This has been researched. Not only is the rock dodgy but huge numbers of first world war munitions were dumped there.

My wife and I plan to go to Ireland later this year. We shall take the train to Holyhead and the ferry. The one way fare for this is pounds 33, including train and ferry. That fare is from any station in Britain.

[lpetrich]

Those are problems for a submerged-tube tunnel, but a sufficiently-deep bored tunnel will avoid those problems. But that's likely a more expensive option.

From South Korea, Railway Gazette: National plan to put cities 90 min apart

The Ministry of Land, Transport and Maritime Affairs has published plans for national network of high speed and upgraded lines which would put the most of the major cities within 90 min travel time of each other by 2020.

... The proposed network is based on ‘an X within a square’, with the KTX high speed lines from Seoul to Busan and Mokpo forming the centrepiece of the X. ...

Noting that air travel between Seoul and Busan fell by more than 30% in the first four months following the opening of KTX Phase II between Daegu and Busan in November 2010, ministry officials said the rail expansion package would reduce carbon emissions by 7·7 million tonnes a year and increase annual GDP by 91tr won.

[BostonUrbEx]

I think there's a lot of over thinking here. High speed rail will only compete with airlines for regional service for quite some decades (centuries, even) to come. No matter how much you want to eliminate gas guzzling airbuses, they're incredibly cost effective due to the minimal infrastructural needs. We can easily slash overall air traffic around the globe by perhaps 50%, but to take a train from NYC to London/Paris? Come on, let's be focus on getting NYC to DC up to par first. If you want to fantasize, that's great, perhaps a fantasy thread would be lovely, but this thread should be discussing legitimate options to compete with airlines.

New York to DC in 90 minutes should be our goal, or something entirely new such as LA to SF in 2 hours.

Even a straight shot from NYC to Paris at 300MPH would be 12 hours! How would you even maintain such a system? 6 hours for MOW workers to reach the center. And that's assuming the MOW equipment could even attain such speeds as high as 300MPH.