The Future Of Canal Transport–Take The Elevated 2: A Use For 4-6 Kilometer Tall Towers–Sky Canals And Superrapid Transport ETT

A guest article by Joseph Friedlander

This is  a sequel to
In my post on supertall towers
it was obvious that there was an awkward spot in the low supertall tower range.  Above a kilometer (pretty much today’s capability and below the tens of kilometers– there were few compelling and unique applications.

Well, I was thinking about the canal post and the tower post and it was light bulb time. What if we could build a world-wide network of transport canals which would be at high altitude for level travel on barges as well as contain tubes for anything from super-rapid vacuum transit and other cool engineering uses?

I have been accused of designing hundred billion dollar butter churns. I think this is more like a 8 trillion dollar butter churn but  let’s analyze the system to see where it goes.
A reminder about the prototype elevated canal (on mere tens of meters elevation to ground level) to save you going to the previous article: Highlighting in blue:

The Magdeburg  Canal Bridge in Germany might technically be called a navigable aquaduct but it differs in both feel and practicality from earlier efforts. Compare the smaller ones shown on this web page

with the Magdeburg water bridge and you will be comparing a toy tool with a professional power tool.

Germany: The Magdeburg Water Bridge – Wasserstraßenkreuz Magdeburg

Facts about Magdeburg  Canal Bridge
  • six years to build
  • cost of 500 million Euros
  • 918 meters long
  • 545000 Euros per meter
  • Width   34 m
  • Water depth     4.25 m
  • Longest span    106 m
  • Total length       918 m (690 m over land and 228 m over water)
  • Clearance below           90.00 m x 6.25 m
  • 68,000 cubic meters of concrete and 24,000 metric tons of steel
  • Connects Hannover and Berlin directly
  • Connects Berlin’s inland harbor network and Rhine river ports.

This being Next Big Future, we take the obvious and intuitive step of wondering what would happen if this sort of thing became widespread.

Forget 6 years for a unique product to be engineered and built, and postulate Broad Group style mass manufacturing and rapid deployment.

As built the 4.25 water depth is larger than many areas with rough bottom on the Mississippi, for example. But the more reliable clearance on the bottom can enable larger cargoes with confidence.

Although our fantasies would like a super canal bridge able to float the heaviest supertankers, don’t forget a mere 34 m wide bargeway can float at a guess ten thousand ton barge clusters, depending how straight it is and if you operate it as a single lane vs double lane highway. Barges can be grouped and assembled into trains with multiple tugs providing huge power.

Assuming 5 meters a second throughput, 200 seconds to clear each kilometer span, and one 10,000 ton barge per interval, we see that a (for example) 100 kilometer stretch of such elevated canal could move a million tons of cargo each 6 hours. Individual barge clusters could be assembled that could hold not merely a Saturn 5, but a Nova or possibly a Sea Dragon.

As for the cost per meter, this was a unique project, with all charges piled on top, the engineering, the extra costs of a flagship project, and so forth. We can imagine 200,000 tons of structure a kilometer being built for 200 million Euro. This is still a lot but one can imagine the economic benefits to many small countries which could use a few hundred kilometers of such “navigable river’ say from a seaport to an interior city.

To go higher of course you would build locks that would be elevators…
OK, we’ve had our flashback. Now we can consider reader comments from the previous post.  Some pointed out correctly that there are other means than simple flooding locks to raise a ship.  The Falkirk wheel was specifically mentioned.
The allowable dimensions of the Falkirk Wheel is as follows:

Length 70′ 0″ 21.33m Beam 19′ 8″ 6.0m

Headroom 9′ 0″ 2.74m

Draught 4′ 6″ 1.37m

The caissons can hold 300 tonnes of water so I guess the maximum weight of a craft would be around 200 tonnes

Quite true, as an illustrative guide we can imagine a big pan with four cables on it into which a barge sails and which is then locked against drainage and then the four cables are reeled in till the barge is at the tower level– 4 kilometers up!

A mention about the altitude:  Why did I pick 4-6 kilometers altitude?  As an optimax:  If you have a worldwide network of 4 km high pillar supported elevated canals, you want a height that will allow essentially straight line travel between major nodes in the network of cities. Or at least great circle routes.  If you look at a map of elevations you will see that above 4 kilometers there are essentially two big obstacles in the world– the high Andes and the high Himalayas– like islands to be avoided.  There are individual peaks elsewhere but no huge ranges that high. Why not build 9 km high and avoid every possible obstacle and butter our toast on board a barge overlooking Mount Everest?  Um, because then the system is over twice as expensive. I’d rather have a hugely more extensive network than a brute force conquest of the Earth just to prove an engineering point. Also at 6 km you can just kinda breathe if you don’t exert yourself, at 8 km plus you are in the death zone without oxygen to supplement your breathing. We could build with iron easily up to 4 km, to 8 km would involve tapering because iron’s strength of materials limits are being reached. And so on. Extreme altitude is not free. It could be that a 3 km altitude route can be found across most of the high traffic world, or even 2 km (if you build across the oceans).  Certainly an equatorial routing, while adding to the kilometer count, would enable wonderful ballistic paths to many space trajectories. The actual length of the 6 km segments might be quite limited.

There is a company named ETT  that has posted online dreams of a world wide vacuum tube network.

There is a popular video by Next Media; (see: ) that claims to be about our Evacuated Tube Transport ™ technology; however the next media video has many gross inaccuracies. Next Media never contacted us or consulted with us about the production of the video, and use of our tradename was not authorized.
Some of the next media inaccuracies are:
The capsules are depicted larger than the optimal 1.3m (51in) diameter X 4.95m (16.24ft) length; The tubes are shown larger than the optimal 1.5m (5ft;) diameter; The video fails to show the airlocks that prevent air from entering the system; The video indicates the use of impractical routes crossing large expanses of ocean; The video uses our tradenames without attribution.–Ofo
ETT Patent

Notice the 4000 mph projected top speed– in a magnetic levitation vacuum system –this picture from Salter (below) not ETT

you can actually go to orbit or even escape velocity (if you don’t mind plenty of Gs and superwide curves) because you get regenerative power back in braking.  And if you have an escape tube pointed to the sky, during idle times (power supply  and tube safety considerations) and in remote locations (sonic boom considerations )  you can launch huge uploads directly to space for very little added cost. That is the real killer app of this kind of network– blasting through the atmosphere at escape velocity is far easier going from vacuum to 4-6 km altitude than trying to do it at sea level. (at the upper end of that range you are above half the atmosphere by mass)

Note in the 1970s there was a concept caled PlanetTran detailed here  which was the work of Robert Salter

This should not be confused with the San Francisco website of the same name 
Planet Tran brings you SmartTransport: reliable, convenient and eco-friendly chauffeur-driven transportation in greater Boston and the San Francisco Bay area.

which brings to mind Jerry Pournelle’s definition of a dark age as not that you know you could no longer build that but you never knew it was once possible.  As you’ll see from the map below, once it would theoretically have been possible to travel from Boston to the Bay area in 21 minutes. 

 On the general subject of an old techie grousing I remember when Beechcraft came out with a light plane called the Starship.    No. Sorry. You want a starship, build a starship.–and it better be capable of a hop to Alpha C or you’ll be hearing from me. Don’t call a light plane a starship. Don’t call a limousine service PlanetTran unless you are chauffering around Supercar  But I suppose I digress…

More on the vaccum subway concepts of Robert Slater Space launch version note that 6 km is their selected exit height–
In the reference design, the exit is on the surface of a mountain peak of 6,000 metres (20,000 ft) altitude, where 8.78 kilometres per second (5.46 mi/s) launch velocity at a 10 degree angle takes cargo capsules to low earth orbit when combined with a small rocket burn providing 0.63 kilometres per second (0.39 mi/s) for orbit circularization. With a bonus from Earth’s rotation if firing east, the extra speed, well beyond nominal orbital velocity, compensates for losses during ascent including 0.8 kilometres per second (0.50 mi/s) from atmospheric drag.[1][13]

A 40-ton cargo craft, 2 metres (6 ft 7 in) diameter and 13 metres (43 ft) length, would experience briefly the effects of atmospheric passage. With an effective drag coefficient of 0.09, peak deceleration for the mountain-launched elongated projectile is momentarily 20 g but halves within the first 4 seconds and continues to decrease as it quickly passes above the bulk of the remaining atmosphere.
In the first moments after exiting the launch tube, the heating rate with an optimal nose shape is around 30 kW/cm2 at thestagnation point, though much less over most of the nose, but drops below 10 kW/cm2 within a few seconds.[1] Transpiration water cooling is planned, briefly consuming up to  100 liters/m2 of water per second. Several percent of the projectile’s mass in water is calculated to suffice.[1]
Gen-2 variant of the StarTram is supposed to be for reusable manned capsules, intended to be low g-force, 2 to 3 g acceleration in the launch tube and an elevated exit at such high altitude (22 kilometres (14 mi)) that peak aerodynamic deceleration becomes  1g.[1] Though NASA test pilots have handled multiple times those g-forces,[17] the low acceleration is intended to allow eligibility to the broadest spectrum of the general public.
With such relatively slow acceleration, the Gen-2 system requires 1,000 to 1,500 kilometres (620 to 930 mi) length. A 280 megaamp current in ground cables creates a magnetic field of 30 Gauss strength at 22 kilometres (14 mi) above sea level (somewhat less above local terrain depending on site choice), while cables on the elevated final portion of the tube carry 14 megaamps in the opposite direction, generating a repulsive force of 4 tons per meter; it is claimed that this would keep the 2 ton/meter structure strongly pressing up on its angled tethers, a tensile structure on grand scale.[3][21] In the example of niobium-titanium superconductor carrying 2 x 105 amps per cm2, the levitated platform would have 7 cables, each 23 cm2 (3.6 sq in) of conductor cross-section when including copper stabilizer.[4]

The Generation 2 goal is $13,000 per person. Up to 4 million people could be sent to orbit per decade per Gen-2 facility if as estimated  Startram website. Note the inventors also worked on the MITEE concept for nuclear rocket engines

At highest speeds the G forces in turns would have been a real consideration.  By the way about the time of the 1991 Los Angeles riots my joke about the future of civilization was something like this,”Do you wonder at what an amazing culture we live in? You can flee underground in LA to escape the riots and be mugged and left for dead just outside the Times Square Station a mere 21 minutes later!” 
Slater’s map above was based on straight-line routes and tunnels but I am pretty sure with an elevated tubeway under the elevated barge canal you could get even better routings between major traffic modes.

Note that as Salter correctly notes a chord of the Earth’s circumference between  LA and New York would have a nadir 200 miles beneath the Earth’s surface at midpoint– or 200 miles above on high towers. Maybe another article– what it comes down to is these are straight line routes but not true ballistic courses and there is considerable G force in cornering. At 5500 mph a 400 mile turn will give you a 1 G force.

 5200 2600 cubic foot a minute roughing pumps would be needed to maintain the vacuum over 5000 miles.

Here is a picture from the ETT  patent showing near great circle routes which I have modified a bit for illustrative purposes Notice that there are great circle routes available between India and China that avoid the 4 km plus area of the Himalayas.
Notice too that even though they avoid the oceans if you are building 4 km high towers there is no serious problem to building them to the bottom of the ocean (various buoyancy options are available subsurface so the height/strength of material problems begin not from the bottom of the sea but when you break surface. On the other hand you need to shield against wave action so nothing is free)
Why not go down to ocean level and tow the barges there?  Well, if you have the postulated network you avoid the storms and collision problems at the ocean surface.

This is not a trivial concern in heavy winter seas  Even if you had the icing at altitude you would not have the heaving that could break your ship in two in the right circumstances. (But that wind at 4-6 km would be wicked cold…)

 If the network flows smoothly and quickly there are often advantages to staying in the network. As I said in the first article

As built the 4.25 water depth is larger than many areas with rough bottom on the Mississippi, for example. But the more reliable clearance on the bottom can enable larger cargoes with confidence.

Although our fantasies would like a super canal bridge able to float the heaviest supertankers, don’t forget a mere 34 m wide bargeway can float at a guess ten thousand ton barge clusters, depending how straight it is and if you operate it as a single lane vs double lane highway. Barges can be grouped and assembled into trains with multiple tugs providing huge power.

Assuming 5 meters a second throughput, 200 seconds to clear each kilometer span, and one 10,000 ton barge per interval, we see that a (for example) 100 kilometer stretch of such elevated canal could move a million tons of cargo each 6 hours. 

10000 km across the ocean could then theoretically move a billion tons of cargo plus a week– but there are only maximum 10s of billions a year moved theoretically so yes one canal across the ocean could theoretically move everything that has to be moved. The question is how cheaply but remember that we might be able to use electric power in place of petroleum for propulsion.

data on canal depths and navigability for barges
Class Tonnage (t) Draught (m) Length (m) Width (m) Air Draught (m) Description
Class III 1,000
Class IV 1,000–1,500 2.5 80–85 9.5 5.2–7.0 Johann Welker[1]
Class Va 1,500–3,000 2.5–2.8 95–110 11.4 5.2–7.0–9.1 Large Rhine[1]
Class VIb 6,400–12,000 3.9 140 15 9.1 [1]
Class VII 14,500–27,000 2.5–4.5 275–285 33.0–34.2 9.1 [1]
Classification Tonnage (t) Length (m) Breadth (m) Draught (m) Air Draft (m) Notes
RA 5.5 2.00 0.50 2.00 “Open boat”
RB 9.5 3.00 1.00 3.25 Cabin cruiser
RC 15.0 4.00 1.50 4.00 “Motor yacht”
RD 15.0 4.00 2.10 30.00 “Sailing boat”
I 250–400 38.5 5.05 1.80–2.20 3.70 “Péniche”
II 400–650 50.0–55.0 6.60 2.50 3.70–4.70 Euro-barge
III 650–1,000 67.0–80.0 8.20 2.50 4.70 “Gustav Koenigs”
IV 1,000–1,500 80.0–85.0 9.50 2.50 4.50; 6.70 “Johann Welker”
Va 1,500–3,000 95.0–110.0 11.40 2.50–4.50 4.95; 6.70; 8.80 “Large Rhine”
Vb 3,200–6,000 172.0–185.0 11.40 2.50–4.50 4.95; 6,70; 8,80 1×2 convoy
VIa 3,200–6,000 95.0–110.0 22.80 2.50–4.50 6.70; 8.80 2×1 convoy
VIb 6,400–12,000 185.0–195.0 22.80 2.50–4.50 6.70; 8.80 2×2 convoy
VIc 9,600–18,000 270–280 22.80 2.50–4.50 8.80 2×3 convoy
9,600–18,000 195–200 33.00–34.20 2.50–4.50 8.80 3×2 convoy

VII 14,500–27,000 285 33.00–34.20 2.50–4.50 8.80 3×3 convoy

Wiki: This map of shipping routes illustrates the relative density of commercial shipping in the world’s oceans

Now imagine how that would change if point to point routing between major supply/demand nodes were possible.

Reader  Andrew Jones correctly pointed in the original article showing possible canal routings around the world–

Why nothing in the Southwest. I am pretty sure those dry states would appreciate some extra water and think it mighty nice to be connected to that Midwest area. There is some heavy mining in those areas, especially construction grade stones, which is not feathery light.
  • The simple answer is that in the previous article I was focusing mostly on the possibility and starting slow to get peoples minds used to it. But the key problem with a high altitude system is not merely altitude (high tower design, elevator issues) but also temperature and icing issues, exactly like those old 40’s and 50’s movies where the propeller plane flying at low altitude (4-6 km encounters icing and is running out of de-icing fluid)– This is not as big a deal as on an aircraft because you don’t need to worry about
  • friction and drag (increased fuel consumption) unless of course the canal surface freezes up– thus the heating/cover canopy.
  • loss of airspeed and lift and maneuverability
  • increase of weight from ice

but you do need to worry about loss of engine power if using combustion for fuel as well as general icing issues (walking a slippery deck)

Without heating of the canal (and an overhead canopy to cut down radiation into space at night you would get this kind of icing (not to mention freezing of the canal itself)  

Reader Chris Phoenix suggested

2 years ago
OK, we have an idea, now let’s optimize it. 1) Don’t use tugs – use mechanical linkages to the side (cables, rows of gear teeth). More efficient traction. 2) Water has one main advantage. Low static friction. It has several major disadvantages: It’s heavy, and it has high dynamic friction, and it’s corrosive, and it promotes growth of unwanted organisms, and it freezes. So why not just build a surface for an air-bearing? Those use very little energy, can support immense weight, and can sustain far higher speeds. Basically, I’m suggesting building tracks for hovercraft – except that the hovercraft would be optimized for high load and low skirt leakage. A barge is very low-tech – just an open-topped box, in principle. But the devil is in the details, including maintenance. An air-bearing slab would be pretty low tech also, by today’s standards. Or we could turn it around, and build the air supply into the surface, like an air hockey table. Then anything flat you put on the surface would float. Simple sheets of iron, like those they use as temporary patches in road work, could be your “barges.”

Yes, around 1962 CERN was moving very large machines on air bearings, but because of the icing problem at altitude it was not clear to me we could keep the roadway effectively level. If totally enclosed, yes. If open– no and I am not sure what the final configuration could be. 
We can imagine a dehumidified enclosed pathway at altitude but I am thinking of worsecasing, such as weather damage that rips off the roof at one point and that icing incident blocks the canal till you clear it.  If that can be avoided We can imagine covered roadways of such precision that the sea barges are floated to them then lowered on the ‘hoverbarges’ which are elevator-ed to altitude. That hockey table like surface by the way was used for a hospital bed to ‘float’ burn victims.  A lot of clever thinking went into that…

Assuming it is a canal–if I am wrong that’s a whole different article–with the superhigh canal system  we can go in straight lines from most destinations to most destinations. The Western USA may indeed be crossed back and forth with many routes to bring access to vast amounts of minerals..

Interesting data on canals and shipping  in general

So what is this system so far? It is far more involved than the system proposed for individual countries in the last article.

  • Supertall towers in vast numbers elevating bargeway canals world wide
  • Molten salt reactors and electricity cables (superconducting?) to distribute power in a worldwide network. The waste heat helps keep the canal deiced (note that it may still have to be enclosed with an arched roof) And the inter-hemispheric power exchange makes possible load levelling around the world.
  • Great circle routing to give the capability of suspending vacuum subways underneath the canal And several space launch stretches of the network optimized between orbital track, (for example to low equatorial orbit) power availability and network idle spaces (where good launch times are available a good fraction of the time without interfering with the network flow)
  • Note also we use nuclear power for intercontinental oceanic and air travel and space travel without actually putting risky nuclear motors on the air vehicles themselves. This is a much better solution for safety engineering than trying to make a nuclear powered launch vehicle. It also cuts substantially into domestic air and surface travel, greatly conserving petroleum resources. (If you want a justification for spending $8 trillion on this that is basically it)

Now we come to questions of cost. The original prototype with all the custom expenses of a unique engineering project was 500 million Euros for a kilometer with all charges piled on top, the engineering, the extra costs of a flagship project, and so forth.
Forget 6 years for a unique product to be engineered and built, and postulate Broad Group style mass manufacturing and rapid deployment.
Supposing for a second that price held for 4-6 km high towers (in mass production and well amortized) then we would be looking at 8 trillion Euro for a 16000 km system.
Supposing the cost were not 500 million euros per kilometer but 5 times lower still (through a combination of vastly cheaper raw materials and huge mass production efficiencies) then such a map as we see in this article might be practical. 80000 kilometers of canal then would cost 8 trillion Euro for such a network.

Remember from the first article that the average height of the continents is a mere a mere 840 meters 2760 feet and we can understand that in each continent there will be numerous local systems interfacing with the interconinental system at equal speed but lower altitude. One can see the North European plain having hypersonic transport from Paris to Moscow, but also a line that can go intercontinental nearby, perhaps in London.  One can imagine the horsetrading that will go on if this network is ever actually built.  

As for material considerations We might imagine a pair of towers each 300 meters each together massing 100 kilotons or about 300 kilotons a kilometer. For 10000 kilometer stretch, 3 billion tons of steel. Remember that this is a multi decade infrastructure build and you will see that that is not out of the question. Especially as China completes her infrastructure build and looks for new markets for metal.

 The map is based on the NGDC world elevation map at This was the low altitude map from the first article

(For geography fanatics a 13 megabyte enlargement of the underlay map is available at
and here is a 2-4 km altitude optimized routing–note the equatorial stretch between Brazil and Singapore– frankly optimized for the space launch route (no you don’t need that length, but if you need a transoceanic route with good sunlight anyway…) the routes are debatable.
Here is the ETT map again showing a speed optimized route suitable for 4-6 km altitude–I frankly think it is easier to go across the equatorial ocean than to battle permafrost in foundations across Siberia but I don’t want to insult our Russian Alaskan and Canadian readers so I have included the routes even though a hub and spoke across the oceans may turn out to be the best routing of all.  Who knows? The idea of the concept is to get us thinking.
{Rather than trying for a high altitude route across Siberia the best Siberian development plan would be a series of low-altitude towers working with the natural river system and connecting resources to points on those rivers.  Such a system could aid enormously in developing the mineral wealth of Siberia at about 1% the cost of a grand plan. But that is another article.) 
 In the last part of this series we discuss an out of this world idea that makes this one look tame—- the building of liquid sodium elevated canals on the Moon itself. Of course  we would need the space launch system in this second part to build it:)
When will the next great era of canal building begin?
If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks