It’s Not A Bridge, And Not A Tunnel. Or, Maybe It’s Both?

The gist of the idea is to suspend an underwater tunnel from floating pontoons. By the time you finished reading that sentence, you probably already had a list of things in your head that seem to make this a terrible idea. After all, it does seem to combine the worst aspects of both underwater tunnels and bridges. But, the idea may actually be a good one, and it’s already being seriously considered in Norway.

You see, Norwegians have a fairly serious obstacle to long distance travel: fjords. The tradition way to cross a fjord is by ferry. One simply can’t ford a fjord – they’re just too deep. The deepest fjord in Norway, the Sogn, is almost a mile deep. That’s the same reason that a traditional underwater tunnel is impractical; putting a tunnel that deep underwater is inadvisable for a number of reasons, such as construction challenges, pressures at those depths, and unknowns about the seafloor itself.

A regular ol’ bridge could work, but presents difficulties of its own. The harsh weather would make travel across such a bridge dangerous, and you’d still have the age-old problem of getting ships under it. For those reasons, Norwegians usually stick to ferries for crossing their fjords – of which there are more than a thousand in Norway alone. But, an untested idea for a submerged floating tunnel may be just the solution they need.

It’s no secret that humans have an innate distrust of bridges, and that fear is dramatically amplified by the idea of being stuck under the surface of the frigid Norwegian Sea. But, if we can get past our fears, the proposed idea has a number of benefits. The tunnel being underwater protects it from the poor weather above, and ships would have no trouble passing above. As an added bonus, it maintains the picturesque beauty of the Norwegian countryside. However, it’s all for naught if no one will drive through it. Would you?

[via reddit]

94 thoughts on “It’s Not A Bridge, And Not A Tunnel. Or, Maybe It’s Both?

          1. Just to re-mix those, “XKCD, always self-righteous”. “A webcomic of smugness, out-pedanting pedants, and more smugness”. “And shitty stickmen”.

      1. Sigh.
        Speed limits in the tunnel could be posted in knots.
        If it sinks, it’s all for naught.
        I think it will not be built.

        I believe that unties the word knots above.

    1. I don’t know about the rest of you, but I would prefer to be corrected, in non-judgmental way as Bob did, than continue with my ignorance. If one doesn’t follow rules of grammar, eventually we will not be able to communicate. I know grammar usage changes over time, but not instead of naught is an obvious misuse of the word.

      1. While I can agree with the sentiment, I would like to point out (per your request) in the most non-judgmental way I can that this was not a failure to follow the rules of grammar, but, a failure of vocabulary and/or spelling in stead.

        Written language involves at least the four skills of grammar, syntax, vocabulary, and spelling. Using the correct word or not is almost always a matter of vocabulary. Comprising said word of the correct sequence of letters is Spelling. Grammar is the sequencing of or combining of words in an order such that they convey the intended meaning, ideally in a way that does not seem awkward or confusing to the recipient (reader in this case). As I understand it, Syntax is very closely related to grammar, but focuses more on the rules and the structure (including things such as punctuation) whereas grammar focuses more on ordering the words and having the clause make sense and convey the intended meaning.

    1. I’ll propose an answer to the rhetorical question. It probably offers access to electrical as well as an escape option were there to be a need. It looks like there is the hint of a platform shown at the far left of the image. There might even be shelter to help people withstand the weather while waiting to be rescued.

      1. More likely it leads to the tube for the opposite direction traffic which is likely part of the same tunnel complex and not visible in the picture. Likely this is via the center area which would, in fact, most likely be used for electrical and other utilities.

        Often on subterranean or underwater tunnel systems, such as the trans-bay tube for the BART system in the San Francisco Bay Area, these doors do serve an emergency exit purpose leading via a utility corridor to the tunnel normally used for opposite direction travel. Presumably the train causing the emergency is blocking the track on the emergency side, so a rescue train can be dispatched via the opposite tunnel and the people can traverse from one tunnel to the other via said door.

        Obviously in this case, it wouldn’t involve trains, but the overall principle is the same.

    2. To the tunnelbridge-tube beside the one you see.

      You know, they want to travel in both directions, so a tube for each direction if you as on the picture has two lanes in the same direction in one tube…

  1. The question is, why do they have to put the tunnels at the mouth of the fjords, can’t they move the highways a little bit up the creek where the fjords get narrower and the surrounding landscape gets higher, so they can build suspension bridges – and there isn’t the sort of maritime weather around.

    You know – don’t try to cross the river where it’s widest and deepest.

    1. Perhaps because this would require many miles of hazardous roadway be added to the journey to reach the bridge anchorages and then the bridges would be quite hazardous themselves due to the inclement weather that is common in Norway.

    2. Ever seen a fjord? They can go for miles with the same high cliffs and deep channels; it’s like being in a hallway for architecture. The maritime weather extends pretty much across the whole country and if you think you would be spooked by a floating tunnel try to imagine yourself crossing a quarter mile open span roughly the same height in the air with 70 mile an hour crosswinds and blowing sleet.

      1. yes, and the high cliffs are a perfect place to hang a suspension bridge high enough that lets ships pass underneath. Part of the description as to why they would sink the tubes under is so ships would pass freely, which implies the crossing is at the mouth of the fjord where the cliffs are low.

        For the high winds… isn’t it a bit obvious? Don’t have an open span – put a tube around it.

        1. That’s gotta have at least as many engineering hurdles as a brunnel below the draught of a modem freighter. Round objects in the wind like to behave as wings and sails.
          Sure, we’ve learned a few lessons since Tacoma narrows but strong sustained winds put huge stains on structures.

          1. Gotta love that aero-elastic flutter. Anyone remember Shawn Frayne? He was supposed to build a power harvesting device from this concept and then he just disappeared. This was several years ago. The wind belt generator I believe.

        2. I’ll just assume you didn’t read the whole article prior to your first comment as you completely missed the weather part. Your second comment caught it, but the engineering required to suspend a tube in the air hundreds of feet above a wind-swept inclement-weather-prone fjord would be daunting, to say the least.

    3. Because the mouth of the fjord is where most of the people are. Also, the sides of the fjord above the water are often just as steep as those below the water. I.e. building the road back to where the fjord is narrow enough to bridge will probably require nailing a highway to the side of a cliff for many miles.

      1. Point taken. Would be a bit of a roundabout.

        They’d need to make a miles long grade uphill along the side of the fjord to the crossing and then another downhill to the other side. Would be awesome for longboarding though.

  2. If they put them anywhere else, they would spoil Slartibartfast best work.

    “You know the fjords in Norway? I got a prize for creating those, you know.”
    -Slartibartfast [hitchhikers guide to the galaxy]

  3. Saw an article decades ago in Popular mechanics with a similar tunnel but it was moored to the bottom so it didn’t flow up. This one needs to have negative buoyancy so if something breaks off or there is a leak, it’ll sink! Scary as hell. What about ice and tide?

    1. If a bottom-moored tunnel breaks and gets a leak, it will quickly sink because it gets filled with water. There’s nothing keeping it up once it starts taking water.

      A float moored tunnel can at least be kept up by using oversized floats that are partially filled with water, which would keep them more stable in the waves to start with and allow them to be inflated more if the tube below loses buoyancy.

    2. Well it could be made to be slightly negatively buoyant for its depth. It will however be a tube filled with air, so I’d guess it would have to use some kind of ballast to retain its negative buoyancy. That ballast could be released in the event of an emergency.

  4. Bad idea…ships will hit it if it is close to the surface, and if it is deeper the British will just crash their most expensive nuclear sub into it… This is the kind of project that will make a few contractors and politicians rich and never be finished.

  5. why float it on pontoons at all? design it to float at the required depth. you already need to pump air through it so people can breathe so adjust the air pressure to counter the varying load.

    1. I think that a mechanism similar to a floating dock interface to land will be used at each end. These are well understood and well known structures that are not difficult to implement and do not suffer abnormal wear and tear from tidal processes.

      As to keeping it at a particular depth through buoyancy control, you are clearly underestimating the difficulty and risks associated with such an endeavor.

      First, Buoyancy is not from pressure, it is from displaced volume. The reason a diver has to add air to his BCD or Dry Suit as he descends is to keep the volume consistent in order to keep his buoyancy neutral. If there were such a thing as an incompressible lightweight fluid, then the fluid could be loaded into the BCD to the point of neutral buoyancy and the diver would be able to move anywhere in the water column without requiring additional adjustment.

      Inflating a plastic bag (BCD) or a rubber sack (dry suit) is relatively easy. Inflating and deflating a concrete, iron, or other rigid structure is stressful and very limited in its elasticity.

      1. For first approximations Boyles law can be used, where pressure is inverse to volume, so it’s not unreasonable to talk in terms of pressure for this example.
        More specifically buoyancy is a function of relative density.
        So, getting the structure to near neutral buoyancy would be the goal so your control system doesn’t have to move as much gas to moderate its depth. At any rate the structure is them at the whim of the power grid. A passive design is ideal.

        1. Boyle law only works if you have an expandable (non-rigid) container (e.g. plastic bag). If you try to add air to a gas cylinder, for example, the cylinder expands very little, and instead, PV=NRT (ideal gas law) takes over and instead with the increase in N, you gain P and T rather than V.

          While you are technically correct about “relative density”, the fact of the matter is when displacement is a fixed value, then weight becomes an equivalent term to relative density because one need only compare the weight of the water displaced to the weight contained within the displaced volume.

          An additional complicating factor is that fjords are seasonally affected tidal estuaries. (The salinity, especially near the mouth, is highly variable, meaning that the weight of the displaced water is not constant).

          The difference in weight between a steel SCUBA cylinder pressurized to 15 PSI (1 bar) and one pressurized to 3000 PSI (200 bar) is roughly 6 pounds. Since we’re talking about human exposure here and not wanting to give them a nasty case of the bends, you probably don’t want to get much beyond 2.5 atmospheres for safety. Let’s assume a 5 mile long tunnel consisting of two cylindrical spaces 40 feet in diameter joined by a 20×20 foot rectangular utility/emergency/maintenance passage.

          The total area of a cross section is 2π(20^2)+20*20 =2π(400)+400 = 6.3(400)+400 = 2,520+400 = 2920.
          We can now compute the volume (I’m being generous with the value of π and not pretending the entire space is air, not accounting for additional displacement for walls, etc., so these are very idealistic (unrealistically so) numbers) as the length (5*5280) multiplied by the area of the cross-section (2920). 5*5280*2920=77,088,000. For convenience, let’s just call that 77 million cubic feet.

          At 1 atmosphere (we’ll assume standard temperature and pressure for now), 77 million cubic feet of air weighs approximately 6.2 million pounds. If we pump that up to 2.5 atmospheres, we’ll reach approximately 15 million pounds.

          The weight of the same volume of fresh water is 4.8 billion pounds.
          The weight of the same volume of salt water is 4.9 billion pounds.

          So the difference between fresh and salt water alone is more than 11 times the maximum weight difference you can achieve with air pressure in the tunnels.

          To make matters worse, if you do not change the weight or volume of an object, it’s buoyancy will not change with depth. (At least not significantly). Since water is a relatively incompressible fluid, it doesn’t get denser with pressure (at least not much). Boyle’s law says that, if anything, as an object descends (external pressure increases), it will shrink, thus exacerbating the negative buoyancy of the object. The reverse is also true. If it starts to float (exhibit positive buoyancy), then it will tend to expand on the way up, thus exacerbating its positive buoyancy tendencies.

          So, again, no, pressure doesn’t help here. It just creates additional strain on the concrete (and the automobiles and passengers).

          You could use inflatable bladders attached to the tunnels with extremely rapid controls, but even then, when you consider the vertical momentum of the tunnels, tidal and current effects, other lifting forces, the variations of weight in different sections as vehicles pass through, etc., it would be extremely complex to manage the inflation and deflation of these bladders in a manner that didn’t lead to substantial oscillations.

          You are talking about an incredibly complex control system that would have an excessive momentum and would have negative static stability characteristics. (If you don’t understand negative static stability, think of a wrench hung from a pegboard hook. If you hang the wrench downward from the hook, you have positive static stability. (it will tend to return towards its original position after being perturbed). OTOH, if you carefully balance the wrench pointing straight up from the hook, it will have negative static stability… It will tend to amplify even a small perturbation from it’s original position.)

          A passive system is much simpler, much less expensive, and much more reliable.

          1. How else would you control buoyancy if not air bladders or ballast tanks?
            You’re certainly not going to be able to build a brunnel that you can pressurize, yet still accommodate a reasonable pace of traffic. Nevermore the bends or the impracticality of an airlock for traffic, why would you put internal combustion engines in a sealed container with people?
            A brunnel with active buoyancy would by necessity have to borrow from ROV and submarine technology. Ballast tanks that flood and drain as necessary.

            As for salinity concerns, it’s trivial to track the halocline and simply build a sufficient distance away from it so the brunnel is either always fresh or always salty. There are likely fjords that don’t follow this regime which would limit a brunnels suitability here. Even then ballast could be added to account for seasonal salinity differences if they’re stable enough.

            As I said in my first post, the amount of energy to accomplish active buoyancy control year-round, makes this solution impractical at best. However an interesting thought experiment.

    2. Keeping it neutrally buoyant to a specific depth would be a nightmare.

      There’s basically two different proposals: Make it negative buoyant and have large floats, and make it positive buoyant and anchor it to the bottom. In the latter case it would be completely stationary, even with a tide (the height above it would change, but the distance to the bottom would be constant).

        1. Actually, they don’t. Submarines use salt water ballast tanks that they pump water in and out of in order to stay somewhere near neutral buoyancy and then they use thrusters, control surfaces, and motor(s) to keep the submarine in a desired attitude at a desired depth.

          Once something is neutrally buoyant, it will stay neutral at any depth unless volume or weight changes.

          Submarines are constantly changing depth and then compensating. That’s fine if you’re not attached to land at both ends, but it’s quite a different problem when you are tethered. When docked, submarines are kept positively buoyant so that they float at the same level relative to the floating dock they are attached to.

      1. Yeah, especially if there’s a river emptying into fjjord and salinity can vary with rainfall and wind shoving seawater against flow so that the actual specific gravity of the water varies.

    3. Isn’t this kind of a moot point? In a submerged structure, maintaining an altitude is going to be dependant on the height/properties of the water column ABOVE it, not below – So regardless of whether or not a pontoon system is used, the same variables will be at play. Just a thought?

  6. I guess it somehow solves the temperature and the snow problems. Winters in Norway wouldn’t be anything to joke about. I wonder if the temperature under water would remain steady above 0°C even in winter. Would icebergs be a problem? Will the sea freeze?

    1. Water’s an unusual substance, in that it’s most dense at 4C. So 4C water sinks. That’s why bodies of water freeze over on top, but stay liquid beneath. It’d have to get really, really cold to freeze the whole sea solid. And there’s the whole ocean flowing about to distribute temperature.

      As far as icebergs, I suppose you just build it where there aren’t any.

  7. The norwegians are really good with building a road where most wouldnt really consider it or it being that usefull. THey have some really impressive roads, bridges and tunnel solutions. A 3-5km tunnel isnt really something you consider special after a driving a lot in norway (there is a 5.5km, and a 11km tunnel before/after laerdal tunnel, 24.5km), and they have several 7km+ tunnels that goes down 200m under fjords. See: and

    But they have big cities at the mouths of fjords and in the archipelago, and it takes a long time to drive along the roads winding there way in/up the fjords and out the other side. Quite often the fjord is too wide and to deep at the mouth for a bridge. Thats why there are plenty of ferries, and they take time, limited throughput over longer distances, and weather is a factor.

    Another problem with norway is they tend to focus om building the impressive projects, not the most usefull. The road from Hardanger bridge to Odda? Road 13? It has places where two trucks/lorries/busses cant pass eachother, one have to stop and reverse… with other vechicles piling up behind both directions. Norwegian roads can be a horrible with summer traffic and good weather just because of this. Going short distances can take a long time.

    If I need to drive to Narvik i northern Norway, I dont use the E6 (European way 6), I go thru “smaller” roads in Sweden. Its faster and much less frustrating. And E6 is the main road going thru Norway all the way to Russia. If you can choose a “smaller” road over a mountain its often faster and better than following the valleys.

    And you dont want to get a speeding ticket in norway.

  8. One of their Fijords is partially blocked from the wreckage of the first Troll gas drilling platform. There were going to be two of the massive, bottom resting, all concrete platforms.

    When the base of the first one was mostly complete, a submergence test was done. *CRUNCH* It imploded and the debris hit bottom so hard it set off a minor earthquake.

    After some ohshitting by the engineers the designs were adjusted. Troll #2 was built and tested without incident. It had to be maneuvered around the debris of #1 on its way out of the Fijord to its resting place in the north sea.

  9. The article’s premise is that to go from Kristiansand to Trondheim, you need 21 hours. $25 billion will get you there in 10 1/2. Google says, sure, but if you just take the inland route, you’ll get there in ten hours, anyways.

    I could see trying a tube from Stavanger north and eventually connecting to Bergen, but I doubt there’s a pressing need beyond that.

  10. Aren’t fjord the place where half the world hides its submarines? Seems a bit tricky collision-wise.

    But if they insist, why not also make it a vacuum and turn it into a hyperloop.

  11. While we’re calculating air mass and salinity and bouyancy and whatnot, how about we take a moment to consider the *vehicles*. The dynamics between rush hour and late night have got to be 2 or 3 orders of magnitude greater than any of those other factors.

    Also, how exactly do you make a water-tight tunnel for road vehicles that is limber enough to handle shifting in three dimensions with traffic, tides, displacement, etc?

    1. My thoughts were along these lines to.

      Without rigidity, imagine a large truck going through. To maintain perfect buoyancy you would have to be continuously moving ballast away from it in the same direction. It is much easier to move a truck rolling on wheels than it is to move water through pumps.

      Now, of course, there must be some rigidity and ballast movements only have to average the buoyancy due to the rigidity factor but still so much energy would be required to keep this thing stable that the energy cost alone would be prohibitive to the consumer who would be expected to pay the toll.

  12. What on earth would it need pontoons for? Surely displacing that mass of water would mean the structure would already be too buoyant. An inverted suspension bridge would be more appropriate.

  13. I’d suggest both suspension from floaters and anchorage by weights. Make the tunnels slightly buoyant, cable them down to masses sitting on the floor, and connect them by cables or normally sealed tunnels to floats on the surface. Vulnerable to both surface ships and submarines, but neither alone need be the cause of immediate destruction. Tunnels to floaters provide some possibility of escape and means of maintenance access, otherwise I’d say just dispense with them altogether and anchor the tunnels to the bottom. Catastrophic flooding may not be much better than sinking. Provide escape pods attached to the tops of the tunnels.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.