Undersea Cable Repair Via 19th Century Tech

An anonymous reader writes to mention a story going across the wires about an old-fashioned way to fix a modern convenience. Taiwanese boaters are using simple hooks to fish up the fiber-optic cables damaged in an earthquake late last year. The outage that resulted kept millions of users offline in half a dozen countries around the Pacific rim. From the article: "They work 24 hours a day but the weather can hinder their progress. Walters said one ship is waiting for 30 to 40 mile-an-hour winds (48 to 64 kilometres- an-hour) to die down in the Bashi Channel. The winds have stirred up 10 to 12 metre waves ... After arriving at the scene they survey the ocean bottom to assess whether the contour has changed, and the degree of sediment movement. Then the traditional tools are brought out. A rope with a grapnel on the end is played out, down into the depths, and towed over the sea floor until tension registers on a graph on the ship, indicating contact has been made with the cable. Today's fibre optic cables are just 21 millimetres in diameter."

Comments

No Electronics?

[SageMusings]

I beg to differ. There are devices known as OTDRs (optical time-domain reflectometers). Essentially, you shoot a pulse of light down the carrier and start a high-speed counter. The difference in refraction (say, a break in the cable or the end of the cable) causes a reflection that is detected at the device. Using the elapsed time, you get the distance to the break.

Check out this link http://en.wikipedia.org/wiki/Optical_time_domain_r eflectometer

I also saw a documentary on the Science channel about these ships. The whole process of fixing the break is sterile and professional. They use fusion splicers, which fuse the two ends with an electric arc. Fascinating stuff.

Re:No Electronics?

[rjforster]

Disclaimer: I used to build the pump lasers for the submarine repeaters so I know something about this. I also have a fibre optics PhD and have made more fusion splices than I care to count.

OTDR only works if there are no isolators in the path. (Gizmos which let light pass in one direction only)

In some submarine cable designs at each repeater there is a return path (ie a fibre loops back) going back the way the light came. I seem to remember this being at an out of channel wavelength (so it passes through some wavelength dependent isolators). Anyway, once you know how many repeaters you do get light back from along this return path you know more about where the break is.

I was surprised by the comment about the cable thickness for working at 2.5 miles depth. The repeater chassis I've seen are steel, coffin sized, and the walls are 21mm thick.

I also have a feeling that todays technology is the same as that of 4 or 5 years ago. There hasn't been that much investment (or new jobs) in new submarine cable tech since the dot-com crash. Maybe it's picking up again now but it will take a while to get the momentum back that we had in the R&D team 6 years ago.

Re:huh

[amorsen]

It's reasonably easy to find a cable. At those depths the cables lie directly on the sea floor, and if you run a hook across the sea floor from one side of the cable to the other, you'll cross it by definition. Sure you might then end up some distance from the break, but they are apparently reasonably good at handling that. Such techniques don't work with shipwrecks -- and even if they did, there may not be anything the right shape for a hook to catch.

Anyway, there's an old but interesting article on undersea cables by Neal Stephenson.

And what exactly would be a modern way?

[Josef Meixner]

When a method works well, there is not really a need for something else, it will only be used when it is superior. So what would those alternatives be, that are cheaper overall?

A submarine robot repairing on site is probably not possible (I have problems believing it would be able to fuse the fibers), so you only could use it to more quickly find the cable and perhaps make it easier to get the hook onto it as you can see what you are doing. But honestly, how much faster would it be, I guess a hook and the cable for it can just be tossed into the water, an expensive robot probably would take a bit longer to reach the ground. And it probably has limitations on the depth it can operate. Additionally I am quite sure they don't just drive out there and plow the sea bed, they probably have a very good idea where the cable is supposed to be. And don't forget, to find a very long cable is much simpler than finding a wreck. I don't have to find a particular piece, any piece before and after the break is ok, as I can just pull the part up and then follow it, no need to exactly grab the end.

Also the strong winds and high waves probably would make fusing the cables very hard as well, even if they could bring up the cable. So the only thing 'old fashioned' I can see is, that they use a hook. The rest is probably quite up to date and the hook is simply the easiest, most reliable and cheapest way. Why use expensive technology if something simple is perfectly adequate?